Event-Triggered Swarm Optimization of Formation Control Policies for Unmanned Aerial Vehicle Networks with Limited Bandwidth

Abstract

Unmanned aerial vehicle networks are increasingly deployed for persistent sensing, inspection, and emergency response in environments where wired infrastructure is unavailable and wireless bandwidth is scarce. Formation control is a practical coordination pattern that simplifies collision avoidance and coverage, yet it imposes continual exchange of state estimates that can saturate shared links. Event-triggered approaches reduce communication by transmitting only when local errors cross meaningful thresholds. Swarm optimization strategies, inspired by collective search and assignment dynamics, offer a decentralized way to tune these thresholds and policies without relying on a single coordinator. This paper explores event-triggered swarm optimization for formation control policies in bandwidth-limited UAV networks. The approach integrates local triggers for state and topology variation with population-based optimization that adapts controller gains, trigger sensitivities, and role assignments to prevailing channel load and motion dynamics. The study formulates a graph-based model for formation maintenance with explicit radio constraints, introduces an event mechanism that couples estimation and control errors to message release conditions, and develops a swarm-driven tuner that searches viable parameter regions under changing wind profiles and packet loss. The analysis discusses stability and performance in terms of formation error, message rate, and resilience to link dropouts, and articulates design choices that trade communication savings for transient overshoot and settling time. The results emphasize neutral system properties and practical guidelines for applying event-triggered swarm tuning in multi-tier aerial networks that must operate with limited bandwidth and intermittent connectivity.