High resolution automotive imaging using MIMO radar and Doppler beam sharpening
Scott L. Cassidy, Sukhjit Pooni, M. Cherniakov, E. G. Hoare, Marina Gashinova
Abstract
Highly detailed sensing of a vehicle's surrounding environment is a key requirement for the advancement of autonomous driving technology. While conventional automotive radar sensors remain robust under challenging weather conditions, poor cross-range resolution and high sidelobe levels present significant challenges. In this paper we propose an approach which combines MIMO beamforming with Doppler beam sharpening. We demonstrate a significant improvement in terms of cross-range resolution and, importantly, nearly 20 dB sidelobe suppression compared to conventional MIMO processing. This approach is investigated in detail and validated through theoretical analysis, simulation and experiment using data recorded on a moving vehicle. We demonstrate performance which is comparable to a high resolution mechanically scanned radar, using a commercially available MIMO sensor.