• Home
  • Textbooks
  • Digital Signal Processing: Fundamentals and Applications
  • Infinite Impulse Response Filter Design

Digital Signal Processing: Fundamentals and Applications

Lizhe Tan, Jean Jiang

Chapter 8

Infinite Impulse Response Filter Design - all with Video Answers

Educators


Chapter Questions

Problem 1

Given an analog filter with the transfer function
$$
H(s)=\frac{1000}{s+1000},
$$
convert it to the digital filter transfer function and difference equation using the BLT if the DSP system has a sampling period of $T=0.001 \mathrm{~s}$.

Check back soon!

Problem 2

The lowpass filter with a cutoff frequency of $1 \mathrm{rad} / \mathrm{s}$ is given as
$$
H_P(s)=\frac{1}{s+1}
$$
(a) Use $H_P(s)$ and the BLT to obtain a corresponding IIR digital lowpass filter with a cutoff frequency of $30 \mathrm{~Hz}$, assuming a sampling rate of $200 \mathrm{~Hz}$.
(b) Use MATLAB to plot the magnitude and phase frequency responses of $H(z)$.

Check back soon!

Problem 3

The normalized lowpass filter with a cutoff frequency of $1 \mathrm{rad} / \mathrm{s}$ is given as
$$
H_P(s)=\frac{1}{s+1}
$$
(a) Use $H_P(s)$ and the BLT to obtain a corresponding IIR digital highpass filter with a cutoff frequency of $30 \mathrm{~Hz}$, assuming a sampling rate of $200 \mathrm{~Hz}$.
(b) Use MATLAB to plot the magnitude and phase frequency responses of $H(z)$.

Check back soon!

Problem 4

Consider the normalized lowpass filter with a cutoff frequency of $1 \mathrm{rad} / \mathrm{s}$ :
$$
H_P(s)=\frac{1}{s+1}
$$
(a) Use $H_P(s)$ and the BLT to design a corresponding IIR digital notch (bandstop) filter with a lower cutoff frequency of $20 \mathrm{~Hz}$, an upper cutoff frequency of $40 \mathrm{~Hz}$, and a sampling rate of $120 \mathrm{~Hz}$
(b) Use MATLAB to plot the magnitude and phase frequency responses of $H(z)$.

Check back soon!

Problem 5

Consider the following normalized lowpass filter with a cutoff frequency of $1 \mathrm{rad} / \mathrm{s}$ :
$$
H_P(s)=\frac{1}{s+1}
$$
(a) Use $H_P(s)$ and the BLT to design a corresponding IIR digital bandpass filter with a lower cutoff frequency of $15 \mathrm{~Hz}$, an upper cutoff frequency of $25 \mathrm{~Hz}$, and a sampling rate of $120 \mathrm{~Hz}$.
(b) Use MATLAB to plot the magnitude and phase frequency responses of $H(z)$.

Check back soon!

Problem 6

Design a first-order digital lowpass Butterworth filter with a cutoff frequency of $1.5 \mathrm{kHz}$ and a passband ripple of $3 \mathrm{~dB}$ at a sampling frequency of $8000 \mathrm{~Hz}$.
(a) Determine the transfer function and difference equation.
(b) Use MATLAB to plot the magnitude and phase frequency responses.

Check back soon!

Problem 7

Design a second-order digital lowpass Butterworth filter with a cutoff frequency of $1.5 \mathrm{kHz}$ and a passband ripple of $3 \mathrm{~dB}$ at a sampling frequency of $8000 \mathrm{~Hz}$.
(a) Determine the transfer function and difference equation.
(b) Use MATLAB to plot the magnitude and phase frequency responses.

Check back soon!

Problem 8

Design a third-order digital highpass Butterworth filter with a cutoff frequency of $2 \mathrm{kHz}$ and a passband ripple of $3 \mathrm{~dB}$ at a sampling frequency of $8000 \mathrm{~Hz}$.
(a) Determine the transfer function and difference equation.
(b) Use MATLAB to plot the magnitude and phase frequency responses.

