Topik Khusus

Pengolahan Sinyal

Discrete-Time Filter Design

Sudirman Maliang
D41106007
Overview
One of the most powerful operations of
discerete-time signal processing (DSP)
is that of filtering.

There are two types of the discrete-

time filters:
 FIR (Finite Impulse Response) Filters.
 IIR (Infinite Impulse Response) Filters.
Filter  Digital
Masing-masing FIR dan IIR ini dapat melakukan hal yang serupa
dengan filter analog.

input analog  h(t)  output analog

input analog  A/D  h(n)  D/A  output analog

Filter analog mengambil input analog dan menghasilkan output

analog. Filter digital, dengan adanya perangkat pencuplikan dan
konverter, melakukan hal yang sama dengan filter analog.
FIR
(Finitive Impulse Response) Filter
IIR
(Infinitive Impulse Response) Filter
Merancang Filter Digital
Menentukan aplikasi / spesifikasi filter.
Menerapkan konsep-konsep secara
matematis difference equation or H(z)
Mengimplementasikan konsep yang telah
dibuat.

Ada dua metode untuk merancang filter FIR

yaitu
 Metode Windowing dan
 Metode Multiband dengan Band transisi.
Magnitude Specifications

1. Absolute spesifications:
Magnitude Response Function  |H(ej)|
2. Relative spesifications:
j
H
H (e j))max
( e
dB scale  20
dB scale 20log
log j
max 00
H
H((ee ))
j
Block Scheme
1+1
Passband
1
)

ripple
1-1
j
H(e

Transition Stopband
band rippel
-s
(a) 0 
0 p s π
0 
Rp
Decibels

As
(b
)

FIR filter specification : (a) Absolute (b) Relative

Rp is the passband ripple in dB

1  1
R p  20 log10
1  1
As is the stopband attenuation in dB
2
As  20 log10
1  1
Contoh 1:

Jika diketahui Rp = 0.25 dB dan As = 50 dB maka tentukanlah

1 dan 2

Jawab:

1  1
R p  0.25  20 log10  1  0.0144
1  1

2
As  50  20 log10   2  0.0032
1  1
Contoh 2:

Jika diketahui toleransi passband 1 = 0.01 dan toleransi stopband

2 =0.001 maka tentukanlah passband ripple dan stopband attenuation

Jawab:

1  1
R p  20 log10  0.1737 dB
1  1

2
As  20 log10  60 dB
1  1
Linear-Phase FIR Filters
Transfer function
M 1 M 1
H ( z)   h( n) z
n 0
n
z  ( M 1)
 h( n) z
n0
M 1 n

Frequency response function

M 1
H (e ) 
j
 h ( n )e
n 0
 j n
,     
Pembagian Fasa Linier dari h(n):

A. Symmetrical impulse response:

h(n) = h(M-1-n) n = 0, 1, 2, …, M-1

B. Antisymmetrical impulse response:

h(n) = - h(M-1-n) n = 0, 1, 2, …, M-1

The length of the impulse response of the FIR

filter (M) can be even or odd.

13
 A.1 Symmetrical Impulse Response, M: Even
h( n) h(7)=h(8) M  16

h(2)=h(13)

h(1)=h(14)

h(0)=h(15)

n
14
 A.2 Symmetrical Impulse Response, M: Odd
h( n) M  15
“h(7)=h(7)”

h(6)=h(8)

h(1)=h(13)

h(0)=h(14)

n
15
B.1 Antisymmetrical Impulse Response, M: Even
h( n) M  16

h(1)=-h(14)

h(0)=-h(15)

h(7)=-h(8)
n
16
B.2 Antisymmetrical Impulse Response, M: Odd
h( n) M  17

h(1)=-h(15)

h(8)=-h(8)=0
h(0)=-h(16)

h(7)=-h(9)
n
17
Contoh 1
Jika impulse response h(n) = {1,1,1}
maka tentukan dan gambarkanlah
response frekuensinya.
Jawab:
Respon frekuensi:
2
H (e j )   h(n)e  jn  1  e  j  e  j 2
n 0

 {e j  1  e  j }e  j  {1  2 cos }e  j

Magnitude and phase responses:

H (e j )  1  2 cos  , 0 
   , 0    2 / 3
H (e )  j

   , 2 / 3    
 Magnitude Resonanse Amplitude Response
|H|

Hr
0 0

0 2/3 1 0 2/3 1
Frequency in pi units Frequency in pi units
Contoh 2:
Jika h(n) = {-4,1,-1,-2,5,6,5,-2,-1,1,-4}.
Tentukanlah Amplitude Response Hr()
dan zeroes dari H(z)
Contoh 3:
Jika h(n) = {-4,1,-1,-2,5,6,6,5,-2,-1,1,-4}.
Tentukanlah Amplitude Response Hr()
dan zeroes dari H(z)
Impulse responce

% IIR filter coeffs

a=[1 -0.8];
b=[1];
% Plot h(n)
impz(b,a,30);

1
H ( z) 
1  0.8  z 1 h(n)  0.8n  u (n)
1st order IIR filter: stability

a=0.8

a=0.8

a=1.2

||hh((nn))| nn
|||aa11 || If a  1 then h(n)  
Amplitude-frequency response

1 2
r=0.9;
a1=-2*r*0.5;
a2=r*r;

3
1) r=0.2; cos() = 0.5
2) r=0.9; cos() = 0.5
3) r=0.9; cos() = 0.5
Impulse responce function

% IIR filter coeffs

a=[1 -0.8 + 0.9];
b=[1 0.2];
% Responce func
impz(b,a,100);

sin  (n  1) 
h(n)  r n
u ( n)
sin( )
Perbandingan filter FIR dan IIR
FIR IIR

M1 M N
y (n)   bk x(n  k )   ak y (n  k )
y ( n)   b x( n  k )
k  M 2
k k 0 k 1
M

M1 b z k
k

k H ( z) 
k k 0
H ( z)  b z N
k  M 2 1   ak z  k
k 1