Abstract:

To solve the problem that the ultra wideband-impulse radio (UWB-IR) signal detection performance of the traditional detection method based on FitzHugh-Nagumo (FHN) will be deteriorated in α stable distribution noise, the triple scale parameter criterion for eliminating exceptional values is introduced into the FHN model,a UWB-IR detection method based on triple scale parameters and the FHN model is proposed. In addition, the sample quintiles method is used to estimate the parameter of α stable distribution and improve the traditional triple scale parameter criterion, it overcomes the defect that the traditional triple scale parameter criteria rely on the prior information of α stable distribution parameters. Simulation results show that the proposed algorithm can suppress the strong pulse effectively in α stable distributed noise, the UWB-IR signal waveform can be reconstructed preferably, the UWB-IR signal can be detected in strong α stable distributed noise effectively.