Вы находитесь на странице: 1из 41

Anand Academy of Engineering, Tel 2642 5305 1

Anand Academy of Engineering


Bandra, Andheri, Borivali,
Dadar, Thane, Mulund,Vashi,

NAME :

Sr TOPIC
No
PAGE

1. Discrete Time signals 2

Discrete Fourier Transform and Fast 3


2.
Fourier Transform

3.
Analysis of DT system using Z-Transform 18

4. Digital Filter Design 24

DSP Help Line : 99 870 30 881


--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 2

1. DISCRETE TIME SIGNALS

FAQ ” What & Why V 

(1) What are the classification of signals ?


(2) What do you mean by Causal signal, Anti-causal Signal and Both-sided signal ?
(3) Give one example of each signal.
(4) What do you mean by real time signal ? Give example.
(5) What is an energy signal ? Give example.
(6) What is power signal ? Give example.
(7) What is symmetric signal ? Give example.
(8) What is Anti-symmetric signal ? Give example.
(9) What is an Even signal ? Give example.
(10) What is an odd signal ? Give example.
(11) What is the sum of odd signal values ?
(12) How to check whether the given signal is periodic or not ?
(13) What is the concept of digital frequency f ?
(14) What is the range of w and f ?
(15) What is the unit of digital frequency w and f ?
(16) Classify the following signal : Finite Length or Infinite Length :-
x[n] = u[n] + 2 u[n-1] – 3 u[n-5]

(17) What is correlation ?


Ans : Correlation gives a measure of similarity between two data sequences. In this process, two signals
are compared and the degree to which the two signals are similar is computed.

(18) What are the applications of Correlation ?


Ans : Typical applications of correlation include speech processing, image processing and radar systems.
In a radar system, the transmitted signal is correlated with the echo signal to locate the position of
the target. Similarly, in speech processing systems, different waveforms are compared for voice
recognition.
(19) What are the properties of Convolution ?
Ans :
i) Commutative
x [n] * h[n] = h[n] * x[n]

ii) Associative
( x [n] * h1[n] * h2[n] ) = ( x [n] * h1[n] ) * h2[n]

iii) Distributive
x[n] * [h 1[n ] + h 2 [ n ]] = x [n] * h1[n] + x[n] * h2[n].

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 3

2. DFT and FFT

FAQ ” What & Why V 

(1) Define Discrete Fourier Transform of x[n].


N −1
X [k ] = ∑
nk
Ans : x[n ] W n
n = 0

(2) What is the interpretations of DFT coefficients ?


Ans : DFT gives N values of Fourier Transforms of DT signal x[n]
2πk
at w = for k = 0,1, 2, ...... N − 1 .
N

They are equally spaced with frequency spacing of
N
(3) How many complex multiplications and additions are required to find DFT ?
Ans : By DFT
(i) Complex Multiplications = N2
(ii) Complex Additions = N ( N − 1)

(4) How many real multiplications and additions are required to find DFT of 32 point signal.?
Ans : By DFT
(i) Real Multiplications = 4 N 2 = 4(32) 2 = 4096
(i) Real Additions = 4 N 2 − 2 N = 40321

(5) How many complex multiplications and additions are required to find FFT ?
Ans : By DFT
N
(i) Complex Multiplications = log 2 N
2
(ii) Complex Additions = N log 2 N

(6) How many real multiplications and additions are required to find DFT of 32 point signal using
FFT algorithm?
Ans : By FFT
(i) Real Multiplications = 2 N log 2 N = 320
(ii) Real Additions = 3 N log 2 N = 480

(7) What is Scaling and Linearity property of DFT ?

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 4

Ans : Scaling Property : If signal is multiplied by constant Then DFT is also multiplied by the same
constant. i.e. DFT { a x 1 [n] } = a X 1 [k]

Linearity Property : If signals are added, Then DFT’s are also added.
i.e. DFT { a x 1 [n] + b x 2 [n] } = a X 1 [k] + b X 2 [k

(8) What is the DFT of δ[n] ?


Ans : DFT { δ[n] } = 1

(9) What is the DFT of N pt signal u[n] ?


Ans : DFT {u[n] } = N δ[k]

(10) What is the DFT of 4 pt x[n] where x[n] = δ[n] + u[n] ?


Ans : X[k] = 1+ 4 δ[k]
= { 5, 1, 1, 1 }

(11) What is periodicity property of DFT ?


Ans : DFT equation produces periodic results with period = N
i.e. X[k] = X[k+N] = X[k MOD N] = X[((k))]
Inverse DFT equation produces periodic results with period = N
i.e. x[n] = x[n+N] = x[n MOD N] = x[((n))]

(12) Why DFT results are periodic ?


Ans : DFT results are periodic because twiddle factor is periodic with period = N

(13) DFT gives discrete spectrum or continuous spectrum ? Justify ?


Ans : DFT gives discrete spectrum.
If the signal is periodic then spectrum is discrete and if the signal is non-periodic then spectrum
is continuous. DFT assumes that input signal is periodic and therefore DFT gives discrete
spectrum.

(14) Find DFT of x[n] where x[n] = u[n] + 2 u[n-2] – 3 u[n-4]


Ans : Here x[n] = { 1, 1, 3, 3 } By DFT X[k] = {8, -2+2j, 2, -2-2j }

(15) Find DFT of 10 pt x[n] where x[n] = δ[n] + δ[n-5] ?

Ans : X [k ] = 1 + W N5k = 1 + (−1) k

(16) What is Time shift and frequency shift property of DFT ?

Ans : DFT {x [n − m] } = W Nmk X [k ]

{ }
DFT W N− mn x [n] = X [k − m]

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 5

(17) What is symmetry property of DFT ?


Ans : If x[n] ÍÆ X[k] Then X[k] = X*[-k]. i.e. If x[n] is real valued signal, then real part of X[k] is
symmetric about k = N/2 and Imaginary part of X[k] is Anti-symmetric about k = N/2.

(18) What is DFT property of EVEN signal ?


Ans : If x[n] is Even , Then X[k] is also Even
i.e.. If x[n] = x[–n] Then X[k] = X[–k]

(19) What is the DFT of real and even signal.?


Ans : If x[n] is Real and Even, Then X[k] is also Real and Even
Eg. x[n] = { 1, 2, 3, 2 }
X[k] = { 8, -2, 0, -2 }

(20) What is the DFT of Imaginary and Even signal ?


Ans : If x[n] is Imaginary and Even
Then X[k] is also Imaginary and Even
Eg. x[n] = { j, 2j, 3j, 2j }
X[k] = { 8j, -2j, 0, -2j }

(21) What is DFT property of ODD signal ?


Ans : If x[n] = – x[–n] Then X[k] = – X[–k]
i.e. If x[n] is Odd , Then X[k] is also Odd.

(22) What is the DFT of real and Odd signal ?


Ans : If x[n] is Real and Odd, Then X[k] is also Imaginary and Odd
Eg. x[n] ={ 0, 2, 0, –2 }
X[k] = { 0, – 4j, 0, 4j }

(23) What is the DFT of Imaginary and Odd signal ?


Ans : If x[n] is Imaginary and Odd
Then X[k] is also Real and Odd
Eg. x[n]={ 0, 2j, 0, – 2j }
X[k] = { 0, 4, 0, – 4 }

(24) How to find energy of signal from its DFT ?


Ans : According to parseval’s energy theorem, Energy of the signal is given by,
1 N −1
E= ∑
N k =0
| X [k ] | 2

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 6

(25) If DT signal is expanded in time domain what will be the effect in frequency domain?
Ans : Expansion in time domain corresponds to Compression in frequency domain.
Eg. x[n] = {1,2,3,2 } X[k] = { 8, –2, 0, –2}
Let p[n] = {1, 0, 2, 0, 3, 0, 2,0 } Then P[k] = { 8, –2, 0, –2, 8, –2, 0, –2}

0 0.5π π 1.5π 2π 0 0.5π π 1.5π 2π

(26) If DT signal is compressed in time domain what will be the effect in frequency domain?
Ans : Compression in time domain corresponds to Expansion in frequency domain.
Eg. x[n] = {1, 0, 2, 0, 3, 0, 2,0 } X[k] = { 8, –2, 0, –2, 8, –2, 0, –2}
Let p[n] = {1,2,3,2 } Then P[k] = { 8, –2, 0, –2 }

(27) What is convolution property of DFT ?


Ans : Convolution in time domain corresponds to multiplication in frequency domain.
If x[n] X[k] and
h[n] H[k]
Then
DFT { x[n] ⊗ h[n] } = X[k] H[k]

(28) What is correlation property of DFT ?


Ans : If x[n] X[k] and
h[n] H[k]
Then
DFT { x[n] o h[n] } = X[k] H * [k]

(29) What do you mean by decimation ?


Ans : Decimation means sampling.

(30) Which algorithm is more powerful : DIT-FFT or DIF-FFT ?


Ans : Computationally, both the algorithms are exactly same.

(31) What is the length of linearly convolved signals ?


Ans : Length of linearly convolved signal is always equal to N = L + M – 1 where L is length of first
signal and M is length of second signal.

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 7

(32) FFT is faster than DFT . Justify.


Ans : FFT produces fast results because calculations are reduced by decomposition technique.
In FFT, N pt DFT is decomposed into two N/2 pt DFT’s, N/2 pt DFT is decomposed into N/4 pt
DFT’s and so on… Decomposition reduces calculations. FFT algorithms are implemented
using parallel processing techniques. Because calculations are done in parallel, FFT produces
fast results.

Complex Multiplications
DFT FFT
N
N2 N
log 2 N
2
16 256 32
32 1,024 80
64 4,096 192
256 65,536 1,024
512 2,62,144 2,304
1024 10,48,576 5,120

(33) What are the applications of FFT. ?


