基于SFFT的宽带信号互谱法测向算法

A Cross-Spectral Direction Finding Algorithm for Broadband Signals Based on SFFT

在无线电频谱监测中,随着数据采集能力和采样频率的不断提高,对算法的时效性提出了更高要求。对于宽带信号测向系统,提出基于稀疏快速傅里叶变换的互谱法相位测量算法,该算法利用信号频域的稀疏特性,通过频谱重排、滤波、降采样和估值,能快速计算出频谱中K(信号稀疏度)个拥有最大值的傅里叶系数。利用这K个大值点计算平均时延,在保证与传统快速傅里叶变换有相同精度的同时,降低算法的时间复杂度。分析表明,该算法的时间复杂度与信号稀疏度K呈亚线性关系。该方法提高了算法效率。仿真分析对比了基于稀疏快速傅里叶变换的互谱法和基于快速傅里叶变换的互谱法的误差,表明了该算法的有效性。

In the radio spectrum monitoring, with the constant improvement of the data acquisition ability and the sampling frequency, higher requirements of the algorithm efficiency are put forward. For broadband signal direction finding system, a cross-spectral method measurement algorithm based on SFFT is proposed. The algo- rithm used the sparse feature of the signal frequency domain, through the spectrum rearrangement, filtering, sampling and valuation, can quickly calculate K （ sparse degree） Fourier coefficients that has a maximum in the spectrum. And these big K value points are used to calculate the average delay, while guaranteeing traditional fast Fourier transfomre, reducing the time complexity of algorithm. The analysis shows that the algorithm＇s time complexity is sublinear with the signal sparee degree K. The efficiency of the algorithm is improved. The simu- lation analysis compares the ereoors of the cross spectrum based on SFFT and FFT, which indicate the effective- ness of the algorithm.