Indonesi
an
Journa
l
of El
ect
ri
cal Engineer
ing
an
d
Comp
ut
er
Scie
nce
Vo
l.
11
,
No.
2
,
A
ugus
t
201
8
,
pp
.
5
42
~
5
48
IS
S
N: 25
02
-
4752, DO
I: 10
.11
591/ijeecs
.v1
1.i
2.
pp
5
42
-
5
48
542
Journ
al h
om
e
page
:
http:
//
ia
es
core.c
om/j
ourn
als/i
ndex.
ph
p/ij
eecs
RF Simu
lations f
or
AA
β
Crypt
osystem,
an Asymm
etric
Encrypti
on Schem
e
Sy
ed
F
arid S
ye
d Adn
an
,
Mo
hd Anu
ar M
at Is
a,
Habi
bah
Hashim
Inform
at
ion
Se
c
urity
and Trusted Infra
stru
ct
ure
s
La
bor
at
or
y
(InSTIL)
,
Facul
t
y
of
El
e
ct
r
ic
a
l
Eng
ineeri
ng
,
Univer
si
ti Te
kn
ologi
MA
RA
,
S
hah
Alam,
Sela
n
gor,
Mal
a
y
s
ia
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Feb
1,
2018
Re
vised A
pr 21
, 2
018
Accepte
d
Apr 2
7,
2018
Inte
rne
t
of
Thi
n
gs
(IoT
)
is
a
wa
y
of
providi
ng
d
at
a
with
th
e
ph
y
sica
l
thi
n
g
tha
t
in
te
rc
onn
ecte
d
to
the
n
et
w
ork,
which
is
th
e
Inte
rn
et
.
The
IoT
devi
c
es
conne
c
te
d
to
th
e
interne
t,
broa
dca
st
of
the
data
to
the
broke
r
or
a
serv
er
,
bec
om
es
an
ope
n
route
for
attackers
to
gai
n
th
e
dat
a
and
m
aki
n
g
the
data
bec
om
es
vulne
r
abl
e
.
Thus,
the
dat
a
coul
d
b
e
al
t
ere
d
or
spo
ofe
d
b
y
an
at
t
ac
ker
which
l
ed
to
sec
ur
ity
issues
espe
cially
on
dat
a
int
egr
ity.
The
re
for
e
,
the
d
at
a
sec
ur
ity
col
l
ecte
d
fr
om
t
he
sensors
is
as
important
as
on
the
serv
ers
tha
t
eve
ntu
al
l
y
b
ec
om
e
the
b
ig
d
at
a
.
How
eve
r,
m
ost
sensors
are
l
ow
powere
d
devi
c
es
in
te
rm
of
CP
U,
storage
,
m
emor
y
and
bat
t
eri
es
that
cr
y
ptogr
aphic
al
gorit
hm
computat
ions
m
ight
give
over
he
ad
to
the
sensors
and
re
duce
th
e
bat
t
eri
es
ev
en
fa
ster
tha
n
it
i
s
suppos
ed
t
o
be.
Inste
ad
of
looki
ng
a
t
s
y
m
m
et
ric
enc
r
ypti
on
sch
eme,
th
is
pape
r
tries
to
expl
ore
the
c
apabili
t
ie
s
of
the
as
y
m
m
etric
sche
m
e
on
re
s
ourc
e
constra
in
e
d
devi
c
es
comm
unic
at
ions
.
Thus,
th
is
pap
er
pre
sen
ts
an
RF
comm
unic
ation
an
aly
sis
of
a
low
consum
pti
on
asy
m
m
et
r
ic
en
cr
ypti
on,
th
e
AA
β
(AA
-
Bet
a)
tha
t
prom
ising
to
implement
on
the
IoT
devi
c
es
to
sec
ure
the
IoT
net
works
.
