Yixuan Wu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Yixuan Wu has a total of 32 co-authors. Here are the top co-authors by number of shared publications:
- Laisen Nie
School of Electronics and Information, Northwestern Polytechnical University, Xian, China
7 shared publications - Zhaolong Ning
School of Communications and Information Engineering, Chongqing University of Posts and Telecommunications, Chongqing, China
4 shared publications Feng Yang
3 shared publications- Lin Yang
Systems Engineering Research Institute, Academy of Military Sciences, Beijing, China
3 shared publications Shengtao Li
School of Information Science and Engineering, Shandong Normal University, Jinan, China
3 shared publicationsLin Yang
2 shared publications
Latest Publications
- Open Access
- 14 August 2026
- Published by Elsevier in SSRN Electronic Journal
Dual-Deviation-Aware Heterogeneous Expert Fusion for Interpretable Insider Threat Detection
Insider threats rank among the most damaging yet elusive security risks: malicious intent rarely surfaces directly, manifesting instead as weak, fragmented traces dispersed across heterogeneous behavioral logs. Most existing approaches, however, collapse multi-source activities into a unified representation processed by a monolithic predictor, thereby blurring source-specific behavioral semantics and biasing learning toward prevalent attack patterns. Moreover, abnormality is typically assessed solely against a user's own history, a single reference frame that overlooks malicious behavior masked by persistently elevated personal baselines. To address these limitations, we propose the Heterogeneous Channel–Sequence–Graph Mixture-of-Experts (HCSG-MoE), a dual-deviation-aware framework for interpretable insider threat detection. Its core idea is to characterize each user-day relative to two complementary behavioral references—individual history and organizational peers—while enabling heterogeneous experts to specialize in distinct sources of threat evidence. Specifically, raw behavioral features are augmented with self-relative and peer-relative deviations, after which five channel-specific experts, a temporal sequence expert, and an organizational graph expert encode source-specific, sequential, and relational evidence, respectively. A dense gating network then fuses the expert outputs through sample-adaptive weighting, yielding both the final threat prediction and an expert-level attribution that summarizes the relative emphasis assigned to each evidence source for a given alert. Under a leakage-controlled, user-disjoint evaluation protocol, HCSG-MoE equipped with an optional tree-fusion expert achieves PR-AUC values of 0.953 and 0.974 on CERT r4.2 and CERT r5.2, respectively, while the core model maintains strong recall at stringent low-FPR operating points. Peer-relative deviations further expose high-baseline anomalies that remain invisible to purely self-relative modeling.
- 14 August 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessDual-Deviation-Aware Heterogeneous Expert Fusion for Interpretable Insider Threat Detection
Insider threats rank among the most damaging yet elusive security risks: malicious intent rarely surfaces directly, manifesting instead as weak, fragmented traces dispersed across heterogeneous behavioral logs. Most existing approaches, however, collapse multi-source activities into a unified representation processed by a monolithic predictor, thereby blurring source-specific behavioral semantics and biasing learning toward prevalent attack patterns. Moreover, abnormality is typically assessed solely against a user's own history, a single reference frame that overlooks malicious behavior masked by persistently elevated personal baselines. To address these limitations, we propose the Heterogeneous Channel–Sequence–Graph Mixture-of-Experts (HCSG-MoE), a dual-deviation-aware framework for interpretable insider threat detection. Its core idea is to characterize each user-day relative to two complementary behavioral references—individual history and organizational peers—while enabling heterogeneous experts to specialize in distinct sources of threat evidence. Specifically, raw behavioral features are augmented with self-relative and peer-relative deviations, after which five channel-specific experts, a temporal sequence expert, and an organizational graph expert encode source-specific, sequential, and relational evidence, respectively. A dense gating network then fuses the expert outputs through sample-adaptive weighting, yielding both the final threat prediction and an expert-level attribution that summarizes the relative emphasis assigned to each evidence source for a given alert. Under a leakage-controlled, user-disjoint evaluation protocol, HCSG-MoE equipped with an optional tree-fusion expert achieves PR-AUC values of 0.953 and 0.974 on CERT r4.2 and CERT r5.2, respectively, while the core model maintains strong recall at stringent low-FPR operating points. Peer-relative deviations further expose high-baseline anomalies that remain invisible to purely self-relative modeling.
