Range-based cooperative localization with single landmark

Abstract

We consider the problem of range-based cooperative localization for multiple agents in a GPS-denied environment. For the minimal reference purpose, there exists only one single landmark whose global position is known to a subset of agents. All agents which are equipped with Ultra Wideband (UWB) radars are able to measure the relative distance between themselves and their communicating neighbours. We propose two approaches for the purpose of cooperative localization. We first design a covariance intersection (CI) based estimator which needs the information exchange among the agents to enhance their localization precision and show that the estimate is stable under mild conditions. We also design an alternative approach based on estimate consensus which differs from CI approach that primarily focuses on covariance fusion. Simulation results show the effectiveness of both the proposed approaches.