Completion Time Minimization for Content Delivery in Cache-Enabled Multi-UAV Networks

This paper studies the completion time minimization of unmanned aerial vehicle (UAV) for content delivery in cache-enabled multi-UAV networks. Due to the limited cache space of UAV, uncached contents need to be fetched from the ground base station (GBS) via wireless backhaul, which causes additional delay to mission completion. To reduce the content delivery time, we formulate a non-convex minimization problem by optimizing cache placement, content scheduling, bandwidth allocation and UAVs' trajectories, which, however, is challenging to tackle. To handle this, the problem is decomposed into four subproblems, and we prove that the subproblem of cache placement is NP-hard for the sake of seeking the appropriate solving methods. Then block coordinate descent (BCD) and convex approximation is used to obtain a suboptimal solution to the primal problem. Numerical results show that our algorithm effectively decreases the completion time of content delivery and outperforms the benchmark schemes.