A Signal Blind Detection Method Based on Wavelet Denoising

Abstract

In this paper, a blind detection method is proposed based on wavelet denoising for short burst signals under low SNR (Signal-to-Noise Ratio), and a definite method to determine the decision threshold is given. The method first performs wavelet denoising on the received data, and then uses the energy detector based on a short window to detect the existence of the signal. Simulation results show that for 2FSK, MSK, QPSK and 16QAM signals, when SNR is −12dB, the detection probability can be higher than 85% and false alarm probability is basically maintained below 20%.