KIIT POLYTECHNIC

= ∑  .  


=  +    +    +    +   + ⋯

(c)  = 

  = ∑   
∞
∞

= ∑  

   
= +  +   +⋯
 

= 
− 

(d)  =  − 

  = ∑   −   
∞
∞

= ∑  

   
= +  +  + ⋯
 
   
=    + +   + ⋯
  
  
= (  ) =   ( )
 − −
 

(4) Scaling Property:

If    
↔
       
↔
(a)  = 

  = ∑  
∞
∞

= ∑  

 
= 
=
−  −

(b)  =  

Digital Signal Processing 24 Abhiram Pradhan

 KIIT POLYTECHNIC

  = ∑    
∞
∞

= ∑   

=  +    +   
=  +    + .    + ⋯
 
=  =  −  
−

 
G.P  =   , = =  < 
 

=
−
(5) Differentiation property

If    
↔

Then    − 
↔ 
×
Or       1
↔ 
Ex:-   =  
∞ ∞

  = ∑     = ∑ .  
∞ 
=  +   +   +   + ⋯ 
=    +   +   + ⋯ 
   
=  
  =    −   
   −  
=         +  
(6) Convolution property

If ,     and     

↔ ↔
Then
∞

  = ,  ×   = ∑    − 

∞
   =  .  
↔

3.3. Discus Z- transform

3.3.1. Explain poles and Zones:
 is a national function if this is ratio of two polynomials in   
   +    + ⋯ +   
  = =
   +    + ⋯ +   

   
= 
   
Zeros : The values of Z at which   =  are eared as Zero
Poles: The values of z at which   =  are called as poles
 In representation of  graphically by a pole –Zero plot (pattern) in the complier which shows
the location of poles by crosses (x) and zeros by circles (0)

 Roc should not contain any pole.

Digital Signal Processing 25 Abhiram Pradhan
 KIIT POLYTECHNIC

3.3.2. Determine Pole location and time domain behavior for causal signals.

 Here we will discuss Z-plane location of pole pair and the form (Shope) of the corresponding
signal in the time domain.

Transform Transform
(1) 
(2)  −   
(3)   

=
− −
(4) −  −   

=
− −
(5)     
 
=
 −    − 
(6)     

=
 −  −
(7)   − −  
−
(8)     
 −  
(9)   
 −  
(10)     + 
 − 
(11)    
 −   
(12) sin    sin 
 −  cos  + 
(13) cos     − cos  
 −  cos  + 
(14)   sin     sin 
  −   cos  +  
(15)   cos    − acos  
 −  cos  +  

Q-1 Find inverse Z-transform of


 =
  −  + 
By fraction method
 
 −
 
 = +
− −


Digital Signal Processing 26 Abhiram Pradhan

 KIIT POLYTECHNIC

   
∴   =    −    
  

Q-2 Find inverse Z transform of  = , ||
 

 −   + 
  =  
=
 +  +   +  + 
 
⇒  = −   
+ +
∴  −   − −  

Q-3 Find inverse Z transform of  = 
∴ By partial fraction
  
 = − +
− − −
∴  =   −   +  − − 

Q-4  =   
 
Q-5  = 
Q-1 Find inverse –Z Transform of

 = 
 −  + 
 
= 
  −  + 
  
⇒ = 
  −  −  + 

=
 −  −   − 

=
 −  − 
Now by partial fraction method
  
= +
 −  −   −   − 
  −  +  − 
⇒ =
 −  −   −  − 
⇒  =   −  +   − 
⇒  =  −  +  − 
=   +  −  − 
By Comp
⇒  =  +  −  − 
By Comparing coefficient of Z & constants
We get
 +  =  & − −  = 

 −  +  =  ⇒ − =  + 
⇒ − +  +  =  ⇒  = −
⇒ − = − 
 = —
 
⇒=  

=− =
 

 
    
 
 = −
  −   − 
Digital Signal Processing 27 Abhiram Pradhan
 KIIT POLYTECHNIC

     
⇒ =  −    
  −   − 
   
⇒   =   −    
 −   − 
   
⇒   =   − ( )
 − × −


   
=   − ( )
 −  −

By formula
   
 =     −     
  

Digital Signal Processing 28 Abhiram Pradhan

 KIIT POLYTECHNIC

CHAPTER- 4

DISCUSS FOURIER TRANSFORM: ITS APPLICATIONS PROPERTIES

4.1. Discuses Discrete furriers transform.

 Discrete Fourier Transform is a computational or mathematical tool for analyzing discreet time signal
in frequency domain.

 DFT consents   (Discrete time domain signal of infinite length to discrete frequency sequence
 of finite length.

 DFT is obtained by sampling one period of the Fourier transform at finite number of frequency points.

