Mahdi Jafari Siavoshani
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Mahdi Jafari Siavoshani has a total of 49 co-authors. Here are the top co-authors by number of shared publications:
- Seyed Pooya Shariatpanahi
School of Electrical and Computer Engineering, College of Engineering, University of Tehran, Tehran, Iran
9 shared publications - Ali Pourmiri
School of Computer Science Institute For Research in Fundamental Sciences (IPM) Tehran Iran
4 shared publications - Amir Hossein Jahangir
Department of Computer Engineering, Sharif University of Technology, Tehran, Iran
3 shared publications - Amir Ali Gatmiri
Information, Network, and Learning Lab (INL) Computer Science and Engineering Department, Sharif University of Technology, Tehran, Iran
2 shared publications - Arash Jamshidi
Department of Computer Engineering Department of Computer Engineering Department of Computer Engineering Department of Computer Engineering Institute for Research in Fundamental Sciences (IPM Sharif University of Technology Sharif University of Technology Sharif University of Technology Sharif University of Technology
2 shared publications - Behnam Bahrak
Tehran Institute for Advanced Studies, Khatam University, Tehran 19916-33357, Iran
2 shared publications
Latest Publications
- 17 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. PP  (00189545) , 1-6
Meta Reinforcement Learning Empowered Orchestration of SIM and RIS for Downlink Multiuser Communications
The concept of Stacked Intelligent Metasurface (SIM) stands out as an avant-garde signal processing paradigm, empowering the instantaneous manipulation of electromagnetic waves at the speed of light. This paper explores the performance of a wireless system, where a reconfigurable intelligent surface (RIS) assists the communication between a SIM-deployed base station (BS) and downlink terminals. In contrast to conventional communication systems, the downlink multiuser beamforming is performed in the wave domain by appropriately designing the electromagnetic response of the SIM. The performance of this system is evaluated by formulating a resource allocation optimization problem, aimed at maximizing the system data rate, constrained by the quality-of-service (QoS) of terminals and the transmit power budget of the BS. Owing to its intricately coupled variables and non-convex characteristics, we adeptly reformulate this problem into the form of a Markov decision process (MDP), which mirrors its dynamicity. Subsequently, we deploy a twin delayed deep deterministic policy gradient (TD3) agent to optimize its decision variables, namely, the transmit power at the BS, the electromagnetic response at the SIM, and the reflection coefficient at the RIS, in a holistic manner. Furthermore, considering the dynamic mobility of terminals in real-time scenarios, we augment the adaptability of the trained TD3 model via meta-learning. Simulation findings reveal that incorporating the RISs into SIM systems leads to a considerable improvement—reaching as high as 20%—in the average data rate experienced by users, particularly when the BS transmit power budget lies within a moderate range.
- 17 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. PP  (00189545) , 1-6
Meta Reinforcement Learning Empowered Orchestration of SIM and RIS for Downlink Multiuser Communications
The concept of Stacked Intelligent Metasurface (SIM) stands out as an avant-garde signal processing paradigm, empowering the instantaneous manipulation of electromagnetic waves at the speed of light. This paper explores the performance of a wireless system, where a reconfigurable intelligent surface (RIS) assists the communication between a SIM-deployed base station (BS) and downlink terminals. In contrast to conventional communication systems, the downlink multiuser beamforming is performed in the wave domain by appropriately designing the electromagnetic response of the SIM. The performance of this system is evaluated by formulating a resource allocation optimization problem, aimed at maximizing the system data rate, constrained by the quality-of-service (QoS) of terminals and the transmit power budget of the BS. Owing to its intricately coupled variables and non-convex characteristics, we adeptly reformulate this problem into the form of a Markov decision process (MDP), which mirrors its dynamicity. Subsequently, we deploy a twin delayed deep deterministic policy gradient (TD3) agent to optimize its decision variables, namely, the transmit power at the BS, the electromagnetic response at the SIM, and the reflection coefficient at the RIS, in a holistic manner. Furthermore, considering the dynamic mobility of terminals in real-time scenarios, we augment the adaptability of the trained TD3 model via meta-learning. Simulation findings reveal that incorporating the RISs into SIM systems leads to a considerable improvement—reaching as high as 20%—in the average data rate experienced by users, particularly when the BS transmit power budget lies within a moderate range.
