Edge Computing and Wireless Power Transfer for Integrated Radar and Communication-Equipped IoT Systems

In this paper, we investigate the wireless power-enabled mobile edge computing (WP-MEC) for an integrated radar and communication (IRC)-equipped internet of things (IoT) system. The system allows an IoT device to harvest energy from a power station (PS) and offload its computation task to the PS. Furthermore, the system enables the IoT device to leverage the offloading bits for the radar tracking. Orthogonal frequency division multiplexing (OFDM) technique is used for the task offloading of the IoT devices. We aim to maximize computation efficiency over the IoT devices subject to their radar performance requirements by optimizing the energy transfer time, the OFDM subcarrier allocation to the devices, and the transmit power. Due to the stochastic and dynamic nature of the computing resource and the targets, we leverage a deep reinforcement learning (DRL) algorithm, namely Advantage Actor Critic (A2C), to solve the problem. Simulation results are provided to evaluate the effectiveness and improvement of the A2C algorithm.