Cooperative Perception in Connected Vehicle Traffic under Field-of-View and Participation Variations

Connected vehicles can improve on their perception of their environment beyond their local sensors' field-of-view (FoV) by constructing and maintaining state estimates of out of FoV objects from communicated information. This paper outlines a decentralized cooperative perception framework with a vehicle-to-vehicle (V2V) communication model that considers intermittent communications as well as with a model that takes into account dynamic changes in an individual vehicle's sensors' FoV in accordance with the prevailing traffic conditions. In this framework, each vehicle asynchronously broadcasts its local perception consisting of estimates of the motion states of self and neighboring vehicles along with the corresponding uncertainty measures of the estimates. Each participating receiver vehicle acts as an independent fusing node where data association and fusion occur. To evaluate the performance of the proposed cooperative perception scheme, we adapt and use the optimal subpattern assignment (OSPA) metric. Ground truth data were obtained from microscopic traffic simulation of multi-lane highway scenarios at different traffic conditions and varying rates of participation of the vehicles in the cooperative perception scheme. The results showed that the proposed cooperative perception scheme achieves the lowest OSPA error at moderate participation rates. In our settings, the minima occur near 30% participation for moderate to high traffic flow settings, and the OSPA error seems insensitive to traffic density near this minimum at all traffic flow settings tested.