U.S. flag

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

  • Publications
  • Account settings
  • Advanced Search
  • Journal List
  • Sensors (Basel)

Logo of sensors

Intelligent Techniques for Detecting Network Attacks: Review and Research Directions

Malak aljabri.

1 Computer Science Department, College of Computer and Information Systems, Umm Al-Qura University, Makkah 21955, Saudi Arabia; as.ude.uqu@iritomhs

2 SAUDI ARAMCO Cybersecurity Chair, Department of Computer Science, College of Computer Science and Information Technology, Imam Abdulrahman Bin Faisal University, P.O. Box 1982, Dammam 31441, Saudi Arabia; as.ude.uai@4807000812 (S.M.); as.ude.uai@5017000812 (F.M.A.); as.ude.uai@0917000812 (M.A.); as.ude.uai@3222000812 (H.S.A.)

Sumayh S. Aljameel

3 Department of Computer Science, College of Computer Science and Information Technology, Imam Abdulrahman Bin Faisal University, P.O. Box 1982, Dammam 31441, Saudi Arabia; as.ude.uai@leemajlas

Rami Mustafa A. Mohammad

4 Department of Computer Information Systems, College of Computer Science and Information Technology, Imam Abdulrahman Bin Faisal University, P.O. Box 1982, Dammam 31441, Saudi Arabia; as.ude.uai@dammahommr

Sultan H. Almotiri

Samiha mirza, fatima m. anis, menna aboulnour, dorieh m. alomari.

5 SAUDI ARAMCO Cybersecurity Chair, Department of Computer Engineering, College of Computer Science and Information Technology, Imam Abdulrahman Bin Faisal University, P.O. Box 1982, Dammam 31441, Saudi Arabia; as.ude.uai@9807000812 (D.M.A.); as.ude.uai@5217000812 (D.H.A.)

Dina H. Alhamed

Hanan s. altamimi.

The significant growth in the use of the Internet and the rapid development of network technologies are associated with an increased risk of network attacks. Network attacks refer to all types of unauthorized access to a network including any attempts to damage and disrupt the network, often leading to serious consequences. Network attack detection is an active area of research in the community of cybersecurity. In the literature, there are various descriptions of network attack detection systems involving various intelligent-based techniques including machine learning (ML) and deep learning (DL) models. However, although such techniques have proved useful within specific domains, no technique has proved useful in mitigating all kinds of network attacks. This is because some intelligent-based approaches lack essential capabilities that render them reliable systems that are able to confront different types of network attacks. This was the main motivation behind this research, which evaluates contemporary intelligent-based research directions to address the gap that still exists in the field. The main components of any intelligent-based system are the training datasets, the algorithms, and the evaluation metrics; these were the main benchmark criteria used to assess the intelligent-based systems included in this research article. This research provides a rich source of references for scholars seeking to determine their scope of research in this field. Furthermore, although the paper does present a set of suggestions about future inductive directions, it leaves the reader free to derive additional insights about how to develop intelligent-based systems to counter current and future network attacks.

1. Introduction and Background

Rapid advancements in technology have made the Internet easily accessible and it is now actively used by the majority of people for a plethora of professional and personal tasks. Various sensitive activities including communication, information exchange, and business transactions are carried out using the Internet. The Internet helps foster connection and communication, but the integrity and confidentiality of these connections and information exchanges can be violated and compromised by attackers who seek to damage and disrupt network connections and network security. The number of attacks targeting networks are increasing over time, leading to the need to analyze and understand them and develop more robust security protection tools. Every organization, industry, and government requires network security solutions to protect them from the ever growing threat of cyber-attacks. The need for more effective and stable network security systems to protect business and client data is rising as there is no network immune to network attacks.

Several techniques have been proposed over the years to handle and classify network traffic attacks. One is the port-based technique, which includes identifying port numbers among the ones registered by the Internet Assign Number Authority (IANA) [ 1 ]. However, due to the growing number of applications, the number of unpredictable ports has increased and this technique has proven to be ineffective. Furthermore, this technique does not cover account applications that do not register their ports with the IANA or applications that use dynamic port numbers. Another technique that has been proposed is the payload-based technique, also known as deep packet inspection (DPI), where the network packet contents are observed and matched with an existing set of signatures stored in the database [ 1 ]. This method provides more accurate results than the port-based technique, but does not work on network applications using encrypted data. Furthermore, this technique has been proven to be complex, involving high computational costs and a high processing load [ 1 ]. Behavioral classification techniques analyze the entire network traffic received at the host in order to identify the type of application [ 2 ]. The network traffic patterns can be analyzed graphically as well as by examining heuristic information, for example, transport layer protocols and the number of distinct ports contacted. Although behavioral techniques yield good results as they are able to detect unknown threats, they are resource-intensive and are prone to false positives. Another technique, called the rationale-based technique or the statistical technique [ 2 ] examines the statistical characteristics of traffic flow, namely, the number of packets and the maximum, mean, and minimum of the packet size. These statistical characteristics are used to identify different applications since these measurements are unique for every application. However, there is a growing need to incorporate this approach with techniques that could improve the accuracy and speed up the process of classifying the statistical patterns. The correlation-based classification [ 2 ] accumulates packets into flows; that is, it collects data packets with the same source and destination IP, port, and protocol. These are classified according to the correlation between network flows. Multiple flows are usually accumulated further into a Bag of Flow (BoF). Although this technique has proven to perform better than statistical techniques as it overcomes the problem of feature redundancy, it has a high computational overhead for feature matching. Therefore, the need to create techniques that could overcome the rising challenges persist.

At the onset of the 21st century, the concepts of intelligent techniques, namely machine learning (ML) and deep learning (DL) became widespread. Researchers widely acknowledged that these techniques could greatly increase the calculation potential since they focus on using statistical methods and data to make computers think the way humans think. Hence, these intelligent techniques started being used by computer scientists in network security as they addressed the limitations of the non-intelligent techniques. In the field of network security, ML or DL algorithms can be trained with network data to recognize traffic type as normal or malicious and thus protect the network from intruders. Furthermore, the algorithms can be trained to identify the attack type if the network traffic is malicious and trigger appropriate action to prevent the attack. By analyzing past cyber-attacks, the model can be taught to prepare individual defensive reactions. These applications of intelligent methods in network security, which is the focal point of this research paper, can be useful in big businesses, organizations, law enforcement agencies, and banks that store sensitive information as well as in personal networks.

In the past, most of the developed network attack detection techniques actively depended on a set of pre-defined signature-based attacks. This was a major setback since the database of the attacks needed to be constantly updated as the attackers found new ways to exploit network security. However, with the evolution of intelligent-based techniques such as ML and DL, the predictive accuracy of identifying and classifying network attacks has been greatly improved. Therefore, using intelligent-based techniques in network security is a thriving field for research that needs to be explored.

Although several review articles exploring how intelligent-based systems have been applied to detect network attacks have been published in the last few years, none have been found that are as comprehensive as this article. This article covers almost one hundred research articles produced from 2010 to 2021 on a range of network attacks. It will provide clear insights into the race between developing intelligent systems to counter network attacks and how these attacks have evolved to circumvent intelligent systems, thus highlighting gaps in the research and indicating potential future research areas. This research also applied a different taxonomy that, to the best of our knowledge, has not been used in any previous research. It sets up several criteria against which the articles being reviewed could be assessed and compared including:

  • (i) What is/are the classification algorithms implemented?
  • (ii) What is/are the datasets employed for developing the intelligent systems?
  • (iii) Furthermore, this research article compared the results obtained using different evaluation metrics.

It then discusses the answers to the following main questions:

  • (i) Which algorithm(s) was/were commonly implemented and in which kind of attacks?
  • (ii) Which dataset(s) is/are considered more reliable based on the results obtained?

The resulting comparisons and discussions will help future researchers to identify the directions to take in their research, that is, to either improve the intelligent-based algorithms or consider other algorithms, to identify the features that should be added or removed when building the training dataset, and to indicate the evaluation metrics that should be adopted to evaluate the created intelligent systems.

The outcomes of this paper provide valuable directions for further research and applications in the field of applying effective and efficient intelligent techniques in network analytics.

This article is organized into four sections. The first section provides an introduction and background to the research area. A brief overview of network attacks is presented in Section 2 . Section 3 discusses intelligent network attack mitigation techniques where all the reviewed research papers, the network attacks they address using ML and DL techniques, and their findings are presented. Finally, the last section provides a discussion of the findings and the ideas presented in the papers reviewed and sets out promising research directions.

2. Network Attacks

For decades, networking technologies have been used to improve data transfer and circulation. Their continuous improvements have facilitated a wide range of new services.

The Internet of Things (IoT) is a powerful tool for improving communication by connecting different devices to the Internet and collecting data. The information gathered assists firms in the analysis and forecasting of consumer behavior to enhance the quality of their products. Nowadays, ML and DL are being used to construct network systems that can conduct advanced analytics and automation. This technology is transforming the users’ networking experiences by simulating human intellect and gathered data with built-in algorithms [ 3 ].

The emerging cloud computing technologies have brought about remarkable evolutions in network technology where different applications, services, and computing and storage resources are offered on demand to a large number of users via the Internet, thus offering tremendous advantages including flexibility, minimal administrative efforts, cost effective resource utilization, high accessibility, efficiency, and reliability [ 4 ].

A new global wireless standard is the 5th generation (5G) mobile network, which represents a logical network type that connects essentially anything including machines, objects, and gadgets. Not only does 5G offer faster speeds and a greater number of linked devices, it also enables network slicing. Network slicing is the process of dividing several virtual networks operating on the same network infrastructure to create subnetworks that meet the demands of various applications. From entertainment and gaming to school and community safety, the 5G network technology has the ability to develop anything. 5G has the potential to provide higher download rates, real-time replies, and improved connection over time, allowing companies and consumers to explore new innovations [ 5 ].

Such an exponential growth in network technologies has offered many advantages and has greatly improved communications. However, each emerging network technology presents new security challenges and triggers the need for the development of detection tools and countermeasures to meet the new demands. The following subsections briefly discuss the main types of network attacks.

2.1. Types of Network Attacks

A network attack is an approach to hurt, reveal, change, destroy, steal, or obtain illegal access to a network system resource. The attack could come from inside (internal attack) or from outside (external attack). Table 1 lists and describes a number of different types of network attacks that disrupt communication, classifying them as either active or passive attacks, bitcoin attacks, account attacks, or a security breach [ 6 ].

Types of network attacks.

2.2. Network Attack Detection and Prevention Techniques

Security and defense systems are designed to identify, defend, and recover from network assaults. Confidentiality, availability, and integrity are the three primary aims of network security systems. Network intrusion detection and prevention techniques can be classified based on the approach used to detect network threats, prevent them, or a combination of both. These techniques are developed as software, hardware, or a combination of both. They can be classified into two classes: intrusion detection systems (IDS), and intrusion prevention systems (IPS) [ 6 , 7 ].

  • Intrusion Detection System (IDS): Referred to also as network-based IDS (NIDS). This system intensely monitors malicious network activities and notifies officials if an attack is detected with no prevention abilities. Signature-based and anomaly-based detection are the two most prevalent approaches used by IDS to identify threats. Signature-based procedures are applied to detect only known threats, relying on a database containing a list of pre-existing characteristics of known attacks (attacks signatures) to identify suspicious events. The database needs to be continuously updated to include emerging attacks. On the other hand, anomaly-based procedures attempt to differentiate malicious traffic from real traffic based on a change in the network traffic; thus, they can detect unknown threats. Inconsistencies such as high-size traffic, network latency, traffic from uncommon ports, and abnormal system performance, all represent changes in the normal behaviors of the system and can indicate the presence of network attacks.
  • Intrusion Prevention System (IPS): Known also as intrusion detection and prevention systems (IDPS). It scans the network continuously for the presence of illegal or rogue control points that are detected on the basis of changes in behavior. The system automatically takes countermeasures to tackle the threats and defend the system. The primary objective of an IDPS is to keep malicious or undesired packets and attacks from causing any harm. IDPS is more effective than IDS as it not only detects threats, but is able to take action against them. There are two types of IDPS: network-based intrusion detection and prevention systems (NIDPS) that analyze the network protocol to identify any suspicious activities and host-based intrusion detection and prevention systems (HIDPS) that are used to monitor host activities for any suspicious events within the host.

To identify attacks effectively and efficiently, a variety of detection approaches are constantly being developed based on intelligent techniques including ML and DL, which have recently gained immense popularity in the network security field.

3. Intelligent Network Attack Mitigation Techniques

In this section, research studies that used intelligent models to detect different cyber-attack types are reviewed and their findings summarized. Several ML algorithms have been used in these studies including classification, regression, and clustering techniques such as logistic regression (LR), decision trees (DT), etc. Some used random forest (RF), an ensemble of DT, in order to visually represent the sequences of the decision-making process in the form of a tree. Support vector machine (SVM) was widely used in classification due to its ability to distinctly classify the data points by building a hyperplane in an n-dimensional space, where n represents the number of features. Another ML classifier that has been widely used is naïve Bayes (NB), a supervised learning model that uses Bayes’ theorem of probability. Finally, some researchers have used the K-nearest neighbor (KNN) for classification and K-means clustering, an unsupervised approach. Further details about these algorithms can be found in [ 8 ].

DL is a subset of ML, which is a subset of artificial intelligence (AI). A number of DL techniques have been used to build the detection models in some studies, primarily the artificial neural network (ANN), which is an information-processing system that consists of several layers that work best with non-linear dependence and recurrent neural network (RNN), a type of ANN that contains memory function to maintain previous content. Another commonly used DL technique is the convolutional neural network (CNN), which is also a type of ANN that mimics human vision. Furthermore, deep neural network (DNN), a supervised learning type of ANN that finds correct mathematical manipulation to turn input into output, has been used by some authors. Long-short term memory (LSTM), a type of RNN designed to model temporal sequences more accurately, and multi-layer perceptron (MLP), a type of ANN that consists of many layers in directed graphs, have also been widely used. Finally, the gated recurrent unit (GRU), which, though a variant of LSTM and is considered to be more efficient than LSTM as it uses comparatively less memory and executes faster, has also been used. More information about the mentioned algorithms can be found in [ 9 ].

3.1. Problem Domains of the Reviewed Articles

The papers were classified according to the cyber-attack type on which they focused. The different attack types mentioned in this section are insider threat, DDoS attacks, zero-day attacks, phishing attacks, malware attacks, and botnet attacks. We then reviewed articles that did not target specific attacks, but aimed to identify attacks at IoT networks, classify the malicious traffic to different attacks, and identify attacks at the DNS level. Finally, we also mention papers targeting the detection of intrusions in the network.

3.1.1. Insider Threat

Cybersecurity measures have tended to focus on threats outside an organization rather than threats inside that can cause harmful effects. Therefore, researchers have started to look at different techniques to identify insider threats. Tuor et al. [ 10 ] built a model using principal component analysis (PCA) for feature selection, and unsupervised DL namely, DNN, RNN, SVM, isolation forest, DNN-Ident, DNN-Diagnosis LSTM-Ident, LSTM-Diagnosis, among others, that use system logs to detect anomalous activities in the network. The dataset used was synthetic CERT insider threat v6.2 [ 11 ], which was taken from the event log lines of a network of a simulated organization’s computer. The researchers targeted two prediction approaches: the “next time step” and the “same time step”. The results of the experiments showed that the “same time step” approach resulted in higher performance, and that the isolation forest model was the strongest model. To evaluate the proposed model, recall was used and DNN-Diagnosis, LSTM-Diagnosis, and the isolation forest model all obtained 100% recall. In future work, the researchers may apply the proposed model to a wider range of streaming tasks and explore different granularities of time.

Similarly, LSTM and CNN techniques were used by Yuan et al. [ 12 ] to build a model to detect insider threats. They applied the model on the CERT insider threat v4.2 dataset [ 13 ], which contained 32 M log lines among which 7323 were anomalous activities. The advantage of this version of the CERT dataset was that it contained more samples of insider threats than other versions. The train–test split was 70–30%. The researchers first used LSTM to extract the user behavior, abstracted temporal features, and produced the feature vectors. After that, the researchers transformed the feature vectors into fixed-size matrices. Finally, CNN was used to classify the feature matrices into anomaly or normal. The proposed model resulted in an area under the curve (AUC) of 94.49%.

Hu et al. [ 14 ] used DL methods to build a user authentication model based on characteristics of mouse behaviors that could be used to monitor and detect insider authentications. They used an open-source dataset called the Balabit Mouse Dynamics Challenge dataset [ 15 ], and CNN algorithm. CNN showed high performance in user authentication based on mouse features with a false acceptance rate (FAR) of 2.94% and a false rejection rate (FRR) of 2.28%.

3.1.2. DDoS Attacks

One of the most harmful threats in network security is distributed denial of service (DDoS) attacks that attempt to disrupt the availability of services. Since DDoS is easy to launch but not easy to detect, as in most cases the attack traffic is very similar to legitimate traffic, some researchers have focused solely on detecting them using different ML approaches.

Yuan et al. [ 16 ] proposed DeepDefense, which is a DL-based DDoS attack detection approach that can study network traffic sequence patterns and trace the network attack activities. They used the UNB ISCX intrusion detection evaluation 2012 (ISCX2012) dataset [ 17 ], and the RNN algorithm to build the model. From ISCX2012, the team extracted 20 network traffic fields to generate a 3-D feature map using a sliding time window. Data14 and data15 were extracted from ISCX2012, which contained 9.6 M packets and 34.9 M packets, respectively. The total number of training samples in data14 and data15 were 15,176 and 233,450, respectively. The experiment results showed that the DL models reduced the error rate by 39.69% compared to ML methods in a small dataset. For large datasets, the reduction in the error rate ranged from 7.517% to 2.103%. For future work, they suggested increasing the diversity of DDoS vectors and system settings to test the DeepDefense model as well as compare DeepDefense against other ML algorithms.

A study proposing a model for analyzing and detecting DDoS attacks on the network-level and service levels of the bitcoin ecosystem was carried out by Baek et al. [ 18 ]. The dataset consisted of real DDoS attacks [ 19 ] and contained the service affected, date of the attack, category of service, number of posts, etc. From the bitcoin block data, the researchers extracted statistical data such as maximum, minimum, summation, and standard variation. The researchers used PCA to perform feature extraction. MLP was used to detect DDoS while the training set, validation set, and testing set were divided according to the ratio 6:2:2. The results showed that the accuracy of DDoS attack detection was about 50% and the accuracy for classifying normal block data was about 70% while setting the unit of epoch to 100. In future work, the researchers wish to find out how to extract the features that impact the characteristics of the blocks made when a DDoS attack occurs.

Sabeel et al. [ 20 ] used DNN and LSTM for binary prediction of unknown DoS and DDoS attacks. To train the models, they used the CICIDS2017 dataset (size 283 MB) [ 17 ]. For testing, a new dataset called ANTS2019 (size 330 MB), which mimics real-life attacks, was generated in a simulated environment to measure performance. In feature engineering, 78 features were used for the training set and 77 for testing (the ‘Fwd Header length’ feature was dropped). The train–test split was 75–25%. When the model was trained using CICIDS2017 and part of ANTS2019, the highest evaluation accuracy of 99.68% for DNN was obtained. When the researchers demonstrated the retraining of the models on a dataset with new unknown attacks, the true positive rate (TRP) obtained was 99.8% and 99.9% for DNN and LSTM, respectively. To maintain performance, it was concluded that the models should be updated with new attacks at regular intervals.

An intrusion detection system (IDS) used against DDoS attacks called DDoSNet was built by Elsayed et al. [ 21 ], which was a combination of autoencoder (AE) with RNN. In their paper, the researchers evaluated their classifier using the newly released CICDDoS2019 dataset [ 22 ], which contained 80 flow features. For feature engineering, PCA was applied, and the input features were 77. The total number of samples for training, validation, and testing sets were 161,523, 46,150, and 23,000, respectively. When the model was evaluated, the results indicated an accuracy of 99%, outperforming all compared ML methods—SVM, DT, NB, RF, Booster, and LR. In future work, the researchers intend to test the performance of their model in different datasets and extend the work to multiclass classification, since, in this research, a binary classification framework was applied.

A model that exploited the characteristics of CNN to classify the traffic flows as either benign or malicious was proposed by Doriguzzi-Corin et al. [ 23 ]. The CICIDS2018, CICIDS2017, and ISCX2012 datasets, which can be obtained through the Canadian Institute for Cybersecurity of the University of New Brunswick (UNB), were used by the researchers. They extracted 37,378 DDoS flows, and 37,378 randomly selected benign flows from ISCX2012. Then, they repeated the process for CICIDS2017 with 97,718 for benign and 97,718 for DDoS flows, and again for CICIDS2018 [ 17 ] with 360,832 for benign and 360,832 for DDoS flows. Following the pre-processing phase, each dataset was split as 90–10% train–test sets. The results showed that the accuracy for each dataset was 99.87%, 99.67%, and 98.88%, respectively. The UNB201X dataset was then constructed by combining splits from every year and the accuracy for the model with the UNB201X dataset was 99.46%. In future work, the researchers would like to optimize the pre-processing tool, rather than the detection model and also extend the dataset’s labels.

Ahuja et al. [ 24 ] used various DL algorithms to detect the DDOS attacks: CNN, RNN, LSTM, CNN-LSTM, support vector classifier-self organizing map (SVC-SOM), and stacked autoencoder-multi layer perceptron (SAE-MLP). The team used the dataset provided by leading India Project Mentor [ 25 ], which consists of 22 features. Two different optimizers were used: stochastic gradient descent (SGD) for the first 10 epochs and Adam for the next 150 epochs. For an unencrypted network, using a CNN, traffic features can be extracted automatically. Finally, they evaluated the model using the following metrics: accuracy, precision, recall, F-score, false positive rate (FPR), and false negative rate (FNR). The highest classification accuracy of 99.75% was achieved with the SAE-MLP algorithm.

A study conducted by Shi et al. [ 26 ] focused on using DL for both packet-wise and period-wise methods for traffic DDoS attack detection. They proposed a model that leveraged a DL approach for DDoS detection, which was DeepDDoS. It used spark as a big data processing framework. Additionally, for feature selection, maximal information coefficient and mutual information were used. The LSTM model was used for the training phase due to its better performance in longer sequences. The proposed work tried to filter out the abnormal flow with the least computational costs. The dataset used was CICIDS2017 (Size 283 MB). The results showed that the model achieved over 99% accuracy when receiving five packets in a continuous flow.

A model that used DL for the detection of multi-vector DDoS on a software-defined network was construed by Quamar Niyaz et al. [ 27 ]. An SAE-based DL approach was applied and the team collected network traffic from a real network (packets for normal traffic were captured from network connected to the Internet) and a private network (packets with DDoS attacks were captured from a private lab network) for the evaluation of the model. They divided the dataset files into training and testing, and then normalized them using max–min normalization. For comparison, models with soft-max and neural networks (NN) were also developed. The result showed that SAE performed better than the soft-max and NN model. The model achieved 95.65% accuracy. The researchers intend to develop a NIDS in future to detect the DDoS along with other attacks as well as the use of DL for feature extraction from raw bytes.

Pande et al. [ 28 ] aimed to build a ML model to detect DDoS attacks. To build the proposed model, a DDoS attack was performed using the ping of death technique and detected using RF. The dataset used by the researchers was the NSL-KDD [ 29 ] dataset containing a training set of 125,973 records and testing set of 22,544 instances and 41 attributes. The building time of the model was 8.71 s and the testing time was 1.28 s. The proposed model built using the random forest (RF) algorithm resulted in 99.76% accuracy. For future work, the researchers will implement the DL technique to classify the instances.

Radivilova et al.’s [ 30 ] goal was to analyze the main methods of identifying DDoS attacks through network traffic using the SNMP-MIB dataset [ 31 ]. They used RF as the classification method. The experiments began with the training and evaluation of a time series classifier. Recurrence analysis was used to extract features and the Hurst exponent was set at 10 intervals during the experiment. The main evaluation metrics were accuracy, FNR, and TPR. A numerical experiment showed that early detection is plausible when the average attack ratio represents 15–20% of the average traffic.

Likewise, Filho et al. [ 32 ] presented a smart detection system for DoS using ML. The goal was to detect both high- and low-volume DDoS attacks. The researchers used RF, perceptron, AdaBoost, DT, SGD, and LR. Since RF achieved higher precision while using 28 variables, it was used for classifying the network traffic. The evaluation of the proposed system was based on four intrusion detection benchmark datasets, namely, CICIDS2017, CICDoS2017 [ 33 ], CICIDS2018, and customized datasets. To evaluate the proposed model, recall, precision, and F-measure (F1) were used. In the CICIDS2018 and CIC-DoS2017 datasets, the proposed system achieved precision and a detection rate (DR) of more than 93% with a false alarm rate (FAR) of less than 1%. The researchers intend to include an analysis of DDoS attacks of Heartbleed and brute force attacks in their future work and to evolve methods for correlating triggered alarms.

Correspondingly, Vijayanand et al. [ 34 ] proposed a detection system of novel DoS attacks using multi-layer deep algorithms arranged in hierarchical order to detect the attacks accurately by analyzing the smart meter network traffic. The suggested technique addresses issues arising as a result of a large amount of input data and the complexity of input features. To evaluate the designed model, 9919 records from the CICIDS2017 dataset were used. The performance of the proposed system was analyzed by comparing it with simple multi-layer DL algorithms and hierarchical SVM algorithms, obtaining efficiency values of 39.78% and 99.99%, respectively.

An improved rule induction (IRI) based model was put forth by Mohammed et al. [ 35 ] for detecting DDoS attacks. UNSW-NB15 [ 36 ] dataset was used and, following the application of under-sampling without replacement and further pre-processing as well as correlation-based feature selection, the final dataset ended up with eight attributes. The suggested algorithm, called IRI for detecting DDoS attacks (IRIDOS), eliminates all insignificant items during the model creation and reduces the searching space to create the classification rules. Furthermore, the algorithm stops learning a rule after reaching a ‘rule-power’ threshold. The proposed technique was evaluated on 13 datasets from the UCI repository. IRI obtained a F1 score of 93.90% on UNSW-NB15. The model attained promising results, especially when compared to other data mining algorithms such as PRISM (divide-conquer knowledge-based approach), PART (a rule-based classification algorithm), and OneRule (OR).

An evaluation and comparison of the performance of different supervised ML algorithms on the CAIDA DoS attack dataset [ 37 ] were carried out by Robinson and Thomas [ 38 ]. Other datasets used were CAIDA Conficker, and KDD-99 [ 39 ]. The different ML algorithms included NB, RF, MLP, BayesNet, J.48, IBK, and Voting. It was observed that since the CAIDA Conficker dataset contained DDoS attacks generated from large botnets with flooding-attack vectors that were easily distinguishable with more bias, all ML algorithms, except NB, achieved an accuracy rate of more than 99% in this dataset.

Research that used the same CAIDA dataset was conducted by Barati et al. [ 40 ] who developed a hybrid ML technique to detect DDoS attacks. The CAIDA USCD 2007 was used for the attack as it contained an hour of anonymized traces from a DDoS attack on 4 August 2007. For normal traffic, the CAIDA Anonymized 2013 was used as it contained passive traces from CAIDA passive monitors in 2013. For feature selection and attack detection, genetic algorithm (GA) and ANN were used, respectively, and to select the most efficient feature wrapper method, GA was applied. The attack detection method was improved by deploying the MLP method of ANN. While building the model, the 10-fold cross-validation technique was used. The results showed that the proposed method obtained an excellent AUC of 99.91%. The researchers’ future work will include performing more experiments to detect the robustness of the model on different datasets.

Kim et al. [ 41 ] developed a model based on a convolutional neural network (CNN) for DoS attacks. They used two different datasets: the KDD-99 dataset and the CICIDS2018 dataset. They generated two types of intrusion images, RGB and grayscale. They considered the number of convolutional layers and the size of the kernel when they designed their CNN model. They performed both binary classification and multiclass classification. Moreover, the performance of the proposed model was evaluated by comparing it to the recurrent neural network (RNN) model. The best results were achieved with the KDD dataset by the CNN model that showed 99% or more accuracy in the binary and multiclass classifications. The RNN showed 99% accuracy in the binary classification. The CNN model proposed by the researchers was better able to identify specific DoS attacks with similar characteristics than the RNN model.

