Adaptive nulling array direction-of-arrival estimation based on adaptive kernel width mixture correntropy under impulsive noise

In practical underwater detection scenarios, impulsive noise caused by ice cracking, marine biological activity, and noncooperative sonar transmissions poses significant challenges to accurate direction-of-arrival (DOA) estimation. To address this, an efficient adaptive DOA estimation algorithm is proposed, which integrates bias-compensated adaptive nulling array technology with an adaptive kernel width updating method based on mixture correntropy. In this approach, the kernel width is adaptively optimized along the steepest gradient of the prediction error, ensuring the fastest convergence in each iteration. Combined with a bias compensation strategy, the proposed algorithm achieves a low steady-state error even under limited snapshot conditions. Extensive simulations are conducted to evaluate the algorithm's performance in terms of DOA estimation accuracy, root mean square error, and resolution probability. Finally, real validation is performed using sea trial data, further demonstrating the algorithm's robustness and effectiveness in practical underwater acoustic environments.