Joint task scheduling and multi-UAV deployment for aerial computing in emergency communication networks

This article studies mobile edge computing technologies enabled by unmanned aerial vehicles (UAVs) in disasters. First, considering that the ground servers may be damaged in emergency scenarios, we proposed an air-ground cooperation architecture based on ad-hoc UAV networks. We defined the system cost as the weighted sum of task delay and energy consumption because of different delay sensitivity and energy sensitivity tasks in emergency communication networks. Then, we formulated the system cost-minimization problem of task scheduling and multi-UAV deployments. To solve the proposed mixed integer nonlinear programming problem, we decomposed it to two sub-problems that were solved by proposing a swap matching-based task scheduling sub-algorithm and a successive convex approximation-based multi-UAV deployment sub-algorithm. Accordingly, we propose a joint optimization algorithm by iterating the two sub-algorithms to obtain a low complexity sub-optimal solution. Finally, the simulation results show that (i) the proposed algorithm converges in several iterations, and (ii) compared with the benchmark algorithms, the proposed algorithm has better performance of reducing task delay and energy consumption and achieves a good trade-off between them for diverse tasks.