Joint Trajectory and Computation Offloading Optimization for UAV-assisted MEC with NOMA

Abstract

Unmanned aerial vehicles (UAVs) offer cost-effective services in wireless communication with no or limited infrastructure. In this paper, a UAV-assisted mobile edge computing system is studied, in which UAV serves as a relay between the offloading users and the Base Station (BS). Furthermore, nonorthogonal multiple access (NOMA) is introduced to improve the spectrum efficiency. To enable effective computation offloading, the trajectory of UAV and multi-dimensional resource allocation play an important role in the offloading procedure. However, there still lack studies on this topic in such a network. This paper aims at joint trajectory and computation offloading optimization while considering the energy consumption of UAV and quality of services (QoS) of the users. This problem is formulated to minimize the total delay of all users. To enable tractable analysis, we first divide the original problem into two subproblems: joint optimization of trajectory and power allocation, and user schedule. Then the first problem is addressed by leveraging successive convex approximations. And the difference of convex programming theory is adopted to obtain a sub-optimal solution. With the obtained results, the user schedule indicator is finally optimized. Extensive simulation results are presented to demonstrate the effectiveness of the proposed solution.