Multi-Vehicle Tracking and ID Association Based on Integrated Sensing and Communication Signaling
Zhiqin Wang, Kaifeng Han, Jiamo Jiang, Fan Liu, Weijie Yuan
Abstract
The integrated sensing and communication (ISAC) technology, which performs vehicle state sensing and communications simultaneously, has been widely regarded as a key to enable future intelligent vehicular networks. However, the communications information is in general vehicle-specific, which requires the RSU to distinguish vehicle identities (IDs). Conventionally, the vehicles feed back their IDs to the RSU, since they are not contained in the sensing echoes. This letter develops a novel approach for multi-vehicle tracking and ID association using the ISAC signals. In particular, the ID association is done by comparing the similarity of the distributions of estimated and predicted locations based on the Kullback-Leibler divergence (KLD). With the aid of the proposed approach, frequent uplink ID feedback is avoided, leading to a reduced communication overhead as well as latency. We further design a high-efficiency predictive beamforming scheme, which predicts the angular parameters in the following time instant relying on the estimated states. Simulation results show that our proposed scheme can effectively associate different vehicle IDs and improve the communications performance.