- Open Access
A Hybrid TGN-SEAL Model for Dynamic Graph Link Prediction
Predicting links in sparse, continuously evolving networks is a central challenge in network science. Conventional heuristic methods and deep learning models, including Graph Neural Networks (GNNs), are typically designed for static graphs and thus struggle to capture temporal dependencies. Snapshot-based techniques partially address this issue but often encounter data sparsity and class imbalance, particularly in networks with transient interactions such as telecommunication call detail records (CDRs). Temporal Graph Networks (TGNs) model dynamic graphs by updating node embeddings over time; however, their predictive accuracy under sparse conditions remains limited. In this study, we improve the TGN framework by extracting enclosing subgraphs around candidate links, enabling the model to jointly learn structural and temporal information. Experiments on a sparse CDR dataset show that our approach increases average precision by 2.6% over standard TGNs, demonstrating the advantages of integrating local topology for robust link prediction in dynamic networks.
Open AccessA Hybrid TGN-SEAL Model for Dynamic Graph Link Prediction
Predicting links in sparse, continuously evolving networks is a central challenge in network science. Conventional heuristic methods and deep learning models, including Graph Neural Networks (GNNs), are typically designed for static graphs and thus struggle to capture temporal dependencies. Snapshot-based techniques partially address this issue but often encounter data sparsity and class imbalance, particularly in networks with transient interactions such as telecommunication call detail records (CDRs). Temporal Graph Networks (TGNs) model dynamic graphs by updating node embeddings over time; however, their predictive accuracy under sparse conditions remains limited. In this study, we improve the TGN framework by extracting enclosing subgraphs around candidate links, enabling the model to jointly learn structural and temporal information. Experiments on a sparse CDR dataset show that our approach increases average precision by 2.6% over standard TGNs, demonstrating the advantages of integrating local topology for robust link prediction in dynamic networks.
- Open Access
Extended Deep Submodular Functions
We introduce a novel category of set functions called Extended Deep Submodular functions (EDSFs), which are neural network-representable. EDSFs serve as an extension of Deep Submodular Functions (DSFs), inheriting crucial properties from DSFs while addressing innate limitations. It is known that DSFs can represent a limiting subset of submodular functions. In contrast, through an analysis of polymatroid properties, we establish that EDSFs possess the capability to represent all monotone submodular functions, a notable enhancement compared to DSFs. Furthermore, our findings demonstrate that EDSFs can represent any monotone set function, indicating the family of EDSFs is equivalent to the family of all monotone set functions. Additionally, we prove that EDSFs maintain the concavity inherent in DSFs when the components of the input vector are non-negative real numbers-an essential feature in certain combinatorial optimization problems. Through extensive experiments, we illustrate that EDSFs exhibit significantly lower empirical generalization error than DSFs in the learning of coverage functions. This suggests that EDSFs present a promising advancement in the representation and learning of set functions with improved generalization capabilities.
Open AccessExtended Deep Submodular Functions
We introduce a novel category of set functions called Extended Deep Submodular functions (EDSFs), which are neural network-representable. EDSFs serve as an extension of Deep Submodular Functions (DSFs), inheriting crucial properties from DSFs while addressing innate limitations. It is known that DSFs can represent a limiting subset of submodular functions. In contrast, through an analysis of polymatroid properties, we establish that EDSFs possess the capability to represent all monotone submodular functions, a notable enhancement compared to DSFs. Furthermore, our findings demonstrate that EDSFs can represent any monotone set function, indicating the family of EDSFs is equivalent to the family of all monotone set functions. Additionally, we prove that EDSFs maintain the concavity inherent in DSFs when the components of the input vector are non-negative real numbers-an essential feature in certain combinatorial optimization problems. Through extensive experiments, we illustrate that EDSFs exhibit significantly lower empirical generalization error than DSFs in the learning of coverage functions. This suggests that EDSFs present a promising advancement in the representation and learning of set functions with improved generalization capabilities.
