A Motion-Based Virtual Aperture Approach for Bistatic FMCW Automotive Radar

Frequency-modulated continuous wave (FMCW) automotive radar is a key sensing instrument in advanced driver assistance systems given its ability for accurate range, angle and Doppler estimation, low cost, and robustness to adverse weather conditions. Its compact bumper-mounted antennas, however, offer limited spatial resolution. This paper presents a motion-based virtual aperture (MBVA) technique that exploits the movement of a sensing vehicle in a bistatic FMCW configuration. By combining data collected from multiple consecutive coherent processing intervals (CPIs) as the vehicle moves, the proposed method synthesizes a much larger virtual array, thereby achieving higher angular resolution. Computer simulations show that this MBVA approach consistently outperforms state-of-the-art baselines with much improved angle estimation result.