Check back soon!

Problem 9

Design a second-order digital bandpass Butterworth filter with a lower cutoff frequency of $1.9 \mathrm{kHz}$, an upper cutoff frequency $2.1 \mathrm{kHz}$, and a passband ripple of $3 \mathrm{~dB}$ at a sampling frequency of $8000 \mathrm{~Hz}$.
(a) Determine the transfer function and difference equation.
(b) Use MATLAB to plot the magnitude and phase frequency responses.

Check back soon!

Problem 10

Design a second-order digital handstop Butterworth filter with a center frequency of $1.8 \mathrm{kH}$, a bandwidth of $200 \mathrm{~Hz}$, and a passband ripple of $3 \mathrm{~dB}$ at a sampling frequency of $8000 \mathrm{~Hz}$.
(a) Determine the transfer function and difference equation.
(b) Use MATLAB to plot the magnitude and phase frequency responses.

Check back soon!

Problem 11

Design a first-order digital lowpass Chebyshev filter with a cutoff frequency of $1.5 \mathrm{kHz}$ and $1 \mathrm{~dB}$ ripple on passband at a sampling frequency of $8000 \mathrm{~Hz}$.
(a) Determine the transfer function and difference equation.
(b) Use MATLAB to plot the magnitude and phase frequency responses.

Check back soon!

Problem 12

Design a second-order digital lowpass Chebyshev filter with a cutoff frequency of $1.5 \mathrm{kHz}$ and $0.5 \mathrm{~dB}$ ripple on passband at a sampling frequency of $8000 \mathrm{~Hz}$. Use MATLAB to plot the magnitude and phase frequency responses.

Check back soon!

Problem 13

Design a third-order digital highpass Chebyshev filter with a cutoff frequency of $2 \mathrm{kHz}$ and $1 \mathrm{~dB}$ ripple on the passband at a sampling frequency of $8000 \mathrm{~Hz}$.
(a) Determine the transfer function and difference equation.
(b) Use MATLAB to plot the magnitude and phase frequency responses.

Check back soon!

Problem 14

Design a second-order digital bandpass Chebyshev filter with the following specifications: Center frequency of $1.5 \mathrm{kHz}$
Bandwidth of $200 \mathrm{~Hz}$
0.5-dB ripple on passband
Sampling frequency of $8000 \mathrm{~Hz}$.
(a) Determine the transfer function and difference equation.
(b) Use MATLAB to plot the magnitude and phase frequency responses.

Check back soon!

Problem 15

Design a second-order bandstop digital Chebyshev filter with the following specifications: Center frequency of $2.5 \mathrm{kHz}$
Bandwidth of $200 \mathrm{~Hz}$
1-dB ripple on stopband
Sampling frequency of $8000 \mathrm{~Hz}$.
(a) Determine the transfer function and difference equation.
(b) Use MATLAB to plot the magnitude and phase frequency responses.

Check back soon!

Problem 16

Design a fourth-order lowpass digital Butterworth filter with a cutoff frequency of $2 \mathrm{kHz}$, and the passband ripple of $3 \mathrm{~dB}$ at a sampling frequency at $8000 \mathrm{~Hz}$.
1. Determine transfer function and difference equation.
2. Use MATLAB to plot the magnitude and phase frequency responses.

Check back soon!

Problem 17

Design a fourth-order digital lowpass Chebyshev filter with a cutoff frequency of $1.5 \mathrm{kHz}$ and a $0.5-\mathrm{dB}$ ripple at a sampling frequency of $8000 \mathrm{~Hz}$.
(a) Determine the transfer function and difference equation.
(b) Use MATLAB to plot the magnitude and phase frequency responses.

Check back soon!

