Distributed relative cooperative positioning in Vehicular Ad-Hoc Networks

Abstract

We propose a distributed Relative Cooperative Positioning algorithm (ReCoP) that computes the position in relation to other vehicles in the network. ReCoP does not depend on fixed reference node/s or infrastructure, and it is independent of Global Navigation Satellite System (GNSS) reading. Instead one node in each group of one-hop connected neighbours establishes relative map for group members. Then when messages (e.g., warning messages) arrive to the vehicle with gateway responsibility, the coordinates transform to be recognized by different relative maps. This approach can be utilized in data dissemination to reduce broadcast storm problem due to redundant retransmissions. It has been compared with the Local Self Positioning (LSP) in which each vehicle individually builds its own local relative map. The performance of the proposed algorithm, ReCoP, has been evaluated with respect to different traffic density, transmission range, and speed. The simulation results illustrate that ReCoP computes the relative position with more precision than LSP. It also demonstrates the scalability, robustness, and flexibility of the proposed ReCoP compared to LSP.