Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
9
, No
.
3
,
J
un
e
201
9
, pp.
174
2
~
174
9
IS
S
N:
20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v
9
i
3
.
pp
1742
-
174
9
1742
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
Improve
perform
ance of t
he digit
al sin
usoi
dal gene
rator in
FPGA
by m
emor
y usage
optimiz
ation
Aiman Z
ak
w
an Jidin
,
Irn
a Nadi
ra M
ah
z
an
,
A
.
S
ha
m
s
ul R
ah
im
i
A
S
ubki
,
Wan H
asz
eri
la
Wan
H
as
s
an
Ce
ntre fo
r
Tel
e
com
m
un
ic
at
ion
Resea
rch &
I
nnovat
ion, Fa
kult
i Tek
no
l
og
i
Keju
ru
te
raa
n E
le
ktrik
&
Ele
ktronik (
FT
KEE
),
U
niv
er
s
it
i Tekn
ikal M
al
ay
sia
Mel
aka
, Mal
ay
sia
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
A
ug
12, 201
8
Re
vised Dec
20,
2018
Accepte
d
Ja
n 2
, 2019
Thi
s
pap
er
pre
s
ent
ed
the
improvem
ent
in
the
per
form
anc
e
of
the
d
igi
t
al
sinusoidal
signal
gene
rat
or
,
whic
h
was
implemen
te
d
in
FP
GA
,
by
opti
m
iz
ing
the
usage
o
f
the
ava
i
la
bl
e
m
emory
onbo
ard
.
The
s
ine
wave
was
ge
ner
ated
b
y
using
a
L
ookup
Ta
bl
e
m
et
hod,
w
her
e
i
ts
pre
-
c
alc
ula
t
ed
val
u
es
we
re
stored
in
the
onboar
d
m
emory
,
and
it
s
fr
eque
nc
y
c
an
be
adj
ustable
b
y
c
hangi
ng
th
e
inc
rement
al
st
ep
val
ue
of
th
e
m
e
m
ory
addr
ess.
In
thi
s
proposed
r
e
sea
rch
,
the
m
emory
stores
o
nl
y
25000
sam
pl
es
of
th
e
f
irst
qu
art
er
from
a
per
i
od
of
a
sin
e
wave
and
thus
,
the
outpu
t
sig
nal
a
cc
ur
acy
w
as
inc
re
ase
d
an
d
the
outpu
t
fre
quency
ran
g
e
was
expanded,
compare
d
to
t
he
pre
v
ious
res
ea
rch
.
The
proposed
design
was
succ
essfull
y
d
eve
lop
ed
and
implemente
d
in
ALTE
RA
C
y
c
lone
I
II
D
E0
FP
GA
Deve
lopment
Board
,
and
it
s
func
t
i
onal
ity
was
val
id
at
ed
vi
a
func
ti
on
al
sim
ula
ti
on
in
Mode
lsim
and
al
so
har
dware
expe
riment
al re
s
ult
s observation in
SignalTap
II.
Ke
yw
or
d
s
:
FPGA
Functi
on g
e
ne
r
at
or
Mem
or
y o
pti
m
iz
at
ion
Sinu
s
oi
dal
Copyright
©
201
9
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
Aim
an
Zakw
a
n
Ji
din
,
Ce
ntre fo
r
Tel
e
com
m
u
nicat
ion
Resea
rch &
I
nnovat
ion,
Fakult
i Tek
no
l
og
i
Ke
j
ur
utera
an
Ele
ktrik &
Ele
ktronik (
FT
KEE
),
Un
i
ver
sit
i Te
knikal
Ma
la
ysi
a Me
la
ka
,
Hang T
ua
h
Jay
a, 76
100 D
ur
ia
n
T
unggal
, Me
la
ka,
Mal
ay
sia
.
Em
a
il
: aim
anzak
wa
n@utem
.ed
u.m
y
1.
INTROD
U
CTION
Sign
al
ge
ne
rator
or
s
om
eti
m
es
cal
le
d
as
a
fu
nctio
n
ge
ne
ra
tor,
is
an
e
qu
i
pm
ent
wh
ic
h
is
ver
y
wi
dely
us
e
d
in
orde
r
t
o
gen
e
rate
the
input
sig
nals
f
or
te
sts
a
nd
e
xperim
ents
in
m
any
industria
l
app
li
cat
ions,
li
ke
in
el
ect
ro
nics
m
e
asur
em
ent,
te
le
com
m
un
ic
at
ion
syst
e
m
s
and
m
achine
co
ntr
ols,
as
well
as
it
s
ver
y
popula
r
us
a
ge
inside
the
la
bo
rator
ie
s
for
te
achin
g
an
d
le
ar
ning
pur
poses
[1
]
.
A
si
gnal
g
ener
at
or
is
ty
pical
ly
a
dev
ic
e
wh
ic
h
is
capab
le
of
c
on
st
ru
ct
in
g
a
nd
delive
rin
g
re
petit
ive
sign
al
s
,
w
hich
ty
pe
c
an
be
sel
ect
ed
for
dif
fer
e
nt
opti
ons
avail
able
su
c
h
as
the
sine
w
ave,
the
squar
e
wav
e
,
an
d
t
he
tria
ng
ular
wav
e
.
Be
sides
,
it
is
al
so
cap
able
to
pr
oduce
sig
nal
s
with
a
sp
eci
fic
fr
eq
ue
ncy
and
am
plit
ud
e
values
accu
ra
te
ly
[2
]
.
This
will
al
low
us
ers
to
rep
li
cat
e the i
nput
sig
nal tar
ge
te
d
f
or the
ci
r
cuits u
nder test
s.
A
f
un
ct
io
n
ge
ner
at
or
ca
n
be
i
m
ple
m
ented
on
pro
gr
am
m
a
ble
de
vices
suc
h
as
a
m
ic
ro
c
on
t
ro
ll
e
r
[
3].
Fo
r
e
xam
ple,
by
us
in
g
a
hig
h
-
le
vel
progr
a
m
m
ing
la
ngua
ge
li
ke
C
la
ngua
ge,
desig
ner
s
ca
n
dep
l
oy
th
e
pr
e
def
i
ned
si
n
()
f
unct
ion
i
n
orde
r
to
ge
ne
rate
the
sine
wa
ve
.
Des
pite
it
s
design
sim
plici
t
y,
the
execu
ti
on
tim
e
of
m
ic
ro
co
ntroll
ers
are
ge
neral
ly
qu
it
e
slow
and
m
ay
deg
r
ade
the
pe
rfo
r
m
ance.
