An adaptive method of pulse detection based on frequency-domain CFAR

Aiming at pulse detection in practical complex environment applications, a new method based on frequency-domain constant false alarm rate (CFAR) detector is proposed in this paper. First, short time Fourier transform (STFT) is carried out. Then frequency-domain CFAR detector is used to judge whether signal exists or not in each STFT window. Finally the complete pulse is detected by sliding the STFT window. The threshold of the method is obtained and updated automatically according to the current noise level. The proposed method is robust to signal to noise ratio (SNR) and can work well without artificially setting detection threshold. Simulation results and real data applications validate the effectiveness of the proposed method.