System Realization

4
4.1 INTRODUCTION
Systems may be continuous-time systems or discrete-time systems. Discrete-time systems
may be FIR (Finite Impulse Response) systems or IIR (Infinite Impulse Response) systems.
FIR systems are the systems whose impulse response has finite number of samples and IIR
systems are systems whose impulse response has infinite number of samples. Realization of a
discrete-time system means obtaining a network corresponding to the difference equation or
transfer function of the system. In this chapter, various methods of realization of discrete-
time systems are discussed.

4.2 REALIZATION OF DISCRETE-TIME SYSTEMS

To realize a discrete-time system, the given difference equation in time domain is to be
converted into an algebraic equation in z-domain, and each term of that equation is to be
represented by a suitable element (a constant multiplier or a delay element). Then using
adders, all the elements representing various terms of the equation are to be connected to
obtain the output. The symbols of the basic elements used for constructing the block diagram
of a discrete-time system (adder, constant multiplier and unit delay element) are shown in
Figure 4.1.

Figure 4.1 (a) Adder (b) Constant multiplier and (c) Unit delay element.
277
278 h Digital Signal Processing

Adder: An adder is used to add two or more signals. The output of adder is equal to the
sum of all incoming signals.
Constant multiplier: A constant multiplier is used to multiply the signals by a constant. The
output of the multiplier is equal to the product of the input signal and the constant of the
multiplier.
Unit delay element: A unit delay element is used to delay the signal passing through it by
one sampling time.

EXAMPLE 4.1 Construct the block diagram for the discrete-time systems whose input-
output relations are described by the following difference equations:
(a) y( n) = 0.7x (n) + 0.3 x (n  1)
(b) y(n) = 0.5y(n  1) + 0.8 x (n) + 0.4 x (n  1)

Solution:
(a) Given y(n) = 0.7x (n) + 0.3 x (n  1)

The system may be realized by using the difference equation directly or by using
the Z-transformed version of that. The individual terms of the given difference
equation are 0.7x(n) and 0.3x(n – 1). They are represented by the basic elements as
shown in Figure 4.2.
Alternatively
Taking Z-transform on both sides of the given difference equation, we have
Y(z) = 0.7X(z) + 0.3z–1X(z)

The individual terms of the above equation are: 0.7X ( z ) and 0.3 z 1 X (z ).
They are represented by the basic elements as shown in Figure 4.2.

Figure 4.2 Block diagram representation of (a) 0.7X(z) and (b) 0.3z–1X(z).

The input to the system is X(z) [or x(n)] and the output of the system is Y(z) [or
y(n)]. The above elements are connected as shown in Figure 4.3 to get the output
Y(z) [or y(n)].
 System Realization h 279

Figure 4.3 Realization of system described by y(n) = 0.7x(n) + 0.3x(n–1).

(b) Given y( n) = 0.5y ( n  1) + 0.8x ( n) + 0.4x ( n  1)

The individual terms of the above equations are 0.5 y(n  1), 0.8x ( n) and 0.4x (n  1) .
They are represented by the basic elements as shown in Figure 4.4.
Alternatively
Taking Z-transform on both sides of the given difference equation, we have
Y (z ) = 0.5 z 1Y (z) + 0.8X ( z ) + 0.4 z 1 X (z)

The individual terms of the above equation are 0.5 z 1Y (z), 0.8X ( z) and 0.4 z 1 X (z) .
They are represented by the basic elements as shown in Figure 4.4.

Figure 4.4 Block diagram representation of (a) 0.8X(z) (b) 0.4z–1X(z) and (c) 0.5z–1Y(z).

The input to the system is X(z)[or x(n)] and the output of the system is Y(z)[or y(n)].
The above elements are connected as shown in Figure 4.5 to get the output Y(z)[or y(n)].

Figure 4.5 Realization of the system described by y(n) = 0.5y(n–1) + 0.8x(n) + 0.4x(n–1).
280 h Digital Signal Processing

Discrete-time LTI systems may be divided into two types: IIR systems (those that have
an infinite duration impulse response) and FIR systems (those that have a finite duration
impulse response).

4.3 STRUCTURES FOR REALIZATION OF IIR SYSTEMS

IIR systems are systems whose impulse response has infinite number of samples. They are
designed by using all the samples of the infinite duration impulse response. The convolution
formula for IIR systems is given by

y(n) =  h(k ) x(n  k )
k 0

Since this weighted sum involves the present and all the past input samples, we can say that
the IIR system has an infinite memory.
A system whose output y(n) at time n depends on the present input and any number of
past values of input and output is called a recursive system. The past outputs are
y(n – 1), y(n – 2), y(n – 3), ...
Hence, for recursive system, the output y(n) is given by
y( n) = F[y(n  1), y(n  2), ..., y(n  N ), x ( n), x (n  1), ..., x ( n  M )]

In recursive system, in order to compute y(n0), we need to compute all the previous values
y(0), y(1), y(2), ..., y(n0 – 1) before calculating y(n0). Hence, output of recursive system has
to be computed in order [y(0), y(1), y(2), ... ].

Transfer function of IIR system

In general, an IIR system is described by the difference equation
N M
y( n) =   ak y(n  k ) +  bk x (n  k )
k 1 k 0

i.e. in general, IIR systems are those in which the output at any instant of time depends not
only on the present and past inputs but also on the past outputs. Hence, in general, an IIR
system is of recursive type.
On taking Z-transform of the above equation for y(n), we get
N M
Y (z) =   ak z  k Y (z ) +  bk z  k X (z)
k 1 k 0

N M
i.e. Y (z ) +  ak z  k Y (z) =  bk z  k X (z )
k 1 k 0
 System Realization h 281

The system function or the transfer function of the IIR system is:
M

 bk z  k
Y ( z) k 0 b0 + b1 z 1 + b2 z 2 + ... + bM z  M
= H ( z) = =
X (z ) N
1 + a1 z 1 + a2 z 2 + ... + aN z  N
1+  ak z  k
k 1

The above equations for Y(z) and H(z) can be viewed as a computational procedure (or
algorithm) to determine the output sequence y(n) from the input sequence x(n). The
computations in the above equation can be arranged into various equivalent sets of difference
equations with each set of equations defining a computational procedure or algorithm for
implementing the system.
For each set of equations, we can construct a block diagram consisting of delays,
adders and multipliers. Such block diagrams are referred to as realization of the system or
equivalently as structure for realizing the system.
The main advantage of re-arranging the sets of difference equations (i.e. the main
criteria for selecting a particular structure) is to reduce the computational complexity,
memory requirements and finite word length effects in computations.
So the factors that influence the choice of structure for realization of LTI system are:
computational complexity, memory requirements and finite word length effects in
computations.
Computational complexity refers to the number of arithmetic operations required to
compute the output value y(n) for the system.
Memory requirements refer to the number of memory locations required to store the
system parameters, past inputs and outputs and any intermediate computed values.
Finite-word-length effects or finite precision effects refer to the quantization effects that
are inherent in any digital implementation of the system either in hardware or in software.
Although the above three factors play a major role in influencing our choice of the
realization of the system, other factors such as whether the structure lends itself to parallel
processing or whether the computations can be pipelined may play a role in selecting a
specific structure.
The different types of structures for realizing IIR systems are:
1. Direct form-I structure 2. Direct form-II structure
3. Transposed form structure 4. Cascade form structure
5. Parallel form structure 6. Lattice structure
7. Ladder structure

4.3.1 Direct Form-I Structure

Direct form-I realization of an IIR system is nothing, but the direct implementation of the
difference equation or transfer function. It is the simplest and most straight forward
realization structure available.
282 h Digital Signal Processing

The difference equation governing the behaviour of an IIR system is

N M
y( n) =   ak y(n  k ) +  bk x (n  k )
k =1 k =0

i.e. y(n) =  a1 y(n  1)  a2 y(n  2)  ...  aN y(n  N ) + b0 x (n) + b1 x (n  1) + ... + bM x ( n  M )

On taking the Z-transform of the above equation for y(n), we get

Y ( z) =  a1 z 1Y ( z )  a2 z 2Y (z)  ...  aN z  N Y (z) + b0 X (z ) + b1 z 1 X (z) + ... + bM z  M X ( z)

The equation for Y(z) [or y(n)] can be directly represented by a block diagram as
shown in Figure 4.6 and this structure is called Direct form-I structure. This structure uses
separate delays (z–1) for input and output samples. Hence, for realizing this structure more
memory is required. The direct form structure provides a direct relation between time domain
and z-domain equations.
The structure shown in Figure 4.6 is called a non-canonical structure because the
number of delay elements used is more than the order of the difference equation.
If the IIR system is more complex, that is of higher order, then introduce an
intermediate variable W(z) so that
M
W (z ) =  bk z k X (z) = b0 X (z) + b1 z 1 X (z) + ... + bM z  M X (z)
k=0

M
or w( n) =  bk x (n  k ) = b0 x (n) + b1 x (n  1) + . . . + bM x (n  m)
k 0

\ Y ( z ) =  a1 z 1Y ( z)  a2 z 2Y ( z )  ... + W ( z)

or y( n) =  a1 y(n  1)  a2 y( n  2)  ... + w(n )

So, the direct form-I structure is in two parts. The first part contains only zeros [that is, the
input components either x(n) or X(z)] and the second part contains only poles [that is, the
output components either y(n) or Y(z)]. In direct form-I, the zeros are realized first and poles
are realized second.

Limitations of direct form-I

· Since the number of delay elements used in direct form-I is more than (double) the
order of the difference equation, it is not effective.
· It lacks hardware flexibility.
· There are chances of instability due to the quantization noise.
 System Realization h 283

Figure 4.6 Direct form-I structure.

4.3.2 Direct Form-II Structure

The Direct form-II structure is an alternative to direct form-I structure. It is more
advantageous to use direct form-II technique than direct form-I, because it uses less number
of delay elements than the direct form-I structure.
In direct form-II, an intermediate variable is introduced and the given transfer function
is split into two, one containing only poles and the other containing only zeros. The poles
[that is, the output components y(n) or Y(z) which is the denominator part of the transfer
function] are realized first and the zeros [that is, the input components either x(n) or X(z),
which is the numerator part of the transfer function] second.
If the coefficient of the present output sample y(n) or the non-delay constant at
denominator is non unity, then transform it to unity. The systematic procedure is given as
follows:
284 h Digital Signal Processing

Consider the general difference equation governing an IIR system

N M
y ( n) =   ak y( n  k ) +  bk x ( n  k )
k 1 k 0

i.e. y(n) =  a1 y(n  1)  a2 y(n  2)  a3 y(n  3)  ...  aN y(n  N )

+ b0 x (n) + b1 x (n  1) + b2 x ( n  2) + ... + bM x (n  M )

On taking Z-transform of the above equation and neglecting initial conditions, we get
Y ( z) =  a1 z 1Y ( z)  a2 z 2Y ( z )  ...  aN z  N Y ( z ) + b0 X ( z ) + b1 z 1 X ( z) + ... + bM z  M X ( z )

i.e. Y ( z) + a1 z 1Y ( z) + a2 z 2Y ( z) + ... + aN z  N Y ( z ) = b0 X ( z) + b1 z 1 X ( z ) + ... + bM z  M X ( z)

i.e. Y ( z )[1 + a1 z 1 + a2 z 2 + ... + aN z  N ] = X ( z )[b0 + b1 z 1 + b2 z 2 + ... + bM z  M ]

Y ( z) b + b1 z 1 + b2 z 2 + ... + bM z  M
i.e. = 0
X (z ) 1 + a1 z 1 + a2 z 2 + ... + aN z  N

Y (z ) Y (z ) W (z )
Let =
X ( z ) W ( z) X ( z)

W ( z) 1
where =
X ( z) 1 + a1 z 1 + a2 z 2 + ... + a N z  N

Y ( z)
and = b0 + b1 z 1 + b2 z 2 + ... + bM z  M
W ( z)
On cross multiplying the above equations, we get

W ( z ) + a1 z 1W ( z) + a2 z 2W ( z ) + ... + aN z  N W ( z ) = X ( z)

\ W ( z ) = X ( z )  a1 z 1W ( z)  a2 z 2W ( z)  ...  aN z  N W ( z )
and Y ( z) = b0W ( z) + b1 z 1W ( z) + b2 z 2W ( z) + ... + bM z  M W ( z)
The realization of an IIR system represented by these equations in direct form-II is shown in
Figure 4.7.

Advantage of the direct form-II over the direct form-I

The number of delay elements used in direct form-II is less than that of direct form-I.

Limitations of direct form-II

· It also lacks hardware flexibility
· There are chances of instability due to the quantization noise
 System Realization h 285

Figure 4.7 Direct form-II structure of IIR system for M = N.

