基于频率着色的稀疏贝叶斯宽带波达角估计方法

Wideband direction of arrival estimation in an interference environment via the sparse Bayesian learning based on frequency coloring

为了提升稀疏贝叶斯(Sparse Bayesian Learning,SBL)算法在干扰环境下对目标信号的检测能力,提出将频率着色技术(Frequency Coloring,FC)推广至SBL算法中。在SBL-FC算法中,首先将阵列接收信号通过傅里叶变换转换至各个子带,在各子带内利用SBL算法进行波达角估计,输出功率谱。不同于常规的SBL算法仅将各子带的功率谱进行简单地叠加,算法考虑干扰和目标频谱结构的差异性,对各子带进行不同的着色,使得干扰和目标轨迹在方位时间历程图上对应于不同的颜色,从而使得目标轨迹更易被提取。数值仿真和实验数据分析表明,利用目标和干扰频谱结构的差异性可有效提升SBL算法在干扰环境下对目标信号的检测能力。

In this paper,the frequency coloring technique is extended to the sparse Bayesian learning(SBL)algorithm to improve its performance of weak target detection in an interference environment.By this SBL-FC method,the array-received data are transformed into different frequency bins through Fourier transformation,and the SBL is used to estimate the directions of arrivals in each frequency bin for obtaining the power spectrum.Unlike the conventional SBL that sums the results in all frequency bins,the frequency spectrum difference between the interference and the target is considered,and the result in each frequency bin is colored differently.Based on this,the tracks of the interference and the target are shown in the bearing and time record(BTR)with different colors,making the target easily to be detected.Simulation and experimental results confirm that the performance of the SBL algorithm for target detection is improved by considering the frequency spectrum difference between the interference and the target.