      =   

→



  = ∑     


⇒   =  

 
  = ∑    


⇒   =   
Both  &  are ranging from 0 to N-1
  time index since if denotes time constant  ≤  ≤  − 
 frequency index since if denotes frequency constant  ≤  ≤  − 

 =    = Twiddle factor
 −  Equally spaced sample points.
Ex 3:1 = Find DFT
  = ,,,
=


  = ∑     , = … − 


=
 

  = ∑  .  = ∑   
 
=   +   +  +   =  +  +  +  = 
 
1 
   
 = ,  = ∑   .  4 = ∑   .  2

 
 12 .3
        
=  .  +  .  2 +   −  2 + .  2

=
Digital Signal Processing 29 Abhiram Pradhan
 KIIT POLYTECHNIC

  =

  =
Ex.3.2 Find DFT of   =  for  ≤  ≤   = ,  
=  otherwise
Ex.33 Find 6 point DFT of  = ,,,,, , 

4.2. Relate DFT to other transform

Relate to fierier transform

The Fourier transform    of a finite duration sequence  having length N is given by

 
 = ∑     … . . 

Where     is a continuous transition of .
The DFT of   is given by

2 
  
  = ∑    − − − − − 

 = ,, −
By comparing with (1) & (2) we get.
 =   |2

Relate to Z- transform :-
Z- transform of finite duration ‘N’ sequence  is given by.


  = ∑     − − − − − 



But by IDFT

 2
  = ∑     


By putting   in equation (1) from equation (2) we get
 
 2
  = ∑  ∑        

 
 
 2/ 
⇒   = ∑   ∑       

 

 2 2  2 

= ∑   [ +   .   +    .    + ⋯ +      ]


2 

  [ −  .    ]

  
=

∑ 2   
   −   .   
 

  −     
= ∑    2


  −   .  

  
⇒   =  −    ∑ 2

−    
 
 Sum of first ‘n’ terms of geometries sequence is

Digital Signal Processing 30 Abhiram Pradhan

 KIIT POLYTECHNIC

1  
 = ,  = Finet number

 =  
 Sum of infinite G.P. series, is

∞ = ,  =  

4.3. Discuss property of DFT

4.4. Discuss periodicity linearity & symmetry property

Symmetry – if DFT  = 

Then DFT  ∗  =  ∗  −  =  ∗ −
It is also called as complex conjugate property
Periodicity :- If   &   one on  point DFT pair then.
  +   =   for all .
  +   =  for all 
 
Linearity :- if   ↔   &   ↔  
 

   +    ↔    +   

Where  &  are two arbitrary constants.
Multiplication of tow DFTS:
Let   &   be two finite duration sequences of length N . and their DN-point DFTS are.


  = ∑       ,  = , , … .  − 



  = ∑       ,  = , , … .  − 

Let   =  .  .
By IDFT of   is

 2
   = ∑    


  
 2 2 2
⇒    = ∑ ∑       ∑         

  
  
 
⇒    = ∑   ∑    ∑    

  



⇒    = ∑     −  − − − 


This above expression continually that multiplication of two DFTS of two sequences is equivalent to the
circular convolution of two sequences in time domain.
 −  circular convolution.

4.5. Explain multiplication of two DFT. & circular convolution

4.6. Let   &   one finite duration sequence both of length .

  &   be DFTs of   &   respecivally

Let   be another sequence whose DFT is  
Where   =   

Digital Signal Processing 31 Abhiram Pradhan

 KIIT POLYTECHNIC

From convolution theorem we know



  = ∑   .   −   − − − 

(For N number of Samples)
The equation (1) can be represented as.
  =    
Hence DFT     =   
Multiplication
If DFT   =  
DFT   =  

Then DFT    =   +  

  = cos  −  sin 
  = cos  +  sin 

Digital Signal Processing 32 Abhiram Pradhan

 KIIT POLYTECHNIC

CHAPTER- 5
FAST FOURIER TRANSFORM ALGORITHM & DIGITAL FILTERS

5.1 Compute DFT & FFT. Algorithm

 FFT is a providence for computing

 DFT of a finite series easily

 It is nothing only set of algorithm

 It is used in digital spectral analysis filter simulation, auto connection and pattern recognition.

 FFT is based on decomposition and breaking the transform into smaller transforms and combining
them to get the total transform.


  =   = Twoddle factor

 FFT algorithm basically bits two properties of twiddle factor.



(i)  2
= 

(ii)  = 

 There are two types of FFT algorithms

(i) Decimation in –time

(ii) Decimation in frequency

 In decimation –in time algorithm, the sequence for which we need the DFT is successively divided
into smaller sequences and the DFTs of these entire sequences are combined in a certain portion to
obtain the required DFT of entire sequence.

