Energy Optimization for Cellular-Connected Multi-UAV Mobile Edge Computing Systems with Multi-Access Schemes

In this paper, a cellular-connected unmanned aerial vehicle (UAV) mobile edge computing system is studied where several UAVs are associated to a terrestrial base station (TBS) for computation offloading. To compute the large amount of data bits, a part of computation task is migrated to TBS and the other part is locally handled at UAVs. Our goal is to minimize the total energy consumption of all UAVs by jointly adjusting the bit allocation, power allocation, resource partitioning as well as UAV trajectory under TBS's energy budget. For deeply comprehending the impact of multi-UAV access strategy on the system performance, four access schemes in the uplink transmission is considered, i.e., time division multiple access, orthogonal frequency division multiple access, one-by-one access and non-orthogonal multiple access. The involved problems under different access schemes are all formulated in non-convex forms, which are difficult to be tackled optimally. To solve this class of problem, the successive convex approximation technique is employed to obtain the suboptimal solutions. The numerical results show that the proposed scheme save significant energy consumption compared with the benchmark schemes.