Finally, an approach to detect DDoS attacks using GRU was carried out by Rehman et al. [ 42 ]. The team produced a high-efficiency approach called DIDDOS to detect real world DDoS attacks using GRU, a form of RNN. Different classification models, namely GRU, RNN, NB, and SMO, were applied on the CICDDoS2019 dataset. For DDoS classification in the case of reflection attacks, the highest accuracy level of 99.69% was achieved while for the DDoS classification in the case of exploitation attacks, the highest accuracy level of 99.94% was achieved using GRU.

3.1.3. Phishing Attacks

Some studies have focused on training models and testing them to detect phishing attacks. For instance, the main goal of Alam et al. [ 43 ] was to defend against phishing attacks by developing an attack detection model using RF and DT, which are ML algorithms. For ML processing, a traditional phishing attack dataset from Kaggle that contained 32 features was used. To analyze the dataset characteristics, the intended model used PCA, a type of feature selection algorithm. An accuracy level of 97% through RF was reached. With less change and variance in RF, the over-fitting obstacle was controlled. Future studies will include the prediction of phishing attacks from the registered attacks in a dataset by applying CNN and implementing the IDS.

To identify phishing website attacks, a self-structuring neural network based on ANN was developed by Mohammad et al. [ 44 ]. Phishing-related features are crucial in detecting the kind of web pages that are extremely dynamic, thus the structure of the network should be constantly improved. The proposed approach addresses this issue by automating the network structuring process and demonstrating high acceptance for noisy input, fault tolerance, and significant prediction accuracy. This was accomplished by increasing the learning rate and expanding the hidden layer with additional neurons. The goal of the developed model was to obtain generalization ability, which means that the training and testing classification accuracy should be as similar as possible. The dataset included 600 legal and 800 phishing websites, with 17 characteristics retrieved using their own tool [ 45 , 46 ]. The accuracy of the training, validation, and testing sets were 94.07%, 91.31%, 92.18% for 1000 epochs, respectively. The principle of the model was to use an adaptive scheme with four processes including structural simplicity, learning rate adaptation, structural design adaptation, and an early stopping approach based on validation faults.

Trial and error is one of the most popular techniques used to train a neural network, but it has a significant drawback in that it takes a very long time to set the parameters and might even require the assistance of a domain expert. Rather than trial and error, a better self-structuring neural network anti-phishing model, which makes it simpler to structure NN classifiers, was proposed by Thabtah et al. [ 47 ]. The goal of the technique was to build a large enough structure from the training dataset to develop models that can be generalized to the testing dataset. During the training phase, the algorithm dynamically modifies the structural parameters in order to generate accurate non-overfitting classifiers. With a dataset of over 11,000 websites from UCI, the neural network characteristics were updated as the classification model was being built, but they were largely dependent on the computed error rate, intended error rate, and underlying technologies. When compared to Bayesian networks and DT, the findings indicated that the dynamic neural network anti-phishing model had a higher prediction accuracy. The highest average accuracy achieved was 93.06% when information gain was used for pre-processing.

A two-layered detection framework to identify phishing web attacks by using features derived from domain and DNS packet-level data was built by Rendall et al. [ 48 ] using four ML models, namely MLP, SVM, NB, and DT. The team investigated the use of the approach where a phishing domain was classified multiple times, with additional classification being carried out only when it scored below a predefined confidence level set by the owner of the system. The model was evaluated on a dataset created by the team, and it contained 5995 phishing records and 7053 benign records. After applying the models in the two-layered architecture, the highest accuracy of 86% was achieved by MLP and DT.

Li et al. [ 49 ] built a stacking model using URL and HTML features to detect phishing web pages. They used lightweight HTML and URL features as well as HTML string embeddings to make it possible to detect phishing in real-time. The 50K-PD dataset that contained around 49,947 samples as well as the 50K-IPD dataset that contained 53,103 web page samples were made and used. The stacking model was made by combining GBDT, XGBoost, and LightGBM in multiple layers. The model achieved an accuracy of 97.30% on the 50K-PD dataset and an accuracy of 98.60% on the 50K-IPD dataset.

Phishpedia, an ensemble deep learning model described in [ 50 ], addresses major technological difficulties in phishing detection by identifying and matching brand logo variations. Three different datasets were used for this experiment. First, researchers collected the first dataset by subscribing to a service; then they collected the second one from a top-ranked Alexa list, and finally, to evaluate the detection model, they collected the third dataset from a benign dataset. As a Siamese neural network converts image to vector, which assists in estimating the correlation between two visuals, this model was chosen by the researchers for their project. A better accuracy level and less runtime cost were achieved with Phishpedia. Unlike many other approaches, phishing data are not required for training. With an accuracy of 99.2%, Phishpedia outperformed the state-of-art approaches such as LogoSENSE, EMD, and PhishZoo by a large margin. In the future, the researchers plan to expand Phishpedia by adding a system to monitor phishing online.

Supervised machine learning models were used to detect phishing attacks based on novel combination features that were extracted from the URL by Batnaru et al. [ 51 ]. The researchers used a dataset from Kaggle [ 52 ] and PhishTank [ 53 ] containing 100,000 URLs that consisted of 40,000 benign URLs from Kaggle and 60,315 phishing URLs from PhishTank for the training. They used five ML models, namely MLP, RF, SVM, NB, and DT. In terms of model selection, RF was found to be the best candidate based on F1 scores. The evaluation process was performed using an unbalanced dataset that consisted of 305,737 benign URLs and 74,436 phishing URLs to evaluate the selected model in a realistic scenario. The achieved accuracy was 99.29%. The results were compared with the performance of Google Safe Browsing (GSB), which is the default protection that is available through popular web browsers. The model outperformed the GSB. In future work, the researchers’ aim is to explore the effectiveness of their model on other datasets as well as experiment with more features. They also plan to assess the robustness of the methodology against adversarial attacks that are mostly used by malicious parties.

PhishDump, a new mobile app based on a mix of LSTM and SVM algorithms, was suggested by Rao et al. [ 54 ] to detect genuine and fake websites in mobile platforms. Because PhishDump concentrates on extracting characteristics of URLs, it offers important benefits in comparison with previous efforts including quick calculation, class independence, and resistance to unintentional malware installation. The data were gathered from three separate inputs: Alexa, OpenPhish, and PhishTank. The application’s positive aspect is that it is free of external code and databases, allowing for the identification of malicious websites in as little as 621 ms. The characteristics extracted from the LSTM model are supplied as input for URL classification to SVM using a python code. Using several datasets, this application was compared against current baseline classifiers. PhishDump surpassed all previous studies with an accuracy of 97.30%. This approach has limitations such as the chance that an intruder might circumvent the approach by implementing structural modifications to the URL, and the system could miss phishing websites with shortened URLs.

Marchal et al. [ 55 ] reviewed phishing attack problems. The researchers provided guidelines for designing and evaluating phishing webpage detection techniques. They also presented the strengths and weaknesses of various design and implementation alternatives with regard to deployability and ease of use. Moreover, they provided a list of guidelines to evaluate the proposed solutions following the selection of representative ground truth, appropriate use of the dataset, and the relevant metrics. These recommendations can also enable comparison of the accuracy of different phishing detection technologies. The researchers state that academic research in phishing detection should adopt design and evaluation methods that are relevant to real-world publication.

Similarly, Das et al. [ 56 ] also reexamined the existing research on phishing and spear phishing from the perspective of different security domains such as real-time detection, dataset quality, active attacker, and base rate fallacy. They elucidated on the challenges faced and surveyed the existing solutions to phishing and spear phishing. Their work helps guide the development of more robust solutions by examining all the existing research on phishing.

3.1.4. Zero-Day Attacks

Interestingly, some researchers have focused on identifying zero-day attacks. One such study was conducted by Beaver et al. [ 57 ] who used ML methods that are able to distinguish between normal and malicious traffic. In their study, they used the adaptive boosting (AdaBoost) ensemble learner with DT in order to distinguish and classify the type of traffic on the KDD-99 dataset. The implementation that was tested in this study had four levels: (1) the top-level model that puts a cap on the FPR; (2) the first internal model that includes the AdaBoost ensemble, (3) the second internal model that implements the DT, and the lowest model that provides a judgment on whether the traffic was normal and relies on an anomaly detection algorithm. The system was able to detect 82% of the attacks that were previously missed by the signature-based sensor, detected 89% of attacks that it had not been trained to detect, and had a DR of 94% and a 1.8% false alarm rate. The future goals of the researchers are to scale the performance, which will require more parallelism in the architecture and modification of the training in order to accommodate larger datasets.

Ahmed et.al. [ 58 ] proposed a DL model that was used for identifying zero-day botnet attacks in real-time with a feed-forward backpropagation ANN technique and DNN. An important factor for obtaining high performance is a reliable dataset and hence the CTU-13 dataset [ 59 ] was obtained from the Botnet Capture Facility. There were nine input layer features and the dataset size was 10,000 randomly chosen flows. The first step was to normalize the whole data followed by the application of Adam’s optimizer in the model. The train–test split was 80–20%. The result showed that the accuracies achieved were over 99.6% after 300 epochs and that the model outperformed the NB, SVM, and backpropagation algorithms. In future work, the researchers suggest examining the efficiency of the proposed model with various other datasets.

3.1.5. Malware Attacks

Barut et al. [ 60 ] aimed to compare the ML algorithms, namely SVM, RF, and MLP, to determine the most accurate and the fastest method to detect malware encrypted data. Two datasets were generated: dataset1, which was produced using Stratosphere IPS [ 61 ] extracting 20 types of malware classes (Adload, Ransom, Trickbot, etc.), and dataset2, which used CICIDS2017. In feature engineering, 200 flow features were extracted and the chi-square was used. The researchers concluded that RF was the best performing algorithm as its results showed a DR of 99.996% and a FAR of 2.97%. Generally, the results showed that the SVM, RF, and MLP models are the most accurate, with some trade-offs. For dataset1, the RF model was the best performing across all evaluation metrics except for the prediction speed, which was higher when using the SVM model. For dataset2, the SVM model was the most accurate.

Marin et al. [ 62 ] developed a model for malware traffic detection of an encrypted network using DL. The specific DL model proposed in this study was the DeepMAL, which automatically discovered the best features/data representation from raw data. The dataset used was the USTCTFC2016 [ 63 ], which comprised two sections labelled malicious or normal traffic and 10 types of malware traffic. Two types of representations were used for the raw data: packets and flows. It was concluded that using raw flows representation of the input for the DL models achieved better results. The results showed that DeepMAL detected Rbot botnet with an accuracy of 99.9%, while Neris and Virut achieved 63.5% and 54.7% each. Despite the low rates achieved, they still performed better than RF.

Park et al. [ 64 ] evaluated the recognition performance of various types of attacks including IDS, malware, and shellcode using the RF algorithm and the Kyoto 2006+ [ 65 ] dataset (total size 19.8 GB). The dataset consisted of three class types: attack, shellcode, and normal. For the first two classes, there are three attack types: IDS, malware, and shellcode. This dataset contains the traffic data collected from November 2006 to December 2015. In the data preparation step, the researchers selected one month of data (May 2014) to train the model and another month (April 2014) to test the model. In the experiment, Park et al. considered 17 features and normalized the data. The overall performance was 99% for F-Score. However, it was observed that the performance of detecting different attacks differed. They propose to further evaluate the performance of the detection of various attack types using the same dataset but varying the training conditions.

In order to classify new malware variants accurately, David et al. [ 66 ] used DL to build a model using a deep belief network (DBN) algorithm that could generate and classify a malware signature automatically. The dataset used to build the proposed model was collected by the authors and contained 1800 instances and six malware categories (Zeus, Carberp, Spy-Eye, Cidox, Andromeda, and DarkCome) with 300 variants for each category. The DBN had eight layers with the output layer containing 30 neurons. The training process was unsupervised with 1200 vectors for training and 600 vectors for testing. To denoise the autoencoders, the noise ratio was 0.2 and training epochs was 1000. The model resulted in an accuracy of 98.6% when evaluated.

Reinforcement learning continuously mimics attackers to produce new malware samples, thereby giving viable attack models for defenders, as Wu et al. [ 67 ] explained. They suggested the gym-plus model, where gym-malware is improved by adding additional activities to the action space and allowing it to modify harmful portable executable files. Additionally, it retrains the algorithm using the public EMBER [ 68 ] dataset to substantially increase the DR. In gym-plus, the DQN, SARSA, and Double DQN algorithms were used, and DQN established better policies than the other algorithms. Through retraining on the adversarial instances provided by the DQN agent, malware detection accuracy increased from 15.75% to 93.5%.

Another dataset called MTA KDD 19 [ 69 ] was explored by Letteri et al. [ 70 ], who applied dataset optimization strategies to detect malware traffic. Two dataset optimization strategies, namely dimensional reduction technique based on autoencoders (AE-optimized) as well as feature selection technique based on rank relevance weight (RRw-optimized) and sensibility enhancement on the MLP algorithm were used. In RRw, feature selection consisted of two steps: dataset tampering where 5-fold cross-validation was applied, and backward feature elimination. In the AE-optimized technique, 33 input and output neurons were made and the train–validation split was 85–15%. The training set was further split to 15% testing. The highest accuracy of 99.60% was achieved in the RRw-optimized MTA KDD 19 dataset.

3.1.6. Malware Botnet Attacks

A novel scheme using supervised learning algorithms and an improved dataset to detect botnet traffic was carried out by Ramos et al. [ 71 ]. Five ML classifiers were evaluated namely, DT, RF, SVM, NB, and KNN on two datasets: CICIDS2018 and ISOT HTTP [ 72 ] Botnet (total size 420 GB). A network flow metrics analysis and feature selection was carried out on both datasets after which the ISOT dataset had 20 attributes including sources, destination port numbers, and transfer protocols among the selected features, and CICIDS2018 had 19 similar kinds of attributes. Five-fold cross-validation was applied and 80% of botnet instances were used for training and the remaining for testing. For the CICIDS2018 dataset, RF and DT achieved the highest accuracy of 99.99%. For ISOT HTTP, again, RF and DT achieved a high accuracy of 99.94% and 99.90%, respectively.

Using a similar dataset, Pektas and Akerman [ 73 ] utilized DL techniques and flow-based botnet discovery methods to identify botnet using two datasets: CTU-13 and ISOT HTTP, containing both normal and botnet data. They combined two DL algorithms namely, MLP and LSTM. In feature extraction, a flow graph was constructed where all flow data were processed to extract the features. The ISOT dataset consisted of two types of botnets, namely Waledac and Zeus, whereas CTU-13 contained seven botnet families. For the ISOT dataset, the approach achieved an F-score of 98.8%, and for CTU-13, an F-score of 99.1%.

3.1.7. Detecting Attacks over IoT Networks

As the Internet of Things (IoT) has become an important aspect of our lives, concerns about its security have increased, motivating researchers to focus their efforts on identifying new techniques to detect different attacks and increase the security of IoT. One such study was conducted by Abu Al-Haija et al. [ 74 ], where they developed an intelligent detection and classification DL-based system by leveraging the power of CNN for cyber-attacks in IoT communication networks. For evaluation, the NSL-KDD, which includes all the key IoT computing attacks, was employed. This system was validated and evaluated using K-fold and confusion matrix parameters, respectively. The outcome was an efficient and intelligent deep-learning-based system that can detect the mutations of IoT cyberattacks with an accuracy level that is greater than 99.3% and 98.2% for the binary-class and the multiclass, respectively. Discussions on future work include developing new software that catches and investigates data packets that communicate through the IoT environment and updating the existing dataset for more attacks.

By utilizing unique computing resources in a regular IoT space and applying an instance of extreme learning machine (ELM), a blockchain-based efficient solution for safe and secure IoT was proposed by Khan et al. [ 75 ]. This approach analyzes the credibility of the blockchain-based smart home in terms of the fundamental security objectives of confidentiality, accessibility, and integrity. The simulation outputs were provided to show that ELM’s overheads were minor in comparison to the cybersecurity advantages it brings. The ELM architecture is made up of input layers, numerous hidden layers, and a final output layer, with hidden layers consisting of fixed neurons to boost the network’s efficiency. To minimize the error rate, the backpropagation approach is combined with a feed-forward mechanism to modify the network weights. After pre-processing the data, to remove abnormalities and lessen the risk of faults, input data from NSL-KDD was mainly split into 85% training and 15% validation. The researchers aim to investigate more datasets and architectures in the future, because the presented ELM surpassed previous ML algorithms and achieved an accuracy of 93.91%.

Ullah et al. [ 76 ] aimed to detect malware-infected files and pirated software across the IoT network using the DL approach. The dataset used was collected by Google Code Jam (GCJ) [ 77 ]. The combined DL-based approach comprised two steps. First, to detect the pirated features, the TensorFlow neural network was proposed. The unwanted details were removed using the tokenization process and extra features were mined using stemming, root words, and frequency constraints. Second, to detect the malware, a new methodology based on CNN was proposed. The raw binary files were converted to a color image to solve the detection of malware by using an image classification problem. Grayscale visualization was gained by transforming the color images, which were then used to classify malware types. The results showed that this method performed better than modern methods when it came to measuring cybersecurity threats in IoT. In future work, the researchers intend to put forward an algorithm that can detect unknown malware families.

A model that was used for the classification of attacks in IoT networks and anomaly detection was created by Tama and Rhee [ 78 ] using a DNN. The team used CIDDS-001 [ 79 ], UNSW-NB15, GPRS-WEP [ 80 ], and GPRS-WPA2 [ 80 ] datasets and compared the results. The results showed a good performance in attack detection. The average performance of DNN was validated using 10-fold cross-validation on the UNSW-NB15, CIDDS-001, GPRS-WEP, and GPRS-WPA2 datasets that resulted in 94.17%, 99.99%, 82.89%, and 94% accuracy, respectively. In future work, the researchers want to investigate a larger value of trial repetition given the unaffected performance of the different validation methods.

To mitigate IoT cybersecurity threats in a smart city, Alrashdi et al. [ 81 ] proposed an anomaly detection-IoT system using the RF model of ML. The UNSW-NB15 dataset was selected for this project, which includes 49 features and nine attack classifications to revise normal and abnormal behaviors. The resulting model could detect cyber-attacks at fog nodes in a smart city by monitoring the network traffic that passed through each node. After detection, it alerted the security cloud services to analyze and update their system. This solution achieved the highest classification accuracy of 99.34% with the lowest FPR while detecting compromised IoT devices at distributed fog nodes. Using open sources of distributed computing to distribute the model in fog nodes to detect IoT attack networks and using n-fold cross validation to evaluate performance metrics of design are some of the researchers’ future goals.

3.1.8. Malicious Traffic Classification

In order to protect organizations and individuals against cyber-attacks, network traffic first needs to be analyzed and classified so that anomaly and malicious attacks can be detected. As the role of malicious traffic classification is very important, many researchers have sought to improve classification techniques using the power of AI. Some studies have focused on anomaly and abnormal traffic. Yang et al. [ 82 ] built a model that found hidden abnormal traffic in the network to detect attacks using DL techniques. The dataset used was NetFlow campus information, which is a collection of data gathered by campus routers. For the pre-processing stage, the authors transformed the data into standardized format, and then the RNN algorithm was applied. The proposed model resulted in an accuracy of 98%. For future work, the authors propose to search for more critical features that could help in detecting further cyber-attacks.

Chou et al. [ 83 ] used AI algorithms through TensorFlow to train the system by providing it with rules and signatures to distinguish between normal and abnormal traffic behavior. The researchers developed a framework of a DL model on TensorFlow by combining multiple layers of non-linear features and training the system to learn the normal behavior using a forward propagation algorithm on the NSL-KDD dataset. The results were promising, showing high accuracy during testing of up to 97.65% in the detection of probing attacks and 98.99% in the detection of DDoS attacks. In future work, improvements need to be made in the training characteristics in TensorFlow as the present model could not predict user to root (U2R—attacker tries to gain unauthorized access posing as a normal user) and remote to local (R2L—attacker tries to gain unauthorized access by exploiting network vulnerabilities) attacks since the dataset sample was too monotonous, leading to over-learning.

An ensemble deep model to detect and classify anomalies at both the network and host levels was presented by Dutta et al. [ 84 ]. The datasets used were IoT-23 [ 61 ], LITNET-2020 [ 85 ], and NetML-2020 [ 86 ] and the DL techniques applied were DNN, long short-term memory (LSTM), and a meta-classifier (i.e., LR). A deep sparse autoencoder (DSAE) was used as the feature engineering technique and a stacking ensemble learning approach was used for classification. After testing on three heterogenous datasets, the researchers concluded that the suggested approach outperformed individual and meta-classifiers such as RF and SVM. In future work, the researchers suggest conducting experiments on more sophisticated datasets and using advanced computational methods to boost processing speed.

Sun et al. [ 87 ] built a traffic classification model using DL techniques, focusing on web and peer-to-peer (P2P) traffic. The dataset used to train the proposed model was collected by the authors by capturing traffic from the network using a distributed host-based traffic collection platform (DHTCP). In the training process, the dataset was divided by 5:5, 7:3, and 10-fold cross-validation for the first, second, and third experiment, respectively, and radial basis function neural network (RBFNN), SVM, and probabilistic neural network (PNN) were applied. The results showed that the highest accuracy was 88.18% when using PNN and dividing the dataset as 7:3 for training and testing.

Some researchers have focused on investigating the effects of network data representation on the intelligent models. Millar et al. [ 88 ] devised and compared three ways of network data representation for malicious traffic classification to deep learners: payload data, flow image, and flow statistics. They showed that malicious classes can be predicted using just 50 bytes of a packet’s payload. Since DL benefits from an extensive and large dataset, the UNSW-NB15 dataset was selected for the experiment. The payload-based method was found to have the best performance. However, all methods failed to accurately identify DDOS attacks. Since different malicious attacks exhibit different defining characteristics, there is no ‘one size fits all’ solution for identifying all attacks. Hence, in future work, the researchers propose to research the combination of payload-based and statistical inputs to identify malicious traffic.

Yang et al. [ 89 ] aimed to develop a model for malicious traffic detection of an encrypted network using DL. The model proposed was developed based on a residual neural network (ResNet), which can automatically identify features and effectively isolate contextual information of the encrypted traffic. Moreover, the CTU-13 dataset was used to train the model and, in the pre-processing stage, the data were converted into the IDX format, then traffic refinement, traffic purification, data length unification, and IDX file generation were performed. Then, deep Q-network (DQN) reinforcement learning, and deep convolution generative adversarial networks (DCGAN) were used to generate the encrypted traffic adversarial sample. This resolved the issue of unbalanced and insufficient or small samples. The model achieved a high accuracy of 99.94%. In future, the researchers will focus on delivering advanced genetic algorithms into DCGAN to enhance generator efficiency.

A new framework using ML for hardware-assisted malware detection by monitoring and memory access pattern classification was introduced by Xu et al. [ 90 ]. They proposed in-processor monitoring to obtain virtual address trace and addressed this by dividing accesses into epochs and summarizing the memory access patterns of each epoch into features, after which they are fed to ML classifiers, namely RF and LR. It was concluded that the best performing classifier was RF for both kernel rootkits and memory corruption attacks. Its accuracy in kernel rootkits detection reached a 100% TPR, with less than 1% FPR. As for user-level memory corruption attacks, the algorithm demonstrated a 99.0% DR with less than 5% FPR.

De Lucia et al. [ 91 ] proposed a malicious network traffic detection mechanism of encrypted traffic using two techniques—SVM and CNN. To conduct the experiments, the team leveraged a public dataset [ 92 ], which consisted of malicious and normal TLS network traffic packets. In data pre-processing, the desired TLS features were extracted from the packet captures using a custom program written in the PcapPlusPlus framework [ 93 ]. The train–test split was 70–30%. Both methods successfully achieved a high F-score and accuracy and a low FPR. However, SVM outperformed CNN by achieving a lower FPR and a slightly higher F-score, precision, accuracy, and recall.

While building ML models for the detection of normal or malicious traffic, it was observed that questions arise regarding the selection of the right features. With this in mind, Shafiq et al. [ 94 ] proposed a ML algorithm called weighted mutual information_ area under the curve (WMI_AUC), a hybrid feature selection algorithm, that helps in selecting the effective features in the traffic flow. The databases used in the study were the HIT Trace 1, which was captured by the authors from WeChat messenger using Wireshark, and the NIMS dataset, which was collected by the authors from their research-tested network. To build the final model, the researchers used 11 different ML algorithms. The model built using the partial decision tree (PART) algorithm resulted in an accuracy of 97.88% using the HIT Trace 1 dataset. For the NIMS dataset, RF resulted in an accuracy of 100%.

Another field that was also covered by researchers was the detection of malicious virtual private network (VPN) traffic. Miller et al. [ 95 ] proposed a computational model to address the current limitations in detecting VPN traffic and aid in the detection of VPN technologies that are being used to hide an attacker’s identity. A model was built to detect VPN usage by using a MLP trained neural network by flow statistics found in the captured network packets’ TCP header. The experiment using OpenVPN was able to identify VPN traffic with an accuracy of 93.71% and identify Stunnel OpenVPN with an accuracy of 97.82% when using 10-fold cross-validation. Future studies could be carried out to detect unauthorized user access and research organizational security, which is essential for a business.

Since the spread of malicious websites, research emphasis has been on factor analysis of the site category and the correct identification of unlabeled data in order to distinguish between benign and dangerous websites to mitigate the risk of malicious websites. Wang et al. [ 96 ] demonstrated the use of the NB model to classify malicious websites. A self-learning system was developed to categorize websites based on their features, with NB being used to divide the websites into two categories: malicious or benign. The dataset used was the ISCX2016 [ 97 ] dataset, which contains over 100,000 URLs and 50 features for each URL. A higher accuracy of up to 90% was achieved after applying factor identification of datasets and accomplishing website classification using the NB classifier, demonstrating that the NB classifier can perform well when it comes to website classification.

Finally, Ongun et al. [ 98 ] used the CTU-13 dataset to build ensemble models for malicious traffic detection. The algorithms used to build the model were LR, RF, and gradient boosting (GB). The first representation was connection-level representation where the features were extracted from the raw connection logs. The second representation was aggregated traffic statistics where the authors compared between raw features in the first representation and the features obtained by time aggregation in this representation. The last representation was temporal features, where the authors considered the time interval with the features obtained by time aggregation in the second representation. The best performance achieved by the model built using RF and GB and resulted in high AUC of 99% when applying it on the features of the third representation.

Malicious Traffic in a Cloud Environment

Using a dataset constructed from a real cloud environment, Alshammari and Aldribi [ 99 ] built ML models to detect malicious traffic in cloud computing. The dataset used was the new ISOT CID [ 100 ], a publicly available cloud-specific dataset where the training data contained 17,296 instances and testing had 7411 instances. Their aim was to add some significant features, prepare the training data, and test the dataset against different ML models, namely DT, KNN, NNet, SVM, NB, and RF. The dataset contained 89,364 instances among which 44,569 were malicious and 44,795 were normal instances. They performed both cross-validation (5-, 10-, 15-folds) and split–validation (90–10%, 80–20%, 70–30%). For cross-validation (all 5-, 10-, 15-folds), DT, RF, and KNN all obtained an accuracy of 100%. In the case of split validation (for all 90%, 80%, and 70% splitting), both DT and RF achieved an accuracy of 100%.

Using the same cloud dataset, Sethi et al. [ 101 ] proposed an IDS to protect cloud networks from cyber-attacks. The algorithm applied was double deep Q-learning (DDQN). The datasets used were the ISOT CID dataset, and the standard NSL-KDD dataset. The total size of ISOT is 8 TB, but for the purposes of the experiment, only the network traffic data portion was used. For the feature selection phase, the team applied a chi-square feature selection algorithm. The selected features were 164 and 36 for ISOT CID and NSL-KDD, respectively. The accuracy for the proposed model tested for NSL-KDD was 83.40%, whereas for ISOT CID, it was 96.87%. After measuring the robustness of their model against an adversarial attack, the accuracy obtained was 79.77% for NSL-KDD and 92.17% for ISOT CID.

Xie et al. [ 102 ] used a class SVM technique based on a short sequence model. They used the Australian Defense Force Academy (ADFA) dataset [ 103 ], which contains thousands of normal traces taken from a host setup to simulate a modern Linux server as well as hundreds of anomalous traces caused by six different types of cyber-attacks. As it was a short sequence, duplicate entries were removed, leading to an improved separability between the normal and abnormal. The k values chosen for this experiment were k = 3, 5, 8, 10, with k = 5 providing the greatest results and an accuracy of 70% attained at an FPR of roughly 20%. Although the experimental result showed a significant reduction in computing cost, the rate of an individual kind of attack mode recognition was low.

