DOA Estimation from One-Bit Magnitude-Only Measurements via Sign-Consistency Optimization

Abstract

The direction-of-arrival (DOA) estimation problem using one-bit quantized magnitude-only measurements is studied, where magnitude-only measurements offer robustness against phase errors, thereby avoiding the need for array calibration, while one-bit quantization significantly reduces hardware cost and system complexity. As their direct combination results in meaningless constant measurements, we formulate a sign-consistency optimization problem using a smooth logistic surrogate with ell_2,1-norm regularization to promote joint sparsity. To solve this problem, a proximal-gradient algorithm is developed with guaranteed convergence to a critical point. Numerical results demonstrate that the proposed method achieves accuracy comparable to coherent one-bit baselines under ideal conditions, while maintaining robust performance under severe phase errors that substantially impair coherent methods.