Agents Collaboration and Service Computing

智能体协同与服务计算

With the rapid advances in foundation models, retrieval-augmented generation, tool use, and autonomous planning, intelligent agents are evolving from standalone inference units into service entities capable of environment perception, task decomposition, tool invocation, collaborative decision-making, and continuous execution. This trend is reshaping service computing from conventional service composition and invocation toward agent-oriented workflow orchestration, resource-aware deployment, execution efficiency optimization, and end-to-end QoS assurance. In this context, several key challenges have emerged, including multi-agent collaboration, integration of heterogeneous models and service components, elastic deployment across cloud-edge-end environments, and the joint optimization of latency, throughput, reliability, and cost. This special session focuses on the emerging intersection of agent collaboration and service computing. It solicits original research on agent workflow orchestration and execution, multi-agent coordination mechanisms, service composition and resource scheduling, inference acceleration and deployment optimization, performance modeling, and QoS assurance. The goal is to bring together recent advances in theories, methodologies, systems, and applications, and to promote efficient, reliable, and scalable agent systems for complex service scenarios.
随着大模型、检索增强生成、工具调用与自主规划能力的快速发展，智能体正从单一推理单元演化为能够感知环境、分解任务、调用工具、协同决策并持续执行的智能服务载体。这一趋势推动服务计算从传统的服务组合与调用，进一步迈向面向智能体的工作流编排、资源感知部署、执行效率优化与服务质量保障。在这一背景下，如何实现多智能体协同、异构模型与服务组件集成、跨云边端的弹性部署，以及面向时延、吞吐、可靠性、成本等QoS指标的联合优化，已成为智能化系统与服务领域的重要研究方向。本专题聚焦智能体协同与服务计算的交叉前沿，关注智能体工作流编排与执行、多智能体协同机制、服务组合与资源调度、推理加速与部署优化、性能建模与QoS保障等关键问题，旨在汇聚相关领域的最新理论、方法、系统与应用成果，推动智能体系统在复杂业务场景中的高效、可靠与可扩展落地。

Topics:
-多智能体协同与协调机制 (Multi-agent collaboration and coordination mechanisms)
-智能体工作流建模、规划与编排 (Agentic workflow modeling, planning, and orchestration)
-智能体工作流的任务调度与资源分配 (Task scheduling and resource allocation for agentic workflows)
-面向 QoS 的智能体服务执行、调度与保障 (QoS-aware execution, scheduling, and assurance for agent services)
-智能体系统的性能建模与效率优化 (Performance modeling and efficiency optimization for agent systems)
-智能体系统的训练、微调与适配 (Training, fine-tuning, and adaptation for agent systems)
-智能体推理服务与性能优化 (Inference serving and performance optimization for agent services)
-智能体系统的基准测试、性能分析与评估 (Benchmarking, profiling, and performance evaluation for agent systems)

ICET 2026 will take place at Chengdu, China, from May 29-31, 2026. This key electronics event should be a "must attend" on your conference calendar.

Call for Papers Timeline / 征稿时间

• Submission Deadline: April 10th, 2026
  全文截稿时间：2026年04月10日
  
• Registration Deadline: April 25th, 2026
  注册截止时间：2026年04月25日
  

Oragnizing Chair

Danyang Zheng is an Associate Professor at Southwest Jiaotong University. His research focuses on AI for networking and agent service computing. He has published more than 70 papers in leading journals and conferences in computer networks and service computing, including IEEE/ACM Transactions on Networking, IEEE Transactions on Services Computing, IEEE Transactions on Network Science and Engineering, IEEE Transactions on Network and Service Management, and IEEE INFOCOM. He serves as an Associate Editor of Big Data Mining and Analytics and has held organizing roles in several international conferences, including ICCC, WCCCT, and IEEE ICNC. He was selected for a provincial high-level youth talent program in Sichuan and has also served as a Competition Manager for WorldSkills International.