Kite-Inspired Deadlock-Free Deploying for UAV Swarm-to-Swarm in GNSS-Denied Environment

In a global navigation satellite system (GNSS) denied environment, unmanned-aerial-vehicle (UAV) clusters and adversarial targets do not have precise positional information. In contrast, targets keep high mobility and try to hide from surveillance. These accumulated complex scenarios can cause difficulty in tracking, monitoring, and attacking (or rescuing) target swarms. To solve these problems, we organize the self-organizing UAV clusters to involve topology construction, routing, communication deployment, and collaborative control. We aim to reduce reliance on control bases and achieve coordinated completion of relevant tasks. We propose a series of task scheduling and UAV deployment strategies to address the dynamic changes of obscured target positions in real scenarios. Meanwhile, to confront poorly understood target swarms, we introduce an UAV network connectivity structure like a kite, which only fixes a single node - "the control base." Based on the distance between control bases and targets, we propose two modified classical evolution algorithms to search for the optimal solution set to expand the kite-like topology step by step. At the same time, a depth-first search-based unlocking algorithm is designed to unlock the target matrix that satisfies the UAV model constraints while considering a new deadlock scenario of multitarget saturation task scheduling. Simulation results demonstrate that the proposed comprehensive multitask scheduling algorithm exhibits good convergence and provides high-efficiency online and offline solutions for deployment.