SPECIAL SESSIONS

 

Special Session 2: Distributed Large Language Model Training/Inference Systems over Edge Networks

Flyer Download

The deployment and training of Large Language Models (LLMs) over edge networks introduce substantial challenges, including constrained resources, heterogeneous infrastructures, and stringent latency constraints. This Special Session solicits research contributions on distributed systems and optimization techniques for resource-efficient LLM inference and convergence-aware LLM training in edge computing environments. Two core technical thrusts include:
(1) Inference deployment resource optimization, encompassing expert selection and placement, token-aware routing, memory- and bandwidth-constrained scheduling, KV-cache management, and SLO-compliant serving strategies across distributed edge nodes;
(2) Training convergence optimization, including communication-efficient gradient synchronization, sparsity-aware update mechanisms, adaptive parallelism strategies (e.g., MoE), and convergence acceleration under limited communication bandwidth.

 

Topics of interest include but are not limited to:
– Joint computation–communication–energy co-optimization;
– Model partitioning and orchestration across edge–cloud hierarchies;
– Federated or hierarchical training with provable convergence guarantees;
– System frameworks for scalable, low-latency LLM inference.
The session welcomes both theoretical results and system-level designs, particularly those validated on realistic edge platforms or standardized LLM benchmarks.

 

Please choose Special Session 2 to submit. Submission Link: https://www.zmeeting.org/submission/wccct2026

 

Special Session 2 Chair

   Assoc. Prof. Danyang Zheng, Southwest Jiaotong University, China
   Email: dzheng5@swjtu.edu.cn
   Research Areas: Services Computing, AI on Network, Network Reliability, and Network Security

Biography: Danyang Zheng received the B.S. degree in computer science from the University of Electronic Science and Technology of China, Chengdu, China, in 2016, and the Ph.D. degree in computer science from the Georgia State University, Atlanta, GA, USA, in 2021. He is currently an Associate Professor at Southwest Jiaotong University (SWJTU), Chengdu, China. His research interests include Services Computing, AI on Network, Network Reliability and Security, In-Network Computation, and Combinational Optimization. He has published more than 70 technical works in his research fields and is serving as the Youth Associate Editor for IEEE Big Data Mining and Analytics. He served as Publicity Chair of ICCC 2024-2025, WCCCT 2025-2026, CNCIT 2025 and Symposium chair of WCCCT 2024, IEEE ICNC 2025-2026.

 

Co-Chairs

   Assoc. Prof. Chengzong Peng, Chengdu University of Information Technology, China
   Email: pcz@cuit.edu.cn
   Research Areas: Network Function Virtualization, Network Security

Biography: Chengzong Peng is a tenured Associate Professor and master’s supervisor at the School of Cyberspace Security, Chengdu University of Information Technology (Chengdu University of Information Engineering). He earned his Ph.D. in Computer Science from Georgia State University, USA, and completed his undergraduate studies in Mathematics at the University of California, Irvine. His research interests focus on cybersecurity, network reliability, and AI‑enabled defense in edge networks. He has published in major international venues and serves on program committees of conferences such as ICCC.

 

   Assoc. Prof. Shaohua Cao, University of Petroleum (East China), China
   Email: shaohuacao@upc.edu.cn
   Research Areas: Edge Intelligence, Federated and Distributed Learning

Biography: Shaohua Cao is an Associate Professor at the College of Computer Science and Technology, China University of Petroleum (East China). His research focuses on edge computing, federated learning, distributed AI systems, and resource-efficient inference under communication and energy constraints. He has authored over 40 peer-reviewed publications and actively serves as a reviewer for journals such as IEEE Access and FGCS. Dr. Cao has led and participated in several national and provincial research projects in intelligent computing and industrial AI. He also supervises graduate students and teaches courses related to computer systems and emerging intelligent technologies.

 

   Asst. Prof. Chen Yang, Southwest Jiaotong University, China
   Email: yangc@swjtu.edu.cn
   Research Areas: Services Computing, Edge Intelligence, Federated Learning

Biography: Chen Yang received Ph.D. degree in computer science at the State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications. Currently, he is an assistant professor in School of Computing and Artificial Intelligence, Southwest Jiaotong University. He has published papers in Mobicom, TMC, TSC, and so on. His research interests include edge intelligence, federated learning and service computing.