- 11 April 2025
- Published by Elsevier in Computers & Security
- Vol. 155
IR-IDS: A network intrusion detection method based on causal feature selection and explainable model optimization
- 11 April 2025
- Published by Elsevier in Computers & Security
- Vol. 155
IR-IDS: A network intrusion detection method based on causal feature selection and explainable model optimization
- 5 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 20 (15566013) , 1965-1980
Intrusion Detection for Internet of Things: An Anchor Graph Clustering Approach
Intrusion detection systems are a crucial technique for securing the Internet of Things (IoT) from malicious attacks. Additionally, due to the continuous emergence of new vulnerabilities and unknown attack types, only a small number of attack samples in the IoT environments can be captured for analysis. In this work, we introduce an anchor graph clustering (AGC) method for intrusion detection to address the challenge of limited labeled samples in the IoT environments. AGC initially transforms the raw data into the embedding space to obtain more representative anchors. Then, AGC unifies anchor graph construction, anchor graph learning, and graph clustering into a unified framework, solving the resulting optimization problem through an iterative solution algorithm. Finally, AGC leverages the powerful analytical capabilities of graph learning to achieve fine-grained classification of low-quality labels. Experimental results on both real and synthetic datasets confirm that AGC can identify intrusions with high precision, while also being time-efficient in detection.
- 5 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 20 (15566013) , 1965-1980
Intrusion Detection for Internet of Things: An Anchor Graph Clustering Approach
Intrusion detection systems are a crucial technique for securing the Internet of Things (IoT) from malicious attacks. Additionally, due to the continuous emergence of new vulnerabilities and unknown attack types, only a small number of attack samples in the IoT environments can be captured for analysis. In this work, we introduce an anchor graph clustering (AGC) method for intrusion detection to address the challenge of limited labeled samples in the IoT environments. AGC initially transforms the raw data into the embedding space to obtain more representative anchors. Then, AGC unifies anchor graph construction, anchor graph learning, and graph clustering into a unified framework, solving the resulting optimization problem through an iterative solution algorithm. Finally, AGC leverages the powerful analytical capabilities of graph learning to achieve fine-grained classification of low-quality labels. Experimental results on both real and synthetic datasets confirm that AGC can identify intrusions with high precision, while also being time-efficient in detection.
- 1 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (12) , 20970-20982
Intrusion Detection for Unmanned Aerial Vehicles Security: A Tiny Machine Learning Model
Unmanned aerial vehicles (UAVs) are vulnerable to network attacks. Designing an effective intrusion detection system (IDS) for UAVs is crucial. However, UAVs have limited computing resources and need to deal with massive amounts of network data, which further increases the difficulty of detection. Moreover, most existing IDSs have large parameters. In this study, we develop a tiny machine learning-based IDS to solve the above issue. We first establish an improved fuzzy rough set (FRS) model based on adaptive neighborhoods. Then, using the proposed FRS model, we employ a feature selection (FS) method to select optimal features and reduce the overall computational cost of the IDS. Furthermore, we proposed a tiny intrusion detection model that attains high-precision detection via shallow deep learning. Additionally, the proposed method can address intrusion detection problems in scenarios with partial data missing. According to the evaluations, the proposed method can effectively address intrusion detection in UAVs.
- 1 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (12) , 20970-20982
Intrusion Detection for Unmanned Aerial Vehicles Security: A Tiny Machine Learning Model
Unmanned aerial vehicles (UAVs) are vulnerable to network attacks. Designing an effective intrusion detection system (IDS) for UAVs is crucial. However, UAVs have limited computing resources and need to deal with massive amounts of network data, which further increases the difficulty of detection. Moreover, most existing IDSs have large parameters. In this study, we develop a tiny machine learning-based IDS to solve the above issue. We first establish an improved fuzzy rough set (FRS) model based on adaptive neighborhoods. Then, using the proposed FRS model, we employ a feature selection (FS) method to select optimal features and reduce the overall computational cost of the IDS. Furthermore, we proposed a tiny intrusion detection model that attains high-precision detection via shallow deep learning. Additionally, the proposed method can address intrusion detection problems in scenarios with partial data missing. According to the evaluations, the proposed method can effectively address intrusion detection in UAVs.