Vanhoenshoven et al. [ 104 ] addressed a variety of ML approaches to solve the challenge of detecting malicious URLs as a binary classification problem including multi-layer perceptron, DT, RF, and KNN. The researchers used Ma et al.’s dataset [ 105 ], called the Malicious URLs Dataset, which consists of 121 sets gathered over 121 days. There are 2.3 million URLs and 3.2 million features in the overall dataset. The researchers divided the URLs into three groups based on their characteristics. Each of the methods was used to classify these sets. The models were assessed based on their accuracy, precision, and recall, with features such as blacklists and WHOIS information taken into account. The article implies that all of its approaches achieved high accuracy, with RF being the most convenient approach to use, obtaining an accuracy of roughly 97% based on experimental results. The method also had great precision and recall, demonstrating its reliability.

For the purpose of detecting harmful URLs, Yuan et al. [ 106 ] introduced a parallel neural joint model approach. The semantic and text features were included in the method by integrating a parallel joint neural network incorporating capsule network (CapsNet) and independent RNN (IndRNN) to improve the detection accuracy. The malicious URLs data were gathered from two sources: an anti-phishing website called PhishTank and a malware domain list that collects a blacklist of harmful websites. The 5-fold cross-validation technique was applied and unified performance metrics were used to evaluate the model’s performance. According to the results of the experiments, the model performed best when the dimension of the feature was 185 and the number of IndRNN layers was 2. The accuracy and recall rates both reached 99.78% and 99.98%, respectively, resulting in a performance that exceeded traditional models.

By utilizing ML on the latest and more advanced dataset for IoT networks called IoTID 20 [ 107 ], Maniriho et al. [ 108 ] proposed an approach for anomaly-based intrusion detection in IoT networks. The ML algorithm applied was RF. The dataset had three subsets: subset 1 contained normal and DoS instances; subset 2 contained normal and man-in-the-middle (MITM), and subset 3 contained normal and scan traffic. A 10-fold cross-validation and train–test split of 70–30% were applied. The overall accuracy for each subset attack was DoS—99.95%, Scan—99.96%, and MITM—99.9761% using cross-validation while using the percentage split DoS—99.94%, Scan—99.93%, and MITM—99.9647.

Since the security of IoT networks is a major concern for researchers and decision-makers, some other researchers have used the same IoTID 20 dataset in order to build an IDS for in-home devices. A three-stage strategy that includes clustering with oversampling, reduction, and classification using a single hidden layer feed-forward neural network (SLFN) was provided by Qaddoura et al. [ 109 ]. The paper’s significance lies in the data reduction and oversampling techniques used to provide relevant and balanced training data as well as the hybrid combination of supervised and unsupervised techniques for identifying intrusion activities. With a ratio of 0.9 and a k value of 3 for the k-means++ clustering technique, the results showed that using the SLFN classification technique and using the SVM and synthetic minority oversampling technique (SVM-SMOTE) yielded more accurate results than using other values and classification techniques. Similarly, a deep multi-layer classification strategy was suggested by Quddoura et al. [ 110 ], which consisted of two phases of detection. The first phase entails detecting the presence of an intrusion and the second phase identifies the kind of intrusion. In preprocessing, the oversampling technique was carried out to enhance classification results. Furthermore, the most optimal model was built, which contained 150 neurons for the single-hidden layer feed-forward neural network (SLFN) (phase 1), and 150 neurons and two layers for LSTM (phase 2). When the findings were compared to well-known classification approaches, the suggested model outscored them by 78% with regard to the G-mean.

3.1.9. Attacks at DNS Level

In order to improve the user’s privacy, a new protocol called DNS over HTTP (DoH) was recently created. This protocol can be used instead of traditional DNS for domain name translation with the benefit of encryption. However, security tools depend on readable information from DNS to detect attacks such as malware and botnet. Hence, Singh and Roy [ 111 ] aimed to use ML algorithms to detect malicious DoH traffic. The five ML algorithms used were GB, NB, RF, KNN, and LR. The team conducted the experiment on the benchmark MoH dataset—CIRA-CIC-DoHBrw-2020, which was recently developed and shared publicly [ 112 ]. It contained a benign file that had 19,807 instances and a malicious file that had 249,836 instances. The DoHMeter tool [ 113 ], which was developed in Python and is freely available, was used to extract important features from the PCAP files. To build the model, the data were split into a train–test ratio of 70–30%. The experimental results showed that RF and GB attained the maximum accuracy of 100%.

3.1.10. Intrusion Detection

NIDS analyzes and monitors the whole network to detect malicious traffic. The following studies used the NSL-KDD dataset. Al-Qatf et al. [ 114 ] proposed self-taught learning (STL)-IDS using the DL approach in an unsupervised manner as a feature selection technique to reduce the testing and training time and effectively enhance the accuracy of the prediction for the SVM model. In the pre-processing phase, a 1-n encoding system was applied before STL. Max–min normalization was used to map all features into a specific range. The results obtained through the proposed model represented the classification accuracy of improved SVM compared with algorithms such as J.48, NB, and RF. Moreover, it performed well in five-category (normal and five types of attacks) and two-category (attacks and normal traffic) classification.

Similarly, to develop a flexible and efficient NIDS, Niyaz et al. [ 115 ] proposed a self-taught learning (STL) based on sparse autoencoder (AE) and soft-max regression (SMR) on the NSL-KDD dataset. The authors applied 10-fold cross validation on the training data for STL and applied the dataset directly for SMR. The results showed a high-performance accuracy rate of 98% for STL.

Following the same principle of using DL for intrusion detection, Zhang et al. [ 116 ] proposed an approach using the NSL-KDD dataset, consisting of normal and different forms of abnormal traffic. By first applying feature selection to remove the unrelated features and noise, the autoencoder was implemented to learn the features of the input data and extract the key features. Soft-max regression classification was then applied. The measures for evaluation used were accuracy, precision, recall, and F-score. Finally, the model achieved F-score and recall values of 76.47% and 79.47%, respectively.

Some studies have focused on multi-layer DL algorithms. Wu and Guo [ 117 ] proposed a LuNet model, which is a hierarchical CNN and RNN neural network, applied on the NSL-KDD and UNSW-NB15 dataset. They started by converting the categorical features using the ‘get dummies’ function in Pandas, then they applied standardization to scale input data and concluded by employing K-fold cross-validation. To evaluate LuNet, the following evaluation criteria were used: accuracy, FPR, and DR. The performance in binary classification achieved on average 99.24% on the NSL-KDD dataset and 97.40% accuracy on the UNSW-NB15 dataset. The performance in multiclass classification was an average of 99.05% accuracy on NSL-KDD, and 84.98% accuracy on UNSW-NB15. In future work, the researchers intend to investigate worms and backdoors as these were wrongly classified in the model.

To detect network intrusions efficiently, Hasan et al. [ 118 ] used an ANN. Different backpropagation algorithm training approaches were employed to detect the attacks and non-attack connections. The DARPA 1998 [ 119 ] intrusion detection dataset was used for training and testing purposes. To train the model, the researchers used the backpropagation learning algorithm, letting it detect intrusions in the following three modes: batch gradient descent with momentum (BGDM), batch gradient descent (BGD), and resilient backpropagation (RP). Finally, they used the DR and the FPR to determine the performance of intrusion detection. The total attack detection performance and the efficiency measure support the RP method of training, which obtained an accuracy of 92%. Further changes in the network architecture can be made to enable the efficient use of the network with other approaches.

Likewise, Devikrishna et al. [ 120 ] proposed an approach that used ANN as a pattern recognition technique to classify normal and attack patterns. The dataset used was the KDD-99 dataset. The feature extraction process consisted of feature selection and feature construction. An MLP was used for intrusion detection. MLP was a layered feed-forward ANN network typically trained with backpropagation. Accuracy was a goal that largely improved the overall effectiveness of the IDS. A possible future research direction could be to incorporate more attack scenarios in the dataset.

Abuadlla et al. [ 121 ] also proposed an IDS based on flow data built in two stages. The first stage involved the detection of abnormal traffic on the network. The second stage involved detecting and classifying the attack types in the network traffic. The NetFlow dataset made by network captures was employed to train the proposed system. To build the proposed model, a multilayer feedforward neural network and the radial basis function network (RBFN) were used. The proposed model resulted in a higher accuracy of 94.2% for the abnormal traffic detection stage, and 99.4% for the attack detection and classification stage. Although the multilayer feedforward neural network resulted in higher accuracy, it consumed more time and memory in comparison with RBFN, which makes RBFN a better choice for real-time detection. In future work, the researchers aim to build a faster and more accurate model for real-time detection with a smaller number of features.

Utilizing the KDD-99 dataset, Alrawashdeh et al. [ 122 ] aimed to build a DL model for anomaly detection in real-time. The researchers began by transforming categorical features into numerical features for convenience. Then, they removed the duplicated records to reduce computational time and improve performance. Three models were built: first using the restricted Boltzmann machine (RBM), the second using deep belief network (DBN), and the third using DBN with LR. The model that was built using DBN and LR resulted in the best performance with an accuracy of 97.9% and a FN rate of 2.47%.

In addition, Al-Janabi et al. [ 123 ] proposed a model based on ANN using the KDD-99 dataset and incorporated three scenarios: detection mode, detection and classification mode, and detailed classification mode. The researchers performed their experiment for each scenario by training the models using different number of features in each. The best results achieved were a 91% DR and 3% FP rate using 44 features with the detection only scenario. The results showed that performance decreased as a higher level of classification was performed.

Belavagi et al. [ 124 ] evaluated the different ML algorithms used to classify the network data traffic as normal traffic or intrusive (malicious) traffic. By using the NSL-KDD dataset consisting of internet traffic record data, supervised ML classifiers, namely LR, SVM, Gaussian NB, and RF were applied to identify four simulated attacks. After converting all the categorical data to numerical form in the pre-processing stage, the predicted labels from these models were compared with the actual labels, and TPR and FPR were computed. From the observed results, it was concluded that the RF classifier outperformed other classifiers for the considered dataset, with an accuracy of 99%. The researchers suggested that the work can be further extended by considering the classifiers for multiclass classification and considering only the important attributes for intrusion detection.

Additionally, Almseidin et al. [ 125 ] evaluated the different ML algorithms, keeping the focus on FNR (identifying an attack as normal traffic) and FPR (identifying normal traffic as an attack) performance metrics to improve the DR of the IDS. They used several algorithms, namely J.48, RF, random tree, decision table, multi-layer perception (MLP), NB, and Bayes network. The KDD-99 dataset was imported to SQL server 2008 to implement statistical measurement values such as attack types and occurrence ratios. Then, 148,753 record instances were extracted for training data. A wide range of results was obtained by using Weka tools that demonstrated that the RF achieved the highest average accuracy and the decision table achieved the lowest FNR.

Choudhury et al. [ 126 ] implemented ML algorithms to categorize network traffic as normal or anomalous. Algorithms such as BayesNet, LR, instance-based knowledge (IBK), J.48, PART, JRip, random tree, RF, REPTree, boosting, bagging, and blending were incorporated and compared. The researchers used the NSL-KDD dataset and Weka tools to model and compare the algorithms. The results showed that RF achieved the highest accuracy of 91.523%, and the lowest accuracy of 84.96% resulted from LR.

Similarly, the objective of the system proposed by Thaseen et al. [ 127 ] was to detect any intrusions in the network using ML by classifying different packets without decrypting their content. For intrusion detection analysis, packets were generated and transmitted over a network and were captured by Wireshark. This captured data was organized into a dataset. By implementing ML algorithms such as NB, SVM, RF, and KNN, the data were classified with an accuracy of 83.63%, 98.23%, 99.81%, and 95.13%, respectively. Future work to this study includes the plan to use DL algorithms to enhance the performance and accuracy of recognition and classifying different types of packets transmitted over a network.

Likewise, Churcher et al. [ 128 ] proposed several ML models to cope with the increase in the number of network attacks. The researchers highlighted several ML methods that were used in IDS such as DT, SVM, NB, RF, KNN, LR, and ANN. The Bot-IoT dataset [ 129 ] containing ten CSV files that have records of IoT network attacks and 35 features was used. In the pre-processing stage, the undesirable features were removed. The results of the model showed that in RF, the accuracy for DDoS attacks was 99% in binary classification and its performance was superior in the context of all types of attacks. However, KNN achieved 99% accuracy and outperformed other ML algorithms in the multiclass classification. In conclusion, KNN and ANN are more accurate when used in weighted and non-weighted datasets, respectively, for multiclass classification.

A comparative analysis of two commonly used classification methods, SVM and NB, to evaluate the accuracy and misclassification rate was conducted by Halimaa et al. [ 130 ] using the NSL-KDD dataset. For comparative analysis, the Weka tool’s randomized filter was used to ensure the random selection of 19,000 cases. The results showed that SVM attained an accuracy of 93.95% and NB achieved an accuracy of 56.54%. The researchers plan to work with larger amounts of data and construct a cross multistage model to create the ability to categorize additional attacks with accuracy and better performance.

Ghanem et al. [ 131 ] assessed the performance of their existing IDS against 1- and 2-class SVMs by applying both straight and non-linear forms. For the first step of data collection, they collected five datasets from the IEEE 802.11 network testbed and another dataset was collected in Loughborough University from an ethernet local area network office. All this traffic was collected in the PCAP structure using tcpdump. The results demonstrated that the linear 2-class SVM presented generally highly accurate findings. In addition to reaching a 100% success rate over four out of five of the metrics, it required training datasets. Meanwhile, the linear 1-class SVM’s performance was nearly as good as the best technique and did not require training the dataset. Overall, it was concluded that the existing unsupervised anomaly-based IDS can benefit from using any of the two ML techniques to improve accuracy in detection and its analysis of traffic, especially when it is comprised of non-homogeneous features.

Mehmood et al. [ 132 ] focused on supervised learning algorithms to make a comparison of three ML algorithms, namely SVM, J.48, NB, and decision table for anomaly-based detection. These algorithms were trained using the short version of the KDD-99 dataset as it has many records. The performance measures used in this comparison were FPR, TPR, and precision. The results highlighted a limitation when it came to DR, as not a single algorithm had a high DR for all the tested attacks in the KDD-99 dataset. However, the J.48 had a low misclassification rate. Hence, it was concluded that this algorithm performed best out of all the other algorithms.

An approach that boosts the capacities of wireless network IDS was introduced by AlSubaie et al. [ 133 ]. The dataset used was WSN-DS [ 134 ], which included 23 attributes and five potential outputs (four attacks (DoS attack): flooding, grayhole, blackhole, and scheduling and one normal state (no attack)). The ML algorithms used here were ANN and the J.48. Additionally, the data noise was calculated as it affects the accuracy of the ML algorithms. The amount of noise permissible for the ML model to be deemed trustworthy was determined. The results determined that J.48 performed better than the ANN when noise was not considered, obtaining the highest accuracy rate of 99.66%. With datasets having more noise, ANN was more tolerable.

In order to determine which of the models could handle large amounts of data and still produce accurate predictions, Ahmad et al. [ 135 ] used the SVM linear and radial basis function (RBF), RF, and ELM methods and compared their performance on the NSL-KDD dataset. The results demonstrated that when using the full dataset, the ELM outperformed the other algorithms in terms of all the metrics being tested in all experiments including accuracy, which reached 99.5%. On the other hand, when using half and a quarter of the dataset, SVM performed better overall, with an accuracy of around 98.5%. Hence, it was concluded that ELM is best suited for intrusion detection when dealing with large amounts of data. The researchers plan to further explore ELM and experiment with it using different selection and feature transformation techniques and their impact on its performance.

Amira et al. [ 136 ] found MLP to be the most effective and appropriate classifier to increase detection accuracy. The data pre-processing phase was carried out using the equal width binning algorithm. The sequential floating forward selection (SFFS) feature selection technique was applied, resulting in the selection of 26 features. Using the NSL-KDD dataset, Amira et al. then applied a multi-agent, 2-layer classification algorithm. The different classifiers that were tested and compared were: NB and DT, namely NBTree, BFTree, J.48, and RF Tree. NBTree and BFTree gave better results than RF and J.48. MLP gave good results in terms of classifying normal and DoS attacks compared to identifying the R2L and U2R attacks. Overall, it was concluded that a single classifier is not sufficient to classify the attack class. Therefore, to increase classification accuracy, multiple classifiers must be involved.

Rather than comparing different techniques, Gogoi et al. [ 137 ] focused on evaluating the clustering approach to detect network traffic anomalies on different datasets. The proposed method was evaluated using TUIDS [ 138 ] datasets, the NSL-KDD dataset, and the KDD-99 datasets. The real-life TUIDS intrusion datasets consist of three datasets: flow level, packet level, and port scan. After the pre-processing stage, they applied a combination of supervised clusters and unsupervised incremental clusters which labelled the training data into different profiles (or rules). The prediction was undertaken using a supervised classification algorithm. Using the TUIDS dataset, the packet level had the highest accuracy of 99.42%. When using the KDD-99 dataset, the accuracy achieved was 92.39%. Finally, using NSL-KDD, the accuracy achieved was 98.34%.

Aiming to classify real-time traffic by using 12 features of network traffic data to classify 17 attack types of DoS, probing as well as normal was conducted by Wattanapongsakorn et al. [ 139 ]. Supervised ML techniques—DT, ripple rule, back-propagation neural network, and Bayesian network—were applied. In the pre-processing stage, the team used a packet sniffer and a built-in Jpcap library to collect and store network records over a period of time. Then, in the classification part, training and testing were performed using Weka tool, and results were observed. The DT approach achieved the highest DR of 85.7%. In the second experiment, some attack types were grouped together, and training data consisted of 9000 records with 600 records of each attack type (so 600 × 15). In this case, the DR was much higher, with the DT being 95.5%.

Further research that worked on enhancing an existing algorithm for intrusion detection was done by Cui et al. [ 140 ], who worked on enhancing the Bayes classifier (BC). The proposed method seeks to integrate the spatiotemporal patterns of measurement into a flexible BC to detect cyber-attacks. Spatiotemporal patterns were captured by the graph Laplacian matrix for system measurements. After the evaluation of the developed method’s performance, it was concluded that the flexible BC showed the largest TPR compared with the naïve BC, SVM, and DT methods, which verified the effectiveness of the developed method. For future work, DL techniques will be involved by mapping the spatiotemporal patterns to a linear space using the LSTM network for better detection accuracy of cyber-attacks.

Moreover, Kumar et al. [ 141 ] focused on enhancing the detection efficiency by combining three algorithms—RF, JRIP, PART—to identify threats of mobile devices. The dataset used contained around 600 samples that were captured by the researchers from the virtual machine using Wireshark. For feature extraction, the researchers used bidirectional flow export using the IP flow information export method (RFC-5103 BiFlow). The challenge the researchers faced was an overfitting problem and concept drift condition, which is caused by choosing low performance giving features. The ensemble model resulted in an accuracy of 98.2% with the ability to identify benign traffic. For future work, the researchers aim to integrate ML with conventional NIDS and to reduce the chance of concept drift by introducing innovative methods.

Similarly, Tahir et al. [ 142 ] constructed a hybrid ML technique for detecting network traffic as normal or intrusive by combining K-means clustering and SVM classification to improve the DR and to reduce the FPR alarm and FNR alarm. The dataset applied in the proposed technique was the NSL-KDD dataset. Pre-processing was performed on the dataset to reduce ambiguity and supply accurate information to the detection engine. After applying the classifier subset evaluator and best-first search algorithms, both the classifiers—K-means and SVM—were then tested and their performance evaluated. The hybrid ML technique results showed that they attained 96.26% as the DR and 3.7% as the FNR. The model showed a comparatively higher detection for DoS, PROBE, and R2L attacks.

One more enhanced technique was proposed by Sharma et al. [ 143 ] to apply efficient data mining algorithms for detecting network traffic as normal or anomalous. The team applied KDD-99, which contains 4.9 M data instances and four class types. In feature selection, they collected basic features such as protocol type, duration, flags, etc. The data was normalized and the classification was carried out using k-means clustering via a NB classifier. The target variable was classified as normal, DoS, U2L, R2L, probing. The DR achieved by using the proposed method was 99%.

Following the same ideology, Lehnert et al. [ 144 ] built their system in steps with more complexity added at each level. They used the KDD-99 dataset and Shogun ML Toolbox to test and train the data. The study’s focus was mainly on using the SVM implementation provided by the toolbox. The key step in this paper was the training phase, which was done using labelled data. The goal was to attempt to choose the most appropriate kernel and minimize the number of features. The results showed that two out of the four available kernels on Shogun tied in the best accuracy. These kernels were Gaussian and Sigmoid, which produced an error of only 2.79%. It was concluded that identifying both the kernel that has the lowest error rate and the subset of the most relevant features leads to an improved version of the algorithm. Ultimately, this can enhance the accuracy and efficiency of the SVM applied for intrusion detection, making it able to predict with higher speed and accuracy.

An innovative feature selection algorithm called the ‘highest wins (HW)’ was proposed by Mohammad and Alsmadi [ 145 ] in order to enhance intrusion detection. This HW algorithm was applied in NB techniques on 10 benchmark datasets from the UCI repository to evaluate its performance. The results showed that the proposed HW algorithm could successfully reduce the dimensionality for most of these datasets compared to other feature selection methods such as chi-square and IG. The team conducted another set of experiments where NB and DT (C4.5) classifiers were built using the HW technique on the NSL-KDD dataset on its binary and multiclass versions. For binary, HW reduced the features of the dataset from 41 to eight and the results gave an accuracy of 99.33% using the reduced features (0.23% decrease compared to using complete features). For multiclass, HW reduced the features of the dataset from 41 to 11, and in terms of time needed for building the model, reduced features had an enhancement of 2.3%. The results demonstrated that instead of using all 41 features of this dataset, using only eight by applying HW could produce classifiers with the same classification performance.

Furthermore, Chawla et al. [ 146 ] proposed a computational efficient anomaly-based IDS that was a combination of CNN and RNN. To detect malicious system calls, they merged stacked CNNs with GRUs. Using the ADFA dataset of system call traces, they obtained a set of equivalent findings with shorter training periods when using GRU. They employed CNN to extract the local features of system call sequences and feed them into the RNN layer, which was then processed through a fully connected SoftMax layer, which generates a probability distribution across the system calls processed by the network. Trained on normal system calls, which predict the likelihood of a subsequent system call, a testing sequence was employed to detect a malicious trace based on a pre-defined threshold. The RNN-based LSTM model’s training time was claimed to be reduced by the researchers.

In addition, Nguyen et al. [ 147 ] used the DL approach for detecting cyber-attacks in a mobile cloud environment. The used datasets were KDD-99, NSL-KDD, and UNSW-NB15 (training = 173,340 records, testing = 82,331 records). The researchers adopted principal component analysis (PCA) to reduce the dimensions for the datasets and the learning process comprised of three layers: the input layer, hidden layers, and output layer. The input layer used Gaussian restricted Boltzmann machine (GRBM) to transform real values to binary code. The hidden layer used restricted Boltzmann machine (RBM) to perform the learning process. The output of the hidden layer was used as input in the output layer (SoftMax regression step). They used accuracy, recall, and precision for measuring performance. The results showed that the accuracy for NSL-KDD, UNSW-NB15, and KDD-99 datasets, respectively, were 90.99%, 95.84%, and 97.11%. For future work, Nguyen et al. proposes implementing the model on real devices to measure the accuracy on a real-time basis and evaluate the energy and time consumed in the detection.

An improved IDS was proposed by Tama et al. [ 148 ] where they used two datasets to evaluate the performance of the model: NSL-KDD and UNSW-NB15. To minimize the feature size, a hybrid feature selection technique was used. The hybrid feature selection consisted of three methods: the ant colony algorithm, particle swarm optimization, and genetic algorithm. Then, the researchers proposed a two-stage classifier ensemble, which was rotation forest and bagging. The proposed model achieved an accuracy of 85.8% with the NSL-KDD dataset and 91.27% with the UNSW-NB15 dataset. For future work, the researchers intend to perform the proposed model to solve the multiclass classification problem.

A novel intrusion detection system was proposed that takes the advantage of both statistical features and payload features by Min et al. [ 149 ]. They used the ISCX2012 dataset, which is more updated and closer to reality, and they utilized word embedding and text-CNN to extract more features from the payloads. Then, the RF algorithm was applied on the combination of payload features and statistical features, where they named the model with TR-IDS. Moreover, the effectiveness of TR-IDS was compared against five ML models, namely SVM, NN, CNN, and RF (RF-1) and RF (RF-2, which used statistical features only). The highest result achieved was by TR-IDS with an accuracy of 99.13%.

Finally, more information on intrusion detection using unsupervised and hybrid methods can be found in a survey paper composed by Nisioti et al. [ 150 ]. They presented and highlighted important issues such as feature engineering methods for IDS. Furthermore, using IDS data to construct and correlate attacks to identify attackers as well as extending the current IDS to identify modern attacks were all addressed by the paper.

Table 2 below presents a summary of all details discussed in this section, giving overview picture of all reviewed articles in terms of research problem domain targeted, dataset used, and intelligent techniques applied as well as the results achieved.

Brief summaries of the reviewed papers.

3.2. Common Intelligent Algorithms Applied

In this literature review, a number of papers were studied between the period of 2010–2021 and a plethora of both ML and DL techniques were utilized in these papers to build or compare models to detect and classify network attacks. Table 3 presents a list of all the respected papers that utilized the different algorithms, highlighting all problem domains where each algorithm was used for as well as the highest performance achieved. Figure 1 presents the number of articles that utilized each algorithm. As seen from the figure and table, RF and SVM were the most widely used algorithms in a good number of papers and ELM was the least applied algorithm. For ML algorithms, the best performing algorithms were DT, RF, and KNN with their accuracy reaching up to 100% and the least utilized algorithms were J.48 and KNN. For DL algorithms, the best performing algorithm was RRN with the highest accuracy of 100% achieved and the least utilized and least popular algorithm was ELM, which is considered to be fast in terms of training as it consists of a single hidden layer, so it is usually applied to simple applications. However, it has recently been extended to be hierarchical to handle more complex problems with higher accuracy [ 152 ].

An external file that holds a picture, illustration, etc.
Object name is sensors-21-07070-g001.jpg

ML and DL algorithms used in the reviewed papers.

ML and DL algorithms evaluated in the reviewed papers.

3.3. Common Datasets Used

There are several datasets used by researchers in the reviewed papers to evaluate their network detection and classification model. The most widely used dataset is NSL-KDD due to the reasonable size of its training and testing sets and is also available publicly. There are 41 features in the NSL-KDD dataset. It is an enhanced version of the KDD dataset and removed the duplication of the records to eliminate the bias of the classifiers. Then, KDD-99 and CICIDS2017 came after NSL-KDD. The KDD-99 dataset was used for the first time in a competition and is an improved version of DARAP98. The CICIDS2017 dataset contains normal and new attacks and was published in 2017 by the Canadian Institute for Cybersecurity (CIC).

After that, the UNSW-NB15 dataset comes next in terms of repeatedly being used. The IXIA tool was used for creating the UNSW-NB15 dataset and it consists of nine types of attacks.

There are many other datasets, however, few researchers have tried to create their datasets. The CTU-13 dataset was captured by CTU University in the Czech Republic. It contains real botnet traffic combined with normal traffic and contains thirteen scenarios including legitimate traffic and attacks such as DoS. The SNMP-MIB dataset consists of about 4998 records with 34 variables. The attacks recorded in the data include six DoS attacks (TCP-SYN, ICMP-ECHO, HTTP flood, UDP flood, Slowloris, Slowpost) and web brute force attacks. The Kyoto 2006+ dataset was built from real traffic data from Kyoto University’s Honeypots over three years, from November 2006 to August 2009. The Kyoto 2006+ dataset consists of 24 features, 14 of which are derived from the KDD-99 dataset and 10 additional features that can be used to analyze and evaluate the IDS network. Honeypots, email server, darknet sensors, and web crawler were used to construct the Kyoto 2006+.

ADFA is an IDS that includes three data types in its structure: (1) normal training data with 4373 traces; (2) normal validation data with 833 traces; and (3) attack data with 10 attacks per vector. As the web became a significant internet criminal activity platform, the security community put in efforts to blacklist malicious URLs. Ma et al.’s dataset [ 153 ] consists of 121 sets with overall 2.3 million URLs and 3.2 million features in the dataset. The researchers divided the URLs into three groups based on their characteristics, with features being identified as binary, non-binary, numerical, or discrete.

