Retrieval Delay Minimization for Reconfigurable Intelligent Surface Empowered Edge Caching

Edge caching can reduce system retrieval latency with effect, while reconfigurable intelligent surface (RIS) can repair losses caused by wireless propagation environments. In this paper, a RIS-empowered edge caching is proposed to minimize the content retrieval latency by optimizing the content placement vectors, transmit beamforming, and RIS's phase shift. To overcome the difficulties in coupled optimization variables and binary content placement, we decouple the underlying problem into two subproblems. Specifically, the mixed integer subproblem of caching strategy is relaxed as the binary constraints of content placement vectors and is solved via Karush-Kuhn-Tucker conditions. The minimum latency can be equivalently transformed as maximization of average sum rate, where the sum-of-logarithms is reformulated as a multiple ratio fractional programming problem, where the transmit beamforming is obtained by using quadratic transformation approach, and the passive beamforming is achieved via alternating direction method of multipliers alogrithm. Simulation results indicate the effectiveness of the RIS-aided edge caching network in reducing system retrieval delay.

I. Introduction

The mobile edge computing (MEC) is considered a novel paradigm for data processing and computing to support dra-matically increased amount of flowing data [1]–[3]. Edge caching is implemented by deploying the MEC server [4], which stores frequently accessed, data and content closer to the user to relieve the backhaul congestion and content retrieval latency. As a result, when a user requests data or content, edge caching can quickly respond and return it to the user.