Forums

wide band aliasing noise in the spectrum of non-uniformly sampled signal

Started by qaisar February 28, 2005
Hello,

       As you know that whenever we perform GDFT( General Discreat Fourier
Transform), on a nonuniformly sampled data, we obtain wideband alysing
noise alongwith signal spectrum, which donot allow us to analyse the
signal with apprepriate approximation.  

If any one have some idea how to solve this problem, please tell me.

I shall be thankfull to you for your help.

Qaisar 
 
		
This message was sent using the Comp.DSP web interface on
www.DSPRelated.com
qaisar wrote:
> Hello, > > As you know that whenever we perform GDFT( General Discreat Fourier > Transform), on a nonuniformly sampled data, we obtain wideband alysing > noise alongwith signal spectrum, which donot allow us to analyse the > signal with apprepriate approximation. > > If any one have some idea how to solve this problem, please tell me. > > I shall be thankfull to you for your help. > > Qaisar > > > This message was sent using the Comp.DSP web interface on > www.DSPRelated.com
Qaisar: This is a good source. OUP --------------------------- Technique for frequency analysis of unevenly sampled radar data House, M.G. Mountcastle, P.D. XonTech Inc., Huntington Beach, CA, USA This paper appears in: Radar Conference, 2002. Proceedings of the IEEE Publication Date: 22-25 April 2002 On page(s): 63 - 67 ISSN: 1097-5659 Number of Pages: xix+508 Inspec Accession Number: 7366627 Abstract: A new technique is described that generalizes the usefulness of the discrete Fourier transform for spectral analysis of radar data to applications where the discrete data points to be analyzed are not sampled at regular intervals and/or do not have equal statistical weight. This method finds a frequency-domain representation which best represents the given time-domain data in the least-square-residual sense. The theoretical implications of such an approach are discussed, and an example of the technique applied to a Doppler processing task is given. The merits of the approach relative to other spectral analysis techniques are discussed.
>qaisar wrote: >> Hello, >> >> As you know that whenever we perform GDFT( General Discreat
Fourier
>> Transform), on a nonuniformly sampled data, we obtain wideband alysing >> noise alongwith signal spectrum, which donot allow us to analyse the >> signal with apprepriate approximation. >> >> If any one have some idea how to solve this problem, please tell me. >> >> I shall be thankfull to you for your help. >> >> Qaisar >> >> >> This message was sent using the Comp.DSP web interface on >> www.DSPRelated.com > >Qaisar: > >This is a good source. > >OUP > >--------------------------- > >Technique for frequency analysis of unevenly sampled radar data >House, M.G. Mountcastle, P.D. >XonTech Inc., Huntington Beach, CA, USA >This paper appears in: Radar Conference, 2002. Proceedings of the IEEE > >Publication Date: 22-25 April 2002 >On page(s): 63 - 67 >ISSN: 1097-5659 >Number of Pages: xix+508 >Inspec Accession Number: 7366627 >Abstract: >A new technique is described that generalizes the usefulness of the >discrete Fourier transform for spectral analysis of radar data to >applications where the discrete data points to be analyzed are not >sampled at regular intervals and/or do not have equal statistical >weight. This method finds a frequency-domain representation which best >represents the given time-domain data in the least-square-residual >sense. The theoretical implications of such an approach are discussed, >and an example of the technique applied to a Doppler processing task is >given. The merits of the approach relative to other spectral analysis >techniques are discussed. >
*************************************************************************** Thank you, for the hint to paper, I shall check it out and then inform you about the progress. Regards. Qaisar This message was sent using the Comp.DSP web interface on www.DSPRelated.com