Table 4 lists all the respected papers that utilized the different datasets, highlighting the main references for all datasets as well as the last year when each dataset was used. Figure 2 presents the number of articles that utilized each dataset.

An external file that holds a picture, illustration, etc.
Object name is sensors-21-07070-g002.jpg

Datasets used in the reviewed papers.

Network traffic datasets used in the reviewed papers.

4. Discussion and Conclusions

Network security is a major concern for individuals, profit, and non-profit organizations as well as governmental organizations. In fact, with the digital explosion that we are witnessing in the present era, ensuring network security is an urgent necessity in order to safeguard society’s acceptance for thousands and thousands of services that rely essentially on the backbone of the digital life, which is the network. Therefore, network security turns out to be an urgent requirement, and not a luxury. Although many protection methods have been introduced, there are still some vulnerabilities that are exploited by hackers, leaving the network security administrators in a continuous race against the network attackers. Techniques that hover around the use of intelligent methods, namely machine learning (ML) and deep learning (DL) have proved their merits in several domains including health care systems, financial analysis, higher education, energy industry, etc. This indeed motivated the people responsible for the network security to further explore the ability of these techniques in providing the required level of network security. Consequently, several intelligent security techniques have been offered in the past few years. Although these techniques showed exceptional performance, the problem has not been resolved entirely. This leaves us in a position to critically evaluate the currently offered solutions to recognize the possible research directions that might lead to building more secured network environments.

The complication of using the right dataset and features or the right ML and DL algorithms to identify the different attack types has proven to be an arduous decision for experts to make. Hence, among the reviewed papers, some researchers focused on comparing different algorithms to determine which algorithm to use for building an intelligent model using a training dataset. As no algorithm has been found to be a silver bullet for identifying and classifying all attacks with high accuracy, it was widely noted that it is not reasonable to accept a single algorithm as a universal model.

When building any intelligent system, the designer should take into account what is/are the algorithm(s) that best fit the domain. Not only this, but the designer should also decide which dataset comprises a set of features that better represent the classification area. Considering the network attacks, this research article found that RF is the most commonly used algorithm and this can be justified due to the fact that it uses an ensemble learning technique, which to some extent might ensure a life-long system due to the exceptional capability to continuously learn new knowledge on the fly. Producing models with reduced overfitting is another motivation behind using the RF. Not only this, but RF can also be effectively applied on both categorical and continuous features, and thus it can be applied to a wide range of datasets. In addition, the exceptional ability to handle missing data puts RF as a first option when building network attack mitigation models taking into account that most of the datasets are susceptible to include missing values. However, since RF produces complex trees, building a real-life system based on RF could be a challenging task because it might require more computational power and resources, while in fact, the main success factor for building a system for detecting network attacks is the quick and instant reaction. SVM is the second most widely used algorithm. However, SVM is applied to a fewer number of network attacks when compared to RF. This can be justified due to the fact that SVM produces complex intelligent models that are difficult to apply in real life. Nevertheless, SVM is considered as the main competitor to RF due to the fact that it shares several advantages with RF such as the exceptional capability to deal with missing values, and the remarkable capability to reduce the overfitting problem. NB ranks in third place, but still did not achieve the same predictive performance as RF and SVM due to the fact that it assumes that the dataset features are independent, which in fact, is not true in most training datasets. DT was employed almost half the time that RF and SVM were used. DT proved its merits in several domains, but in the network security domains, it has not been used very much. This can be justified due to the fact that it produces a set of rules that if exposed to the attackers, they can adopt their attacks by avoiding the rules adopted from the DT models.

Included among the algorithms that conveyed excellent performing results were DL models, namely, DNN and RNN as well as ML models, namely, RF and DT with their accuracies reaching up to 100%. A more promising research direction to explore can increasingly be toward applying hybrid or ensemble models to improve attack detection accuracy; for instance, augmenting DL techniques such as CNN with long short-term memory (LSTM) for automating feature engineering and improving network attack detection accuracy. Furthermore, gated recurrent unit (GRU), initially proposed in 2014, can further be applied by researchers in solving various problem domains in network security as it is considered more efficient than LSTM, and it uses comparatively less memory, and executes faster. They can solve complex problems faster, if trained well, and therefore, they are worth trying in network attack detection, namely for DDoS or in IoT networks.

Since the performance of the intelligent models largely depend on the datasets used for training them, it is important to analyze and evaluate which dataset to use for which type of attack. It is recommended that large datasets are used with a good distribution of each class type to increase the detection and classification accuracy. Moreover, limited availability of such datasets represents a challenge in the development of more robust intelligent-based models and highlights the need for producing and publishing more new datasets in different network attack problem domains. Most of the authors in the reviewed articles used the KDD-99 dataset as well as its latest version, the NSL-KDD dataset. However, the ADFA dataset was also used by some, which was proposed as a replacement for the KDD-99 dataset, ISOT HTTP for botnet, ISOT CID for cloud environments, and IoT20 for IoT environments, so can be explored further and used to build different ML and DL models.

Identifying malicious and benign URLs was also a fundamental research direction carried out by researchers where an important set of features that affected the model accuracy were URL related features. It was found that additional improvements in classifying malicious and benign URLs can be accomplished by deploying a lexical approach, which uses static lexical features extrapolated from the URL, in addition to analyzing the URL contents for instantaneous and reliable results. Hence, using a lexical approach to classify URLs can be an important direction to explore.

Several other problem domains need to be explored as they could be a valuable direction for enhancing network security in the modern world. Namely, with the growing establishment of encrypted network traffic as well as virtual private networks, more research needs to be carried out in detecting malicious traffic in these domains using intelligent techniques as not enough research has been focused in this area. Furthermore, with the rising number of inter-connected devices and the establishments of Internet of Things (IoTs) networks, more investigation needs to be carried out in assessing different intelligent techniques on new datasets such as IoT20 as well as paving ways to developing software that can detect and analyze data packets communicated in IoT environments to update the existing datasets for more attacks. Additionally, a new protocol called DNS over HTTP (DoH) has been created recently for which more research needs to be explored on detecting malicious DoH traffic at this (DNS) level.

Finally, multiple researchers intend in their future work to convert the models they built into a real-time system in order to benefit from them in real-life scenarios such as in attack detection and prevention. There are two levels of real-time ML which are online predictions and online learning. Online prediction means making predictions in real-time. Furthermore, online learning allows for the system to incorporate new data and update the model in real-time. Hence, converting intelligent models into real time systems may be considered as a fundamental direction to probe by more researchers.

Author Contributions

Conceptualization, M.A. (Malak Aljabri), S.S.A., R.M.A.M. and S.H.A.; methodology, M.A. (Malak Aljabri), S.S.A., R.M.A.M. and S.H.A.; software, S.M., F.M.A., M.A. (Mennah Aboulnour), D.M.A., D.H.A. and H.S.A.; validation, M.A. (Malak Aljabri), S.M. and F.M.A.; formal analysis, M.A. (Malak Aljabri), S.M. and F.M.A.; investigation, M.A. (Malak Aljabri), S.M., F.M.A., M.A. (Mennah Aboulnour), D.M.A., D.H.A. and H.S.A.; resources, M.A. (Malak Aljabri), S.M., F.M.A., M.A. (Mennah Aboulnour), D.M.A., D.H.A. and H.S.A.; data curation, S.M. and F.M.A.; writing—original draft preparation, M.A. (Malak Aljabri), S.M., F.M.A., M.A. (Mennah Aboulnour), D.M.A., D.H.A. and H.S.A.; writing—review and editing, M.A. (Malak Aljabri), S.M., F.M.A., S.S.A., R.M.A.M. and S.H.A.; visualization, S.M. and F.M.A.; supervision, M.A. (Malak Aljabri); project administration, M.A. (Malak Aljabri); funding acquisition, M.A. (Malak Aljabri) and S.S.A. All authors have read and agreed to the published version of the manuscript.

We would like to thank SAUDI ARAMCO Cybersecurity Chair for funding this project.

Conflicts of Interest

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Communications and Networking Research Group

Search form.

  • Publications

PUBLICATIONS

Journal articles | other papers | conference papers | book chapters | technical reports, journal articles.

134. Vishrant Tripathi, Nick Jones, Eytan Modiano, Fresh-CSMA: A Distributed Protocol for Minimizing Age of Information, IEEE Journal on Communications and Networks, 2024.

133. Bai Liu, Quang Nguyen, Qingkai Liang, Eytan Modiano, Tracking Drift-Plus-Penalty: Utility Maximization for Partially Observable and Controllable Networks, IEEE/ACM Transactions on Networking, 2024.

132. Xinzhe Fu, Eytan Modiano, Optimal Routing to Parallel Servers with Unknown Utilities – Multi-armed Bandit With Queues, IEEE/ACM Transactions on Networking, January 2022.

131. Bai Liu, Qingkai Liang, Eytan Modiano, Tracking MaxWeight: Optimal Control for Partially Observable and Controllable Networks, IEEE/ACM Transactions on Networking, August 2023.

130. Xinzhe Fu, Eytan Modiano, Joint Learning and Control in Stochastic Queueing Networks with unknown Utilities, Proceedings of the ACM on Measurement and Analysis of Computing Systems, 2023.

129. Vishrant Tripathi, Rajat Talak, Eytan Modiano, Information Freshness in Multi-Hop Wireless Networks, IEEE/ACM Transactions on Networking,” April 2023.

128.  Xinzhe Fu, Eytan Modiano, “ Learning-NUM: Network Utility Maximization with Unknown Utility Functions and Queueing Delay ,”  IEEE/ACM Transactions on Networking,” 2022.

127.  Bai Liu, Qiaomin Xie, Eytan Modiano,  " RL-QN: A Reinforcement Learning Framework for Optimal Control of Queueing Systems ,"  ACM Transactions on Modeling and Performance Evaluation of Computing Systems (TOMPECS), 2022.

126. Xinzhe Fu and E. Modiano,  “ Elastic Job Scheduling with Unknown Utility Functions ,” Performance Evaluation, 2021.

125. Bai Liu and E. Modiano, “ Optimal Control for Networks with Unobservable Malicious Nodes ,”  Performance Evaluation, 2021.

124. Vishrant Tripathi, Rajat Talak, Eytan Modiano, " Age Optimal Information Gathering and Dissemination on Graphs ,”  Transactions on Mobile Computing, April 2021.

123.  Xinyu Wu, Dan Wu, Eytan Modiano, “ Predicting Failure Cascades in Large Scale Power Systems via the Influence Model Framework, ”  IEEE Transactions on Power Systems, 2021.

122.   Roy D. Yates, Yin Sun, D. Richard Brown III, Sanjit K. Kaul, Eytan Modiano and Sennur Ulukus, “ Age of Information: An Introduction and Survey, ”  Journal on Selected Areas in Communications, February 2021.

121.   Jianan Zhang, Abhishek Sinha, Jaime Llorca, Anonia Tulino, Eytan Modiano, “ Optimal Control of Distributed Computing Networks with Mixed-Cast Traffic Flows ,”  IEEE/ACM Transactions on Networking, 2021.

120.   Thomas Stahlbuhk, Brooke Shrader, Eytan Modiano, " Learning Algorithms for Minimizing Queue Length Regret ,”  IEEE Transactions on Information Theory, 2021.

119.   Thomas Stahlbuhk, Brooke Shrader, Eytan Modiano, “ Throughput Maximization in Uncooperative Spectrum Sharing Networks ,”  IEEE/ACM IEEE/ACM Transactions on Networking, Vol. 28, No. 6, December 2020.

118.   Thomas Stahlbuhk, Brooke Shrader, Eytan Modiano, “ Learning algorithms for scheduling in wireless networks with unknown channel statistics ,” Ad Hoc Networks, Vol. 85, pp. 131-144, 2019.

117.   Rajat Talak, Eytan Modiano, “ Age-Delay Tradeoffs in Queueing Systems ,”  IEEE Transactions on Information Theory, 2021.

116.   Rajat Talak, Sertac Karaman, Eytan Modiano, " Improving Age of Information in Wireless Networks with Perfect Channel State Information ,”  IEEE/ACM Transactions on Networking, Vol. 28, No. 4, August 2020.

115.   Igor Kadota and Eytan Modiano, “ Minimizing the Age of Information in Wireless Networks with Stochastic Arrivals ,” IEEE Transactions on Mobile Computing, 2020.

114.   Rajat Talak, Sertac Karaman, Eytan Modiano, “ Optimizing Information Freshness in Wireless Networks under General Interference Constraints ,”  IEEE/ACM transactions on Networking, Vol. 28, No. 1, February 2020.

113.   X. Fu and E. Modiano, " Fundamental Limits of Volume-based Network DoS Attacks ," Proceedings of the ACM on Measurement and Analysis of Computing Systems, Vol. 3, No. 3, December 2019. 

112.   Rajat Talak, Sertac Karaman, Eytan Modiano, “ Capacity and Delay Scaling for Broadcast Transmission in Highly Mobile Wireless Networks ,” IEEE Transactions on Mobile Computing, 2019.

111.   Abhishek Sinha and Eytan Modiano, “ Throughput-Optimal Broadcast in Wireless Networks with Point-to-Multipoint Transmissions , IEEE Transactions on Mobile Computing, Vol. 19, No. 9, September 2020.

110.   Yu-Pin Hsu, Eytan Modiano, Lingjie Duan, “ Scheduling Algorithms for Minimizing Age of Information in Wireless Broadcast Networks with Random Arrivals ,”  IEEE Transactions on Mobile Computing, Vol. 19, No. 12, December 2020.

109.   Xiaolin Jiang, Hossein S. Ghadikolaei, Gabor Fodor, Eytan Modiano, Zhibo Pang, Michele Zorzi, Carlo Fischione, " Low-latency Networking: Where Latency Lurks and How to Tame It ,”  Proceedings of the IEEE, 2019.

108.   Jianan Zhang, Edmund Yeh, Eytan Modiano, “ Robustness of Interdependent Random Geometric Networks ,” IEEE Transactions on Network Science and Engineering, Vol. 6, No. 3, July-September 2019.

107.   Qingkai Liang, Hyang-Won Lee, Eytan Modiano, “ Robust Design of Spectrum-Sharing Networks ,” IEEE Transactions on Mobile Computing, Vol. 18, No. 8, August 2019.

106.   A. Sinha, L. Tassiulas, E. Modiano, “ Throughput-Optimal Broadcast in Wireless Networks with Dynamic Topology ,”  IEEE Transactions on Mobile Computing, Vol. 18, No. 5, May 2019.

105. Igor Kadota, Abhishek Sinha, Eytan Modiano, “ Scheduling Algorithms for Optimizing Age of Information in Wireless Networks With Throughput Constraints ,”  IEEE/ACM Transactions on Networking, August 2019.

104.   Igor Kadota, Abhishek Sinha, Rahul Singh, Elif Uysal-Biyikoglu, Eytan Modjano, “ Scheduling Policies for Minimizing Age of Information in Broadcast Wireless Networks ,” IEEE/ACM Transactions on Networking, Vol. 26, No. 5, October 2018.

103.   Jianan Zhang and Eytan Modiano, “ Connectivity in Interdependent Networks ,”  IEEE/ACM Transactions on Networking, 2018.

102.   Qingkai Liang, Eytan Modiano, “ Minimizing Queue Length Regret Under Adversarial Network Models ,” Proceedings of the ACM on Measurement and Analysis of Computing Systems, Volume 2, Issue 1, April 2018, Article No.: 11, pp 1-32. (same as Sigmetrics 2018).

101.   A. Sinha and E. Modiano, “ Optimal Control for Generalized Network Flow Problems ,”  IEEE/ACM Transactions on Networking, 2018.

100.   Hossein Shokri-Ghadikolaei, Carlo Fischione, Eytan Modiano  “ Interference Model Similarity Index and Its Applications to mmWave Networks ,”  IEEE Transactions on Wireless Communications, 2018.

99.   Matt Johnston, Eytan Modiano, “ Wireless Scheduling with Delayed CSI: When Distributed Outperforms Centralized, ’ IEEE Transactions on Mobile Computing, 2018.

98.   A. Sinha, G. Paschos, E. Modiano, “ Throughput-Optimal Multi-hop Broadcast Algorithms ," IEEE/ACM Transactions on Networking, 2017.

97.   Nathan Jones, Georgios Paschos, Brooke Shrader, Eytan Modiano, " An Overlay Architecture for Throughput Optimal Multipath Routing ,” IEEE/ACM Transactions on Networking, 2017.

96.   Greg Kuperman, Eytan Modiano, “ Providing Guaranteed Protection in Multi-Hop Wireless Networks with Interference Constraints ,” IEEE Transactions on Mobile Computing, 2017.

95.   Matt Johnston, Eytan Modiano, Isaac Kesslassy, “ Channel Probing in Opportunistic Communications Systems ,”  IEEE Transactions on Information Theory, November, 2017.

94.   Anurag Rai, Georgios Paschos, Chih-Ping Lee, Eytan Modiano, " Loop-Free Backpressure Routing Using Link-Reversal Algorithms ", IEEE/ACM Transactions on Networking, October, 2017.

93.   Matt Johnston and Eytan Modiano, “" Controller Placement in Wireless Networks with Delayed CSI ,” IEEE/ACM Transactions on Networking, 2017.

92.   Jianan Zheng, E. Modiano, D. Hay, " Enhancing Network Robustness via Shielding ,”  IEEE Transactions on Networking, 2017.

91.   M. Markakis, E. Modiano, J.N. Tsitsiklis, “ Delay Analysis of the Max-Weight Policy under Heavy-Tailed Traffic via Fluid Approximations ,” Mathematics of Operations Research, October, 2017.

90.   Qingkai Liang and E. Modiano, “ Survivability in Time-Varying Graphs ,”  IEEE Transactions on Mobile Computing, 2017.

89.   A. Sinha, G. Paschos, C. P. Li, and E. Modiano, “ Throughput-Optimal Multihop Broadcast on Directed Acyclic Wireless Networks ," IEEE/ACM Transactions on Networking, Vol. 25, No. 1, Feb. 2017.

88.   G. Celik, S. Borst, , P. Whiting , E. Modiano, “ Dynamic Scheduling with Reconfiguration Delays ,”  Queueing Systems, 2016.

87.  G. Paschos, C. P. Li, E. Modiano, K. Choumas, T. Korakis, “ In-network Congestion Control for Multirate Multicast ,”   IEEE/ACM Transactions on Networking,  2016.

86.   H. Seferoglu and E. Modiano, “ TCP-Aware Backpressure Routing and Scheduling ,” IEEE Transactions on Mobile Computing, 2016.

85.   H. Seferoglu and E. Modiano, “ Separation of Routing and Scheduling in Backpressure-Based Wireless Networks ," IEEE/ACM Transactions on Networking, Vol. 24, No. 3, 2016.

84.   M. Markakis, E. Modiano, J.N. Tsitsiklis, “ Delay Stability of Back-Pressure Policies in the presence of Heavy-Tailed Traffic ,”  IEEE/ACM Transactions on Networking, 2015.

83.   S. Neumayer, E. Modiano,  “ Network Reliability Under Geographically Correlated Line and Disk Failure Models ,” Computer Networks, to appear, 2016.

82.   S. Neumayer, E. Modiano, A. Efrat, “ Geographic Max-Flow and Min-Cut Under a Circular Disk Failure Model ,” Computer Networks, 2015.

81.   Marzieh Parandehgheibi, Hyang-Won Lee, Eytan Modiano, Survivable Path Sets:  A new approach to survivability in multi-layer networks ,”  IEEE Journal on Lightwave Technology, 2015.

80.   G. Kuperman, E. Modiano, A. Narula-Tam, “ Network Protection with Multiple Availability Guarantees ,” Computer Networks, 2015.

79.   G. Kuperman, E. Modiano, A. Narula-Tam, “ Analysis and Algorithms for Partial Protection in Mesh Networks ,” IEEE/OSA Journal of Optical Communications and Networks, 2014.

78.   Krishna Jagannathan, Mihalis Markakis, Eytan Modiano, John Tsitsiklis, " Throughput Optimal Scheduling over Time-Varying Channels in the presence of Heavy-Tailed Traffic ,” IEEE Transactions on Information Theory, 2014.

77.   Chih-Ping Li and Eytan Modiano, “ Receiver-Based Flow Control for Networks in Overload ," IEEE/ACM Transactions on Networking, Vol. 23, No. 2, 2015.

76.   Matthew Johnston, Hyang-Won Lee, Eytan Modiano, “ A Robust Optimization Approach to Backup Network Design with Random Failures ,” IEEE/ACM Transactions on Networking, Vol. 23, No. 4, 2015.

75.   Guner Celik and Eytan Modiano, “ Scheduling in Networks with Time-Varying Channels and Reconfiguration Delay ," IEEE/ACM Transactions on Networking, Vol. 23, No. 1, 2015.

74.   Matt Johnston, H.W. Lee, E. Modiano, “ Robust Network Design for Stochastic Traffic Demands ," IEEE Journal of Lightwave Technology, 2013.

73.   Mihalis Markakis, Eytan Modiano, John Tsitsiklis, “ Max-Weight Scheduling in Queueing Networks With Heavy-Tailed Traffic, ” IEEE/ACM Transactions on Networking, 2014.

72.   Kayi Lee, Hyang-Won Lee, Eytan Modiano, " Maximizing Reliability in WDM Networks through Lightpath Routing ,”  IEEE ACM Transactions on Networking, 2014.

71.   Krishna Jaggannathan and Eytan Modiano, “ The Impact of Queue Length Information on Buffer Overflow in Parallel Queues ,”  IEEE transactions on Information Theory, 2013.

70.   Krishna Jagannathan, Ishai Menashe, Gil Zussman, Eytan Modiano, “ Non-cooperative Spectrum Access - The Dedicated vs. Free Spectrum Choice ,” IEEE JSAC, special issue on Economics of Communication Networks & Systems, to appear, 2012.

69.   Guner Celik and Eytan Modiano, “ Dynamic Server Allocation over Time Varying Channels with Switchover Delay ," IEEE Transactions on Information Theory, to appear, 2012.

68.   Anand Srinivas and Eytan Modiano, " Joint Node Placement and Assignment for Throughput Optimization in Mobile Backbone Networks ,” IEEE JSAC, special issue on Communications Challenges and Dynamics for Unmanned Autonomous Vehicles, June, 2012.

67.   Guner Celik and Eytan Modiano, “ Controlled Mobility in Stochastic and Dynamic Wireless Networks ," Queueing Systems, 2012.

66.   Krishna Jagannathan, Shie Mannor, Ishai Menache, Eytan Modiano, “ A State Action Frequency Approach to Throughput Maximization over Uncertain Wireless Channels ,” Internet Mathematics, Vol. 9, Nos. 2–3: 136–160.

65.   Long Le, E. Modiano, N. Shroff, “Optimal Control of Wireless Networks with Finite Buffers ,” IEEE/ACM Transactions on Networking, to appear, 2012.

64.   K. Jagannathan, M. Markakis, E. Modiano, J. Tsitsiklis, “Queue Length Asymptotics for Generalized Max-Weight Scheduling in the presence of Heavy-Tailed Traffic,” IEEE/ACM Transactions on Networking, Vol. 20, No. 4, August 2012.

63.   Kayi Lee, Hyang-Won Lee, Eytan Modiano, “ Reliability in Layered Networks with Random Link Failures, ” IEEE/ACM Transactions on Networking, December 2011.

62.   Krishna Jagannathan, Eytan Modiano, Lizhong Zheng, “ On the Role of Queue Length Information in Network Control ,” IEEE Transactions on Information Theory, September 2011.

61.   Hyang-Won Lee, Long Le, Eytan Modiano, “ Distributed Throughput Maximization in Wireless Networks via Random Power Allocation, ” IEEE Transactions on Mobile Computing, 2011.

60.   Sebastian Neumayer, Gil Zussman, Rueven Cohen, Eytan Modiano, " Assessing the Vulnerability of the Fiber Infrastructure to Disasters, " IEEE/ACM Transactions on Networking, December 2011.

59.   Kayi Lee, Eytan Modiano, Hyang-Won Lee, “ Cross Layer Survivability in WDM-based Networks ,” IEEE/ACM Transactions on Networking, August 2011.

58.   Emily Craparo, Jon How, and Eytan Modiano, “Throughput Optimization in Mobile Backbone Networks,” IEEE Transactions on Mobile Computing, April, 2011.

57.   Hyang-Won Lee, Kayi Lee, and Eytan Modiano, “Diverse Routing in Networks with Probabilistic Failures,” IEEE/ACM Transactions on Networking, December, 2010.

56.   Guner Celik, Gil Zussman, Wajahat Khan and Eytan Modiano, “MAC Protocols For Wireless Networks With Multi-packet Reception Cabaility ,” IEEE Transactions on Mobile Computing, February, 2010.

55.   Atilla Eryilmaz, Asuman Ozdaglar, Devavrat Shah, and Eytan Modiano, “Distributed Cross-Layer Algorithms for the Optimal Control of Multi-hop Wireless Networks,” IEEE/ACM Transactions on Networking, April 2010.

54.   Murtaza Zafer and Eytan Modiano, “Minimum Energy Transmission over a Wireless Channel With Deadline and Power Constraints ,” IEEE Transactions on Automatic Control, pp. 2841-2852, December, 2009.

53.   Murtaza Zafer and Eytan Modiano, “A Calculus Approach to Energy-Efficient Data Transmission with Quality of Service Constraints,” IEEE/ACM Transactions on Networking, 2009.

52.   Anand Srinivas, Gil Zussman, and Eytan Modiano, “Construction and Maintenance of Wireless Mobile Backbone Networks,” IEEE/ACM Transactions on Networking, 2009.

51.   Andrew Brzezinski, Gil Zussman, and Eytan Modiano, “Distributed Throughput Maximization in Wireless Mesh Networks Via Pre-Partitioning,” IEEE/ACM Transactions on Networking, December, 2008.

50.   Amir Khandani, Eytan Modiano, Jinane Abounadi, Lizhong Zheng, “Reliability and Route Diversity in Wireless Networks,” IEEE Transactions on Wireless Communications, December, 2008.

49.   Alessandro Tarello, Jun Sun, Murtaza Zafer and Eytan Modiano, “Minimum Energy Transmission Scheduling Subject to Deadline Constraints,” ACM Wireless Networks, October, 2008.

48.   Murtaza Zafer, Eytan Modiano, “Optimal Rate Control for Delay-Constrained Data Transmission over a Wireless Channel,” IEEE Transactions on Information Theory, September, 2008.

47.   Andrew Brzezinski and Eytan Modiano, “Achieving 100% Throughput In Reconfigurable IP/WDM Networks,” IEEE/ACM Transactions on Networking, August, 2008.

46.   Michael Neely, Eytan Modiano and C. Li, “Fairness and Optimal Stochastic Control for Heterogeneous Networks,” IEEE/ACM Transactions on Networking, September, 2008.

45.   Amir Khandani, Jinane Abounadi, Eytan Modiano, Lizhong Zheng, “Cooperative Routing in Static Wireless Networks,” IEEE Transactions on Communications, November 2007.

44.   Murtaza Zafer, Eytan Modiano, “Joint Scheduling of Rate-guaranteed and Best-effort Users over a Wireless Fading Channel,” IEEE Transactions on Wireless Communications, October, 2007.

43.   Krishna Jagannathan, Sem Borst, Phil Whiting and Eytan Modiano, “Scheduling of Multi-Antenna Broadcast Systems with Heterogeneous Users,” IEEE Journal of Selected Areas in Communications, September, 2007.Amir Khandani, Jinane

42.   Anand Ganti, Eytan Modiano, and John Tsitsiklis, “Optimal Transmission Scheduling in Symmetric Communication Models with Intermittent Connectivity, ” IEEE Transactions on Information Theory, March, 2007.

41.   Michael Neely and Eytan Modiano, “Logarithmic Delay for NxN Packet Switches Under Crossbar Constraints,” IEEE/ACM Transactions on Networking, November, 2007.

40.   Jun Sun, Jay Gao, Shervin Shambayati and Eytan Modiano, “Ka-Band Link Optimization with Rate Adaptation for Mars and Lunar Communications,”   International Journal of Satellite Communications and Networks, March, 2007.

39.   Jun Sun and Eytan Modiano, "Fair Allocation of A Wireless Fading Channel: An Auction Approach" Institute for Mathematics and its Applications, Volume 143: Wireless Communications, 2006.

