A Q-Learning Based Scheme to Securely Cache Content in Edge-Enabled Heterogeneous Networks

Abstract

With the increasing demands for mobile users to obtain content, the heterogeneous networks (HetNets), which consolidate the backbone networks with mobile edge networks, have been regarded as a promising paradigm to provide mobile users with high quality of experience (QoE). However, the limited network resources and caching capacity become a new challenge to deliver content in HetNets. Therefore, in this paper, a cooperative scheme between edge server and content provider in HetNets is proposed to improve the performance of content delivery. Firstly, a novel framework of content delivery with backbone networks and mobile edge networks is introduced. The edge servers are deployed at the edge of networks and close to users. Secondly, a Q-learning based scheme for content caching is developed to securely cache contents with the cooperation between edge server and content provider. Thirdly, the cooperative interaction between edge server and content provider is modeled as the double auction game. Each player can obtain the maximum utility from the equilibrium strategy. Finally, simulation results show that the proposed scheme can improve the efficiency of content delivery and bring more utilities to edge server and content provider than the conventional schemes.