脉冲噪声下基于中值离差相关熵的DOA估计方法

Direction of Arrival Estimation Method using Deviation from the Median based Correntropy under Impulsive Noise

  • 摘要: 作为阵列信号处理的一个基本问题,波达方向估计在现代通信领域有着广泛的应用。常规估计方法通常基于高斯噪声假设。 而当阵列接收数据包含 Alpha 稳定分布脉冲噪声时算法将完全失效。针对 Alpha 稳定分布噪声下的波达方向估计问题,本文定义了一种改进的相关熵算子——中值离差相关熵,并从理论上证明了它的有界性。 以此为基础结合 MUSIC 算法提出了一种波达方向估计新方法。该方法不需要噪声先验知识,且在脉冲噪声环境中具有鲁棒性。仿真实验表明, 即使在短快拍、强脉冲噪声的恶劣环境下,本文方法依然具有良好的性能。

     

    Abstract: As a fundamental issue of array signal processing, direction of arrival estimation is widely applied in contemporary communication. Conventional direction of arrival estimation schemes are primarily developed under Gaussian noise assumptions. However, when the received signal contains alpha-stable distribution impulsive noise, second order statistics will become unbounded and the conventional direction of arrival estimation schemes will become ineffective. To overcome this defect, a modified correntropy operator——deviation from the median based correntropy is first defined and its boundedness property is proved to ensure its validity in the alpha-stable distribution noise environment. Based on the modified correntropy operator and MUSIC algorithm, a novel direction of arrival estimation method is proposed. The proposed scheme not only requires no prior knowledge of background noise but also shows robustness to the impulsiveness of noise. Comprehensive Monte Carlo simulations verify the superior performance of the proposed method under many impulsive noise scenario, even when there are few snapshots and the impulsiveness of noise is strong.

     

/

返回文章
返回