Organizers 组织者:

• • Zhou Haoxue (周皓雪), Senior Engineer, Chongqing Polytechnic University of Mechatronics
• • He Xiang (何湘), Professor-level Senior Lecturer, Chongqing Polytechnic University of Mechatronics

Introduction 论坛简介:

With the rapid breakthroughs of large language models and generative AI, artificial intelligence is entering a new stage of general intelligence driven by foundation models. Next-generation communication networks face emerging challenges such as ultra-large-scale connectivity, heterogeneous network convergence and intelligent operations and maintenance, calling for large model technologies to reshape network architecture and management paradigms. However, the stringent requirements of communication networks on real-time performance, reliability and security make it difficult to deploy general-purpose large models directly. This track focuses on the core challenges at the intersection of large models and communication networks, addressing key issues such as domain-specific model construction for networking, agent-driven network autonomy, real-time inference optimization and trustworthy deployment. It aims to advance large models from “chat tools” to “network infrastructure”, and to help communication networks evolve toward self-intelligent, autonomous next-generation intelligent networks.

随着大语言模型和生成式 AI 技术的快速突破，人工智能正迈向大模型驱动的通用智能新阶段。下一代通信网络面临超大规模连接、异构网络融合、智能运维等新挑战，亟需大模型技术重塑网络架构与管理范式。然而通信网络对实时性、可靠性与安全性的严苛要求，使通用大模型难以直接落地。本专题聚焦大模型与通信网络交叉领域核心挑战，探讨网络领域模型构建、Agent 驱动网络自治、实时推理优化与可信部署等关键问题，推动大模型从“聊天工具”进化为“网络基础设施”，助力通信网络向自智自治的下一代智能网络演进。

Topics 主题范围:

• Pre-training and Instruction Tuning of Large Models for Communication Networks
  通信网络领域大模型的预训练与指令微调策略
• Large Model-Assisted Network Fault Diagnosis and Intelligent O&M
  大模型辅助的网络故障诊断与智能运维
• Large Model Architectures and Protocol Design for 6G and Future Networks
  面向 6G / 未来网络的大模型架构与协议设计
• Large Models for Network Security and Intelligent Intrusion Detection
  大模型在网络安全与智能入侵检测中的应用
• Efficient Inference Architectures of Large Models for Real-Time Network Scenarios
  面向实时网络场景的大模型高效推理架构
• Safety Alignment and Privilege-Abuse Prevention of Large Models in Network Automation
  大模型在网络自动化中的安全对齐与越权防护
• Modeling and Optimization of Network Constraints on Large Model Inference Latency and Throughput
  大模型推理延迟与吞吐量的网络约束建模与优化
• Physical-Layer Security and Anti-Jamming Technologies for Low-Altitude Networks
  低空网络的物理层安全与抗干扰技术
• UAV Identity Authentication and “Rogue Drone” Detection and Countermeasure Systems
  无人机身份认证与“黑飞”检测反制系统
• AIGC Content Security and Network Governance
  AIGC 内容安全与网络治理
• Device-Level Autonomous Decision-Making and Lightweight Reinforcement Learning
  设备级自主决策与轻量化强化学习
• Multimodal Sensor Fusion and Environment Perception Algorithms
  多模态传感器融合与环境感知算法