Cooperative navigation for UAVs in GNSS-denied area based on optimized belief propagation

Abstract

A multitude of individual unmanned aerial vehicle (UAV) cooperatively self-organized as a swarm attracts increasing attention for exploration and parallel tasks. Accurate position is essential for UAVs to avoid collision and implement control. Although the Global Navigation Satellite System (GNSS) has been integrated in UAV, it is susceptible to disruption. In this paper, we propose a novel cooperative navigation framework based on optimized belief propagation (BP) in the GNSS-denied area. Specifically, we derive a directional uncertainty model to characterize the position information of UAVs with different dynamic motions. With this model, we improve the navigation performance by designing a weighted uncertainty belief propagation (WUBP) based on information evaluation. Simulation and experimental results validate that in the dynamic scenarios, the proposed WUBP can make the swarm reap the benefits of more accurate positional estimate from UAVs with less maneuvers, higher-grade of INS or shorter time interval of GNSS signal lost.