38.   Jun Sun, Eytan Modiano and Lizhong Zhang, “Wireless Channel Allocation Using An Auction Algorithm,” IEEE Journal on Selected Areas in Communications, May, 2006.

37.   Murtaza Zafer and Eytan Modiano, "Blocking Probability and Channel Assignment for Connection Oriented Traffic in Wireless Networks," IEEE Transactions on Wireless Communications, April, 2006.

36.   Alvin Fu, Eytan Modiano, and John Tsitsiklis, "Optimal Transmission Scheduling over a fading channel with Energy and Deadline Constraints" IEEE Transactions on Wireless Communications, March,2006.

35.   Poompat Saengudomlert, Eytan Modiano and Rober Gallager, “On-line Routing and Wavelength Assignment for Dynamic Traffic in WDM Ring and Torus Networks,” IEEE Transactions on Networking, April, 2006.

34.   Li-Wei Chen, Eytan Modiano and Poompat Saengudomlert, "Uniform vs. Non-Uniform band Switching in WDM Networks," Computer Networks (special issue on optical networks), January, 2006.

33.   Andrew Brzezinski and Eytan Modiano, "Dynamic Reconfiguration and Routing Algorithms for IP-over-WDM networks with Stochastic Traffic," IEEE Journal of Lightwave Technology, November, 2005

32.   Randall Berry and Eytan Modiano, "Optimal Transceiver Scheduling in WDM/TDM Networks," IEEE Journal on Selected Areas in Communications, August, 2005.

31.   Poompat Saengudomlert, Eytan Modiano, and Robert G. Gallager, “Dynamic Wavelength Assignment for WDM All-Optical Tree Networks,” IEEE Transactions on Networking, August, 2005.

30.   Ashwinder Ahluwalia and Eytan Modiano, "On the Complexity and Distributed Construction of Energy Efficient Broadcast Trees in Wireless Ad Hoc Networks," IEEE Transactions on Wireless Communications, October, 2005.

29.   Michael Neely, Charlie Rohrs and Eytan Modiano, "Equivalent Models for Analysis of Deterministic Service Time Tree Networks," IEEE Transactions on Information Theory, October, 2005.

28.   Michael Neely and Eytan Modiano, "Capacity and Delay Tradeoffs for Ad Hoc Mobile Networks," IEEE Transactions on Information Theory, May, 2005.

27.   Li-Wei Chen and Eytan Modiano, "Efficient Routing and Wavelength Assignment for Reconfigurable WDM Networks with Wavelength Converters," IEEE/ACM Transactions on Networking, February, 2005. Selected as one of the best papers from Infocom 2003 for fast-track publication in IEEE/ACM Transactions on Networking.

26.   Michael Neely and Eytan Modiano, "Convexity in Queues with General Inputs," IEEE Transactions on Information Theory, May, 2005.

25.   Anand Srinivas and Eytan Modiano, "Finding Minimum Energy Disjoint Paths in Wireless Ad Hoc Networks," ACM Wireless Networks, November, 2005. Selected to appear in a special issue dedicated to best papers from Mobicom 2003.

24.   Michael Neely, Eytan Modiano and Charlie Rohrs, "Dynamic Power Allocation and Routing for Time-Varying Wireless Networks," IEEE Journal of Selected Areas in Communication, January, 2005.

23.   Chunmei Liu and Eytan Modiano, "On the performance of additive increase multiplicative decrease (AIMD) protocols in hybrid space-terrestrial networks," Computer Networks, September, 2004.

22.   Li-Wei Chen and Eytan Modiano, "Dynamic Routing and Wavelength Assignment with Optical Bypass using Ring Embeddings," Optical Switching and Networking (Elsevier), December, 2004.

21.   Aradhana Narula-Tam, Eytan Modiano and Andrew Brzezinski, "Physical Topology Design for Survivable Routing of Logical Rings in WDM-Based Networks," IEEE Journal of Selected Areas in Communication, October, 2004.

20.   Randall Berry and Eytan Modiano, "'The Role of Switching in Reducing the Number of Electronic Ports in WDM Networks," IEEE Journal of Selected Areas in Communication, October, 2004.

19.   Jun Sun and Eytan Modiano, "Routing Strategies for Maximizing Throughput in LEO Satellite Networks,," IEEE JSAC, February, 2004.

18.   Jun Sun and Eytan Modiano, "Capacity Provisioning and Failure Recovery for Low Earth Orbit Satellite Networks," International Journal on Satellite Communications, June, 2003.

17.   Alvin Fu, Eytan Modiano, and John Tsitsiklis, "Optimal Energy Allocation and Admission Control for Communications Satellites," IEEE Transactions on Networking, June, 2003.

16.   Michael Neely, Eytan Modiano and Charles Rohrs, "Power Allocation and Routing in Multi-Beam Satellites with Time Varying Channels," IEEE Transactions on Networking, February, 2003.

15.   Eytan Modiano and Aradhana Narula-Tam, "Survivable lightpath routing: a new approach to the design of WDM-based networks," IEEE Journal of Selected Areas in Communication, May 2002.

14.   Aradhana Narula-Tam, Phil Lin and Eytan Modiano, "Efficient Routing and Wavelength Assignment for Reconfigurable WDM Networks," IEEE Journal of Selected Areas in Communication, January, 2002.

13.   Brett Schein and Eytan Modiano, "Quantifying the benefits of configurability in circuit-switched WDM ring networks with limited ports per node," IEEE Journal on Lightwave Technology, June, 2001.

12.   Aradhana Narula-Tam and Eytan Modiano, "Dynamic Load Balancing in WDM Packet Networks with and without Wavelength Constraints," IEEE Journal of Selected Areas in Communications, October 2000.

11.   Randy Berry and Eytan Modiano, "Reducing Electronic Multiplexing Costs in SONET/WDM Rings with Dynamically Changing Traffic," IEEE Journal of Selected Areas in Communications, October 2000.

10.   Eytan Modiano and Richard Barry, "A Novel Medium Access Control Protocol for WDM-Based LANs and Access Networks Using a Master-Slave Scheduler," IEEE Journal on Lightwave Technology, April 2000.

9.   Eytan Modiano and Anthony Ephremides, "Communication Protocols for Secure Distributed Computation of Binary Functions," Information and Computation, April 2000.

8.   Angela Chiu and Eytan Modiano, "Traffic Grooming Algorithms for Reducing Electronic Multiplexing Costs in WDM Ring Networks," IEEE Journal on Lightwave Technology, January 2000.

7.   Eytan Modiano, "An Adaptive Algorithm for Optimizing the Packet Size Used in Wireless ARQ Protocols," Wireless Networks, August 1999.

6.   Eytan Modiano, "Random Algorithms for Scheduling Multicast Traffic in WDM Broadcast-and-Select Networks," IEEE Transactions on Networking, July, 1999.

5.   Eytan Modiano and Richard Barry, "Architectural Considerations in the Design of WDM-based Optical Access Networks," Computer Networks, February 1999.

4.   V.W.S. Chan, K. Hall, E. Modiano and K. Rauschenbach, "Architectures and Technologies for High-Speed Optical Data Networks," IEEE Journal of Lightwave Technology, December 1998.

3.   Eytan Modiano and Anthony Ephremides, "Efficient Algorithms for Performing Packet Broadcasts in a Mesh Network," IEEE Transactions on Networking, May 1996.

2.   Eytan Modiano, Jeffrey Wieselthier and Anthony Ephremides, "A Simple Analysis of Queueing Delay in a Tree Network of Discrete-Time Queues with Constant Service Times," IEEE Transactions on Information Theory, February 1996.

1.   Eytan Modiano and Anthony Ephremides, "Communication Complexity of Secure Distributed Computation in the Presence of Noise," IEEE Transactions on Information Theory, July 1992.

Other Papers

5.  Eytan Modiano, "Satellite Data Networks," AIAA Journal on Aerospace Computing, Information and Communication, September, 2004.

4.  Eytan Modiano and Phil Lin, "Traffic Grooming in WDM networks," IEEE Communications Magazine, July, 2001.

3.  Eytan Modiano and Aradhana Narula, "Mechanisms for Providing Optical Bypass in WDM-based Networks," SPIE Optical Networks, January 2000.

2.  K. Kuznetsov, N. M. Froberg, Eytan Modiano, et. al., "A Next Generation Optical Regional Access Networks," IEEE Communications Magazine, January, 2000.

1.  Eytan Modiano, "WDM-based Packet Networks," (Invited Paper) IEEE Communications Magazine, March 1999.

Conference Papers

246. Xinyu Wu, Dan Wu, Eytan Modiano, “ Overload Balancing in Single-Hop Networks With Bounded Buffers ,” IFIP Networking, 2022.

245.  Xinzhe Fu, Eytan Modiano, “ Optimal Routing for Stream Learning Systems ,”  IEEE Infocom, April 2022.

244.  Vishrant Tripathi, Luca Ballotta, Luca Carlone, E. Modiano, “ Computation and Communication Co-Design for Real-Time Monitoring and Control in Multi-Agent Systems ,”  IEEE Wiopt, 2021.

243. Eray Atay, Igor Kadota, E. Modiano, “ Aging Wireless Bandits: Regret Analysis and Order-Optimal Learning Algorithm ,”  IEEE Wiopt 2021.

242. Xinzhe Fu and E. Modiano,  “ Elastic Job Scheduling with Unknown Utility Functions ,” IFIP Performance, Milan, 2021.

241. Bai Liu and E. Modiano, “ Optimal Control for Networks with Unobservable Malicious Nodes ,”  IFIP Performance, Milan, 2021.

240. Bai Liu, Qiaomin Xie,  Eytan Modiano, “ RL-QN:  A Reinforcement Learning Framework for Optimal Control of Queueing Systems ,”  ACM Sigmetrics Workshop on Reinforcement Learning in Networks and Queues (RLNQ), 2021.

239. Xinzhe Fu and E. Modiano,  “ Learning-NUM: Network Utility Maximization with Unknown Utility Functions and Queueing Delay ,  ACM MobiHoc 2021.  

238. Vishrant Tripathi and Eytan Modiano,  “ An Online Learning Approach to Optimizing Time-Varying Costs of AoI ,”  ACM MobiHoc 2021. 

237.   Igor Kadota, Muhammad Shahir Rahman, and Eytan Modiano, " WiFresh: Age-of-Information from Theory to Implementation ,”  International Conference on Computer Communications and Networks (ICCCN), 2021.

236. Vishrant Tripathi and Eytan Modiano, “ Age Debt: A General Framework For Minimizing Age of Information ,”  IEEE Infocom Workshop on Age-of-Information, 2021.

235. Igor Kadota, Eytan Modiano, “ Age of Information in Random Access Networks with Stochastic Arrivals ,” IEEE Infocom, 2020.

234. Igor Kadota, M. Shahir Rahman, Eytan Modiano, Poster: Age of Information in Wireless Networks: from Theory to Implementation , ACM Mobicom, 2020.

233. Xinyu Wu, Dan Wu, Eytan Modiano, “ An Influence Model Approach to Failure Cascade Prediction in Large Scale Power Systems ,” IEEE American Control Conference, July, 2020.

232. X. Fu and E. Modiano, " Fundamental Limits of Volume-based Network DoS Attacks ," Proc. ACM Sigmetrics, Boston, MA, June 2020.

231. Vishrant Tripathi, Eytan Modiano, “ A Whittle Index Approach to Minimizing Functions of Age of Information ,” Allerton Conference on Communication, Control, and Computing, September 2019.

230. Bai Liu, Xiaomin Xie, Eytan Modiano, “ Reinforcement Learning for Optimal Control of Queueing Systems ,” Allerton Conference on Communication, Control, and Computing, September 2019.

229. Rajat Talak, Sertac Karaman, Eytan Modiano, “ A Theory of Uncertainty Variables for State Estimation and Inference ,” Allerton Conference on Communication, Control, and Computing, September 2019.

228. Rajat Talak, Eytan Modiano, “ Age-Delay Tradeoffs in Single Server Systems ,” IEEE International Symposium on Information Theory, Paris, France, July, 2019.

227. Rajat Talak, Sertac Karaman, Eytan Modiano, “ When a Heavy Tailed Service Minimizes Age of Information ,” IEEE International Symposium on Information Theory, Paris, France, July, 2019.

226. Qingkai Liang, Eytan Modiano, “ Optimal Network Control with Adversarial Uncontrollable Nodes ,” ACM MobiHoc, Catania, Italy, June 2019.

225. Igor Kadota, Eytan Modiano, “ Minimizing the Age of Information in Wireless Networks with Stochastic Arrivals ,” ACM MobiHoc, June 2019.

224. Maotong Xu, Jelena Diakonikolas, Suresh Subramaniam, Eytan Modiano, “ A Hierarchical WDM-based Scalable Data Center Network Architecture ,” IEEE International Conference on Communications (ICC), Shanghai, China, June 2019.

223. Maotong Xu, Min Tian, Eytan Modiano, Suresh Subramaniam, " RHODA Topology Configuration Using Bayesian Optimization

222.   Anurag Rai, Rahul Singh and Eytan Modiano, " A Distributed Algorithm for Throughput Optimal Routing in Overlay Networks ,”  IFIP Networking 2019, Warsaw, Poland, May 2019.

221.   Qingkai Liang and Eytan Modiano, " Optimal Network Control in Partially-Controllable Networks ,”  IEEE Infocom, Paris, April 2019.

220.   Xinzhe Fu and Eytan Modiano, " Network Interdiction Using Adversarial Traffic Flows ,”  IEEE Infocom, Paris, April 2019.

219.   Vishrant Tripathi, Rajat Talak, Eytan Modiano, " Age Optimal Information Gathering and Dissemination on Graphs ,”  IEEE Infocom, Paris, April 2019.

218.   Jianan Zhang, Hyang-Won Lee, Eytan Modiano, " On the Robustness of Distributed Computing Networks ,”  DRCN 2019, Coimbra, Portugal, March, 2019.

217.   Hyang-Won Lee, Jianan Zhang and Eytan Modiano, " Data-driven Localization and Estimation of Disturbance in the Interconnected Power System ,”  IEEE Smartgridcomm, October, 2018.

216.   Jianan Zhang and Eytan Modiano, " Joint Frequency Regulation and Economic Dispatch Using Limited Communication ,”  IEEE Smartgridcomm, October, 2018.

215.   Rajat Talak, Sertac Karaman, Eytan Modiano, " Scheduling Policies for Age Minimization in Wireless Networks with Unknown Channel State ,”  IEEE International Symposium on Information Theory, July 2018.

214.   Thomas Stahlbuhk, Brooke Shrader, Eytan Modiano, " Online Learning Algorithms for Minimizing Queue Length Regret ,”  IEEE International Symposium on Information Theory, July 2018.

213.   Rajat Talak, Sertac Karaman, Eytan Modiano, " Distributed Scheduling Algorithms for Optimizing Information Freshness in Wireless Networks ,”  IEEE SPAWC, Kalamata, Greece, June, 2018.

212.   Rajat Talak, Sertac Karaman, Eytan Modiano, " Optimizing Information Freshness in Wireless Networks under General Interference Constraints ,”  ACM MobiHoc 2018, Los Angeles, CA, June 2018.

211.   Thomas Stahlbuhk, Brooke Shrader, Eytan Modiano, " Learning Algorithms for Scheduling in Wireless Networks with Unknown Channel Statistics ,”  ACM MobiHoc, June 2018.

210.   Khashayar Kamran, Jianan Zhang, Edmund Yeh, Eytan Modiano, " Robustness of Interdependent Geometric Networks Under Inhomogeneous Failures ,”  Workshop on Spatial Stochastic Models for Wireless Networks (SpaSWiN), Shanghai, China, May 2018.

209.   Rajat Talak, Sertac Karaman, Eytan Modiano, " Optimizing Age of Information in Wireless Networks with Perfect Channel State Information ,”  Wiopt 2018, Shanghai, China, May 2018.

208.   Abhishek Sinha, Eytan Modiano, " Network Utility Maximization with Heterogeneous Traffic Flows ,”  Wiopt 2018, Shanghai, China, May 2018.

207.   Qingkai Liang, Eytan Modiano, " Minimizing Queue Length Regret Under Adversarial Network Models ,”  ACM Sigmetrics, 2018.

206.   Jianan Zhang, Abhishek Sinha, Jaime Llorca, Anonia Tulino, Eytan Modiano, " Optimal Control of Distributed Computing Networks with Mixed-Cast Traffic Flows ,”  IEEE Infocom, Honolulu, HI, April 2018.

205.   Qingkai Liang, Eytan Modiano, " Network Utility Maximization in Adversarial Environments ,”  IEEE Infocom, Honolulu, HI, April 2018.

204.   Igor Kadota, Abhishek Sinha, Eytan Modiano, " Optimizing Age of Information in Wireless Networks with Throughput Constraints ,”  IEEE Infocom, Honolulu, HI, April 2018.

203.   QIngkai Liang, Verina (Fanyu) Que, Eytan Modiano, " Accelerated Primal-Dual Policy Optimization for Safe Reinforcement Learning ,”  NIPS workshop on “Transparent and interpretable machine learning in safety critical environments,"December 2017.

202.   Rahul Singh, Xueying Guo,Eytan Modiano, " Risk-Sensitive Optimal Control of Queues ,”  IEEE Conference on Decision and Control (CDC), December 2017.

201.   Rajat Talak, Sertac Karaman, Eytan Modiano, " Minimizing Age of Information in Multi-Hop Wireless Networks ,”  Allerton Conference on Communication, Control, and Computing, September 2017.

200.   Abhishek Sinha, Eytan Modiano, " Throughput-Optimal Broadcast in Wireless Networks with Point-to-Multipoint Transmissions ,”  ACM MobiHoc, Madras, India, July 2017.

199.   Rajat Talak, Sertac Karaman, Eytan Modiano, " Capacity and delay scaling for broadcast transmission in highly mobile wireless networks ,”  ACM MobiHoc, Madras, India, July 2017.

198.5 . Y.-P. Hsu, E. Modiano, and L. Duan, " Age of Information: Design and Analysis of Optimal Scheduling Algorithms ,”  IEEE International Symposium on Information Theory (ISIT), 2017.

198.   Qingkai Liang and Eytan Modiano, " Coflow Scheduling in Input-Queued Switches: Optimal Delay Scaling and Algorithms ,”  IEEE Infocom, Atlanta, GA, May 2017.

197.   Jianan Zhang and Eytan Modiano, " Robust Routing in Interdependent Networks ,”  IEEE Infocom, Atlanta, GA, May 2017.

196.   Abhishek Sinha, Eytan Modiano, " Optimal Control for Generalized Network Flow Problems ,”  IEEE Infocom, Atlanta, GA, May 2017.

195.   Rajat Talak*, Sertac Karaman, Eytan Modiano, " Speed Limits in Autonomous Vehicular Networks due to Communication Constraints ,”  IEEE Conference on Decision and Control (CDC), Las Vegas, NV, December 2016.

194.   Marzieh Parandehgheibi*, Konstantin Turitsyn, Eytan Modiano, " Distributed Frequency Control in Power Grids Under Limited Communication ,”  IEEE Conference on Decision and Control (CDC), Las Vegas, NV, December 2016.

193.   Igor Kadota, Elif Uysal-Biyikoglu, Rahul Singh, Eytan Modiano, " Minimizing Age of Information in Broadcast Wireless Networks ,”  Allerton Allerton Conference on Communication, Control, and Computing, September 2016.

192.   Jianan Zhang, Edmund Yeh, Eytan Modiano, " Robustness of Interdependent Random Geometric Networks ,”  Allerton Conference on Communication, Control, and Computing, September 2016.

191.   Abhishek Sinha, Leandros Tassiulas, Eytan Modiano, " Throughput-Optimal Broadcast in Wireless Networks with Dynamic Topology ,”  ACM MobiHoc'16, Paderborn, Germany, July, 2016. (winner of best paper award)

190.   Abishek Sinha, Georgios Paschos, Eytan Modiano, " Throughput-Optimal Multi-hop Broadcast Algorithms ,”  ACM MobiHoc'16, Paderborn, Germany, July, 2016.

189.   Thomas Stahlbuhk, Brooke Shrader, Eytan Modiano, " Throughput Maximization in Uncooperative Spectrum Sharing Networks ,”  IEEE International Symposium on Information Theory, Barcelona, Spain, July 2016.

188.   Thomas Stahlbuhk, Brooke Shrader, Eytan Modiano, " Topology Control for Wireless Networks with Highly-Directional Antennas ,”  IEEE Wiopt, Tempe, Arizona, May, 2016.

187.   Qingkai Liang, H.W. Lee, Eytan Modiano, " Robust Design of Spectrum-Sharing Networks ,”  IEEE Wiopt, Tempe, Arizona, May, 2016.

186.   Hossein Shokri-Ghadikolae, Carlo Fischione and Eytan Modiano, " On the Accuracy of Interference Models in Wireless Communications ,”  IEEE International Conference on Communications (ICC), 2016.

185.   Qingkai Liang and Eytan Modiano, " Survivability in Time-varying Networks ,”  IEEE Infocom, San Francisco, CA, April 2016.

184.   Kyu S. Kim, Chih-Ping Li, Igor Kadota, Eytan Modiano, " Optimal Scheduling of Real-Time Traffic in Wireless Networks with Delayed Feedback ,”  Allerton conference on Communication, Control, and Computing, September 2015.

183.   Marzieh Parandehgheibi, Eytan Modiano, " Modeling the Impact of Communication Loss on the Power Grid Under Emergency Control ,”  IEEE SmartGridComm, Miami, FL, Nov. 2015.

182.   Anurag Rai, Chih-ping Li, Georgios Paschos, Eytan Modiano, " Loop-Free Backpressure Routing Using Link-Reversal Algorithms ,”  Proceedings of the ACM MobiHoc, July 2015.

181.   Longbo Huang, Eytan Modiano, " Optimizing Age of Information in a Multiclass Queueing System ,”  Proceedings of IEEE ISIT 2015, Hong Kong, Jun 2015.

180.   M. Johnston, E. Modiano, " A New Look at Wireless Scheduling with Delayed Information ,”  Proceedings of IEEE ISIT 2015, Hong Kong, June 2015.

179.   M. Johnston, E. Modiano, " Scheduling over Time Varying Channels with Hidden State Information ,”  Proceedings of IEEE ISIT 2015, Hong Kong, June 2015.

178.   M. Johnston and E. Modiano, " Controller Placement for Maximum Throughput Under Delayed CSI ,”  IEEE Wiopt, Mombai, India, May 2015.

177.   A. Sinha, G. Paschos, C. P. Li, and E. Modiano, " Throughput Optimal Broadcast on Directed Acyclic Graphs ,”  IEEE Infocom, Hong Kong, April 2015.

176.   J. Zheng and E. Modiano, " Enhancing Network Robustness via Shielding ,”  IEEE Design of Reliable Communication Networks, Kansas City, March 2015.

175.   H. W. Lee and E. Modiano, " Robust Design of Cognitive Radio Networks ,”  Information and Communication Technology Convergence (ICTC), 2014.

174.   Greg Kuperman and Eytan Modiano, " Disjoint Path Protection in Multi-Hop Wireless Networks with Interference Constraints ,”  IEEE Globecom, Austin, TX, December 2014.

173.   Marzieh Parandehgheibi, Eytan Modiano, David Hay, " Mitigating Cascading Failures in Interdependent Power Grids and Communication Networks ,”  IEEE Smartgridcomm, Venice, Italy, November 2014.

172.   Georgios Paschos and Eytan Modiano, " Throughput optimal routing in overlay networks ,”  Allerton conference on Communication, Control, and Computing, September 2014.

171.   Nathan Jones, George Paschos, Brooke Shrader, Eytan Modiano, " An overlay architecture for Throughput Optimal Multipath Routing ,”  ACM MobiHoc, August 2014.

170.   Matt Johnston, Eytan Modiano, Yuri Polyanskiy, " Opportunistic Scheduling with Limited Channel State Information: A Rate Distortion Approach ,”  IEEE International Symposium on Information Theory, Honolulu, HI, July 2014.

169.   Chih-Ping Li, Georgios Paschos, Eytan Modiano, Leandros Tassiulas, " Dynamic Overload Balancing in Server Farms ,”  Networking 2014, Trondheim, Norway, June, 2014.

168.   Hulya Seferonglu and Eytan Modiano, " TCP-Aware Backpressure Routing and Scheduling ,”  Information Theory and Applications, San Diego, CA, February 2014.

167.   Mihalis Markakis, Eytan Modiano, John Tsitsiklis, " Delay Stability of Back-Pressure Policies in the presence of Heavy-Tailed Traffic ,”  Information Theory and Applications, San Diego, CA, February 2014.

166.   Kyu Soeb Kim, Chih-ping Li, Eytan Modiano, " Scheduling Multicast Traffic with Deadlines in Wireless Networks ,”  IEEE Infocom, Toronto, CA, April 2014.

165.   Georgios Paschos, Chih-ping Li, Eytan Modiano, Kostas Choumas, Thanasis Korakis, " A Demonstration of Multirate Multicast Over an 802.11 Mesh Network ,”  IEEE Infocom, Toronto, CA, April 2014.

164.   Sebastian Neumayer, Eytan Modiano, " Assessing the Effect of Geographically Correlated Failures on Interconnected Power-Communication Networks ,”  IEEE SmartGridComm, 2013.

163.   Marzieh Parandehgheibi, Eytan Modiano, " Robustness of Interdependent Networks: The case of communication networks and the power grid ,”  IEEE Globecom, December 2013.

162.   Matt Johnston, Eytan Modiano, " Optimal Channel Probing in Communication Systems: The Two-Channel Case ,”  IEEE Globecom, December 2013.

161.   Mihalis Markakis, Eytan Modiano, John N. Tsitsiklis, " Delay Analysis of the Max-Weight Policy under Heavy-Tailed Traffic via Fluid Approximations ,”  Allerton Conference, October 2013.

160.   Matthew Johnston, Isaac Keslassy, Eytan Modiano, " Channel Probing in Communication Systems: Myopic Policies Are Not Always Optimal ,”  IEEE International Symposium on Information Theory, July 2013.

159.   Krishna P Jagannathan, Libin Jiang, Palthya Lakshma Naik, Eytan Modiano, " Scheduling Strategies to Mitigate the Impact of Bursty Traffic in Wireless Networks ,”  11th International Symposium on Modeling and Optimization in Mobile, Ad Hoc and Wireless Networks Wiopt 2013, Japan, May 2013. (Winner – Best Paper Award).

158.   Hulya Seferoglu and Eytan Modiano, " Diff-Max: Separation of Routing and Scheduling in Backpressure-Based Wireless Networks ,”  IEEE Infocom, Turin, Italy, April 2013.

157.   Chih-Ping Li, Eytan Modiano, " Receiver-Based Flow Control for Networks in Overload ,”  IEEE Infocom, Turin, Italy, April 2013.

156.   Nathan Jones, Brooke Shrader, Eytan Modiano, " Distributed CSMA with Pairwise Coding ,”  IEEE Infocom, Turin, Italy, April 2013.

155.   Greg Kuperman and Eytan Modiano, " Network Protection with Guaranteed Recovery Times using Recovery Domains ,”  IEEE Infocom, Turin, Italy, April 2013.

154.   Greg Kuperman and Eytan Modiano, " Providing Protection in Multi-Hop Wireless Networks ,”  IEEE Infocom, Turin, Italy, April 2013.

153.   Greg Kuperman, Eytan Modiano, Aradhana Narula-Tam, " Network Protection with Multiple Availability Guarantees ,”  IEEE ICC workshop on New Trends in Optical Networks Survivability, June 2012.

152.   Nathaniel Jones, Brooke Shrader, Eytan Modiano, " Optimal Routing and Scheduling for a Simple Network Coding Scheme ,”  IEEE Infocom, Orlando, Fl, March, 2012.

151.   Mihalis Markakis, Eytan Modiano, John Tsitsiklis, " Max-Weight Scheduling in Networks with Heavy-Tailed Traffic ,”  IEEE Infocom, Orlando, Fl, March, 2012.

150.   Guner Celik and Eytan Modiano, " Scheduling in Networks with Time-Varying Channels and Reconfiguration Delay ,”  IEEE Infocom, Orlando, Fl, March, 2012.

149.   Sebastian Neumayer, Alon Efrat, Eytan Modiano, " Geographic Max-Flow and Min-cut Under a Circular Disk Failure Model ,”  IEEE Infocom (MC), Orlando, Fl, March, 2012.

