On message filtering for cooperative localisation of vehicles in an urban environment

In this paper we present cooperative localisation of vehicles within an urban environment using short-range radio communication technologies such as Dedicated Short Range Communication (DSRC) and IEEE 802.11 (WiFi) etc. In our application environment vehicles are moving along streets and the geometric combination that they form is mostly degenerate for the purpose of position estimation. We explore different scenarios where incorporation of information received from all other vehicles might be suboptimal for the purpose of position estimation. We calculate a confidence measure based on the geometric configuration of different sub-groups of vehicles and propose that the position estimates from each sub-group be weighted accordingly to arrive at the final estimate. Simulation results illustrate that the proposed method can effectively filter out degenerate configurations and achieve considerable performance gain over standard averaging of position estimates.