Efficient Wideband DoA Estimation with a Robust Iterative Method for Uniform Circular Arrays

TL;DR

This paper introduces a robust iterative wideband 2D DoA estimation method for uniform circular arrays, improving accuracy and efficiency by focusing on partial frequency points and constrained MUSIC spectrum, with performance benchmarks derived.

Contribution

The paper presents a novel iterative coherent signal-subspace method for wideband 2D DoA estimation that enhances robustness and efficiency over existing techniques.

Findings

Better estimation accuracy than benchmark methods

Improved efficiency in wideband DoA estimation

Performance close to the Cramér-Rao bound

Abstract

Direction-of-arrival (DoA) is a critical parameter in wireless channel estimation. With the ever-increasing requirement of high data rate and ubiquitous devices in wireless communication systems, effective wideband DoA estimation is desirable. In this paper, an iterative coherent signal-subspace method including three main steps in each iteration is proposed for wideband two-dimensional (2D) DoA estimation with a uniform circular array. The first step selects partial frequency points for the subsequent focusing process. The second step performs the focusing process, where the angle intervals are designed to generate focusing matrices with robustness, and the signal-subspaces at the selected frequency points are focused into a reference frequency. The third step estimates DoAs with the multiple signal classification (MUSIC) algorithm, where the range of the MUSIC spatial spectrum is constrained by the aforementioned angle intervals. The key parameters of the proposed method in the current iteration are adjusted based on the estimation results in the previous iterations. Besides, the Cram\'er-Rao bound of the investigated scenario of DoA estimation is derived as a performance benchmark, based on which the guidelines for practical application are provided. The simulation results indicate the proposed method enjoys better estimation performance and preferable efficiency when compared with the benchmark methods.