Bayesian Learning for Sparse Parameter Estimation in OTFS-aided mmWave MIMO Radar Systems

This paper proposes an orthogonal time-frequency space (OTFS) modulation aided millimeter wave (mmWave) multiple-input multiple-output (MIMO) phased-array radar (OmM-PAR) system for sparse radar target parameter estimation. Initially, we derive the delay-Doppler (DD)-domain end-to-end input-output model for the OmM-PAR system, which employs a single RF chain (RFC) both at the radar transmitter and receiver (R-TRX). Subsequently, a Bayesian learning (BL)-based procedure is developed for improved sparse radar target parameter estimation. Finally, our simulation results illustrate the enhanced performance of the proposed parameter learning framework for OmM-PAR systems. Furthermore, the performance of the proposed scheme is also benchmarked against the Bayesian Cramer-Rao lower bounds (BCRLB).