Since the number of delay elements used in direct form-II is the same as that of the
order of the difference equation, direct form-II is called a canonical structure.
The comparison of direct form-I and direct form-II structures is given in Table 4.1

TABLE 4.1 Comparison of direct form-I and direct form-II structures

Direct form-I structure Direct form-II structure

This realization uses separate delays (memory) This realization uses a single delay (memory)
for both the input and output signal samples. for both the input and output signal samples.
For the (M – 1)th or (N – 1)th order IIR system, For the (M – 1)th or (N – 1)th order IIR system,
direct form-I requires M + N – 1 multipliers, direct form-II requires M + N – 1 multipliers,
M + N – 2 adders and M + N – 2 delays. M + N – 2 adders and max [(M – 1), (N – 1)]
delays.
It is also called non-canonical, because it It is called canonical, because it requires a
requires more number of delays. minimum number of delays.
It is not efficient in terms of memory require- It is more efficient in terms of memory require-
ments compared to direct form-II. ments.
Direct form-I can be viewed as two linear Direct form-II can also be viewed as two linear
time-invariant systems in cascade. The first time-invariant systems in cascade. The first
one is non-recursive and the second one one is recursive and the second one non-
recursive. recursive.
286 h Digital Signal Processing

Conversion of direct form-I structure to direct form-II structure

The direct form-I structure can be converted to direct form-II structure by considering the
direct form-I structure as cascade of two systems H1(z) and H2(z) as shown in Figure 4.8(a).
By linearity property, the order of cascading can be interchanged as shown in Figure 4.8(b).

Figure 4.8 (a) Direct form-I structure as cascade of two systems (b) Direct form-I structure after
interchanging the order of cascading.

In Figure 4.8(b), we can observe that the inputs to the delay elements in H1(z) and
H2(z) are the same and so the outputs of the delay elements in H1(z) and H2(z) are same.
Therefore, instead of having separate delays for H1(z) and H2(z), a single set of delays can be
used. Hence, the delays can be merged to combine the cascaded systems to a single system.
The resultant structure will be direct form-II structure as that of Figure 4.7. The process of
converting direct form-I structure to direct form-II structure is shown in Figure 4.9.

Figure 4.9 Conversion of direct form-I structure to direct form-II.

 System Realization h 287

4.3.3 Transposed Form Structure Realization of IIR System

It is practically true that if we reverse the direction of all the branch transmittances and
interchange the input and the output in the structure or signal flow graph, the system remains
unchanged. The transposed structure or transpose form or reverse structure is obtained by
reversing the direction of all branch transmittances and interchanging the input and output in
the direct form structure.
The transposed structure remains valid, provided:
1. The branch transmittances are untouched.
2. The direction of all the branches in the structure is reversed.
3. The roles of the input and output are reversed.
By these steps, the system function remains unchanged.

Procedure to realize transposed form structure of IIR system

1. First realize the given difference equation or transfer function by using the direct
form structure.
2. Reverse or transpose the direction of signal flow and interchange the input and
output nodes.
3. Replace the junction points by adders and adders by junction points.
4. Fold the structure, which is the transposed form realization of an IIR system.
In general, the transposed structure realization of an IIR system is advantageous only if
it is implemented over a direct form-II structure, because the number of components used or
the number of additions and multiplications get reduced when it is used over a direct form-II
structure. However, the higher the order of the systems, the better would be its advantages. It
has no advantage for a direct form-I structure.
The realization of a general Nth order IIR system in Direct form-I is shown in Figure 4.6.
Its transposed version is shown in Figure 4.10(a).
The realization of a general Nth order IIR system in Direct form-II is shown in
Figure 4.7. Its transposed version is shown in Figure 4.10(b).

EXAMPLE 4.2 Realize an FIR system

y( n) + 2y( n  1) + 3y( n  2) = 4x ( n) + 5x (n  1) + 6x ( n  2)
using the transposed form structure.
Solution: Taking Z-transform on both sides of the given difference equation and neglecting
initial conditions, we get

Y ( z) + 2z 1Y ( z) + 3z 2Y ( z) = 4X (z ) + 5z 1 X ( z) + 6z 2 X (z)

Therefore, the transfer function of the given IIR system is

Y ( z) 4 + 5z 1 + 6z 2
H (z ) = =
X ( z ) 1 + 2z 1 + 3z 2
288 h Digital Signal Processing

Figure 4.10 (a) General transposed structure realization of IIR system through direct form-I,
(b) General transposed structure realization of IIR system through direct form-II.

The direct form-II realization structure, the recovered realization structure and the
transposed form realization structure of this system are shown in Figure 4.11[(a), (b) and (c)
respectively].
a1 = 2, a2 = 3, b0 = 4, b1 = 5, b2 = 6

4.3.4 Cascade Form Realization

The cascade form structure is nothing, but a cascaded or series interconnection of the sub
transfer functions or sub system functions which are realized by using the direct form
structures (either direct form-I or direct form-II or a combination of both).
Hence, in cascade form realization, the given transfer function H(z) is expressed as a
product of a number of second order or first order sections as indicated below:

3 Hi (z )
k
Y (z )
H ( z) = =
X (z ) i 1
 System Realization h 289

Figure 4.11 (a) Direct form-II realization (b) Recovered realization structure and (c) Transposed
realization structure (Example 4.2).

C0 i + C1i z 1 + C2 i z 2
where H i (z) = [second order section]
d0 i + d1i z 1 + d2 i z 2

C0 i + C1i z 1
or H i ( z) = [first order section]
d0 i + d1i z 1

Each of these sections is realized separately and all of them are connected in cascade
(series). Therefore, the cascade form realization is also called a series structure in which one
sub transfer function is the input to the other transfer function and so on.
The cascade form realization is shown in Figure 4.12.

X(z) H1(z) H2(z) Hk(z) Y(z)

Figure 4.12 Cascade realization of IIR system.

290 h Digital Signal Processing

The difficulties in cascade structure are:

1. Decision of pairing poles and zeros.
2. Deciding the order of cascading the first and second order sections.
3. Scaling multipliers should be provided between individual sections to prevent the
filter variables from becoming too large or too small.

4.3.5 Parallel Form Realization

Parallel form structure is nothing, but the parallel connection of sub-transfer functions or
sub-system functions, which is decomposed by using the partial fraction method.
In parallel form realization, by partial fraction expansion, the transfer function H(z) is
expressed as a sum of first and second order sections.
k
Y (z )
H (z ) =
X (z )
=C +  H i ( z)
i 1

C0 i + C1i z 1
where, H i (z) = [second order section]
d0 i + d1i z 1 + d2 i z 2

C0 i
or H i ( z) = [first order section]
d0 i + d1i z 1
Each first order and second order section is realized either in direct form-I structure or in
direct form-II structure and the individual sections are connected in parallel to obtain the
over all system as shown in Figure 4.13. As the filter operation is performed in parallel, i.e.
the processing is performed simultaneously, the parallel form structure is used for high speed
filtering application.
The difficulty with this method is expressing the transfer function in partial fraction
form is not easy for higher order systems.

Figure 4.13 Parallel form realization of IIR system.

 System Realization h 291

4.3.6 Lattice Structure Realization of IIR Systems

The IIR system consists of both zeros and poles. Therefore, the poles and zeros will be
considered as separate sub-transfer functions in cascade, that is

M  
 bk z k  
 
k  
M
1
=   bk z  
k 0
H ( z) = N   N  = H z (z ) H p ( z)
1+  ak z k k 0  1 +
  ak z k 

k 1  k 1 

M
where H z (z) =  bk z  k = Zeros
k 0

1
H p (z) = N
= Poles
1+  ak z k

k 1

The above equation for Hp(z) can be rewritten as

1 Yp (z )
H p (z ) = N
=
X p (z)
1+  ak z  k
k =1

or X p (z) = Yp ( z ) +  ak z  k Yp ( z )
k 1

Taking the inverse Z-transform on both sides and rearranging it, we get
N
y p (n) = x p (n)   ak y p ( n  k )
k 1

The output of the sub-transfer function Hz(z) is the input to the sub-transfer function Hp(z).
So yz(n) = xp(n).
While realizing the IIR system using the lattice structure, the zeroes, i.e. Hz(z) should
be realized first and then the poles, i.e. Hp(z) which is realized in cascade.
For the realization of the poles, a lattice structure consists of two paths, xp(n) and x¢p(n)
through which the input xp(n) or yz(n) is processed. However, those two different paths are
opposite in direction to each other. A single-stage lattice structure for a pole is shown in
Figure 4.14.
292 h Digital Signal Processing

Figure 4.14 Single-stage lattice structures for a pole.

The output from the single-stage lattice structure is

y(n) = x p1 (n) + K1p y(n  1)
and the feedback response is
xp1 (n) = K1p y(n) + y(n  1)
Similarly, the two-stage lattice structure for a pole is shown in Figure 4.15.

Figure 4.15 Two-stage lattice structure for a pole.

The intermediate output of the two-stage lattice structure is

x p1 (n) = x p2 (n) + K2p x p1 (n  1)

Substituting the value of x p1 (n) in the expression for xp(n), we have

x p1 (n) = x p 2 (n) + K 2p [ K1p y( n  1) + y( n  2)]

= x p 2 (n) + K1p K2p y( n  1) + K 2p y( n  2)
Substituting this value of xp1(n) in the expression for y(n), we get
y( n) = x p 2 (n) + K1p K2p y( n  1) + K 2p y( n  2) + K1p y(n  1)
= x p 2 (n) + K1p [1 + K 2p ] y( n  1) + K 2p y(n  2)

Therefore, the overall lattice structure for a second order IIR system (both poles and zeros
are of second order) can be in general, realized as shown in Figure 4.16.
In a similar fashion, the (M – 1)th order or (N – 1)th order IIR system can be realized
by using the lattice structure.
 System Realization h 293

Figure 4.16 A general realization of second order IIR system using the all-zero-all-pole lattice structure.

Procedure to realize the lattice structure of IIR systems

1. Find the order of the difference equation and compare the coefficients with the
reflection coefficients K1p , K2p , K 3p , ... and K1z , K 2z , K 3z , ... of the same order lattice
structure output. This applies both for the poles and zeros, separately.
2. Assign the calculated values of the reflection coefficients K1p , K 2p , K 3p , ... and
K1z , K 2z , K 3z , ... and construct the structures.
3. Cascade the structures of both zeroes and poles.

EXAMPLE 4.3 Determine the lattice coefficients corresponding to an IIR filter described
2 1 1
by y( n)  y( n  1) + y(n  2) = x (n) + x ( n  1) and realize it.
5 5 4
Solution: The given system described by the difference equation

2 1 1
y( n) = y( n  1)  y( n  2) + x (n) + x ( n  1)
5 5 4

has first order zeroes and second order poles. Hence, the proposed lattice structure is given
in Figure 4.17.

Figure 4.17 A proposed lattice structure (Example 4.3).

The output y(n) from Figure 4.17 is

y(n) = y1 (n) + K1p [1 + K 2p ] y(n  1) + K 2p y(n  2)

i.e. y(n) = x (n) + K1z x ( n  1) + K1p [1 + K 2p ] y( n  1) + K 2p y (n  2)

294 h Digital Signal Processing

Comparing this equation with the given difference equation, we have

1 1
K1z = , K 2p = 
4 5
p 2
and K1 [1 + K 2p ] =
5
2/5 1
or K1p = =
1  (1/5) 2

Hence, the lattice structure realization of the given IIR filter is shown in Figure 4.18.
The realization obtained in Figure 4.18 is called all-zero-all-pole lattice structure
realization.

Figure 4.18 All-zero-all-pole lattice structure (Example 4.3).

4.3.7 Ladder Structure Realization of IIR Systems

Ladder structure realization is possible only for an IIR system. In this structure, the
numerator polynomial will be divided by the denominator polynomial sequentially and the
result substituted at the ladder fashion shown below. Two cases are considered for the
realization of the ladder structure.
CASE-I: If the negative order of the numerator polynomial is more than the negative order
of the denominator polynomial, that is,

bN +1z ( N 1) + ... + b3 z 3 + b2 z 2 + b1z 1 + b0

H ( z) =
aN z  N + ... + a3 z 3 + a2 z 2 + a1 z 1 + a0
then, the transfer function of a ladder structure is shown below.
1
H ( z) = B1 z 1 +
1
C1 +
1
B2 z 1 + 1
C2 +
1
B3 z 1 +
C3 + ...
 System Realization h 295

The ladder structure realization of the above equation for H(z) is shown in Figure 4.19(a).

Figure 4.19 (a) Ladder structure realization of IIR system for case-I (b) Ladder structure realization of
IIR system for case-II.

CASE-II: If the order of both the numerator and denominator polynomials is equal, that is
bN z  N + ... + b3 z 3 + b2 z 2 + b1 z 1 + b0
H ( z) =
aN z  N + ... + a3 z 3 + a2 z 2 + a1 z 1 + a0
then, the transfer function for a ladder structure is shown below.

1
H ( z) = B 0 +
1
C1z 1 +
1
B1 + 1
C2 z 1 +
1
B2 +
C3 z 1
+ ...

The ladder structure realization of the above equation for H(z) is shown in Figure 4.19(b).

Procedure to realize the ladder structure of IIR systems

1. Express the numerator and denominator polynomials of the transfer function in
descending order of negative powers of z and assess whether it falls under case-I or
case-II.
296 h Digital Signal Processing

2. Divide the numerator polynomial by the denominator polynomial sequentially and

find the quotient at each division.
3. Compare the quotient obtained with that of the general transfer function for the
corresponding ladder structure and find the corresponding parameters a and b.
4. Realize the ladder structure with the obtained values of parameters.

EXAMPLE 4.4 Realize the IIR filter

3 z 2 + 5z + 4
H ( z) =
z 2 + 6z + 8
using ladder structure.

3 z 2 + 5z + 4 3 + 5 z 1 + 4 z 2 4 z 2 + 5 z 1 + 3
Solution: Given H ( z) = = =
z 2 + 6z + 8 1 + 6 z 1 + 8 z 2 8 z 2 + 6 z 1 + 1
Here, the negative order of the numerator and denominator is equal. So it falls under case-II.
Performing sequential division operation.

