Noise Cancellation Without reference signal

Anyone here have done the Noise Cancellation Without reference signal. Please share the Matlab code. Thank you

[Neil Taylor]

An old post but DD above is correct. Also look for Adaptive self-tuning filters, and self-tunning Kalman filters.

Answer 1

[Image Analyst]

There are lots of ways to reduce noise without any reference signal. For example convolving with a box/rect filter.

smoothedSignal = conv(noisySignal, ones(1,11)/11, 'same');

Did you have some specific method in mind?

[Da Da]

I am asking about the filtering work Without the reference signal. With the reference I did already.

[Walter Roberson]

I checked Google Scholar, but I do not find any papers on the topic "Noise Cancellation Without Reference Signal" ?

[Image Analyst]

There is no reference signal in my code, just the noisy signal, a box filter (a rect function), and an output signal.

Answer 2

[Umair Nadeem]

You can use a digital lowpass filter to reduce the high frequency noise peaks. Any filter would do the job but a Butterworth filter has a better response because of minimum ripples in the passband.

Sample code:

[num den] = butter(3, wn, 'low');

output = filter(num, den, input);

here take wn = signal_freq/sampling_freq, if you dont know the sampling frequency then take

wn = any value between 0.1 and 0.9 and check the response for the most suitable wn value.

Hope this helps. Your feedback will be appreciated. Thanks

Answer 3

[jnaumann]

It sounds like you are after an Adaptive Linear Enhancer (ALE) - some info given here

http://www.mathworks.co.uk/help/dsp/examples/adaptive-line-enhancer-ale.html