As
the
m
at
te
r
of
fact,
al
l
the
instru
ct
io
n
set
s
of
the
m
ic
ro
co
ntr
ollers
a
r
e
exec
uted
se
qu
e
ntial
ly
and
in
m
os
t
of
t
he
cases
,
onl
y
on
e
instru
ct
io
n
ca
n
be
e
xecu
te
d
at
a
tim
e.
Hen
ce,
this
wea
kness
can
be
ove
rcom
e
by
us
in
g
a
Fiel
d
Pro
gr
am
m
able
Gate
A
rr
ay
(FPG
A)
.
It
is
an
ad
eq
uate
s
olu
ti
on
f
or
hi
gh
-
perform
ance
com
pu
ta
ti
ons
a
nd
i
t
is
widely
u
se
d
m
any
high
-
s
pee
d
a
ppli
cat
ion
s,
owing
t
o
it
s
low
c
os
t,
it
s
abili
ty
to
i
m
ple
m
ent
pip
el
ine
d
a
nd
par
al
le
l
com
puta
ti
on
s,
and it
s capa
bili
ty
to
opera
te
at
h
ig
h
-
f
reque
nc
y cl
ock
s
[4],
[5].
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Impr
ove
performa
nce
of the
d
i
gital sin
usoid
al
g
ene
ra
t
or
in
FPGA
…
(
Aim
an Z
akw
an Jidi
n
)
1743
In
fact,
the
re
a
re
di
ff
e
ren
t
wa
ys
of
de
velo
pin
g
a
nd
im
ple
m
enting
the
si
gnal
gen
e
rato
r
in
FPGA.
O
ne
of
the
popula
r
m
et
ho
ds
is
to
gen
e
rate
sig
na
ls
by
us
in
g
the
Direct
Digi
ta
l
Synthesize
r
(DDS),
a
te
chn
i
qu
e
whi
c
h
is
a
ble
to
pr
oduce
outp
ut
sig
nals
with
high
-
fr
e
quency
range
a
nd
acc
ur
at
e
fr
e
qu
e
nc
y
adjustm
ent.
In
this
te
chn
iq
ue,
t
he
analo
g
outp
ut
sign
al
is
pr
oduc
ed
by
gen
e
rat
ing
the
ti
m
e
-
var
yi
ng
si
gn
al
i
n
a
di
gital
for
m
,
then
conve
rted
int
o
the
anal
og
sig
na
l
via
dig
it
al
-
to
-
a
nalo
g
c
onve
rsion.
Its
pri
nc
iple
is
to
va
ry
the
f
reque
ncy
of
t
he
cl
ock
,
w
hich
i
s
us
e
d
t
o
read
the
pr
e
-
cal
cul
at
ed
wa
ve
for
m
a
m
p
li
tud
e
dat
a
di
gital
ly
store
d
in
a
m
e
m
or
y.
The
n,
t
he data
wh
ic
h hav
e
b
e
en read
is co
nv
erted t
o becom
e the a
nalo
g
si
gn
al
[
6]
-
[
8].
Me
anwhil
e,
re
search
in
[
9]
pro
po
se
d
a
m
et
hod
wh
ic
h
i
s
al
m
os
t
si
m
i
l
ar
to
the
D
D
S
te
ch
nique.
Howe
ver,
the
f
reque
ncy
of
th
e
op
e
rati
ng
cl
oc
k
is
fixe
d
an
d
the
gen
e
rated
sign
al
f
reque
nc
y
is
tun
able
sim
pl
y
by adj
us
ti
ng th
e increm
ental
step v
al
ue of
th
e ad
dr
e
s
s c
ount
er in
t
he pr
opose
d
si
gn
al
ge
ne
rator.
Be
sides,
resea
rch
in
[
10
]
ha
d
im
ple
m
ente
d
the
wav
e
for
m
gen
erato
r
i
n
Xili
nx
Virte
x
II
FP
GA,
by
us
i
ng
the
e
m
bed
ded
m
ic
ro
pr
ocess
or
.
I
n
this
resea
rch,
a
soft
pr
ocess
or
cal
le
d
Mi
cr
oB
la
ze
,
w
hich
c
on
t
ro
l
the
syst
e
m
operati
on,
is
inte
rf
ace
d
to
per
i
pherals
s
uc
h
as
m
e
m
or
ie
s
an
d
D
AC.
Howe
ver,
to
ac
hieve
high
-
band
width
si
gnal
ge
ner
at
or
,
it
is
req
uire
d
to
us
e
high
-
en
d
FP
G
A
su
c
h
as
Virtex
FP
GA
w
hich
c
ost
ver
y
exp
e
ns
i
ve.
This
pa
pe
r
pre
sents
the
im
p
ro
vem
ent
of
t
he
perform
ance
of
t
he
di
gital
sinu
s
oid
al
ge
ne
rator
w
hich
was
de
velo
ped
an
d
im
ple
m
ented
in
FP
GA.
The
im
pr
ovem
ent
wa
s
a
ble
to
be
m
ade
by
optim
iz
at
ion
the
us
a
ge
of
the
a
vaila
ble
m
e
m
or
y
reso
ur
ces
on
-
bo
a
rd.
In
this
pr
opose
d
resea
rch,
the
sine
wa
ve
is
gen
e
rated
by
us
i
ng
the
lookup
ta
ble
m
et
ho
d,
wh
e
re
the
pre
-
cal
culat
ed
sig
nal
data
are
s
tore
d
in
a
m
e
m
or
y
and
the
sign
al
fr
e
qu
e
ncy
is
a
dju
sta
ble
by
m
od
i
fyi
ng
t
he
va
lue
of
the
a
dd
ress
c
ounter
i
nc
rem
ental
ste
ps
.
The
sine
wa
ve
ca
n
be
c
onfig
ur
a
bl
e
within
a
ra
ng
e
of
1
kHz
to
10
MHz
,
with
the
f
re
qu
e
ncy
r
esolutio
n
of
1
kH
z
.
I
n
t
his
pa
per,
no
dig
it
al
-
to
-
anal
og
c
onve
rsion
is
inv
ol
ved
a
nd
thu
s
,
the
pro
po
s
ed
res
earc
h
pr
od
uce
d
the
dig
it
al
sine
wa
ve
with
accurate
fr
e
que
ncy.
2.
SIN
E
WA
VE
GENER
ATIO
N MET
HO
D
2.1.