8z–2 + 6z–1 + 1) 4z–2 + 5z–1 + 3 (1/2

4z–1 + 3z–1 + 1/2
2z–1 + 5/2) 8z–2 + 6z–1 + 1 (4z–1
8z–2 + 10z–1
– 4z–1 + 1) 2z–1 + 5/2 (–1/2
2z–1 – 1/2
3) –4z–1 + 1 (– 4/3z–1
– 4z–1
1) 3 (3
3
0
Comparing with H(z) of case-II, we get

1 1 4
B0 = , C1 = 4, B1 =  , C2 =  and a2 =3
2 2 3
Hence, the required transfer function for realization is

1 1
H (z ) = +
2 1
4z 1 +
1 1
 +
2 4 1 1
 z +
3 3
Thus, the realization of the given IIR filter using the ladder form structure is shown in
Figure 4.20.
 System Realization h 297

EXAMPLE 4.5 Realize the IIR filter

5z 3 + 3 z 2 + 4 z + 2
H (z ) =
z [2 z 2 + 3z + 1]
using ladder structure
Solution: Given

5z 3 + 3z 2 + 4 z + 2 z 3 [5 + 3z 1 + 4z 2 + 2z 3 ] 2 z 3 + 4z 2 + 3z 1 + 5
H (z ) = = =
z [2z 2 + 3z + 1] z 3 [2 + 3z 1 + z 2 ] z 2 + 3z 1 + 2
Here the negative order of the numerator polynomial is greater than that of the denominator.
So it falls under case-I.
Performing the sequential division operation

z–2 + 3z–1 + 2) 2z–3 + 4z–2 + 3z–1 + 5 (2z–1

2z–3 + 6z–2 + 4z–1

–2z–2 – z–1 + 5) z–2 + 3z–1 + 2 (–1/2

z–2 + 1/2z–1 – 5/2
5/2z–1 + 9/2) –2z–2 – z–1 + 5 (– 4/5 z–1
–2z–2 – 18/5 z–1

13/5z–1 + 5) 5/2z–1 + 9/2 (25/26

5/2z–1 + 125/26

–8/13) 13/5z–1 + 5 (–169/40z–1

13/5z–1
5) –8/13 (–8/65
–8/13
0
Comparing with H(z) of the case-I, we get

1 4 25 169 8
B1 = 2, C1 =  , B 2 =  , C2 = , B3 =  , C3 = 
2 5 26 40 65
Hence, the required transfer function for realization is

1
H (z) = 2z 1 +
1 1
 +
2 4 1
 z 1 +
5 25 1
+
26 169 1 1
 z +
40 8

65
298 h Digital Signal Processing

Figure 4.20 Ladder structure realization (Example 4.4).

Thus, the realization of the given IIR filter using the ladder form structure is shown in Figure 4.21.

EXAMPLE 4.6 Obtain the direct form-I, direct form-II, cascade and parallel form realiza-
tions of the LTI system governed by the equation

13 9 1
y(n) =  y( n  1)  y( n  2)  y( n  3) + x ( n) + 4x (n  1) + 3x (n  2)
12 24 24

Solution:
Direct form-I
13 9 1
Given y(n) =  y( n  1)  y( n  2)  y( n  3) + x ( n) + 4x (n  1) + 3x (n  2)
12 24 24
Taking Z-transform on both sides, we get
13 1 9 2 1 3
Y (z ) =  z Y (z)  z Y (z )  z Y (z ) + X (z) + 4z 1 X (z ) + 3z 2 X (z )
12 24 24
The direct form-I structure can be obtained from the above equation as shown in Figure 4.22.
Direct form-II
Taking Z-transform of the given difference equation, we have

13 1 9 2 1 3
Y (z ) =  z Y (z)  z Y (z )  z Y (z ) + X ( z) + 4z 1 X (z ) + 3z 2 X (z )
12 24 24
 System Realization h 299

Figure 4.21 Ladder structure realization (Example 4.5).

13 1 9 2 1 3
i.e. Y ( z) + z Y (z ) + z Y ( z) + z Y (z) = X ( z ) + 4z 1 X ( z ) + 3z 2 X (z )
12 24 24

 13 1 9 2 1 3 
i.e. Y ( z) 1 + z + z + z  = X ( z ) [1 + 4z 1 + 3z 2 ]
 12 24 24 
Therefore, the transfer function of the system is

Y (z ) 1 + 4 z 1 + 3 z 2
H ( z) = =
X ( z) 1 + (13/12) z 1 + (9/24) z 2 + (1/24) z 3

Y (z ) Y (z ) W (z )
Let =
X ( z ) W ( z) X ( z)

W ( z) 1
where =
X (z ) 1 + (13/12) z + (9/24) z 2 + (1/24) z 3
1
300 h Digital Signal Processing

Figure 4.22 Direct form-I realization structure (Example 4.6).

Y ( z)
and = 1 + 4z 1 + 3z 2
W (z )
On cross multiplying the above equations, we get

 13 1 9 2 1 3 
W ( z ) 1 + z + z + z  = X (z )
 12 24 24 

13 1 9 2 1 3
i.e. W ( z ) = X ( z)  z W (z )  z W ( z)  z W ( z)
12 24 24

and Y ( z) = W (z ) + 4 z 1 W ( z) + 3z 2W ( z)
The above equations for W(z) and Y(z) can be realized by a direct form-II structure as shown
in Figure 4.23.

Cascade form
Y ( z) 1 + 4z 1 + 3z 2
The transfer function is H ( z ) = =
X ( z ) 1 + (13/12) z 1 + (9/24) z 2 + (1/24) z 3
Factorizing the numerator and denominator, we have

(1 + z 1 ) (1 + 3z 1 )
H ( z) =
[1 + (1/2) z 1 ] [1 + (1/3) z 1 ] [1 + (1/4) z 1 ]

Since there are three first order factors in the denominator of H(z), H(z) can be expressed as
product of three sections.
 System Realization h 301

Figure 4.23 Direct form-II realization structure (Example 4.6).

Let H (z ) = H1 (z ) H 2 (z) H 3 (z)

1 + z 1 1 + 3z 1 1
where H1 (z ) = , H ( z ) = and H3 ( z) =
1 + (1/2) z 1 1 + (1/3)z 1 1 + (1/4)z 1
2

The transfer function H1(z) can be realized in direct form-II structure as shown in Figure 4.24(a).

Y1 ( z ) Y ( z) W1 (z) 1 + z 1
Let H1 (z ) = = 1 . =
X ( z) W1 ( z) X ( z ) 1 + (1/2)z 1
W1 ( z) 1 Y (z )
where = 1
and 1 = 1 + z 1
X ( z) 1 + (1/2) z W 1 ( z )

1 1
\ W1 ( z) = X ( z )  z W1 (z)
2

and Y1 ( z) = W1 (z) + z 1W1 (z)

The transfer function H2(z) can be realized in direct form-II structure as shown in Figure 4.24(b).

Y2 (z ) Y (z ) . W2 (z ) 1 + 3z 1
Let H 2 (z ) = = 2 =
Y1 (z) W2 (z) Y1 (z ) 1 + (1/3)z 1

W2 (z) 1 Y (z )
where = 1
and 2 = 1 + 3 z 1
Y1 ( z ) 1 + (1/3) z W2 ( z)
302 h Digital Signal Processing

Figure 4.24 Direct form-II structure of (a) H1(z), (b) H2(z) and (c) H3(z) (Example 4.6).

1 1
\ W2 (z) = Y1 (z)  z W2 (z)
3
and Y2 ( z ) = W2 (z) + 3z 1W2 (z )

The transfer function H3(z) can be realized in direct form-II structure as shown in Figure 4.24(c).
Y ( z) 1
Let H 3 ( z) = =
Y2 ( z) 1 + (1/4) z 1

1 1
\ Y (z) = Y2 (z)  z Y (z)
4
The cascade structure of the given system is obtained by connecting the individual sections
shown in Figures 4.24(a), (b) and (c) in cascade as shown in Figure 4.25.

Figure 4.25 Cascade realization of the system (Example 4.6).

Parallel form
(1 + z 1 ) (1 + 3z 1 )
Consider the equation H ( z) =
[1 + (1/2) z 1 ] [1 + (1/3) z 1 ] [1 + (1/4) z 1 ]
By partial fraction expansion, we have
A B C
H (z ) = + +
1 + (1/2) z 1 1 + (1/3) z 1 1 + (1/4) z 1
 System Realization h 303

where, the coefficients A, B and C are

(1 + z 1 )(1 + 3z 1 ) (1  2) (1  6) 5
A= z 1 =  2
= = = 30
 1 1   1 1   2 2 1.1
1 + z  1 + z  1   1  
 3  4   3 4 3 2

(1 + z 1 ) (1 + 3z 1 ) (1  3) (1  9) 16
B= z 1 =  3
= = =  128
 1 1   1 1   3 3

1.1
1 + z  1 + z  1   1  
 2  4   2 4 2 4

(1 + z 1 ) (1 + 3z 1 ) (1  4) (1  12) 33
C= z 1   4
= = = 99
 1 1   1 1   4 1
1 + z   1 + z   
1  2 1    1 . 
 2  3   3 3

30 128 99
\ H ( z) =  +
1 1
1 + (1/2) z 1 + (1/3) z 1 + (1/4) z 1

Y ( z)
Let H ( z) =
X (z )

Y ( z) 30 128 99
\ =  +
1 1
X ( z ) 1 + (1/2) z 1 + (1/3) z 1 + (1/3) z 1

30 128 99
\ Y ( z) = 1
X ( z)  1
X (z ) + X (z )
1 + (1/2) z 1 + (1/3) z 1 + (1/4) z 1
Let Y ( z) = Y1 ( z ) + Y2 ( z ) + Y3 (z)
30 128 99
where Y1 (z ) = X ( z ); Y2 (z ) =  X ( z ) and Y3 ( z) = X (z )
1 + (1/2) z 1 1 + (1/3) z 1 1 + (1/4) z 1
Y1 ( z) 30 Y (z) 128
Let H1 ( z ) = = ; H 2 (z ) = 2 = 
X ( z ) 1 + (1/2) z 1 X (z ) 1 + (1/2) z 1

Y3 (z ) 99
and H3 (z ) = =
X ( z ) 1 + (1/4) z 1

The transfer function H1(z) can be realized in direct form-I structure as shown in Figure 4.26(a).
Y1 ( z) 30
Let H1 ( z ) = =
X ( z ) 1 + (1/2) z 1

On cross multiplying and rearranging, we get

1 1
Y1 ( z ) = 30X ( z )  z Y (z )
2
304 h Digital Signal Processing

Figure 4.26 Direct form-I structure of (a) H1(z), (b) H2(z) and (c) H3(z) (Example 4.6).

The transfer function H2(z) can be realized in direct form-I structure as shown in Figure 4.26(b).

Y2 ( z ) 128
Let H2 (z ) = = 
X ( z) 1 + (1/3) z 1

On cross multiplying and rearranging, we get

1 1
Y2 ( z ) =  128 X ( z )  z Y2 (z )
3
The transfer function H3(z) can be realized in direct form-I structure as shown in Figure 4.26(c).

Y3 ( z ) 99
Let H 3 (z) = =
X ( z) 1 + (1/4) z  1
On cross multiplying and rearranging, we get

1 1
Y3 ( z) = 99 X ( z)  z Y3 (z)
4
The overall structure is obtained by connecting the individual sections in Figures 4.26(a), (b)
and (c) in parallel as shown in Figure 4.27.

EXAMPLE 4.7 Find the direct form-I and direct form-II realizations of a discrete-time
system represented by the transfer function

3z 3  5z 2 + 9z  3
H ( z) =
[z  (1/2)][z 2  z + (1/3)]

Y ( z)
Solution: Let H ( z) =
X (z )
where Y(z) = Output and X(z) = input
 System Realization h 305

Figure 4.27 Parallel form realization (Example 4.6).

Y (z ) 3z 3  5z 2 + 9z  3 3 z 3  5z 2 + 9z  3
\ H ( z) = = =
X ( z) [z  (1/2)] [z 2  z + (1/3)] z 3  (3/2)z 2 + (5/6) z  (1/6)

z 3 [3  5z 1 + 9z 2  3z 3 ] 3  5z 1 + 9z 2  3z 3
= =
z 3 [1  (3/2) z 1 + (5/6) z 2  (1/6) z 3 ] 1  (3/2) z 1 + (5/6) z 2  (1/6) z 3

Direct form-I
On cross multiplying the above equation for Y(z)/X(z), we have
3 1 5 1
Y (z )  z Y ( z) + z 2Y ( z)  z 3Y ( z ) = 3X ( z)  5z 1 X ( z) + 9z 2 X ( z)  3z 3 X ( z)
2 6 6

3 1 5 1
\ Y ( z) = z Y (z)  z 2Y (z ) + z 3Y ( z) + 3X ( z )  5z 1 X ( z) + 9z 2 X ( z)  3 z 3 X ( z)
2 6 6
306 h Digital Signal Processing

The direct form-I structure of the above equation for Y(z) can be obtained as shown in Figure 4.28.

Figure 4.28 Direct form-I structure (Example 4.7).

