基于稀疏表示和近似(e)0范数约束的宽带信号DOA估计

针对宽带信号的波达方向(DOA)估计问题,在稀疏框架下提出一种近似(e)0范数约束的宽带信号DOA估计新算法.首先对宽带信号进行预处理,得到同一参考频率点下的接收数据,然后对其协方差矩阵元素进行加和平均运算,得到一个低维的观测向量,并在稀疏框架下进行稀疏表示,最后利用截断(e)1函数设定权值,构造逼近(e)0范数约束的稀疏重构方法,进而重构信号,获得宽带信号的DOA估计.仿真结果表明,相比于传统的宽带信号DOA估计算法,所提算法具有更高的分辨率和估计精度.

Broadband signal DOA estimation based on sparse representation and (e)0-norm approximation

Based on (e)0-norm approximation,an efficient algorithm is proposed to deal with the localization of the broadband signal under the sparse framework.First,by preprocessing broadband signal,the received data under the same frequency is obtained.Then a sum-average operation to array covariance matrix elements of the received data is made in order to get a low dimensional observation vector and the sparse representation of the new model under sparse framework is built.Finally,exploiting truncated (e)0 function as the weight coefficients to construct (e)0-norm penalty sparse reconstruction method and then reconstruct the broadband signal to obtain DOA estimation.The simulation results demonstrate that comparing to the traditional broadband signal DOA estimate algorithms,the proposed algorithm is able to provide higher resolution and estimation accuracy.