Loo
k
up t
ab
le
The
lo
okup
ta
ble
co
ntains
th
e
data
wh
ic
h
r
epr
ese
nt
the
sa
m
ples
of
the
sine
wa
vefor
m
,
wh
ic
h
wa
s
pre
-
cal
culat
ed
offli
ne
by
us
i
ng
a
data
pr
oce
ssing
to
ol
li
ke
the
Mi
cro
s
oft
Excel.
In
the
previ
ou
s
resea
rc
h
[
9],
the
lo
okup
ta
bl
e
was
util
iz
ed
to
sto
red
20
000
16
-
b
it
data,
w
hich
represe
nts
the
1
kHz
sine
wa
ve
form
data
sam
pled
at
ever
y
50
ns.
T
hus,
al
l
the
20
000
sam
pling
wi
ll
m
ake
a
com
plete
cy
cl
e
with
a
per
io
d
of
1
Ms
.
Howe
ver,
in
this
pr
op
os
ed
res
earch
,
the
look
up
ta
ble
co
ntains
2500
0
data
wh
ic
h
was
sa
m
pled
fr
om
on
ly
the
first
quarte
r
of
a
com
plete
cy
cl
e
of
a
sin
e
wav
e
.
As
t
he
m
atter
of
fact,
this
is
doable
ow
i
ng
t
o
the
char
act
e
risti
c of a si
ne wave
wh
ic
h
is sy
m
m
et
ric.
By
loo
king
at
t
he
wa
ve
in
Fig
ur
e
1,
the
re
gion
2
of
the
wa
ve
is
sy
m
m
e
tric
to
the
reg
io
n
1,
w
hile
t
he
reg
i
on
3
a
nd
r
egio
n
4
are
sym
m
e
tric
to
the
reg
io
n
1
an
d
reg
i
on
2
with
resp
ect
to
the
x
-
a
xis,
res
pect
ively
.
Ther
e
f
or
e,
the
reg
i
on
2
of
the
wa
vefor
m
can
be
obta
ined
si
m
pl
y
by
co
un
t
ing
dow
n,
inst
ead
of
c
ou
ntin
g
up,
the
m
e
m
or
y
a
ddress
c
ounter
,
from
the
m
a
xi
m
u
m
to
the
m
ini
m
u
m
add
r
ess
of
t
he
lo
okup
ta
ble.
Wh
il
e
the
wav
e
f
or
m
in
r
egio
n
3
an
d
re
gion
4
are
sim
ply
the
neg
at
ion
of
the
val
ue
s
ob
ta
ine
d
from
the
reg
io
n
1
an
d
reg
i
on 2, r
e
sp
e
ct
ively
.
Fig
ure
1
.
The
four
reg
i
ons
of
a sine
wa
ve
Ow
i
ng
to
this,
the
m
e
m
or
y
reso
urces
util
iz
at
ion
can
be
opti
m
iz
ed
and
in
this
case,
since
a
qu
a
rter
of
the
sine
wa
ve
f
or
m
is
rep
rese
nted
by
2500
0
data,
there
a
r
e
in
fact
a
total
of
10
0000
data
sam
ples
fo
r
the
com
plete
cy
cl
e
of
t
he
sine
wa
ve.
In
a
ddit
ion
,
ever
y
pr
e
-
cal
culat
ed
data
sto
red
i
n
this
lo
ok
up
ta
ble
are
wri
tt
en
on
ly
in
13
bits.
Ther
e
f
or
e
,
in
t
he
case
wh
e
re
it
s
a
m
plit
ud
e
is
ad
j
us
ta
ble
by
+/
-
10,
a
16
-
bit
ou
t
pu
t
si
gn
al
sh
al
l
be
e
nough t
o
e
ncode t
he
m
axim
u
m
o
r
the m
i
nim
u
m
v
al
ue
of the
g
e
ner
at
e
d si
gnal
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
3
,
June
201
9
:
174
2
-
174
9
1744
2.2.
Mem
ory ini
tia
liz
at
ion
in
FP
GA
In
order
to
use
the
onboar
d
m
e
m
or
y
in
th
e
FPGA
boar
d,
a
m
e
m
or
y
in
it
ia
li
za
ti
on
file
is
need
ed
.
This
file
c
onta
ins
the
necess
ary
inf
orm
ation
s
uch
as
t
he
m
e
m
or
y
dep
t
h,
the
siz
e
of
e
ach
data,
a
nd
al
so
the
data
to
be
sto
red
in
the
m
e
m
or
y
wh
ic
h
can
be
w
ritt
en
ei
ther
in
bin
a
r
y
or
hex
a
deci
m
al
fo
rm
a
t.
Fo
r
this
researc
h
pu
rpo
se,
the
ty
pe
of
m
e
m
or
y
us
ed
i
s
the
Re
ad
-
O
nly
Mem
or
y
(ROM)
becau
se
al
l
the
data
stored
i
n
the m
e
m
or
y are pre
-
cal
c
ulate
d
a
nd their
v
al
ues
s
hall rem
ain
un
c
ha
ng
e
d w
hen the
syst
e
m
is operate
d.
2.3.
Mem
ory addr
ess counte
r
All
the
data
sto
red
i
n
the
lo
okup
ta
ble
ca
n
be
acce
ssed
by
usi
ng
t
heir
a
ddr
esses
as
s
how
n
in
Ta
ble
1
.
In
this
pr
opos
e
d
syst
e
m
,
a
m
e
m
or
y
add
ress
c
ounter
is
us
e
d
to
increase
t
he
address
value
a
t
ever
y
cl
ock
c
yc
le
.
The
co
unte
r
be
ha
vi
or
dep
e
nds
o
n
the
reg
i
on
of
the si
ne
wa
ve
form
, as
sh
own
in Fig
ure
1.
I
t wil
l cou
nt up un
ti
l
it
reaches
or
al
m
os
t
reaches
the
m
axi
m
u
m
a
ddress
i
n
re
gion
1
an
d
reg
i
on
3,
w
hile
it
will
count
do
wn
unti
l
it
reaches
or alm
os
t reac
hes
the
m
ini
m
u
m
ad
dress in
re
gi
on
3 and re
gion
4.
Table
1.
T
he
L
ookup Ta
ble fo
r
Si
ne Wave
Generati
on
Ad
d
ress
Data
0
0
1
2
2
4
…
…
1
2
4
9
9
3275
1
2
5
0
0
3276
1
2
5
0
1
3275
…
…
2
4
9
9
7
6
2
4
9
9
8
4
2
4
9
9
9
2
Her
e
,
the
ch
oi
ce
of
the
cl
oc
k
fr
e
quen
cy
is
s
ub
sta
ntial
in
order
to
pro
duce
accu
rate
sig
na
l
fr
e
quency
at
the
ou
t
pu
t.
