Joint Task Offloading and Resource Allocation for RIS-assisted UAV for Mobile Edge Computing Networks

Unmanned aerial vehicles (UAVs) can assist mobile edge computing (MEC) networks to enhance ground users’ communication in emergencies. However, how to provide good quality of service (QoS) to ground users for a long time is still a tricky problem due to the limited battery capacity and computing power of UAVs. Therefore, we take advantage of reconfigurable intelligent surface (RIS) for UAV-MEC networks and propose a joint task offloading and resource allocation strategy. The strategy aims to minimize the energy consumption of the UAV by jointly optimizing task offloading decisions, allocation of UAVs’ computing resources, communication resource allocation, and phase shift of RIS. Considering the non-convex optimization and computational complexity of the above optimization problem, we first model the target problem using the Markov decision process (MDP) and then solve it efficiently using a double deep Q network (DDQN). Simulation results show that our proposed solution outperforms other benchmark test solutions.

I. Introduction

Mobile edge computing (MEC) networks [1] have been considered a promising key technology to enhance the quality of service (QoS) for computation-intensive and latency-critical tasks. A MEC network equipped with UAVs can take full advantage of the flexible deployment of UAVs [2] to provide extended communication coverage and reliable connectivity to mobile users promptly, especially when terrestrial communication infrastructure is destroyed or massive users proliferate in a short time. However, due to the limitations of UAVs’ energy and computing resources, the deployment of more UAVs requires high cost and power consumption in MEC networks. Fortunately, the emerging new technique called reconfigurable intelligent surface (RIS) is able to provide an energy-efficient alternative to enhance the network capacity.