Problem 18

Design a fourth-order digital bandpass Chebyshev filter with a center frequency of $1.5 \mathrm{kHz}, \mathrm{a}$ bandwidth of $200 \mathrm{~Hz}$, and a $0.5-\mathrm{dB}$ ripple at a sampling frequency of $8000 \mathrm{~Hz}$.
(a) Determine the transfer function and difference equation.
(b) Use MATLAB to plot the magnitude and phase frequency responses.

Check back soon!
00:38

Problem 19

Consider the following Laplace transfer function:
$$
H(s)=\frac{10}{s+10}
$$

Hast Aggarwal
Hast Aggarwal
Numerade Educator

Problem 19

(a) Determine $H(z)$ and the difference equation using the impulse invariant method if the sampling rate $f_s=10 \mathrm{~Hz}$.
(b) Use MATLAB to plot the magnitude frequency response $|H(f)|$ and the phase frequency response $\phi(f)$ with respect to $H(s)$ for the frequency range from 0 to $f_s / 2 \mathrm{~Hz}$.
(c) Use MATLAB to plot the magnitude frequency response $\left|H\left(e^{j \Omega}\right)\right|=\left|H\left(e^{i 2 n f T}\right)\right|$ and the phase frequency response $\phi(f)$ with respect to $H(z)$ for the frequency range from 0 to $f_s / 2 \mathrm{~Hz}$.

Check back soon!

Problem 20

Consider the following Laplace transfer function:
$$
H(s)=\frac{1}{s^2+3 s+2}
$$
(a) Determine $H(z)$ and the difference equation using the impulse invariant method if the sampling rate $f_s=10 \mathrm{~Hz}$.
(b) Use MATLAB to plot the magnitude frequency response $|H(f)|$ and the phase frequency response $\phi(f)$ with respect to $H(s)$ for the frequency range from 0 to $f_s / 2 \mathrm{~Hz}$.
(c) Use MATLAB to plot the magnitude frequency response $\left|H\left(e^{j \Omega}\right)\right|=\left|H\left(e^{j 2 n f T}\right)\right|$ and the phase frequency response $\phi(f)$ with respect to $H(z)$ for the frequency range from 0 to $f_s / 2 \mathrm{~Hz}$.

Check back soon!

Problem 21

Consider the following Laplace transfer function:
$$
H(s)=\frac{s}{s^2+4 s+5}
$$
(a) Determine $H(z)$ and the difference equation using the impulse invariant method if the sampling rate $f_s=10 \mathrm{~Hz}$.
(b) Use MATLAB to plot the magnitude frequency response $|H(f)|$ and the phase frequency response $\phi(f)$ with respect to $H(s)$ for the frequency range from 0 to $f_s / 2 \mathrm{~Hz}$;
(c) Use MATLAB to plot the magnitude frequency response $\left|H\left(e^{j \Omega}\right)\right|=\left|H\left(e^{j 2 \pi f T}\right)\right|$ and the phase frequency response $\phi(f)$ with respect to $H(z)$ for the frequency range from 0 to $f_s / 2 \mathrm{~Hz}$.

Check back soon!

Problem 22

A second-order bandpass filter is required to satisfy the following specifications:
Sampling rate $=8000 \mathrm{~Hz}$
3-dB bandwidth: $B W=100 \mathrm{~Hz} \mathrm{~Hz}$
Narrow passband centered at $f_0=2000 \mathrm{~Hz}$
Zero gain at 0 and $4000 \mathrm{~Hz}$.
Find the transfer function and difference equation by the pole-zero placement method.

Check back soon!

Problem 23

A second-order notch filter is required to satisfy the following specifications:
Sampling rate $=8000 \mathrm{~Hz}$
3-dB bandwidth: $B W=200 \mathrm{~Hz}$
Narrow passband centered at $f_0=1000 \mathrm{~Hz}$.
Find the transfer function and difference equation by the pole-zero placement method.

Check back soon!

Problem 24

A first-order lowpass filter is required to satisfy the following specifications:
Sampling rate $=8000 \mathrm{~Hz}$
3-dB cutoff frequency: $f_c=200 \mathrm{~Hz}$
Zero gain at $4000 \mathrm{~Hz}$.
Find the transfer function and difference equation using the pole-zero placement method.

