Indonesian
Journal
of
Electrical
Engineering
and
Computer
Science
V
ol.
14,
No
.
3,
J
une
2019,
pp
.
1272
๎
1280
DOI:
10.11591/ijeecs
.v14.i3.pp
1272-1280.
๎
1272
On
the
P
erf
ormance
of
Coded
Copper
Wire
Systems
f
or
G-fast
Comm
unications
Ali
J
.
Al-Asker
y
*
and
Ali
Mehdi
Hammadi
Depar
tment
of
Control
and
A
utomation
T
echnical
Engineer
ing
Electr
ical
Engineer
ing
T
echnical
College/
Middle
T
echnical
Univ
ersity
Baghdad,
Ir
aq/Phone:
+964
7817569229
*
e-mail:
alialaskar
y1@y
ahoo
.com
Ar
tic
le
Inf
o
ABSTRA
CT
Ar
tic
le
histor
y
:
Receiv
ed
Dec
25,
2018
Re
vised
J
an
21,
2019
Accepted
F
eb
22,
2019
This
paper
aims
to
in
v
estig
ate
the
perf
or
mance
of
diff
erent
chan-
nel
coding
techniques
o
v
er
copper
wire
channels
.
First,
Chen
model
is
used
to
sim
ulate
the
copper
wire
channel
model,
then
the
perf
or
mance
of
LDPC
coded
system
with
D
VB-S2
standard
is
compared
to
the
perf
or
mance
of
turbo
coded,
and
con
v
olu-
tional
coded
systems
o
v
er
wired
channels
.
In
addition,
the
per-
f
or
mance
of
the
systems
under
consider
ation
is
fur
ther
in
v
esti-
gated
in
the
presence
of
impulsiv
e
noise
(IN)
with
IN
cancellation
technique
.
Matlab
sim
ulations
are
utiliz
ed
to
sim
ulate
these
sys-
tems
f
or
diff
erent
quadr
ature
amplitude
modulations
(QAM).
The
results
ha
v
e
sho
wn
that
LDPC
coded
system
with
large
b
loc
k
length
ha
v
e
outperf
or
ms
the
perf
or
mance
of
the
considered
cod-
ing
techniques
.
K
e
yw
or
ds
:
Chen
model
copper
wired
channel
Coding
techniques
Cop
yr
ight
2019
Institute
of
Adv
anced
Engineer
ing
and
Science
.
All
r
ights
reser
v
ed.
Corresponding
A
uthor:
Dr
.Ali
Al-Ask
er
y
Electr
ical
Engineer
ing
T
echnical
College
Middle
T
echnical
Univ
ersity
Mobile:
009647817569229
Email:
alialaskar
y1@y
ahoo
.com
1.
Intr
oduction
Digital
subscr
iber
line
(DSL)
is
a
technology
f
or
tr
ansmitting
and
receiving
the
data
o
v
er
copper
wire
lines
.
Due
to
the
e
xpensiv
e
cost
of
replacing
the
infr
astr
ucture
of
the
twisted
pair
copper
wire
,
DSL
has
been
e
v
olv
ed
o
v
er
the
y
ears
to
suppor
t
the
requiremen
ts
of
v
er
y
high
bit
r
ate
tr
ansmission
o
v
er
these
lines
.
The
G.f
ast
is
a
standardization
f
or
DSL
technolog
y
that
suppor
ts
up
to
106
MHz
broad
spectr
um
which
is
e
xpected
to
be
e
xpanded
to
212
MHz
[1].
In
addition,
discrete
m
ultitone
(DMT)
tr
ansmission
con
v
er
ts
the
tr
ansmitted
data
into
a
lo
w
r
ate
m
ulti-carr
ier
sequence
,
which
mak
es
it
more
con
v
enient
and
ef๏ฌcient
in
the
data
tr
ansmission
[2].
Channel
coding
has
been
used
widely
to
impro
v
e
the
tr
ansmission
reliability
and
to
en-
hance
the
bite
error
r
ate
(BER)
perf
or
mance
[3โ5].
In
[6],
turbo
coded
ADSL-DMT
system
with
64
quadr
ature
amplitude
modulation
(QAM)
is
proposed
to
impro
v
e
the
system
perf
or
mance
.
The
perf
or
mance
has
been
compared
to
the
trellis
coded
modulation
(TCM)
to
sho
w
the
impro
v
ement
Jour
nal
homepage:
http://iaescore
.com/jour
nals/inde
x.php/ijeecs
Evaluation Warning : The document was created with Spire.PDF for Python.
Indonesian
J
Elec
Eng
&
Comp
Sci
ISSN:
2502-4752
๎
1273
Binar
y
Data
Input
Channel
Encoding
DMT
Modula
tion
Copper
Channel
DMT
Demod
Channel
Decodi
ng
Data
Output
A
WGN+IN
Linear
Detecto
r
QAM
Modula
tion
LLR
IN
Cancel
lat
ion
Figure
1.
