217 / 2020-01-02 06:12:00

Cognitive Antenna Selection for DOA Estimation with BZB Guided Multi-Step Planning

Bobrovsky-Zakai bound (BZB); CLEAN; cognitive radar; multi-step planning; multiple signal classification (MUSIC)

IEEE SAM 2020: Special Sessions > Automotive Radar Sensing

Cognitive antenna selection approach for direction of arrival (DOA) estimation based on conditional Bobrovsky-Zakai bound (BZB) has shown superior accuracy over non-adaptive approaches in automotive radar applications. However, its high computational complexity bounded the method to use a single-target model and to select the antenna array switching matrix in a greedy manner. The current work extends this approach to a multi-target model using BZB-CLEAN method, which significantly improves the receiver dynamic range at feasible computational complexity. Furthermore, this work introduces a multi-step planning method which is implemented using trajectory search-trees with BZB-based pruning. It is shown via simulations that the proposed approach outperforms other tested switching matrix selection methods in terms of DOA estimation error cumulative distribution and convergence rate.