摘要
边缘缓存能够有效降低服务时延、缓解回程链路流量压力以及提升用户体验质量,可用于解决现有移动通信网络架构难以支撑的数据流量极速增长,满足用户对高质量网络服务的需求。但是复杂的网络状况和未知内容流行度给边缘缓存策略研究带来很大挑战。首先从边缘缓存的优势挑战、应用场景、核心要素等方面介绍边缘缓存网络的基础架构和流程,对研究现状进行分析总结。随后简述深度强化学习技术和DQN算法,分类介绍基于不同缓存系统架构和基于不同深度强化学习方法的边缘缓存策略。最后重点对基于深度强化学习的缓存模型进行解析,提出下一步研究方向。
Edge cache can effectively reduce service delay,relieve traffic pressure of backhaul link and improve users’experience quality.With these advantages,it can be used to solve the rapid growth of data traffic that is difficult to be supported by existing mobile communication network architecture and meet users’demand for high-quality network services.However,complex network conditions and unknown content popularity bring great challenges to the research of edge caching strategy.Firstly,the infrastructure and process of edge cache network were introduced from the advantages and challenges,application scenarios,core elements and other aspects,and the research status were analyzed and summarized.Secondly,deep reinforcement learning technology and DQN algorithm were briefly introduced,and edge caching strategies based on different cache architecture and different deep reinforcement learning methods were classified.Finally,the cache model based on deep reinforcement learning was analyzed and the next research direction was proposed.
作者
毛鹏强
谢钧
夏士明
骆西建
MAO Pengqiang;XIE Jun;XIA Shiming;LUO Xijian(College of Command&Control Engineering,Army Engineering University of PLA,Nanjing 210007,China;Army Engineering University of PLA,Nanjing 210007,China)
出处
《陆军工程大学学报》
2022年第6期56-64,共9页
Journal of Army Engineering University of PLA
关键词
无线网络
边缘内容缓存
深度强化学习
缓存策略
wireless network
content caching on edge network
deep reinforcement learning
cache strategy