Ans : (i) Linear Filtering i.e. to find output of digital filter for any given input sequence x[n].
(ii) Spectral Analysis i.e. to find magnitude spectrum and phase spectrum
(iii) Circular Correlation ie to find degree of similarity between two signals.

(34) How to find CC using DFT ?


Ans : To find CC of x[n] and h[n] using DFT,
(i) Select N
Let N = max(L,M) where L is the length of x[n] and M is length of h[n],
(ii)Append x[n] by (N-L) zeros and Append h[n] by (N-M) zeros
N −1
(iii) Find X[k] where X [ k ] = ∑ x[n] W Nnk
n= 0
N −1
(iv) Find H[k] where H [ k ] = ∑ h[n] W Nnk
n= 0
(v) Let Y[k] = X[k] H[k].

i N −1
(vi) Find y[n] where y[n] = ∑ Y [k ] WN− nk
N K=0

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 8

(35) How to find CC using FFT ?


Ans : To find CC of x[n] and h[n] using FFT,
(i) Select N
Let N = max(L,M) where L is the length of x[n] and M is length of h[n],
(ii) Append x[n] by (N-L) zeros and Append h[n] by (N-M) zeros
(iii) Find X[k] by using N point DIT-FFT / DIF-FFT flowgraph
(iv) Find H[k] by using N point DIT-FFT / DIF-FFT flowgraph
(v) Let Y[k] = X[k] H[k].
(vi) Find y[n] by Inverse FFT.

By Inverse FFT, y [n] =


1
N
( { })
FFT Y * [k ]
*

ÆÆ Always explain wrt diagram.

(36) How to find LC using CC ?


Ans : To find LC of x[n] and h[n] using CC,
(i) Select N:
Let N ≥ L + M – 1 where L is the length of x[n] and M is length of h[n],
(ii) Append x[n] by (N-L) zeros and Append h[n] by (N-M) zeros
(iii) Perform N point Circular convolution of x[n] and h[n]

(37) How to find LC using DFT /FFT ?


Ans : : To find LC of x[n] and h[n] using DFT/FFT,
(i) Select N
Let N ≥ L + M – 1 where L is the length of x[n] and M is length of h[n],
(ii) Append x[n] by (N-L) zeros and Append h[n] by (N-M) zeros
(iii) Perform N point Circular convolution of x[n] and h[n] using DFT/FFT.
Find N point X[k] and H[k]
Let Y[k] = X[k] H[k].
Find y[n] by Inverse DFT/FFT.

ÆÆ Always explain wrt diagram.

(38) Let x[n] = { 1, 2, 3, 4 }, and h[n] = { 5, 6, 7 }. Both are non-periodic finite length sequences.
Give step by step procedure to obtain linear convolution using FFT–IFFT
(39) How to find output of the filter using DFT ?
Ans : Output of the filter is Linear convolution of impulse response with the input of the signal.
To find output means to find LC by DFT.

(40) What is periodic convolution ?


Ans : Periodic convolution is convolution of two periodic signals of the same period. When two
periodic signals are periodic with common period, periodic convolution is similar to circular
convolution.

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 9

(41) How to find output of the FIR filter using FFT ?


Ans : In FIR filter length if h[n] is finite. Output of the filter is always Linear convolution of impulse
response with the input of the signal. To find output i.e. to find LC by FFT
(i) Select N
Let N ≥ L + M – 1 where L is the length of x[n] and M is length of h[n],
(ii) Append x[n] by (N-L) zeros and Append h[n] by (N-M) zeros
(iii) Perform N point Circular convolution of x[n] and h[n] using FFT.
* Find N point X[k] and H[k] by using FFT flowgraph.
* Let Y[k] = X[k] H[k].

* Find y[n] by Inverse FFT. y [n] =


1
N
( { })
FFT Y * [k ]
*

ÆÆ Always explain wrt diagram.

(42) What is the difference between circular convolution and periodic convolution ?
Ans : In periodic convolution input signals are originally periodic with common value of period.
In circular convolution, if input signals are not periodic then they are assumed to be periodic with
period = N where N = max(L,M) where L is the length of x[n] and M is the length of h[n].

(43) What do you mean by aliasing in circular convolution ?


Ans : In circular convolution if value of N < L+M-1 then last M-1 values of y[n] wraps around gets
added with first M-1 values of y[n]. This is called aliasing.

(44) Why FFT is used to find output of FIR filter ? Justify.


Ans : FFT produces fast results because in practical applications FFT algorithms are implemented using
parallel processing techniques. Because calculations are done in parallel, FFT produces fast
results.

(45) What are the limitations of filtering by FFT algorithms? Justify.


Ans : (i) NOT suitable for real time applications :
FFT algorithms are implemented using parallel processing techniques. When FFT is used
input is applied in parallel i.e simultaneously. For real time applications entire input signal
is not available. So FFT algorithms can not be used.
(ii) NOT suitable for Long Data Sequence.
As the length of the input sequence increases, the no of stages in FFT will also increase
proportionally and so the delay increases, processing time at each stage increases.

(46) How to find output of FIR filter for long input sequence.
Ans : To find output of digital FIR filter FFT technique is used. But for Long data sequence, direct FFT
technique is not suitable.
For long data sequence, Overlap Add Method using FFT and Overlap Save Method using FFT is
used.

(47) Explain Overlap Add Method.


(48) Explain Overlap Save Method .

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 10

(49) DFT gives continuous spectra or discrete spectra.


Ans : When signal is periodic spectrum is Discrete. If the signal is not-periodic
then spectrum is always continuous.
DFT assumes that signals ar e periodic and therefore DFT gives discrete
spectra.

(50) What is DTFT ?


Ans : DTFT is Fourier Transform of DT signal that converts the sampled DT signal from time domain
to frequency domain. Frequency domain representation parameters are magnitude and phase.
DTFT gives frequency response that includes magnitude response and phase response.

(51) Describe the relation between DFT and DTFT.


Ans : DFT is frequency sampling of DTFT. When DTFT is sampled in frequency domain with


frequency spacing of w = we get DFT coefficients. i.e. X [ k ] = X ( w) 2πk
N w=
N

(52) How to find DFT of infinite length sequence ?


Ans : To find DFT of infinite length sequence x[n]:

(i) Find DTFT of x[n] i.e. X ( w) = ∑ x[n] e − jnw
n = −∞

(ii) Find DFT by frequency sampling DTFT. i.e. X [ k ] = X ( w) 2πk


w=
N
DFT coefficients can be obtained by evaluating DFT equation.

(53) What is Power Density Spectrum of Periodic DT Signals ?


Ans :
N −1
1 2
The average power of periodic DT signal is given by P =
N
∑ x [n]
n=0
2
1 N −1 2
N −1
According to Parseval’s theorem, P= ∑
N n=0
x [ n] = ∑ C k
k= 0
2
The coefficients Ck for k =0, 1, 2…..N-1 is the distribution of power as a function of
frequency is called the power density spectrum of the DT periodic signal

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 11

(54) What is Energy Density Spectrum of DT Aperiodic Signals


Ans :

2
The energy of DT signal x[n] is E = ∑ x [ n]
n = −∞
π 2

2 1
According to parseval’s theorem, E = ∑ x [n] =
2π ∫ X ( w) dw
−∞ −π
2 *
Let S x ( w) = X ( w) = X ( w) X ( w)
Sx (w) is the function of frequency and it is called energy density spectrum of x [n].
∞ π
1
∑ ∫
2
E= x [n] = = Sx ( w) dw.
−∞ 2π
−π

(55) Find DTFT and Energy Density Spectrum of x[n] = u[n].


Ans : Energy of u[n] is infinite. Therefore u[n] is not energy signal.
Fourier Transform is defined only for energy signal.

(56) What is the necessary condition to find DTFT of any signal. ?


Ans : To find DTFT of any signal the necessary condition is, signal must be an energy signal. It must
be absolutely summable.

(57) What is the effect of increasing length of signal by padding zeros on DFT results.?
Ans : As the length of signal increases, the frequency spacing decreases. The number of points per unit
length i.e. resolution of the spectrum increases. Therefore the approximation error in the
representation of the spectrum decreases.

(58) Derive DFT equation . [Refer notes ]


(59) What is the difference between DFT and DTFS ? [Refer notes ]
(60) What is the relation between DFT and DTFS ? [Refer notes ]
(61) What is the relation between DFT and DTFT ? [Refer notes ]
(62) What is the relation between DTFT and ZT ? [Refer notes ]
(63) What is the relation between DFT and ZT ? [Refer notes ]
(64) What is the order of input and output sequence in 8 pt DIT-FFT ?
(65) What is the order of input and output sequence in 8 pt DIF-FFT?
(66) What is bit reversal technique ?

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 12

Short Answer Questions on DFT + FFT


------------------------------------------------------------------------
(67) Let x[n] be 4 point sequence X[k] = {1, 2, 3, 4}. Find the FFT of the following
sequences using X[k] and not otherwise.

1) x[n-1] 4) x[-n+1] 7) 2δ[n] + x[n] 10) ejnπ x[n]


2) x[n+1] 5) x[-n-1] 8) 2 + x[n] 11) ej(n-2)π x[n-2]
3) x[-n] 6) x[n] * x[n] 9) x*[n] 12) e j n π x[n-2]

1] x[n – 1] = ?
Let y[n] = x[n – 1]
Y [k ] = W NK X [k ]
⎡ 1⎤ ⎡ 1⎤
⎢ − j⎥ ⎢ 2 ⎥⎥
=⎢ ⎥⎢
⎢ −1 ⎥ ⎢ 3⎥
⎢ ⎥⎢ ⎥
⎣ j⎦⎣ 4⎦
⎡ 1 k =0
⎢ −2 j

Y(k) = ⎢ −3

⎣ 4j

2] x[n + 1] = ?

