EE 608: Adaptive Signal Processing
Problem Statement:
2. This problem deals with adaptive noise cancellation using the LMS
algorithm. You will need two data streams: one corresponding to signal plus
noise and the other corresponding to noise. You could synthesize such data but
this would be too far from the real situation. Thus (a) Use the speech data
files provided on the course web page. The files are for speech plus noise, and
only noise. You will notice that you cannot decipher the speech when you play
the speech plus noise file; but upon proper implementation of the noise
cancellation algorithm you will be able to decipher the speech
. i. noise3.wav and
signal_noise3.wav form one pair.
ii. snp.wav and n.wav from one
pair.
iii. look for other such pairs on the web; these can also be included in the
noise cancellation simulation data set.
1.For signal_noise3 and noise3 pair.
filter order: M
mue
output
5
0.05
out_speech1.wav
10
0.02
out_speech2.wav
2.For snp and n pair
filter order: M
mue
output
5
0.05
out_sound1.wav
10
0.02
out_sound2.wav