Task-Driven Delay Minimization for AAV-Assisted Mobile Crowdsensing Networks: A Joint Optimization Approach

In this work, we investigate a task-driven delay minimization problem for autonomous aerial vehicle (AAV) enabled mobile crowdsensing (MCS) networks. Our focus is to reduce overall latency and improve data collection efficiency for delay-sensitive tasks through jointly optimizing the sensing data size, bandwidth allocation, and AAV hovering position. The formulated problem is a mixed-integer programming problem, which is also nonconvex. We introduce an efficient alternating optimization algorithm to address this challenge. Specifically, the original minimization problem is decomposed into two subproblems. The first subproblem focuses on bandwidth and sensing data allocation. By leveraging the latent structure property of the problem, we derive the optimal sensing data allocation given a bandwidth allocation policy. Based on it, we reveal that the first optimization subproblem can be converted into a maximum weighted matching problem in a bipartite graph, which can be optimally solved using the Hungarian algorithm. To address the optimization of the AAV’s hovering position subproblem, we employ the successive convex approximation (SCA) technique, which transforms it into a convex problem that can be efficiently solved by standard convex optimization solvers. We also analyze the convergence and the time complexity for the developed joint optimization algorithm. Afterward, we approximate the global optimal solutions in closed form for several specific cases of the problem. Extensive simulations confirm the superior performance of our proposed scheme compared to various benchmark strategies in terms of both delay and energy consumption.