The
r
esult
show
s
onl
y
14%
si
ze
i
ncr
ea
sed
in
ci
ph
ert
ex
t
from
pla
i
nte
xt
and
the
R
F
sim
ula
ti
on
comm
unic
at
ions
show
a
b
et
t
er
r
esult
in
Raspbi
a
n
OS
en
vironme
nt
compare
to
windows
environm
ent
eve
n
th
ough
with
sam
e
conf
igurations.
Ke
yw
or
d
s
:
In
te
r
net
of Th
i
ng
s
IoT
Encr
y
ption
Publi
c k
ey
Copyright
©
201
8
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
:
Syed
Fa
rid
Sye
d Adna
n
,
Inform
at
ion
Se
cur
it
y an
d
T
r
ust
ed
I
nfrastr
uctu
res
La
borato
ry
(InST
IL)
,
Faculty
of Elec
tric
al
Engineer
ing
,
Un
i
ver
sit
i Te
knol
og
i M
ARA
,
Sh
a
h Alam
, S
el
angor,
Ma
la
ysi
a
.
Em
a
il
:
far
rid
85@g
m
ai
l.co
m
1.
INTROD
U
CTION
In
t
he
w
or
l
d
of
the
I
ntern
et
,
a
lot
of
ta
s
ks
has
been
m
igrated
to
onli
ne
s
uch
as
onli
ne
ba
nk
ing
,
onli
ne
trans
port
booki
ng
a
nd
e
ven
onli
ne
sho
pp
i
ng
exp
e
rience.
A
s
we
cha
ng
e
to
wards
an
a
utono
m
ou
s
w
orl
d
wh
e
re
al
l
thing
s
are
un
ifie
d
on
the
internet
,
this
info
rm
at
ion
is
colle
ct
ed
and
e
xch
a
nge
d,
in
a
way
we
are
un
i
ntentio
nally
op
e
n
to
a
dv
e
rsar
ie
s
the
li
ke
li
ho
ods
of
da
ta
vu
lne
ra
ble.
The
at
ta
ck
c
ou
l
d
be
by
il
l
egall
y
gaining
th
e
dat
a
or
b
y
ta
m
per
ing
t
he
data
i
ntegr
it
y
regar
dless
of
the d
at
a
le
vel
sig
nifican
t
or
ty
pe
of
t
hin
gs
ar
e
on
t
he
inter
net
.
This
sta
te
m
e
nt
co
uld
be
suppo
rted
by
He
wlet
t
Packard
(H
P
)
re
searc
h
carried
out
by
the
IoT
Re
search
stu
dy
[1]
wh
ic
h
fou
nd
that
70
pe
rc
ent
of
I
oT
de
vi
ces
runn
i
ng
on
un
e
ncr
y
pted
ne
twork
ser
vice
s
and
80 p
e
rce
nt of t
he devices
elev
at
ed
up
pri
vac
y worri
es
.
In
te
r
net
of
Th
ing
s
(
IoT)
en
ables
interco
nnect
ed
of
thin
gs
eq
uipped
with
em
bed
de
d
su
c
h
as
in
veh
ic
le
s,
house
s
an
d
eve
n
ci
ti
es
that
al
lows
the
thi
ng
s
aut
onom
ou
sly
sha
r
e
the
data
on
the
inter
net
[
2]
.
In
t
he
ye
ar
2025,
100
bill
ion
I
oT
de
vices
co
nnect
ed
is
pro
j
ect
ed
[
3]
.
I
oT
a
pp
li
ca
ti
on
exam
ple
are
by
Or
ti
z
et
al
[4]
that
m
easur
e
he
art
rates
to
get
her
with
wav
e
form
s.
The
dat
a
is
then
tra
nsm
itted
to
the
web
sit
e
a
nd
da
ta
base.
Anothe
r
e
xam
ple
is
by
Suba
sh
ini
et
al
.
