Multi-Step Knowledge-Aided Iterative Conjugate Gradient for Direction Finding

Abstract

In this work, we propose a Krylov subspace-based algorithm for direction-of-arrival (DOA) estimation, referred to as multi-step knowledge-aided iterative conjugate gradient (CG) method (Multi-Step KAI-CG), which achieves more accurate estimates than those of prior work. Differently from existing knowledge-aided methods, which make use of available known DOAs to improve the estimation of the covariance matrix of the input data, the proposed Multi-Step KAI-CG exploits knowledge of the structure of the covariance matrix and its perturbation terms and the gradual incorporation of prior knowledge, which is obtained on line. Simulation results illustrate the improvement achieved by the proposed method and the influence of iterations on its performance.