远距离通信语音信号噪声抑制仿真研究

Simulation of Noise Suppressing Mechanism for Speech Signal Communication Over Long Distance Transmission

为消除无线传播过程中强高斯白噪声对语音信号包络的影响,提高语音信号在接收端的可懂度,提出一种远距离无线通信中噪声抑制基带语音压扩方法。上述方法针对语音信号频带内噪声与语音信号无法在频域和时域分离、难以使用传统低通滤波方法滤除导致包络严重失真的难点问题,在发送端使用基带语音高频段专门承载语音包络调频信号,并在接收端进行专门的包络平滑滤波,提高包络抗噪性能。经语音压扩算法仿真平台验证,改进方法较传统方法在信号接收端能够获得更高的信噪比,更好的抑制了无线通信接收端基带语音信号频带内噪声强度。实验证明,噪声抑制基带语音压扩方法能够更精确的恢复有声段的语音信号包络和并提高语音信号的可懂度。

To eliminate the effect of strong Gaussian noise attached during wireless transmission on speech signal intelligibility, a noise suppressing mechanism for baseband speech signal in wireless communication is proposed. This mechanism recovers the original speech and suppresses the noises by preprocessing the speech signal on the transmit- ter side, which includes compressing the frequency band and modulating the envelope of the signal, and recovering the waveform by a reverse process on the receiver side. To verify the effectiveness of this mechanism, a simulation platform for this speech companding algorithm has been set up. Simulation results show that this mechanism can sig- nificantly suppress the in - band noise and improve the signal to noise ratio, and accurately recover the envelope and intelligibility of the speech signal.