基于DOA估计的RLS-CMA算法

恒模算法被广泛地应用到盲自适应波束形成中,除了传输信号波形具有恒定的包络外,恒模算法不需要先验知识。提出一种基于来波方向估计的递推最小二乘恒模算法,基于恒模阵列级联的结构,由递推最小二乘算法决定恒模的初始权向量,同时通过对权向量多项式求根获得下一级的初始权向量,再利用最小二乘恒模迭代几步获得准确的结果。系统可以分离多个同信道信源,在干扰信号较强时,仍有稳定的信干比输出,并对阵列幅相差不敏感。计算机仿真证明了算法的正确性和有效性。

A RLS-CMA Algorithm Based on DOA

Constant modulus algorithm is a well-known blind adaptive beamforming method as it requires no pre-knowledge about the signal except that the transmitted signal waveform has a constant envelope. The paper propose a recursive least square-constant modulus algorithm（RLS-CMA） based on direction of arrival. The main idea is to use RLS to determine the initial weight vector for CMA operation, and then, by weight vector polynomial rooting, the next initial weight vector is obtained. The initial weight vectors are set in the directions of the other CM signals based on CM array＇s cascade framework. The LS-CMA iteration several steps are used to arrive accurate result. The method can recover several narrowband co-channel sig- nals and extract the desired signal in despite of the interfering signal stronger than the desired signal. The output SINR is increased. It is not sensitive to the amplitude difference and the phase difference of array. Computer simulations demonstrate the correctness and validity of the algorithm.