Base Station Social-Aware Caching Strategy for 5G Ultra Dense Networks

Abstract

The ultra dense network (UDN) is a promising technique to meet the requirements of the explosive data traffic in next generation (5G) wireless networks. As the small cells (SCs) are densely located in UDN, how to deploy cache resources on SCs to improve the network throughput and energy efficiency should be considered clearly. In this paper, a base station socialaware caching strategy for UDN is proposed. Firstly, we utilize the social network theory to study the social characteristics of base stations (BSs) as the traffic variation of BSs has an obvious pattern in the temporal-spatial and content domain, and take advantage of the social-tie factor (STF) modeled in our previous study which reflects the social characteristic of the correlation of traffic variation among BSs. A few very important base stations (VIBS) are selected with higher average STF values and the rest SCs are divided into two groups according to the STF with VIBSs: normal SCs with higher STF which can be served by VIBSs and unique SCs with lower STF that shares limited caching capacity and backhaul capacity with VIBSs. Afterward, the network throughput, power consumption, energy efficiency (EE) and network latency are analyzed as performance indicators with the tool of stochastic geometric process. By optimizing the number of the selected VIBSs and the value of STF threshold, our proposed caching strategy can always achieve better performance compared with two baselines. With our social-aware based caching strategy, the maximum gain on the network throughput can achieve approximately 83% and 33% respectively.