Introduction - If you have any usage issues, please Google them yourself
Frequency domain analysis (2) plus random noise, and frequency-domain analysis with noisy speech (3) design filters for filtering noisy speech (wavread, fft, awgn or randn record a voice (1) when its , filter, etc.)