System
model
f
or
the
copper
line
wired
system
in
the
coding
gain.
In
[7],
a
diff
erential
e
v
olution
algor
ithm
aided
turbo
channel
estimation
(DEA-
CE)
has
been
proposed
f
or
the
G.f
ast
systems
to
achie
v
e
near
Shannon
capacity
limit
with
a
reduced
comple
xity
.
In
addition,
the
perf
or
mance
of
G.f
ast
coded
copper
wire
channel
has
been
tak
en
into
consider
ation
in
[8]
in
the
presence
of
impulsiv
e
noise
(IN),
in
which,
the
coded
system
has
been
designed
as
a
cascaded
trellis
code
as
an
outer
code
with
non
binar
y
Reed-Solomon
(RS)
code
as
an
inter
nal
code
.
Lo
w-density
par
ity-chec
k
(LDPC)
coding
has
been
used
in
[9]
as
an
inner
code
f
or
DSL
system
with
DMT
mo
dulation,
while
the
outer
code
w
as
selected
as
RS
code
with
Galois
๏ฌeld
GF(
2
8
).
The
code-w
ord
length
w
as
selected
betw
een
529
and
4489
bits
with
m
ultiple
M-QAM
scheme
,
while
the
IN
eff
ect
on
the
perf
or
mance
has
not
been
considered.
An
algor
ithm
to
cal-
culate
the
IN
statist
ics
has
been
proposed
in
[10]
to
calculate
an
accur
ate
log-lik
elihood
r
atios
(LLRs)
f
or
the
LDPC
coded
DSL
system.
DMT
modulation
has
been
used
in
this
paper
with
a
precoding
technique
and
the
results
ha
v
e
been
v
er
i๏ฌed
utilizing
the
EXIT
char
t.
On
the
other
hand,
a
gener
ator
matr
ix
with
lo
w
density
has
been
proposed
in
[11]
which
is
more
con
v
enient
in
the
video
tr
ansmitting
and
receiving.
In
addition,
a
fur
ther
e
xtension
to
the
proposed
matr
ix
has
been
demonstr
ated
based
on
rearr
anging
the
matr
ix
columns
which
result
in
a
better
imm
unity
to
the
b
urst
errors
.
Se
v
er
al
techniques
to
mitigate
the
IN
ha
v
e
been
tak
en
into
consider
ation
including
clip-
ping
and
b
lanking.
In
[12],
an
IN
mitigation
technique
has
been
proposed
to
remo
v
e
the
eff
ect
of
the
IN
model
of
the
English
and
Ger
man
wired
systems
.
A
compar
ison
of
diff
erent
IN
mitigation
techniques
ha
v
e
also
been
pro
vided
to
sho
w
the
v
alidity
of
the
proposed
technique
.
In
order
to
enab
le
the
DSL
systems
to
suppor
t
wide
spectr
um
and
suppor
t
high
quality
video
streaming,
a
po
w
erful
channel
coding
is
required
that
imp
ro
v
e
s
the
BER
perf
or
mance
and
increase
the
tr
ansmission
reliability
.
LDPC
coding
with
digital
video
broadcasting-satellite-second
gener
ation
(D
VB-S2)
is
an
LDPC
sta
ndard
with
64800
bits
code-w
ord
with
irregular
encoding
and
decoding
procedure
has
near
Shannon
limit
compared
to
other
channel
coding
techniques
[13,
14].
Theref
ore
,
the
contr
ib
ution
of
this
paper
is
to
in
v
estigate
the
perf
or
mance
of
LDPC
coded
copper
wire
with
D
VB-S2
standard
and
compar
ing
the
system
perf
or
mance
to
oth
er
channel
cod-
ing
techniques
such
as
turbo
coding
and
con
v
olutional
coding
with
diff
erent
M-QAM
schemes
.
The
perf
or
mance
of
the
techniques
under
consider
ation
is
in
v
estigated
o
v
er
G.f
ast
system
appli-
cations
with
DMT
tr
ansmission
and
Middleton
class
A
model
f
or
IN
representation.
Finally
,
the
IN
eff
ect
is
mitigated
from
the
receiv
ed
signal
utilizing
clipping
technique
.
This
paper
is
organiz
ed
as
f
ollo
ws
,
in
Section
2.
the
system
model
is
presented
in
details
,
including
the
channel
model
of
the
wired
system,
Middleton
class
A
distr
ib
ution
as
IN
model,
the
LLR
equations
are
presented
as
par
t
of
computing
the
soft
decoding
f
or
the
receiv
ed
coded
signal,
and
the
sum-product
algor
ithm
(SP
A)
f
or
LDPC
decoder
with
D
VB-S2
standard.
The
sim
ulation
and
results
are
discussed
in
Section
3.
