Resource Allocation for Energy Efficient STAR-RIS Aided MEC Systems

Abstract

In this letter, we investigate a novel simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) aided mobile edge computing (MEC) system. Our object is to minimize the sum energy consumption of all users by jointly optimizing the transmission and reflection time and coefficients of the STAR-RIS, as well as the transmit power and the size of offloading data for each user. Since the formulated optimization problem is non-convex, we decompose it into two subproblems, and then solve the subproblems alternately to obtain the solution. Simulation results demonstrate that users prefer to offload more task bits with the deployment of STAR-RIS. Further, the STAR-RIS in MEC outperforms conventional reflecting/transmiting-only RIS in terms of sum energy consumption, and the performance gain is more significant for a larger amount of task-input bits.