Let y(n) = x[n + 1]


Y [k ] = W N− K X [k ]
⎡ 1⎤ ⎡1⎤
⎢ j⎥ ⎢2⎥
=⎢ ⎥ ⎢ ⎥
⎢ −1⎥ ⎢3⎥
⎢ ⎥ ⎢ ⎥
⎣ − j⎦ ⎣4⎦
⎡ 1 k=0
⎢ 2j
= ⎢
⎢−3

⎣− 4 j

3] x(-n) = ?

Let y[n] = x(-n)


Y[k] = X[-k]

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 13

⎡1 k = 0
⎢4
Y[k] = ⎢
⎢3

⎣2

4] x(-n + 1) = ?

Let y(n) = x(-n)


Put n = n – 1
y(n–1) = x(–n + 1)
WNK Y[k ] = DFT {x (−n + 1)}
DFT {x (− n + 1)} = WNk Y[k ]
DFT {x(− n + 1)} = W Nk X [−k ]

⎡ 1⎤ ⎡1⎤
⎢ − j⎥ ⎢ 4 ⎥
DFT {x (−n + 1)} = ⎢ ⎥⎢ ⎥
⎢ −1 ⎥ ⎢ 3 ⎥
⎢ ⎥⎢ ⎥
⎣ j ⎦ ⎣2⎦
⎡ 1 k =0
⎢− 4 j
DFT {x (−n + 1)} = ⎢
⎢ −3

⎣ 2j
5] x(–n – 1) = ?
Let y(n) = x(– n)
Put n = n + 1
y[n + 1]= x(–n–1)

W N− k Y [ k ] = DFT {x[ −n − 1]}


−k
DFT {x[-n -1]}= W N Y [ k ]

= WN−k X[-k]
⎡ 1⎤ ⎡1⎤
⎢ j ⎥ ⎢4⎥
=⎢ ⎥ ⎢ ⎥
⎢ −1⎥ ⎢3⎥
⎢ ⎥ ⎢ ⎥
⎣ − j⎦ ⎣2⎦

⎡ 1 k = 0⎤
⎢ 4j ⎥
= ⎢ ⎥
⎢−3 ⎥
⎢ ⎥
⎣− 2 j ⎦
6] x[n] * x[n]=?

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 14

Let y[n] = x[n] * x[n]


By DFT,
Y[k] = X[k] X[k]
⎡1⎤ ⎡1⎤
⎢2⎥ ⎢2⎥
=⎢ ⎥ ⎢ ⎥
⎢3⎥ ⎢3⎥
⎢ ⎥ ⎢ ⎥
⎣4⎦ ⎣4⎦
⎡1 k = 0⎤
⎢4 ⎥
=⎢ ⎥
⎢9 ⎥
⎢ ⎥
⎣16 ⎦⎥

7] y[n] = 2δ[n] + x[n]

By DFT,
Y[k] = 2 DFT {δ[n] } + X[k]
⎡1 ⎤ ⎡ 1 ⎤
⎢1 ⎥ ⎢ 2 ⎥
=2 ⎢ ⎥ + ⎢ ⎥
⎢1 ⎥ ⎢ 3 ⎥
⎢ ⎥ ⎢ ⎥
⎣1 ⎦ ⎣ 4 ⎦
⎡3 k = 0⎤
⎢4 ⎥
Y[k] = ⎢ ⎥
⎢5 ⎥
⎢ ⎥
⎣6 ⎦
8] Let y[n] = 2 + x[n]

By DFT,
y[n] = 2 u[n] + x[n]
Y[k] = 2 DFT {u[n] + X[k]
= 2 . 4δ[k] + X[k]
⎡1⎤ ⎡1⎤
⎢0⎥ ⎢2⎥
= 8 ⎢ ⎥+⎢ ⎥
⎢0⎥ ⎢3⎥
⎢ ⎥ ⎢ ⎥
⎣0⎦ ⎣4⎦
⎡8⎤ ⎡1⎤
⎢0⎥ ⎢2⎥
= ⎢ ⎥+⎢ ⎥
⎢0⎥ ⎢3⎥
⎢ ⎥ ⎢ ⎥
⎣0⎦ ⎣4⎦

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 15

⎡9 k = 0⎤
⎢2 ⎥
= ⎢ ⎥
⎢3 ⎥
⎢ ⎥
⎣4 ⎦
9] y[n] = x*[n]

Y[k] = X*[-k]
⎡1 k = 0⎤
⎢4 ⎥
Y[k] = ⎢ ⎥
⎢3 ⎥
⎢ ⎥
⎣2 ⎦
jn π
10] y[n] = e x[n]
By DFT, Frequency shift property,
y[n] = W N− mn x[n]
Y[k] = X[k – m]

To find m :
−mn
⎛ − j2π ⎞
e jnπ
= WN−mn = 1 − mn
(WN ) = ⎜e N ⎟
⎜ ⎟
⎝ ⎠
−mn
⎛ − j2π ⎞
= ⎜e N ⎟
⎜ ⎟
⎝ ⎠
π
j mn
jnπ
=e
2
e
By comparing we get m = 2
By substituting in Y[k] we get,

Y(k) = X [k - 2]
⎡3 k = 0⎤
⎢4 ⎥
Y(k) = ⎢ ⎥
⎢1 ⎥
⎢ ⎥
⎣2 ⎦

11] y[n] = e j ( n − 2)π x (n − 2)


put n – 2 = m
n=m+2
y[m + 2] = e j mπ x(m)
By DFT

W N−2k Y [k ] = X [k − 2]

Y [k ] = W N2 K X [k − 2]

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 16

⎡ WN0 ⎤ ⎡ 3 ⎤
⎢ 2⎥⎢ ⎥
W ⎥ ⎢4⎥
=⎢ N
⎢ W4 ⎥ ⎢ 1 ⎥
⎢ N6 ⎥ ⎢ ⎥
⎢⎣ WN ⎥⎦ ⎣ 2 ⎦
⎡ 1⎤ ⎡3⎤
⎢ −1⎥ ⎢ 4 ⎥
=⎢ ⎥⎢ ⎥
⎢ 1⎥ ⎢ 1 ⎥
⎢ ⎥⎢ ⎥
⎣ −1⎦ ⎣ 2 ⎦
⎡ 3 k = 0⎤
⎢− 4 ⎥
Y[k] = ⎢ ⎥
⎢ 1 ⎥
⎢ ⎥
⎣− 2 ⎦

12] y[n] = e jnπ x(n − 2)

put n – 2 = m
n=m+2

y[m + 2] = e j ( m + 2)π x(m)


jmπ j2π
= e e x(m)
jmπ
y[m + 2] = e x(m)
By DFT

WN−2k Y[k] = X[k − 2]

Y[k ] = WN2 k X (k − 2)

⎡ 3 k = 0⎤
⎢− 4 ⎥
=⎢ ⎥
⎢ 1 ⎥
⎢ ⎥
⎣− 2 ⎦

y[m + 2] = e jmπ x(m)

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305 17

--------------------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305

(68) Let x[n] = { 1, 2, 3, 4 } and x[n] ÅÆ X[k]. Find inverse DFT of the
following without using DFT/iDFT equations.

1) X[k-2] 4) X[-k+2] 7) 2 X[K] 10) e j π k X[k-2]


2) X[k+2] 5) X[-k-2] 8) 8 + X[K] 11) e j 1 . 5 π k X[–k]
3) X[-k] 6) X 2 [k] 9) e j π k X[k] 12) X*[–k]

Solution :

1] Y[k] = X [k – 2]
⎡ 1 k =0
By i DFT , frequency shift property ⎢ −4
iDFT { X [-k + 2] } = ⎢
y[n] = WN−2n x(n) ⎢ 3

⎣ −2
⎡ 1⎤ ⎡1⎤
⎢ −1⎥ ⎢ 2 ⎥
=⎢ ⎥ ⎢ ⎥ 5] Let Y[k] = X[-k]
⎢ 1⎥ ⎢ 3 ⎥ put k = k + 2
⎢ ⎥⎢ ⎥
⎣ −1⎦ ⎣ 4 ⎦ Y[k + 2] = X[– k– 2]
⎡ 1 n = 0⎤
⎢− 2 ⎥ WN2n y(n) = iDFT {X[−k − 2]}
y[n] = ⎢ ⎥
⎢ 3 ⎥ iDFT {X[-k-2] = WN2 n y(n )
⎢ ⎥
⎣− 4 ⎦ = WN2 n x (− n )
2] Y[k] = X [ k + 2] ⎡ 1⎤ ⎡1⎤
y(n) = WN2n x(n) ⎢ −1⎥ ⎢ 4 ⎥
=⎢ ⎥ ⎢ ⎥
⎡ 1⎤ ⎡1⎤ ⎢ 1⎥ ⎢ 3 ⎥
⎢ −1⎥ ⎢ 2 ⎥ ⎢ ⎥⎢ ⎥
=⎢ ⎥ ⎢ ⎥ ⎣ −1⎦ ⎣ 2 ⎦
⎢ 1⎥ ⎢ 3 ⎥ ⎡ 1 n = 0⎤
⎢ ⎥⎢ ⎥ ⎢− 4 ⎥
⎣ −1⎦ ⎣ 4 ⎦
= ⎢ ⎥
⎡ 1 n = 0⎤ ⎢ 3 ⎥
⎢− 2 ⎥ ⎢ ⎥
y[n] = ⎢ ⎥ ⎣− 2 ⎦
⎢ 3 ⎥
⎢ ⎥ 6] Y[k] = X2[k]
⎣− 4 ⎦
Y[k] = X[k] X[k]
3] Y[k] = X[-k]
By iDFT Time Reversal Property, iDFT, Circular Convolution property
y(n) = x(-n) = { 1, 4, 3, 2 }
y(n) = x(n) * x(-n)
N −1
4] Let Y[k] = X[-k] = ∑ x(m) x[n − m]
put k = k – 2 m =0