While
the
conclusions
are
dr
a
wn
in
Section
4.
On
the
P
erf
or
mance
of
Coded
Copper
Wire
(Ali
J
.
Al-Ask
er
y)
Evaluation Warning : The document was created with Spire.PDF for Python.
1274
๎
ISSN:
2502-4752
2.
System
Model
This
paper
considers
the
G.f
ast
system
model
that
is
illustr
ated
in
Figure
1.
The
binar
y
data
B
is
๏ฌrst
gener
ated
and
encoded
with
the
rele
v
ant
channel
encoder
to
gener
ate
the
code-
w
ord
C
.
Then
the
code-w
ord
is
modulated
wit
h
M-QAM
and
DMT
modulator
to
produce
the
tr
ansmitted
data
D
that
tr
a
v
els
o
v
er
the
copper
line
wire
that
is
represented
as
Chen
mo
del
[15].
At
the
receiv
er
,
the
additiv
e
white
Gaussian
noise
(A
WGN)
and
the
IN
is
added
to
the
receiv
ed
signal
such
that
Y
=
H
๎
D
+
N
+
I
;
(1)
where
,
Y
2
C
1
๎
K
and
D
2
C
1
๎
K
are
the
receiv
ed
and
tr
an
s
m
itted
v
ectors
with
K
as
the
length
of
the
tr
ansmitted
signal.
H
2
C
1
๎
K
and
N
2
C
1
๎
K
are
the
channel
and
the
noise
par
ame-
ter
v
ectors
,
respectiv
ely
,
and
I
2
C
1
๎
K
is
the
IN
v
ector
.
The
receiv
ed
signal
is
equaliz
ed
with
minim
um-mean
squared-error
(MMSE)
detector
such
that,
W
mmse
=
H
๎
=
(
j
H
2
j
+
๎
2
=E
s
)
(2)
where
H
๎
is
the
comple
x
conjugate
of
the
v
ector
H
,
and
E
s
is
the
symbol
energy
.
F
ollo
wing
signal
equalization,
IN
compensation,
and
DMT
demodulation,
the
signal
~
Q
is
utiliz
ed
to
calculate
the
log-lik
elihood
r
atio
(LLR)
f
or
the
soft
decoding
of
the
selected
f
orw
ard
error
correcting
code
(FECC).
Finally
,
the
decoded
signal
~
B
is
compared
to
the
binar
y
gener
ated
sequence
at
the
tr
an
smitter
to
calculate
the
BER.
It
is
w
or
th
mention
that
the
channel
is
assumed
to
be
kno
wn
at
the
receiv
er
.
2.1.
Channel
Model
As
mentioned
bef
ore
,
the
channel
model
of
this
paper
is
based
on
Chen
model
to
rep-
resent
the
copper
wire
model,
which
is
suitab
le
f
or
up
to
200
MHz
frequency
r
ange
[16].
The
atten
uation
constant
๎
and
the
phase
constant
๎
are
calculated
with
Chen
model
constants
k
1
,
k
2
,
and
k
3
,
and
f
or
the
frequency
f
as
๎
(
f
)
=
k
1
p
f
+
k
2
f
;
๎
(
f
)
=
k
3
f
:
(3)
The
channel
par
ameters
can
be
wr
itten
using
(3)
and
the
cab
le
length
L
f
or
a
speci๏ฌc
frequency
as
H
=
e
๎
L๎
(
f
)
;
(4)
where
,
the
propagation
constant
๎
(
f
)
=
๎
(
f
)
+
j
๎
(
f
)
and
L
is
measured
in
meter
.
2.2.
Impulsive
Noise
Model
Middleton
class
A
distr
ib
ution
[17,
18]
is
used
in
this
paper
to
sim
ulate
the
IN
of
the
copper
wire
line
comm
unication.
The
probability
density
function
(PDF)
of
this
model
can
be
wr
itten
as
p
(
๎
)
=
1
X
๎
=0
e
๎
A
A
๎
๎
!
๎
e
(
๎
๎
2
2
๎
2
)
p
2
๎
๎
2
๎
;
(5)
where
the
impulsiv
e
inde
x
is
A
,
and
the
smallest
v
alue
of
A
will
result
in
a
v
er
y
high
peak,
while
the
distr
ib
ution
approaches
Gaussian
distr
ib
ution
when
A
is
v
er
y
high.
The
noise
po
w
er
c
a
n
be
calculated
as
๎
2
=
๎
2
G
+
๎
2
I
,
where
๎
2
G
and
๎
2
I
are
the
v
ar
iance
of
the
A
WGN
and
the
IN,
respectiv
ely
.
The
v
ar
iance
๎
2
๎
can
be
calculated
as
,
๎
2
๎
=
๎
2
๎
๎
A
+
๎
1
+
๎
๎
;
(6)
and
๎
=
๎
2
G
=๎
2
I
.
