Direction Finding in Partly Calibrated Arrays With Hybrid Imperfections

Abstract

Direction finding in partly calibrated arrays has received significant attention in recent years. From the perspective of the calibration imperfections, existing literature predominantly focuses on inter-subarray displacement, while there has been relatively limited discussion regarding subarray inclination. This article introduces a comprehensive signal model that accounts for hybrid imperfections, including inter-subarray displacement, gain/phase uncertainty, and particularly subarray inclination. Building on this, we propose a gridless two-stage method based on block-structured steering matrix recovery for direction finding, termed BSSMR. Initially, we leverage the covariance matrix of the measurements to recover the steering matrix, followed by the estimation of direction-of-arrivals (DOA) based on the block-structured array manifold. We provide a sufficient condition for BSSMR and conduct an analysis of its asymptotic convergence behavior. Numerical simulations demonstrate the robust performance of our proposed method in scenarios of hybrid imperfections, which reaches the Cramér-Rao lower bound (CRLB).