Yi Zhao
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
Research Subjects
Yi Zhao has a total of 85 co-authors. Here are the top co-authors by number of shared publications:
- Ke Xu
Tsinghua University, Beijing, China
10 shared publications - Qi Li
Tsinghua University, Beijing, China
6 shared publications - Liehuang Zhu
Beijing Institute of Technology
4 shared publications - Rui Zhang
Xi’an Jiaotong-Liverpool University, Suzhou, China
3 shared publications - Zijian Zhang
Beijing Institute of Technology
3 shared publications Haiyang Wang
Department of Computer Science, University of Minnesota Duluth, Duluth, MN, USA
2 shared publications
Latest Publications
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Dependable and Secure Computing
- Vol. 23 (5) , 11598-11613
Graph Unlearning for MLaaS: Toward Flexible Privacy Adjustment via Influenced Subgraph
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Dependable and Secure Computing
- Vol. 23 (5) , 11598-11613
Graph Unlearning for MLaaS: Toward Flexible Privacy Adjustment via Influenced Subgraph
- 17 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 21 (15566013) , 4460-4474
Early-Stage Detection of Encrypted Malware Traffic via Multi-Flow Temporal Graph Learning
Malware widely adopts network traffic encryption techniques to conceal malicious activities. Recent research has demonstrated the effectiveness of machine learning (ML)-, deep learning (DL)-, and pre-training-based malware traffic detection methods. However, a vast majority of these methods rely on the collected complete traffic during the malware attack. While certain methods can operate on partial traffic, their detection accuracy often significantly decreases when the available data is restricted to the extreme early stage, where information is most sparse. In this paper, we propose DawnGuard, an effective early-stage encrypted malware traffic detection framework through multi-flow temporal graph learning. Specifically, based on the temporal packet density distribution analysis, DawnGuard innovatively proposes a self-adjusting data augmentation strategy for early-stage malware traffic, which can force the model to focus on the early-stage interaction phase with more distinguishable properties. Meanwhile, considering that temporal-topological correlations among multiple flows can provide more distinguishable properties in a malware attack, we further develop a temporal graph learning framework to extract features, which can form Multi-Flow Graph Features (MGF). By utilizing MGF, Dawn-Guard implements a Vision Transformer-based detection mechanism, enabling accurate and precise encrypted malware traffic detection with early-stage traffic by capturing both local and global contextual relationships. Extensive experiments with two real-world datasets demonstrate that DawnGuard outperforms the state-of-the-art (SOTA) methods in three typical scenarios: varying early-stage time windows, imbalanced data, and unseen malware detection. Particularly, DawnGuard achieves an average F1 of 95.11%, 8.7% higher than the SOTA method, by only utilizing the first 20% loading ratio of complete traffic.
- 17 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 21 (15566013) , 4460-4474
Early-Stage Detection of Encrypted Malware Traffic via Multi-Flow Temporal Graph Learning
Malware widely adopts network traffic encryption techniques to conceal malicious activities. Recent research has demonstrated the effectiveness of machine learning (ML)-, deep learning (DL)-, and pre-training-based malware traffic detection methods. However, a vast majority of these methods rely on the collected complete traffic during the malware attack. While certain methods can operate on partial traffic, their detection accuracy often significantly decreases when the available data is restricted to the extreme early stage, where information is most sparse. In this paper, we propose DawnGuard, an effective early-stage encrypted malware traffic detection framework through multi-flow temporal graph learning. Specifically, based on the temporal packet density distribution analysis, DawnGuard innovatively proposes a self-adjusting data augmentation strategy for early-stage malware traffic, which can force the model to focus on the early-stage interaction phase with more distinguishable properties. Meanwhile, considering that temporal-topological correlations among multiple flows can provide more distinguishable properties in a malware attack, we further develop a temporal graph learning framework to extract features, which can form Multi-Flow Graph Features (MGF). By utilizing MGF, Dawn-Guard implements a Vision Transformer-based detection mechanism, enabling accurate and precise encrypted malware traffic detection with early-stage traffic by capturing both local and global contextual relationships. Extensive experiments with two real-world datasets demonstrate that DawnGuard outperforms the state-of-the-art (SOTA) methods in three typical scenarios: varying early-stage time windows, imbalanced data, and unseen malware detection. Particularly, DawnGuard achieves an average F1 of 95.11%, 8.7% higher than the SOTA method, by only utilizing the first 20% loading ratio of complete traffic.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 21, 9064-9079
Highly Reliable Anonymous Cloud Synchronization
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 21, 9064-9079
Highly Reliable Anonymous Cloud Synchronization
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 21 (15566013) , 3944-3959