Zhidk
o
v
compensation
algor
ithm
is
used
in
this
paper
to
remo
v
e
the
IN
eff
ect
as
sho
wn
in
[8].
Indonesian
J
Elec
Eng
&
Comp
Sci
V
ol.
14,
No
.
3,
J
une
2019
:
1272
๎
1280
Evaluation Warning : The document was created with Spire.PDF for Python.
Indonesian
J
Elec
Eng
&
Comp
Sci
ISSN:
2502-4752
๎
1275
Figure
2.
Constellation
map
of
the
16-QAM
scheme
.
2.3.
LLR
Calculations
The
LLRs
f
or
the
system
under
consider
ation
can
be
calculated
based
on
the
QAM
con-
stellation
inde
x
such
that
[14,
19],
2.3.1.
4-QAM
Sc
heme
The
LLR
of
this
constellation
can
be
calculated
as
,
L
0
=
ln
๎
p
(
N
j
D
I
=1
)
p
(
N
j
D
I
=
๎
1
)
๎
;
L
1
=
ln
๎
p
(
N
j
D
Q
=1
)
p
(
N
j
D
Q
=
๎
1
)
๎
;
(7)
where
L
0
;
L
1
are
the
LLR
of
the
tw
o
bits
B
0
;
B
1
,
respectiv
ely
,
and
p
(
N
)
represents
the
PDF
of
the
A
WGN.
The
supe
rscr
ipts
I
and
Q
are
to
indicate
the
in-phase
and
the
quadr
ature
par
ts
of
the
signal.
2.3.2.
16-QAM
Sc
heme
The
LLR
of
the
16-QAM
constellation
presented
in
Figure
2
can
be
calculated
using,
L
0
=
ln
๎
p
(
N
I
j
D
I
=1
)
p
(
N
I
j
D
I
=
๎
1
)
๎
;
L
1
=
min
๎
ln
๎
p
(
N
I
j
D
I
=
๎
1
)
p
(
N
I
j
D
I
=
๎
3
)
๎
;
ln
๎
p
(
N
I
j
D
I
=1
)
p
(
N
I
j
D
I
=3
)
๎๎
;
L
2
=
ln
๎
p
(
N
Q
j
D
Q
=1
)
p
(
N
Q
j
D
Q
=
๎
1
)
๎
;
L
3
=
min
๎
ln
๎
p
(
N
Q
j
D
Q
=
๎
1
)
p
(
N
Q
j
D
Q
=
๎
3
)
๎
;
ln
๎
p
(
N
Q
j
D
Q
=1
)
p
(
N
Q
j
D
Q
=3
)
๎๎
;
(8)
On
the
P
erf
or
mance
of
Coded
Copper
Wire
(Ali
J
.
Al-Ask
er
y)
Evaluation Warning : The document was created with Spire.PDF for Python.
1276
๎
ISSN:
2502-4752
where
L
0
;
L
1
;
L
2
;
L
3
are
the
LLR
of
the
4
bits
B
0
;
B
1
;
B
2
:B
3
,
respectiv
ely
.
0
5
10
15
20
10
โ6
10
โ5
10
โ4
10
โ3
10
โ2
10
โ1
10
0
E
b
/
No
(
d
B
)
B
E
R
L
D
P
C
R
a
t
e
1
/
2
C
o
n
v
o
l
u
t
i
o
n
a
l
R
a
t
e
1
/
2
T
u
r
b
o
R
a
t
e
1
/
2
Figure
3.
Sim
ulation
results
at
4-QAM
scheme
f
or
the
turbo
coded,
con
v
olutional
coded
and
LDPC
coded
systems
.
2.4.
LDPC
Codes
LDPC
code
with
D
VB-S2
standard
and
1/2
r
ate
has
a
b
loc
k
length
of
N
=
64800
bits
with
binar
y
data
b
loc
k
of
k
=
32400
bits
.
In
addition,
irregular
LDPC
codes
has
an
impro
v
ed
perf
or
mance
compared
to
regular
codes
in
which,
the
BER
perf
or
mance
app
roaches
the
Shannon
capacity
limit
closely
[20,
21].
The
implementation
of
the
LDPC
decoder
in
this
w
or
k
is
based
on
the
ef๏ฌcient
sum-product
algor
ithm
(SP
A)
that
ha
v
e
been
proposed
in
[22].