148.   Marzieh Parandehgheibi, Hyang-Won Lee, and Eytan Modiano, " Survivable Paths in Multi-Layer Networks ,”  Conference on Information Science and Systems, March, 2012.

147.   Greg Kuperman, Eytan Modiano, and Aradhana Narula-Tam, " Partial Protection in Networks with Backup Capacity Sharing ,”  Optical Fiber Communications Conference (OFC), Anaheim, CA, March, 2012.

146.   Krishna Jagannathan, Libin Jiang, Eytan Modiano, " On Scheduling Algorithms Robust to Heavy-Tailed Traffic ,”  Information Theory and Applications (ITA), San Diego, CA, February 2012.

145.   M. Johnston, H.W. Lee, E. Modiano, " Robust Network Design for Stochastic Traffic Demands ,”  IEEE Globecom, Next Generation Networking Symposium, Houston, TX, December 2011.

144.   S. Neumayer, E. Modiano, " Network Reliability Under Random Circular Cuts ,”  IEEE Globecom, Optical Networks and Systems Symposium, Houston, TX, December 2011.

143.   H.W. Lee, K. Lee, E. Modiano, " Maximizing Reliability in WDM Networks through Lightpath Routing ,”  IEEE Globecom, Optical Networks and Systems Symposium, Houston, TX, December 2011.

142.   Guner Celik, Sem Borst, Eytan Modiano, Phil Whiting, " Variable Frame Based Max-Weight Algorithms for Networks with Switchover Delay ,”  IEEE International Symposium on Information Theory, St. Petersburgh, Russia, August 2011.

141.   Krishna Jaganathan, Ishai Menache, Eytan Modiano, and Gil Zussman, " Non-cooperative Spectrum Access - The Dedicated vs. Free Spectrum Choice ,”  ACM MOBIHOC'11, May 2011.

140.   Krishna Jagannathan, Shie Mannor, Ishai Menache, Eytan Modiano, " A State Action Frequency Approach to Throughput Maximization over Uncertain Wireless Channels ,”  IEEE Infocom (Mini-conference), Shanghai, China, April 2011.

139.   Guner Celik, Long B. Le, Eytan Modiano, " Scheduling in Parallel Queues with Randomly Varying Connectivity and Switchover Delay ,”  IEEE Infocom (Mini-conference), Shanghai, China, April 2011.

138.   Gregory Kuperman, Eytan Modiano, Aradhana Narula-Tam, " Analysis and Algorithms for Partial Protection in Mesh Networks ,”  IEEE Infocom (Mini-conference), Shanghai, China, April 2011.

137.   Matthew Johnston, Hyang-Won Lee, Eytan Modiano, " A Robust Optimization Approach to Backup Network Design with Random Failures ,”  IEEE Infocom, Shanghai, China, April 2011.

136.   Krishna Jagannathan, Mihalis Markakis, Eytan Modiano, John Tsitsiklis, " Queue Length Asymptotics for Generalized Max-Weight Scheduling in the presence of Heavy-Tailed Traffic ,”  IEEE Infocom, Shanghai, China, April 2011.

135.   Guner Celik and Eytan Modiano, " Dynamic Vehicle Routing for Data Gathering in Wireless Networks ,”  In Proc. IEEE CDC'10, Dec. 2010..***

134.   Long B. Le, Eytan Modiano, Changhee Joo, and Ness B. Shroff, " Longest-queue-first scheduling under the SINR interference model ,”  ACM MobiHoc, September 2010..***

133.   Krishna Jagannathan, Mihalis Markakis, Eytan Modiano, John Tsitsiklis, " Throughput Optimal Scheduling in the Presence of Heavy-Tailed Traffic ,”  Allerton Conference on Communication, Control, and Computing, September 2010..**

132.   Delia Ciullo, Guner Celik, Eytan Modiano, " Minimizing Transmission Energy in Sensor Networks via Trajectory Control ,”  IEEE Wiopt 2010, Avignon, France, June 2010, (10 pages; CD proceedings – page numbers not available).

131.   Sebastian Neumayer and Eytan Modiano, " Network Reliability with Geographically Correlated Failures ,”  IEEE Infocom 2010, San Diego, CA, March 2010, (9 pages; CD proceedings – page numbers not available).**

130.   Long Le, Eytan Modiano, Ness Shroff, " Optimal Control of Wireless Networks with Finite Buffers ,”  IEEE Infocom 2010, San Diego, CA, March 2010, (9 pages; CD proceedings – page numbers not available).

129.   Kayi Lee, Hyang-Won Lee, Eytan Modiano, " Reliability in Layered Network with Random Link Failures ,”  IEEE Infocom 2010, San Diego, CA, March 2010, (9 pages; CD proceedings – page numbers not available).**

128.   Krishna Jagannathan, Eytan Modiano, " The Impact of Queue length Information on Buffer Overflow in Parallel Queues ,”  Allerton Conference on Communication, Control, and Computing, September 2009, pgs. 1103 -1110 **

127.   Mihalis Markakis, Eytan Modiano, John Tsitsiklis, " Scheduling Policies for Single-Hop with Heavy-Tailed Traffic ,”  Allerton Conference on Communication, Control, and Computing, September 2009, pgs. 112 – 120..**

126.   Dan Kan, Aradhana Narula-Tam, Eytan Modiano, " Lightpath Routing and Capacity Assignment for Survivable IP-over-WDM Networks ,”  DRCN 2009, Alexandria, VA October 2009, pgs. 37 -44..**

125.   Mehdi Ansari, Alireza Bayesteh, Eytan Modiano, " Opportunistic Scheduling in Large Scale Wireless Networks ,”  IEEE International Symposium on Information Theory, Seoul, Korea, June 2009, pgs. 1624 – 1628.

124.   Hyang-Won Lee, Eytan Modiano and Long Bao Le, " Distributed Throughput Maximization in Wireless Networks via Random Power Allocation ,”  IEEE Wiopt, Seoul, Korea, June 2009. (9 pages; CD proceedings – page numbers not available).

123.   Wajahat Khan, Eytan Modiano, Long Le, " Autonomous Routing Algorithms for Networks with Wide-Spread Failures ,”  IEEE MILCOM, Boston, MA, October 2009. (6 pages; CD proceedings – page numbers not available).**

122.   Guner Celik and Eytan Modiano, " Random Access Wireless Networks with Controlled Mobility ,”  IEEE Med-Hoc-Nets, Haifa, Israel, June 2009, pgs. 8 – 14.**

121.   Hyang-Won Lee and Eytan Modiano, " Diverse Routing in Networks with Probabilistic Failures ,”  IEEE Infocom, April 2009, pgs. 1035 – 1043.

120.   Kayi Lee and Eytan Modiano, " Cross-layer Survivability in WDM-based Networks ,”  IEEE Infocom, April 2009, pgs. 1017 -1025..**

119.   Krishna Jagannathan, Eytan Modiano, Lizhong Zheng, " On the Trade-off between Control Rate and Congestion in Single Server Systems ,”  IEEE Infocom, April 2009, pgs. 271 – 279.**

118.   Sebastian Neumayer, Gil Zussman, Rueven Cohen, Eytan Modiano, " Assessing the Vulnerability of the Fiber Infrastructure to Disasters ,”  IEEE Infocom, April 2009, pgs. 1566 – 1574.**

117.   Long Le, Krishna Jagannathan and Eytan Modiano, " Delay analysis of max-weight scheduling in wireless ad hoc networks ,”  Conference on Information Science and Systems, Baltimore, MD, March, 2009, pgs. 389 – 394.**

116.   Krishna Jagannathan, Eytan Modiano, Lizhong Zheng, " Effective Resource Allocation in a Queue: How Much Control is Necessary? ,”  Allerton Conference on Communication, Control, and Computing, September 2008, pgs. 508 – 515.**

115.   Sebastian Neumayer, Gil Zussman, Rueven Cohen, Eytan Modiano, " Assessing the Impact of Geographically Correlated Network Failures ,”  IEEE MILCOM, November 2008. (6 pages; CD proceedings – page numbers not available).**

114.   Emily Craparo, Jonathan P. How, and Eytan Modiano, " Simultaneous Placement and Assignment for Exploration in Mobile Backbone Networks ,”  IEEE conference on Decision and Control (CDC), November 2008, pgs. 1696 – 1701 **

113.   Anand Srinivas and Eytan Modiano, " Joint node placement and assignment for throughput optimization in mobile backbone networks ,”  IEEE INFOCOM'08, pp. 1130 – 1138, Phoenix, AZ, Apr. 2008, pgs. 1130 – 1138.**

112.   Guner Celik, Gil Zussman, Wajahat Khan and Eytan Modiano, " MAC for Networks with Multipacket Reception Capability and Spatially Distributed Nodes ,”  IEEE INFOCOM'08, Phoenix, AZ, Apr. 2008, pgs. 1436 – 1444.**

111.   Gil Zussman, Andrew Brzezinski, and Eytan Modiano, " Multihop Local Pooling for Distributed Throughput Maximization in Wireless Networks ,”  IEEE INFOCOM'08, Phoenix, AZ, Apr. 2008, pgs 1139 – 1147.**

110.   Emily Craparo, Jonathan How and Eytan Modiano, " Optimization of Mobile Backbone Networks: Improved Algorithms and Approximation ,”  IEEE American Control Conference, Seattle, WA, June 2008, pgs. 2016 – 2021.**

109.   Atilla Eryilmaz, Asuman Ozdaglar, Devavrat Shah, Eytan Modiano, " Imperfect Randomized Algorithms for the Optimal Control of Wireless Networks ,”  Conference on Information Science and Systems, Princeton, NJ, March, 2008, pgs. 932 – 937.

108.   Anand Srinivas and Eytan Modiano, " Optimal Path Planning for Mobile Backbone Networks ,”  Conference on Information Science and Systems, Princeton, NJ, March, 2008, pgs. 913 – 918.

107.   Kayi Lee and Eytan Modiano, " Cross-layer Survivability in WDM Networks with Multiple Failures ,”  IEEE Optical Fiber Communications Conference, San Diego, CA February, 2008 (3 pages; CD proceedings – page numbers not available).

106.   Andrew Brzezinski, Gil Zussman and Eytan Modiano, " Local Pooling Conditions for Joint Routing and Scheduling ,”  Workshop on Information Theory and Applications, pp. 499 – 506, La Jolla, CA, January, 2008, pgs. 499 – 506.

105.   Murtaza Zafer and Eytan Modiano, " Minimum Energy Transmission over a Wireless Fading Channel with Packet Deadlines ,”  Proceedings of IEEE Conference on Decision and Control (CDC), New Orleans, LA, December, 2007, pgs. 1148 – 1155.**

104.   Atilla Eryilmaz, Asuman Ozdaglar, Eytan Modiano, " Polynomial Complexity Algorithms for Full Utilization of Multi-hop Wireless Networks ,”  IEEE Infocom, Anchorage, AK, April, 2007, pgs. 499 – 507.

103.   Murtaza Zafer and Eytan Modiano, " Delay Constrained Energy Efficient Data Transmission over a Wireless Fading Channel ,”  Workshop on Information Theory and Application, University of California, San Diego, CA, February, 2007, pgs. 289 – 298.**

102.   Atilla Eryilmaz, Eytan Modiano, Asuman Ozdaglar, " Randomized Algorithms for Throughput-Optimality and Fairness in Wireless Networks ,”  Proceedings of IEEE Conference on Decision and Control (CDC), San Diego, CA, December, 2006, pgs. 1936 – 1941.

101.   Anand Srinivas, Gil Zussman, and Eytan Modiano, " Distributed Mobile Disk Cover - A Building Block for Mobile Backbone Networks ,”  Proc. Allerton Conf. on Communication, Control, and Computing, Allerton, IL, September 2006, (9 pages; CD proceedings – page numbers not available).**

100.   Krishna Jagannathan, Sem Borst, Phil Whiting, Eytan Modiano, " Scheduling of Multi-Antenna Broadcast Systems with Heterogeneous Users ,”  Allerton Conference on Communication, Control and Computing, Allerton, IL, September 2006, (10 pages; CD proceedings – page numbers not available).**

99.   Andrew Brzezinski, Gil Zussman, and Eytan Modiano, " Enabling Distributed Throughput Maximization in Wireless Mesh Networks - A Partitioning Approach ,”  Proceedings of ACM MOBICOM'06, Los Angeles, CA, Sep. 2006, (12 pages; CD proceedings – page numbers not available).**

98.   Eytan Modiano, Devavrat Shah, and Gil Zussman, " Maximizing Throughput in Wireless Networks via Gossiping ,”  Proc. ACM SIGMETRICS / IFIP Performance'06, Saint-Malo, France, June 2006, (12 pages; CD proceedings – page numbers not available). (best paper award)

97.   Anand Srinivas, Gil Zussman, and Eytan Modiano, " Mobile Backbone Networks – Construction and Maintenance ,”  Proc. ACM MOBIHOC'06, Florence, Italy, May 2006, (12 pages; CD proceedings – page numbers not available).**

96.   Andrew Brzezinski and Eytan Modiano, " Achieving 100% throughput in reconfigurable optical networks ,”  IEEE INFOCOM 2006 High-Speed Networking Workshop, Barcelona, Spain, April 2006, (5 pages; CD proceedings – page numbers not available).**

95.   Krishna P. Jagannathan, Sem Borst, Phil Whiting, Eytan Modiano, " Efficient scheduling of multi-user multi-antenna systems ,”  Proceedings of WiOpt 2006, Boston, MA, April 2006, (8 pages; CD proceedings – page numbers not available).**

94.   Andrew Brzezinski and Eytan Modiano, " Greedy weighted matching for scheduling the input-queued switch ,”  Conference on Information Sciences and Systems (CISS), Princeton, NJ, March 2006, pgs. 1738 – 1743.**

93.   Murtaza Zafer and Eytan Modiano, " Optimal Adaptive Data Transmission over a Fading Channel with Deadline and Power Constraints ,”  Conference on Information Sciences and Systems (CISS), Princeton, New Jersey, March 2006, pgs. 931 – 937.**

92.   Li-Wei Chen and E. Modiano, " A Geometric Approach to Capacity Provisioning in WDM Networks with Dynamic Traffic ,”  Conference on Information Science and Systems (CISS), Princeton, NJ, March, 2006, pgs. 1676 – 1683, **

91.   Jun Sun and Eytan Modiano, " Channel Allocation Using Pricing in Satellite Networks ,”  Conference on Information Science and Systems (CISS), Princeton, NJ, March, 2006, pgs. 182 – 187.**

90.   Jun Sun, Jay Gao, Shervin Shambayatti and Eytan Modiano, " Ka-Band Link Optimization with Rate Adaptation ,”  IEEE Aerospace Conference, Big Sky, MN, March, 2006. (7 pages; CD proceedings – page numbers not available).

89.   Alessandro Tarello, Eytan Modiano and Jay Gao, " Energy efficient transmission scheduling over Mars proximity links ,”  IEEE Aerospace Conference, Big Sky, MN, March, 2006. (10 pages; CD proceedings – page numbers not available).

88.   A. Brzezinski and E. Modiano, " RWA decompositions for optimal throughput in reconfigurable optical networks ,”  INFORMS Telecommunications Conference, Dallas, TX, March 2006 (3 pages; CD proceedings – page numbers not available).**

87.   Li Wei Chen and E. Modiano, " Geometric Capacity Provisioning for Wavelength Switched WDM Networks ,”  Workshop on Information Theory and Application, University of California, San Diego, CA, February, 2006. (8 pages; CD proceedings – page numbers not available).**

86.   Murtaza Zafer and Eytan Modiano, " Joint Scheduling of Rate-guaranteed and Best-effort Services over a Wireless Channel ,”  IEEE Conference on Decision and Control, Seville, Spain, December, 2005, pgs. 6022–6027.**

85.   Jun Sun and Eytan Modiano, " Opportunistic Power Allocation for Fading Channels with Non-cooperative Users and Random Access ,”  IEEE BroadNets – Wireless Networking Symposium, Boston, MA, October, 2005, pgs. 397–405.**

84.   Li Wei Chen and Eytan Modiano, " Uniform vs. Non-uniform Band Switching in WDM Networks ,”  IEEE BroadNets-Optical Networking Symposium, Boston, MA, October, 2005, pgs. 219– 228.**

83.   Sonia Jain and Eytan Modiano, " Buffer Management Schemes for Enhanced TCP Performance over Satellite Links ,”  IEEE MILCOM, Atlantic City, NJ, October 2005 (8 pages; CD proceedings – page numbers not available).**

82.   Murtaza Zafer and Eytan Modiano, " Continuous-time Optimal Rate Control for Delay Constrained Data Transmission ,”  Allerton Conference on Communications, Control and Computing, Allerton, IL, September, 2005 (10 pages; CD proceedings – page numbers not available).**

81.   Alessandro Tarello, Eytan Modiano, Jun Sun, Murtaza Zafer, " Minimum Energy Transmission Scheduling subject to Deadline Constraints ,”  IEEE Wiopt, Trentino, Italy, April, 2005, pgs. 67–76. (Winner of best student paper award).**

80.   Amir Khandani, Eytan Modiano, Jinane Abounadi, Lizhong Zheng, " Reliability and Route Diversity in Wireless Networks ,”  Conference on Information Science and System, Baltimore, MD, March, 2005, (8 pages; CD proceedings – page numbers not available).**

79.   Andrew Brzezinski, Iraj Saniee, Indra Widjaja, Eytan Modiano, " Flow Control and Congestion Management for Distributed Scheduling of Burst Transmissions in Time-Domain Wavelength Interleaved Networks ,”  IEEE/OSA Optical Fiber Conference (OFC), Anaheim, CA, March, 2005, pgs. WC4-1–WC4-3.

78.   Andrew Brzezinski and Eytan Modiano, " Dynamic Reconfiguration and Routing Algorithms for IP-over-WDM Networks with Stochastic Traffic ,”  IEEE Infocom 2005, Miami, FL, March, 2005, pgs. 6–11.**

77.   Murtaza Zafer and Eytan Modiano, " A Calculus Approach to Minimum Energy Transmission Policies with Quality of Service Guarantees ,”  IEEE Infocom 2005, Miami, FL, March, 2005, pgs. 548–559.**

76.   Michael Neely and Eytan Modiano, " Fairness and optimal stochastic control for heterogeneous networks ,”  IEEE Infocom 2005, Miami, FL, March, 2005, pgs. 1723 – 1734.**

75.   Aradhana Narula-Tam, Thomas G. Macdonald, Eytan Modiano, and Leslie Servi, " A Dynamic Resource Allocation Strategy for Satellite Communications ,”  IEEE MILCOM, Monterey, CA, October, 2004, pgs. 1415 – 1421.

74.   Li-Wei Chen, Poompat Saengudomlert and Eytan Modiano, " Optimal Waveband Switching in WDM Networks ,”  IEEE International Conference on Communication (ICC), Paris, France, June, 2004, pgs. 1604 – 1608.**

73.   Michael Neely and Eytan Modiano, " Logarithmic Delay for NxN Packet Switches ,”  IEEE Workshop on High performance Switching and Routing (HPSR 2004), Phoenix, AZ, April, 2004, pgs. 3–9.**

72.   Li-Wei Chen and Eytan Modiano, " Dynamic Routing and Wavelength Assignment with Optical Bypass using Ring Embeddings ,”  IEEE Workshop on High performance Switching and Routing (HPSR 2004), Phoenix, Az, April, 2004, pgs. 119–125.**

71.   Randall Berry and Eytan Modiano, " On the Benefits of Tunability in Reducing Electronic Port Counts in WDM/TDM Networks ,”  IEEE Infocom, Hong Kong, March 2004, pgs. 1340–1351.

70.   Andrew Brzezinski and Eytan Modiano, " A new look at dynamic traffic scheduling in WDM networks with transceiver tuning latency ,”  Informs Telecommunications Conference, Boca Raton, FL, March 2004, pgs. 25–26.**

69.   Chunmei Liu and Eytan Modiano, " Packet Scheduling with Window Service Constraints ,”  Conference on Information Science and System, Princeton, NJ, March, 2004, pgs. 178–184.**

68.   Jun Sun, Eytan Modiano, and Lizhong Zheng, " A Novel Auction Algorithm for Fair Allocation of a Wireless Fading Channel ,”  Conference on Information Science and System, Princeton, NJ, March, 2004, pgs. 1377–1383.**

67.   Murtaza Zafer and Eytan Modiano, " Impact of Interference and Channel Assignment on Blocking Probability in Wireless Networks ,”  Conference on Information Science and System, Princeton, NJ, March, 2004, pgs. 430–436.**

66.   Chunmei Liu and Eytan Modiano, " An Analysis of TCP over Random Access Satellite Links ,”  IEEE Wireless Communications and Networking Conference (WCNC), Atlanta, GA, February, 2004, pgs. 2033–2040..**

65.   Randall Berry and Eytan Modiano, " Using tunable optical transceivers for reducing the number of ports in WDM/TDM Networks ,”  IEEE/OSA Optical Fiber Conference (OFC), Los Angeles, CA, February, 2004, pgs. 23–27.

64.   Aradhana Narula-Tam, Eytan Modiano and Andrew Brzezinski, " Physical Topology Design for Survivable Routiing of Logical Rings in WDM-based Networks ,”  IEEE Globecom, San francisco, CA, December, 2003, pgs. 2552–2557.

63.   Jun Sun, Lizhong Zheng and Eytan Modiano, " Wireless Channel Allocation Using an Auction Algorithm ,”  Allerton Conference on Communications, Control and Computing, October, 2003, pgs. 1114–1123..**

62.   Amir Khandani, Jinane Abounadi, Eytan Modiano, Lizhong Zhang, " Cooperative Routing in Wireless Networks ,”  Allerton Conference on Communications, Control and Computing, October, 2003, pgs. 1270–1279.**

61.   Poompat Saengudomlert, Eytan Modiano and Robert Gallager, " Dynamic Wavelength Assignment for WDM all optical Tree Networks ,”  Allerton Conference on Communications, Control and Computing, October, 2003, 915–924.**

60.   Aradhana Narula-Tam and Eytan Modiano, " Designing Physical Topologies that Enable Survivable Routing of Logical Rings ,”  IEEE Workshop on Design of Reliable Communication Networks (DRCN), October, 2003, pgs. 379–386.

59.   Anand Srinivas and Eytan Modiano, " Minimum Energy Disjoint Path Routing in Wireless Ad Hoc Networks ,”  ACM Mobicom, San Diego, Ca, September, 2003, pgs. 122–133.**

58.   Michael Neely and Eytan Modiano, " Improving Delay in Ad-Hoc Mobile Networks Via Redundant Packet Transfers ,”  Conference on Information Science and System, Baltimore, MD, March, 2003 (6 pages; CD proceedings – page numbers not available).**

57.   Michael Neely, Eytan Modiano and Charles Rohrs, " Dynamic Power Allocation and Routing for Time Varying Wireless Networks ,”  IEEE Infocom 2003, San Francisco, CA, April, 2003, pgs. 745–755.**

56.   Alvin Fu, Eytan Modiano, and John Tsitsiklis, " Optimal Energy Allocation for Delay-Constrained Data Transmission over a Time-Varying Channel ,”  IEEE Infocom 2003, San Francisco, CA, April, 2003, pgs. 1095–1105.**

55.   Poompat Saengudomlert, Eytan Modiano and Rober Gallager, " On-line Routing and Wavelength Assignment for Dynamic Traffic in WDM Ring and Torus Networks ,”  IEEE Infocom 2003, San Francisco, CA, April, 2003, pgs. 1805–1815.**

54.   Li-Wei Chen and Eytan Modiano, " Efficient Routing and Wavelength Assignment for Reconfigurable WDM Networks with Wavelength Converters ,”  IEEE Infocom 2003, San Francisco, CA, April, 2003, pgs. 1785–1794. Selected as one of the best papers of Infocom 2003 for fast track publication in IEEE/ACM Transactions on Networking.**

53.   Mike Neely, Jun Sun and Eytan Modiano, " Delay and Complexity Tradeoffs for Dynamic Routing and Power Allocation in a Wireless Network ,”  Allerton Conference on Communication, Control, and Computing, Allerton, Illinois, October, 2002, pgs. 157 –159.**

52.   Anand Ganti, Eytan Modiano and John Tsitsiklis, " Transmission Scheduling for Multi-Channel Satellite and Wireless Networks ,”  Allerton Conference on Communication, Control, and Computing, Allerton, Illinois, October, 2002, pgs. 1318–1327.**

51.   Poompat Saengudomlert, Eytan Modiano, and Robert G. Gallager, " Optimal Wavelength Assignment for Uniform All-to-All Traffic in WDM Tree Networks ,”  Allerton Conference on Communication, Control, and Computing, Allerton, Illinois, October, 2002, pgs. 528–537.**

50.   Hungjen Wang, Eytan Modiano and Muriel Medard, " Partial Path Protection for WDM Networks: End-to-End Recovery Using Local Failure Information ,”  IEEE International Symposium on Computer Communications (ISCC), Taormina, Italy, July 2002, pgs. 719–725.**

49.   Jun Sun and Eytan Modiano, " Capacity Provisioning and Failure Recovery in Mesh-Torus Networks with Application to Satellite Constellations ,”  IEEE International Symposium on Computer Communications (ISCC), Taormina, Italy, July 2002, pgs. 77–84.**

48.   Alvin Fu, Eytan Modiano, and John Tsitsiklis, " Optimal Energy Allocation and Admission Control for Communications Satellites ,”  IEEE INFOCOM 2002, New York, June, 2002, pgs. 648–656.**

47.   Michael Neely, Eytan Modiano and Charles Rohrs, " Power and Server Allocation in a Multi-Beam Satellite with Time Varying Channels ,”  IEEE INFOCOM 2002, New York, June, 2002, pgs. 1451–1460..**

46.   Mike Neely, Eytan Modiano and Charles Rohrs, " Tradeoffs in Delay Guarantees and Computation Complexity for N x N Packet Switches ,”  Conference on Information Science and Systems, Princeton, NJ, March, 2002, pgs. 136–148.**

45.   Alvin Fu, Eytan Modiano and John Tsitsiklis, " Transmission Scheduling Over a Fading Channel with Energy and Deadline Constraints ,”  Conference on Information Science and System, Princeton, NJ, March, 2002, pgs. 1018–1023.**

44.   Chunmei Liu and Eytan Modiano, " On the Interaction of Layered Protocols: The Case of Window Flow Control and ARQ ,”  Conference on Information Science and System, Princeton, NJ, March, 2002, pgs. 118–124.**

43.   Mike Neely, Eytan Modiano and Charles Rohrs, " Packet Routing over Parallel Time-varying Queues with Application to Satellite and Wireless Networks ,”  Conference on Information Science and System, Princeton, NJ, March, 2002, pgs. 360–366.**

42.   Ahluwalia Ashwinder, Eytan Modiano and Li Shu, " On the Complexity and Distributed Construction of Energy Efficient Broadcast Trees in Static Ad Hoc Wireless Networks ,”  Conference on Information Science and System, Princeton, NJ, March, 2002, pgs. 807–813.**

41.   Jun Sun and Eytan Modiano, " Capacity Provisioning and Failure Recovery for Satellite Constellations ,”  Conference on Information Science and System, Princeton, NJ, March, 2002, pgs. 1039–1045.**

40.   Eytan Modiano, Hungjen Wang, and Muriel Medard, " Partial Path Protection for WDM networks ,”  Informs Telecommunications Conference, Boca Raton, FL, March 2002, pgs. 78–79.**

39.   Poompat Saengudomlert, Eytan H. Modiano, and Robert G. Gallager, " An On-Line Routing and Wavelength Assignment Algorithm for Dynamic Traffic in a WDM Bidirectional Ring ,”  Joint Conference on Information Sciences (JCIS), Durham, North Carolina, March, 2002, pgs. 1331–1334.**

38.   Randy Berry and Eytan Modiano, " Switching and Traffic Grooming in WDM Networks ,”  Joint Conference on Information Sciences (JCIS), Durham, North Carolina, March, 2002, pgs. 1340–1343.

37.   Eytan Modiano, Hungjen Wang, and Muriel Medard, " Using Local Information for WDM Network Protection ,”  Joint Conference on Information Sciences (JCIS), Durham, North Carolina, March, 2002, pgs. 1398–1401.**

36.   Aradhana Narula-Tam and Eytan Modiano, " Network architectures for supporting survivable WDM rings ,”  IEEE/OSA Optical Fiber Conference (OFC) 2002, Anaheim, CA, March, 2002, pgs. 105–107.

