Collaboration in the Sky: A Distributed Framework for Task Offloading and Resource Allocation in Multi-Access Edge Computing

Recently, unmanned aerial vehicles (UAVs)-assisted multi-access edge computing (MEC) systems emerged as a promising solution for providing computation services to mobile users outside of terrestrial infrastructure coverage. As each UAV operates independently, however, it is challenging meet the demands due limited capacity at UAV’s MEC server well energy constraint. Therefore, collaboration among UAVs needed. In this article, collaborative multi-UAV-assisted system integrated with an MEC-enabled base station (BS) proposed. Then, problem minimizing total latency experienced by in proposed studied optimizing offloading decision allocation communication and resources while satisfying constraints both UAVs. The shown be nonconvex, mixed-integer nonlinear programming (MINLP) that intractable. formulated decomposed into three subproblems: 1) tasks problem; 2) resource 3) UAV-assisted problem. Lagrangian relaxation alternating direction method multipliers (ADMMs) methods are applied solve problems, alternatively. Simulation results show approach reduces average up 40.7% 4.3% compared greedy exhaustive search methods.