UC Davis

Millimeter wave (mmWave) sensing has emerged as a powerful technology with a wide range of applications in areas that require high-resolution detection and precise distance measurements. Among various mmWave sensing techniques, Frequency Modulated Continuous Wave (FMCW) radar stands out due to its ability to accurately measure a target’s distance, velocity, and angle. FMCW radar is especially useful in environments where reliability, speed, and spatial awareness are critical, such as autonomous driving. However, given the high sensitivity of mmWave signals to environmental factors, careful optimization is essential to ensure mmWave sensing is effective for vehicular communication.

In this dissertation, we present solutions for enhancing mmWave sensing using FMCW radar in vehicular communication. First, our work aims to establish reliable vehicle-to-vehicle (V2V) links for 6G communication by utilizing FMCW radar to optimize a beamforming protocol for IEEE 802.11ad. Experimental results show that our proposed protocol significantly improves throughput performance compared to the default 802.11ad protocol in real-world scenarios. Secondly, we utilize multiple FMCW radars to improve vehicular sensing for cooperative perception. We introduce the first vehicular cooperative dataset based on radar point clouds, V2VRP, developed with a novel experimental setup to synchronize multiple sensors. This dataset fills a critical gap in the real-world V2V cooperative perception field. Finally, we demonstrate that V2VRP dataset can be effectively utilized by the existing deep learning model. Our preliminary result based on our real-world dataset shows the cooperative perception approach offers a 70% improvement in maximum 3D IoU over the single perception approach.