Estimation of Sinusoids in White Noise PowerPoint PPT Presentation

presentation player overlay
1 / 10
About This Presentation
Transcript and Presenter's Notes

Title: Estimation of Sinusoids in White Noise


1
Estimation of Sinusoids in White Noise
  • Selami Ciftci

2
Looking for Sinusoids
  • Given N data points of a signal y(n)
  • Find
  • M, the number of sinusoids
  • f, the frequencies
  • A, the amplitudes
  • s, the variance of the noise

3
Our Good Old Friend DFT
  • Is this the solution?
  • Non-Parametric Approach
  • Pluses?
  • The estimated M, fs and As

4
1,2 or more?
  • Essential for MUSIC and ESPRIT
  • MDL Algorithm

5
Modified Covariance Method
  • Higher spectral resolution than Yule-Walker
    Method
  • Comparison with Burg algorithm
  • Not having any line-splitting problems
  • Less sensitivity to initial phase
  • Noise variance is 0.6838

6
Power Spectrum Estimate
7
MUSIC
  • An eigenspace method
  • Based on a grid search
  • Noise variance is 0.6879

8
Power Spectrum Estimate
9
ESPRIT
  • An eigenspace method
  • Noise variance is 0.6860

10
Last Words
  • MDL worked, how about AIC and their AR model
    versions
  • Power computation
  • Loading the data
  • Computation time
Write a Comment
User Comments (0)
About PowerShow.com