First,
the
initialization
to
the
symbol
node
is
the
LLR
equations
that
ha
v
e
been
calculated
in
Equations
(7)
and
(8)
such
that,
L
(
q
n
!
m
)
=
L
k
(9)
Ne
xt,
the
chec
k
to
node
update
f
or
each
m
and
f
or
n
2
N
(
m
)
is
perf
or
med
using,
L
(
r
m
!
n
)
=
2
tanh
๎
1
๎
Y
n
0
2N
(
m
)
n
tanh[
1
2
L
(
q
n
0
!
m
)
๎
:
(10)
The
third
step
in
v
olv
es
updating
the
node
to
chec
k
f
or
n
and
f
or
m
2
M
(
n
)
utilizing,
L
(
q
n
!
m
)
=
L
k
+
X
m
0
2M
(
n
)
m
L
(
r
m
0
!
n
)
:
(11)
Then,
the
decision
on
the
decoded
bits
can
be
made
b
y
calculating
the
f
ollo
wing
equation,
L
(
u
k
)
=
L
k
+
X
m
0
2M
(
n
)
L
(
r
m
0
!
n
)
;
(12)
such
that
u
k
=
0
f
or
L
(
u
k
)
๎
0
and
u
k
=
1
otherwise
.
These
steps
will
be
repeated
se
v
er
al
times
until
the
condition
H
^
u
=
0
is
achie
v
ed
or
the
maxim
um
n
umber
of
iter
ations
is
reached.
Indonesian
J
Elec
Eng
&
Comp
Sci
V
ol.
14,
No
.
3,
J
une
2019
:
1272
๎
1280
Evaluation Warning : The document was created with Spire.PDF for Python.
Indonesian
J
Elec
Eng
&
Comp
Sci
ISSN:
2502-4752
๎
1277
0
5
10
15
20
10
โ5
10
โ4
10
โ3
10
โ2
10
โ1
10
0
E
b
/
No
(
d
B
)
B
E
R
L
D
P
C
R
a
t
e
1
/
2
C
o
n
v
o
l
u
t
i
o
n
a
l
R
a
t
e
1
/
2
T
u
r
b
o
R
a
t
e
1
/
2
Figure
4.
Sim
ulation
results
at
16-QAM
scheme
f
or
the
turbo
coded,
con
v
olutional
coded
and
LDPC
coded
systems
.
3.
Sim
ulation
and
Results
This
section
is
used
to
illustr
ate
and
discuss
the
results
obtained
from
the
sim
ulations
of
the
wired
copper
system.
The
channel
length
w
as
selected
such
that
L
=
100
m,
which
is
substituted
in
(4).
T
w
o
QAM
schemes
ha
v
e
bee
n
used
in
this
sim
ulation
which
are
4-QAM
and
16-QAM,
respectiv
ely
.
MMSE
detector
ha
v
e
been
used
to
equaliz
e
the
receiv
ed
signal
to
remo
v
e
the
eff
ect
of
the
channel.
The
turbo
decoder
w
as
l
og
๎
M
AP
type
with
1
=
2
r
ate
and
r
andomly
inter
lea
v
ed
with
b
loc
k
length
4096
bit.
The
eIRA
LDPC
decoder
with
r
ate
1
=
2
has
a
length
of
64800
bits
.
In
Figure
3,
LDPC
coded
system
with
4-QAM
modulation
outperf
or
ms
the
perf
or
mance
of
both
con
v
olutional
coded
and
turbo
coded
systems
at
lo
w
signal
to
noise
r
atio
(SNR).
It
is
obser
v
ed
that
there
is
an
impro
v
ement
of
2.5
dB
in
the
perf
or
mance
compared
to
the
turbo
coded
system
at
10
๎
4
BER,
while
the
impro
v
ement
w
as
almost
5
dB
compared
to
the
con
v
olutional
coded
system.
In
Figure
4,
it
is
obser
v
ed
th
at
the
perf
or
mance
of
the
LDPC
coded
system
ha
v
e
better
perf
or
mance
compared
to
the
con
v
olutional
coded
and
the
turbo
coded
systems
.
The
impro
v
e-
ment
in
the
BER
w
as
2
dB
compared
to
the
turbo
coded
while
it
w
as
4
db
with
respect
to
the
con
v
olutional
coded
system
at
10
๎
4
of
BER.
In
Figures
5
and
6,
the
eff
ect
of
the
IN
on
the
perf
or
man
c
e
of
the
coded
systems
can
be
obser
v
ed
such
that
a
deg
r
adation
in
the
BER
perf
or
mance
of
almost
10
dB
is
achie
v
ed
f
or
the
LDPC
coded
system
at
4-QAM
and
16-QAM
schemes
.
Ho
w
e
v
er
,
the
BER
deg
r
adation
w
as
13
dB
f
or
the
turbo
coded
and
con
v
olutional
coded
systems
.
The
IN
compensation
system
ha
v
e
im-
pro
v
ed
the
perf
or
mance
of
the
LDPC
and
turbo
coded
systems
with
4-QAM
scheme
b
y
6
dB
while
the
impro
v
ement
w
as
8
dB
f
or
the
con
v
olutional
coded
system.
Ho
w
e
v
er
,
at
16-QAM
scheme
,
the
impro
v
ement
f
or
all
the
coded
systems
w
as
4
dB
.
On
the
P
erf
or
mance
of
Coded
Copper
Wire
(Ali
J
.
