Self-Organizing UAV Swarm Placement via Layered Loose Coupling and User Prioritization

In this paper, we focus on a self-organizing horizontal placement problem of unmanned aerial vehicles (UAVs) which offer network coverage to selected user devices (UDs) in large areas. Specifically, we apply a virtual force field (FF) concept to allow a swarm of UAVs to autonomously locate themselves in proximity to UDs. Unlike the previous work on the FF approach for the selected scenario, we address the incomplete connectivity of network nodes in terms of temporal and spatial couplings under the condition where the UD associations need to be prioritized. We propose a layered FF approach that combines two loosely coupled but complementary FFs. These include a remote FF (RFF) built upon outdated network status information of the whole network and a proximate FF (PFF) that relies on the prompt status of a local subset of nodes in proximity to a particular UAV. We also derive balancing conditions for both RFF and PFF that prevent the UAV swarm to escape from the vicinity of prioritized UDs. The simulation results first characterize the ability of the PFF and RFF to approach the prioritized UDs and then demonstrate the potential gain to improve the coverage probability using a layered FF structure.