Direct form-II
Y (z ) 3  5z 1 + 9z 2  3z 3
The equation for H(z) is H (z) = =
X ( z) 1  (3/2) z 1 + (5/6) z 2  (1/6) z 3

Let Y (z) Y ( z) . W ( z ) 3  5z 1 + 9z 2  3z 3
H (z ) = = =
X ( z) W ( z) X ( z ) 1  (3/2) z 1 + (5/6) z 2  (1/6) z 3

W (z ) 1
where =
X (z ) 1  (3/2) z + (5/6) z 2  (1/6) z 3
1

Y (z )
and = 3  5z 1 + 9z 2  3 z 3
W (z )

On cross multiplying the above equations, we get

3 1 5 1
W (z )  z W ( z ) + z 2W ( z )  z 3W (z ) = X ( z )
2 6 6

3 1 5 1
i.e. W ( z ) = X (z ) + z W ( z )  z 2W (z ) + z 3W ( z)
2 6 6
 System Realization h 307

and Y ( z) = 3W ( z )  5z 1W ( z) + 9z 2W ( z)  3z 3W (z )

The above equations for Y(z) and W(z) can be realized by a direct form-II structure as shown
in Figure 4.29.

Figure 4.29 Direct form-II structure (Example 4.7).

EXAMPLE 4.8 Find the digital network in direct and transposed form for the system
described by the difference equation
y(n) = 2x(n) + 0.3 x( n  1) + 0.5 x (n  2)  0.7 y( n  1)  0.9 y(n  2)

Solution: Given difference equation is:

y(n) = 2x(n) + 0.3 x( n  1) + 0.5 x (n  2)  0.7 y( n  1)  0.9 y(n  2)
Taking Z-transform on both sides, we have

Y ( z) = 2X ( z ) + 0.3 z 1 X ( z) + 0.5 z 2 X ( z)  0.7 z 1Y ( z)  0.9 z 2Y ( z )

Direct form
The direct form-I digital network can be realized using the above equation for Y(z) as shown
in Figure 4.30.

Transposed form
On rearranging the equation for Y(z), we get

Y ( z) + 0.7 z 1Y (z ) + 0.9z 2Y (z ) = 2X ( z ) + 0.3z 1 X ( z) + 0.5z 2 X ( z )

308 h Digital Signal Processing

Figure 4.30 Direct form structure (Example 4.8).

i.e. Y ( z)[1 + 0.7z 1 + 0.9z 2 ] = X ( z )[2 + 0.3z 1 + 0.5z 2 ]

\ Y ( z) 2 + 0.3 z 1 + 0.5 z 2
H (z ) = =
X (z ) 1 + 0.7z 1 + 0.9 z 2

Y (z ) Y (z ) . W ( z ) 2 + 0.3z 1 + 0.5z 2
Let = =
X ( z) W (z ) X ( z ) 1 + 0.7z 1 + 0.9z 2

W ( z) 1 Y ( z)
where = 1 2
and = 2 + 0.3 z 1 + 0.5 z 2
X ( z) 1 + 0.7 z + 0.9z W ( z )

Cross multiplying the above equations, we get

W ( z ) + 0.7z 1W ( z ) + 0.9 z 2W ( z ) = X ( z)

i.e. W ( z ) = X ( z)  0.7 z 1W (z )  0.9 z 2W (z)

and Y ( z) = 2W ( z ) + 0.3z 1W (z) + 0.5 z 2W ( z )

The transposed form of digital network is realized using equations for W(z) and Y(z) as
shown in Figure 4.31. The recovered realization structure and the transposed forms structure
obtained from the direct form-II are shown in Figure 4.31[(b) and (c)].
 System Realization h 309

Figure 4.31 (a) Direct form-II structure (b) Recovered realization structure (c) Transposed form structure
(Example 4.8).

EXAMPLE 4.9 Realize the given system in cascade and parallel forms

1 + (1/3) z 1
H (z ) =
[1  (1/2) z 1 + (1/3) z 2 ] [1  (1/3) z 1 + (1/2) z 2 ]

Solution:
Cascade form
Let us realize the system as cascade of two second order sections.

1 + (1/3) z 1 1
H ( z) = = H1 (z ) H2 ( z )
1  (1/2) z  (1/ 3) z 1  (1/3) z + (1/2) z 2
1 2 1
310 h Digital Signal Processing

1 + (1/3) z 1
where H1 ( z ) =
1  (1/2) z 1  (1/ 3) z 2

1
and H2 (z ) = 1
1  (1/3) z + (1/2) z 2

Y1 ( z ) Y ( z) W1 ( z) 1 + (1/3) z 1
Let H1 (z ) = = 1 =
X1 ( z ) W1 ( z) X1 ( z ) 1  (1/2) z 1 + (1/3) z 2

W1 ( z ) 1
where =
X1 ( z) 1  (1/2) z + (1/3) z 2
1

Y1 ( z ) 1
and = 1 + z 1
W1 ( z ) 3
On cross multiplying the above equations, we get

1 1 1
W1 ( z)  z W1 ( z ) + z 2W1 ( z ) = X1 ( z)
2 3

1 1 1
i.e. W1 ( z) = X1 ( z ) + z W1 ( z )  z 2W1 ( z)
2 3
1 1
and Y1 ( z) = W1 ( z) + z W1 ( z)
3
Using the above equations for W1(z) and Y1(z), the system H1(z) can be realized in direct
form-II structure as shown in Figure 4.32(a).

Figure 4.32 Direct form-II structure of (a) H1(z) and (b) H2(z) (Example 4.9).
 System Realization h 311

Y2 ( z ) 1
Let H 2 (z ) = =
Y1 ( z ) 1 1
1  z 1 + z 2
3 2
On cross multiplying, we get
1 1 1
Y2 ( z )  z Y2 ( z ) + z 2Y2 ( z ) = Y1 ( z)
3 2

1 1 1
i.e. Y2 ( z) = Y1 ( z ) + z Y2 (z )  z 2Y2 ( z)
3 2
The above equation for Y2(z) can be realized in direct form-II structure as shown in Figure 4.32(b).
The cascade structure of H(z) is obtained by connecting the structures of H1(z) and
H2(z) in cascade as shown in Figure 4.33.

Figure 4.33 Cascade structure of H(z) (Example 4.9).

Parallel realization
1 + (1/3) z 1
Given H (z ) =
[1  (1/2) z 1 + (1/3) z 2 ] [1  (1/3) z 1 + (1/2) z 2 ]

z 1 [z + (1/3)]
=
z 4 [z 2  (1/2) z + (1/3)][ z 2  (1/3) z + (1/2)]

z 3 [ z + (1/3)]
=
[ z 2  (1/2) z + (1/3)] [ z 2  (1/3) z + (1/2)]

z [z 3 + (1/3) z 2 ]
=
[z 2  (1/2) z + (1/3)][ z 2  (1/3) z + (1/2)]
312 h Digital Signal Processing

By partial fraction expansion, we can write

H (z) [z 3 + (1/3) z 2 ] Az + B Cz +D
= 2 = 2 + 2
z [z  (1/2) z + (1/3)] [z  (1/3) z + (1/2)] z  (1/2) z + (1/3)
2
z  (1/3) z + (1/2)
On cross multiplying the above equation, we get

z 3 + (1/3) z 2 = (Az + B)[z 2  (1/3) z + (1/2)] + (Cz + D ) [z 2  (1/2) z + (1/3)]

1 2 1 1 1 1 1 1 1
= A z3  A z + A z + B z 2  B z + B + Cz 3  Cz 2 + Cz + Dz 2  Dz + D
3 2 3 2 2 3 2 3
1 1 1 1 1 1  1 1 
= (A + C ) z 3 + (B + D  A  C) z2 +  A  B + C  D  z +  B + D 
3 2  2 3 3 2   2 3 
On equating the coefficients on both sides and solving for A, B, C and D, we get
13 14 2 21
A= , B=  , C =  and D =
11 11 11 11

\ H (z ) (13/11) z  (14/11)  (2/11) z + (21/11)

= 2 + 2
z z  (1/2) z + (1/3) z  (1/3) z + (1/2)

(13/11) z 2  (14/11)z  (2/11) z 2 + (21/11)z

\ H ( z) = +
z 2  (1/2) z + (1/3) z 2  (1/3) z + (1/2)

z 2 [(13/11)  (14/11)z 1 ] z 2 [  (2/11) + (21/11) z 1 ]

= +
z 2 [1  (1/2) z 1 + (1/3)z 2 ] z 2 [1  (1/3) z 1 + (1/2) z 2 ]

Y ( z) (13/11)  (14/11)z 1  (2/11) + (21/11) z 1

So H (z ) = = +
X ( z ) 1  (1/2) z 1 + (1/3)z 2 1  (1/3) z 1 + (1/2) z 2

(13/11)  (14/11) z 1  (2/11) + (21/11) z 1

\ Y (z ) = X ( z ) + X ( z)
1  (1/2) z 1 + (1/3) z 2 1  (1/3) z 1 + (1/2) z 2

Let Y(z) = Y1(z) + Y2(z)

(13/11)  (14/11) z 1
where Y1 ( z) = X ( z)
1  (1/2) z 1 + (1/3) z 2

 (2/11) + (21/11) z 1
and Y2 (z ) = X ( z)
1  (1/3)z 1 + (1/2) z 2

Y1 ( z ) Y ( z) W1 ( z) (13/11)  (14/11) z 1
Let H1 (z ) = = 1 . =
X ( z) W1 ( z) X ( z ) 1  (1/2) z 1 + (1/3) z 2
 System Realization h 313

W1 ( z ) 1
where =
X ( z) 1  (1/2) z + (1/3) z 2
1

Y1 ( z ) 13 14 1
and =  z
W1 ( z ) 11 11

On cross multiplying the above equations, we get

1 1 1
W1 ( z)  z W1 ( z) + z 2W1 (z ) = X ( z )
2 3
1 1 1
i.e. W1 ( z) = X (z) + z W1 ( z )  z 2W1 ( z )
2 3
13 14 1
and Y1 ( z) = W1 ( z )  z W1 ( z)
11 11
The direct form-II structure of system H1(z) can be realized using the above equations for
W1(z) and Y1(z) as shown in Figure 4.34(a).

Figure 4.34 Direct form-II structure of (a) H1(z) and (b) H2(z) (Example 4.9).

Y2 (z ) Y (z ) W2 ( z)  (2/11) + (21/11) z 1
Let H2 (z ) = = 2 . =
X ( z) W2 ( z) X ( z) 1  (1/3) z 1  (1 / 2) z 2

W2 ( z ) 1
where =
X ( z) 1  (1/3) z + (1/2) z 2
1

Y2 ( z ) 2 21 1
and =  + z
W2 ( z) 11 11
314 h Digital Signal Processing

On cross multiplying the above equations, we get

1 1
W2 (z)  z 1W2 (z ) + z 2W2 (z ) = X ( z)
3 2

1 1 1
i.e. W2 ( z) = X ( z ) + z W2 ( z )  z 2W2 (z)
3 2

2 21 1
and Y2 (z ) =  W2 ( z ) + z W2 (z )
11 11

The direct form-II structure of the system H2(z) can be realized using the above equations for
W2(z) and Y2(z) as shown in Figure 4.34(b).
The parallel form structure of H(z) is obtained by connecting the direct form-II
structures of H1(z) and H2(z) in parallel as shown in Figure 4.35.

Figure 4.35 Parallel form realization of system H(z) (Example 4.9).

 System Realization h 315

EXAMPLE 4.10 An LTI system is described by the equation

y(n) + 2y(n  1)  y(n  2) = x (n)
Determine the cascade and parallel realization structures of the system

Solution: Given y(n) + 2y(n  1)  y(n  2) = x (n)

Taking Z-transform on both sides, we have
Y ( z) + 2 z 1Y (z)  z 2Y (z) = X (z)

i.e. Y ( z) [1 + 2z 1  z 2 ] = X (z )

Therefore, the transfer function of the system is

Y ( z) 1 1
H (z ) = = 1 2
=
X (z ) 1 + 2 z  z (1  0.414 z ) (1 + 2.414 z 1 )
1

Let H ( z) = H1 (z) H 2 ( z )
1 1
where H1 (z ) = 1
and H 2 (z ) =
1  0.414 z 1 + 2.414 z 1

Cascade realization
Y1 ( z ) 1
Let H1 ( z ) = =
X ( z) 1  0.414 z 1
On cross multiplying, we get
Y 1 ( z)  0.414 z 1Y1 ( z ) = X ( z )

i.e. Y 1 ( z ) = X ( z) + 0.414z 1Y1 ( z )

The direct form-I structure of H1(z) is obtained using the above equation for Y1(z) as shown
in Figure 4.36(a).

(a) (b)
Figure 4.36 Direct form-I structure of (a) H1(z) and (b) H2(z) (Example 4.10).
316 h Digital Signal Processing

Y ( z) 1
Let H2 (z ) = =
Y1 ( z ) 1 + 2.414 z 1
On cross multiplying, we get
Y ( z ) + 2.414z 1Y (z ) = Y1 ( z )

i.e. Y ( z ) = Y1 ( z )  2.414z 1Y (z )

The direct form-I structure of H2(z) is obtained using the above equation for Y(z) as shown
in Figure 4.36(b).
The cascade structure is obtained by connecting the direct form structures of H1(z) and
H2(z) as shown in Figure 4.37.

Figure 4.37 Cascade structure of H(z) (Example 4.10).

