Joint Optimization of Energy and Task Scheduling in Wireless-Powered IRS-Assisted Mobile-Edge Computing Systems

Abstract

This paper studies the design of an intelligent reflecting surface (IRS)-assisted mobile edge computing (MEC) system, which consists of a mobile user, an IRS equipped with radio-frequency (RF) energy harvesting (EH) circuits, and a hybrid access point (HAP) connected with an MEC server. The IRS is deployed to reflect the users task offloading signals to enhance the received power at HAP, and it needs to harvest energy from the RF signals emitted by the HAP before reflecting signals. To save energy consumption at the user, we first propose a novel MEC protocol, in which the system is enabled to operate in three modes, namely an EH mode, an IRS-assisted task offloading mode, and an IRS-inactive task offloading mode, so that the energy at IRS and the tasks generated at user can be flexibly scheduled within a finite time horizon, depending on channel conditions, IRS energy states and users task queue states. Under the protocol, we optimize the operation mode selection and resource allocation in each mode with a task execution delay constraint. Due to the randomness of wireless channels and task arrivals, the optimization problem is a stochastic programming. To solve this problem, we first transform it into a deterministic one by assuming that noncausal channel state information (CSI) and task state information (TSI) are available, and then derive a practical algorithm where only causal CSI and TSI are required. Simulation results verify that our proposed design can save at most 80% energy consumption as compared with existing baseline schemes.