Byzantine-Robust Asynchronous Federated Learning via Feature Fingerprinting
Asynchronous federated learning (AFL) accelerates collaborative training across heterogeneous devices compared to synchronous federated learning, but increases vulnerability to Byzantine attacks due to its asynchronous aggregation. Existing defenses rely on parametric similarity between models and temporal consistency of updates, which are compromised by data and device heterogeneity, leading to ineffective robustness. To address this limitation, we propose Belisa, a Byzantine-robust AFL framework that enhances fidelity, robustness, and efficiency under heterogeneous scenarios. Belisa introduces novel discrepancies between feature representations of local models to distinguish malicious models from benign ones. By leveraging a reference model trained on publicly available data, Belisa quantifies these discrepancies, referred to as feature fingerprints, and filters out malicious models through clustering. Extensive experiments on six datasets from three types of tasks under five advanced Byzantine attacks demonstrate Belisa’s superiority. Notably, Belisa consistently outperforms existing approaches across both attack and non-attack settings. Under attack scenarios, it lowers the average test error rate to 0.42× that of baseline methods. Furthermore, Belisa accelerates the aggregation process by an average of 12.3× compared to other methods. To the best of our knowledge, Belisa is the first Byzantine-robust AFL framework, which provides a broadly applicable countermeasure in heterogeneous scenarios which are more prevalent in real-world settings.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 21 (15566013) , 3944-3959
Byzantine-Robust Asynchronous Federated Learning via Feature Fingerprinting
Asynchronous federated learning (AFL) accelerates collaborative training across heterogeneous devices compared to synchronous federated learning, but increases vulnerability to Byzantine attacks due to its asynchronous aggregation. Existing defenses rely on parametric similarity between models and temporal consistency of updates, which are compromised by data and device heterogeneity, leading to ineffective robustness. To address this limitation, we propose Belisa, a Byzantine-robust AFL framework that enhances fidelity, robustness, and efficiency under heterogeneous scenarios. Belisa introduces novel discrepancies between feature representations of local models to distinguish malicious models from benign ones. By leveraging a reference model trained on publicly available data, Belisa quantifies these discrepancies, referred to as feature fingerprints, and filters out malicious models through clustering. Extensive experiments on six datasets from three types of tasks under five advanced Byzantine attacks demonstrate Belisa’s superiority. Notably, Belisa consistently outperforms existing approaches across both attack and non-attack settings. Under attack scenarios, it lowers the average test error rate to 0.42× that of baseline methods. Furthermore, Belisa accelerates the aggregation process by an average of 12.3× compared to other methods. To the best of our knowledge, Belisa is the first Byzantine-robust AFL framework, which provides a broadly applicable countermeasure in heterogeneous scenarios which are more prevalent in real-world settings.
- Open Access
- 27 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 692-706
Pegasus: A Universal Framework for Scalable Deep Learning Inference on the Dataplane
- 27 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 692-706
Open AccessPegasus: A Universal Framework for Scalable Deep Learning Inference on the Dataplane
- 20 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 33 (5) , 2336-2351
Expediting Federated Learning on Non-IID Data by Maximizing Communication Channel Utilization
Federated learning (FL) is at the core of intelligent Internet architecture. It allows clients to jointly train a model without direct data sharing. In such a process, clients and the central server share information through communication channels formed by parameters. However, the non-iid training data in clients significantly impacts global model convergence and brings difficulties for the evaluation of local contributions. Most of existing studies try to expand the communication channel by improving consistency with variance reduction or regularization, but such methods neglect an important factor, i.e., channel utilization, hence their capability for sharing information is under-utilized. Moreover, the issue of contribution evaluation is still unsolved. In this paper, we simultaneously solve the former two challenges (i.e., model convergence and contribution evaluation) by modeling the indirect data sharing of FL as a problem of information communication. We prove that FL with non-iid data forms noisy communication channels, which have limited capability for information transmission, i.e., limited channel capacity. The main factor in deciding the channel capacity is the Gradient Signal to Noise Ratio (GSNR). Through analyzing GSNR, we further prove that channel capacity can be reached by optimal local updates and propose a method FedGSNR to calculate it, which allows us to maximize channel utilization in FL, leading to faster model convergence. Moreover, as the contribution of the local dataset depends on the amount of provided information, the derived GSNR allows the server to accurately evaluate the contributions of different clients (i.e., the quality of local datasets).