Al-Ask
er
y)
Evaluation Warning : The document was created with Spire.PDF for Python.
1278
๎
ISSN:
2502-4752
0
5
10
15
20
25
30
10
โ6
10
โ5
10
โ4
10
โ3
10
โ2
10
โ1
10
0
E
b
/
No
(
d
B
)
B
E
R
L
D
P
C
+
I
N
C
a
n
c
e
l
l
a
t
i
o
n
L
D
P
C
+
I
N
C
o
n
v
o
l
u
t
i
o
n
a
l
+
I
N
C
a
n
c
e
l
l
a
t
i
o
n
C
o
n
v
o
l
u
t
i
o
n
a
l
+
I
N
T
u
r
b
o
R
a
t
e
+
I
N
C
a
n
c
e
l
l
a
t
i
o
n
T
u
r
b
o
+
I
N
Figure
5.
Sim
ulation
results
at
4-QAM
scheme
f
or
the
turbo
coded,
con
v
olutional
coded
and
LDPC
coded
systems
in
the
presence
of
IN.
0
5
10
15
20
25
30
35
10
โ6
10
โ5
10
โ4
10
โ3
10
โ2
10
โ1
10
0
E
b
/
No
(
d
B
)
B
E
R
L
D
P
C
+
I
N
C
a
n
c
e
l
l
a
t
i
o
n
L
D
P
C
+
I
N
C
o
n
v
o
l
u
t
i
o
n
a
l
+
I
N
C
a
n
c
e
l
l
a
t
i
o
n
C
o
n
v
o
l
u
t
i
o
n
a
l
+
I
N
T
u
r
b
o
+
I
N
C
a
n
c
e
l
l
a
t
i
o
n
T
u
r
b
o
+
I
N
Figure
6.
Sim
ulation
results
at
16-QAM
scheme
f
or
the
turbo
coded,
con
v
olutional
coded
and
LDPC
coded
systems
in
the
presence
of
IN.
Indonesian
J
Elec
Eng
&
Comp
Sci
V
ol.
14,
No
.
3,
J
une
2019
:
1272
๎
1280
Evaluation Warning : The document was created with Spire.PDF for Python.
Indonesian
J
Elec
Eng
&
Comp
Sci
ISSN:
2502-4752
๎
1279
4.
Conc
lusion
In
conclusion,
the
perf
or
mance
of
coded
wired
line
systems
ha
v
e
been
e
xamined
with
three
channel
coding
t
echniques
and
tw
o
modulation
inde
x
es
.
The
results
ha
v
e
sho
wn
that
LDPC
coded
systems
outperf
or
ms
the
other
coding
techniques
and
especially
at
high
SNR
v
alues
.
It
is
also
obser
v
ed
that
the
w
ater
f
all
f
or
both
iter
ativ
e
codes
are
v
er
y
steep
,
while
the
con
v
olutional
coded
system
has
smoot
h
perf
or
mance
cur
v
e
o
v
er
the
SNR
r
ange
due
to
their
super
ior
abilit
y
to
detect
and
correct
the
errors
.
The
BER
deg
r
adation
resulting
from
including
the
IN
ha
v
e
been
e
xamined
with
Zhidk
o
v
compensation
algor
ithm
as
an
IN
mitigation
technique
.
As
a
result,
the
perf
or
mance
of
the
coded
systems
ha
v
e
deg
r
aded
b
y
almost
6
dB
e
v
en
after
remo
ving
the
eff
ect
of
the
IN.
Ho
w
e
v
er
,
the
perf
or
mance
of
LDPC
coded
systems
remains
on
the
lead
compared
to
the
considered
coding
techniques
.
Ref
erences
[1]
P
.
Odling,
T
.
Magesacher
,
S
.
Host,
P
.
O
.
Borjesson,
M.
Berg,
and
E.
Areizaga,
โThe
f
our
th
gener
ation
broadband
concept,
โ
IEEE
Comm
un.
Mag.
,
v
ol.
47,
no
.
1,
pp
.
62โ69,
2009.
[2]
B
.
R.
Saltzberg,
โCompar
ison
of
single-carr
ier
and
m
ultitone
digital
modulation
f
or
ADSL
applications,
โ
IEEE
Comm
un.
Mag.
,
v
ol.
36,
no
.
11,
pp
.
114โ121,
1998.
[3]
N.
A.
Khair
i,
A.
B
.
J
ambek,
and
R.
C
.
Ismail,
โP
erf
or
mance
e
v
aluation
of
ar
ithmetic
coding
data
compression
f
or
inter
net
of
things
applications
,
โ
Indonesian
Jour
nal
of
Electr
ical
Engi-
neer
ing
and
Computer
Science
,
v
ol.
13,
no
.
2,
pp
.
591โ597,
2019.
[4]
N.
T
elagam,
S
.
