Scribd
Upload
36 views

Presentation On Power Spectrum Estimation: Course: Digital Signal Processing 2

This presentation discusses power spectrum estimation methods for digital signal processing. It introduces the key concepts of power spectrum and compares parametric and non-parametric spectral estimation methods. Non-parametric methods discussed include the periodogram, modified periodogram, Bartlett's method, and Welch's method. Parametric methods discussed include autoregressive, moving average, and eigen-based methods like MUSIC. The presentation analyzes the resolution, variance, and other performance aspects of different estimation methods. It concludes by presenting example problems from previous digital signal processing exams.

Uploaded by

Bayev Nazar
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
36 views

Presentation On Power Spectrum Estimation: Course: Digital Signal Processing 2

This presentation discusses power spectrum estimation methods for digital signal processing. It introduces the key concepts of power spectrum and compares parametric and non-parametric spectral estimation methods. Non-parametric methods discussed include the periodogram, modified periodogram, Bartlett's method, and Welch's method. Parametric methods discussed include autoregressive, moving average, and eigen-based methods like MUSIC. The presentation analyzes the resolution, variance, and other performance aspects of different estimation methods. It concludes by presenting example problems from previous digital signal processing exams.

Uploaded by

Bayev Nazar
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 28

Presentation on Power Spectrum Estimation

Course: Digital Signal Processing 2

Submitted by : Submitted to :
Md. Istiaq Ansari Md. Hadiur Rahman KhanStudent
Student Id: 1406143 Lecturer
EEE C1 EEE, BUET
Date: 12-18-2018
Spectrum Estimation
What is Power Spectum
Methods of spectral estimation

Power Spectrum

Non Parametric Parametric


Method Method

Classical Non Classical

Minimum
1. Variance Model base
method Eigen Based methods
Periodigram methods
(High 1. Pisarenko Harmonic
2. Modified resolution)
1. Auto Regressive Decomposition
Periodogram 2. Moving Average 2.MUSIC
3. Bartlett’s 3. Moving Average 3. Minimum normal
Method Auto regressive Method(super
resolution)
4. Welch’s
Method (High Resolution
Methods)
5.Blackman-
Tuky
The periodogram
The periodogram
Performance of Periodogram
Performance of Periodogram

Bias
Performance of Periodogram

Resolution :

Variance :
Modified periodogram
Modified periodogram

Variance :
Bartlett’s method
Bartlett’s method

Resolution :

Variance :
Welch’s Method

For a fixed resolution as in Bartlett method variance


Minimum variance method
Basically a filter bank method
Parametric Methods
Requires priori knowledge about how the signal was
generated or experimental results about how well the
model works.
•

It can be used to store huge data as a model and


generate the signal whenever needed using the model
•

It gives high resolution


Parametric Methods

B(z) = b0 + b1z-1 + b2z-2 + ……… + bqz-q , b0 = 1 (Moving average)

A(z) = 1 + a1z-1 + a2z-2 + ……… + apz-p (Auto Regressive)


Problems from previous term finals
Problems from previous term finals
Problems from previous term finals
Problems from previous term finals
Problems from previous term finals
Problems from previous term finals
Problems from previous term finals
Problems from previous term finals
Problems from previous term finals
Problems from previous term finals
Thank you

You might also like