Check back soon!

Problem 25

A first-order lowpass filter is required to satisfy the following specifications:
Sampling rate $=8000 \mathrm{~Hz}$
3-dB cutoff frequency: $f_c=3800 \mathrm{~Hz}$
Zero gain at $4000 \mathrm{~Hz}$.
Find the transfer function and difference equation by the pole-zero placement method.

Check back soon!

Problem 26

A first-order highpass filter is required to satisfy the following specifications:
Sampling rate $=8000 \mathrm{~Hz}$
3-dB cutoff frequency: $f_c=3850 \mathrm{~Hz}$
Zero gain at $0 \mathrm{~Hz}$.
Find the transfer function and difference equation by the pole-zero placement method.

Check back soon!

Problem 27

A first-order highpass filter is required to satisfy the following specifications:
Sampling rate $=8000 \mathrm{~Hz}$
3-dB cutoff frequency: $f_c=100 \mathrm{~Hz}$
Zero gain at $0 \mathrm{~Hz}$.
Find the transfer function and difference equation by the pole-zero placement method.

Check back soon!

Problem 29

Given a filter transfer function
$$
H(z)=\frac{0.3430 z^2+0.6859 z+0.3430}{z^2+0.7075 z+0.7313}
$$
(a) Realize the digital filter using direct-form I and direct-form II.
(b) Determine the difference equations for each implementation.

Check back soon!

Problem 29

Given a fourth-order filter transfer function
$$
H(z)=\frac{0.3430 z^2+0.6859 z+0.3430}{z^2+0.7075 z+0.7313} \times \frac{0.4371 z^2+0.8742 z+0.4371}{z^2-0.1316 z+0.1733}
$$
(a) Realize the digital filter using the cascade (series) form via second-order sections using the direct-form II.
(b) Determine the difference equations for implementation.

Check back soon!

Problem 30

Given a DSP system with a sampling rate of $1000 \mathrm{~Hz}$, develop a $200-\mathrm{Hz}$ single tone generator using the digital IIR filter by completing the following steps:
(a) Determine the digital IIR filter transfer function.
(b) Determine the DSP equation (difference equation).

Check back soon!

Problem 31

Given a DSP system with a sampling rate of $8000 \mathrm{~Hz}$, develop a 250 -Hz single tone generator using the digital IIR filter by completing the following steps:
(a) Determine the digital IIR filter transfer function.
(b) Determine the DSP equation (difference equation).

Check back soon!

Problem 32

Given a DSP system with a sampling rate of $8000 \mathrm{~Hz}$, develop a DTMF tone generator for key 9 using the digital IIR filters by completing the following steps:
(a) Determine the digital IIR filter transfer functions.
(b) Determine the DSP equations (difference equation).

Check back soon!

Problem 33

Given a DSP system with a sampling rate $8000 \mathrm{~Hz}$, develop a DTMF tone generator for key 3 using the digital IIR filters by completing the following steps:
(a) Determine the digital IIR filter transfer functions.
(b) Determine the DSP equations (difference equation).

Check back soon!
06:34

Problem 34

Given $x(0)=1, x(1)=2, x(2)=0, x(3)=-1$, using the Goertzel algorithm to compute the following DFT coefficients and their amplitude spectrum:
(a) $X(0)$
(b) $|X(0)|^2$
(c) $A_0$ (single side)
(d) $X(1)$
(e) $|X(1)|^2$
(f) $A_1$ (single side)

Kajal Gautam
Kajal Gautam
Numerade Educator
01:07

Problem 35

Repeat Problem 8.34 for $X(1)$ and $X(3)$.

Amy Jiang
Amy Jiang
Numerade Educator
02:34

Problem 36

Given the digital data sequence of length 4 as $x(0)=4, x(1)=3, x(2)=2$, and $x(3)=1$, use the modified Goertzel algorithm to compute the spectral amplitude at the frequency bin $k=0$ and $k=2$.