As
pr
e
viously
m
entioned,
t
he
base
sig
nal
f
r
equ
e
ncy
for
t
he
pro
posed
sys
tem
is
F
base
=
1
kH
z
.
Hen
ce
, th
e
b
a
s
e p
e
rio
d
T
base
=
1
m
s.
The
refore
, in
orde
r
to
det
erm
ine the sa
m
pl
ing
clo
ck
perio
d:
wh
e
re
n
is
the
nu
m
ber
of
sa
m
ples.
Hen
ce,
wh
e
n
n
is
eq
ual
to
10
0000
sam
ples,
the
sa
m
pling
cl
oc
k
per
i
od
sh
al
l
be
eq
ual
to
10
ns.
S
ubs
equ
e
ntly
,
a
10
0
MH
z
sam
pli
ng
cl
oc
k
m
us
t
be
us
e
d.
In
thi
s
cas
e,
at
eve
r
y
10
ns,
the
ad
dr
e
ss
c
ounte
r
val
ue
will
be
cha
ng
e
d
by
an
i
ncr
em
ental
ste
p
value
,
w
hich
is
e
qu
al
to
the
f
re
quency
of
the
sine
wa
ve
to
be
ge
ne
rated
in
kHz.
F
or
exam
ple,
in
order
to
pro
du
ce
a
500
kHz
sine
wa
ve,
the
a
ddre
s
s
counter
m
us
t be incr
ease
d
or
decr
ease
d by
500 at
ev
e
ry sa
m
pl
ing
clo
ck
c
yc
le
.
3.
PROP
OSE
D SYSTE
M AR
CHI
TE
CT
U
R
E
Figure
2
il
lustr
at
es
the
blo
ck
diag
ram
of
the
pr
op
os
e
d
syst
e
m
fo
r
gen
e
rat
ing
sine
wa
ve
dig
it
al
ly
i
n
FPGA.
I
n
fact
,
it
con
sist
s
thr
ee
m
ai
n
blo
ck
s:
ph
ase
-
loc
k
loop
(
PLL)
,
ad
dr
e
ss
c
ounte
r
and
si
ne
wa
ve
ta
ble.
In
t
his
project
,
the
PLL
bl
ock
serv
e
s
as
t
he
cl
ock
f
reque
ncy
m
ul
ti
plier.
As
t
he
m
at
te
r
of
fa
ct
,
the
fr
e
qu
e
nc
y
of
the
cl
ock
os
ci
l
la
tor
w
hich
is
avail
able
on
th
e
FPGA
bo
a
rd
us
ed
is
50MH
z.
The
refor
e
,
in
orde
r
to
pro
du
ce
a
100
MHz
cl
oc
k
sig
nal
as
the
sa
m
pling
cl
oc
k
(clk
_100MH
z),
the
f
re
qu
e
nc
y
of
the
50
MHz
cl
ock
(s
ys_cl
k)
needs to
b
e
m
ulti
plied b
y
2 by us
i
ng the
PLL
b
loc
k.
The
sam
pling
cl
ock
f
ro
m
the
PLL
is
con
ne
ct
ed
to
bo
t
h
th
e
address
co
unte
r
and
the
sin
e
wav
e
ta
ble
.
The
f
orm
e
r
wil
l i
ncr
ease o
r de
crease the a
ddr
ess b
y t
he
ste
p value it
r
ecei
ve
s f
r
om
the ex
te
rn
al
. Ne
xt,
the
ne
w
value o
f
the
address
is
passe
d t
o
the lat
te
r
i
n order
to
acc
ess
the si
ne wave
data w
hich
h
a
d bee
n
st
or
e
d
in
side.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Impr
ove
performa
nce
of the
d
i
gital sin
usoid
al
g
ene
ra
t
or
in
FPGA
…
(
Aim
an Z
akw
an Jidi
n
)
1745
Figure
2.
The
blo
c
k diag
ram
wh
ic
h rep
rese
nt
s the
propose
d sy
stem
arch
it
ect
ur
e
The
a
ddress
c
ounte
r
operati
on
is
co
ntr
olled
by
two
finite
sta
te
m
achines.
The
first
sta
te
m
achine
is
essenti
al
to
ge
ner
at
e
t
he
c
orr
ect
m
e
m
or
y
address
a
nd
al
s
o
the
outp
ut
si
gn
al
value
si
gn,
de
pendin
g
on
the
reg
i
on
of
t
he
wav
e
f
or
m
.
Fig
ur
e
3
de
picts
t
he
sta
te
diag
ra
m
of
the
sta
te
m
achine,
w
hich
is
c
onsist
ed
of
f
our
sta
te
s,
w
hich r
epr
ese
nt the
fo
ur r
e
gions:
ONE, T
WO, T
HR
EE and
FOUR.
In
sta
te
O
NE,
the
ad
dress
sta
rts
f
ro
m
0
a
nd
it
will
be
inc
r
eased
by
ste
p
value
at
eac
h
cl
ock
cy
cl
e
.
On
ce
the
s
um
of
it
s
current
address
an
d
the
ste
p
val
ue
is
la
rg
er
tha
n
or
e
qu
al
to
th
e
m
axi
m
u
m
me
m
ory
address
,
w
hich
is
24999,
the
sta
te
m
achine
is
transiti
ng
to
sta
te
T
W
O
.
At
this
sta
te
,
the
counter
will
de
creas
e
t
he
ad
dress
by
the
ste
p
val
ue,
the
s
ub
tract
i
on
of
the
cu
rr
e
nt
address
an
d
t
he
ste
p
value
is
le
sser
or
e
qual
to
0,
wh
e
re
the
sta
te
m
achine
is
tra
ns
it
ing
t
o
sta
te
THREE.
The
sam
e
pr
oducers
ta
ke
place
for
the
tra
ns
it
io
n
from
sta
te
THREE
to
F
O
UR
a
nd
f
r
om
sta
te
FO
U
R
to
ONE.
I
n
a
dd
it
io
n,
duri
ng
sta
te
s
T
HREE
an
d
FOUR,
th
e
dat
a
read f
ro
m
the
m
e
m
or
y wil
l b
e m
ult
ipli
ed
by
-
1
in
ord
e
r
t
o hav
e
the
ne
gative
value o
f
the
g
e
ner
at
e
d
sin
e
w
a
ve.
