Yannan Hu
Research Subjects
Research Interests
Yannan Hu has a total of 32 co-authors. Here are the top co-authors by number of shared publications:
- Yong Cui
Tsinghua University, Beijing, China
5 shared publications - Boyan Huang
Tsinghua University
2 shared publications - Chuyi Wang
Tsinghua University
2 shared publications - Jinzhou Liu
Zhongguancun Laboratory
2 shared publications - Tongze Wang
Tsinghua University
2 shared publications - Wenduo Wang
Tsinghua University
2 shared publications
Latest Publications
- Open Access
NetMamba+: A Framework of Pre-trained Models for Efficient and Accurate Network Traffic Classification
With the rapid growth of encrypted network traffic, effective traffic classification has become essential for network security and quality of service management. Current machine learning and deep learning approaches for traffic classification face three critical challenges: computational inefficiency of Transformer architectures, inadequate traffic representations with loss of crucial byte-level features while retaining detrimental biases, and poor handling of long-tail distributions in real-world data. We propose NetMamba+, a framework that addresses these challenges through three key innovations: (1) an efficient architecture considering Mamba and Flash Attention mechanisms, (2) a multimodal traffic representation scheme that preserves essential traffic information while eliminating biases, and (3) a label distribution-aware fine-tuning strategy. Evaluation experiments on massive datasets encompassing four main classification tasks showcase NetMamba+'s superior classification performance compared to state-of-the-art baselines, with improvements of up to 6.44\% in F1 score. Moreover, NetMamba+ demonstrates excellent efficiency, achieving 1.7x higher inference throughput than the best baseline while maintaining comparably low memory usage. Furthermore, NetMamba+ exhibits superior few-shot learning abilities, achieving better classification performance with fewer labeled data. Additionally, we implement an online traffic classification system that demonstrates robust real-world performance with a throughput of 261.87 Mb/s. As the first framework to adapt Mamba architecture for network traffic classification, NetMamba+ opens new possibilities for efficient and accurate traffic analysis in complex network environments.
Open AccessNetMamba+: A Framework of Pre-trained Models for Efficient and Accurate Network Traffic Classification
With the rapid growth of encrypted network traffic, effective traffic classification has become essential for network security and quality of service management. Current machine learning and deep learning approaches for traffic classification face three critical challenges: computational inefficiency of Transformer architectures, inadequate traffic representations with loss of crucial byte-level features while retaining detrimental biases, and poor handling of long-tail distributions in real-world data. We propose NetMamba+, a framework that addresses these challenges through three key innovations: (1) an efficient architecture considering Mamba and Flash Attention mechanisms, (2) a multimodal traffic representation scheme that preserves essential traffic information while eliminating biases, and (3) a label distribution-aware fine-tuning strategy. Evaluation experiments on massive datasets encompassing four main classification tasks showcase NetMamba+'s superior classification performance compared to state-of-the-art baselines, with improvements of up to 6.44\% in F1 score. Moreover, NetMamba+ demonstrates excellent efficiency, achieving 1.7x higher inference throughput than the best baseline while maintaining comparably low memory usage. Furthermore, NetMamba+ exhibits superior few-shot learning abilities, achieving better classification performance with fewer labeled data. Additionally, we implement an online traffic classification system that demonstrates robust real-world performance with a throughput of 261.87 Mb/s. As the first framework to adapt Mamba architecture for network traffic classification, NetMamba+ opens new possibilities for efficient and accurate traffic analysis in complex network environments.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 34, 7361-7376
NetMamba+: A Framework of Pre-Trained Models for Efficient and Accurate Network Traffic Classification
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 34, 7361-7376
NetMamba+: A Framework of Pre-Trained Models for Efficient and Accurate Network Traffic Classification
- 9 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 34, 1988-2003
TitanLog: Hierarchical and Elastic Logging for High-Speed Network Data Stream
Logging network traffic plays a crucial role as it serves as the foundation for various network applications. As network scale continues to expand, contemporary network traffic becomes increasingly high-speed, high-volume, and dynamic. This growth poses challenges to traditional server-based solutions. In this paper, we propose TitanLog, a hierarchical and elastic logging system designed specifically for large-scale network traffic. TitanLog utilizes the hierarchical logging methodology, which aims to identify the importance of each packet in real-time and log packet data of different importance at different levels. To enhance efficiency, we propose a co-design of the emerging programmable switch and the server, incorporating sketches and RDMA to boost performance. To achieve elasticity, we design mechanisms for run-time adjustments and monitoring for resource insufficiency. TitanLog possesses the capability to switch between these modes at run-time. We fully implement TitanLog on a testbed and conduct extensive evaluations. The experimental results demonstrate that TitanLog supports logging of 100Gbps traffic with a zero packet loss rate and reduces the log volume by up to 96.28%.
