基于时频方差聚类的FH信号参数盲估计

为解决复杂电磁环境下跳频(frequency hopping, FH)参数的盲估计问题,提出了基于时频方差聚类的算法。考虑在低信噪比(signal-to-noise ratio, SNR)和定频干扰同时存在的情况下,通过短时傅里叶变换(short time Fourier transform, STFT)将信号变换到时频域,利用遗传算法对信号的时频区间进行提取,根据时频方差对其进行k-means聚类,消除噪声和定频干扰并提取时频脊线,然后运用Haar小波对该时频脊线进行奇异点检测,进而估计出FH信号的FH周期、跳速和FH频率等参数。仿真结果表明,所提算法在SNR低于-5 dB且存在定频干扰的情况下,能够实现对FH参数的精确估计,参数估计正确概率达到90%以上。

FH signal parameter blind estimation based on time-frequency variance clustering

In order to solve the blind estimation problem of the frequency hopping (FH) parameters in the complex electromagnetic environment, an algorithm based on time-frequency variance clustering is proposed. Considering the case where both low signal-to-noise ratio (SNR) and fixed frequency interference exist simultaneously, the signal is transformed into the time-frequency domain by short-time Fourier transform (STFT). The time-frequency interval of the signal is extracted by genetic algorithm, and k-means clustering is performed according to the time-frequency variance. Eliminating noise and fixed-frequency interference, and then extracting time-frequency ridge, the Haar wavelet is used to detect its singularities, furthermore, to estimate parameters such as FH period, hopping speed and hopping frequency. The simulation results show that the proposed algorithm can accurately estimate the parameters such as FH period when the SNR is lower than -5 dB. The correct probability of parameter estimation is over 90%.