Parallel realization
Y (z ) 1 A B
H ( z) = = = +
X ( z) (1  0.414 z 1 ) (1 + 2.414 z 1 ) 1  0.414 z 1 1 + 2.414 z 1

1
where A = (1  0.414 z 1 ) H ( z ) 1 = = 0.146
z 1 = 1
0.414 1 + 2.414 
0.414

1
B = (1 + 2.414 z 1 ) H ( z ) 1 1 = = 0.853
z = 1
2.414 1 + 0.414 
2.414
Y ( z) 0.146 0.853
\ = +
X ( z ) 1  0.414 z 1 1 + 2.414 z 1

\ 0.146 0.853
Y (z ) = X (z ) + X (z )
1  0.414 z 1 1 + 2.414 z 1
 System Realization h 317

Let Y ( z) = Y1 ( z) + Y2 (z)

where 0.146 0.853

Y1 ( z) = 1
X (z) and Y2 ( z ) = X ( z)
1  0.414 z 1 + 2.414 z 1

Y1 ( z ) 0.146 Y2 ( z ) 0.853
\ = H1 ( z ) = and = H2 (z ) =
X ( z) 1  0.414 z 1 X ( z) 1 + 2.414 z 1

Realizing H1(z) and H2(z) in direct form-I

Y1 ( z ) 0.146
=
X ( z) 1  0.414 z 1

\ Y1 ( z )  0.414 z 1Y1 (z ) = 0.146 X ( z)

i.e. Y1 ( z ) = 0.146 X ( z ) + 0.414 z 1Y1 ( z )
Realization of H1(z) in direct form-I structure is shown in Figure 4.38(a).

(a) (b)
Figure 4.38 Direct form-I structure of (a) H1(z) and (b) H2(z) (Example 4.10).

Y2 ( z ) 0.853
=
X ( z) 1 + 2.414 z 1

\ Y2 ( z ) + 2.414 z 1Y2 ( z ) = 0.853X ( z )

i.e. Y2 ( z ) = 0.853X ( z )  2.414 z 1Y2 ( z )

Realization of H2(z) in direct form-I structure is shown in Figure 4.38(b).
The parallel structure is obtained by connecting the direct form-I structures of H1(z)
and H2(z) in parallel as shown in Figure 4.39.

EXAMPLE 4.11 Obtain the cascade realization of the system

3 + 2 z 1 + z 2
H ( z) =
[1 + (1/3) z 1 ] [1  (1/3) z 1 ] [1 + (1/3) z 1 ]
318 h Digital Signal Processing

Figure 4.39 Parallel form realization of H(z) (Example 4.10).

3 + 2 z 1 + z 2
Solution: Given H ( z) =
[1 + (1/3) z 1 ] [1  (1/3) z 1 ] [1 + (1/3) z 1 ]
The roots of the numerator polynomial

2  4  12
z 1 =
6
are complex conjugate. Hence, H(z) can be realized as cascade of one first order and one second
order system.

\ 1 3 + 2 z 1 + z 2
H ( z) =
1  (1/3) z 1 [1 + (1/3) z 1 ] [(1 + (1/3) z 1 ]

1 3 + 2 z 1 + z 2
=
1  (1/3) z 1 + (2/3) z 1 + (1/9) z 2
1

Let H(z) = H1(z) . H2(z)

1 3 + 2 z 1 + z 2
where H1 ( z ) = and H2 (z ) =
1  (1/3) z 1 1 + (2/3) z 1 + (1/9) z 2

Y1 ( z) 1
Let H1 ( z ) = =
X ( z) 1  (1/3) z 1

On cross multiplying, we get Y1 ( z )  (1/3) z 1Y1 (z ) = X ( z )

 System Realization h 319

i.e. Y1 ( z ) = X (z ) + (1/3) z 1Y1 ( z )

The direct form-II structure of H1(z) can be obtained from the equation for Y1(z) as shown in
Figure 4.40(a).

Figure 4.40 Direct form structure of (a) H1(z) and (b) H2(z) (Example 4.11).

Y (z ) 3 + 2 z 1 + z 2
Let H2 (z ) =
Y1 ( z) 1 + (2/3) z 1 + (1/9) z 2

and Y (z ) Y ( z) W1 ( z)
=
Y1 ( z) W1 ( z) Y1 ( z )

W1 ( z) 1 Y (z )
where = and = 3 + 2 z 1 + z 2
Y1 ( z) 1 + (2/3) z + (1/9) z 2
1 W1 ( z )

On cross multiplying the above equations, we get

2 1 1
W1 ( z) + z W1 ( z) + z 2W1 (z ) = Y1 ( z)
3 9

2 1 1
i.e. W1 ( z) = Y1 ( z)  z W1 ( z)  z 2W1 ( z )
3 9

and Y ( z) = 3W1 ( z) + 2 z 1W1 ( z)  z 2W1 ( z)

The direct form-II structure of H2(z) can be obtained by using the above equations for W1(z)
and Y(z) as shown in Figure 4.30(b).
320 h Digital Signal Processing

The cascade realization of H(z) is obtained by connecting the direct form-II structures
of H1(z) and H2(z) as shown in Figure 4.41.

Figure 4.41 Cascade realization of H(z) (Example 4.11).

EXAMPLE 4.12 The transfer function of a system is given by

(1 + z 1 )3
H ( z) =
[1  (1/2) z 1 ][1 + z 1 + (1/3) z 2 ]

Realize the system in cascade and parallel structures.

Solution:
Cascade realization

Given (1 + z 1 )3
H ( z) =
[1  (1/2) z 1 ] [1 + z 1 + (1/3) z 2 ]

The roots of the quadratic factor 1 + z–1 + (1/3)z–2 in the denominator

1  1  (4/3)
z 1 =
2
are complex conjugate. Hence, the system has to be realized as the cascading of one first
order section and one second order section.
(1 + z 1 ) (1 + z 1 ) 2
\ H (z ) =
[1  (1/2) z 1 ] [1 + z 1 + (1/3) z 2 ]

(1 + z 1 ) (1 + 2 z 1 + z 2 )
=
[1  (1/2) z 1 ] [1 + z 1 + (1/3) z 2 ]
 System Realization h 321

Let H ( z ) = H1 ( z ) H 2 ( z)

(1 + z 1 ) (1 + 2 z 1 + z 2 )
where H1 ( z ) = and H ( z ) =
[1  (1/2) z 1 ] [1 + z 1 + (1/3) z 2 ]
2

Y1 ( z) 1 + z 1
Let H1 ( z ) = =
X ( z) 1  (1/2) z 1

Let Y1 ( z ) Y ( z) W1 ( z ) 1 + z 1
= 1 =
X ( z) W1 ( z ) X ( z) 1  (1/2) z 1

W1 ( z) 1 Y ( z)
where = and 1 = 1 + z 1
X ( z) 1  (1/2) z 1 W1 ( z)

On cross multiplying the above equations, we get

1 1
W1 ( z)  z W1 (z ) = X ( z)
2
1 1
i.e. W1 ( z) = X ( z) + z W1 (z)
2

and Y1 ( z ) = W1 ( z ) + z 1W1 (z )

The direct form-II structure of H1(z) can be realized using the above equations for W1(z) and
Y1(z) as shown in Figure 4.42(a).

Figure 4.42 Direct form-II structure of (a) H1(z) and (b) H2(z) (Example 4.12).

Y ( z) 1 + 2 z 1 + z 2
Let H 2 ( z) = =
Y1 ( z) 1 + z 1 + (1/3) z 2
322 h Digital Signal Processing

Y (z ) Y ( z ) W2 ( z)
and =
Y1 ( z ) W2 ( z ) Y1 ( z )

W2 ( z) 1 Y ( z)
where = 1 2
and = 1 + 2 z 1 + z 2
Y1 ( z ) 1 + z + (1/3) z W2 ( z)

On cross multiplying the above equations, we get

1 2
W2 (z) + z 1W2 (z ) + z W2 ( z) = Y1 ( z )
3

1 2
i.e. W2 ( z) = Y1 ( z)  z 1W2 ( z )  z W2 ( z )
3
and Y ( z) = W2 ( z ) + 2z 1 W2 ( z) + z 2W2 (z)

The direct form-II structure of H2(z) can be realized using the above equations for W2(z) and
Y(z) as shown in Figure 4.42(b).
The cascade realization of H(z) is obtained by connecting the direct form-II structures
of H1(z) and H2(z) in cascade as shown in Figure 4.43.

Figure 4.43 Cascade realization of H(z) (Example 4.12).

Parallel realization
(1 + z 1 )3
Given H ( z) =
[1  (1/2) z 1 ] [1 + z 1 + (1/3) z 2 ]

z 3 (z + 1)3 z 3 + 3z 2 + 3z + 1
= =
 1  1 1 1 1
z 1  z   z 2  z 2 + z +  z3 + z 2  z 
 2  3 2 6 6
 System Realization h 323

1 2 1 1
z3 + z  z  ) z3 + 2 z2 + 3 z + 1 ( 1
2 6 6
1 1 1
z3 + z2  z 
2 6 6
5 2 19 7
z + z+
2 6 6

\ (5/2) z 2 + (19/6) z + (7/6) (5/2) z 2 + (19/6) z + (7/6)

H (z ) = 1 + = 1 +
z 3 + (1/2) z 2  (1/6) z  (1/6) [z  (1/2)] [ z 2 + z + (1/3)]
By partial fraction expansion, we can write

(5/2) z 2 + (19/6) z + (7/6) A Bz + C 81/26  (8/13) z  (10/39)

= + 2 = +
[z  (1/2)] [z + z + (1/3)]
2
z  (1/2) z + z + (1/3) z  (1/2) z 2 + z + (1/3)

81/26  (8/13) z  (10/39)

\ H (z ) = 1 + +
z  (1/2) z 2 + z + (1/3)

(81/26) z 1  (8/13) z 1  (10/39) z 2

=1+ +
1  (1/2) z 1 1 + z 1 + (1/3) z 2

Y (z) (81/26) z 1  (8/13) z 1  (10/39) z 2

Let H (z ) = =1+ +
X ( z) 1  (1/2) z 1 1 + z 1 + (1/3) z 2
Y ( z ) = X (z ) + Y1 (z ) + Y2 (z )

(81/26) z 1
where Y1 ( z ) = X (z )
1  (1/2) z 1

 (8/13) z 1  (10/39) z 2
and Y2 ( z) = X (z )
1 + z 1 + (1/3) z 2

Y1 ( z) Y ( z)
Also = H1 (z) and 2 = H 2 (z )
X (z ) X (z )

On cross multiplying the above equations for Y1(z) and Y2(z), we have
1 1 81 1
Y1 ( z )  z Y1 ( z ) = z X (z )
2 26

81 1 1
i.e. Y1 ( z ) = z X (z ) + z 1Y1 ( z)
26 2

1 2 8 10 2
and Y2 ( z) + z 1Y2 (z ) + z Y2 (z) =  z 1 X (z )  z X (z )
3 13 39
324 h Digital Signal Processing

8 1 10 2 1
\ Y2 (z ) =  z X (z)  z X (z )  z 1Y2 (z )  z 2Y2 ( z )
13 39 3
Using the equation for Y1(z), the direct form-I structure of H1(z) can be drawn as shown in
Figure 4.44(a).
Using equation for Y2(z) the direct form-I structure of H2(z) can be drawn as shown in
Figure 4.44(b).

Figure 4.44 Direct form-I structure of (a) H1(z) and (b) H2(z) (Example 4.12).

The parallel structure of H(z) is obtained by connecting the direct form-I structures of H1(z)
and H2(z) as shown in Figure 4.45.

4.4 STRUCTURES FOR REALIZATION OF FIR SYSTEMS

FIR (Finite duration impulse response) systems are the systems whose impulse response consists
of finite number of samples. They are designed by using only a finite number of samples of the
infinite duration impulse response. The convolution formula for FIR system is given by
N 1
y(n) =  h(k ) x ( n  k )
k 0

where h(n) = 0 for n < 0 and n ³ N. This implies that, the impulse response selects only N
samples of the input signal. In effect, the system acts as a window that views only the most
recent N input signal samples in forming the output. It neglects or simply forgets all prior
input samples. Thus, a FIR system has a finite memory of length N samples.
A system whose output depends only on the present and past inputs and not on past
outputs is called a non-recursive system. Hence, for non-recursive systems, the output y(n) is
given by
y(n) = F[ x (n), x (n  1), ..., x (n  M )]
Hence, in general, an FIR system is of non-recursive type.
 System Realization h 325

Figure 4.45 Parallel structure of H(z) (Example 4.12).

In non-recursive system, y(n0) can be computed immediately without having y(n0 – 1),
y(n0 – 2), ... . Hence, the output of non-recursive system can be computed in any order.
[i.e. y(50), y(5), y(2), y(100), ...]

Transfer function of FIR system

In general, an FIR system is described by the difference equation
N 1
y(n) =  bk x (n  k )
k 0

i.e. in general, in Finite Impulse Response (FIR) systems, the output at any instant depends
only on the present and the past inputs. It does not depend on the past outputs.
Taking Z-transform on both sides of equation for y(n), we get
N 1
y(z) =  bk z  k X ( z )
k 0
326 h Digital Signal Processing

The transfer function of the FIR system is:

" +b
N 1
Y ( z)
H (z ) =
X (z )
=  bk z  k = b0 + b1z 1 + b2 z 2 + N 1 z  ( N  1)
k 0

Also, we know that H(z) = Z[h(n)], where h(n) is the impulse response of the FIR system.
Let us replace the index n by k.