Y[k – 2] = X[- k + 2]
ANS : y(n) = { 26, 28, 23, 20 }
By iDFT freq Shift property,

WN−2n y(n) = iDFT { X[-k+2]} 7] Y[k] = 2 X [k]


By DFT,
iDFT { X [-k + 2] }= WN−2 n y(n ) y[n] = 2 x(n)
= 2 [ 1,2,3,4 ]
= WN−2 n x (− n )
⎡ 1⎤ ⎡1 ⎤ y[n] = { 2,4,6,8 }
⎢ −1⎥ ⎢4 ⎥
= ⎢ ⎥ ⎢ ⎥
⎢ 1⎥ ⎢3 ⎥
⎢ ⎥ ⎢ ⎥
⎣ −1⎦ ⎣2 ⎦
Anand Academy of Engineering, Tel 2642 5305
y(n) = W-2n x(n-2)
8] Y[k] = 8 + X[k]
⎡ 1⎤ ⎡3⎤
By iDFT ⎢ −1 ⎥ ⎢4⎥
=⎢ ⎥ ⎢ ⎥
y[n] = 8 δ (n) + x(n) ⎢ 1⎥ ⎢1⎥
⎢ ⎥ ⎢ ⎥
⎡1 ⎤ ⎡ 1 ⎤ ⎣ −1 ⎦ ⎣2⎦
⎢0 ⎥ ⎢ 2 ⎥
=8 ⎢ ⎥+⎢ ⎥ ⎡ 3 n = 0⎤
⎢0 ⎥ ⎢ 3 ⎥ ⎢−4 ⎥
⎢ ⎥ ⎢ ⎥ y ( n) = ⎢ ⎥
⎣0 ⎦ ⎣ 4 ⎦ ⎢ 1 ⎥
⎢ ⎥
⎡9 n = 0⎤ ⎣−2 ⎦⎥
⎢2 ⎥
y(n) = ⎢ ⎥
⎢3 ⎥

⎢ ⎥ j k
⎣4 ⎦ 11] Y[k] = e 2
X[−k ]

= WN−mk X[−k]
9] Y[k] = e jπk X [k ]
Y[k] W Nmk = X [−k ]
= WN−mk X[k] By iDFT,
By iDFT , Time shift property, y[n – m] = x[–n]
y[n] = x[n – m] put n – m = n
To find m : n= n+m
e jπk = WN−mk = (WN
1 −mk
) y[n] = x[–n – m]

−mk To find m :
⎛ − j 2π ⎞
= ⎜e N ⎟





e
j
2
k
= WN− mk = WN1 ( ) − mk

−mk
π ⎛ − j 2π ⎞
= ⎜e N ⎟
j mk
e jπk = e2 ⎜ ⎟
⎝ ⎠
m=2
−mk
By substituting, ⎛ − j 2π ⎞
= ⎜e 4 ⎟
y[n] = x[n – 2] ⎜ ⎟
⎝ ⎠
y[n] = { 3, 4, 1, 2 } 3π π
j k j mk
e 2
= e 2

10] Y[k] = e jπk X [k − 2] By comparing we get m = 3

put p = k – 2 By substituting,

k=p+2 y[n] = x[–n –3]

Y(p + 2) = e jπ ( p + 2) X [ p ] where x[-n] = {1, 4, 3, 2 }


Then y[n] = {2, 1, 4, 3}
= e jπp e j 2π X [ p ]

But e j 2 x = 1
12] Y[k] = X * [-k]
jπp
Y [p + 2] = e X [ p] y[n] = x * [n]
By iDFT = { 1, 2, 3, 4 } ANS
W N2 n y (n) = iDFT {e jπp X [ p]}

WN2n y(n) = iDFT {e jπp X[p]}

WN2n y(n) = x[n − 2]

20
Anand Academy of Engineering, Tel 2642 5305

(69) Given x(n) = { 1, 1 , 1, 1, 0, 0, 0, 0}. Let X[k] be 8 point DFT of X[k].


Find DFT of the following sequences in terms of X[k].
A). a [n] = { 0, 0, 0, 0, 1, 1, 1, 1 } E). e [n] = { 1, 1, 1, 1, 1, 1, 1, 1}
B). b [n] = { 1, 0, 0, 0, 0, 1, 1, 1} F). f [n] = { 0, 0, 1, 1, 1, 1, 0, 0}
C). c [n] = { 1, 0, 0, 0, –1, 0, 0, 0 } G). g [n] = { 1, –1, 1, –1, 0, 0, 0, 0}
D). d[n] = { 1, 1, 1, 1, –1, –1, –1, –1} H). p [n] = { 1, 0.5, 0.5, 0.5, 0, 0.5, 0.5, 0.5}
Solution :
(A) Let a[n] = x[n – 4] (E) Let e[n] = x [n] + a[n]
By DFT Time shift property, E[k] = X [k] + A[k]

A[k ] = W N4k X [k ]
(F) Let f(n) = x(n – 2 )
A [k] = (–1)k X [k]
By DFT Time shift property,

(B) Let b[n] = x[– n] F [k ] = W N2k X [k ]


By DFT and Time reversal (G) Let g(n) = (-1)n x(n)
property, B[k] = X [– k] = WN4n x(n)

By DFT frequency shift property


(C) Let c[n] = b[n] – a [n]
G [k] = X[k – 4]
By DFT Linearity property,
1
C[k] = B[k] – A[k] (H) Let p[n] = [ x ( n) + x ( − n) ]
2
By DFT,
(D) Let d[n] = x[n] – a [n] 1
P[k ] = [ X (k ) + X (− k ) ]
By DFT, Linearity property 2
=Xe[k] = Real {X[k]}
D[k] = X[k] – A[k]

(70) Consider the finite length sequence x[n] = δ[n] + 2 δ[n-5]. Find 10-point DFT of x[n]
Solution : x[n] = {1, 0, 0, 0, 0, 2, 0, 0, 0, 0}
N −1
By DFT, X [k ] = ∑ x[n] W N = 10
nk
N ,
n=0
9
∴ X[k ] = ∑ x[n ]WN nk
n =0
X[k] = x[0] WN0 + x[5]WN5k
X[k] = 1 + 2 W105k
X[k] = 1 + 2 (W105 )k
X[k] = 1 + 2 (– 1 )k
X[k] = { 3, –1, 3, –1, 3, –1, 3, –1, 3, –1 } ANS
--------------------------------------------------------------------------------------------------------------
(71) Let X[k] = { 1, –2, 1–j, 2j, 0, ….} is the 8 point DFT of a real valued sequence.
What is the 8 point DFT Y[k] such that y[n] = (–1)n x [n] ?
Solution :
By symmetry property of real sequence, X[k] = X*[-k]
X[k] = { 1, –2, 1–j, 2j, 0, -2j, 1+j, -2}
To find Y[k] : y[n] = (–1)n x[n]

= (WN4 ) n x[n]

y[n] = W 4 n x[n]
N

21
Anand Academy of Engineering, Tel 2642 5305

By DFT frequency shift DFT { Wn4mn x[n] } = X[ k + m]


Y[ k] = X[k + 4]
∴Y[k] = { 0, -2j, 1+j, -2, 1, –2, 1–j, 2j } ANS
(72) x1[n] and x2[n] are sequences of length four such that x1[n] is time reversed version x2[n] and
x1[n] = {1, 2, 3, 4}. If x[n] = x1[n] – j x2[n]. Find X[k] in terms of X1[k].
Solution :
x1[n] is time reversed version x2[n] i.e. x2[n] = x1[–n]
By DFT, Time Reversal Property, X2[k] = X1[–k]

Given that x[n] = x1[n] – j x2[n]


By DFT Linearity property,
X[k]= X1[k] – j X2[k]
 X[k] = X1[k] – j X1[–k] ANS
--------------------------------------------------------------------------------------------------------------
(73) Let x[n] = {a, b, c, d} and the corresponding DFT X[k] = {A, B, C, D}.
Find the DFT of the following sequences using X[k] only and not otherwise.

Solution : (a) p[n] = {a, 0, b, 0, c, 0, d, 0}

Let p[2r] = {a, b, c, d} = x[n] and P[2r + 1] = {0, 0, 0, 0} = 0


N
2
−1 N
2
−1

P[k ] = ∑ p[2r ]W N + ∑ p[2r + 1]WN


2 rk ( 2 r +1) k

r =0 r =0
P[k] = X[k]

ANS : P[k] = { A, B, C, D, A, B, C, D }
(b) q[n] = {a, 0, 0, b, 0, 0, c, 0, 0, d, 0 , 0}

Let q[3r] = {a, b, c, d} = x[n]


q[3r + 1] = {0, 0, 0, 0} = 0
q[3r + 2] = {0, 0, 0, 0} = 0

3 3 3
Q[k ] = ∑ q[3r ]WN 3rk + ∑ q[3r + 1]WN (3r +1) k + ∑ q[3r + 2]WN (3r + 2) k
r =0 r =0 r =0
3 3 3
 Q[k ] = ∑ q[3r ]W N rk
+ W Nk ∑ q[3r + 1] W N rk +
2k
WN ∑ q[3r + 2] W N rk
r =0 3 r =0 3 r =0 3
2k
Q[k ] = DFT{ q[3r ] } + W Nk DFT { q[3r + 1]} + W N DFT{ q[3r + 2] }
2k
Q[k ] = DFT{ x[n] } + W Nk DFT { 0 } + W N DFT{ 0 }
Q[k] = X[k]

ANS : Q[k] = { A, B, C, D, A, B, C, D, A, B, C, D }
------------------------------------------------------------------------------------------------

