求解阵列天线自适应滤波问题的一种调比随机逼近算法

提出了求解阵列天线自适应滤波问题的一种调比随机逼近算法.每一步迭代中,算法选取调比的带噪负梯度方向作为新的迭代方向.相比已有的其他随机逼近算法,这个算法不需要调整稳定性常数,在一定程度上解决了稳定性常数选取难的问题.数值仿真实验表明,算法优于已有的滤波算法,且比经典Robbins-Monro （RM）算法具有更好的稳定性.

Scaled stochastic approximation algorithm for antenna arrays adaptive filtering problem

A scaled stochastic approximation algorithm for solving antenna arrays adaptive filtering problems is proposed in this paper. At each iteration step, this algorithm employs a scaled noisy negative gradient direction as the iteration direction. Compared with other existing stochastic approximation algorithms, there is no stable parameter necessary to be adjusted in this new algorithm, and to a certain extent, it solves the problem of choosing the stable parameter. Numerical experiments show that the new algorithm outperforms the classical adaptive filtering algorithm, and it has better stability than the Robbins-Monro （RM） algorithm.