Computing Delay Minimization for UAV-enabled Mobile Edge Computing Systems with URLLC-based Offloading

Unmanned aerial vehicle (UAV)-enabled mobile edge computing (MEC) systems can improve their edge computing performance by taking the advantage of flexible deployment of UAVs and the ability to establish high-quality communication links between UAVs and ground users. The existing works on UAV-enabled MEC systems usually assume that the blocklength of the offloading transmission is long, and their results cannot be directly applied to MEC scenarios with strict requirements on computing delay. A UAV-enabled MEC system with ultra-reliable and low-latency communication (URLLC)-based task offloading is considered, in which a UAV carrying a computing server provides MEC service for multiple ground users and the users offload parts of their computing tasks to the UAV through URLLC. The UAV's deployment location, the offloading bandwidths of the users, and the computing central processing unit (CPU) frequencies of the UAV and users are jointly optimized to minimize the computation delay of the system. To solve the resulting non-convex optimization problem, the block coordinate descent method is used to decompose the problem into two subproblems that optimize the UAV's deployment location, and the offloading bandwidths and CPU frequencies, respectively, and the two subproblems are solved alternately. In solving the two subproblems, logarithmic functions are used to make nonlinear approximation to the expression of URLLC offloading rate to simplify the subproblems, and the two subproblems are both transformed into convex optimization problems by applying the successive convex approximation method. Simulation results show that the proposed algorithm can effectively balance the communication capability and computing capability of the system and reduce the system’s computing delay compared to other benchmark schemes.