- 10 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Consumer Electronics
- Vol. 70 (1) , 2004-2017
Incremental Update Intrusion Detection for Industry 5.0 Security: A Graph Attention Network-Enabled Approach
The Industry 5.0 (I5) incorporates numerous emerging technologies that enable the perception, management, interaction, and control of real physical objects. As cyber attacks are becoming more intelligent and accessible, I5 security issues are gaining increasing prominence. An intrusion detection system plays an important role in network security by detecting and identifying various security threats. However, owing to the limited computing and storage resources of users, it is difficult for I5 nodes to support large-scale network data collection and simultaneous transmission of the collected data. To address the above problems, this study investigates an intelligent intrusion detection technology for I5 security and develops a low-loss intrusion detection algorithm based on abnormal attacks. We first establish an incremental update feature selection method, which realises feature selection using a fuzzy rough set and a fuzzy knowledge distance. Subsequently, to reduce the decision loss of the selected features, we implement feature correlation using the fuzzy knowledge distance. Based on the selected features, we achieve incremental update intrusion detection using a graph attention network and a random forest. Simultaneously, we prove the convergence of the designed graph attention network in a specific scene. Finally, the experimental results show that our designed method has higher accuracy than existing methods.
- 10 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Consumer Electronics
- Vol. 70 (1) , 2004-2017
Incremental Update Intrusion Detection for Industry 5.0 Security: A Graph Attention Network-Enabled Approach
The Industry 5.0 (I5) incorporates numerous emerging technologies that enable the perception, management, interaction, and control of real physical objects. As cyber attacks are becoming more intelligent and accessible, I5 security issues are gaining increasing prominence. An intrusion detection system plays an important role in network security by detecting and identifying various security threats. However, owing to the limited computing and storage resources of users, it is difficult for I5 nodes to support large-scale network data collection and simultaneous transmission of the collected data. To address the above problems, this study investigates an intelligent intrusion detection technology for I5 security and develops a low-loss intrusion detection algorithm based on abnormal attacks. We first establish an incremental update feature selection method, which realises feature selection using a fuzzy rough set and a fuzzy knowledge distance. Subsequently, to reduce the decision loss of the selected features, we implement feature correlation using the fuzzy knowledge distance. Based on the selected features, we achieve incremental update intrusion detection using a graph attention network and a random forest. Simultaneously, we prove the convergence of the designed graph attention network in a specific scene. Finally, the experimental results show that our designed method has higher accuracy than existing methods.
- 11 January 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 18 (11) , 7475-7483
A Multitask Learning-Based Network Traffic Prediction Approach for SDN-Enabled Industrial Internet of Things
With the rapid advance of industrial Internet of Things (IIoT), to provide flexible access for various infrastructures and applications, software-defined networks (SDNs) have been involved in constructing current IIoT networks. To improve the quality of services of industrial applications, network traffic prediction has become an important research direction, which is beneficial for network management and security. Unfortunately, the traffic flows of the SDN-enabled IIoT network contain a large number of irregular fluctuations, which makes network traffic prediction difficult. In this article, we propose an algorithm based on multitask learning to predict network traffic according to the spatial and temporal features of network traffic. Our proposed approach can effectively obtain network traffic predictors according to the evaluations by implementing it on real networks.
- 11 January 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 18 (11) , 7475-7483
A Multitask Learning-Based Network Traffic Prediction Approach for SDN-Enabled Industrial Internet of Things
With the rapid advance of industrial Internet of Things (IIoT), to provide flexible access for various infrastructures and applications, software-defined networks (SDNs) have been involved in constructing current IIoT networks. To improve the quality of services of industrial applications, network traffic prediction has become an important research direction, which is beneficial for network management and security. Unfortunately, the traffic flows of the SDN-enabled IIoT network contain a large number of irregular fluctuations, which makes network traffic prediction difficult. In this article, we propose an algorithm based on multitask learning to predict network traffic according to the spatial and temporal features of network traffic. Our proposed approach can effectively obtain network traffic predictors according to the evaluations by implementing it on real networks.
- 13 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 10 (4) , 3094-3106
Intelligent Intrusion Detection for Internet of Things Security: A Deep Convolutional Generative Adversarial Network-Enabled Approach
With the rapid advance of Internet of Things (IoTs), it is difficult for cloud-centric computing to meet the requirements of low latency and ease of use. As an open and distributed system, edge computing integrates computing, networking, storage, and applications. It provides intelligent services on the edge of an IoT. The edge network is comprised of various wireless and wired networks, and the computing and storage resources of edge nodes are limited. These conditions make the edge network expose to a variety of cyber attacks. Additionally, it is difficult for an IoT edge node to support large-scale network data collection and detection for IoT security. Although big data-enabled intrusion detection algorithms can ensure the high accuracy of intrusion detection systems, it is stressful for resource-limited edge nodes to implement those algorithms in IoTs. Motivated by these challenges, we propose an intelligent intrusion detection algorithm implemented by big data mining based on fuzzy rough set, Generative Adversarial Network (GAN), and Convolutional Neural Network (CNN). In our method, we first propose a fuzzy rough set-based algorithm to perform feature selection for big data via IoTs. Then, we take advantage of the efficient feature extraction capabilities of CNN for implementing intrusion detection based on selected features. Furthermore, after combining CNN and GAN, we propose an intelligent algorithm to realize intrusion detection in a variety of scenarios. Finally, the proposed method is compared with existing methods for evaluation. Simulation results show that our method has up to 4% higher accuracy than existing methods.