⎛ 2πn k ⎞
(74) Compute the energy of N pt sequence x [n] = cos ⎜ ⎟, 0 ≤ n ≤ N − 1
⎝ N ⎠
j 2 π nk −j2 π n k − j 2 π nk j2 π n k ⎤
1⎡ ⎤
* 1⎡
Solution : Let x(n) = ⎢ e N +e N ⎥ and x [n] = ⎢ e N +e N ⎥
2⎢ ⎥ 2⎢ ⎥
⎣ ⎦ ⎣ ⎦
j 4 π nk −j4 π n k
* 1⎡ ⎤
x[n] x [n] = ⎢ 2 + e N +e N ⎥
4⎢ ⎥
⎣ ⎦

E = ∑ x[n] x*[n]
n = −∞
N −1 j 4 π nk − j4 πn k
1
=∑ [2 + e N
+e N
]
n =0 4

22
------------------------------------------------------------------------------------------------------
--DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305
N −1
1 ⎡ ⎛ 4 π nk ⎞⎤
= ∑ 4 ⎢2 + cos⎜ N

⎟⎥
⎠⎦
n=0 ⎣
N
E=
2
--------------------------------------------------------------------------------------------------------------
(75) A four point DT signal x(n) is given by x(n) = [1, 2, 0, 2]. A student found the DFT of this sequence
as X[k]= [ 5, (-1 + j2), -3, (-1 –j2) ] Guess whether this answer is correct or not, without
performing DFT. Justify your guess.
Solution : Here x[n] = x[–n]. Therefore by even signal property of DFT,
If x[n] = x[–n] then X[k] = X[–k]
That means if x[n] is Even Then X[k] is also Even.
But given X[k] is NOT Even { X[k] ≠ X[–k] }
So, Answer is NOT correct.
--------------------------------------------------------------------------------------------------------------
(76) Derive FFT flowgraph for N=2
(77) Derive FFT flowgraph for N= 3
--------------------------------------------------------------------------------------------------------------
⎛ 2πn ⎞ ⎛ 2πn ⎞
(78) Given x1 (n) = cos ⎜ ⎟ x 2 (n) = sin ⎜ ⎟
⎝ N ⎠ ⎝ N ⎠
Determine the N point Circular convolution of x1(n) & x2 (n)
⎛ j 2π n − j 2π n ⎞
1⎜ N ⎟
Let x1 (n) = +e N
⎟ u[n]
2 ⎜⎜
e

⎝ ⎠
N
By DFT, X 1 [k ] = [δ (k − 1) + δ (k + 1)]
2
j2π n −j2 π n ⎞
1 ⎛⎜ N
Similarly, x 2(n) = e − e N ⎟ u[n]
2j ⎜ ⎟
⎝ ⎠
N
By DFT, X 2 [k ] = [δ (k − 1) − δ (k + 1)]
2j

Let x3[n] = x1[n] * x2[n]


By DFT circular convolution property,

N2
X3[k] = X1[k] X2[k] = [ δ (k − 1) − δ (k + 1) ]
4j

N ⎛ 2πn ⎞
By inverse DFT, x 3 [n ] = sin ⎜ ⎟
2 ⎝ N ⎠

23
------------------------------------------------------------------------------------------------------
--DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305

(79) Determine N pt DFT of the following sequences in terms of X[k].


⎛ 2πn k ⎞ ⎛ 2πn k ⎞
(a) p [n] = x[n] cos ⎜ ⎟ (b) q [n] = x[n] sin ⎜ ⎟
⎝ N ⎠ ⎝ N ⎠
Solution (a) Solution (b)
⎡ ⎛ 2π n ko ⎞⎤
q(n) = x(n) ⎢ sin ⎜ ⎟
p [n] = x[n] cos ⎜
⎛ 2πn k ⎞
⎟ ⎣ ⎝ N ⎠⎥⎦
⎝ N ⎠ ⎡ 1 ⎛ j 2π n k o − j 2π n k o
⎞⎤
⎡1 ⎛ j2πk o q[n] = x[n]⎢ ⎜ N
−e N ⎟⎥
− j2 π k o
⎞⎤ 2 ⎜e ⎟
p [n ] = x [n ] ⎢ ⎜⎜ e N
+ e N ⎟⎟ ⎥ ⎣⎢ j ⎝ ⎠⎦⎥
⎢⎣ 2 ⎝ ⎠ ⎥⎦ j 2π n k o − j 2π n k o
1 ⎡ ⎤
j 2π n k o − j 2π k o q[n] = ⎢e
N
x[n] − e N x(n) ⎥
1⎡ ⎤ 2j ⎣
p [n] = ⎢e N x[n] + e N x(n)⎥ ⎦
2⎣ ⎦ 1
q[n] = ( W N− nko x[n] − W Nnko x[n] )
1 2j
= ( W N− nko x[n] + W Nnko x[n] )
2
By DFT Frequency shift property, By DFT Frequency shift property
1
1 ANS : Q (k ) = ( X [k − ko] − X [k + ko] )
ANS : P [k ] = ( X [k − k o ] + X [k + k o ] ) 2j
2
--------------------------------------------------------------------------------------------------------------

(80) Find DFT of the following signals and plot magnitude spectrum.
{
(a) x2 [n] = 1 2 3 4

} x [ n] = { 1
3 2 3 4

}
Solution :

(a) By DTFT, X 2 ( w) = ∑ x 2 [ n] e − jnw
n = −∞

X 2 (w) = x2 [−1] e j w + x2 [0] + x2 [1] e − j w + x2 [2] e − j 2 w


X 2 ( w) = e j w + 2 + 3 e − j w + 4 e − j 2 w
X 2 ( w) = [ 2 + 4 cos ( w) + 4 cos (2 w)]− j [ 2 sin ( w) + 4 sin (2 w )]

But X [k ] = X ( w) w= 2π k where N =4
N

X 2 [k ] = X 2( w) w= π k
2

⎡ ⎛ πk⎞ ⎤ ⎡ ⎛π k ⎞ ⎤
X 2 [k ] = ⎢ 2 + 4 cos ⎜ ⎟ + 4 cos (π k )⎥ − j ⎢ 2 sin ⎜ ⎟ + 4 sin (π k )⎥
⎣ ⎝ 2 ⎠ ⎦ ⎣ ⎝ 2 ⎠ ⎦
⎡ 10 k =0 ⎤
⎢ −2 − 2j ⎥
X 2 [k ] = ⎢ ⎥
⎢ 2 ⎥
⎢ ⎥
⎣ −2 + 2j ⎦

Magnitude Spectrum Æ

0 0.5π π 1.5π 2π
------------------------------------------------------------------------------------------------------------------------

24
------------------------------------------------------------------------------------------------------
--DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305


(b) By DTFT, X 3 ( w) = ∑ x 3 [ n] e − jnw
n = −∞
X 3 ( w ) = x 3 [ − 2 ] e j 2 w + x 3 [ − 1] e j w
+ x 3 [ 0 ] + x 3 [1] e − j w

X3 (w) = e j w +2 + 3 e− j w + 4 e− j 2w
X 3 ( w) = [ 2 + 4 cos ( w) + 4 cos (2 w)] − j [ 2 sin ( w) + 4 sin (2 w )]

But X [k ] = X ( w) w= 2π k where N =4
N

X 3 [k ] = X 3( w) w= π k
2
⎡ ⎛ πk⎞ ⎤ ⎡ ⎛π k ⎞ ⎤
X 3 [k ] = ⎢ 2 + 4 cos ⎜ ⎟ + 4 cos (π k )⎥ − j ⎢ 2 sin ⎜ ⎟ + 4 sin (π k )⎥
⎣ ⎝ 2 ⎠ ⎦ ⎣ ⎝ 2 ⎠ ⎦

⎡ 10 k =0 ⎤
⎢ −2 − 2j ⎥
X 1 [k ] = ⎢ ⎥
⎢ 2 ⎥
⎢ ⎥
⎣ −2 + 2j ⎦

Magnitude Spectrum Æ

0 0.5π π 1.5π 2π

25
------------------------------------------------------------------------------------------------------
--DSP Help Line : 99 870 30 881---
3 . ANALYSIS OF DISCRETE TIME SYSTEM

FAQ ” What & Why V 

(81) W hy ZT i s u s ed f or f re q u e nc y dom a i n a na l ysi s of D T s y s te ms i n s tea d of D TFT ?


Ans : DTFT of every input signal is not possible. DTFT of u[n] is not possible because u[n] is not an energy
signal. However ZT of u[n] is possible. Therefore ZT is used for analysis.
(82) W ha t is t he Z T of δ[n] a nd u[n]
Ans : ZT {δ[n]} =1 and ZT{u[ n]} = z /(z-1)
(83) W ha t is t he Z T of x[ n] = (2) n u[n]
Ans : X(z) = z /(z-2) RO C : |z | > 2 wh er e z = e j w

(84) W ha t i s t he c on ce p t of R O C ?
Ans : ROC gives the set of values of Z for which X(z) is finite. Every value of Z in the ROC gives X(z) finite.
(85) W ha t is the R O C co n diti o n f or causa l si gna l. ? W hy ? Jus tify wi th e xampl e.
Ans : ROC is |z| > | Largest value of POLE |
Ex x[n] = (2)n u[n] + (3)n u[n]

(86) Wha t is the R O C conditi o n f or Anti- causa l signal ? Why ? Ju stify with exa mple.
Ans : ROC is |z| < | Lowest value of POLE |
Ex x[n] = (2)n u[-n] + (3)n u[-n]

(87) Wha t is the R O C conditi o n f or Both-si ded s i g na l . ? W hy ? J u s ti f y w i t h e xa mp l e.


