I have a data set that is complex that i generated by using a hilbert transform. I wanted to add guassian noise (complex) to the data and then do some processing. When i do this, is the correct method to add the hilbert tranform of the data to the hilbert transform of the noise? Or do i use a 'different' noise sequence and create complex additive noise to keep the I and Q of the noise uncorrelated? If your a matlab user.. The question is kind of asking the difference of data = [1:10] complexData = hilbert(data) noise = randn(1,10); complexNoise = hilbert(noise) OR---------Is noise modeled as complexNoise2 = randn(1,10)+i*randn(1,10) Thanks in advance
Complex Data/ Complex Noise Simulation
Started by ●June 4, 2009
Reply by ●June 4, 20092009-06-04
westocl wrote:> I have a data set that is complex that i generated by using a hilbert > transform. I wanted to add guassian noise (complex) to the data and then do > some processing. > > When i do this, is the correct method to add the hilbert tranform of the > data to the hilbert transform of the noise? Or do i use a 'different' noise > sequence and create complex additive noise to keep the I and Q of the noise > uncorrelated? > > If your a matlab user.. The question is kind of asking the difference of > > data = [1:10] > complexData = hilbert(data)I don't think that's right. I would say (be careful to time align the two terms) complexData = data + j*hilbert(data) I noise and Q noise should be uncorrelated. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������