基于内容流行度的卫星网络缓存策略研究  

Research on Satellite Network Caching Strategy Based on Content Popularity

作　　者：刘治国[1,2] 李怡帆[1,2] 潘成胜 LIU Zhi-guo;LI Yi-fan;PAN Cheng-sheng(School of Information Engineering,Dalian University,Dalian Liaoning 116600,China;Key Laboratory of Communication and Network,Dalian University,Dalian Liaoning 116600,China;School of Electronics and Information Engineering,Nanjing University of Information Science and Technology,Nanjing Jiangsu 211800,China)

机构地区：[1]大连大学信息工程学院,辽宁大连116600 [2]大连大学通信与网络重点实验室,辽宁大连116600 [3]南京信息工程大学电子与信息工程学院,南京211800

出　　处：《计算机仿真》2024年第4期50-54,共5页Computer Simulation

基　　金：国家自然科学基金资助项目(61931004)。

摘　　要：针对传统卫星网络缓存策略中只考虑全局内容流行度未考虑到不同区域流行内容差异较大,从而无法准确找到高流行度内容进行放置导致用户获取时延较大的问题,提出了一种基于全局流行度预测与局部流行度计算的混合卫星网络缓存策略。策略使用LSTM模型对全局内容流行度进行预测,并且对局部内容流行度进行计算,同时将卫星缓存节点分区,定义缓存节点分区比例因子,分别缓存全局和局部流行内容,以最小化用户内容平均获取时延为优化目标,采用BPSO求解缓存放置问题。仿真结果表明,与传统的随机缓存放置及最流行缓存放置相比,所提出的策略能够降低用户内容获取时延,提高缓存命中率。Aiming at the problem that the traditional satellite network caching strategy only considers the global content popularity,and does not take into account the large differences in popular content in different regions,so that the high popularity content cannot be accurately found and placed,resulting in a large user acquisition delay.A satellite network caching strategy based on a mixture of global popularity prediction and local popularity calculation is proposed.This strategy uses the LSTM model to predict the popularity of global content,and calculate the popularity of local content,partition satellite cache nodes,define the partition scale factor of cache nodes,cache global and local popular content respectively,minimize the average acquisition delay of user content as the optimization goal,and solve the cache placement problem by using BPSO.Simulation results show that compared with the traditional random cache placement and the most popular cache placement,the proposed strategy can reduce the user content acquisition latency and improve the cache hit ratio.

关 键 词：卫星网络 流行度预测 缓存策略 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]