You are on page 1of 62

Digital Filter

Structures
Singapuri Harikrishna M.
ECED, SVNIT, Surat

Digital Filter Structures


The input-output relation involves a
finite sum of products:
y[n] k 1 ak y[n k ] k 0 bk x[n k ]
N

In terms of Z transform,
M

H ( z)

k
b
z
k
k 0
N

1 ak z k
k 1

Digital Filter Structures


The major factors that affects or
influence realization are:
Complexity
Memory requirements
Finite Word Length

Complexity
Referred as The number of Arithmetic
Operations
Multiplications
Divisions
Additions

Key element to measure computational


Complexity
Recent days, various measuring term is
used:
No. of programmable DSP chips
No. of times a data fetch from Memory

Memory Requirements
No. of Memory Locations Required to
store
System parameters
Past inputs
Past outputs
Intermediate values

Finite Word Length


The actual implementation of an LTI
digital filter can be either in software
or hardware form, depending on
applications.
In either case, the signal variables
and the filter coefficients cannot
represented with finite precision

Finite Word Length


A direct implementation of a filter based
on
Difference equation or
Finite convolution sum

may not provide satisfactory performance


due to the finite precision arithmetic
Interest to develop alternate realizations
and choose the structure that
provides satisfactory performance under finite
precision arithmetic

Digital Filter Structures


A
structural
representation
using
interconnected basic building blocks is
the first step in the hardware or
software implementation of an LTI
digital filter
The structural representation provides
the key relations between some
pertinent internal variables with the
input and output that in turn provides
the key to the implementation

Basic Building Blocks


The computational algorithm of an
LTI
digital
filter
can
be
conveniently represented in block
diagram form using the basic
A
building
blocks
shown
below

y[n]
x[n]
y[n]
x[n]
w[n]

Multiplier

Adder

x[n]

z 1
Unit delay

x[n]

x[n]
y[n]

x[n]
Pick-off node

Digital Filter Structures


The convolution sum description of
an LTI discrete-time system can be
used, in principle, to implement the
system
For an IIR finite-dimensional system,
this approach is not practical as here
the impulse response is of infinite
length
However, a direct implementation of

Block Diagram Representation


In the time domain, the input-output
relations of an LTI digital filter is given
by the convolution
sum

y[n] k h[k ] x[n k ]


or, by the linear constant coefficient
difference equation
y[n]

N
k 1 d k y[ n k ]

M
k 0 pk x[ n k ]

Block Diagram Representation


For the implementation of an LTI digital
filter, the input-output relationship must
be described by a valid computational
algorithm
To illustrate what we mean by a
computational algorithm, consider the
causal first-order LTI digital filter shown
below

Block Diagram Representation


The filter is described by the
difference equation
y[n] d1 y[n 1] p0 x[n] p1x[n 1]
Using the above equation we can
compute
knowing the
n 0y[n] for
n 1and the
yinitial
[1] condition
input x[n] for
:

Block Diagram Representation


Advantages of block diagram representation
Easy to write down the computational Algorithm
by inspection
Easy to analyze the block diagram to determine
the explicit relation between the output and input
H/W requirements can easily determined
Variety of Representations can be developed
based on transfer function
Easy to manipulate a block diagram to derive
other equivalent block diagrams yielding
different computational algorithms

Basic Filter Types


There are basically two types of
Filters used:
Finite Impulse Response (FIR) Filter
Infinite Impulse Response (IIR) Filter

Basic FIR Digital Filter Structures


In the time-domain the input-output
relation of the above FIR filter is
M 1
given by
y (n) h(k ) x ( n k )
k 0

A causal FIR filter of order N is


1
characterized Mby
a transfer
function
n
H ( z ) h( n) z
H(z) given by n 0
-1

Basic FIR Digital Filter Structures


The unit sample response of the FIR
system is identical to the coefficients
{bk},
0 n M 1
bn ,
h( n)
otherwise
0,

Where M = Length of the FIR Filter System

FIR Filter Structures


Various types of forma available for
system realizations are;
Direct Form
Cascade Form
Lattice Form
Linear Phase Form
Poly Phase Form

Direct Form FIR Digital Filter Structures


An FIR filter of order N is
characterized by
N+1 coefficients
and, in general, require
N+1
multipliers and N two-input adders
Structures in which the multiplier
coefficients
are
precisely
the
coefficients of the transfer function
are called direct form structures