- Open Access
Differentially Private Machine Learning-powered Combinatorial Auction Design
We present a new approach to machine learning-powered combinatorial auctions, which is based on the principles of Differential Privacy. Our methodology guarantees that the auction mechanism is truthful, meaning that rational bidders have the incentive to reveal their true valuation functions. We achieve this by inducing truthfulness in the auction dynamics, ensuring that bidders consistently provide accurate information about their valuation functions. Our method not only ensures truthfulness but also preserves the efficiency of the original auction. This means that if the initial auction outputs an allocation with high social welfare, our modified truthful version of the auction will also achieve high social welfare. We use techniques from Differential Privacy, such as the Exponential Mechanism, to achieve these results. Additionally, we examine the application of differential privacy in auctions across both asymptotic and non-asymptotic regimes.
Open AccessDifferentially Private Machine Learning-powered Combinatorial Auction Design
We present a new approach to machine learning-powered combinatorial auctions, which is based on the principles of Differential Privacy. Our methodology guarantees that the auction mechanism is truthful, meaning that rational bidders have the incentive to reveal their true valuation functions. We achieve this by inducing truthfulness in the auction dynamics, ensuring that bidders consistently provide accurate information about their valuation functions. Our method not only ensures truthfulness but also preserves the efficiency of the original auction. This means that if the initial auction outputs an allocation with high social welfare, our modified truthful version of the auction will also achieve high social welfare. We use techniques from Differential Privacy, such as the Exponential Mechanism, to achieve these results. Additionally, we examine the application of differential privacy in auctions across both asymptotic and non-asymptotic regimes.
- Open Access
- 1 May 2024
- Published by Springer Nature in Cybersecurity
- Vol. 7  (1) , 1-25
A multi-agent adaptive deep learning framework for online intrusion detection
The network security analyzers use intrusion detection systems (IDSes) to distinguish malicious traffic from benign ones. The deep learning-based (DL-based) IDSes are proposed to auto-extract high-level features and eliminate the time-consuming and costly signature extraction process. However, this new generation of IDSes still needs to overcome a number of challenges to be employed in practical environments. One of the main issues of an applicable IDS is facing traffic concept drift, which manifests itself as new (i.e. , zero-day) attacks, in addition to the changing behavior of benign users/applications. Furthermore, a practical DL-based IDS needs to be conformed to a distributed (i.e. , multi-sensor) architecture in order to yield more accurate detections, create a collective attack knowledge based on the observations of different sensors, and also handle big data challenges for supporting high throughput networks. This paper proposes a novel multi-agent network intrusion detection framework to address the above shortcomings, considering a more practical scenario (i.e., online adaptable IDSes). This framework employs continual deep anomaly detectors for adapting each agent to the changing attack/benign patterns in its local traffic. In addition, a federated learning approach is proposed for sharing and exchanging local knowledge between different agents. Furthermore, the proposed framework implements sequential packet labeling for each flow, which provides an attack probability score for the flow by gradually observing each flow packet and updating its estimation. We evaluate the proposed framework by employing different deep models (including CNN-based and LSTM-based) over the CIC-IDS2017 and CSE-CIC-IDS2018 datasets. Through extensive evaluations and experiments, we show that the proposed distributed framework is well adapted to the traffic concept drift. More precisely, our results indicate that the CNN-based models are well suited for continually adapting to the traffic concept drift (i.e. , achieving an average detection rate of above 95% while needing just 128 new flows for the updating phase), and the LSTM-based models are a good candidate for sequential packet labeling in practical online IDSes (i.e. , detecting intrusions by just observing their first 15 packets).