35.   Michael Neely, Eytan Modiano, Charles Rohrs, " Packet Routing over Parallel Time-Varying Queues with Application to Satellite and Wireless Networks ,”  Allerton Conference on Communication, Control, and Computing, Allerton, Illinois, September, 2001, pgs. 1110-1111.**

34.   Eytan Modiano and Randy Berry, " The Role of Switching in Reducing Network Port Counts ,”  Allerton Conference on Communication, Control, and Computing, Allerton, Illinois, September, 2001, pgs. 376-385.

33.   Eytan Modiano, " Resource allocation and congestion control in next generation satellite networks ,”  IEEE Gigabit Networking Workshop (GBN 2001), Anchorage, AK, April 2001, (2 page summary-online proceedings).

32.   Eytan Modiano and Aradhana Narula-Tam, " Survivable Routing of Logical Topologies in WDM Networks ,”  IEEE Infocom 2001, Anchorage, AK, April 2001, pgs. 348–357.

31.   Michael Neely and Eytan Modiano, " Convexity and Optimal Load Distribution in Work Conserving */*/1 Queues ,”  IEEE Infocom 2001, Anchorage, AK, April 2001, pgs. 1055–1064.

30.   Eytan Modiano and Randy Berry, " Using Grooming Cross- Connects to Reduce ADM Costs in Sonet/WDM Ring Networks ,”  IEEE/OSA Optical Fiber Conference (OFC) 2001, Anaheim, CA March 2001, pgs. WL1- WL3.

29.   Eytan Modiano and Aradhana Narula-Tam, " Designing Survivable Networks Using Effective Rounting and Wavelenght Assignment (RWA) ,”  IEEE/OSA Optical Fiber Conference (OFC) 2001, Anaheim, CA March 2001, pgs. TUG5-1 – TUG5– 3.

28.   Roop Ganguly and Eytan Modiano, " Distributed Algorithms and Architectures for Optical Flow Switching in WDM networks ,”  IEEE International Symposium on Computer Communications (ISCC 2000), Antibes, France, July 2000, pgs. 134–139.

27.   Aradhana Narula-Tam, Philip J. Lin and Eytan Modiano, " Wavelength Requirements for Virtual topology Reconfiguration in WDM Ring Networks ,”  IEEE International Conference on Communications (ICC 2000), New Orleans, LA, June 2000, pgs. 1650–1654.

26.   Eytan Modiano, "Optical Flow Switching for the Next Generation Internet,”  IEEE Gigabit Networking Workshop (GBN 2000), Tel-aviv, March 2000 (2 page summary-online proceedings).

25.   Aradhana Narula and Eytan Modiano, " Dynamic Reconfiguration in WDM Packet Networks with Wavelength Limitations ,”  IEEE/OSA Optical Fiber Conference (OFC) 2000, Baltimore, MD, March, 2000, pgs. 1210–1212.

24.   Brett Schein and Eytan Modiano, " Quantifying the benefits of configurability in circuit-switched WDM ring networks ,”  IEEE Infocom 2000, Tel Aviv, Israel, April, 2000, pgs.1752–1760..***

23.   Aradhana Narula-Tam and Eytan Modiano, " Load Balancing Algorithms for WDM-based IP networks ,”  IEEE Infocom 2000, Tel Aviv, Israel, April, 2000, pgs. 1010–1019.

22.   Nan Froberg, M. Kuznetsov, E. Modiano, et. al., " The NGI ONRAMP test bed: Regional Access WDM technology for the Next Generation Internet ,”  IEEE LEOS ’99, October, 1999, pgs. 230–231.

21.   Randy Berry and Eytan Modiano, " Minimizing Electronic Multiplexing Costs for Dynamic Traffic in Unidirectional SONET Ring Networks ,”  IEEE International Conference on Communications (ICC ’99), Vancouver, CA, June 1999, pgs. 1724–1730..***

20.   Brett Schein and Eytan Modiano, "Increasing Traffic Capacity in WDM Ring Networks via Topology Reconfiguration,”  Conference on Information Science and Systems, Baltimore, MD, March 1999, pgs. 201 – 206.

19.   Eytan Modiano and Richard Barry, " Design and Analysis of an Asynchronous WDM Local Area Network Using a Master/Slave Scheduler ,”  IEEE Infocom ’99, New York, NY, March 1999, pgs. 900–907.

18.   Randy Berry and Eytan Modiano, " Grooming Dynamic Traffic in Unidirectional SONET Ring Networks ,”  IEEE/OSA Optical Fiber Conference (OFC) ’99, San Diego, CA, February 1999, pgs. 71–73.

17.   Angela Chiu and Eytan Modiano, " Reducing Electronic Multiplexing Costs in Unidirectional SONET/WDM Ring Networks Via Efficient Traffic Grooming ,”  IEEE Globecom '98, Sydney, Australia, November 1998, pgs. 322–327.

16.   Eytan Modiano, " Throughput Analysis of Unscheduled Multicast Transmissions in WDM Broadcast-and-Select Networks ,”  IEEE International Symposium on Information Theory, Boston, MA, September 1998, pg. 167.

15.   Eytan Modiano and Angela Chiu, "Traffic Grooming Algorithms for Minimizing Electronic Multiplexing Costs in Unidirectional SONET/WDM Ring Networks,”  Conference on Information Science and Systems, Princeton, NJ, March 1998, 653–658.

14.   Eytan Modiano and Eric Swanson, " An Architecture for Broadband Internet Services over a WDM-based Optical Access Network ,”  IEEE Gigabit Networking Workshop (GBN '98), San Francisco, CA, March 1998 (2 page summary-online proceedings).

13.   Eytan Modiano, " Unscheduled Multicasts in WDM Broadcast-and-Select Networks ,”  IEEE Infocom '98, San Francisco, CA, March 1998, pgs. 86–93.

12.   Eytan Modiano, Richard Barry and Eric Swanson, " A Novel Architecture and Medium Access Control (MAC) protocol for WDM Networks ,”  IEEE/OSA Optical Fiber Conference (OFC) '98, San Jose, CA, February 1998, pgs. 90–91.

11.   Eytan Modiano, " Scheduling Algorithms for Message Transmission Over a Satellite Broadcast System ,”  IEEE MILCOM 97, Monterey, CA, November 1997, pgs. 628–634.

10.   Eytan Modiano, " Scheduling Packet Transmissions in A Multi-hop Packet Switched Network Based on Message Length ,”  IEEE International Conference on Computer Communications and Networks (IC3N) Las Vegas, Nevada, September 1997, pgs. 350–357.

9.   Eytan Modiano, "A Simple Algorithm for Optimizing the Packet Size Used in ARQ Protocols Based on Retransmission History,”  Conference on Information Science and Systems, Baltimore, MD, March 1997, pgs. 672–677.

8.   Eytan Modiano, " A Multi-Channel Random Access Protocol for the CDMA Channel ,”  IEEE PIMRC '95, Toronto, Canada, September 1995, pgs. 799–803.

7.   Eytan Modiano Jeffrey Wieselthier and Anthony Ephremides, " A Simple Derivation of Queueing Delay in a Tree Network of Discrete-Time Queues with Deterministic Service Times ,”  IEEE International Symposium on Information Theory, Trondheim, Norway, June 1994, pg. 372.

6.   Eytan Modiano, Jeffrey Wieselthier and Anthony Ephremides, "An Approach for the Analysis of Packet Delay in an Integrated Mobile Radio Network,”  Conference on Information Sciences and Systems, Baltimore, MD, March 1993, pgs. 138-139.

5.   Eytan Modiano and Anthony Ephremides, " A Method for Delay Analysis of Interacting Queues in Multiple Access Systems ,”  IEEE INFOCOM 1993, San Francisco, CA, March 1993, pgs. 447 – 454.

4.   Eytan Modiano and Anthony Ephremides, " A Model for the Approximation of Interacting Queues that Arise in Multiple Access Schemes ,”  IEEE International Symposium on Information Theory, San Antonio, TX, January 1993, pg. 324.

3.   Eytan Modiano and Anthony Ephremides, " Efficient Routing Schemes for Multiple Broadcasts in a Mesh ,”  Conference on Information Sciences and Systems, Princeton, NJ, March 1992, pgs. 929 – 934.

2.   Eytan Modiano and Anthony Ephremides, " On the Secrecy Complexity of Computing a Binary Function of Non-uniformly Distributed Random Variables ,”  IEEE International Symposium on Information Theory, Budapest, Hungary, June 1991, pg. 213.

1.   Eytan Modiano and Anthony Ephremides, "Communication Complexity of Secure Distributed Computation in the Presence of Noise,”  IEEE International Symposium on Information Theory, San Diego, CA, January 1990, pg. 142.

Book Chapters

  • Hyang-Won Lee, Kayi Lee, Eytan Modiano, " Cross-Layer Survivability " in Cross-Layer Design in Optical Networks, Springer, 2013.
  • Li-Wei Chen and Eytan Modiano, " Geometric Capacity Provisioning for Wavelength-Switched WDM Networks ," Chapter in Computer Communications and Networks Series: Algorithms for Next Generation Networks, Springer, 2010.
  • Amir Khandani, Eytan Modiano, Lizhong Zhang, Jinane Aboundi, " Cooperative Routing in Wireless Networks ," Chapter in Advances in Pervasive Computing and Networking, Kluwer Academic Publishers, 2005.
  • Jian-Qiang Hu and Eytan Modiano, " Traffic Grooming in WDM Networks ," Chapter in Emerging Optical Network Technologies, Kluwer Academic Publishers, to appear, 2004.
  • Eytan Modiano, " WDM Optical Networks ," Wiley Encyclopedia of Telecommunications (John Proakis, Editor), 2003.
  • Eytan Modiano, " Optical Access Networks for the Next Generation Internet ," in Optical WDM Networks: Principles and Practice, Kluwer Academic Prublishers, 2002.
  • Eytan Modiano, Richard Barry and Eric Swanson, " A Novel Architecture and Medium Access Control protocol for WDM Networks ," Trends in Optics and Photonics Series (TOPS) volume on Optical Networks and Their Applications, 1998.
  • Eytan Modiano and Kai-Yeung Siu, "Network Flow and Congestion Control," Wiley Encyclopedia of Electrical and Electronics Engineering, 1999.

Technical Reports

  • Amir Khandani, Eytan Modiano, Jinane Abounadi, Lizhong Zheng, "Reliability and Route Diversity in Wireless Networks, " MIT LIDS Technical Report number 2634, November, 2004.
  • Anand Srinivas and Eytan Modiano, "Minimum Energy Disjoint Path Routing in Wireless Ad Hoc Networks, " MIT LIDS Technical Report, P-2559, March, 2003.
  • Eytan Modiano and Aradhana Narula-Tam, "Survivable lightpath routing: a new approach to the design of WDM-based networks, " LIDS report 2552, October, 2002.
  • Michael Neely, Eytan Modiano and Charles Rohrs, "Packet Routing over Parallel Time-Varying Queues with Application to Satellite and Wireless Networks," LIDS report 2520, September, 2001.
  • Jun Sun and Eytan Modiano, "Capacity Provisioning and Failure Recovery in Mesh-Torus Networks with Application to Satellite Constellations," LIDS report 2518, September, 2001.
  • Hungjen Wang, Eytan Modiano and Muriel Medard, "Partial Path Protection for WDM Networks: End-to-End Recovery Using Local Failure Information, " LIDS report 2517, Sept. 2001.
  • Alvin Fu, Eytan Modiano, and John Tsitsiklis, "Optimal Energy Allocation and Admission Control for Communications Satellites, " LIDS report 2516, September, 2001.
  • Michael Neely, Eytan Modiano and Charles Rohrs, "Power and Server Allocation in a Multi-Beam Satellite with Time Varying Channels, " LIDS report 2515, September, 2001.
  • Eytan Modiano, "Scheduling Algorithms for Message Transmission Over the GBS Satellite Broadcast System, " Lincoln Laboratory Technical Report Number TR-1035, June 1997.
  • Eytan Modiano, "Scheduling Packet Transmissions in A Multi-hop Packet Switched Network Based on Message Length, " Lincoln Laboratory Technical Report number TR-1036, June, 1997.

network security Recently Published Documents

Total documents.

  • Latest Documents
  • Most Cited Documents
  • Contributed Authors
  • Related Sources
  • Related Keywords

A Survey on Ransomware Malware and Ransomware Detection Techniques

Abstract: is a kind of malignant programming (malware) that takes steps to distribute or hinders admittance to information or a PC framework, for the most part by scrambling it, until the casualty pays a payoff expense to the assailant. As a rule, the payoff request accompanies a cutoff time. Assuming that the casualty doesn't pay on schedule, the information is gone perpetually or the payoff increments. Presently days and assailants executed new strategies for effective working of assault. In this paper, we center around ransomware network assaults and study of discovery procedures for deliver product assault. There are different recognition methods or approaches are accessible for identification of payment product assault. Keywords: Network Security, Malware, Ransomware, Ransomware Detection Techniques

Analysis and Evaluation of Wireless Network Security with the Penetration Testing Execution Standard (PTES)

The use of computer networks in an agency aims to facilitate communication and data transfer between devices. The network that can be applied can be using wireless media or LAN cable. At SMP XYZ, most of the computers still use wireless networks. Based on the findings in the field, it was found that there was no user management problem. Therefore, an analysis and audit of the network security system is needed to ensure that the network security system at SMP XYZ is safe and running well. In conducting this analysis, a tool is needed which will be used as a benchmark to determine the security of the wireless network. The tools used are Penetration Testing Execution Standard (PTES) which is one of the tools to become a standard in analyzing or auditing network security systems in a company in this case, namely analyzing and auditing wireless network security systems. After conducting an analysis based on these tools, there are still many security holes in the XYZ wireless SMP that allow outsiders to illegally access and obtain vulnerabilities in terms of WPA2 cracking, DoS, wireless router password cracking, and access point isolation so that it can be said that network security at SMP XYZ is still not safe

A Sensing Method of Network Security Situation Based on Markov Game Model

The sensing of network security situation (NSS) has become a hot issue. This paper first describes the basic principle of Markov model and then the necessary and sufficient conditions for the application of Markov game model. And finally, taking fuzzy comprehensive evaluation model as the theoretical basis, this paper analyzes the application fields of the sensing method of NSS with Markov game model from the aspects of network randomness, non-cooperative and dynamic evolution. Evaluation results show that the sensing method of NSS with Markov game model is best for financial field, followed by educational field. In addition, the model can also be used in the applicability evaluation of the sensing methods of different industries’ network security situation. Certainly, in different categories, and under the premise of different sensing methods of network security situation, the proportions of various influencing factors are different, and once the proportion is unreasonable, it will cause false calculation process and thus affect the results.

The Compound Prediction Analysis of Information Network Security Situation based on Support Vector Combined with BP Neural Network Learning Algorithm

In order to solve the problem of low security of data in network transmission and inaccurate prediction of future security situation, an improved neural network learning algorithm is proposed in this paper. The algorithm makes up for the shortcomings of the standard neural network learning algorithm, eliminates the redundant data by vector support, and realizes the effective clustering of information data. In addition, the improved neural network learning algorithm uses the order of data to optimize the "end" data in the standard neural network learning algorithm, so as to improve the accuracy and computational efficiency of network security situation prediction.MATLAB simulation results show that the data processing capacity of support vector combined BP neural network is consistent with the actual security situation data requirements, the consistency can reach 98%. the consistency of the security situation results can reach 99%, the composite prediction time of the whole security situation is less than 25s, the line segment slope change can reach 2.3% ,and the slope change range can reach 1.2%,, which is better than BP neural network algorithm.

Network intrusion detection using oversampling technique and machine learning algorithms

The expeditious growth of the World Wide Web and the rampant flow of network traffic have resulted in a continuous increase of network security threats. Cyber attackers seek to exploit vulnerabilities in network architecture to steal valuable information or disrupt computer resources. Network Intrusion Detection System (NIDS) is used to effectively detect various attacks, thus providing timely protection to network resources from these attacks. To implement NIDS, a stream of supervised and unsupervised machine learning approaches is applied to detect irregularities in network traffic and to address network security issues. Such NIDSs are trained using various datasets that include attack traces. However, due to the advancement in modern-day attacks, these systems are unable to detect the emerging threats. Therefore, NIDS needs to be trained and developed with a modern comprehensive dataset which contains contemporary common and attack activities. This paper presents a framework in which different machine learning classification schemes are employed to detect various types of network attack categories. Five machine learning algorithms: Random Forest, Decision Tree, Logistic Regression, K-Nearest Neighbors and Artificial Neural Networks, are used for attack detection. This study uses a dataset published by the University of New South Wales (UNSW-NB15), a relatively new dataset that contains a large amount of network traffic data with nine categories of network attacks. The results show that the classification models achieved the highest accuracy of 89.29% by applying the Random Forest algorithm. Further improvement in the accuracy of classification models is observed when Synthetic Minority Oversampling Technique (SMOTE) is applied to address the class imbalance problem. After applying the SMOTE, the Random Forest classifier showed an accuracy of 95.1% with 24 selected features from the Principal Component Analysis method.

Cyber Attacks Visualization and Prediction in Complex Multi-Stage Network

In network security, various protocols exist, but these cannot be said to be secure. Moreover, is not easy to train the end-users, and this process is time-consuming as well. It can be said this way, that it takes much time for an individual to become a good cybersecurity professional. Many hackers and illegal agents try to take advantage of the vulnerabilities through various incremental penetrations that can compromise the critical systems. The conventional tools available for this purpose are not enough to handle things as desired. Risks are always present, and with dynamically evolving networks, they are very likely to lead to serious incidents. This research work has proposed a model to visualize and predict cyber-attacks in complex, multilayered networks. The calculation will correspond to the cyber software vulnerabilities in the networks within the specific domain. All the available network security conditions and the possible places where an attacker can exploit the system are summarized.

Network Security Policy Automation

Network security policy automation enables enterprise security teams to keep pace with increasingly dynamic changes in on-premises and public/hybrid cloud environments. This chapter discusses the most common use cases for policy automation in the enterprise, and new automation methodologies to address them by taking the reader step-by-step through sample use cases. It also looks into how emerging automation solutions are using big data, artificial intelligence, and machine learning technologies to further accelerate network security policy automation and improve application and network security in the process.

Rule-Based Anomaly Detection Model with Stateful Correlation Enhancing Mobile Network Security

Research on network security technology of industrial control system.

The relationship between industrial control system and Internet is becoming closer and closer, and its network security has attracted much attention. Penetration testing is an active network intrusion detection technology, which plays an indispensable role in protecting the security of the system. This paper mainly introduces the principle of penetration testing, summarizes the current cutting-edge penetration testing technology, and looks forward to its development.

Detection and Prevention of Malicious Activities in Vulnerable Network Security Using Deep Learning

Export citation format, share document.

Toward Scalable Docker-Based Emulations of Blockchain Networks for Research and Development

  • Pennino, Diego
  • Pizzonia, Maurizio

Blockchain, like any other complex technology, needs a strong testing methodology to support its evolution in both research and development contexts. Setting up meaningful tests for permissionless blockchain technology is a notoriously complex task for several reasons: software is complex, large number of nodes are involved, network is non ideal, etc. Developers usually adopt small virtual laboratories or costly real devnets, based on real software. Researchers usually prefer simulations of a large number of nodes, based on simplified models. In this paper, we aim to obtain the advantages of both approaches, i.e., performing large, realistic, inexpensive, and flexible experiments, using real blockchain software within a virtual environment. To do that, we tackle the challenge of running large blockchain networks in a single physical machine, leveraging Linux and Docker. We analyze a number of problems that arise when large blockchain networks are emulated and we provide technical solutions for all of them. Finally, we describe two experiences of emulating fairly large blockchain networks on a single machine, adopting both research oriented and production oriented software, and involving up to more than 3000 containers.

  • Computer Science - Distributed, Parallel, and Cluster Computing;
  • Computer Science - Performance

IEEE Account

  • Change Username/Password
  • Update Address

Purchase Details

  • Payment Options
  • Order History
  • View Purchased Documents

Profile Information

  • Communications Preferences
  • Profession and Education
  • Technical Interests
  • US & Canada: +1 800 678 4333
  • Worldwide: +1 732 981 0060
  • Contact & Support
  • About IEEE Xplore
  • Accessibility
  • Terms of Use
  • Nondiscrimination Policy
  • Privacy & Opting Out of Cookies

A not-for-profit organization, IEEE is the world's largest technical professional organization dedicated to advancing technology for the benefit of humanity. © Copyright 2024 IEEE - All rights reserved. Use of this web site signifies your agreement to the terms and conditions.

Research and application of multimedia compression technology in online physical education teaching task

  • Original Paper
  • Published: 24 February 2024

Cite this article

  • Yuan Xue 1 &
  • Ning Li 1  

Traditional physical education teaching methods are facing increasingly severe challenges. This article studies how to use network technology more reasonably in college physical education teaching and better serve college physical education teaching. The purpose is to comprehensively analyze the requirements of physical education teaching and teaching management in universities, and to plan in detail the functions of the physical education teaching management system to better meet the needs of the development of physical education teaching in universities. Starting from the current development status of online teaching management platforms in Chinese universities, this paper conducts research on the existing problems of online teaching management platforms in higher physical education institutions. Using a questionnaire survey method, this study investigates and analyzes the current situation of online teaching platforms and course resources in 12 domestic higher sports colleges and universities. Starting from the actual situation of online education in physical education teaching in universities, a management platform model that organically combines online teaching platforms and digital teaching resource libraries is proposed. This study indicates that multimedia compression technology has important research value and broad application prospects in online physical education teaching. Through the application of this technology, the quality of online physical education teaching can be greatly improved, teaching costs can be reduced, and a more high-quality and efficient learning experience can be provided for students.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

network technology research paper

Data availability

No datasets were generated or analysed during the current study.

Hai, W.-P., Zhong, C.-L., Li, Y.-L.: The application of multimedia technology in physical education. Int. J. Electr. Eng. Educ. (2020)

Guo, L., Huang, J., Zhang, Y.: Education development in China: education return, quality, and equity. Sustainability 11 (13), 3750 (2019)

Article   Google Scholar  

Zhang, J.: Reform and innovation of artificial intelligence technology for information service in university physical education. J. Intell. Fuzzy Syst. 40 (2), 3325–3335 (2021)

Zhang, H., et al.: Ex-ante online risk assessment for building emergency evacuation through multimedia data. PLoS ONE 14 (4), e0215149 (2019)

Article   MathSciNet   CAS   PubMed   PubMed Central   Google Scholar  

Da-Wei, C., Chao, L., Shun, W., Xun-Ling, W., Wen-fang, W.: Research and application of multimedia digital platform in the teaching of college physical education course. J. Intell. Fuzzy Syst. 34 (2), 893–901 (2018)

Widjaja, G., Aslan, A.: Blended learning method in the view of learning and teaching strategy in geography study programs in higher education. Nazhruna: Jurnal Pendidikan Islam 5 (1), 22–36 (2022)

Battisti, F., Boato, G., Carli, M., Neri, A.: Teaching multimedia data protection through an international online competition. IEEE Trans. Educ. 54 (3), 381–386 (2010)

Yan, B.: Problems and countermeasures in the process of applying multimedia technology in basketball teaching. Wirel. Commun. Mob. Comput.. Commun. Mob. Comput. 2021 , 1–8 (2021)

CAS   Google Scholar  

Teng, Y., Jiang, P.-X., Wang, K.: Moral education teaching in colleges and universities based on the application of multimedia technology. Int. J. Electr. Eng. Educ.Electr. Eng. Educ. 60 , 0020720920940594 (2020)

Google Scholar  

Zhou, P., Zhou, Y., Wu, D., Jin, H.: Differentially private online learning for cloud-based video recommendation with multimedia big data in social networks. IEEE Trans. Multimedia 18 (6), 1217–1229 (2016)

Shen, Y., Yu, H.: Multimedia network teaching platform and its application in mechanical design. Int. J. Electr. Eng. Educ. 0020720920983558 (2021)

Korcz, A., et al.: Physical education teachers’ opinion about online teaching during the COVID-19 pandemic—Comparative study of European countries. Sustainability 13 (21), 11730 (2021)

Article   CAS   Google Scholar  

Ayres, L.B., Lopes, F.S., Garcia, C.D., Gutz, I.G.R.: Integrated instrumental analysis teaching platform with smartphone-operated fluorometer. Anal. Methods 12 (33), 4109–4115 (2020)

Article   CAS   PubMed   Google Scholar  

Zhang, Z., Min, H.: Analysis on the construction of personalized physical education teaching system based on a cloud computing platform. Wirel. Commun. Mob. Comput.. Commun. Mob. Comput. 2020 , 1–8 (2020)

Ba, Y., Qi, L.: Construction of WeChat mobile teaching platform in the reform of physical education teaching strategy based on deep neural network. Mob. Inf. Syst. 2021 , 1–12 (2021)

Qi, S., Li, S., Zhang, J.: Designing a teaching assistant system for physical education using web technology. Mob. Inf. Syst. 2021 , 1–11 (2021)

Wang, Y.: Physical education teaching in colleges and universities assisted by virtual reality technology based on artificial intelligence. Math. Probl. Eng.Probl. Eng. 2021 , 1–11 (2021)

Yuejue, Y., Xinze, S., Bingyue, L., Xiaoyuan, W.: Construct a teaching system combining image linguistics and multimedia technology. Wirel. Commun. Mob. Comput.. Commun. Mob. Comput. 2021 , 1–11 (2021)

Xu, D., Rappaport, T.S.: Construction on teaching evaluation index system of track and field general course for physical education major in light of wireless network technology. J. Intell. Fuzzy Syst. 37 (3), 3435–3443 (2019)

Viciana, J., Mayorga-Vega, D.: Innovative teaching units applied to Physical Education–changing the curriculum management for authentic outcomes. Kinesiology 48 (1), 142–152 (2016)

Download references

Author information

Authors and affiliations.

Department of Sports, Qinhuangdao Vocational and Technical College, Qinhuangdao, 066100, Hebei, China

Yuan Xue & Ning Li

You can also search for this author in PubMed   Google Scholar

Contributions

NL: Writing-Original draft preparation, Conceptualization, Supervision, Project administration. YX: Methodology, Software, Validation.

Corresponding author

Correspondence to Ning Li .

Ethics declarations

Conflict of interest.

The authors declare that they have no conflict of interest.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Xue, Y., Li, N. Research and application of multimedia compression technology in online physical education teaching task. SIViP (2024). https://doi.org/10.1007/s11760-024-03012-8

Download citation

Received : 28 November 2023

Revised : 06 January 2024

Accepted : 08 January 2024

Published : 24 February 2024

DOI : https://doi.org/10.1007/s11760-024-03012-8

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

  • Multimedia data
  • Physical education
  • Compression technique
  • Find a journal
  • Publish with us
  • Track your research

Our next-generation model: Gemini 1.5

Feb 15, 2024

The model delivers dramatically enhanced performance, with a breakthrough in long-context understanding across modalities.

SundarPichai_2x.jpg

A note from Google and Alphabet CEO Sundar Pichai:

Last week, we rolled out our most capable model, Gemini 1.0 Ultra, and took a significant step forward in making Google products more helpful, starting with Gemini Advanced . Today, developers and Cloud customers can begin building with 1.0 Ultra too — with our Gemini API in AI Studio and in Vertex AI .

Our teams continue pushing the frontiers of our latest models with safety at the core. They are making rapid progress. In fact, we’re ready to introduce the next generation: Gemini 1.5. It shows dramatic improvements across a number of dimensions and 1.5 Pro achieves comparable quality to 1.0 Ultra, while using less compute.

This new generation also delivers a breakthrough in long-context understanding. We’ve been able to significantly increase the amount of information our models can process — running up to 1 million tokens consistently, achieving the longest context window of any large-scale foundation model yet.

Longer context windows show us the promise of what is possible. They will enable entirely new capabilities and help developers build much more useful models and applications. We’re excited to offer a limited preview of this experimental feature to developers and enterprise customers. Demis shares more on capabilities, safety and availability below.

Introducing Gemini 1.5

By Demis Hassabis, CEO of Google DeepMind, on behalf of the Gemini team

This is an exciting time for AI. New advances in the field have the potential to make AI more helpful for billions of people over the coming years. Since introducing Gemini 1.0 , we’ve been testing, refining and enhancing its capabilities.