Figure
3.
The
fi
nite st
at
e
m
ac
hin
e t
o determ
i
ne
the
cou
nter beha
vior
base
d
on the
r
e
gion
Wh
il
e the
seco
nd stat
e m
achi
ne,
as
sho
wn in F
i
gure
4
, is
use
d
to
en
s
ure that
the syste
m
will
g
ene
rat
e
correct
an
d
acc
ur
at
e
sig
nal
ou
tpu
t
w
he
nev
e
r
the
ste
p
value
is
updated
.
Thi
s
sta
te
m
achine
has
th
ree
di
fferent
sta
te
s:
COUNT,
UPD
ATE
a
nd
RE
SET.
In
COU
NT
sta
te
,
the
ad
dr
es
s
counter
is
in
i
ts
norm
al
op
er
at
ion,
wh
e
re
it
increa
ses
or
de
crease
d
the
ad
dr
es
s
by
the
giv
e
n
ste
p
value
.
The
n,
wh
e
n
the
ste
p
value
is
cha
ng
i
ng,
the
sta
te
m
achine
will
up
date
the
ne
w
c
ount
ste
p
val
ue
in
UPDAT
E
sta
te
,
an
d
it
will
re
set
back
t
he
a
ddress
value
t
o
0
in
t
he
RESET
sta
te
.
It
will
sta
y
in
this
sta
te
f
or
50
ns
befor
e
tra
ns
it
ing
t
o
th
e
COU
NT
sta
te
,
wh
e
re
the add
ress
c
ounte
r
re
su
m
es it
s nor
m
al
o
perat
ion
with t
he newly
updated
count ste
p valu
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
3
,
June
201
9
:
174
2
-
174
9
1746
Fig
ure
4.
The
fi
nite
sta
te
m
ac
hin
e
w
hich
c
on
trols the
ad
dres
s cou
nter op
e
ra
ti
on
wh
e
n
the step
v
al
ue
i
s updat
ed
4.
RESU
LT
S
AND DI
SCUS
S
ION
The
pro
posed
desig
n
was
s
uc
cessf
ully
devel
op
e
d
by
us
in
g
Ver
il
og
H
D
L
cod
e
.
In
or
de
r
to
validat
e
the
co
rr
ect
ne
ss
of
th
e
pro
pose
d
syst
em
fu
nctionalit
ie
s,
both
functi
onal
si
m
ulati
on
a
nd
ha
r
dw
a
re
e
xp
e
rim
ental
te
sts
are
perfor
m
ed.
I
n
both
c
ases,
validat
io
n
was
e
xec
uted
on
f
our
dif
fe
ren
t
set
s
of
des
ired
fr
e
quency
values:
125 k
Hz,
667
kH
z
, 200
0 kH
z
, and 7
500 k
Hz
.
4.1.
Functi
onal si
mulati
on
In
this
r
e
searc
h,
the
sim
ulatio
n
was
exec
ut
ed
by
us
i
ng
M
ento
r
Grap
hic
Mod
el
Sim
Alt
era
Editi
on
so
ft
war
e
. Bef
ore r
unni
ng the
si
m
ulati
on
, th
e
test
ben
c
h for t
he
te
sts
was wr
it
te
n
in V
e
rilo
g.
Fig
ur
e
5
s
hows
th
e
resu
lt
s
of
the
f
un
ct
io
nal
sim
u
la
ti
on
with
fiv
e
diff
e
ren
t
f
r
e
qu
e
ncy
val
ues.
Fr
om
these
wav
ef
or
m
s,
the
ou
t
put
sign
al
f
re
qu
e
nc
ie
s
can
be
obta
ined
by
m
e
asur
i
ng
t
he
pe
rio
d
of
eac
h
s
ign
al
usi
ng
tw
o
di
ff
e
ren
t
c
urso
rs.
The
m
easur
ed
fr
e
qu
e
ncies
ar
e
then
c
om
par
ed
to
the
desir
ed
f
reque
ncies,
as
li
ste
d
in
Table
2
.
Ba
se
d
on
t
his
ob
s
er
vation,
th
e
fr
eq
ue
ncy
of
the
gen
e
rated
s
ine
wav
e
was
ver
y
accu
rate
f
or
the
fi
rst
thre
e
sign
al
s.
H
ow
ever,
it
p
r
oduce
d
a s
m
al
l erro
r
of
1kHz
(
0.2%
)
when gene
rati
ng t
he 750
0 kH
z
si
ne wave
.
(a)
(b)
(c)
(d)
Fig
ure
5.
The
wav
e
f
or
m
r
esu
lt
s f
or fu
nctiona
l sim
ulati
on
by
u
sin
g dif
fer
e
nt freq
ue
ncy va
lues:
(a)
125 k
Hz, (
b)
667 kHz,
(c)
2
MHz
, (d
) 7.
5 M
Hz
Table
2.
C
om
par
iso
n betwee
n
Desire
d Fre
qu
ency v
s Meas
ured
Fr
e
quency
Desired F
requ
en
cy
Measu
red Pe
riod
Measu
red F
requ
en
cy
1
2
5
kHz
8
0
0
0
.0 n
s
1
2
5
.0 k
Hz
6
6
7
kHz
1
4
9
9
.5 n
s
6
6
6
.9 k
Hz
2
0
0
0
kHz
5
0
0
.0 n
s
2
0
0
0
.0 k
Hz
7
5
0
0
kHz
1
3
3
.3 n
s
7
5
0
1
kHz
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Impr
ove
performa
nce
of the
d
i
gital sin
usoid
al
g
ene
ra
t
or
in
FPGA
…
(
Aim
an Z
akw
an Jidi
n
)
1747
4.2.
Ha
rdw
are e
xper
im
enta
l
tes
t
The
de
velo
ped
Ve
rilog
co
de
f
or
the prop
os
e
d
desig
n
was
s
uccess
fu
ll
y
co
m
pi
le
d
by A
lt
era Q
ua
rtus I
I
so
ft
war
e
an
d
then
im
ple
m
ent
ed
i
n
Alte
ra
C
yc
lon
e
III
DE
0
FP
GA
dev
el
opm
ent
bo
a
rd.
I
n
a
ddit
ion
,
sin
ce
there
is
no
dig
it
al
-
to
-
a
nalo
g
c
onve
rter
in
volve
d
in
this
re
sear
ch,
t
he
ge
ne
ra
te
d
outp
ut
sig
nal
was
obser
ved
an
d
analy
zed
by us
ing
t
he
Si
gn
al
Tap II
Logic
A
naly
zer.