Direct Form FIR Digital Filter Structures


A direct form realization of an FIR
filter can be readily developed from
the convolution sum description as
indicated below for N = 4

Direct Form FIR Digital Filter Structures


The direct form structure shown on
the previous slide is also known as a
tapped delay line or a transversal
filter

Direct Form FIR Digital Filter Structures


The transpose of the direct form
structure shown earlier is indicated
below

Both direct form structures are


canonic with respect to delays

Cascade Form FIR Digital Filter


Structures
A higher-order FIR transfer function can also
be realized as a cascade of second-order
FIR sections and possibly a first-order
section
To this endMwe
express H(z) as
/2
H ( z ) ( 0 k 1k z 1 2 k z 2 )
k 1

Where M is even, 2k = 0
H(z) has odd number of real roots

Cascade Form FIR Digital Filter Structures

A cascade realization for N = 6 is shown


below

Each second-order section in the above


structure can also be realized in the
transposed direct form

Linear-Phase FIR Structures


The symmetry (or ant-symmetry)
property of a linear-phase FIR filter
can be exploited to reduce the
number of multipliers into almost
half of that in the direct form
implementations
Consider a length-7 Type 1 FIR
1
2
3
H ( z ) h[0function
] h[1]z with
h[2]z a symmetric
h[3]z
transfer
impulse hresponse:
[2]z 4 h[1]z 5 h[0]z 6

Linear-Phase FIR Structures


Rewriting H(z) in the form
6
1
5
H ( z ) h[0](1 z ) h[1]( z z )
2
4
3
h[2]( z z ) h[3]z
we obtain the realization shown
below

Linear-Phase FIR Structures


A similar decomposition can be applied
to a Type 2 FIR transfer function
For example, a length-8 Type 2 FIR
transfer function can be expressed as

H ( z ) h[0](1 z
h[2]( z

) h[1]( z
5

z ) h[3]( z

z
3

The corresponding realization is shown


on the next slide

Linear-Phase FIR Structures

Note:
The Type 1 linear-phase structure for
a length-7 FIR filter requires 4
multipliers, whereas a direct form
realization requires 7 multipliers

Linear-Phase FIR Structures


Note:
The Type 2 linear-phase structure for
a length-8 FIR filter requires 4
multipliers, whereas a direct form
realization requires 8 multipliers
Similar savings occurs in the
realization of Type 3 and Type 4
linear-phase FIR filters with
antisymmetric impulse responses

Lattice Structures
Used extensively in
Digital speech processing
Implementation of adaptive Filters

Consider a sequence of FIR filters


with system
H m ( z ) Am ( z ), m 0,1,2,..., M 1
Where

Am ( z ) 1 m (k ) z k ,

and A0(z)=1.

k 1

m 1

Lattice Structures

Lattice Structures

Basic IIR Digital Filter Structures


The causal IIR digital filters we are concerned with in
this course are characterized by a real rational transfer
function of z-1 or, equivalently by a constant coefficient
difference equation.
From the difference equation representation, it can be
seen that the realization of the causal IIR digital filters
requires some form of feedback.

IIR Filter Structures


Various kind of system realization
are:
Direct Form-I Structures
Direct-Form-II Structures
Signal Flow graph and Transposed
Structure
Cascade Form Structure
Parallel Form Structure
Lattice and Lattice Adder Structure

Basic IIR Digital Filter


Structures
An N-th order IIR digital transfer function is
characterized by 2N+1 unique coefficients,
and in general, requires 2N+1 multipliers
and 2N two-input adders for implementation
Direct form IIR filters: Filter structures in
which the multiplier coefficients are
precisely the coefficients of the transfer
function

Direct Form IIR Digital Filter


Structures
Consider for simplicity a 3rd-order IIR filter
with a transfer function

We can implement H(z) as a cascade of two


filter sections as shown on the next slide

Direct Form IIR Digital Filter


Structures
X
(z)

where

W
H1 ( z ) ( z )

H 2( z )

Y
(z)

Direct Form IIR Digital Filter


Structures
The filter section H ( z) can be seen to be
1
an FIR filter and can be realized as shown
below

Direct Form IIR Digital Filter


Structures
The time-domain representation of H ( z) is
2
given by

Realization of H 2
( z)
follows from the
above equation
and is shown on

Direct Form Structures

Direct form I

Direct form II

Direct Form Structures


Direct Form-I

Direct FormII

Multiplication

M+N+1

M+N+1

Addition

M+N

M+N

Memory
Locations

M+N+1

Max {M,N}

Note:
Direct-Form II reduces the memory requirements but not
used in practical applications.