Ans : ROC condition for both sided signal is bounded between two POLES.
Ex x[n] = (2)n u[n] + (3)n u[-n]

1) If x[n] is right handed sequence, the ROC extends outward from the
outermost finite pole in X ( z ) to z = ∞
Sequence ROC

1 x[n] = { 1, 0, 0, 0 } Entire Z-plane

2 x[n] = { 1, 2, 3, 4 } |Z| > 0

3 x[n] = an u[n] |Z| > |a|


4 x[n] = an u[n] + bn u[n] |Z|> max{ |a |,|b| }
5 x[n] = (-3)n u[n] + (2)n u[n] |Z| > 3

2) If x[n] is Left handed sequence, the ROC extends inward from the
innermost finite pole in X(z) to z = 0

Sequence ROC
1 x[n] = { 1, 2, 3, 0 } |Z| < ∞
2 x[n] = an u[-n-1] |Z| < |a|
3 x[n] = an u[-n-1] + bn u[-n-1] |Z| < min { |a|, |b| }
4 x[n] = (-3)n u[-n-1] + (2)n u[-n-1] |Z| < 2
Anand Academy of Engineering, Tel 2642 5305

3) If x[n] is two sided sequence, the ROC consist of a ring in the Z plane,
bounded by interior and exterior pole.
Sequence ROC
1 x[n] = an u[n] + bn u[-n-1] |b| > |z| > |a|
2 x[n] = (2)n u[n] + (3)n u[-n-1] 3 > |z| > 2

3 x[n] = (3)n u[n] + (2)n u[-n-1] Not possible

4 x[n] = (2)n u[n] + (3)n u[n] + 4 > |z| > 3


(–4)nu[-n-1] + (5)nu[-n-1]

(88) What is canonic strucure ?


Ans : If the number of delays in the relaization block diagram is equal to the order of
the transfer function, then the relaization structure is called canonic otherwise
it is called non-canonic.
(89) What is DT sy stem ?
Ans : A DT system is a device or algorithm that operates on a DT signal according to some well defined
rule, to produce – another DT signal. In general a DT system can be thought as a set of operations
performed on the input signal x[n] to produce the output signal y[n].

(90) What are the classification of DT systems ?


Ans : Systems are classified as,

(1) Static (Memorylees ) / Dynamic (Memory System) :-


(2) Linear / Non Linear System.
(3) Causal / Non Causal System
(4) Time Invariant / Time Variant System.
(5) Stable / Unstable system

CLASSIFICATION OF DT SYSTEMS :-

(1) Static (Memorylees ) / Dynamic (Memory System) :-


A DT system is called static or memoryless if it output at any instant depends on the input sample
at the same time and not on past or future samples of the input. If the system is not static then it is
dynamic.
(2) Linear / Non Linear System.
A system that satisfies the superposition principle is called Linear System.
If a system is Linear then,
T { a . x1[n] + b x2[n] } = a1 T {x1 [n]} + a2 T {x2 [n] }
If a system does not satisfy the superposition principle then it is Non Linear System.
(3) Causal / Non Causal System
A system is said to be causal if the output of the system at any time depends only on present and
past values of input and does not depend on future values of input.
If the system is not causal then it is Non casual. For non causal system output depends on future
values of input.
(4) Time Invariant / Time Variant System.
A system is called Time Invariant if a time shift in the input signal causes a time shift in the output
signal.Otherwise the system is Time Variant System.

27
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305
(5) Stable / Unstable system.
A system is said to be bounded input, bounded o/p stable if and only if every bounded input
produces a bounded output.

(91) What is Impulse response ? Step response ?


Ans : Impulse Response is output of the system when input is δ[n].
Step Response is output of the system when input is u[n].

(92) What is zero input response ?


Ans : If the initial state of the system is NOT zero and the input x[n] = 0 to all n, then the output of the
system with zero input is called the zero input response or natural response or free response of the
system.
(93) What is zero state response ?
Ans : If the initial state of the system is zero and the input x[n] ≠ 0 then the output of the system with non
zero input is called the zero state response or forced response of the system.

(94) What is zero step response ?


Ans : If the initial state of the system is zero and the input x[n]=u[n] then the output of the system is
called zero step response of the system.

(95) What is Transient response ?


Ans : Transient response of the system is the response of the system that decays to zero after infinite
intervals of time.

(96) What is Steady State Response ?


Ans : Everlasting response of the system that depends on magnitude response and phase response of the
system is steady state response of the system.

(97) What is Infinite Impulse Response ?


Ans : When length of h[n] is infinite it is called infinite impulse response. E.g. h[n] = ( ½ )n u[n]

(98) What is Finite Impulse Response ?


Ans : When length of h[n] is finite it is called finite impulse response, E.g. h[ n ] = { 1 2, 3, 4 }

(99) What is frequency response ?
Ans : Frequency response means magnitude response and phase response.

(100) What is magnitude response ?


Magnitude of numerator
Ans : Magnitude Response =
Magnitude of Deno min ator

Where Magnitude = (Re al) 2 + (Im aginary) 2

28
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305
(101) What is phase response?
Ans : Phase Response = Angle of Numerator – Angle of denominator

⎡ −1 ⎛ Im aginary ⎞
⎢ tan ⎜ ⎟ When Re al > 0
Where angle = ⎢ ⎝ Re al ⎠
⎢ −1 ⎛ Im aginary ⎞
⎢180 + tan ⎜ ⎟ When Re al < 0
⎣ ⎝ Re al ⎠

(102) Magnitude spectrum is continuous or discrete ?


Ans : If the signal is periodic then magnitude spectrum is discrete and If the signal is not-periodic then
spectrum is continuous function of w.

(103) What is a digital resonator ?


Ans : A digital resonator is essentially a narrowband bandpass filter.
(104) What is eigen value of the system ?
Ans : Eigen-function of a system is an input signal that produces an output that differs
from the input by a constant multiplicative factor. The multiplicative factor is
called an eigen value of the system.

(105) How to find value of DT signal at infinity. ?


⎛ z −1⎞
Ans : By final value theorem we can find x[∞]. x(∞) = lim ⎜ ⎟ X ( z)
z →1 ⎝ z ⎠
(106) What is Transfer function of DT system ?
Ans : The Z – Transform H(z) of an impulse response h[n] is known as the system function or
transfer function of the system
(107) What are different realization methods of digital filters ?
Ans :
IIR FILTER LINEAR PHASE FIR FILTER.
1 Direct Form Realization Direct Form Realization
-DF-I -DF-I
-DF-II -DF-II
2 Lattice Realization Lattice Realization
3 Linear Phase Realization
4 Frequency Sampling Realization

(108) What is the advantage of direct form –II method of relalization ?


Ans : DF-II method of realization requires LESS no of delay block.
(109) What is the advantage of Linear Phase Realization ?
Ans : Linear Phase method of realization requires LESS no of multipliers.
(110) What is the advantage of cascade connection of systems?
Ans : In casade form the shift from the actual POLE location due to quantization is
LESS. So, quantization error is less.
(111) What is difference equation of DT system ?
Ans : Output in terms of past/present , input/output of the system is called difference of the system.

29
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305

Short Answer Questions on ZT + DT System


--------------------------------------------------------------------------
4z
(112) Given X ( z ) = | z | > 0.5 Find Z-Transform of the following using properties and
(z + 0.5)2
specify ROC in each case.

(i) y [n] = x[n – 2] (ii) d [n]= 2n x[n]


Y (z) = z–2 x(z) ⎛z⎞
4 D(z) = X ⎜ ⎟
Y (z) = ⎝2⎠
z(z + 0.5)
4 ( Z / 2) 2z
ROC : | z | > 0.5 = =
2 2
⎛z ⎞ ⎛ z +1⎞
⎜ + 0.5 ⎟ ⎜ ⎟
⎝2 ⎠ ⎝ 2 ⎠
8z
D(z) = ROC: | z | > 1
(z + 1) 2

(iii) e[n] = x [– n] (iv) s [n] = x[n] * x[n]


By ZT, Time Reversal property, s (z) = x(z) ⋅ x(z)
E(z) = x(z–1) 4z 4z
=
4z −1 (z + 0 ⋅ 5) 2
(z + 0 ⋅ 5) 2
E (z) =
(z −1 + 0 ⋅ 5) 2 16z 2 1
ROC: | z | > 2 s (z) = ROC: | z | >
(z + 0 ⋅ 5) 2 2

z 2 + 0 ⋅ 25
(113) Given H (z) = Draw pole – zero diagram of the system and indicate whether the
z 2 − 0 ⋅ 4 z − 0 ⋅ 05
system is of minimum phase or maximum phase type
z 2 + 0 ⋅ 25 z 2 + 0 ⋅ 25
Solution : H (z) = =
z 2 − 0 ⋅ 4z − 0 ⋅ 05 (z − 0 ⋅ 5) (z + 0 ⋅ 1)

Zeros:
z 2 + 0 ⋅ 25 = 0
z 2 = −0 ⋅ 25
z 2 = (−1) (0 ⋅ 5) 2
O z0
z 2 = e jπ (0 ⋅ 5) 2 e j2 πk
z 2 = (0 ⋅ 5) 2 e jπ( 2 k +1) p2 p1
⎛ 2 k +1 ⎞ O z1
jπ ⎜ ⎟
2 ⎠
Zk = 0 ⋅ 5 e ⎝
jπ / 2
k = 0, z0 = 0⋅5 e
j 3π / 2
k = 1, z1 = 0⋅5 e

Poles : P1 = 0⋅5 Since all zeros are inside unit circle.


P2 = 0⋅1 System is minimum phase type.