Suzanne W.
Suzanne W.
Numerade Educator
02:17

Problem 37

Repeat Problem 8.36 for $X(1)$ and $X(3)$.
Use MATLAB to solve Problems $8.38-8.50$.

James Kiss
James Kiss
Numerade Educator

Problem 38

A speech sampled at $8000 \mathrm{~Hz}$ is corrupted by a sine wave of $360 \mathrm{~Hz}$. Design a notch filter to remove the noise with the following specifications:
Chebyshev notch filter
Center frequency: $360 \mathrm{~Hz}$
Bandwidth: $60 \mathrm{~Hz}$
Passband ripple: $0.5 \mathrm{~dB}$
Stopband attenuation: $5 \mathrm{~dB}$ at 355 and $365 \mathrm{~Hz}$, respectively
Determine the transfer function and difference equation.

Check back soon!

Problem 39

In Problem 8.38, if the speech is corrupted by a sine wave of $360 \mathrm{~Hz}$ and its third harmonics, cascading two notch filters can be applied to remove noise signals. The possible specifications are given as follows:
Chebyshev notch filter 1
Center frequency: $360 \mathrm{~Hz}$
Bandwidth: $60 \mathrm{~Hz}$
Passband ripple: $0.5 \mathrm{~dB}$
Stopband attenuation: $5 \mathrm{~dB}$ at 355 and $365 \mathrm{~Hz}$, respectively
Chebyshev notch filter 2
Center frequency: $1080 \mathrm{~Hz}$
Bandwidth: $60 \mathrm{~Hz}$
Passband and ripple: $0.5 \mathrm{~dB}$
Stopband attenuation: $5 \mathrm{~dB}$ at 1075 and $1085 \mathrm{~Hz}$, respectively
Determine the transfer function and difference equation for each filter (Fig. 8.58).

Check back soon!

Problem 40

In a speech recording system with a sampling frequency of $10,000 \mathrm{~Hz}$, the speech is corrupted by random noise. To remove the random noise while preserving speech information, the following specifications are given:
Speech frequency range: $0-3000 \mathrm{~Hz}$
Stopband range: $4000-5000 \mathrm{~Hz}$
Passband ripple: $3 \mathrm{~dB}$
Stopband attenuation: $25 \mathrm{~dB}$
Butterworth IIR filter
Determine the filter order and transfer function.

Check back soon!

Problem 41

In Problem 8.40, if we use a Chebyshev IIR filter with the following specifications:
Speech frequency range: $0-3000 \mathrm{~Hz}$
Stopband range: $4000-5000 \mathrm{~Hz}$
Passband ripple: $1 \mathrm{~dB}$
Stopband attenuation: $35 \mathrm{~dB}$
Chebyshev IIR filter
Determine the filter order and transfer function.

Check back soon!

Problem 42

Given a speech equalizer to compensate midrange frequency loss of hearing (Fig. 8.59) and the following specifications:
Sampling rate: $8000 \mathrm{~Hz}$
Second-order bandpass IIR filter
Frequency range to be emphasized: $1500-2000 \mathrm{~Hz}$
Passband ripple: $3 \mathrm{~dB}$
Pole-zero placement design method
Determine the transfer function.

Check back soon!

Problem 43

In Problem 8.42, if we use an IIR filter with following specifications:
Sampling rate: $8000 \mathrm{~Hz}$
Butterworth IIR filter
Frequency range to be emphasized: $1500-2000 \mathrm{~Hz}$
Lower stopband: $0-1000 \mathrm{~Hz}$
Upper stopband: $2500-4000 \mathrm{~Hz}$
Passband ripple: $3 \mathrm{~dB}$
Stopband attenuation: $20 \mathrm{~dB}$
Determine the filter order and filter transfer function.

Check back soon!