- 13 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 10 (4) , 3094-3106
Intelligent Intrusion Detection for Internet of Things Security: A Deep Convolutional Generative Adversarial Network-Enabled Approach
With the rapid advance of Internet of Things (IoTs), it is difficult for cloud-centric computing to meet the requirements of low latency and ease of use. As an open and distributed system, edge computing integrates computing, networking, storage, and applications. It provides intelligent services on the edge of an IoT. The edge network is comprised of various wireless and wired networks, and the computing and storage resources of edge nodes are limited. These conditions make the edge network expose to a variety of cyber attacks. Additionally, it is difficult for an IoT edge node to support large-scale network data collection and detection for IoT security. Although big data-enabled intrusion detection algorithms can ensure the high accuracy of intrusion detection systems, it is stressful for resource-limited edge nodes to implement those algorithms in IoTs. Motivated by these challenges, we propose an intelligent intrusion detection algorithm implemented by big data mining based on fuzzy rough set, Generative Adversarial Network (GAN), and Convolutional Neural Network (CNN). In our method, we first propose a fuzzy rough set-based algorithm to perform feature selection for big data via IoTs. Then, we take advantage of the efficient feature extraction capabilities of CNN for implementing intrusion detection based on selected features. Furthermore, after combining CNN and GAN, we propose an intelligent algorithm to realize intrusion detection in a variety of scenarios. Finally, the proposed method is compared with existing methods for evaluation. Simulation results show that our method has up to 4% higher accuracy than existing methods.
- 16 April 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Green Communications and Networking
- Vol. 5 (2) , 778-788
Intrusion Detection in Green Internet of Things: A Deep Deterministic Policy Gradient-Based Algorithm
Internet of Things (IoT) is an expanded application of Internet, which are used to provide various services for users. Up to now, IoTs have received more attention, because it can provide ubiquitous connectivity for various devices. With the development of IoTs, its security has become a main issue. Attackers use various techniques to implement cyber attacks for the IoT, which threats the users' privacy seriously. As a security mechanism, intrusion detection techniques can detect various illicit behaviors before attackers invade the network. An intrusion detection system can implement effective defense functions to keep the network away from attacks. This paper proposes an intrusion detection algorithm based on deep reinforcement learning, which pursued the trends of traffic flows by extracting statistical features of prior network traffic for traffic prediction at first. Then, we use traffic predictors to employ intrusion detection. The evaluations verify the effectiveness of our algorithm in detecting Distributed Denial of Service attacks (DDoS).
- 16 April 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Green Communications and Networking
- Vol. 5 (2) , 778-788
Intrusion Detection in Green Internet of Things: A Deep Deterministic Policy Gradient-Based Algorithm
Internet of Things (IoT) is an expanded application of Internet, which are used to provide various services for users. Up to now, IoTs have received more attention, because it can provide ubiquitous connectivity for various devices. With the development of IoTs, its security has become a main issue. Attackers use various techniques to implement cyber attacks for the IoT, which threats the users' privacy seriously. As a security mechanism, intrusion detection techniques can detect various illicit behaviors before attackers invade the network. An intrusion detection system can implement effective defense functions to keep the network away from attacks. This paper proposes an intrusion detection algorithm based on deep reinforcement learning, which pursued the trends of traffic flows by extracting statistical features of prior network traffic for traffic prediction at first. Then, we use traffic predictors to employ intrusion detection. The evaluations verify the effectiveness of our algorithm in detecting Distributed Denial of Service attacks (DDoS).