30
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305

(114) If the step response of the system is given by s[n] = (½)n u[n], find an impulse
response of the system without using ZT and iZT technique.
Solution:
u [n] LT I s[n]

s[n-1] Time Invariant system


u [n-1] LT I

u [n]- u [n-1] LT I s[n]- s[n-1] Linearity Properly

δ [n] LT I h[n] Impulse Response of the


system.
where h[n] = s[n]- s[n-1]
h[n] = (½)n u[n]- (½)n-1 u[n-1],

(115) Find the difference equation of the system, which generates the following output.
y[n] = { 1, 1, 2, 3, 5, 8, 13 - - - - - } for n ≥ 0
Solution:
Let y[n] = δ [n] + δ [n–1] + 2 δ [n–2] + 3 δ [n–3] + 5 δ [n-4] + 8 δ [n-5] + ----------(1)
y[n–1] = δ [n-1] + δ [n–2] + 2 δ [n–3] + 3 δ [n–4] + 5 δ [n-5] + +--------(2)
By equation (1) – (2),
y[n] – y[n-1] = δ [n] + δ [n–2] + δ [n–3] + 2 δ [n-4] + 3 δ [n-5] + ---------------(3)
from eq. (1)
y[n–2] = δ [n-2] + δ [n–3] + 2 δ [n–4] + 3 δ [n–5] + 5 δ [n-6] + +-----------------(4)
By equation (3) – (4),
y[n] – y[n-1] – y[n–2] = δ [n]
Let δ [n] = x[n].
y[n] – y[n-1] – y[n–2] = x[n]
y[n] = y[n-1] + y[n–2] + x[n]
-------------------------------------------------------------------------------------------------------------------

31
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305

4. DIGITAL FILTERS
FAQ ” What & Why V 

(116) What is Digital filter ?


Ans :
Digital filter is a discrete time System which produces a discrete time output sequence y[n] for the
discrete time input sequence x [n]. Digital filter is nothing but mathematical algorithm implemented
in hardware or software.

(117) What is Real time Digital filter?


Ans : Real time digital filter consist of processing of real time signal using digital device called digital
processor.
(118) What are the Advantages of digital filters-?
Ans : The following list gives some of the main advantages of digital over analog filters.
1. A digital filter is programmable, i.e. its operation is determined by a program stored in the
processor's memory. This means the digital filter can easily be changed without affecting the
circuitry (hardware). An analog filter can only be changed by redesigning the filter circuit.

Æ ( i.e. Flexibility in parameter setting )

2. Digital filters are easily designed, tested and implemented on a general-purpose computer or
workstation.

3. The characteristics of analog filter circuits (particularly those containing active components)
are subject to drift and are dependent on temperature. Digital filters do not suffer from these
problems, and so are extremely stable with respect both to time and temperature.

(119) What are Advantages of FIR Filters-? [Refer Theory Notes ]


Ans:
1) They can easily be designed to be "linear phase"

2) They are suited to multi-rate applications.

3) They have desirable numeric properties.

4) They can be implemented using fractional arithmetic.

5) They are simple to implement.


(120) What are the disadvantages of FIR Filters (compared to IIR filters)?

Ans : Compared to IIR filters, FIR filters sometimes have the disadvantage that they require more memory
and/or calculation to achieve a given filter response characteristic.

(121) What are the advantages of IIR filters (compared to FIR filters)?
Ans : IIR filters can achieve a given filtering characteristic using less memory and calculations than a
similar FIR filter.

32
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305

(122) What are the disadvantages of IIR filters (compared to FIR filters)?
Ans : 1) They are more susceptible to problems of finite-length arithmetic, such as noise generated by
calculations, and limit cycles. (This is a direct consequence of feedback: when the output isn't
computed perfectly and is fed back, the imperfection can compound.)

2) They are harder to implement using fixed-point arithmetic.

3) They don't offer the computational advantages of FIR filters for multirate (decimation and
interpolation) applications.

(123) Compare FIR filters and IIR filters

FIR filter IIR filter


1 Provides exact linear phase. Not linear phase.
2 Provides good stability. Stability is not guaranteed.
3 Order required is higher. Order required is lower.
4 Computationally not efficient. Computationally more efficient.
5 More memory required for the storage of Less memory required fro storage of coefficients.
coefficients.
6 Requires more processing time. Requires less processing time.
7 Requires N multiplications per output Requires 2N + 1 multiplications per output sample.
sample

(124) What is the relation between Analog filter pole and digital filter pole when impulse invariant
technique is used for filter design.
Ans : Z = e ST

(125) What is the relationship between Analog filter frequency and digital filter frequency when
impulse invariant technique is used for filter design.
Ans : W = ΩT

(126) Why Impulse Invariant method is not suitable for HPF / BPF design?
Ans : The the mapping from the analog frequency Ω to the freq. variable w in the digital domain is
many to one. which reflects the effect of aliasing due to sampling. A one to one mapping is thus
π π
possible only if freq. Ω lies in the principle range of − ≤ Ω ≤ .
T T
π
That means if cut off frequency of analog filter Ω c is greater than . then one to one mapping
T
from analog filter frequency to digital filter frequency is not possible. Therefore the filter such
π
as HPF or BPF with cut off frequency of analog filter Ω c greater than . can not be
T
designed using impulse invariant method.

(127) What do you mean by invariant ?


Ans : Invariant means, Not variant, ie. Doesn’t change.

(128) What is the relation between Analog filter pole and digital filter pole when BLT method is used for
filter design.

33
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305
2 ( z − 1)
Ans : S =
T ( z + 1)

(129) What is the relationship between Analog filter frequency and digital filter frequency when BLT
method is used for filter design.
2 ⎛ w⎞
Ans : Ω= tan⎜ ⎟
T ⎝2⎠
(130) Explain frequency warping in BLT.
Ans : [ Refer theory notes ]

(131) Frequency warping is needed to perform in BLT technique but not in impulse invariance
technique
(132) In BLT there is no a lia sing
Ans : Bilinear Transformation is a mapping of points from s-plane to corresponding points in the z-
plane. The BLT transforms, the entire j Ω axis in the s-plane into one revolution of the unit circle
in the z-plane ie. only once and therefore avoids the aliasing of frequency components.

(133) Explain the Mapping of points from s-plane to z–plane when Impulse Invariant Method is used
for filter design.

Case-I When σ = 0, r =1
Analog poles which lies on imaginary axis gets mapped onto the unit circle in the z-plane.

Case-II When σ < 0, r < 1,


Analog poles that lies on LEFT half of s-plane gets mapped INSIDE the unit circle in the z–plane.

Case–III When σ > 0, r > 1.


Analog poles that lies on RIGHT half of s-plane gets mapped OUTSIDE the unit circle in the
z–plane.
Æ Æ Always explain wrt diagram.

(134) What is notch filter? Give Applications of Notch filter :

Ans : A notch filter is a filter that contain one or more deep notches or ideally perfect nulls in its frequency
response characteristic.
They are useful in application where specific frequency components must be eliminated. For
example instrumentation and recording systems required that the power line frequency of 60 Hz
and its harmonics to be eliminated.

(135) What is comb filter? Give Applications of comb filter :

Ans : A comb filter can be viewed as a notch filter in which the null occur periodically across the
frequency band.
. Comb filters find applications in a wide range of practical systems such as in the rejection of power
line harmonics, is the separation of solar and lunar components from ionosphere measurements of
electron concentration and is the suppression of cluster from fixed objects in moving target
indicates (MTI) radars.

34
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305

Always Remember This……..

O rd e r N o rma l i ze d A n a lo g No rma lize d A na log


Butterworth LPF Butterworth HPF
N = 1 1 s
H (s) = H (s) =
s +1 s +1

N = 2 1
H (s) = H (s) =

N = 3 1
H (s) = H (s) =

(1) For Linear Phase filter h[n] must be either Symmetric or Antisymmetric.

Examples of Linear phase filters Examples of non Linear phase filters


h[n] = { 3, 2, 1, 2, 3 } h[n] = { 1, 2, 3, 1, 2, 3 }
h[n] = { 1, 2, 2, 1 } h[n] = { 3, 2, 1, 2, 3 }
h[n] = { 1, -2, 0, 2, -1 } h[n] = { 3, 2, 1, -2, -3 }
h[n] = δ[n] + δ[n-3] h[n] = { 3, -2, 2, 3 }

(2) For linear Phase FIR filter.

a) ZEROS are always in reciprocal order (ie linear Phase)


b) POLES are always only at origin (ie FIR)
⎛ 1⎞
⎜ z − ⎟ (z − 2)
2⎠
ex h [n] = { 1, –2.5, 1} H (z) = ⎝
Z2
(z + 1) (z − 1)
h [n] = { 1, 0, -1} H (z) =
Z2
⎛ 1⎞
(z + 1) ⎜ Z − ⎟ (z − 2)
⎝ 2⎠
h [n] = {1, -1.5, -1.5, 1} H (z) = 3
Z

(3) When zeros of the filter are INSIDE the unit circle filter is called Minimum Phase Filter.
Concept : For Minimum Phase filter φ(π) - φ(0) = 0

(4) When all zeros of the filter are OUTSIDE the unit circle filter is called maximum phase
filter.
Concept : For Maximum Phase filter φ(π) - φ(0) = ± m π

35
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305
(5) When all zeros of the filter are INSIDE and OUTSIDE the unit circle filter is called mixed
phase filter.

(6) When all zeros of the FIR filter are LEFT side of Z-plane, filter is LOW PASS FIR
FILTER.

(7) When all zeros of the FIR filter are RIGHT side of Z-plane, filter is HIGH PASS FIR
FILTER.

(8) When poles and Zeros of the filter are in reciprocal order, filter is ALL PASS FILTER.
Z−2
Eg. H (z) = POLE P1 = 0.5 ZERO Z1 =2
Z − 0.5

(1) For Linear Phase FIR filter h[n] must be either Symmetric OR Antisymmetric.
(2) When h[n] is either Symmetric OR Antisymmetric, ZEROS of the filter are always in
Reciprocal order.
1
i.e. If Z1 is ZERO of the filter, Then is also a ZERO of the filter.
z1
(3) If ZEROS of the filter are in reciprocal order, then filter is Linear Phase FIR filter.

(136) What is a linear phase filter?


Ans : "Linear Phase" refers to the condition where the phase response of the filter is a linear (straight-
line) function of frequency

(137) What is the advantage of Linear Phase ?


Ans : This results in the delay through the filter being the same at all frequencies. Therefore, the filter does
not cause "phase distortion" or "delay distortion".

