Learning based Content Caching for Wireless Networks
2020 International Conference on Information Networking (ICOIN)(2020)
摘要
In recent years, the proliferation of smart mobile devices and multimedia applications has resulted in explosively increasing multimedia traffic over next-generation wireless networks. To tackle such an explosive traffic load, offloading network traffic and proactively caching contents at the edge of the network i.e., small BS (SBS) has been envisioned as an effective solution. However, proactive caching requires accurate predictions on user requests where it is unlikely to provide accurate predictions on dynamic content requests in highly mobile wireless networks. To alleviate this problem, we propose a learning-based content caching scheme, where each small BS learns space-time popularity dynamics by exploiting a multi-armed bandit learning agent in a distributed manner. The small BS determines popularity of requested contents using Zipf's distribution, shares its local popularity knowledge with immediate connected (i.e., one hop neighbor) small BSs, and then performs learning to estimate space-time popularity dynamics of contents for proactive caching based on the long-term observations of the requests.
更多查看译文
关键词
content caching,wireless cellular networks,multi-armed bandits
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络