Lakshmi,
and
K.
Nehr
u,
โBER
analysis
of
concatenated
le
v
els
of
encoding
in
GFDM
system
using
lab
vie
w,
โ
Indonesian
Jour
nal
of
Electr
ical
Engineer
ing
and
Computer
Science
,
v
ol.
14,
no
.
1,
pp
.
80โ91,
2019.
[5]
A.
Haff
ane
,
A.
H.
M.
Kheli๏ฌ,
and
B
.
Kadr
i,
โCon
v
ergence
analysis
of
the
unpunctured
turbo
trellis-coded
modulation
(UTTCM),
โ
Indonesian
Jour
nal
of
Electr
ical
Engineer
ing
and
Com-
puter
Science
,
v
ol.
13,
no
.
2,
pp
.
447โ452,
2019.
[6]
L.
Zhang
and
A.
Y
ongacoglu,
โT
urbo
coding
f
or
tr
ansmission
o
v
er
ADSL,
โ
in
Proc.
Int.
Conf
.
Comm
un.
T
echnol.
WCC-ICCT2000.
,
v
ol.
1.
IEEE,
2000,
pp
.
124โ131.
[7]
J
.
Zhang,
S
.
Chen
,
R.
Zhang,
A.
F
.
A.
Ra
wi,
and
L.
Hanz
o
,
โDiff
erential
Ev
olution
Algor
ithm
Aided
T
urbo
Channel
Estimation
and
Multi-User
Detection
f
or
G.
F
ast
Systems
in
the
Pres-
ence
of
FEXT,
โ
IEEE
Access
,
v
ol.
6,
pp
.
33
111โ33
128,
2018.
[8]
I.
Al-Neami,
C
.
T
.
Healy
,
M.
Johnston,
and
C
.
Tsimenidis
,
โIn
v
estigat
ion
into
impulsiv
e
noise
techniques
f
or
a
G.
F
AST
system,
โ
in
Proc.
11th
Int.
Sympo
.
Comm
un.
Syst.,
Net.
&
Digital
Signal
Process
.
(CSNDSP)
.
IEEE,
2018,
pp
.
1โ5.
[9]
E.
Elefther
iou
and
S
.
Olcer
,
โLo
w-density
par
ity-chec
k
codes
f
or
digital
subscr
iber
lines,
โ
in
Proc.
IEEE
Int.
Conf
.
Comm
un.
,
v
ol.
3.
IEEE,
2002,
pp
.
1752โ1757.
[10]
T
.
Bai,
C
.
Xu,
R.
Zhang,
A.
F
.
Al
Ra
wi,
and
L.
Hanz
o
,
โJoint
Impulsiv
e
Noise
Estimation
and
Data
Detection
Conceiv
ed
f
or
LDPC-Coded
DMT
-Based
DSL
Systems,
โ
IEEE
Access
,
v
ol.
5,
pp
.
23
133โ23
145,
2017.
[11]
L.
Al-Jobour
i,
F
.
Casu,
M.
Fleur
y
,
and
J
.
Cabrer
a,
โVideo
o
v
er
DSL
with
LDG
M
Codes
f
or
Inter
activ
e
Applications,
โ
Computers
,
v
ol.
5,
no
.
2,
pp
.
1โ18,
2016.
[12]
R.
F
antacci,
A.
T
ani,
and
D
.
T
archi,
โImpulse
noise
mitigation
techniques
f
or
xDSL
systems
in
a
real
en
vironment,
โ
IEEE
T
r
ans
.
Consum.
Electron.
,
v
ol.
56,
no
.
4,
2010.
[13]
R.
A.
Carr
asco
and
M.
Johnston,
Non-binar
y
error
control
coding
f
or
wireless
comm
unication
and
data
stor
age
.
John
Wile
y
&
Sons
,
2008.
[14]
A.
J
.
Al-Ask
er
y
,
C
.
C
.
Tsimenidis
,
S
.
Boussakta,
and
J
.
A.
Chambers
,
โImpro
v
ed
coded
mas-
siv
e
MIMO
OFDM
detection
using
LLRs
der
iv
ed
from
comple
x
r
atio
distr
ib
utions
,
โ
in
Proc.
20th
IEEE
Int.
W
or
kshop
Comput.
Aided
Modelling
Design
of
Comm
un.
Links
Net.,
CAMAD
2015,
Guildf
ord,
United
Kingdom,
September
7-9,
2015
,
2015,
pp
.
64โ68.
[15]
D
.
Acatauassu,
S
.
Host,
C
.
Lu,
M.
Berg,
A.
Klautau,
a
nd
P
.
O
.
Borjesson,
โSimple
and
causal
twisted-pair
channel
models
f
or
G.f
ast
systems
,
โ
in
Proc.
IE
EE
Global
Comm
un.
Conf
.
(GLOBECOM)
.
IEEE,
2013,
pp
.
