北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2020, Vol. 43 ›› Issue (1): 80-91.doi: 10.13190/j.jbupt.2019-054

• 论文 • 上一篇    下一篇

基于本地内容流行度预测的主动缓存策略

任佳智, 田辉, 聂高峰   

  1. 北京邮电大学 网络与交换技术国家重点实验室, 北京 100876
  • 收稿日期:2019-04-10 出版日期:2020-02-28 发布日期:2020-03-27
  • 通讯作者: 田辉(1963-),女,教授,博士生导师,E-mail:tianhui@bupt.edu.cn. E-mail:tianhui@bupt.edu.cn
  • 作者简介:任佳智(1987-),男,博士生.
  • 基金资助:
    国家技术重大专项项目(2018ZX03001019-003);国家自然科学基金青年基金项目(61801044)

Proactive Caching Scheme with Local Content Popularity Prediction

REN Jia-zhi, TIAN Hui, NIE Gao-feng   

  1. State Key Laboratory of Networking and Switch Technology, Beijing University of Posts and Telecommunications, Beijing 100876, China
  • Received:2019-04-10 Online:2020-02-28 Published:2020-03-27
  • Supported by:
     

摘要: 已有边缘缓存技术研究假设已知固定的全局流行度,忽略了反映基站接收到的内容请求历史中的流行度地域差异特性和动态特性,为此,提出了一种基于本地内容流行度预测的内容部署策略.首先,考虑流行度的地域特性,将内容请求历史记录相似的小基站分簇;然后,使用线性回归方法预测每个小基站簇群的本地内容流行度,基于预测的本地内容流行度,利用随机几何和凸优化理论求得次优内容部署决策;最后,基于真实数据集的实验验证了所提算法性能以及相应的缓存系统性能.仿真结果表明,所提算法优于对比算法的缓存命中率性能.

关键词: 内容部署, 本地流行度预测, 余弦相似性

Abstract: Considering the problem that most works on content placement so far consider global popularity, neglecting the demand difference between base stations (BSs), a content placement scheme based on similarity between small base stations (SBSs) and local content popularity prediction considering popularities' geographical diversity are proposed. Firstly, SBSs that possess similar historical content requests is identified by similarity measurements. Then the probabilities of future requests are predicted for each similar SBS group utilizing linear regression method. Based on this local popularity, the sub-optimal content placement decision is made according to stochastic geometry and convex optimization. Thereafter, real data sets to verify our prediction algorithm and investigate system performance are used. It is shown that the proposed scheme outperforms the comparison schemes in terms of hit ratio.

Key words: content placement, local popularity prediction, cosine similarity

中图分类号: