基于微分进化的极化WSF信号参数估计算法

Polarization WSF Algorithm Based on Differential Evolution

极化敏感阵列与传统的天线阵列相比,可以同时接收到信号的空间信息和更加完整的电磁信息,由于受信号极化变化的干扰较小,接收增益更高,估计出的极化状态参数可以用于检测、多址等领域,因此具有更加广阔的开发价值。极化加权信号子空间(WSF)算法的精度、分辨率明显优于一般子空间类算法,并且可以处理相干信号,鲁棒性较好,与传统空间谱WSF相比,需要估计的参数多了一倍,计算量问题显得更加突出。针对该问题,首先将遗传算法应用于联合谱WSF,与传统测向不同,性能不佳。微分进化算法简单,收敛速度快,搜索精度高,性能稳定,将该算法应用于极化加权信号子空间算法的多维函数求解,并将它与基于遗传算法的极化WSF进行比较,证明文中算法的有效性。

Compared with traditional antenna array,the polarization sensitive array can receive spatial information and more complete electromagnetic information.It has higher receive gain due to less sensitivity to the variation of signal polarization.The polarization weighted subspace fitting（WSF） algorithm is obviously better in accuracy and resolution than the general subspace algorithm and can process coherent signals.The algorithm has good robustness.But the number of parameters needed to be estimated is twice more than traditional WSF,so computation problem appears more prominent.To deal with this problem,the genetic algorithm is used to polarization WSF.But poor performance is expressed,which is different from traditional WSF.Differential evolution algorithm,features as simplicity,fast convergence,high accuracy,search performance,and stability,is suitable for solving multi-dimensional functions of maximum solution,this paper applies the algorithm to the polarization WSF and compares it with the WSF based on genetic algorithm.Experimental comparison simulation shows the efficiency of the method.