Problem 44

A digital crossover can be designed as shown in Fig. 8.60.
Given audio specifications as
Sampling rate: $44,100 \mathrm{~Hz}$
Crossover frequency: $1000 \mathrm{~Hz}$
Highpass filter: third-order Butterworth type at a cutoff frequency of $1000 \mathrm{~Hz}$
Lowpass filter: third-order Butterworth type at a cutoff frequency of $1000 \mathrm{~Hz}$
Use the MATLAB BLT design method to determine:
(a) the transfer functions and difference equations for the highpass and lowpass filters;
(b) frequency responses for the highpass filter and the lowpass filter;
(c) combined frequency response for both filters.

Check back soon!

Problem 45

Given a DSP system with a sampling rate of $8000 \mathrm{~Hz}$, develop an $800 \mathrm{~Hz}$ single-tone generator using a digital IIR filter by completing the following steps:
(a) Determine the digital IIR filter transfer function.
(b) Determine the DSP equation (difference equation).
(c) Write a MATLAB program using the MATLAB function filter() to generate and plot the $800-\mathrm{Hz}$ tone for a duration of $0.01 \mathrm{~s}$.

Check back soon!

Problem 46

Given a DSP system with a sampling rate set up to be $8000 \mathrm{~Hz}$, develop a DTMF tone generator for key "5" using digital IIR filters by completing the following steps:
(a) Determine the digital IIR filter transfer functions.
(b) Determine the DSP equations (difference equation).
(c) Write a MATLAB program using the MATLAB function filter( ) to generate and plot the DTMF tone for key 5 for 205 samples.

Check back soon!
06:34

Problem 47

Given $x(0)=1, x(1)=1, x(2)=0, x(3)=-1$, using the Goertzel algorithm to compute the following DFT coefficients and their amplitude spectra:
(a) $X(0)$
(b) $|X(0)|^2$
(c) $A_0$ (single sided)
(d) $X(1)$
(e) $|X(1)|^2$
(f) $A_1$ (single sided)

Kajal Gautam
Kajal Gautam
Numerade Educator
11:28

Problem 48

Repeat Problem 8.47 for single-side spectra: $A_2$ and $A_3$.

Mark J
Mark J
Numerade Educator

Problem 49

Given a DSP system with a sampling rate set up to be $8000 \mathrm{~Hz}$ and data size of $205(N=205)$, seven Goertzel IIR filters are implemented for DTMF tone detection. For the frequencies corresponding to key 5 , determine:
(a) the modified Goertzel filter transfer functions
(b) the filter DSP equations for $v_k(n)$
(c) the DSP equations for the squared magnitudes
$$
|X(k)|^2=\left|y_k(205)\right|^2
$$
(d) using the data generated in Problem 8.46 (c), Write a program using the MATLAB function filter() and Goertzel algorithm to detect the spectral values of the DTMF tone for key 5.

Check back soon!

Problem 50

Given an input data sequence:
$$
x(n)=1.2 \cdot \sin (2 \pi(1000) n / 10,000))-1.5 \cdot \cos (2 \pi(4000) n / 10,000)
$$
assuming a sampling frequency of $10 \mathrm{kHz}$, implement the designed IIR filter in Problem 8.41 to filter 500 data points of $x(n)$ with the following specified method, and plot the 500 samples of the input and output data.
(a) Direct-form I implementation
(b) Direct-form II implementation.

Check back soon!

Problem 51

The $60-\mathrm{Hz}$ hum eliminator with harmonics and heart rate detection Given the recorded ECG data (ecgbn.dat) which is corrupted by 60 -Hz interference with its harmonics and the sampling rate is $600 \mathrm{~Hz}$, plot its spectrum and determine the harmonics. With the harmonic frequency information, design a notch filter to enhance the ECG signal. Then use the designed notch filter to process the given ECG signal and apply the zero-cross algorithm to determine the heart rate.

Check back soon!