- 20 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 33 (5) , 2336-2351
Expediting Federated Learning on Non-IID Data by Maximizing Communication Channel Utilization
Federated learning (FL) is at the core of intelligent Internet architecture. It allows clients to jointly train a model without direct data sharing. In such a process, clients and the central server share information through communication channels formed by parameters. However, the non-iid training data in clients significantly impacts global model convergence and brings difficulties for the evaluation of local contributions. Most of existing studies try to expand the communication channel by improving consistency with variance reduction or regularization, but such methods neglect an important factor, i.e., channel utilization, hence their capability for sharing information is under-utilized. Moreover, the issue of contribution evaluation is still unsolved. In this paper, we simultaneously solve the former two challenges (i.e., model convergence and contribution evaluation) by modeling the indirect data sharing of FL as a problem of information communication. We prove that FL with non-iid data forms noisy communication channels, which have limited capability for information transmission, i.e., limited channel capacity. The main factor in deciding the channel capacity is the Gradient Signal to Noise Ratio (GSNR). Through analyzing GSNR, we further prove that channel capacity can be reached by optimal local updates and propose a method FedGSNR to calculate it, which allows us to maximize channel utilization in FL, leading to faster model convergence. Moreover, as the contribution of the local dataset depends on the amount of provided information, the derived GSNR allows the server to accurately evaluate the contributions of different clients (i.e., the quality of local datasets).
- Top Cited Papers
- 16 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Communications Surveys & Tutorials
- Vol. 28, 985-1024
Zero-Knowledge Proof-Based Verifiable Decentralized Machine Learning in Communication Network: A Comprehensive Survey
Over recent decades, machine learning has significantly advanced network communication, enabling improved decision-making, user behavior analysis, and fault detection. Simultaneously, the growth of communication networks has facilitated the efficient collection of large-scale training data. Traditional centralized machine learning, however, requires collecting data from users, raising significant concerns about privacy and security. Decentralized approaches, where participants exchange computation results instead of raw private data, mitigate these risks but introduce challenges related to trust and verifiability. A critical issue arises: How can one ensure the integrity and validity of computation results shared by other participants? Existing survey articles predominantly address security and privacy concerns in decentralized machine learning, whereas this survey uniquely highlights the emerging issue of verifiability. Recognizing the critical role of zero-knowledge proofs in ensuring verifiability, we present a comprehensive review of Zero-Knowledge Proof-based Verifiable Machine Learning (ZKP-VML). To clarify the research problem, we present a definition of ZKP-VML consisting of four algorithms and several key security properties. In addition, we provide an overview of the current research landscape by systematically organizing the research timeline and categorizing existing schemes based on their security properties. Furthermore, through an in-depth analysis of each existing scheme, we summarize their technical contributions and optimization strategies, aiming to uncover common design principles underlying ZKP-VML schemes. Building on the reviews and analysis presented, we identify current research challenges and suggest future research directions. To the best of our knowledge, this is the most comprehensive survey to date on verifiable decentralized machine learning and ZKP-VML.
- 16 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Communications Surveys & Tutorials
- Vol. 28, 985-1024
Top Cited PapersZero-Knowledge Proof-Based Verifiable Decentralized Machine Learning in Communication Network: A Comprehensive Survey
Over recent decades, machine learning has significantly advanced network communication, enabling improved decision-making, user behavior analysis, and fault detection. Simultaneously, the growth of communication networks has facilitated the efficient collection of large-scale training data. Traditional centralized machine learning, however, requires collecting data from users, raising significant concerns about privacy and security. Decentralized approaches, where participants exchange computation results instead of raw private data, mitigate these risks but introduce challenges related to trust and verifiability. A critical issue arises: How can one ensure the integrity and validity of computation results shared by other participants? Existing survey articles predominantly address security and privacy concerns in decentralized machine learning, whereas this survey uniquely highlights the emerging issue of verifiability. Recognizing the critical role of zero-knowledge proofs in ensuring verifiability, we present a comprehensive review of Zero-Knowledge Proof-based Verifiable Machine Learning (ZKP-VML). To clarify the research problem, we present a definition of ZKP-VML consisting of four algorithms and several key security properties. In addition, we provide an overview of the current research landscape by systematically organizing the research timeline and categorizing existing schemes based on their security properties. Furthermore, through an in-depth analysis of each existing scheme, we summarize their technical contributions and optimization strategies, aiming to uncover common design principles underlying ZKP-VML schemes. Building on the reviews and analysis presented, we identify current research challenges and suggest future research directions. To the best of our knowledge, this is the most comprehensive survey to date on verifiable decentralized machine learning and ZKP-VML.
