Spatial And Temporal Smoothing For Covariance Estimation In Super-Resolution Angle Estimation In Automotive Radars

Introduction of Digital Coded Modulation (DCM) radars in automotive applications has allowed large scale MIMO systems to be feasible within the operating cost and space constraints. Even with such large number of transceivers and intelligent array design, it is often necessary to employ super-resolution angle estimation methods like the Multiple Signal Classification (MUSIC) algorithm to resolve closely spaced targets. These methods require accurate estimation of the signal covariance across channels. A further complication arises in automotive radar applications due to the fact that this covariance matrix estimation needs to be done using single snapshot. To mitigate this problem, it is possible to make use of spatial and temporal smoothing. In this paper, we analyze the effect covariance matrix averaging in spatial and temporal domain on the angular resolution that can be obtained with MUSIC. We not only focus on traditional angular resolution with targets of same radar cross section (RCS) but also on high contract resolution (HCR) which analyzes the angular resolution in case of targets with large variation in target RCS. Our results indicate that proper trade-off is needed among these smoothing techniques to get the best angular resolution depending on the choice of antenna configurations.