- Open Access
- 22 March 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Social Systems
- Vol. 9 (1) , 134-145
Intrusion Detection for Secure Social Internet of Things Based on Collaborative Edge Computing: A Generative Adversarial Network-Based Approach
The Social Internet of Things (SIoT) now penetrates our daily lives. As a strategy to alleviate the escalation of resource congestion, collaborative edge computing (CEC) has become a new paradigm for solving the needs of the Internet of Things (IoT). CEC can provide computing, storage, and network connection resources for remote devices. Because the edge network is closer to the connected devices, it involves a large amount of users’ privacy. This also makes edge networks face more and more security issues, such as Denial-of-Service (DoS) attacks, unauthorized access, packet sniffing, and man-in-the-middle attacks. To combat these issues and enhance the security of edge networks, we propose a deep learning-based intrusion detection algorithm. Based on the generative adversarial network (GAN), we designed a powerful intrusion detection method. Our intrusion detection method includes three phases. First, we use the feature selection module to process the collaborative edge network traffic. Second, a deep learning architecture based on GAN is designed for intrusion detection aiming at a single attack. Finally, we propose a new intrusion detection model by combining several intrusion detection models that aim at a single attack. Intrusion detection aiming at multiple attacks is realized through the designed GAN-based deep learning architecture. Besides, we provide a comprehensive evaluation to verify the effectiveness of the proposed method.
- 22 March 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Social Systems
- Vol. 9 (1) , 134-145
Open AccessIntrusion Detection for Secure Social Internet of Things Based on Collaborative Edge Computing: A Generative Adversarial Network-Based Approach
The Social Internet of Things (SIoT) now penetrates our daily lives. As a strategy to alleviate the escalation of resource congestion, collaborative edge computing (CEC) has become a new paradigm for solving the needs of the Internet of Things (IoT). CEC can provide computing, storage, and network connection resources for remote devices. Because the edge network is closer to the connected devices, it involves a large amount of users’ privacy. This also makes edge networks face more and more security issues, such as Denial-of-Service (DoS) attacks, unauthorized access, packet sniffing, and man-in-the-middle attacks. To combat these issues and enhance the security of edge networks, we propose a deep learning-based intrusion detection algorithm. Based on the generative adversarial network (GAN), we designed a powerful intrusion detection method. Our intrusion detection method includes three phases. First, we use the feature selection module to process the collaborative edge network traffic. Second, a deep learning architecture based on GAN is designed for intrusion detection aiming at a single attack. Finally, we propose a new intrusion detection model by combining several intrusion detection models that aim at a single attack. Intrusion detection aiming at multiple attacks is realized through the designed GAN-based deep learning architecture. Besides, we provide a comprehensive evaluation to verify the effectiveness of the proposed method.
- Open Access
- 6 December 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 7, 177954-177964
Anomaly Detection Based on Spatio-Temporal and Sparse Features of Network Traffic in VANETs
Vehicular Ad-Hoc Networks (VANETs) have received a great attention recently due to their potential and various applications. However, the initial phase of the VANET has many research challenges that need to be addressed, such as the issues of security and privacy protection caused by the openness of wireless communication networks among the city-wide applied regions. Specially, anomaly detection for a VANET has become a challenging problem, due to the changes in the scenario of VANETs comparing with traditional wireless networks. Motivated by this issue, we focus on the problem of anomaly detection in VANETs, and propose an effective anomaly detection approach based on the convolutional neural network in this paper. The proposed approach takes into account the spatio-temporal and sparse features of VANET traffic, and it uses a convolutional neural network architecture and a loss function based on Mahalanobis distance to extract and estimate the traffic matrix. Then, reinforcement learning is used to implement anomaly detection. Furthermore, a comprehensive assessment is provided to validate the proposed approach, which illustrates the effectiveness of this approach.
- 6 December 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 7, 177954-177964
Open AccessAnomaly Detection Based on Spatio-Temporal and Sparse Features of Network Traffic in VANETs
Vehicular Ad-Hoc Networks (VANETs) have received a great attention recently due to their potential and various applications. However, the initial phase of the VANET has many research challenges that need to be addressed, such as the issues of security and privacy protection caused by the openness of wireless communication networks among the city-wide applied regions. Specially, anomaly detection for a VANET has become a challenging problem, due to the changes in the scenario of VANETs comparing with traditional wireless networks. Motivated by this issue, we focus on the problem of anomaly detection in VANETs, and propose an effective anomaly detection approach based on the convolutional neural network in this paper. The proposed approach takes into account the spatio-temporal and sparse features of VANET traffic, and it uses a convolutional neural network architecture and a loss function based on Mahalanobis distance to extract and estimate the traffic matrix. Then, reinforcement learning is used to implement anomaly detection. Furthermore, a comprehensive assessment is provided to validate the proposed approach, which illustrates the effectiveness of this approach.