Problem 52

Digital speech and audio equalizer
Design a seven-band audio equalizer using fourth-order bandpass filters with a sampling rate of $44.1 \mathrm{kHz}$. The center frequencies are listed in Table 8.14.
In this project, use the designed equalizer to process a stereo audio ("No9seg.wav"). Plot the magnitude response for each filter bank.
Listen and evaluate the processed audio with the following gain settings:
(a) each filter bank gain $=0$ (no equalization)
(b) lowpass filtered
(c) bandpass filtered
(d) highpass filtered

Check back soon!
02:43

Problem 53

DTMF tone generation and detection Implement the DTMF tone generation and detection according to Section 8.11 with the following specifications:
(a) Input keys: $1,2,3,4,5,6,7,8,9, *, 0, \#$, A, B, C, D (key frequencies are given in Fig. 8.61).
(b) Sampling frequency is $8000 \mathrm{~Hz}$.
(c) Program will respond each input key with its DTMF tone and display the detected key.

Elizabeth Xu
Elizabeth Xu
Numerade Educator

Problem 54

For the second-order IIR notch filter design using the pole-placement method show that the pole placement for 3-dB bandwidth is
$$
\begin{aligned}
\theta & =\frac{f_0}{f_s} \times 360^{\circ} \\
r & =1-\left(\frac{B W_{3 d B}}{f_s}\right) \pi \text { for } 0.9<r<1 .
\end{aligned}
$$

Check back soon!

Problem 55

For the first-order IIR lowpass filter design with $f_s / 4<f_c<f_s / 2$ using the pole-placement method show that the pole placement for $3-\mathrm{dB}$ bandwidth is
$$
\alpha=-\left(1-\pi+\frac{2 \pi f_c}{f_s}\right) \text { for }-1<\alpha<-0.9 .
$$

Check back soon!

Problem 56

For the first-order IIR highpass filter design with $0<f_c<f_s / 4$ using the pole-placement method show that the pole placement for $3-\mathrm{dB}$ bandwidth is
$$
\alpha=1-2 \pi f_c / f_s \text { for } 0.9<\alpha<1 .
$$

Check back soon!

Problem 57

For the first-order IIR highpass filter design with $f_s / 4<f_c<f_s / 2$ using the pole-placement method show that the pole placement for $3-\mathrm{dB}$ bandwidth is
$$
\alpha=-\left(1-\pi+\frac{2 \pi f_c}{f_s}\right) \text { for }-1<\alpha<-0.9 .
$$

Check back soon!

Problem 58

Filter design by Padé approximation
Given the desired impulse response $h_d(n), n \geq 0$ for an IIR filter, the IIR filter to be designed can be written as
$$
H(z)=\frac{\sum_{k=0}^M b_k z^{-k}}{1+\sum_{k=1}^N a_k z^{-k}}=\sum_{k=0}^{\infty} h(k) z^{-k} .
$$
Let $x(n)=\delta(n)$, the impulse response $y(n)=h(n)$ can be expressed as
$$
\begin{aligned}
& h(n)=-a_1 h(n-1)-a_2 h(n-2)-\cdots-a_N h(n-N) \\
& +b_0 \delta(n)+b_1 \delta(n-1)+\cdots+b_M \delta(n-M)
\end{aligned}
$$
(a) Show that for $0 \leq n \leq M$
$$
h(n)=-a_1 h(n-1)-a_2 h(n-2)-\cdots-a_N h(n-N)+b_n
$$
(b) Show that for $n>M$
$$
h(n)=-a_1 h(n-1)-a_2 h(n-2)-\cdots-a_N h(n-N)
$$
(c) Let $h(n)$ match $h_d(n)$ for $0 \leq n \leq N+M$, that is, $h(n)=h_d(n)$, for $0 \leq n \leq N+M$. Show that
$$
\begin{gathered}
{\left[\begin{array}{cccc}
-h_d(M) & -h_d(M-1) & \cdots & -h_d(M+1-N) \\
-h_d(M+1) & -h_d(M) & \cdots & -h_d(M+2-N) \\
\cdots & \cdots & \ddots & \cdots \\
-h_d(M+N-1) & -h_d(M+N-2) & \cdots & -h_d(M)
\end{array}\right]\left[\begin{array}{c}
a_1 \\
a_2 \\
\vdots \\
a_N
\end{array}\right]=\left[\begin{array}{c}
h_d(M+1) \\
h_d(M+2) \\
\vdots \\
h_d(M+N)
\end{array}\right]} \\
{\left[\begin{array}{c}
b_0 \\
b_1 \\
\vdots \\
b_M
\end{array}\right]=\left[\begin{array}{cccc}
h_d(0) & 0 & \cdots & 0 \\
h_d(1) & h_d(0) & \cdots & 0 \\
\cdots & \cdots & \ddots & \cdots \\
h_d(M) & h_d(M-1) & \cdots & h_d(0)
\end{array}\right]\left[\begin{array}{c}
1 \\
a_1 \\
\vdots \\
a_M
\end{array}\right] .}
\end{gathered}
$$