 In Decimation –in- frequency algorithm the frequency sample of the DFT are decomposed into smaller
and smaller subsequences in a certain pattern.

5.2 Explain direct computation of DFT.

DFT of a sequence  is evaluated as follows


2
  = ∑     , ≤ ≤−


Digital Signal Processing 33 Abhiram Pradhan

 KIIT POLYTECHNIC

2
Since  =   


  = ∑   , ≤  ≤ −




= ∑   +       +    


  

=∑     − ∑     +  ∑    

  


+ ∑     

By using the above formula we can complete DFT directly .
5.3 Discuss the Radix – 2 algorithm

 Radix -2 algorithm is also known as radix -2 decimation –in-time (DIT) algorithm.

 In Radix-2 algorithm number of output points (N). can be expressed as power of 2. i.e.  =  we
have  is an integer

The N-Point DFT of   is



  = ∑   ,  = ,, … .  − 


By separating  into even and add values of  we get
 

  = ∑    + ∑   

 
 
 
2 2

= ∑    + ∑  +  
 
 
 
2 2

= ∑    +  ∑  +  

 
 
 
2 2

= ∑    +  ∑    − − − −

 
Since   =   +  
=   +  + 
2 
&  =    
2
 
=  2 = 
2
From equation (1)
 
 
2 2

  = ∑    +  ∑   

2 2
 
̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
 ̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅

     
2 2
=   +   

Digital Signal Processing 34 Abhiram Pradhan

 KIIT POLYTECHNIC


 
For  ≥ ,  2
= −


Now   for  ≥ is given by


  
  =   −  −  2   − 
 
 
for = + + , … … . . ,  − 
 
Steps of Radix – 2 DIT FFT . algorithm :-
2
Twiddle factor  =   
(1)  = 

 
(2)  2 = −
(3)  = 
2
The competing formulas for FFT is given by

 =   +    for  ≤  ≤ − 


   
=   −  −  2   −  , ≤  ≤  − 
  
For a 8 point DFT/FFT for  = , , , , , , ,  the FFT values are as follows.
, , , , , , , 
  =   + 8     =   + 8  

 =   + 8    5 =   + 8  
 =   + 8    6 =   + 8  

 =   + 8    7 =   + 8  
 

  +    = 


    +    = 

Butterfly Diagram

  =   +    
   




    =

=   − 
  

Bit reversal
Bit reversal is useful in arranging the samples for calculating DIT algorithm for N=8
For N=8
put Sample epresentation eversal mple
0 000 000 0

Digital Signal Processing 35 Abhiram Pradhan

 KIIT POLYTECHNIC

1 001 100 4
2 010 010 2
3 011 110 6
4 100 001 1
5 101 101 5
6 110 011 3
7 111 111 7

Radix – 2
Divide the number of input samples by 2 till we reach minimum two samples.
Q- Draw and find FFT for a 8-point sequence
  = ,,,,,,,
 = ,   = ,   = ,   = ,   = ,  5 = , 6 = ,
 7 = 
As per bit reversal
put S1 S2

=


=

=


=

=


=

=


=
5.4 Introduction to digital filter

 Filter is defined as a device which rejects unwanted frequencies from the input signal and allow the
desired frequencies

 When this input signal is a discrete time sequence then this filter is a digital filter

 A filter is a LTI discrete time system.

 Basically, two types (i) FIR filter

(ii) IIR Filter

(i) FIR Filter

This filter whose present output sample depends on the present input sample and previous input
samples.
(ii) IIR Filter

This filter whose present output sample depends on present input, past input samples and output
samples.

Digital Signal Processing 36 Abhiram Pradhan

 KIIT POLYTECHNIC

5.5 Introduction to DSP. Architecture, familiarization of different types of processor.

Ans: Digital signal processors one of two types

1) General purpose digital signal processor

2) Special purpose digital signal processor

Introduction to DSP Architecture

DSP Architecture one of following types

1) Von Neumann architecture
2) Havard architecture
3) Super Havard architecture

1) Von Neumann Architecture :-

Advantage:
Memory
 It is cheap and requires lesser number of pins than
the Havard Architecture. CPU
Data &
 It is simple to use
Instruction

Disadvantage:

 It doesn’t permit multiple memory access.

Havard Architecture

Program Data Memory

Memory
CPU
Instruction Data Only
Only

 The Havard architecture physically separates memory for their instruction & data requiring
dedicated buses for each of them.
 Instructions and operands can therefore be fetched simultaneously
 Most DSP processors are modified Havard architecture with two or three memory buses
 It has multiport memor

Digital Signal Processing 37 Abhiram Pradhan