A High-Accuracy Parameter Estimation Algorithm for Jointless Frequency-Shift Track Circuit

FSK (frequency shift keying) signal, which has advantages of narrow-bandwidth, strong anti-interference ability, long transmission distance and so on, is used in jointless frequency-shift track circuit to send different kinds of control signals. However, as FSK is non-linearly modulated, parameter estimation with high accuracy is hard to realize. Based on the spectrum analysis complement with time-frequency distribution, a high-accuracy FSK signal parameter estimation algorithm is put forward in this paper. According to the signal characteristics, under-sampling and ZFFT are used to improve the accuracy of spectrum analysis, and the base frequency resolution meets system requirement of 0.02 Hz. Wigner-Ville distribution has an excellent time-frequency concentration while serious cross-term interference at the same time. Through the design of kernel function, the cross-terms are almost suppressed and the upper/down side frequency resolution meets the system requirement of 0.2 Hz. The effectivities of all the methods mentioned above have been proved by MATLAB simulation, which pave a solid way for the development of high-accuracy FSK signal parameter estimation devices.