T
he
hard
war
e
expe
rim
ental
setup
a
s sho
wn in Fi
gure
6.
The
ou
t
pu
t
res
ults
of
the
hardw
a
re
e
xp
e
rim
ental
te
sts
in
F
PGA
are
sho
w
n
in
Fig
ur
e
7
.
By
us
ing
the
sam
e
m
et
ho
d
us
e
d
in
t
he
si
m
ula
ti
on
,
th
e
per
i
od
of
the
ge
ner
at
e
d
ou
t
put
sine
wa
ves
a
re
m
easur
ed
a
nd
th
us
,
the
ge
ner
at
e
d
sign
al
f
re
qu
e
nc
ie
s
can
be
ob
ta
ined
a
nd
t
he
n
com
par
e
d
w
it
h
their
ex
pec
te
d
val
ues.
It
can
be
ob
s
er
ved
that
the
res
ults
obt
ai
ned
from
the
ha
r
dware
e
xperim
ental
te
s
t
are
the
sam
e
as
the
funct
ion
al
si
m
ulati
on
, as
pr
e
viously
s
hown in T
able
2
.
Fig
ure
6.
The
hard
war
e e
xpe
rim
ental
setup
(a)
(b)
(c)
(d)
Fig
ure
7.
The
wav
e
f
or
m
r
esu
lt
s f
r
om
the F
P
GA h
a
rdwa
re t
est
observe
d
in
Sig
nalTap
L
ogic
Analy
zer
by
us
in
g dif
fer
e
nt
fr
e
qu
e
ncy
valu
es:
(
a)
12
5 kH
z
, (b) 66
7 kH
z
, (c)
2
M
Hz, (
d)
7.5 MHz
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
3
,
June
201
9
:
174
2
-
174
9
1748
As
the
m
at
te
r
of
fact,
the
sm
al
l
pr
eci
sion
e
r
rors
f
or
high
-
f
r
equ
e
ncy
ge
ne
r
at
ed
sig
nal
obs
erv
e
d
in
th
e
hard
war
e
e
xp
e
rim
ental
te
sts
are
due
to
t
he
lim
it
at
ion
of
the
util
iz
ed
to
ols:
since
the
gen
e
rated
si
gn
al
is
a
discrete
si
gn
al
wh
ic
h
ha
d
bee
n
sam
pled
e
very
10
ns
,
t
he
ti
m
e
un
it
reso
l
ution
is
li
m
it
ed
only
to
10
ns
.
H
ence,
for
the
7.5 M
H
z sine
wa
ve,
t
he
tim
e interval
for
a
pe
rio
d of
the sig
nal is s
how
n
as
13
0 ns, i
ns
te
ad
of
133 ns.
Table
3
sho
ws
the
nu
m
ber
of
the
hard
war
e
resour
ces
us
e
d
in
the
pro
posed
resea
rch,
wh
ic
h
was
ob
ta
ine
d from
the g
e
ne
rated c
om
pilat
ion
r
ep
or
t i
n Alt
era Quar
t
us
II
softw
are.
Table
3.
FP
GA
Ha
rdwar
e
Res
ources
Util
iz
at
i
on for Pr
op
os
e
d
Syst
em
I
m
ple
m
entat
ion
Desig
n
Hardwar
e Reso
u
rc
e Usage
Log
ic E
le
m
en
ts
Reg
ister
On
b
o
ard Me
m
o
r
y
PLL
Blo
ck
Prop
o
sed
des
ig
n
w
ith
o
u
t Sign
alTap
190
(1.2
%)
23
(0.1
%)
3
2
5
0
0
0
(63
.0%
)
1
(25
.0%
)
Prop
o
sed
des
ig
n
w
ith
Sign
alTap
1008
(7.0
%)
739
(4.8
.0%
)
4
3
5
5
9
2
(84
.4%
)
1
(25
.0%
)
Wh
il
e
T
able 4
com
par
es
the
overall
p
er
form
ance o
f
the
pr
opose
d
syst
em
with the p
re
vious work [
9]
.
As
ca
n
be
see
n
f
ro
m
this
ta
ble,
the
f
or
m
e
r
co
nsum
e
le
s
ser
lo
gic
el
em
ents
tha
n
the
l
at
te
r,
but
it
s
m
e
m
or
y
util
iz
at
ion
is
a
li
ttle
bit
hig
he
r
(b
y
1%
m
or
e)
than
the
la
tt
er
.
H
ow
e
ve
r,
the
fr
e
qu
e
ncy
ra
nge
s
upported
by
the
form
er,
wh
ic
h
is
between
1
kHz
to
10
MHz
,
is
m
uch
bette
r
than
the
la
tt
er,
w
hich
ca
n
only
gen
e
rate
a
sign
al
rangin
g from
1
kHz to
1 MHz
.
Table
4.
C
om
par
iso
n
Be
twee
n
P
rop
os
e
d
Sy
stem
V
s P
rev
i
ous
Wor
k
Desig
n
Me
m
o
r
y
Utilizatio
n
(bits
)
Log
ic E
le
m
en
ts
Sa
m
p
lin
g
Rate
(M
Hz)
Frequ
en
cy
Ran
g
e
Prop
o
sed
Sy
ste
m
3
2
5
0
0
0
(
6
3
.0%
)
190
(1.2
%)
100
1
kHz
–
1
0
M
Hz
Previo
u
s W
o
rk
3
2
0
0
0
0
(
6
.0%
)
240
(1.5
%)
20
1
kHz
–
1
M
Hz
Fo
r
f
ur
t
her
im
pro
vem
ents,
in
orde
r
to
im
pr
ov
e
the
te
st
a
nd
validat
io
n
of
t
he
ge
ner
at
ed
sig
nal,
a
dig
it
al
-
to
-
anal
og
co
nverte
r
(
DA
C
)
c
ou
l
d
be
ad
de
d
at
t
he
ou
t
pu
t
an
d
t
hus,
the
outp
ut
si
gn
al
can
be
vis
ualiz
ed
by
us
i
ng
the
osc
il
loscop
e
,
a
too
l
that
is
no
rm
ally
capab
le
of
m
easur
ing
the
f
re
qu
e
ncy
of
the
si
gn
al
m
or
e
accuratel
y.
Be
sides,
the
t
otal
har
m
on
ic
dist
or
ti
on
of
the
ge
ner
at
e
d
sig
nal
cou
l
d
al
so
be
analy
zed
in
order
t
o
ver
ify
it
s
fr
e
qu
ency p
uri
ty
.
