Adaptive Cache Server Selection and Resource Allocation Strategy in Mobile Edge Computing

Abstract

The enormous increase of data traffic generated by mobile devices emanate challenges for both internet service providers (ISP) and content service provider (CSP). The objective of this paper is to propose the cost-efficient design for content delivery that selects the best cache server to store repeatedly accessed contents. The proposed strategy considers both caching and transmission costs. To achieve the equilibrium of transmission cost and caching cost, a weighted cost model based on entropy-weighting-method (EWM) is proposed. Then, an adaptive cache server selection and resource allocation strategy based on deep-reinforcement-learning (DRL) is proposed to place the cache on best edge server closer to end-user. The proposed method reduces the cost of service delivery under the constraints of meeting server storage capacity constraints and deadlines. The simulation experiments show that the proposed strategy can effectively improve the cache-hit rate and reduce the cache-miss rate and content access costs.