Channel phase calibration based on RARE in mode domain for direction of arrival estimation

A novel method is proposed to estimate direction of arrival (DOA) with uniform circular array (UCA) in the presence of phase errors. Inspired by the rank reduction (RARE) theory in the spatial domain, which employs the extra sensors to construct the necessary condition for RARE. i.e., rank reduction vector, a reduction in the radius of UCA is considered for the creation of the rank reduction vector that is only related to the DOA in the UCA mode domain. This indicates that the DOA and the channel phase errors coupling in the steering vector in the spatial domain can be decoupled in the UCA mode domain, and then the optimization function only related to the phase errors is formulated with the rank reduction vector based on the RARE theory employed in the UCA mode domain. Then with the channel phase errors calibrated, the DOA can be estimated via UCA-ESPRIT. The proposed method requires no redundant sensors or calibration sources. Also, it avoids the multi-dimensional iteration, which is superior to the Schur-product based method as it has a larger tolerance for the array aperture. Results verify its effectiveness.

Automated summary

A novel method is proposed to estimate direction of arrival (DOA) with uniform circular array (UCA) in the presence of phase errors. Inspired by the rank reduction (RARE) theory, this method utilizes a reduction in the radius of UCA to create a rank reduction vector that is only related to the DOA in the UCA mode domain. By decoupling the DOA and the channel phase errors in the steering vector, an optimization function only related to the phase errors is formulated. The proposed method requires no redundant sensors or calibration sources.