" + h(n  1) z
N 1
\ H (z ) =  h( k ) z  k = h(0) + h(1) z 1 + h(2) z 2 +  ( N  1)

k 0

On comparing the above two equations for H(z), we can say that bk = h(k) for k = 0, 1, 2, ...,
(N – 1).
The above equations for Y(z) and H(z) can be viewed as a computational procedure to
determine the output sequence y(n) from the input sequence x(n). These equations can be
used to construct the block diagram of the system using delays, adders and multipliers.
This block diagram is referred to as realization of the system or equivalently as a
structure for realizing the system.
The different types of structures for realizing FIR systems are:
1. Direct form realization 2. Transposed form realization
3. Cascade realization 4. Lattice structure realization
5. Linear phase realization

4.4.1 Direct Form Realization of FIR System

Since there are no denominator components or poles in an FIR system, the direct form has
only one structure which is called direct form. It realizes directly either the difference
equation or the system function.
The direct form structure can be obtained from the general equation for Y(z) governing
the FIR system.
N 1
Y (z) =  bk z  k X (z )
k 0

= b0 X (z ) + b1z 1 X (z ) + b2 z 2 X (z) + "+b N 2 z  ( N  2) X (z ) + bN 1z  ( N 1) X (z)

The direct form structure using the above equation for Y(z) can be drawn as shown in
Figure 4.46.
The direct form realization of Figure 4.46 is often called a transversal or tapped-delay-
line filter because the output is just tapping the delayed inputs. This structure is also called a
canonical structure because the number of delay elements used and the order of the filter is
the same.
 System Realization h 327

Figure 4.46 Direct form structure of FIR system.

4.4.2 Transposed Form Structure Realization of FIR System

The transposed form structure realization has already been discussed earlier with respect to
IIR systems. The same holds good for FIR systems also.

Procedure to realize transposed form structure of FIR system

1. First realize the given difference equation or transfer function by using the direct
form structure.
2. Reverse or transpose the direction of signal flow and interchange the input and
output nodes.
3. Replace the junction points by adders and adders by junction points.
4. Fold the structure, which is the transposed form realization of an FIR system.
In general, the transposed structure realization of an FIR system has no advantages
compared to the direct form structure. Whatever be the number of additions, multiplications
and storage components needed for the direct form structure, the same number of elements
are needed for the transposed structure too. The direct form structure and the transposed
form structure of a general Nth order FIR system are shown in Figure 4.47.

EXAMPLE 4.13 Realize the second order FIR system

y(n) = 2x (n) + 4x (n  1)  3x ( n  2)
by using the transposed form structure.
Solution: The given system is:
y(n) = 2x (n) + 4x (n  1)  3x ( n  2)
Taking Z-transform on both sides, we have

Y ( z) = 2X ( z) + 4z 1 X (z)  3 z 2 X (z )
The direct form structure, the recovered realization structure and the transposed structure are
shown in Figure 4.48 [(a), (b) and (c) respectively].
328 h Digital Signal Processing

Figure 4.47 (a) A general direct form structure of FIR system (b) Transposed structure of FIR system.

Figure 4.48 (a) Direct form structure (b) Recovered realization structure and (c) Transposed structure
(Example 4.13).

4.4.3 Cascade Form Structure Realization of FIR System

The cascade form structure is nothing, but a cascade connection or a series connection of
direct form structures. Hence, in cascade structure, the given transfer function H(z) is broken
into the product of many sub-transfer functions H1(z), H2(z), ..., HN(z) and each of these
sub-transfer functions is realized separately in direct form structure and all these are
connected in series or cascade.
 System Realization h 329

The block diagram representation of a cascade form structure is shown in Figure 4.49.

Figure 4.49 Block diagram of cascade structure.

The transfer function of FIR system, H(z) is an (N – 1)th order polynomial in z.

When N is odd, then (N – 1) will be even and so H(z) will have (N – 1)/2 second order
factors. When N is odd, then
N 1
N 1
3 (C0i + C1i z 1 + C2i z 2 ) = H1 H2 . . . H N 1
2
H (z ) =  bk z  k =
i 1
k 0 2

Each second order factor of the above equation for H(z) can be realized in direct form and
all the second order systems are connected in cascade to realize H(z) as shown in Figure 4.50.

Figure 4.50 Cascade structure of FIR system when N is odd.

When N is even, then (N – 1) will be odd and so H(z) will have one first order factor
and (N – 2)/2 second order factors.
When N is even, then
N 1
3 (C0i + C1i z 1 + C2i z 2 )
N/2
H ( z) =  bk z  k = (C0 i + C1i z 1 )
i2
k 0

= H1, H2, . . ., HN/2

In this case, the cascade structure will have one first order section and (N – 2)/2 second order
sections.
Each one of them can be realized in direct form and all of them connected in cascade
as shown in Figure 4.51.
330 h Digital Signal Processing

Figure 4.51 Cascade structure of FIR system when N is even.

4.4.4 Lattice Structure Realization of FIR Systems

The FIR structures discussed till now are used in general. However, the lattice structure is
extensively used in digital speech processing.
The lattice structure consists of two different paths through which the input x(n) is
processed. Hence, the lattice structure has two different output set-ups: y(n) and y¢(n). y(n) is
the real output and y¢(n) is the supporting output, which offers support for obtaining the
output for the next stage. A single stage lattice structure is shown in Figure 4.52, wherein K
is called the reflection coefficient. Therefore, the output from the first stage of the lattice
structure, which is the first order FIR system, is given below.
y1 (n) = x (n) + K1 x (n  1)
y1 (n) = K1 x (n) + x (n  1)

Figure 4.52 Single stage lattice structure.

Similarly, the two-stage lattice structure is shown in Figure 4.53. The output from the second
stage of Figure 4.53, which is the second order FIR system is given as follows:
y2 (n) = y1 (n) + K 2 y1 (n  1)
y2 (n) = K 2 y1 (n) + y1 (n  1)
 System Realization h 331

Figure 4.53 Two-stage lattice structure.

Substituting the values of y1(n) and y¢1(n) in the above equations for y2(n) and y¢2(n) we get

y2 (n) = [x (n) + K1 x ( n  1)] + K 2 [K1 x ( n  1) + x (n  2)]

or y2 (n) = x (n) + K1 [1 + K 2 ] x ( n  1) + K 2 x ( n  2)
y2 (n) = K 2 [x (n) + K1 x ( n  1)] + K1 x ( n  1) + x (n  2)]
or y2 (n) = K 2 x (n ) + K 1 [1 + K 2 ] x ( n  1) + x ( n  2)
Let B 2 (0) = 1, B 2 (1) = K1 (1 + K 2 ) and B 2 (2) = K2
Then the equation for y2(n) changes to

y2 ( n) = B 2 (0) x (n) + B 2 (1) x ( n  1) + B 2 (2) x (n  2)

2
or y2 (n) =  B2 (k ) x(n  k )
k 0

Therefore, in general, the output of the mth order FIR system by using the lattice structure
can be written as:
m
ym (n) =  B m ( k ) x (n  k )
k 0

The above equation is the convolution sum. Using the convolution property it can be
re-written as:
Ym (z ) = B m (z ) X ( z )

Hence, the transfer function of the FIR system can be obtained as

Ym ( z)
B m (z) =
X (z )

The FIR filter having a system function given above is called the forward prediction.
332 h Digital Signal Processing

Procedure to realize the lattice structure of FIR system

1. If the coefficient of the present input x(n) is not unity, convert it to unity by taking
common of the coefficient of the present input.
2. Find the order of the difference equation and compare the coefficients of the given
difference equation with the coefficients of the same order lattice structure output
involving the reflection coefficients K1, K2, K3, ...
3. Assign the calculated values of K1, K2, K3, ... and construct the structure.
In order to realize by using lattice structure, it is enough to find the values of reflection
coefficients. Therefore, it is easily programmable. But the number of components used,
specially adders and multipliers will increase.

EXAMPLE 4.14 Realize a system with H(z) = 5 + 3z–1 by using lattice structure.

Y (z )
Solution: Given H ( z) = = 5 + 3z 1
X (z )

or Y(z) = 5X(z) + 3z–1X(z)

Taking the inverse Z-transform on both sides, we get
y(n) = 5x(n) + 3x(n – 1)

3
= 5[x (n) + x ( n  1)] = 5 p(n)
5
3
This is of first order. So comparing p(n) = x ( n) + x (n  1) with the standard equation for first
5
order lattice structure, i.e. with y(n) = x(n) + K1x(n – 1), we get

3
K1 =
5
Therefore, the lattice structure realization of the given first order FIR system is as shown in
Figure 4.54.

Figure 4.54 Lattice structure (Example 4.14).

 System Realization h 333

EXAMPLE 4.15 Determine the lattice coefficients corresponding to the FIR system with
7 1 3 2
the system function H ( z) = 1 + z + z and realize it.
9 5

Y ( z) 7 3
Solution: Given H ( z) = = 1 + z 1 + z 2
X ( z) 9 5
7 1 3
\ Y (z ) = X ( z ) + z X ( z ) + z 2 X ( z )
9 5
Taking inverse Z-transform on both sides, we get

7 3
y(n) = x (n) + x (n  1) + x ( n  2)
9 5
This corresponds to a second order system. Comparing this with the standard equation for
second order lattice structure, i.e. with y( n) = x (n) + K1 (1 + K 2 ) x (n  1) + K2 x ( n  2) , we get
3
K2 =
5

and 7
K1 (1 + K 2 ) =
9
7/9 35
or K1 = =
1 + (3/5) 72

Therefore, the lattice structure realization of the given FIR system is shown in Figure 4.55.

Figure 4.55 Lattice structure (Example 4.15).

4.4.5 Linear Phase Realizations

An FIR system is said to be linear phase, if it satisfies the condition
h(k ) = ±h( N  1  k )
An FIR system which satisfies the relation h( k ) = h( N  1  k ) is called a symmetric FIR
system and an FIR system which satisfies the relation h(k ) =  h( N  1  k ) is called an
334 h Digital Signal Processing

antisymmetric FIR system. Here we discuss symmetric FIR systems. The symmetric
condition may be viewed as
1. Odd symmetry
2. Even symmetry
The impulse response is symmetrical, i.e. h( k ) = h( N  1  k ) means

N  N
h(0) = h ( N  1); h (1) = h( N  2); . . . ; h   1  = h  
2  2
By using this symmetry condition, it is possible to reduce the number of multipliers required
for the realization of FIR system.

Odd symmetry (N=odd)

A function is said to be of odd symmetry if it has an odd number of samples and satisfies
the condition h(k) = h(N – 1 – k).
For example, if N = 11 (i.e. N is odd), then a linear phase FIR system will have
h(0) = h (9  1  0) = h (8), h(1) = h(9  1  1) = h(7), h(2) = h(9  1  2) = h(6)
h(3) = h(9  1  3) = h(5), h(4) = h(9  1  4) = h(4)
In case of odd symmetry, it is obvious that the central sample is the same for both LHS and
RHS. Hence, in case of odd symmetry, the common or central sample will be lying at
(N – 1)/2 The graphical representation of the above impulse response for N = 11 is shown in
Figure 4.56.

Figure 4.56 Linear phase impulse response sequence for N = 11.

The linear phase FIR system can be remodeled as

N 1
Y (z )
H (z ) =  h(k ) z k = X (z)
k 0
 System Realization h 335

or Y ( z) = [h(0) + h(1) z 1 + h(2) z 2 + " + h(N  2) z  ( N  2)

+ h( N  1) z ( N 1) ] X ( z)

For an odd symmetry linear phase FIR system, we have

 N 3  N + 1  N 1  N 1

h(0) = h ( N  1), h(1) = h ( N  2), ..., h   = h  , h  = h 
 2   2   2   2 

 N 1
The impulse h   will remain single.
 2 
Substituting these in the expression for Y(z), we have the direct form structure of linear
phase FIR system as

Y ( z) = h0 [X (z ) + z  ( N 1) X (z)] + h1 [z 1 X ( z ) + z  ( N 2) X (z)] + "+

  ( N 5) 
( N  3)    ( N 3) 
( N 1)  
( N 1)
h N  5  z 2 X ( z ) + z 2 X ( z )  + h N 3  z 2 X (z) + z 2 X (z )  + h N 1 z 2 X (z )
2    2    2

which is constructed as shown in Figure 4.57.

Figure 4.57 Direct form realization of FIR system when N is odd.

In Figure 4.57
Y0 = h0 [X (z) + z  ( N 1) X (z )]; Y1 = h1 [z 1 X (z) + z  ( N  2) X (z)]

  ( N  5) 
( N  3) 
Y N 5 = h N 5  z 2 X ( z ) + z 2 X ( z) 
2 2   

  (N 3) 
(N  1) 
Y N 3 = h N 3 z 2 X (z ) + z 2 X ( z ) 
2 2  
336 h Digital Signal Processing

Even symmetry
A function is said to have even symmetry if it has even number of samples and satisfies the
condition h(k) = h(N – 1 – k).
For example, if N = 8 (i.e. N is even), then a linear phase symmetric FIR system will
have
h(0) = h (8  1  0) = h (7), h (1) = h(8  1  1) = h(6), h(2) = h (8  1  2) = h (5)
and h(3) = h(8  1  3) = h(4)
In the case of even symmetry, since there is no central sample, a virtual central sample
is selected at k = (N – 1)/2. In this case virtual sample is at k = (8–1)/2 = 3.5.
The graphical representation of the above impulse responses for N = 8 is shown in
Figure 4.58.

