Bayesian Multipath Mitigation Technique With Fully Digital Phased-Array Radar

Abstract

Multipath propagation represents a major source of bias in angle-of-arrival estimation for pulsed radar systems, particularly in low-elevation scenarios. Classical techniques, such as monopulse processing, are known to exhibit systematic errors in the presence of multipath reflections. This article proposes a Bayesian antimultipath approach that jointly estimates the angle of arrival of the target of interest and the angle of arrival of the multipath component, explicitly exploiting a spherical Earth geometry and a four-ray multipath propagation model. In addition to angle estimation, the proposed method enables the separate estimation of the received signal power associated with aggregated signal contributions arriving from above and below the radar horizon. The performance of the estimator is evaluated through numerical simulations and compared with classical monopulse processing and the complex angle (CA) technique under realistic multipath conditions. Results show that the proposed Bayesian estimator substantially reduces the bias in the elevation error in the presence of multipath, whereas monopulse-based angle estimates and CA estimates exhibit significant bias or instability, demonstrating the effectiveness of the proposed approach for robust angle estimation in multipath environments.