Cascade Structures
Consider Transfer Function :

Various different cascade realizations of H(z) can be obtained


by different pole-zero polynomial pairings.
Usually, the polynomials are factored into a product of firstorder and second-order polynomials.
e.g. :

(1 1k z 1 2 k z 2 )
H ( z ) p[0]
,
1
2

k (1 1k z 2 k z )

Cascade Realization

Cascade realization of a IIR transfer function

Parallel Structure
Making use of the partial-fraction
expansion of the transfer function:
H ( z)
k

Qk ( z )
Dk ( z )

The parallel form I:

The parallel form II:

Parallel Structure

Parallel Structure

Aa Example of Parallel realization of a IIR transfer function

Transposed Structure

Lattice structure for an allpole IIR system

Lattice-ladder structure
for the realization of a
pole-zero system

Block Diagram Representation


y[0] d1 y[1] p0 x[0] p1x[1]
y[1] d1 y[0] .p0 x[1] p1x[0]
..
y[2] . d1 y[1] p0 x[2] p1x[1]
..
We can continue this calculation for
any value of the time index n we
desire

Block Diagram Representation


Each step of the calculation
requires a knowledge of the
previously calculated value of the
output sample (delayed value of
the output), the present value of
the input sample, and the previous
value of the input sample (delayed
value of the input)
As a result, the first-order
difference equation can be

Polyphase FIR Structures


The polyphase decomposition of H(z)
leads to a parallel form structure
To illustrate this approach, consider a
causal FIR transfer function H(z) with
N = 8:
H ( z ) h[0] h[1]z 1 h[2]z 2 h[3]z 3 h[4]z 4
h[5]z 5 h[6]z 6 h[7]z 7 h[8]z 8

Polyphase FIR Structures


H(z) can be expressed as a sum of
two terms, with one term
containing the even-indexed
coefficients and the other
2 odd-indexed
4
6
8
containing
the
H ( z ) (h[0] h[2]z h[4]z h[6]z h[8]z )
coefficients:
(h[1]z 1 h[3]z 3 h[5]z 5 h[7]z 7 )
2

(h[0] h[2]z h[4]z h[6]z h[8]z )


1
2
4
6
z (h[1] h[3]z h[5]z h[7]z )

Polyphase FIR Structures


By using the notation
1

E0 ( z ) h[0] h[2]z h[4]z h[6]z h[8]z


1
2
3
E1( z ) h[1] h[3]z h[5]z h[7]z
we can express H(z) as
2

H ( z ) E0 ( z ) z E1( z )

Polyphase FIR Structures


In a similar manner, by grouping the
terms in the original expression for
H(z), we can reexpress it in the form
3
1
3
2
3
H ( z ) E0 ( z ) z E1( z ) z E2 ( z )
where now
1

E0 ( z ) h[0] h[3]z h[6]z


E1( z ) h[1] h[4]z 1 h[7]z 2
1
2
E2 ( z ) h[2] h[5]z h[8]z

Polyphase FIR Structures


The decomposition of H(z) in the
form H ( z ) E ( z 2 ) z 1E ( z 2 )
0

or H ( z ) E0 ( z 3 ) z 1E1( z 3 ) z 2 E2 ( z 3 )
is more commonly known as the
polyphase decomposition

Polyphase FIR Structures


In the general case, an L-branch
polyphase decomposition of an FIR
transfer function of order N is of the
form H ( z ) L 1 z m Em ( z L )
m 0
where

Em ( z )

( N 1) / L

n 0

h[ Ln m]z

with h[n]=0 for n > N

Polyphase FIR Structures


Figures below show the 4-branch,
3-branch, and 2-branch
polyphase realization of a
transfer function H(z)

Note: The Eexpression


for
the
(z
)
m
polyphase components
are

Polyphase FIR Structures


L
The subfilters
in the
Em ( z )
polyphase realization of an FIR
transfer function are also FIR filters
and can be realized using any
methods described so far
However, to obtain a canonic
realization of the overall structure,
the delays in all subfilters must be
shared

Polyphase FIR Structures


Figure below shows a canonic
realization of a length-9 FIR
transfer function obtained using
delay sharing