Check back soon!

Problem 59

Assume the desired unit impulse response is
$$
h_d(n)=\frac{1}{n+1} u(n)
$$
Design an IIR filter using the Pade approximation method with the following form:
$$
H(z)=\frac{\sum_{k=0}^5 b_k z^{-k}}{1+\sum_{k=1}^5 a_k z^{-k}}
$$

Check back soon!

Problem 60

Shanks methods for least-mean squares filter design
Given the desired impulse response $h_d(n), n \geq 0$ for an IIR filter, the IIR filter to be designed can be written as
$$
H(z)=\frac{\sum_{k=0}^M b_k z^{-k}}{1+\sum_{k=1}^M a_k z^{-k}}=\sum_{k=0}^{\infty} h(k) z^{-k} .
$$
Find $b_k$ and $a_k$ such that the sum of squared errors between $h_d(n)$ and $h(n)$ is minimized.
(a) From Padé approximation, we can set
$$
\tilde{h}_d(n)=-\sum_{k=1}^N a_k h_d(n-k)
$$
Show that the equations determine $a_k$ by minimizing the following squared errors:
$$
E_1=\sum_{n=M+1}^{\infty}\left[h_d(n)-\tilde{h}_d(n)\right]^2
$$
That is, for $m=1,2, \cdots, N$
$$
\sum_{k=1}^N a_k \sum_{n=M+1}^{\infty} h_d(n-k) h_d(n-m)=-\sum_{n=M+1}^{\infty} h_d(n) h_d(n-m) .
$$
(b) To determine $b_k$, we first split $H(z)$ into $H_1(z)$ and $H_2(z)$, that is, $H(z)=H_1(z) H_2(z)$ where
$$
H_1(z) \frac{1}{1+\sum_{k=1}^N \tilde{a}_k z^{-k}} \text { and } H_2(z)=\sum_{k=0}^M b_k z^{-k}
$$
Let $v(n)$ be the impulse response of $H_1(z)$, that is,
$$
v(n)=-\sum_{k=1}^N \tilde{a}_k v(n-k)+\delta(n)
$$
Then
$$
\hat{h}_d(n)=\sum_{k=0}^M b_k v(n-k)
$$
After minimizing the sum of squared error
$$
E_2=\sum_{n=0}^{\infty}\left[h_d(n)-\hat{h}_d(n)\right]^2,
$$
Show that for $m=0,1, \cdots, M$
$$
\sum_{k=0}^M b_k \sum_{n=0}^{\infty} v(n-k) v(n-m)=\sum_{n=0}^{\infty} h_d(n) v(n-m) .
$$

Check back soon!

Problem 61

Assume that the desired unit impulse response is
$$
h_d(n)=2\left(\frac{1}{2}\right)^n u(n)
$$
Design an IIR filter using the Shanks method with the following form:
$$
H(z)=\frac{b_0+b_1 z^{-1}}{1+a_1 z^{-1}}
$$

Check back soon!