Sparse Bayesian Learning Based DOA Estimation and Array Gain-Phase Error Self-Calibration

Abstract

This paper proposes a joint estimation algorithm based on sparse-Bayesian learning (SBL) for the gain-phase problem between array antenna channels. The algorithm uses the idea of the iterative method to jointly estimate the direction-of-arrival (DOA) and gain-phase error calibration coefficients in the iterative process, combining self-calibration and calibration with a calibration source. At each iteration, the rough value of DOA is first estimated using SBL, and then the DOA estimate is used to calculate the gain-phase error calibration coefficient. The value obtained in each iteration is brought into the error cost function, which is constructed based on the principle of signal and noise subspace orthogonality. Iterations are continued until convergence to find the minimum value of the cost function. The algorithm does not require a priori knowledge of array perturbations and has good performance in DOA and array gain and phase error estimation. Simulations and experimental measurements show that the method has better calibration performance than other methods based on optimization algorithms, and the algorithm effectively improves the antenna gain.