Figure 4.58 Linear phase impulse response sequence for N = 8.

For an even symmetry linear phase FIR system, we have

N  N
h(0) = h( N  1), h(1) = h( N  2), ..., h   1  = h  
2  2
Substituting these in the expression for Y(z), we have the direct form structure of linear
phase FIR system as:
N 

" +h    1 N

Y (z ) = h0 [X (z) + z  ( N  1) X (z )] + h1[z 1 X ( z) + z  ( N  2) X (z )] + N [z 2 
X (z) + z 2 X ( z )]
1
2

as shown in Figure 4.59.

In Figure 4.59.
Y0 = h0 [X (z) + z  ( N  1) X (z )]; Y1 = h1 [z 1 X (z) + z  ( N  2) X (z )]

   N  2  N 
   1 
YN = hN  z 2 
X (z ) + z 2 
X (z ) 
2 2  
2 2  
 System Realization h 337

Figure 4.59 Direct form realization of FIR system when N is even.

   N  1  N
  
YN = hN  z 2 
X (z ) + z  2  X (z) 
1 1  
2 2  

   N  1  
Y N  1 = h N  1  z  2  X (z) 
2 
 
2 

EXAMPLE 4.16 Draw the direct form structure of the FIR system described by the
transfer function
1  1 3 2 1 3 1 4 1 5
H (z ) = 1 + z + z + z + z + z
5 4 3 7 6
Y (z ) 1 3 1 1 1
Solution: Let H (z) = = 1 + z 1 + z 2 + z 3 + z 4 + z 5
X ( z) 5 4 3 7 6
1 1 3 1 1 1
\ Y (z ) = X (z ) + z X (z ) + z 2 X (z ) + z 3 X (z ) + z 4 X (z ) + z 5 X (z )
5 4 3 7 6
The direct form structure of FIR system can be obtained directly from the above equation for
Y(z) as shown in Figure 4.60.

Figure 4.60 Direct form structure of H(z) (Example 4.16).

338 h Digital Signal Processing

EXAMPLE 4.17 Realize the filter transfer function given by the expression below using
the direct form

H ( z) = (1  z 1 ) (1 + 2 z 1  3 z 2 )

Solution: Since the given filter transfer function has only the numerator part or zeros, it is
an FIR system. To realize in direct form, multiply the factors and obtain a single expression.

Y (z )
Given H ( z) = = (1  z 1 )(1 + 2z 1  3z 2 ) = 1 + z 1  5z 2 + 3z 3
X ( z)
On cross multiplying, we get

Y ( z) = X (z) + z 1 X (z )  5z 2 X (z) + 3z 3 X (z )
The direct form realization of Y(z) is shown in Figure 4.61.

Figure 4.61 Direct form realization of H(z) (Example 4.17).

EXAMPLE 4.18 Realize the following systems with minimum number of multipliers.
1 1 1 2 2 1 3 1 4
(a) H (z ) = + z + z + z + z
3 5 3 5 3
1 1 1 1
(b) H (z ) = + z 1 + z 2 + z 3
2 4 4 2
 1   1 1 2 
(c) H ( z ) =  1 + z 1 + z 2  1 + 5 z + z 
 3   

Solution:
1 1 1 2 2 1 3 1 4
(a) Given H (z ) = + z + z + z + z
3 5 3 5 3
By the definition of Z-transform, we get

 h(n)z n = h(0) + h(1) z 1 + h(2) z 2 + h(3) z 3 + "


H ( z) =
n 0
 System Realization h 339

On comparing the above two equations for H(z), we get

1 1 2 1 1 
Impulse response h( n) =  , , , , 
3 5 3 5 3

Here h(n) satisfies the condition, h(n) = H(N – 1 – n) and so the impulse response
is symmetrical. Hence, the system has linear phase and can be realized with
minimum number of multipliers.

Y (z ) 1 1 2 1 1
Let H ( z) = = + z 1 + z 2 + z 3 + z 4
X ( z) 3 5 3 5 3
1 1 2 1 1
\ Y ( z) = X ( z ) + z 1 X ( z ) + z 2 X ( z ) + z 3 X ( z ) + z 4 X ( z )
3 5 3 5 3

1 1 2
= [ X ( z ) + z 4 X ( z)] + [ z 1 X ( z ) + z 3 X ( z )] + z 2 X ( z )
3 5 3
The direct form structure of linear phase FIR system is constructed using the above
equation for Y(z) as shown in Figure 4.62.

Figure 4.62 Linear phase realization of H(z) (Example 4.18(a)).

1 1 1 1
(b) Given H ( z) = + z 1 + z 2 + z 3
2 4 4 2

Y ( z) 1 1 1 1
Let H (z ) = = + z 1 + z 2 + z 3
X (z ) 2 4 4 2
340 h Digital Signal Processing

1 1 1 1
\ Y ( z) = X ( z) + z 1 X ( z ) + z 2 X ( z ) + z 3 X ( z)
2 4 4 2

1 1
= [ X (z) + z 3 X (z)] + [ z 1 X ( z) + z 2 X ( z)]
2 4
The direct form realization of H(z) with minimum number of multipliers (i.e. linear
phase realization) is obtained using the above equation for Y(z) as shown in Figure 4.63.

Figure 4.63 Linear phase realization of H(z) [Example 4.18(b)].

 1  1 
(c) Given H ( z ) =  1 + z 1 + z 2   1 + z 1 + z 2 
 3  5 
The given system can be realized as cascade of two second order systems. Each system
can be realized with minimum number of multipliers (i.e. linear phase realization).
Let H ( z) = H1 (z) H 2 (z )
1 1 1
where H1 ( z ) = 1 + z + z 2 and H 2 (z ) = 1 + z 1 + z 2
3 5
Y1 ( z ) 1
Let H1 ( z ) = = 1 + z 1 + z 2
X ( z) 3
1 1
\ Y1 ( z ) = X ( z ) + z X (z) + z 2 X (z)
3
1 1
= [X (z ) + z 2 X (z)] + z X ( z)
3
The linear phase realization structure of H1(z) using the above equation for Y1(z) is
as shown in Figure 4.64(a).
 System Realization h 341

Figure 4.64 Linear phase realization of (a) H1(z) and (b) H2(z) [Example 4.18(c)].

Y ( z) 1
Let H2 (z ) = = 1 + z 1 + z 2
Y1 ( z) 5

1 1
\ Y (z) = Y1 ( z) + z Y1 ( z ) + z 2Y1 (z )
5
1 1
= [Y1 (z ) + z 2Y1 (z )] + z Y1 ( z)
5
The linear phase realization structure of H2(z) using the above equation for Y(z) is
as shown in Figure 4.64(b).
The linear phase structure of H(z) is obtained by connecting the systems H1(z)
and H2(z) in cascade as shown in Figure 4.65.

Figure 4.65 Cascade realization of H(z) [Example 4.18(c)].

342 h Digital Signal Processing

TABLE 4.2 Comparison between IIR and FIR systems

IIR system FIR system

· Impulse response has infinite number of Impulse response has finite number of samples.
samples.
· Output depends upon present input sample, Output depends upon present input sample
past input samples and past output samples. and past input samples.
· Both poles and zeros are present. Only zeros are present.
· It is a recursive system. It is a non-recursive system.
· Not always stable. Finite word length Stability is guaranteed, because the system
effect can make it unstable. impulse response is of finite length and
therefore bounded.
· It is a closed loop system. It is an open loop system.
· More complex to implement. Less complex to implement.
· Reliability is not always guaranteed. Reliability is always guaranteed.
· Design is first done in analog domain and Design is done directly in digital domain.
then transformed into a digital domain.
· Magnitude response is better, sharp cut-off It requires higher order system to get sharp
with lower order itself. magnitude response.
· It does not have linear phase characteristics. It has linear phase characteristics.

SHORT QUESTIONS WITH ANSWERS

1. What is cascade form realization?
Ans. Cascade form realization is one in which the given transfer function is
expressed as a product of several transfer functions and each of these transfer
functions is realized in direct form-II and then all of them are cascaded.
2. What is parallel form realization?
Ans. Parallel form realization is one in which the given transfer function is
expressed in its partial fractions and each factor is realized in direct form-II and
then all those realized structures are connected in parallel.
3. What is recursive and non-recursive system? Give an example.
Ans. If the output y(n) at time n of a system depends on present and past inputs
as well as on past outputs, then the system is called a recursive system.
Example: y(n) = y(n – 1) + 0.5x(n) + x(n – 1)
If the output y(n) at time n of a system does not depend on past output values, then
the system is called a non-recursive system. In non-recursive systems, the outputs
are functions of present and past outputs.
Example: y(n) = x(n) + 2x(n – 1) – 1.5x(n – 2)
 System Realization h 343

4. What are the basic elements used to construct the block diagram of a discrete-time
system? Draw their symbols.
Ans. The basic elements used to construct the block diagram of a discrete-time
system are: Adder, Constant multiplier and Unit delay element. The symbols are
given below.

x1(n) x1(n) + x2(n) x(n) “Kx(n) x(n) x(n – 1)

+ “K z–1
X1(z) X1(z) + X2(z) X(z) “KX(z) X(z) X(z – 1)
x2(n)
X2(z)
(a) Adder (b) Multiplier (c) Unit delay element

5. What is an IIR system?

Ans. An IIR system is one which is designed by selecting all the infinite samples
of impulse response.
6. Write the convolution sum formula for IIR system?
Ans. The convolution sum formula for IIR system is

y ( n) =  h( k ) x ( n  k )
k 0

7. Write the general difference equation and transfer function of an IIR system.
Ans. In general, an IIR system is described by the difference equation
N M
y( n) =   ak y ( n  k ) +  bk x (n  k )
k 1 k 0

Its transfer function is

 bk z k
Y (z ) k 0 b0 + b1 z 1 + b2 z 2 + "+b z M

+"+a z
M
H ( z) = = =
X ( z) N
1 + a1 z 1 + a2 z 2 N
1+  ak z  k N

k 1

8. What are the factors that influence the choice of structure for realization of an LTI
system?
Ans. The major factors that influence the choice of structure for realization of
LTI system are: computational complexity, memory requirements and finite word
length effects. Other factors that influence the choice of realization are: whether the
structure lends itself to parallel processing or whether the computations can be
pipelined.
344 h Digital Signal Processing

9. List the different types of structures for realization of IIR systems.

Ans. The different types of structures for the realization of IIR systems are:
(a) Direct form-I structure (b) Direct form-II structure
(c) Transposed form structure (d) Cascade form structure
(e) Parallel form structure (f) Lattice structure
(g) Ladder structure
10. What is the advantage of direct form-II structure when compared to direct form-I
structure? How?
Ans. The advantage of direct form-II structure when compared to direct form-I
structure is:
The direct form-II structure requires less amount of memory when compared to
direct form-I structure. This is because in direct form-II structure, the number of
delay elements required is exactly half of that for direct form-I structure when the
number of poles and zeros are equal.
11. Why Direct form-I is called non-canonical structure?
Ans. Since the number of delay elements used in direct form-I is more than the
order of the difference equation, direct form-I is called a non-canonical structure.
12. Why Direct form-II is called canonical structure?
Ans. Since the number of delay elements used in direct form-II is the same as
that of the order of the difference equation, direct form-II is called a canonical
structure.
13. What are the difficulties in cascade realization?
Ans. The difficulties in cascade realization are:
(a) Decision of pairing poles and zeros.
(b) Deciding the order of cascading the first and second order sections.
(c) Scaling multipliers should be provided between individual sections to prevent
the filter variables from becoming too large or too small.
14. What is the advantage in cascade and parallel realization of IIR systems?
Ans. The advantage in cascade and parallel realization of IIR systems is that the
sensitivity of frequency response characteristics to quantization of the coefficients
is minimized.
15. What is an FIR system?
Ans. An FIR system is one which is designed by selecting only a finite number
of samples of impulse response.
16. Write the convolution sum formula for FIR system?
Ans. The convolution sum formula for FIR system is
N 1
y( n) =  h (k ) x ( n  k )
k 0

17. Write the general difference equation and transfer function of an FIR system.
Ans. In general, an FIR system is described by the difference equation
 System Realization h 345

N 1
y ( n) =  bk x (n  k )
k 0

Its transfer function is

"+b
N 1
Y (z )
H (z ) =
X ( z)
=  bk z k = b0 + b1 z 1 + b2 z 2 + N 1 z
 ( N  1)

k 0

18. List the different types of structures for realizing FIR systems.
Ans. The different types of structures for realizing FIR systems are:
(a) Direct form realization (b) Transposed form realization
(c) Cascade realization (d) Lattice structure realization
(e) Linear phase realization
19. What is the advantage of linear phase realization of FIR systems?
Ans. The advantage of linear phase realization of FIR systems is that the system
can be realized with minimum number of delay elements.

REVIEW QUESTIONS
1. Discuss the basic elements used to construct the block diagram of discrete-time
systems.
2. Explain the factors that influence the choice of structure for realization of a LTI system.
3. Write the difference equations for FIR and IIR system and hence derive the transfer
functions of FIR and IIR systems.
4. Discuss the different methods of realization of IIR systems and explain how
conversion can be made from direct form-I structure to direct form-II structure.
5. Discuss the different methods of realization of FIR systems.
6. Compare FIR and IIR systems.
7. Compare cascade and parallel form realizations.