Today, we’re announcing our next-generation model: Gemini 1.5.

Gemini 1.5 delivers dramatically enhanced performance. It represents a step change in our approach, building upon research and engineering innovations across nearly every part of our foundation model development and infrastructure. This includes making Gemini 1.5 more efficient to train and serve, with a new Mixture-of-Experts (MoE) architecture.

The first Gemini 1.5 model we’re releasing for early testing is Gemini 1.5 Pro. It’s a mid-size multimodal model, optimized for scaling across a wide-range of tasks, and performs at a similar level to 1.0 Ultra , our largest model to date. It also introduces a breakthrough experimental feature in long-context understanding.

Gemini 1.5 Pro comes with a standard 128,000 token context window. But starting today, a limited group of developers and enterprise customers can try it with a context window of up to 1 million tokens via AI Studio and Vertex AI in private preview.

As we roll out the full 1 million token context window, we’re actively working on optimizations to improve latency, reduce computational requirements and enhance the user experience. We’re excited for people to try this breakthrough capability, and we share more details on future availability below.

These continued advances in our next-generation models will open up new possibilities for people, developers and enterprises to create, discover and build using AI.

Context lengths of leading foundation models

Highly efficient architecture

Gemini 1.5 is built upon our leading research on Transformer and MoE architecture. While a traditional Transformer functions as one large neural network, MoE models are divided into smaller "expert” neural networks.

Depending on the type of input given, MoE models learn to selectively activate only the most relevant expert pathways in its neural network. This specialization massively enhances the model’s efficiency. Google has been an early adopter and pioneer of the MoE technique for deep learning through research such as Sparsely-Gated MoE , GShard-Transformer , Switch-Transformer, M4 and more.

Our latest innovations in model architecture allow Gemini 1.5 to learn complex tasks more quickly and maintain quality, while being more efficient to train and serve. These efficiencies are helping our teams iterate, train and deliver more advanced versions of Gemini faster than ever before, and we’re working on further optimizations.

Greater context, more helpful capabilities

An AI model’s “context window” is made up of tokens, which are the building blocks used for processing information. Tokens can be entire parts or subsections of words, images, videos, audio or code. The bigger a model’s context window, the more information it can take in and process in a given prompt — making its output more consistent, relevant and useful.

Through a series of machine learning innovations, we’ve increased 1.5 Pro’s context window capacity far beyond the original 32,000 tokens for Gemini 1.0. We can now run up to 1 million tokens in production.

This means 1.5 Pro can process vast amounts of information in one go — including 1 hour of video, 11 hours of audio, codebases with over 30,000 lines of code or over 700,000 words. In our research, we’ve also successfully tested up to 10 million tokens.

Complex reasoning about vast amounts of information

1.5 Pro can seamlessly analyze, classify and summarize large amounts of content within a given prompt. For example, when given the 402-page transcripts from Apollo 11’s mission to the moon, it can reason about conversations, events and details found across the document.

Reasoning across a 402-page transcript: Gemini 1.5 Pro Demo

Gemini 1.5 Pro can understand, reason about and identify curious details in the 402-page transcripts from Apollo 11’s mission to the moon.

Better understanding and reasoning across modalities

1.5 Pro can perform highly-sophisticated understanding and reasoning tasks for different modalities, including video. For instance, when given a 44-minute silent Buster Keaton movie , the model can accurately analyze various plot points and events, and even reason about small details in the movie that could easily be missed.

Multimodal prompting with a 44-minute movie: Gemini 1.5 Pro Demo

Gemini 1.5 Pro can identify a scene in a 44-minute silent Buster Keaton movie when given a simple line drawing as reference material for a real-life object.

Relevant problem-solving with longer blocks of code

1.5 Pro can perform more relevant problem-solving tasks across longer blocks of code. When given a prompt with more than 100,000 lines of code, it can better reason across examples, suggest helpful modifications and give explanations about how different parts of the code works.

Problem solving across 100,633 lines of code | Gemini 1.5 Pro Demo

Gemini 1.5 Pro can reason across 100,000 lines of code giving helpful solutions, modifications and explanations.

Enhanced performance

When tested on a comprehensive panel of text, code, image, audio and video evaluations, 1.5 Pro outperforms 1.0 Pro on 87% of the benchmarks used for developing our large language models (LLMs). And when compared to 1.0 Ultra on the same benchmarks, it performs at a broadly similar level.

Gemini 1.5 Pro maintains high levels of performance even as its context window increases. In the Needle In A Haystack (NIAH) evaluation, where a small piece of text containing a particular fact or statement is purposely placed within a long block of text, 1.5 Pro found the embedded text 99% of the time, in blocks of data as long as 1 million tokens.

Gemini 1.5 Pro also shows impressive “in-context learning” skills, meaning that it can learn a new skill from information given in a long prompt, without needing additional fine-tuning. We tested this skill on the Machine Translation from One Book (MTOB) benchmark, which shows how well the model learns from information it’s never seen before. When given a grammar manual for Kalamang , a language with fewer than 200 speakers worldwide, the model learns to translate English to Kalamang at a similar level to a person learning from the same content.

As 1.5 Pro’s long context window is the first of its kind among large-scale models, we’re continuously developing new evaluations and benchmarks for testing its novel capabilities.

For more details, see our Gemini 1.5 Pro technical report .

Extensive ethics and safety testing

In line with our AI Principles and robust safety policies, we’re ensuring our models undergo extensive ethics and safety tests. We then integrate these research learnings into our governance processes and model development and evaluations to continuously improve our AI systems.

Since introducing 1.0 Ultra in December, our teams have continued refining the model, making it safer for a wider release. We’ve also conducted novel research on safety risks and developed red-teaming techniques to test for a range of potential harms.

In advance of releasing 1.5 Pro, we've taken the same approach to responsible deployment as we did for our Gemini 1.0 models, conducting extensive evaluations across areas including content safety and representational harms, and will continue to expand this testing. Beyond this, we’re developing further tests that account for the novel long-context capabilities of 1.5 Pro.

Build and experiment with Gemini models

We’re committed to bringing each new generation of Gemini models to billions of people, developers and enterprises around the world responsibly.

Starting today, we’re offering a limited preview of 1.5 Pro to developers and enterprise customers via AI Studio and Vertex AI . Read more about this on our Google for Developers blog and Google Cloud blog .

We’ll introduce 1.5 Pro with a standard 128,000 token context window when the model is ready for a wider release. Coming soon, we plan to introduce pricing tiers that start at the standard 128,000 context window and scale up to 1 million tokens, as we improve the model.

Early testers can try the 1 million token context window at no cost during the testing period, though they should expect longer latency times with this experimental feature. Significant improvements in speed are also on the horizon.

Developers interested in testing 1.5 Pro can sign up now in AI Studio, while enterprise customers can reach out to their Vertex AI account team.

Learn more about Gemini’s capabilities and see how it works .

Get more stories from Google in your inbox.

Your information will be used in accordance with Google's privacy policy.

Done. Just one step more.

Check your inbox to confirm your subscription.

You are already subscribed to our newsletter.

You can also subscribe with a different email address .

Related stories

Gemini models are coming to performance max.

gemma-header

Gemma: Introducing new state-of-the-art open models

What is a long context window.

MSC_Keyword_Cover (3)

How AI can strengthen digital security

Shield

Working together to address AI risks and opportunities at MSC

AI Evergreen 1 (1)

How we’re partnering with the industry, governments and civil society to advance AI

Let’s stay in touch. Get the latest news from Google in your inbox.

  • See us on facebook
  • See us on twitter
  • See us on youtube
  • See us on linkedin
  • See us on instagram

Stanford Medicine study identifies distinct brain organization patterns in women and men

Stanford Medicine researchers have developed a powerful new artificial intelligence model that can distinguish between male and female brains.

February 20, 2024

sex differences in brain

'A key motivation for this study is that sex plays a crucial role in human brain development, in aging, and in the manifestation of psychiatric and neurological disorders,' said Vinod Menon. clelia-clelia

A new study by Stanford Medicine investigators unveils a new artificial intelligence model that was more than 90% successful at determining whether scans of brain activity came from a woman or a man.

The findings, published Feb. 20 in the Proceedings of the National Academy of Sciences, help resolve a long-term controversy about whether reliable sex differences exist in the human brain and suggest that understanding these differences may be critical to addressing neuropsychiatric conditions that affect women and men differently.

“A key motivation for this study is that sex plays a crucial role in human brain development, in aging, and in the manifestation of psychiatric and neurological disorders,” said Vinod Menon , PhD, professor of psychiatry and behavioral sciences and director of the Stanford Cognitive and Systems Neuroscience Laboratory . “Identifying consistent and replicable sex differences in the healthy adult brain is a critical step toward a deeper understanding of sex-specific vulnerabilities in psychiatric and neurological disorders.”

Menon is the study’s senior author. The lead authors are senior research scientist Srikanth Ryali , PhD, and academic staff researcher Yuan Zhang , PhD.

“Hotspots” that most helped the model distinguish male brains from female ones include the default mode network, a brain system that helps us process self-referential information, and the striatum and limbic network, which are involved in learning and how we respond to rewards.

The investigators noted that this work does not weigh in on whether sex-related differences arise early in life or may be driven by hormonal differences or the different societal circumstances that men and women may be more likely to encounter.

Uncovering brain differences

The extent to which a person’s sex affects how their brain is organized and operates has long been a point of dispute among scientists. While we know the sex chromosomes we are born with help determine the cocktail of hormones our brains are exposed to — particularly during early development, puberty and aging — researchers have long struggled to connect sex to concrete differences in the human brain. Brain structures tend to look much the same in men and women, and previous research examining how brain regions work together has also largely failed to turn up consistent brain indicators of sex.

test

Vinod Menon

In their current study, Menon and his team took advantage of recent advances in artificial intelligence, as well as access to multiple large datasets, to pursue a more powerful analysis than has previously been employed. First, they created a deep neural network model, which learns to classify brain imaging data: As the researchers showed brain scans to the model and told it that it was looking at a male or female brain, the model started to “notice” what subtle patterns could help it tell the difference.

This model demonstrated superior performance compared with those in previous studies, in part because it used a deep neural network that analyzes dynamic MRI scans. This approach captures the intricate interplay among different brain regions. When the researchers tested the model on around 1,500 brain scans, it could almost always tell if the scan came from a woman or a man.

The model’s success suggests that detectable sex differences do exist in the brain but just haven’t been picked up reliably before. The fact that it worked so well in different datasets, including brain scans from multiple sites in the U.S. and Europe, make the findings especially convincing as it controls for many confounds that can plague studies of this kind.

“This is a very strong piece of evidence that sex is a robust determinant of human brain organization,” Menon said.

Making predictions

Until recently, a model like the one Menon’s team employed would help researchers sort brains into different groups but wouldn’t provide information about how the sorting happened. Today, however, researchers have access to a tool called “explainable AI,” which can sift through vast amounts of data to explain how a model’s decisions are made.

Using explainable AI, Menon and his team identified the brain networks that were most important to the model’s judgment of whether a brain scan came from a man or a woman. They found the model was most often looking to the default mode network, striatum, and the limbic network to make the call.

The team then wondered if they could create another model that could predict how well participants would do on certain cognitive tasks based on functional brain features that differ between women and men. They developed sex-specific models of cognitive abilities: One model effectively predicted cognitive performance in men but not women, and another in women but not men. The findings indicate that functional brain characteristics varying between sexes have significant behavioral implications.

“These models worked really well because we successfully separated brain patterns between sexes,” Menon said. “That tells me that overlooking sex differences in brain organization could lead us to miss key factors underlying neuropsychiatric disorders.”

While the team applied their deep neural network model to questions about sex differences, Menon says the model can be applied to answer questions regarding how just about any aspect of brain connectivity might relate to any kind of cognitive ability or behavior. He and his team plan to make their model publicly available for any researcher to use.

“Our AI models have very broad applicability,” Menon said. “A researcher could use our models to look for brain differences linked to learning impairments or social functioning differences, for instance — aspects we are keen to understand better to aid individuals in adapting to and surmounting these challenges.”

The research was sponsored by the National Institutes of Health (grants MH084164, EB022907, MH121069, K25HD074652 and AG072114), the Transdisciplinary Initiative, the Uytengsu-Hamilton 22q11 Programs, the Stanford Maternal and Child Health Research Institute, and the NARSAD Young Investigator Award.

About Stanford Medicine

Stanford Medicine is an integrated academic health system comprising the Stanford School of Medicine and adult and pediatric health care delivery systems. Together, they harness the full potential of biomedicine through collaborative research, education and clinical care for patients. For more information, please visit med.stanford.edu .

Artificial intelligence

Exploring ways AI is applied to health care

Stanford Medicine Magazine: AI

  • Newsletters

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

  • Will Douglas Heaven archive page

OpenAI has built a striking new generative video model called Sora that can take a short text description and turn it into a detailed, high-definition film clip up to a minute long.

Based on four sample videos that OpenAI shared with MIT Technology Review ahead of today’s announcement, the San Francisco–based firm has pushed the envelope of what’s possible with text-to-video generation (a hot new research direction that we flagged as a trend to watch in 2024 ).

“We think building models that can understand video, and understand all these very complex interactions of our world, is an important step for all future AI systems,” says Tim Brooks, a scientist at OpenAI.

But there’s a disclaimer. OpenAI gave us a preview of Sora (which means sky in Japanese) under conditions of strict secrecy. In an unusual move, the firm would only share information about Sora if we agreed to wait until after news of the model was made public to seek the opinions of outside experts. [Editor’s note: We’ve updated this story with outside comment below.] OpenAI has not yet released a technical report or demonstrated the model actually working. And it says it won’t be releasing Sora anytime soon. [ Update: OpenAI has now shared more technical details on its website.]

The first generative models that could produce video from snippets of text appeared in late 2022. But early examples from Meta , Google, and a startup called Runway were glitchy and grainy. Since then, the tech has been getting better fast. Runway’s gen-2 model, released last year, can produce short clips that come close to matching big-studio animation in their quality. But most of these examples are still only a few seconds long.  

The sample videos from OpenAI’s Sora are high-definition and full of detail. OpenAI also says it can generate videos up to a minute long. One video of a Tokyo street scene shows that Sora has learned how objects fit together in 3D: the camera swoops into the scene to follow a couple as they walk past a row of shops.

OpenAI also claims that Sora handles occlusion well. One problem with existing models is that they can fail to keep track of objects when they drop out of view. For example, if a truck passes in front of a street sign, the sign might not reappear afterward.  

In a video of a papercraft underwater scene, Sora has added what look like cuts between different pieces of footage, and the model has maintained a consistent style between them.

It’s not perfect. In the Tokyo video, cars to the left look smaller than the people walking beside them. They also pop in and out between the tree branches. “There’s definitely some work to be done in terms of long-term coherence,” says Brooks. “For example, if someone goes out of view for a long time, they won’t come back. The model kind of forgets that they were supposed to be there.”

Impressive as they are, the sample videos shown here were no doubt cherry-picked to show Sora at its best. Without more information, it is hard to know how representative they are of the model’s typical output.   

It may be some time before we find out. OpenAI’s announcement of Sora today is a tech tease, and the company says it has no current plans to release it to the public. Instead, OpenAI will today begin sharing the model with third-party safety testers for the first time.

In particular, the firm is worried about the potential misuses of fake but photorealistic video . “We’re being careful about deployment here and making sure we have all our bases covered before we put this in the hands of the general public,” says Aditya Ramesh, a scientist at OpenAI, who created the firm’s text-to-image model DALL-E .

But OpenAI is eyeing a product launch sometime in the future. As well as safety testers, the company is also sharing the model with a select group of video makers and artists to get feedback on how to make Sora as useful as possible to creative professionals. “The other goal is to show everyone what is on the horizon, to give a preview of what these models will be capable of,” says Ramesh.

To build Sora, the team adapted the tech behind DALL-E 3, the latest version of OpenAI’s flagship text-to-image model. Like most text-to-image models, DALL-E 3 uses what’s known as a diffusion model. These are trained to turn a fuzz of random pixels into a picture.

Sora takes this approach and applies it to videos rather than still images. But the researchers also added another technique to the mix. Unlike DALL-E or most other generative video models, Sora combines its diffusion model with a type of neural network called a transformer.

Transformers are great at processing long sequences of data, like words. That has made them the special sauce inside large language models like OpenAI’s GPT-4 and Google DeepMind’s Gemini . But videos are not made of words. Instead, the researchers had to find a way to cut videos into chunks that could be treated as if they were. The approach they came up with was to dice videos up across both space and time. “It’s like if you were to have a stack of all the video frames and you cut little cubes from it,” says Brooks.

The transformer inside Sora can then process these chunks of video data in much the same way that the transformer inside a large language model processes words in a block of text. The researchers say that this let them train Sora on many more types of video than other text-to-video models, varied in terms of resolution, duration, aspect ratio, and orientation. “It really helps the model,” says Brooks. “That is something that we’re not aware of any existing work on.”

“From a technical perspective it seems like a very significant leap forward,” says Sam Gregory, executive director at Witness, a human rights organization that specializes in the use and misuse of video technology. “But there are two sides to the coin,” he says. “The expressive capabilities offer the potential for many more people to be storytellers using video. And there are also real potential avenues for misuse.” 

OpenAI is well aware of the risks that come with a generative video model. We are already seeing the large-scale misuse of deepfake images . Photorealistic video takes this to another level.

Gregory notes that you could use technology like this to misinform people about conflict zones or protests. The range of styles is also interesting, he says. If you could generate shaky footage that looked like something shot with a phone, it would come across as more authentic.

The tech is not there yet, but generative video has gone from zero to Sora in just 18 months. “We’re going to be entering a universe where there will be fully synthetic content, human-generated content and a mix of the two,” says Gregory.

The OpenAI team plans to draw on the safety testing it did last year for DALL-E 3. Sora already includes a filter that runs on all prompts sent to the model that will block requests for violent, sexual, or hateful images, as well as images of known people. Another filter will look at frames of generated videos and block material that violates OpenAI’s safety policies.

OpenAI says it is also adapting a fake-image detector developed for DALL-E 3 to use with Sora. And the company will embed industry-standard C2PA tags , metadata that states how an image was generated, into all of Sora’s output. But these steps are far from foolproof. Fake-image detectors are hit-or-miss. Metadata is easy to remove, and most social media sites strip it from uploaded images by default.  

“We’ll definitely need to get more feedback and learn more about the types of risks that need to be addressed with video before it would make sense for us to release this,” says Ramesh.

Brooks agrees. “Part of the reason that we’re talking about this research now is so that we can start getting the input that we need to do the work necessary to figure out how it could be safely deployed,” he says.

Update 2/15: Comments from Sam Gregory were added .

Artificial intelligence

Ai for everything: 10 breakthrough technologies 2024.

Generative AI tools like ChatGPT reached mass adoption in record time, and reset the course of an entire industry.

What’s next for AI in 2024

Our writers look at the four hot trends to watch out for this year

  • Melissa Heikkilä archive page

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

Deploying high-performance, energy-efficient AI

Investments into downsized infrastructure can help enterprises reap the benefits of AI while mitigating energy consumption, says corporate VP and GM of data center platform engineering and architecture at Intel, Zane Ball.

  • MIT Technology Review Insights archive page

Stay connected

Get the latest updates from mit technology review.

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at [email protected] with a list of newsletters you’d like to receive.

IMAGES

  1. The Research of Distributed Computer Network Technology Based on Web

    network technology research paper

  2. 😂 Research paper on wireless network. Research paper on wireless

    network technology research paper

  3. 😀 Research paper on web technology. Technology Research Paper. 2019-02-02

    network technology research paper

  4. 📗 Research Paper on Web Technologies

    network technology research paper

  5. (PDF) NETWORKING-HUMAN AREA NETWORK

    network technology research paper

  6. (PDF) Network Analysis Part 2

    network technology research paper

VIDEO

  1. Fundamentals of information technology 1 semester model paper

  2. INTERNATIONAL GUEST LECTURE, PROF. HYUNJIN LEE FROM KOOKMIN UNIVERSITY

  3. Computer Networks Technology

  4. Internet Technologies

  5. Information Technology, What is IT

  6. Computer Networks Technology

COMMENTS

  1. Full article: The past, present, and future of network monitoring: A

    Research in network monitoring, and more generally network science, spans multiple disciplines, including mathematics, statistics, physics, and computer science, to name a few. Describe what you recognize as advantages and disadvantages associated with the interdisciplinary nature of this area, particularly as they pertain to research in, and ...

  2. Study and Investigation on 5G Technology: A Systematic Review

    Abstract. In wireless communication, Fifth Generation (5G) Technology is a recent generation of mobile networks. In this paper, evaluations in the field of mobile communication technology are presented. In each evolution, multiple challenges were faced that were captured with the help of next-generation mobile networks.

  3. A Survey of 5G Network: Architecture and Emerging Technologies

    In the near future, i.e., beyond 4G, some of the prime objectives or demands that need to be addressed are increased capacity, improved data rate, decreased latency, and better quality of service. To meet these demands, drastic improvements need to be made in cellular network architecture. This paper presents the results of a detailed survey on the fifth generation (5G) cellular network ...

  4. A Study on 5G Technology and Its Applications in Telecommunications

    The 5G network is a promising technology that revolutionizes and connects the global world through seamless connectivity. This paper presents a survey on 5G networks on how, in particular, it to address the drawbacks of foregoing cellular standards and be a potential key facilitator for the future as well as the extant technologies such as IoT ...

  5. Evolution and Impact of Wi-Fi Technology and Applications: A ...

    The IEEE 802.11 standard for wireless local area networking (WLAN), commercially known as Wi-Fi, has become a necessity in our day-to-day life. Over a billion Wi-Fi access points connect close to hundred billion of IoT devices, smart phones, tablets, laptops, desktops, smart TVs, video cameras, monitors, printers, and other consumer devices to the Internet to enable millions of applications to ...

  6. The rise of 5G technologies and systems: A quantitative analysis of

    This paper presents a systematic outline of the development of 5G-related research until 2020 as revealed by over 10,000 science and technology publications. The exercise addresses the emergence, growth, and impact of this body of work and offers insights regarding disciplinary distribution, international performance, and historical dynamics.

  7. The impact of 5G on the evolution of intelligent automation and

    Cellular wireless networks have come a long way since the first 1G system was introduced in 1981, with a new mobile generation appearing approximately every 10 years (Pathak 2013; Mishra 2018).In the past 30 years, the mobile industry has transformed society through 4 or 5 generations of technology revolution and evolution, namely 1G, 2G, 3G, and 4G networking technologies (Fig. 1). 1G gave us ...

  8. PDF 5G-Advanced Technology Evolution from a Network Perspective (2021)

    technologies further. Therefore, this white paper proposes that continuous research on the follow-up evolution of 5G networks— 5G-Advanced [1] is required, and full consideration of architecture evolution and function enhancement is needed. This white paper first analyzes the network evolution architecture

  9. Computer Networks

    The International Journal of Computer and Telecommunications Networking Computer Networks is an international, archival journal providing a publication vehicle for complete coverage of all topics of interest to those involved in the computer communications networking area. The audience includes researchers, managers and operators of networks as well as designers and implementors.

  10. Intelligent Techniques for Detecting Network Attacks: Review and

    The first section provides an introduction and background to the research area. A brief overview of network attacks is presented in Section 2. Section 3 discusses intelligent network attack mitigation techniques where all the reviewed research papers, the network attacks they address using ML and DL techniques, and their findings are presented ...

  11. A Review on Implementation Issues in IPv6 Network Technology

    A Review on Implementation Issues in IPv6 Network Technology. International Journal of Engineering Research and General Science, 3(6), 800-809. 10 Pages Posted: 30 Apr 2020. See all articles by Ramesh Chand Meena ... In this review paper, we reviewed research papers presented by the researchers between 2007 and 2015 related to IPv4 & IPv6 ...

  12. PUBLICATIONS

    Communications and Networking Research Group. Search form. Search . Research; ... A new approach to survivability in multi-layer networks," IEEE Journal on Lightwave Technology, ... CA, April, 2003, pgs. 1785-1794. Selected as one of the best papers of Infocom 2003 for fast track publication in IEEE/ACM Transactions on Networking.** 53. ...

  13. IoT: Networking Technologies and Research Challenges

    Abstract and Figures. This paper gives insight into how IoT networking is a unique com-bination of wireless personal area networks, like 6LowPAN, Zig-Bee, and the IEEE 802.15.4 protocol, which ...

  14. A comprehensive review study of cyber-attacks and cyber security

    Research paper. A comprehensive review study of cyber-attacks and cyber security; Emerging trends and recent developments ... Today's world is highly dependent on electronic technology, and protecting this data from cyber-attacks is a challenging issue. ... The Internet has created a vast global network that has generated billions of dollars ...

  15. Computer Network Security and Technology Research

    The rapid development of computer network system brings both a great convenience and new security threats for users. Network security problem generally includes network system security and data security. Specifically, it refers to the reliability of network system, confidentiality, integrity and availability of data information in the system. Network security problem exists through all the ...

  16. (PDF) 6G Wireless Communications: Future Technologies and Research

    These include data rate, capacity, latency, reliability, resources sharing, and energy/bit. To meet these challenging demands, research is focusing on 6G wireless communications enabling different ...

  17. (PDF) Research Paper on Future of 5G Wireless System

    Dr Pooja Nayak S. This paper discusses 5G wireless networks and their upcoming challenges, such as increasing network capacity to support a large number of devices running applications that ...

  18. network virtualization Latest Research Papers

    In light of extending the IoT domain capabilities' with Cloud-based mechanisms akin to the IaaS Cloud model, network virtualization is a fundamental enabler of infrastructure-oriented IoT deployments. Indeed, an IoT deployment without networking resilience and adaptability makes it unsuitable to meet user-level demands and services ...

  19. A review paper on wireless sensor network techniques in Internet of

    Thus, a review is presented in this paper with specific attention to aspects of wireless networking for the preservation of energy and aggregation of data. 2. Role of IoT in WSN. Significant classification opinions and surveys of WSN and IoT-based energy-saving technologies have been supported by several research papers and studies.

  20. network security Latest Research Papers

    Penetration testing is an active network intrusion detection technology, which plays an indispensable role in protecting the security of the system. This paper mainly introduces the principle of penetration testing, summarizes the current cutting-edge penetration testing technology, and looks forward to its development.

  21. Toward Scalable Docker-Based Emulations of Blockchain ...

    Blockchain, like any other complex technology, needs a strong testing methodology to support its evolution in both research and development contexts. Setting up meaningful tests for permissionless blockchain technology is a notoriously complex task for several reasons: software is complex, large number of nodes are involved, network is non ideal, etc. Developers usually adopt small virtual ...

  22. (PDF) ADVANCES IN NETWORK SECURITY: A COMPREHENSIVE ...

    The report proposes new research directions to advance research. This paper discusses network security for secure data communication. ... and Technology, ... is a review of papers with keywords ...

  23. IEEE Network

    IEEE Network. null | IEEE Xplore. Need Help? US & Canada: +1 800 678 4333 Worldwide: +1 732 981 0060 Contact & Support

  24. Research and application of multimedia compression technology in online

    Traditional physical education teaching methods are facing increasingly severe challenges. This article studies how to use network technology more reasonably in college physical education teaching and better serve college physical education teaching. The purpose is to comprehensively analyze the requirements of physical education teaching and teaching management in universities, and to plan in ...

  25. Introducing Gemini 1.5, Google's next-generation AI model

    Gemini 1.5 delivers dramatically enhanced performance. It represents a step change in our approach, building upon research and engineering innovations across nearly every part of our foundation model development and infrastructure. This includes making Gemini 1.5 more efficient to train and serve, with a new Mixture-of-Experts (MoE) architecture.

  26. Stanford Medicine study identifies distinct brain organization patterns

    The lead authors are senior research scientist Srikanth Ryali, PhD, and academic staff researcher Yuan Zhang, PhD. "Hotspots" that most helped the model distinguish male brains from female ones include the default mode network, a brain system that helps us process self-referential information, and the striatum and limbic network, which are ...

  27. (PDF) 5G Technology

    PDF | 5G is the latest generation of radio networks and network technology to be introduced from 2020. Manufacturers of global network hardware and... | Find, read and cite all the research you ...

  28. OpenAI teases an amazing new generative video model called Sora

    OpenAI has built a striking new generative video model called Sora that can take a short text description and turn it into a detailed, high-definition film clip up to a minute long.. Based on four ...