2834โ2839.
[16]
T
.
Magesacher
,
J
.
R.
i
Riu,
M.
J
ak
o
vlje
vi
ยด
c,
M.
Loiola,
P
.
ยจ
Odling,
and
P
.
O
.
B
ยจ
orjesson,
โLimits
On
the
P
erf
or
mance
of
Coded
Copper
Wire
(Ali
J
.
Al-Ask
er
y)
Evaluation Warning : The document was created with Spire.PDF for Python.
1280
๎
ISSN:
2502-4752
of
ultr
a-wideband
comm
unication
o
v
er
copper
,
โ
in
Proc.
Int.
Conf
.
Comm
un.
T
echnol.
,
2006,
pp
.
1โ4.
[17]
D
.
Middleton,
โStatistical-ph
ysical
models
of
electromagnetic
interf
erence
,
โ
IEEE
T
r
ans
.
Elec-
tromagn.
Compat.
,
no
.
3,
pp
.
106โ127,
1977.
[18]
T
.
Shongw
e
,
A.
Han
Vinc
k,
and
H.
C
.
F
erreir
a,
โA
study
on
impulse
noise
and
its
models,
โ
SAIEE
Afr
ica
Research
Jour
nal
,
v
ol.
106,
no
.
3,
pp
.
119โ131,
2015.
[19]
A.
J
.
Al-Ask
er
y
,
C
.
C
.
Tsimenidis
,
S
.
Boussakta,
and
J
.
A.
Chambers
,
โP
erf
or
mance
analy-
sis
of
coded
massiv
e
MIMO-OFDM
systems
using
eff
ectiv
e
matr
ix
in
v
ersion,
โ
IEEE
T
r
ans
.
Comm
un.
,
v
ol.
65,
no
.
12,
pp
.
5244โ5256,
2017.
[20]
T
.
J
.
Richardson
and
R.
L.
Urbank
e
,
โThe
capacity
of
lo
w-density
par
ity-chec
k
codes
under
message-passing
decoding,
โ
IEEE
T
r
ans
.
Inf
.
Theor
y
,
v
ol.
47,
no
.
2,
pp
.
599โ618,
2001.
[21]
S
.-Y
.
Chung,
G.
D
.
F
or
ne
y
,
T
.
J
.
Richardson,
and
R.
Urban
k
e
,
โOn
the
design
of
lo
w-density
par
ity-chec
k
codes
within
0.0045
dB
of
the
Shannon
limit,
โ
IEEE
Comm
un.
lett.
,
v
ol.
5,
no
.
2,
pp
.
58โ60,
2001.
[22]
X.-Y
.
Hu,
E.
Elefther
iou,
D
.-M.
Ar
nold,
and
A.
Dholakia,
โEf๏ฌcient
implementations
of
the
sum-product
algor
ithm
f
or
decoding
LDPC
codes,
โ
in
Proc.
IEEE
Global
T
elecomm
un.
Conf
.
GLOBECOMโ01
,
v
ol.
2,
2001,
pp
.
1036โ1036E.
BIOGRAPHIES
OF
A
UTHORS
Ali
J
.
Al-Ask
er
y
is
a
lecturer
in
the
Electr
ical
Engineer
ing
T
echnical
College
=
Middle
T
echnical
Univ
ersity
,
Baghdad,
Ir
aq
.
He
receiv
ed
the
B
.Sc.
and
M.Sc.
deg
rees
in
Electr
ical
Engineer
ing
from
Al-Mustansir
iy
a
Univ
ersity
,
Baghdad,
Ir
aq,
in
2001
and
2004,
respectiv
ely
.
He
receiv
ed
his
PhD
from
the
School
of
Electr
ical
and
Electronics
Engineer
ing,
Ne
wcastle
Univ
ersity
,
Ne
wcastle
Upon
T
yne
,
U
.K.
His
research
f
ocuses
on
wireless
comm
unications
,
wired
comm
unications
,
OFDM
systems
,
coded
systems
and
receiv
er
design.
Ali
Mhaidi
Hammadi
is
an
assistant
prof
essor
in
the
Electr
ical
Engineer
ing
T
ech-
nical
College
=
Middle
T
echnical
Univ
ersity
,
Baghdad,
Ir
aq.
He
receiv
ed
his
B
.Sc
,
M.Sc
a
nd
PhD
from
the
Univ
ersity
of
T
echnology/
Al-Rasheed
College
of
Engi-
neer
ing
and
Science
,
Ir
aq,
Baghdad,
in
1993,1998
and
2007
respectiv
ely
.
His
research
interest
is
on
optical
๏ฌber
comm
unications
and
wired
comm
unications
.
Indonesian
J
Elec
Eng
&
Comp
Sci
V
ol.
14,
No
.
3,
J
une
2019
:
1272
๎
1280
Evaluation Warning : The document was created with Spire.PDF for Python.