36
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305

(138) Explain the concept of Linear Phase and its importance.


Ans : I. If the Phase Response is Linear the output of the Filter during pass-band is delayed input.
II. If the phase Response is non Linear the output of the filter during pass-band is distorted one

The linear Phase characteristic is important when the phase distortion is not tolerable.

FIR Filter can be designed with linear phase characteristic. In application like data transmission,
speech processing etc phase distortion can not be tolerated and here linear phase characteristic of
FIR filter is useful
(139) Show that if the Phase Response is Linear the output of the Filter during pass-band is delayed
input.
Consider a LPF with frequency response H(ejw) given by
⎧ e − jwα | w | ≤ wc
H (e jw ) = ⎨
⎩ 0 wc < w ≤ π
x[n] y[n]
H(z)
X (w) Y (w)

Let X(w) = DTFT { X[n] } ,


The FT of y[n] is then given by
Y(w) =X(w) . H(w)
Y(w) = X(w) . e–jwα
By iDTFT,
y[n] = x[n – α] ← o/p of filter

(140) What is the condition for linear phase?


Ans : A FIR filter is linear-phase if (and only if) its coefficients are symmetrical OR Anti-symmetric
around the center coefficient

(141) What is the role of window in the design of FIR filter ? Name the few types of windows.
Ans : FIR filter is designed by truncating infinite samples of hd[n] by using window function.
Examples of window function include, Hamming window, Bartlet Window, Hanning window,
Blackman window etc

(142) Why rectangular window is not preferred for FIR filter design ?
Ans : Rectangular window function has As = 21 db which is very small compared to other window
function. Larger value of As desired.

(143) Is the following filter a linear phase filter. If yes, what is the type of filter ? It’s transfer function is
given by H(z) = 1 – z –4 .
Ans : By IZT h[n] = { 1, 0, 0, 0, –1 } Since h[n] is anti-symmetric, filter is a linear phase FIR filter.
Antisymmetric h[n] with N odd is suitable only for Band Pass Filter.
(i) At w = 0, z = 1 : H(w) = 0
(ii) At w = π, z = – 1: H(w) = 0
(iii) At w = π/2, z = j: H(w) = 2

37
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305
(144) What is the advantage of frequency sampling realization ?
Ans : The frequency sampling realization of this filter is computationally more efficient than the direct
form realization.
Justification : When the desired frequency response characterization of the FIR filter is
narrowband, most of the coefficients H[k] are zero. The corresponding filter sections can be
eliminated and only the filters with non zero coefficients need to be retained.
The net result is a filter that requires fewer computations (multiplications and additions) than the
corresponding direct form realization. Thus frequency Sampling realization is more efficient
realizations.
(145) Why an tisymmetric h[n] is no t su itab le for LPF filter d esign ?
(146) Why symme tric h[n ]with N ev en and an ti-symm h [n ] with N odd is no t su itab le for
HPF d esign ?

(147) Explain Linear phase FIR filt er design using window.


(148) Explain frequency sampling method of FIR filter design ?

(149) Why IIR filters are called as recursive filters


Ans : In IIR filter output depends on output values.
e.g. y[n] = x[n] + x[n-1] + y[n] + y[n-1].
Therefore IIR filters are also called as Recursive Filters

(150) Why FIR filters are called as Non-recursive filters?


Ans : In FIR filter output depends only on input values. It doesn’t depend on output values.
e.g. y[n] = x[n] + x[n-1]
Therefore FIR filters are also called as Non-Recursive Filters
(151) Explain how to find output of digital FIR filter in real time application.
Ans : In real time applications, output of FIR filter is obtained using overlap add method / overlap save
method.
(152) Explain how to find output of digital IIR filter in real time application.
Ans : In real time applications, output of IIR filter can be obtained by evaluating difference equation.

(153) Can we use Overlap Add Method and Overlap Save Method to find output of IIR filter for long data
sequence.
Ans : No
:

38
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305

Short Answer Questions


-------------------------------------------------------------------------
(154) Determine following statements are TRUE OR FALSE

(A) Linear phase Filters are always IIR :

FALSE.

For linear phase filter h[n] must be either symmetric or Antisymmetric. For symm h[n] phase
⎛ N −1⎞ π ⎛ N −1⎞
φ = −⎜ ⎟ w and for antisymm h[n] phase φ = − ⎜ ⎟ w where N is length of h[n]. ie
⎝ 2 ⎠ 2 ⎝ 2 ⎠
⎛ N −1⎞
Finite value. Group delay ⎜ ⎟ is finite value. Therefore linear phase filters are always FIR
⎝ 2 ⎠
Filters

(B) A stable filter is always causal.


FALSE
A stable filter can be causal, or Non-causal (i.e. Anti-causal or Both-sided)
For stable filter, ROC must include unit circle. For causal & stable filter all poles must line inside the unit
z
circle eg. H ( z ) = | z | > 0.5 For Anti-causal & stable filter all poles must lie outside the unit
z − 0 .5
z
circle eg H ( z ) = | z| > 2
z−2
z z
Ex. Both sided & stable filter H ( z) = + 2 > | z | > 0.5
z − 0 .5 z − 2

(C) If a linear phase filter having Anti symmetric even number of coefficients, then the filter acts like a
band pass filter only.
FALSE
Anti-symm h[n] with N even has definite zero at z = 1. i.e. w = 0 H ( z) z =1 = 0
w =0
That means low frequency components will get attenuated and zero frequency components will not get
passed. However, the filter can pass high frequency components, therefore it can be used for HPF design
also.

(D) A stable, causal FIR filter has its poles lying anywhere inside the unit circle in the x plane :
FALSE

In case of causal FIR Filter, poles are always only at origin.


i.e. pole = 0 . For causal and stable filter all the poles must lie inside the unit circle. Therefore, FIR
filters are always stable filter with poles only at origin

(E) IIR filters have recursive realization always :


TRUE
In case of IIR Filter, poles can be anywhere n the z plane. Due to pole position, the transfer function
H(z) of IIR filter has the form,
b 0 + b1 + ⋅ ⋅ ⋅ ⋅ ⋅ + b m z
H (z) = .
1 + a1 + ⋅ ⋅ ⋅ ⋅ ⋅ + a N z −N
39
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305
This gives difference equation in the form,
y[n] = b0 x[n] + b1 x [n – 1] + ⋅⋅⋅⋅⋅bm x[n – m] + a1 y [n – 1] + ⋅⋅⋅⋅⋅+ a2 y[n–N] o/p of the filter interms
of past output leads to recurisve realization.
(155) Find the response of the Linear Phase FIR filter with impulse response h[n] = δ[n]+δ[n-10] to the
⎛ π⎞
input x [ n] = 10 + 5 cos ⎜ nπ + ⎟
⎝ 4⎠
ANS : Here input signal is applied at n = – ∞. Then there is no transient response. Output is Steady State
Response of the system. H(z) = 1 – z-10
--- -----------------------------------------------------------------------------------------------------------------------------
(156) Given H(ejw) = e-j3w [ 2 + 1.8 cos(3w) + 1.2 cos(2w)+0.5 cos(w) ]. Find h[n].

Solution : H(ejw) = e-j3w [2 + 1.8 cos(3w) + 1.2 cos(2w) + 0.5 cos(w )


⎡ ⎧z3 + z−3 ⎫ ⎧z3 + z−3 ⎫ ⎧z + z ⎫⎤
∴ H(z) = z− 3 ⎢2 + 1.8⎨ ⎬ + 1 . 2⎨ ⎬ + 0.5⎨ ⎬⎥
⎢⎣ ⎩ 2 ⎭ ⎩ 2 ⎭ ⎩ 2 ⎭⎥⎦

∴H(z) = 2z–3 + 0.9z–6 + 0.6z–1 + 0.6z–5 + 0.25z–2 + 0.25


∴ H(z) = 0.9 + 0.6z–1 + 0.25z–2 + 2z–3 + 0.25z–4 + 0.6z–5 + 0.9z–6
By iZT, h[n] = { 0.9, 0.6, 0.25, 2, 0.25, 0.6, 0.9 } for n ≥ 0

----------------------------------------------------------------------------------------------------------------------------------

(157) Find the response of the second order antisymmetric, linear phase filter to the input
π
x[n] = ( ½ )n cos ( n 3 ) u[n].

ANS : Order = 2 N-1 = 3


For Anti-symmetric h[n] with N odd there are definite zeros at z = 1 and z = –1

( z + 1)( z − 1) z2 −1
So H ( z ) = = = 1 − z −2
2 2
z z
By IZT, h[n] = { 1 0 − 1 }

To find y[n]:
Let Y(z) = H(z) X(z)
= (1– z–2) X(z)
= X(z) – z–2 X(z)
By IZT, y[n] = x[n] – x[n–2]
π π
y[n] = ( ½ )n cos ( n 3 ) u[n] – ( ½ )n-1 cos { (n–1) 3 ) } u[n] ANS

----------------------------------------------------------------------------------------------------------------------------------

40
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---
Anand Academy of Engineering, Tel 2642 5305

(158) Find the response of the second order antisymmetric, linear phase filter to the input
π
x[n] = ( ½ )n cos ( n π + 3 ) .

ANS : Order = 2 N-1 = 3


For Anti-symmetric h[n] with N odd there are definite zeros at z = 1 and z = –1

( z + 1)( z − 1) z2 −1
So H ( z ) = = = 1 − z −2
2 2
z z
By IZT, h[n] = { 1 0 − 1 }

Here input is applied at n = – ∞, Therefore There is no transient response.
Output is only SSR.
To find SSR
At w = π , H(w) = 0
The SSR of the system is then given by y[n] = 0
----------------------------------------------------------------------------------------------------------------------------------

41
-----------------------------------------------------------------------------------------------------
---DSP Help Line : 99 870 30 881---

Вам также может понравиться