- 1 May 2024
- Published by Springer Nature in Cybersecurity
- Vol. 7  (1) , 1-25
Open AccessA multi-agent adaptive deep learning framework for online intrusion detection
The network security analyzers use intrusion detection systems (IDSes) to distinguish malicious traffic from benign ones. The deep learning-based (DL-based) IDSes are proposed to auto-extract high-level features and eliminate the time-consuming and costly signature extraction process. However, this new generation of IDSes still needs to overcome a number of challenges to be employed in practical environments. One of the main issues of an applicable IDS is facing traffic concept drift, which manifests itself as new (i.e. , zero-day) attacks, in addition to the changing behavior of benign users/applications. Furthermore, a practical DL-based IDS needs to be conformed to a distributed (i.e. , multi-sensor) architecture in order to yield more accurate detections, create a collective attack knowledge based on the observations of different sensors, and also handle big data challenges for supporting high throughput networks. This paper proposes a novel multi-agent network intrusion detection framework to address the above shortcomings, considering a more practical scenario (i.e., online adaptable IDSes). This framework employs continual deep anomaly detectors for adapting each agent to the changing attack/benign patterns in its local traffic. In addition, a federated learning approach is proposed for sharing and exchanging local knowledge between different agents. Furthermore, the proposed framework implements sequential packet labeling for each flow, which provides an attack probability score for the flow by gradually observing each flow packet and updating its estimation. We evaluate the proposed framework by employing different deep models (including CNN-based and LSTM-based) over the CIC-IDS2017 and CSE-CIC-IDS2018 datasets. Through extensive evaluations and experiments, we show that the proposed distributed framework is well adapted to the traffic concept drift. More precisely, our results indicate that the CNN-based models are well suited for continually adapting to the traffic concept drift (i.e. , achieving an average detection rate of above 95% while needing just 128 new flows for the updating phase), and the LSTM-based models are a good candidate for sequential packet labeling in practical online IDSes (i.e. , detecting intrusions by just observing their first 15 packets).
- 28 March 2023
- Published by Springer Nature in Soft Computing
- Vol. 27  (11) , 7191-7208
Machine learning interpretability meets TLS fingerprinting
- 28 March 2023
- Published by Springer Nature in Soft Computing
- Vol. 27  (11) , 7191-7208
Machine learning interpretability meets TLS fingerprinting
- Open Access
A Multi-Agent Adaptive Deep Learning Framework for Online Intrusion Detection
The network security analyzers use intrusion detection systems (IDSes) to distinguish malicious traffic from benign ones. The deep learning-based IDSes are proposed to auto-extract high-level features and eliminate the time-consuming and costly signature extraction process. However, this new generation of IDSes still suffers from a number of challenges. One of the main issues of an IDS is facing traffic concept drift which manifests itself as new (i.e., zero-day) attacks, in addition to the changing behavior of benign users/applications. Furthermore, a practical DL-based IDS needs to be conformed to a distributed architecture to handle big data challenges. We propose a framework for adapting DL-based models to the changing attack/benign traffic behaviors, considering a more practical scenario (i.e., online adaptable IDSes). This framework employs continual deep anomaly detectors in addition to the federated learning approach to solve the above-mentioned challenges. Furthermore, the proposed framework implements sequential packet labeling for each flow, which provides an attack probability score for the flow by gradually observing each flow packet and updating its estimation. We evaluate the proposed framework by employing different deep models (including CNN-based and LSTM-based) over the CIC-IDS2017 and CSE-CIC-IDS2018 datasets. Through extensive evaluations and experiments, we show that the proposed distributed framework is well adapted to the traffic concept drift. More precisely, our results indicate that the CNN-based models are well suited for continually adapting to the traffic concept drift (i.e., achieving an average detection rate of above 95% while needing just 128 new flows for the updating phase), and the LSTM-based models are a good candidate for sequential packet labeling in practical online IDSes (i.e., detecting intrusions by just observing their first 15 packets).
Open AccessA Multi-Agent Adaptive Deep Learning Framework for Online Intrusion Detection
The network security analyzers use intrusion detection systems (IDSes) to distinguish malicious traffic from benign ones. The deep learning-based IDSes are proposed to auto-extract high-level features and eliminate the time-consuming and costly signature extraction process. However, this new generation of IDSes still suffers from a number of challenges. One of the main issues of an IDS is facing traffic concept drift which manifests itself as new (i.e., zero-day) attacks, in addition to the changing behavior of benign users/applications. Furthermore, a practical DL-based IDS needs to be conformed to a distributed architecture to handle big data challenges. We propose a framework for adapting DL-based models to the changing attack/benign traffic behaviors, considering a more practical scenario (i.e., online adaptable IDSes). This framework employs continual deep anomaly detectors in addition to the federated learning approach to solve the above-mentioned challenges. Furthermore, the proposed framework implements sequential packet labeling for each flow, which provides an attack probability score for the flow by gradually observing each flow packet and updating its estimation. We evaluate the proposed framework by employing different deep models (including CNN-based and LSTM-based) over the CIC-IDS2017 and CSE-CIC-IDS2018 datasets. Through extensive evaluations and experiments, we show that the proposed distributed framework is well adapted to the traffic concept drift. More precisely, our results indicate that the CNN-based models are well suited for continually adapting to the traffic concept drift (i.e., achieving an average detection rate of above 95% while needing just 128 new flows for the updating phase), and the LSTM-based models are a good candidate for sequential packet labeling in practical online IDSes (i.e., detecting intrusions by just observing their first 15 packets).
