SAR on the Wheels: High-Resolution Synthetic Aperture Radar Sensing on Moving Vehicles

Emerging autonomous vehicles require reliable perceptions of their surroundings at all times. Unfortunately, current mainstream sensors, i.e., camera and Lidar, are vulnerable under challenging lighting and weather conditions. In contrast, despite all-weather operations, today's vehicle Radars are limited to simple localization and speed detection due to their lower sensing resolution. In this paper, we consider enabling Synthetic Aperture Radar (SAR) imaging for vehicle Radars to advance the Radar sensing capability. Specifically, we analyze Radar signal variations caused by movement and realize Radar self-tracking at wavelength-scale precision for SAR. Our evaluation results show that the Radar self-tracking effectively combats motion uncertainties and enables SAR imaging with much higher resolution than existing vehicle Radar solutions.