A New Algorithm on 94 GHz Asymmetric Antenna Bio-radar Conducted Speech Enhancement

To further enhance the detection sensitivity of the bio-radar, a 94 GHz asymmetric antenna radar is used to acquire human speech signal in this paper. However, the radar speech signal are often degraded by sources of noise that include radar harmonic noise, electrocircuit noise, and ambient noise. Thus, this paper proposes a new algorithm, which uses a perceputal wavelet packet, teager energy operator and a modified thresholding function to suppress the noise of the radar speech signal. The spectrogram results show that this method can reduce the noise and significantly improve the performance of radar speech.