More
ov
e
r,
m
or
e
featur
es
nee
d
to
be
add
e
d
to
this
propose
d
des
ig
n
to
m
a
ke
it
m
or
e
us
efu
l
to
us
e
rs
.
Fo
r
exam
ple,
com
m
on
sign
a
l
ty
pes
li
ke
tria
ng
le
,
sawt
oo
t
h
an
d
pu
lse
m
ay
be
ad
de
d
a
s
an
op
ti
on.
B
esi
des,
functi
onal
it
ie
s
su
c
h
as
am
plit
ud
e
a
nd
phase
a
dju
stm
e
nt
will
be
ve
ry
us
e
fu
l
i
n
m
any
app
li
cat
ion
s.
Fu
rt
her
m
or
e,
it
is
al
so
po
ssi
ble
to
hav
e
a
m
ulti
-
channel
fun
ct
ion
ge
ner
at
or,
wh
e
re
tw
o
or
m
or
e
sign
al
s
can
be
gen
e
rated
sim
ultaneo
us
ly
.
B
ut,
this
one
m
ay
depen
d
on
the
FP
G
A
de
vi
ce
capa
bili
ti
es
in
te
rm
of
ha
rdwar
e
resou
rces.
5.
CONCL
US
I
O
N
This
pap
e
r
ha
s
discu
ssed
on
the
im
pr
ov
e
m
ent
of
t
he
im
ple
m
entat
ion
of
t
he
dig
it
al
sine
w
av
e
gen
e
rato
r
in
F
PGA,
w
hich
had
been
ac
hieved
by
op
ti
m
iz
ing
the
m
e
m
or
y
resour
c
es
util
iz
at
ion
.
In
this
pro
po
se
d
rese
arch,
the
ge
ne
rated
f
reque
nc
y
accuracy
and
al
s
o
it
s
frequ
e
ncy
ra
ng
e
was
i
m
pr
ove
d
by
increas
in
g
t
he
nu
m
ber
of
sam
ples
f
or
one
pe
rio
d
of
t
he
sig
na
l
and
th
us
,
i
nc
reasin
g
it
s
sam
pling
rate.
This
was
achieva
ble
owing
t
o
the
sym
m
et
ric
char
act
erist
ic
of
the
si
ne
wa
ve
an
d
t
hu
s
,
only
the
f
irst
qu
a
rter
the
sign
al
need
to
be
sam
pled
an
d
st
or
e
d
in
the
m
e
m
o
r
y.
T
he
pro
pos
ed
researc
h
ha
d
been
s
uccess
fu
ll
y
im
ple
m
e
nted
i
n
Alte
ra
Cy
cl
one
II
I
D
E0
FP
GA
a
nd
the
co
rr
ect
ne
ss
of
it
s
fu
ncti
on
al
it
y
had
been
ver
if
ie
d
by
us
in
g
both
the
functi
onal
sim
ulati
on
a
nd
al
s
o
the
FP
GA
ha
rdwar
e
ex
per
i
m
ental
te
st,
where
the
outp
ut
pro
d
uce
d
by
th
e
la
tt
er
was obse
r
ve
d
i
n
Si
gn
al
Ta
p II
Lo
gic Analy
zer.
ACKN
OWLE
DGE
MENTS
The
a
uthors
w
ish
to
ac
knowl
edg
e
U
niv
e
rsiti
Tekn
i
kal
Ma
la
ysi
a
Me
la
ka
(
UTeM)
a
nd
th
e
Ma
la
ysi
a
Mi
nistry
of
Hi
gh
e
r
Ed
ucati
on
fo
r
the
fi
na
ncial
fu
ndin
g
un
de
r
Gr
a
nt
N
o.
F
RGS/1/
2015/TK0
4/FT
K/03/F
00285
and pr
ov
i
ding i
ns
tr
um
entat
ion
d
e
vices s
uppo
rt for
this
pro
j
e
ct
.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Impr
ove
performa
nce
of the
d
i
gital sin
usoid
al
g
ene
ra
t
or
in
FPGA
…
(
Aim
an Z
akw
an Jidi
n
)
1749
REFERE
NCE
S
[1]
S.
Ding,
A.
An,
and
X.
Gou,
"
Digit
al
W
a
vef
or
m
Gene
rat
or
Based
on
FPGA,"
Re
s.
J
.
Appl
.
S
c
i.
Eng.
Te
chnol
,
vol.
4(
14
)
,
pp
.
2
160
–
2166,
2012
.
[2]
A.
H.
T
irmare
,
M.
S.
R.
Mohite,
V.
A
.
Sur
y
avans
hi,
T
.
C.
Depa
rtment,
B
.
Vid
yape
e
th,
and
E.
Kolhapur
,
"
FP
GA
Based
Functi
on
Gene
rat
or
,
"
Inter
nati
onal
Re
sea
rch
Journal
of
Engi
ne
ering
and
Technol
ogy
(
IRJ
ET)
,
vol.
2(
9
),
pp.
2394
–
2399
,
2015.
[3]
V.
M.
Ibr
ahi
m
,
O.
Ta
iwo,
a
nd
U.
E.
U
y
o
at
a
,
"
Microc
ontroller
Sy
nth
esized
Functi
on
Gene
r
at
o
r,
"
Inte
rnationa
l
Journal
of
Engi
n
ee
ring
Re
search
and
Dev
e
lopmen
t
,
vo
l. 2(
7
)
,
pp.
1
9
–
25,
2012
.
[4]
G.
Brebne
r
and
W
.
Jiang,
"
High
-
Speed
Packet
Proce
ss
ing
Usi
ng
Rec
onfigur
a
ble
Com
puti
ng,
"
IEE
E
Mic
ro
,
vol.
34(
1
)
,
pp
.
8
–
18,
2014
.
[5]
W
.
Zhe
ng
,
R
.
L
i
u,
M.
Zha
ng
,
G.
Zhua
ng,
and
T.
Yuan,
"
Design
o
f
FP
GA
Based
High
-
Speed
Dat
a
Acquisi
ti
on
an
d
Rea
l
-
Ti
m
e
Da
ta
Proce
ss
ing
S
y
st
e
m
on
J
-
TE
XT T
okamak,
"
Fusio
n
Eng. Des
.
,
vol
.
89(
5
)
,
pp.
698
–
701,
2014
.
[6]
S.
Yanbin,
G.
Jian
,
and
C.