- 9 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 34, 1988-2003
TitanLog: Hierarchical and Elastic Logging for High-Speed Network Data Stream
Logging network traffic plays a crucial role as it serves as the foundation for various network applications. As network scale continues to expand, contemporary network traffic becomes increasingly high-speed, high-volume, and dynamic. This growth poses challenges to traditional server-based solutions. In this paper, we propose TitanLog, a hierarchical and elastic logging system designed specifically for large-scale network traffic. TitanLog utilizes the hierarchical logging methodology, which aims to identify the importance of each packet in real-time and log packet data of different importance at different levels. To enhance efficiency, we propose a co-design of the emerging programmable switch and the server, incorporating sketches and RDMA to boost performance. To achieve elasticity, we design mechanisms for run-time adjustments and monitoring for resource insufficiency. TitanLog possesses the capability to switch between these modes at run-time. We fully implement TitanLog on a testbed and conduct extensive evaluations. The experimental results demonstrate that TitanLog supports logging of 100Gbps traffic with a zero packet loss rate and reduces the log volume by up to 96.28%.
- Open Access
- 3 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 169-170
StarTCP: Handover-aware Transport Protocol for Starlink
Legacy transport protocols such as TCP and QUIC suffer from high packet loss and low link utilization in Starlink. From the measurement data, we figure out the ground-satellite link (GSL) handover is mainly to blame. The periodic handovers result in link interruptions and bursty losses with a fixed interval of 15s, which impair TCP’s performance. Based on this finding, we present a handover-aware transport protocol, StarTCP, which proactively stalls transmission during handovers to avoid bursty losses and erroneous congestion signals. Preliminary results indicate that StarTCP can efficiently reduce packet loss and enhance throughput in Starlink.
- 3 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 169-170
Open AccessStarTCP: Handover-aware Transport Protocol for Starlink
Legacy transport protocols such as TCP and QUIC suffer from high packet loss and low link utilization in Starlink. From the measurement data, we figure out the ground-satellite link (GSL) handover is mainly to blame. The periodic handovers result in link interruptions and bursty losses with a fixed interval of 15s, which impair TCP’s performance. Based on this finding, we present a handover-aware transport protocol, StarTCP, which proactively stalls transmission during handovers to avoid bursty losses and erroneous congestion signals. Preliminary results indicate that StarTCP can efficiently reduce packet loss and enhance throughput in Starlink.
- 10 May 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet Computing
- Vol. 27 (3) , 44-49
SAV-D: Defending DDoS with Incremental Deployment of SAV
Large-scale Internet Protocol (IP) spoofing distributed denial-of-service (DDoS) attacks is one of the major cyber threats. Current commercial defenses focus on eliminating attacks at the destination end, which raises concerns about the cost of appliances and the impact on quality of service. As complementaries, source-end schemes using source address validation (SAV) can block IP spoofing traffic from entering the backbone. However, their effectiveness is restricted by incremental deployment. This paper proposes SAV-D, an SAV-based honeynet-like distributed defense architecture against IP spoofing DDoS. Each SAV device functions as a honeypot to capture more threat data. By aggregating these data, SAV-D can accurately detect ongoing attacks and generate defense policies. With the policies, both SAV and non-SAV devices can filter malicious traffic. Our simulation results demonstrate that SAV-D can effectively filter out 80% of attack traffic with a modest deployment ratio of only 10%. To enable broader adoption, we also provide some guidance on standardizing SAV-D.
- 10 May 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet Computing
- Vol. 27 (3) , 44-49
SAV-D: Defending DDoS with Incremental Deployment of SAV
Large-scale Internet Protocol (IP) spoofing distributed denial-of-service (DDoS) attacks is one of the major cyber threats. Current commercial defenses focus on eliminating attacks at the destination end, which raises concerns about the cost of appliances and the impact on quality of service. As complementaries, source-end schemes using source address validation (SAV) can block IP spoofing traffic from entering the backbone. However, their effectiveness is restricted by incremental deployment. This paper proposes SAV-D, an SAV-based honeynet-like distributed defense architecture against IP spoofing DDoS. Each SAV device functions as a honeypot to capture more threat data. By aggregating these data, SAV-D can accurately detect ongoing attacks and generate defense policies. With the policies, both SAV and non-SAV devices can filter malicious traffic. Our simulation results demonstrate that SAV-D can effectively filter out 80% of attack traffic with a modest deployment ratio of only 10%. To enable broader adoption, we also provide some guidance on standardizing SAV-D.