FILL IN THE BLANKS

1. A system whose output y(n) at time n depends only on the present and past inputs
is called a system.
2. A system whose output y(n) at time n depends on the past outputs is called a
system.
3. The basic elements used to construct the block diagram of a discrete-time system
are , , .
4. refers to the number of arithmetic operations required to compute an
output value y(n) for the system.
346 h Digital Signal Processing

5. refers to the number of memory locations required to store the system

parameters, past inputs and outputs and any intermediate computed values.
6. refer to the quantization effects that are inherent in any digital
implementation of the system, either in hardware or in software.
7. In systems, the impulse response consists of infinite number of samples.
8. The convolution sum formula for IIR system is .
9. The difference equation describing an IIR system is .
10. The transfer function of an IIR system is .
11. The different types of structures for realizing IIR systems are , ,
, .
12. structure provides a direct relation between time domain and z-domain
equations.
13. Direct form-II realization of discrete-time systems uses less number of
than direct form-I realization.
14. In systems, the impulse response consists of finite number of samples.
15. The convolution sum formula for FIR systems is .
16. The difference equation describing an FIR system is .
17. The transfer function of an FIR system is .
18. The different types of structures for realizing FIR systems are , ,
.
19. In FIR systems, for linear phase response, the should be symmetrical.
20. Linear phase results in reduction of the number of required for the realiza-
tion of FIR system.

OBJECTIVE TYPE QUESTIONS

1. A system whose output y(n) at time n depends on any number of past output values
is called a
(a) recursive system (b) non-recursive system
(c) causal system (d) non-causal system
2. A system whose output y(n) at time n depends only on present and past input
values is called a
(a) recursive system (b) non-recursive system
(c) causal system (d) non-causal system
3. The structure which uses less number of delay elements is
(a) direct form-I (b) direct form-II
(c) cascade form (d) parallel form
 System Realization h 347

4. The number of multipliers required for the realization of FIR systems is reduced if
we choose
(a) direct form (b) cascade form
(c) parallel form (d) linear phase realization

PROBLEMS
1. Construct the block diagram and signal flow graph of a discrete-time system whose
input-output relations are described by the difference equation
(a) y( n) = 0.5x (n) + 0.5x ( n  1)
(b) y(n) = 0.25y( n  1) + 0.5x ( n) + 0.75x (n  1)
2. Find the digital network in direct and transposed form for the system described by
the difference equation
y( n) = x (n)  0.3x (n  1)  0.7x (n  2) + 0.6y(n  1) + 0.8y(n  2)
3. Determine the direct form-I and direct form-II realizations of the following LTI
systems.
(a) y(n) =  0.5y(n  1) + 0.25y(n  2) + 0.125y(n  3) + x( n) + 0.5x (n  1) + 0.75x (n  2)
3 3 1
(b) y(n) =  y( n  1) + y( n  2) + y( n  3) + x ( n) + 3x (n  1) + 2x (n  2)
8 32 64
4. An LTI system is described by the difference equation
y(n) = a1 y( n  1) + x (n) + b1 x ( n  1)
Realize it in direct form-I structure and convert it to direct form-II structure.
5. Obtain the direct form-I, direct form-II, cascade and parallel realization of the
following LTI systems.
13 9 1
(a) y(n) = y( n  1)  y(n  2) + y( n  3) + x (n)  4x ( n  1) + 3x ( n  2)
12 24 24
1 1 1 1
(b) y( n) = y( n  1) + y(n  2) + x ( n) + x (n  1)  x ( n  2)
6 6 6 6
6. Realize the system in cascade and parallel forms
1 1
1+ z
(a) H (z) = 3
 1 1  1 
1  2z + z 2  1  z 1 + z 2 
 3  5 

1 1
1+ z
(b) H ( z ) = 2
 1 1  1 
1  z + z 2   1  z 1 + z 2 
 4  2 
348 h Digital Signal Processing

7. Realize the following IIR system in cascade and parallel forms

1 1
y( n) + y(n  1)  y(n  2) = x ( n)  2x (n  1) + x (n  2)
4 8
8. Determine the direct form-I, direct form-II, cascade and parallel realization of the
following LTI systems.
z 3  2z 2 + z  1 1 + 0.5 z 1
(a) H ( z) = (b) H ( z) =
(z  0.4) ( z 2 + 2z  0.4) (1  z 1 + 0.5 z 2 ) (1  z 1 + 0.8 z 2 )
9. Realize the IIR filter using ladder structure

2z 2 + 3z + 4 z 3 + 3z 2 + 2z + 5
(a) H ( z) = (b) H ( z) =
z 2 + 5z + 7 2z2 + z + 4
10. Realize the second order IIR system using the transposed form structure
1 1 1 1
y ( n) + y(n  1) + y( n  2) = 2x ( n) + x (n  1) + x ( n  2)
2 4 3 5
11. Determine the lattice coefficients corresponding to the IIR filter described by
1 1 1
y( n) 
y(n  1) + y( n  2) = x (n) + x ( n  1) and realize it.
2 7 5
12. Draw the direct form structure of the FIR system described by the transfer function
1 1 1 2 1 3 1 4 1 5
(a) H ( z) = 1 + z + z + z + z + z
3 2 4 5 7
1 1 1 2 1 3 1 4
(b) H ( z ) = 2 + z + z + z + z
2 4 3 6
1  1 3 2 1 3 1 4 1 5
(c) H ( z ) = 1 + z + z + z + z + z
2 4 4 2 8
13. Realize the following systems with minimum number of multipliers
(a) H (z) = 1 + 3z 1 + 2z 2 + 5z 3 + 2z 4 + 3z 5 + z 6
1 2 3 4 5 6 7 8
(b) H ( z) = 0.2 + 0.6z + 0.7z + 0.8z + 0.9z + 0.8z + 0.7z + 0.6z + 0.2z
(c) H (z) = 0.5 + 0.75z 1 + 0.8z 2 + 0.9z 3 + 2z 4 + 0.9z 5 + 0.8z 6 + 0.75z 7 + 0.5z 8
1 1 3 1 1
(d) H (z) = + z 1 + z 2 + z 3 + z 4
4 2 4 2 4
 1  1 
(e) H (z) =  1 + z 1 + z 2   1 + z 1 + z 2 
 2  4 
 System Realization h 349

14. Realize the following systems in cascade form

 1  1 
(a) H ( z) =  1 + z 1   1 + z 1 + z 2 
 2  3 

 1  1 1 
(b) H (z) = 1 + z 1 + z 2   1 + z 1 + z 2 
 2  5 3 
15. Realize the following second order FIR system using the transposed form structure
y(n) = 3x (n) + 5x (n  1)  2x (n  2)
16. Determine the lattice coefficients corresponding to the FIR system with the system
5 1 2 2
function H ( z) = 1 + z + z and realize it.
12 3
350 h Digital Signal Processing

MATLAB PROGRAMS

Program 4.1
% Parallel form realization of IIR filters

clc; clear all; close all;

num=[2 10 23 34 31 16 4];
den=[36 78 87 59 26 7 1];
[r1 p1 k1]=residuez(num,den);
[r2 p2 k2]=residue(num,den);
disp(‘parallel form 1’)
disp(‘residues are’)
disp(r1)
disp(‘poles are at’)
disp(p1)
disp(‘constant value’)
disp(k1)
disp(‘parallel form II’)
disp(‘residues are’)
disp(r2)
disp(‘poles are at’)
disp(p2)
disp(‘constant value’)
disp(k2)

Output:
parallel form 1
residues are
–0.5952 – 0.7561i
–0.5952 + 0.7561i
–0.5556 – 2.2785i
–0.5556 + 2.2785i
–0.8214 + 4.3920i
–0.8214 – 4.3920i
 System Realization h 351

poles are at
–0.5000 + 0.2887i
–0.5000 – 0.2887i
–0.3333 + 0.4714i
–0.3333 – 0.4714i
–0.2500 + 0.4330i
–0.2500 – 0.4330i
constant value
4
parallel form II
residues are
0.5159 + 0.2062i
0.5159 – 0.2062i
1.2593 + 0.4976i
1.2593 – 0.4976i
–1.6964 – 1.4537i
–1.6964 + 1.4537i
poles are at
– 0.5000 + 0.2887i
– 0.5000 – 0.2887i
– 0.3333 + 0.4714i
– 0.3333 – 0.4714i
– 0.2500 + 0.4330i
– 0.2500 – 0.4330i
constant value
0.0556

Program 4.2
% Direct form to cascade form conversion
clc; clear all; close all;
b=[4 5 6]; %numerator coefficients of direct form
a=[1 2 3]; %denominator coefficients of direct form
% compute gain coefficient
b0=b(1); a0=a(1);
b=b/b0; a=a/a0;
m=length(b);
n=length(a);
if n > m
b= [b zeros(1,n–m)];
352 h Digital Signal Processing

elseif m > n
a=[a zeros(1,m–n)];
end
k=floor(n/2);
B=zeros(k,3); A=zeros(k,3);
if k*2== n
b=[b 0];
a=[a 0];
end
broots=cplxpair(roots(b));
aroots=cplxpair(roots(a));
for i=1:2:2*k
brow=broots(i:1:i+1,:);
brow=real(poly(brow));
B(fix((i+1)/2),:)=brow;
arow=aroots(i:1:i+1,:);
arow=real(poly(arow));
A(fix((i+1)/2),:)=arow;
end
disp(‘numerator coefficients of cascade form’)
disp(brow)
disp(‘denominator coefficients of cascade form’)
disp(arow)

Output:
numerator coefficients of cascade form
1.0000 1.2500 1.5000

denominator coefficients of cascade form

1.0000 2.0000 3.0000
 System Realization h 353

Program 4.3
% Cascade form realization of FIR & IIR filters

clc; clear all; close all;

b=[4 5 6]; %numerator coefficients of direct form
a=[1 2 3]; %denominator coefficients of direct form
% compute gain coefficient
b0=b(1);
x=[1 2 3 8 9 4 6 7 10];% input sequence
[k l]=size(b);
n=length(x);
w=zeros(k+1,n);
w(1,:)=x;
for i=1:1:k
w(i+1,:)=filter(b(i,:),a(i,:),w(i,:));
end
y=b0*w(k+1,:);
disp(‘output of the final filter operation’)

disp(y)

Output:
output of the final filter operation
16 20 24 128 48 –44 336 –212 –140

Program 4.4
% Cascade form to direct form conversion

clc; clear all; close all;

B=[4 5 6]; %numerator coefficients of cascade form
A=[1 2 3]; %denominator coefficients of cascade form
% compute gain coefficient
b0=B(1);
[k l]=size(B);
354 h Digital Signal Processing

b=[1];
a=[1];

for i=1:1:k
b=conv(b,B(i,:));
a=conv(a,A(i,:));
end

b=b*b0;
disp(‘numerator coefficients of direct form’)
disp(b)
disp(‘denominator coefficients of direct form’)
disp(a)

Output:
numerator coeffcients of direct form
16 20 24

denominator coeffcients of direct form

1 2 3

Program 4.5
%Direct form to parallel form conversion

clc; clear all;close all;

b=[4 5 6]; %numerator coefficients of direct form
a=[1 2 3]; %denominator coefficients of direct form
m=length(b);
n=length(a);
[r1 p1 c]=residuez(b,a);
p=cplxpair(p1,10000000*eps);
p2=cplxpair(p1);
I=[];
for j=1:1:length(p2)
 System Realization h 355

for i=1:1:length(p1)
if(abs(p1(i)–p2(j)) < 0.0001)
I=[I,i];
end
end
end
I=I’;
r=r1(I);

K=floor(n/2);B=zeros(K,2); A=zeros(K,3);

if K*2 ==n
for i=1:2:n-2
Brow=r(i:1:i+1,:);
Arow=p(i:1:i+1,:);
[Brow Arow]=residuez(Brow,Arow,[]);
B(fix((i+1)/2),:)=real(Brow);
A(fix((i+1)/2),:)=real(Arow);
end

[Brow Arow]=residuez(r(n–1),p(n–1),[]);
B(K,:)=[real(Brow) 0]; A(K,:)=[real(Arow) 0];

else
for i=1:2:n–1
Brow=r(i:1:i+1,:);
Arow=p(i:1:i+1,:);
[Brow Arow]=residuez(Brow,Arow,[]);
B(fix((i+1)/2),:)=real(Brow);
A(fix((i+1)/2),:)=real(Arow);
end
end

disp(‘numerator coefficients of parallel form’)

disp(B)
disp(‘denominator coefficients of parallel form’)
disp(A)
356 h Digital Signal Processing

Output:
numerator coefficients of parallel form
2.0000 1.0000

denominator coefficients of parallel form

1.0000 2.0000 3.0000

Program 4.6
% Parallel form to direct form conversion

clc;clear all;close all;

C=[0];
A=[1 1 0.9; 1 0.4 –0.4];
B=[2 4; 3 1];

[K,L]=size(A);
R=[]; P=[];
for i=1:1:K
[r p k]=residuez(B(i,:),A(i,:));
R=[R;r];
P=[P;p];
end
[b a]=residuez(R,P,C);
b=b(:)’;
a=a(:)’;

disp(‘numerator coefficients of direct form’)

disp(b)
disp(‘denominator coefficients of direct form’)
disp(a)
 System Realization h 357

Output:
numerator coefficients of direct form

5.0000 8.8000 4.5000 –0.7000 0

denominator coefficients of direct form

1.0000 1.4000 0.9000 –0.0400 –0.3600