- 15 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Dependable and Secure Computing
- Vol. 22 (4) , 3337-3353
Pisces: In-Path Distributed Denial-of-Service Defense via Efficient Authentication Code Embedded in IP Address
High-volume brute-force distributed denial-of-service (DDoS) attack is among the top threats on the Internet. Existing widely deployed methods (e.g., BGP blackhole and scrubbing center) have difficulty achieving legitimate traffic friendliness, low cost, low latency, and high accuracy. We present an in-path DDoS defense mechanism, namely Pisces. Without requiring modifications to existing IP protocols, Pisces embeds authentication information into the IP address. Simultaneously, we design a QUIC-based extension to distribute authentication information. Pisces incorporates a translator module and a filter module, which accurately identifies malicious and legitimate traffic. These multi-dimensional compatibility advantages make it easy to deploy in the real world. We implement Pisces on a high-end commercial router with service processing units. Even without hardware acceleration, a single CPU can achieve $20Gbps$ throughput and the performance can scale linearly with the number of CPUs. The additional latency for the victim-related traffic and other traffic is around $27us$ and $0.5us$, respectively, whose cost is far less than the scrubbing center. Remarkably, Pisces without false positives can provide high-quality datasets for intelligent approaches and form a prominent complementary effect.
- 15 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Dependable and Secure Computing
- Vol. 22 (4) , 3337-3353
Pisces: In-Path Distributed Denial-of-Service Defense via Efficient Authentication Code Embedded in IP Address
High-volume brute-force distributed denial-of-service (DDoS) attack is among the top threats on the Internet. Existing widely deployed methods (e.g., BGP blackhole and scrubbing center) have difficulty achieving legitimate traffic friendliness, low cost, low latency, and high accuracy. We present an in-path DDoS defense mechanism, namely Pisces. Without requiring modifications to existing IP protocols, Pisces embeds authentication information into the IP address. Simultaneously, we design a QUIC-based extension to distribute authentication information. Pisces incorporates a translator module and a filter module, which accurately identifies malicious and legitimate traffic. These multi-dimensional compatibility advantages make it easy to deploy in the real world. We implement Pisces on a high-end commercial router with service processing units. Even without hardware acceleration, a single CPU can achieve $20Gbps$ throughput and the performance can scale linearly with the number of CPUs. The additional latency for the victim-related traffic and other traffic is around $27us$ and $0.5us$, respectively, whose cost is far less than the scrubbing center. Remarkably, Pisces without false positives can provide high-quality datasets for intelligent approaches and form a prominent complementary effect.
- 12 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ACM Transactions on Networking
- Vol. 32 (3) , 1873-1887
FedPAGE: Pruning Adaptively Toward Global Efficiency of Heterogeneous Federated Learning
When workers are heterogeneous in computing and transmission capabilities, the global efficiency of federated learning suffers from the straggler issue, i.e., the slowest worker drags down the overall training process. We propose a novel and efficient federated learning framework named FedPAGE, where workers perform distributed pruning adaptively towards global efficiency, i.e., fast training and high accuracy. For fast training, we develop a pruning rate learning approach generating an adaptive pruning rate for each worker, making the overall update time approximate to the fastest worker’s update time, i.e., no stragglers. For high accuracy, we find that structural similarity between sub-models is essential to global model accuracy in the distributed pruning, and thus propose the CIG_X pruning scheme to ensure maximum similarity. Meanwhile, we adopt the sparse training and design model aggregating of different size sub-models to cope with distributed pruning. We prove the convergence of FedPAGE and demonstrate the effectiveness of FedPAGE on image classification and natural language inference tasks. Compared with the state-of-the-art, FedPAGE achieves higher accuracy with the same speedup ratio.
- 12 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ACM Transactions on Networking
- Vol. 32 (3) , 1873-1887
FedPAGE: Pruning Adaptively Toward Global Efficiency of Heterogeneous Federated Learning
When workers are heterogeneous in computing and transmission capabilities, the global efficiency of federated learning suffers from the straggler issue, i.e., the slowest worker drags down the overall training process. We propose a novel and efficient federated learning framework named FedPAGE, where workers perform distributed pruning adaptively towards global efficiency, i.e., fast training and high accuracy. For fast training, we develop a pruning rate learning approach generating an adaptive pruning rate for each worker, making the overall update time approximate to the fastest worker’s update time, i.e., no stragglers. For high accuracy, we find that structural similarity between sub-models is essential to global model accuracy in the distributed pruning, and thus propose the CIG_X pruning scheme to ensure maximum similarity. Meanwhile, we adopt the sparse training and design model aggregating of different size sub-models to cope with distributed pruning. We prove the convergence of FedPAGE and demonstrate the effectiveness of FedPAGE on image classification and natural language inference tasks. Compared with the state-of-the-art, FedPAGE achieves higher accuracy with the same speedup ratio.