[Paper Review] Analysis and Design of Nonuniform Arrays for Direction Finding

Elie Daher|arXiv (Cornell University)|Dec 7, 2018

Direction-of-Arrival Estimation Techniques3 citations

TL;DR

This paper proposes advanced non-uniform array designs and signal processing techniques for improved direction-of-arrival (DOA) estimation in both narrowband and wideband scenarios. It addresses key challenges like limited degrees of freedom (DOFs), coherent targets, and mutual coupling using sparsity-based methods, multi-frequency operation, and interpolation techniques, achieving enhanced resolution and robustness in practical array configurations.

ABSTRACT

The purpose of this research is to employ non-uniform arrays in different active and passive sensing applications for both narrowband and wideband operations, while providing a multitude of array processing methodologies that assist in dealing with the different encountered challenges. The problem of direction-of-arrival (DOA) estimation using non-uniform arrays is considered. The different challenges that are treated include the reduction of the available degrees-of-freedom (DOFs), the presence of coherent targets, and the mutual coupling effect in practical antenna arrays. Multi-frequency operation is exploited to increase the DOFs that are available for DOA estimation using both high-resolution subspace and sparse reconstruction techniques. In addition, a sparsity-based interpolation technique is presented to perform DOA estimation with increased DOFs. Moreover, a DOA estimation approach for a mixture of coherent and uncorrelated targets based on sparse reconstruction and active non-uniform arrays under narrowband signal platform is proposed. The aforementioned approaches deal with ideal operational scenarios. To address a more practical scenario, various methods for DOA estimation using non-uniform arrays in the presence of mutual coupling are presented. Extensive numerical simulations which validate the different proposed methods are also included.

Motivation & Objective

• To enhance DOA estimation performance in non-uniform arrays by increasing available degrees of freedom (DOFs) beyond uniform array limits.
• To address coherent target interference in DOA estimation using sparse reconstruction and active array configurations.
• To mitigate the practical impact of mutual coupling in real-world antenna arrays through novel calibration and processing techniques.
• To enable high-resolution DOA estimation in wideband and multi-frequency scenarios using non-uniform array geometries.
• To develop a sparsity-based interpolation method that increases effective DOFs for improved estimation accuracy.

Proposed method

• Leverages multi-frequency operation to synthetically increase the aperture and DOFs of non-uniform arrays for wideband DOA estimation.
• Applies sparse reconstruction techniques (e.g., compressive sensing) to estimate DOA in the presence of coherent and uncorrelated sources.
• Introduces a sparsity-based interpolation method to enhance DOF availability by reconstructing virtual array responses from non-uniform sensor positions.
• Proposes a DOA estimation framework for mixed coherent and uncorrelated targets using active non-uniform arrays under narrowband conditions.
• Develops mutual coupling compensation techniques tailored to non-uniform array geometries using calibration and signal modeling.
• Employs subspace-based methods (e.g., MUSIC-like) in conjunction with sparse recovery for high-resolution DOA estimation.

Experimental results

Research questions

• RQ1 How can DOA estimation performance be improved in non-uniform arrays with limited degrees of freedom?
• RQ2 What is the impact of coherent targets on DOA estimation in non-uniform arrays, and how can it be mitigated?
• RQ3 How does mutual coupling degrade DOA estimation accuracy in practical non-uniform arrays, and what methods can correct for it?
• RQ4 Can multi-frequency operation effectively increase the effective aperture and DOFs in non-uniform arrays?
• RQ5 To what extent can sparsity-based interpolation enhance DOA resolution in non-uniform array configurations?

Key findings

• Multi-frequency operation successfully increases the effective DOFs in non-uniform arrays, enabling higher resolution DOA estimation beyond the physical array aperture.
• The proposed sparsity-based interpolation technique improves DOA estimation accuracy by synthesizing a denser virtual array response from sparse non-uniform sensor positions.
• The method for handling mixed coherent and uncorrelated targets achieves accurate DOA estimation using sparse reconstruction in active non-uniform arrays.
• Mutual coupling compensation techniques significantly reduce estimation errors in practical non-uniform array setups, improving robustness.
• Numerical simulations validate the effectiveness of all proposed methods across various scenarios, including high SNR and low SNR conditions.
• The integration of sparse reconstruction with non-uniform array geometry enables DOA estimation with performance approaching that of uniform arrays but with greater flexibility.