About Me
I am currently an Associate Professor at East China University of Science and Technology . I received my Ph.D. in Computer Science (2023) from Tsinghua University , advised by Prof. Yong Jiang and Prof. Qing Li at Peng Cheng Laboratory. I completed my undergraduate studies at China University of Geosciences (Beijing) , earning a bachelor's degree in Computer Science and a second degree in Law in 2017. I was a visiting researcher at Dublin City University (2022), working with Prof. Gabriel-Miro Muntean (IEEE Fellow) and a postdoctoral researcher at Fudan University (2023–2025), working with Prof. Yang Xu .
My group, AI-driven Media and Networking (AimNet Lab), focuses on AI-driven media systems, networking, and intelligent infrastructure. Current research directions include:
- Cloud-edge-end collaborative intelligent network distribution and optimization.
- Enhancement and restoration for videos and AI-generated media content.
- Communication acceleration for AI computing clusters.
- High-performance interconnection and storage optimization for large-model serving.
- Trustworthy AI agents and networked system security.
My research has been published in leading venues including IEEE/ACM Transactions on Networking, IEEE Communications Surveys & Tutorials, ACM WWW, ACM Multimedia, IEEE Transactions on Multimedia, and IEEE/ACM IWQoS. See Research for representative publications and ongoing directions.
Research & Publications
AimNet Lab studies AI-driven media systems, networking, and intelligent infrastructure. The work below starts with representative publications with figures, followed by direction-specific reading lists and remaining publications.
Cloud-Edge-End Video Delivery
Media Enhancement
AI Cluster Communication
Large-Model Serving Systems
Trustworthy AI Agents & Security
Selected Publications with Figures
Six representative papers are shown together before the direction-specific reading lists.
Research Directions
Cloud-Edge-End Video Delivery and Network Optimization
This direction studies how edge intelligence, network adaptation, and resource coordination can improve delivery quality under dynamic bandwidth, device, and data-center conditions.
- LEAP: Learning-Based Smart Edge with Caching and Prefetching for Adaptive Video Streaming - IWQoS 2019 (CCF B)
- NCTM: A Novel Coded Transmission Mechanism for Short Video Deliveries - WWW 2024 (CCF A)
- Poche: A Priority-Based Flow-Aware In-Network Caching Scheme in Data Center Networks - IEEE Transactions on Network and Service Management 2022 (CCF B)
- Modeling and Optimization of the Data Plane in the SDN-based DCN by Queuing Theory - Journal of Network and Computer Applications 2022
Video Enhancement and Restoration
This direction focuses on real-time media enhancement under constrained networks, including frame recovery, super-resolution, and quality improvement for emerging AI-generated media.
- Reparo: QoE-Aware Live Video Streaming in Low-Rate Networks by Intelligent Frame Recovery - ACM Multimedia 2023 (CCF A)
- RTCSR: Zero-latency Aware Super-resolution for WebRTC Mobile Video Streaming - SIGCOMM Workshop 2023 (CCF A)
AI Cluster Communication
This direction explores communication acceleration for distributed training and multi-tenant AI workloads, especially in-network aggregation, resource pooling, and training-parallelism aware network design.
- AIRP: Accelerating Multi-tenant Distributed Learning with In-network Resource Pooling - IEEE Transactions on Networking 2026 (CCF A)
- 3D-INA: An Exploration of Integrating In-Network Aggregation into 3D Parallelism for LLM Training - INFOCOM 2026 (CCF A)
- Rina: Enhancing Ring-AllReduce with In-network Aggregation in Distributed Model Training - ICNP 2024 (CCF B)
Large-Model Serving Systems
This direction studies high-performance interconnection, scheduling, placement, and storage optimization for large-model serving and AI infrastructure. This direction is pursued with collaborators including Dongbiao He, focusing on GPU scheduling, disaggregated memory, and RDMA-based distributed systems.
- Bridging the GPU Utilization Gap: Predictive Multi-Dimensional Resource Scheduling for AI Workloads - EuroSys 2026
- Shard: A Scalable and Resize-optimized Hash Index on Disaggregated Memory - VLDB 2026
- HybridSkipList: A Case Study of Designing Distributed Data Structure with Hybrid RDMA - COMPSAC 2021
Trustworthy AI Agents and Networked System Security
This direction examines security risks that arise when AI agents interact with tools, code, networked systems, and large-model services. This direction is pursued with collaborators including Dr. Qiyang Song, connecting AI-agent security with software security, Web3 security, provenance-based auditing, and multimodal generative AI security.
- Silence False Alarms: Identifying Anti-Reentrancy Patterns on Ethereum to Refine Smart Contract Reentrancy Detection - NDSS 2025
- vCause: Efficient and Verifiable Causality Analysis for Cloud-based Endpoint Auditing - USENIX Security 2026
- Weaponizing Tokens: Backdooring Text-to-Image Generation via Token Remapping - ICME 2025