Infrastructure-free cooperative relative localization for UAVs in GPS-denied environments

Relative positions of unmanned aerial vehicles (UAVs) play a critical role in various multi-agent control applications including swarm and formation control in many real environments where global positioning signals are blocked, and predefined infrastructures are not allowed. In this paper, we focus on the problem on how to realize the inter-UAV relative localization (RL) of a UAV team without using the communicated global position information in the infrastructure-free case. With the aid of ultra-wideband modules, each UAV is able to measure the relative distances to its neighbours and obtain the neighbour's local information. To solve the relative localization problem, a direct relative localization method is proposed first to estimate the relative positions to the directly perceptible neighbours. Benefitting from the sensing topology, a consensus-based cooperative RL approach is presented to fuse the direct and indirect RL estimates to enhance the system robustness. Numerical simulations and flight tests have demonstrated the performance of the proposed strategy.