云边端协同计算与智能

Cloud-Edge-End Collaborative Computing and Intelligence

摘要: 云边端协同计算作为应对数据爆炸与实时智能需求的关键架构，通过将计算任务合理分布于终端、边缘与云端，有效缓解了传统云计算在高时延、带宽压力和隐私安全等方面的瓶颈。文章系统梳理了云边端协同计算与智能的国内外研究进展，重点从云边端协同计算框架构建与分布式智能实现两个维度展开分析，总结了包括任务卸载、资源调度、多模态感知、联邦学习等关键技术路径。文章进一步展望了未来发展趋势，指出算力网络化、边缘智能化、绿色计算与内生安全将成为推动云边端协同走向纵深的核心方向，为我国构建高效、可信、可持续的协同智能基础设施提供理论参考。

Abstract: Cloud-Edge-End collaborative computing has emerged as a critical architecture to address the challenges of data explosion and the demand for real-time intelligence. By efficiently distributing computational tasks across end devices, edge nodes, and the cloud, this paradigm effectively mitigates bottlenecks inherent in traditional cloud computing, such as high latency, bandwidth pressure, and privacy concerns. This paper systematically reviews the state-of-the-art in Cloud-Edge-End collaborative computing and intelligence, analyzing key research from two primary dimensions: the construction of collaborative computing frameworks and the implementation of distributed intelligence. It summarizes pivotal technical approaches, including task offloading, resource scheduling, multi-modal perception, and federated learning. Furthermore, the paper outlines future development trends, identifying computing power networking, edge intelligence, green computing, and intrinsic security as core directions for advancing collaborative intelligent infrastructure, providing a theoretical reference for building efficient, trustworthy, and sustainable systems.