Semantic Communication for Cooperative Perception Using HARQ

Cooperative perception, offering a wider field of view than standalone perception, is becoming increasingly crucial in autonomous driving. This perception is enabled through vehicle-to-vehicle (V2V) communication, allowing connected automated vehicles (CAVs) to exchange sensor data, such as light detection and ranging (LiDAR) point clouds, thereby enhancing the collective understanding of the environment. Based on the semantic communication system for cooperative perception developed in our previous work, we investigate the incorporation of the hybrid automatic repeat request (HARQ) mechanism to further improve the perception accuracy, especially in the low signal-to-noise ratio (SNR) regime. We develop a novel semantic error detection method to accurately identify errors in the task-oriented communication process, which is key to determining if retransmission is warranted in the HARQ mechanism. Simulation results demonstrate that our proposed HARQ methods significantly enhance perception performance compared to that without HARQ, especially in the low SNR regime. Furthermore, in terms of throughput, our proposed HARQ methods exhibit superior efficiency compared to conventional schemes.