Ning,
"
High
Prec
ision
Digit
al
Freque
n
c
y
Signa
l
Source
Based
on
FPGA,"
in
Phy
sics
Proce
dia
,
vo
l. 2
5,
pp
.
1342
–
134
7,
2012
[7]
X.
Ye,
M.
Gao,
and
J
.
Huang
,
"
12
-
W
a
y
High
Acc
ura
c
y
S
ine
Signal
Gene
r
at
o
r
S
y
stem
Based
on
FP
GA
,
"
IEEE
16th
Int
ernati
on
al
Conf
ere
nce o
n
Comm
unic
ati
o
n
Technol
og
y
(
ICCT
)
,
Hangz
hou,
pp
.
833
–
836
,
2015.
[8]
M.
Herre
ro,
J.
J.
Ro
drígue
z
-
Andi
na,
and
J.
Fariñ
a
,
"
FPGA
-
base
d
Design,
Im
ple
m
ent
a
ti
on
and
Evaluat
ion
of
Digita
l
Sinus
oida
l
Gen
e
rat
ors,"
in
I
ECON
Proceedi
ngs (
Industrial
E
le
c
tronic
s
Conf
ere
nc
e)
,
pp.
2459
–
246
4,
2008
.
[9]
A.
Z.
Jidin
,
I.
N.
Mahz
an
,
N.
Hass
im,
and
A.
F.
Kadm
in,
"
Low
-
Cost
and
Portabl
e
Inte
r
active
Si
nusoidal
Digita
l
Signal
Gen
erato
r
b
y
Us
ing
FPG
A,"
Journal
of
T
el
e
communic
ation,
El
e
ct
ronic
a
nd
Computer
Engi
nee
ring
(
JTEC)
,
vol.
10(
1
)
,
pp
.
1
9
–
24,
2018
.
[10]
A.
Gom
an,
"
W
a
vef
orm
Gene
rator
Im
ple
m
ent
ed
in
F
PGA
with
an
Embedde
d
Pro
ce
ss
or,
"
Master
Thesis,
Linköpi
ng
Unive
rs
it
y
,
Sw
e
den,
no
.
LIT
H
-
I
SY
-
EX
-
3412
-
20
03,
2003
.
BIOGR
AP
H
I
ES
OF
A
UTH
ORS
Aiman
Z
a
kw
an
Jidin
obta
ine
d
his
MEng
in
Elec
tron
ic
and
Microe
l
ec
tron
ic
S
ystem
Engi
nee
r
in
g
from
ESIEE
Engi
nee
r
ing
Paris
Franc
e
in
2011.
He
has
2
y
ea
rs
o
f
working
expe
ri
enc
e
in
design
in
g
digi
tal
IC
and
d
igi
tal
s
y
s
te
m
in
FP
GA
at
Alte
ra
Corpora
t
ion
M
al
a
y
si
a,
b
efo
re
j
oini
ng
Univ
ersiti
Te
knik
al
Ma
lays
ia
Mela
k
a
(U
Te
M)
as
l
ectur
er
and
r
ese
ar
ch
er,
in
Elec
tronics
and
Com
put
er
Engi
ne
eri
ng.
His
rese
a
rch
intere
st
s inc
lud
e
FP
GA
Design
and
D
igital S
y
st
em Design
Irna
Nad
ir
a
Mah
z
an
i
s
cur
ren
t
l
y
in
her
fin
al
y
e
ar
of
studie
s
in
Bac
he
lor
of
Com
pute
r
Engi
ne
er
ing
Te
chno
log
y
in
Univer
siti
T
eknikal
Malay
si
a
Mela
k
a
(UTe
M).
Rec
en
tly
,
she
ha
d
bee
n
working
on
the
developm
ent
of
a
low
-
cost
p
orta
bl
e
signal
ge
ner
at
or
b
y
using
FPGA
as
her
fina
l
y
ea
r
proj
ec
t
.
Curre
ntly
,
she
is
under
going
her
final
y
e
a
r
int
er
nship
at
U
TAC
Manufa
ct
ur
ing
Servic
es
Sdn
Bh
d,
which
is
expect
e
d
to
b
e complete
d
in
Jul
y
2018.
A
S
hams
ul
Ra
himi
A
S
ubki
o
bta
in
ed
his
MS
c
in
Microe
le
c
tro
nic
s
from
Unive
rsiti
Keba
ngsaa
n
Malay
s
ia
in
201
1.
He
has
2
y
e
a
rs
of
working
expe
rie
n
ce
as
Photoli
thogr
aph
y
a
nd
W
et
Et
chi
ng
Proce
ss
Engi
ne
e
r
at
ON
Sem
ic
o
nduct
or
Mal
a
y
s
i
a
and
as
Proce
s
s
Inte
gra
t
ion
En
gine
er
at
Si
lT
e
rr
a
Malay
s
ia
foc
usi
ng
on
m
ai
ntaini
ng
proc
ess
par
a
m
et
ers
for
CMO
S
0.
16
nm
t
e
chnol
og
y
,
bef
or
e
joi
ning
Univer
si
ti
Te
knik
al
Mal
a
y
sia
Mel
aka
(U
Te
M)
as
le
ct
ur
er
and
rese
arc
h
er,
in
El
e
ct
roni
cs
and
Com
pute
r
Enginee
ring
.
His
re
sea
rch
in
te
rests
inc
lude
CMO
S
proc
ess
flow
opti
m
iz
ation
a
nd
cha
ra
cteri
z
at
ion
,
and
pow
er el
e
ctronics.
Wan
H
as
z
er
il
a
Wan
Hassan
holds
a
Master
of
El
e
ct
r
ic
a
l,
El
e
ct
roni
c
&
T
el
e
comm
unic
at
i
on
Engi
ne
eri
ng
fro
m
UTM
and
Ba
che
lor
of
E
le
c
trica
l
&
E
lectr
oni
c
Engi
ne
eri
ng
fro
m
UTM.
She
ha
s
pre
vious
working
expe
ri
enc
e
in
sem
ic
onduct
or
industr
y
at
Infin
eon
Technol
ogi
es
Malay
s
ia.
She
joi
n
Univ
ersit
i
T
ekni
ka
l
Mal
a
y
s
i
a
Mel
aka
(UTe
M)
as
lectur
er
o
n
2014.
H
er
r
ese
arc
h
area
inc
lud
es
Photonic
s,
Opt
ical
Com
m
un
ic
ati
on
and
W
ir
el
ess
Com
m
unic
at
ion.
Evaluation Warning : The document was created with Spire.PDF for Python.