EE4000 Prac2 2011

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 1

EE4000:03 Semester 1 2011

Signal Processing 3 O P KENNY

Practical No 2

FIR Filter Design Frequency Sampling Method

Introduction:
This laboratory class is designed to give the student experience in designing and testing a low-pass FIR
filter using the frequency sampling method. The student will use theory given in lectures or text book
and use MATLAB to verify the filter.

The laboratory is designed to be completed and written up in the 3 hour laboratory period. Your report
must be submitted within two weeks of the practical.

Submission:
You need to submit a printed copy of your MATLAB code and a report of your findings.

Assessment:
• Marks for this practical will be divided equally between code and report.
• Your code will be assessed for completeness in performing the nominated tasks.
• Your report will be assessed for its completeness and should include printouts of all the
results, justification for the method used and conclusions.

Exercise 1:
Use the Frequency Sampling Method to design a low pass filter in MATLAB.
Verify your filter performance by using it to filter a broadband white noise signal.

Filter Specifications: Design a length N=23 linear phase FIR low-pass filter to approximate an ideal
response that has a pass-band edge of ω = 0.3π . Assume a unity sampling rate, which gives a
Nyquist frequency of ω = π .

Exercise 2:
From exercise 1, the frequency response is either one or zeros at the frequency sample.

1 0 ≤ ω ≤ ω0
A(ω ) = 
0 ω0 ≤ ω ≤ π

This means that the transition band takes place in one frequency sample interval. This can result in
large amount of oscillations in the frequency response near the band edge. A transition band between
the passband and the stopband gives more flexibility in stating the specifications and can reduce the
size of the oscillations in the amplitude response. The ideal response including a transition band is
defined as,

1 0 ≤ ω ≤ ωp
ω −ω
A(ω ) = s ω p ≤ ω ≤ ωs
ωs − ω p
ωs ≤ ω ≤ π
0

(
Where ω p is the edge of the pass band, ωS the edge of the stop band, and ω0 = ω p − ωs / 2 . )
Modify the ideal frequency response so that transition from the pass-band to the stop band
over two frequency sampling intervals by changing the sample nearest the band edge to 0.5
rather than zero. Redesign the filter and compare the filter response and the impulse response
to that in exercise 1. Verify the design by filtering a broad band noise signal.

You might also like