LAUSR

This paper establishes a new unmanned aerial vehicle (UAV) swarm-enabled wireless inland ship mobile edge computing (MEC) network with time windows, where UAVs are deployed to support time-constrained and low-resource unmanned surface vehicles (USVs) communication and computation. We aim to minimize UAV swarm energy consumption subject to numerous constraints. To tackle the challenging formulated problem, we decompose the original challenging problem into a list of sub-problems according to block coordinate descent (BCD) method and propose a heuristic algorithm to solve subproblems in an iterative manner. Specifically, the optimization of USVs offloading decisions is formulated as mixed-integer non-linear programming and then a modified deferred acceptance algorithm is proposed. The joint consideration of UAVs flight speed and UAV swarm stability is converted into a single-objective optimization problem after determining UAV control system’s latency requirements, and then Lagrangian multiplier method with Karush-Kuhn-Tucker (KKT) conditions is proposed. Successive convex approximations (SCA)-based algorithm is proposed to convert the non-convex constraints to convex ones, and then UAVs hovering coordinates can be obtained by solving the convex approximation problem. Numerical results verify the effectiveness of jointly considering USVs offloading decisions, UAVs flight speed, UAVs horizontal hovering coordinates and the number of UAVs for energy efficiency in comparison with several advanced baseline algorithms without optimization design.