Multipath Ghost Suppression Based on Doppler Velocity Estimation in Traffic Scenarios

Millimeter-wave (MMW) radar has been widely applied in road traffic target detection and tracking. However, targets and strong reflective surfaces such as static road guardrails can lead to multipath ghosts during the propagation of electromagnetic (EM) waves in these complex scenarios. The multipath phenomenon causes interference and many errors. To reduce the probability of false alarms for real target detection, we propose a method based on radar point clouds for multipath ghost recognition and suppression. This method utilizes a multipath propagation geometric model and reflection boundary information to obtain the Doppler velocity of multipath ghosts and implement velocity filtering. Furthermore, experimental results shows that our algorithm can efficiently recognize and suppress multipath ghosts in complex road scenarios with multiple targets, significantly reducing the false alarm rate of real target detection.