基于时频图和波形变换的FSK信号符号速率估计算法

Algorithm for Estimating Symbol Rate of FSK Signal Based on Time-Frequency Graph and Waveform Convertion

针对FSK信号符号速率估计问题,现有的算法大多基于小波变换,但其存在抗噪性能不佳、小波参数不易选取等缺点。文章基于对时频图的研究,首先由时频图获得FSK信号时频变化波形,进行差分变换后得到符号跳变时刻脉冲,将跳变脉冲转换为单极性归零波形,进而提取其定时分量,实现了低信噪比条件下FSK信号的符号速率估计。通过循环迭代不断修正时频图参数,获取合适的时间分辨率,进而提高符号速率估计精度。仿真实验表明,文章算法参数估计精度高,低信噪比条件下性能优于已有基于小波变换的符号速率估计算法。

The existing algorithms for estimating symbol rate of FSK signal are mostly based on wavelet transform.The anti-noise performance of these algorithms is not good enough and it is not easy to select the wavelet parameters.Based on research on time-frequency graph,the time-frequency waveform of FSK signal is obtained,then the pulses are extracted from the differential waveform.With the differential waveform converted to unipolar RZ waveform,the timing components can be obtained.In this way,symbol rate estimation of the FSK signal is achieved under low SNR conditions.The parameters to calculate the time-frequency graph are revised by iteration to improve the estimation accuracy.Simulation results show that the algorithm achieves high accuracy and better performance than the existing symbol rate estimation algorithm based on wavelet transform.