- Open Access
- 26 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 113827-113838
Safe Deep Driving Behavior Detection (S3D)
The human factor is one of the most critical parameters in car accidents and even traffic occurrences. Driving style affected by human factors comprises driving events (maneuvers) and driver behaviors. Driving event detection is the fundamental step of identifying driving style and facilitates predicting potentially unsafe behaviors, preventing accidents, and imposing restrictions on high-risk drivers. This paper proposes a deep hybrid model to detect safe driver behaviors and driving events using real-time smartphone sensor signals. The ensemble of Multi-layer Perceptron, Support-Vector Machine, and Convolutional Neural Network classifiers process each driving event sample. In order to evaluate our model, we develop an Android Application to capture smartphone sensor signal data. We capture about 24000 driving data from 50 drivers. Results indicate that the fusion model performs better than each individual classifier in terms of Accuracy, False Positive Rate (FPR), and Specificity (96.75, 0.004, and 0.996). This research gives insights to Auto-mobile developers to focus on the speed and cost efficiency of smartphone driver monitoring platforms. Although some insurance and freight management companies utilize smartphones as their monitoring platforms, the market share of these use cases is meager and could improve rapidly with the promotion of new smartphones with better processing and storage.
- 26 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 113827-113838
Open AccessSafe Deep Driving Behavior Detection (S3D)
The human factor is one of the most critical parameters in car accidents and even traffic occurrences. Driving style affected by human factors comprises driving events (maneuvers) and driver behaviors. Driving event detection is the fundamental step of identifying driving style and facilitates predicting potentially unsafe behaviors, preventing accidents, and imposing restrictions on high-risk drivers. This paper proposes a deep hybrid model to detect safe driver behaviors and driving events using real-time smartphone sensor signals. The ensemble of Multi-layer Perceptron, Support-Vector Machine, and Convolutional Neural Network classifiers process each driving event sample. In order to evaluate our model, we develop an Android Application to capture smartphone sensor signal data. We capture about 24000 driving data from 50 drivers. Results indicate that the fusion model performs better than each individual classifier in terms of Accuracy, False Positive Rate (FPR), and Specificity (96.75, 0.004, and 0.996). This research gives insights to Auto-mobile developers to focus on the speed and cost efficiency of smartphone driver monitoring platforms. Although some insurance and freight management companies utilize smartphones as their monitoring platforms, the market share of these use cases is meager and could improve rapidly with the promotion of new smartphones with better processing and storage.
- 6 December 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Communications Letters
- Vol. 26  (2) , 273-276
Multi-Sender Index Coding Over Linear Networks
We consider an index coding problem in which several transmitters deliver distinct files to a number of users with minimum delay. Each user has access to a subset of other files from the library, which can be used as side information. The information sent by the transmitters experience a linear transformation before being received at the users. By benefiting from the concept of Zero-Forcing in MIMO systems, we generalize the notion of MinRank characterization and the clique cover algorithm to accommodate this generalized setting. We show that increasing the number of transmitters can substantially reduce the delivery delay.
- 6 December 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Communications Letters
- Vol. 26  (2) , 273-276
Multi-Sender Index Coding Over Linear Networks
We consider an index coding problem in which several transmitters deliver distinct files to a number of users with minimum delay. Each user has access to a subset of other files from the library, which can be used as side information. The information sent by the transmitters experience a linear transformation before being received at the users. By benefiting from the concept of Zero-Forcing in MIMO systems, we generalize the notion of MinRank characterization and the clique cover algorithm to accommodate this generalized setting. We show that increasing the number of transmitters can substantially reduce the delivery delay.
- 23 September 2021
- Published by Springer Nature in International Journal of Information Security
- Vol. 21  (3) , 547-562
A content-based deep intrusion detection system
- 23 September 2021
- Published by Springer Nature in International Journal of Information Security
- Vol. 21  (3) , 547-562
A content-based deep intrusion detection system