[
5]
Io
T
ap
plica
ti
on
to
m
on
it
or
t
he
cr
op
s
hum
idi
ty
,
tem
per
at
ur
e
,
li
gh
t
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
RF
Si
mu
l
atio
ns f
or
AAβ Cry
pt
os
yst
em, a
n A
symm
et
ric
…
(
Sy
ed
F
ar
id
S
ye
d Ad
nan
)
543
and
s
oil
m
ois
ture,
an
d
tran
sm
it
te
d
to
ser
ver
f
or
a
naly
sis
an
d
databa
se
w
hich
ca
n
con
tr
ol
the
growt
h
par
am
et
ers.
Howe
ver,
the
vu
l
ner
a
bili
ty
of
the
Th
i
ngs
is
su
e
m
igh
t
aris
e
with
t
hese
in
te
rconn
ect
e
d
de
vices.
In
a
stud
y
by
[
1]
,
a
la
rg
e
num
ber
of
dev
ic
es
neg
l
ect
ed
th
e
enc
ryption
w
hile
updating
t
he
software
updates
a
nd
no
encr
y
ption i
n
c
omm
un
ic
at
ion
wh
il
e c
onnecti
ng to
t
he
inte
rnet
.
As
sta
te
d
by
R
os
e
et
al
.
[
3]
,
I
oT
op
e
n
t
o
c
ha
ll
eng
es
s
uc
h
as
secu
rity
wh
e
re
IoT
im
ple
m
en
ta
ti
on
s
will
encou
nter
ne
w
and
un
i
qu
e
se
cur
it
y
chall
en
ge
s
.
A
dd
it
io
nally
,
there
are
se
ver
al
at
ta
cks
on
the
sen
sor
it
sel
f
as
m
entioned
by
Anwar
et
.
al
.
[
6]
su
ch
as
rep
l
ay
at
ta
ck,
i
m
p
ersonati
on
an
d
com
pr
om
ise
a
tt
acks.
The
re
for
e,
it
is
a
need
f
or
Io
T
dev
ic
es
to
be
trusted
by
use
rs
in
orde
r
to
pr
e
ve
nt
the
da
ta
fr
om
be
ing
com
pr
om
ise
d
i
n
any
way
[
3]
.
I
n
a
nothe
r
r
esea
rc
h,
Ma
hm
ou
d
et
al
[7]
sai
d
t
hat
the
I
oT
fr
am
e
work
can
be
at
ta
cked
at
eac
h
la
ye
r
giv
in
g rise t
o t
r
e
m
end
ous
secu
rit
y chall
eng
e
s
and the
reb
y i
t i
s cruci
al
to
e
xa
m
ine n
ew
sec
uri
ty
r
eq
uirem
e
nts.
These
sta
te
m
e
nts
are
the
n
s
uppo
rted
in
a
re
port
by
on
e
of
the
la
rg
est
lo
gi
sti
cs
com
pan
y
in
the
w
orl
d,
the
D
HL
.
T
he
DH
L
arti
cl
e
by
Ma
caulay
et
al
.
[8]
m
entione
d
that
a
uto
m
at
ion
opport
un
it
ie
s
can
be
occ
up
ie
d
by
I
oT
wh
ic
h
l
ead
to
secu
rity
vu
l
ner
a
bili
t
ie
s
an
d
it
is
vital
to
e
ns
ure
I
oT
s
ecur
it
y
age
nda
by
the
c
ollab
orat
io
n
of
in
dustry
an
d
gove
r
nm
ents
.
In
te
rc
onne
ct
ion
of
I
oT
dev
i
ces
op
e
ns
a
pa
th
for
cy
ber
c
ri
m
inals,
hack
e
r
s
and
even te
rro
rist m
ay
h
arm
the IoT net
w
ork f
or their
own
ag
e
nd
a
.
More
ov
e
r,
t
o
ge
t
t
he
Io
T
co
nn
ect
ed,
the
thin
gs
m
us
t
be
connecte
d
to
the net
work.
O
ne
th
ing
rem
ai
n
s
i
m
po
rtant
is
th
e
m
ob
il
it
y
of
the
dev
ic
es
or
thin
gs
t
o
be
anyw
her
e
an
d
scat
te
red
a
rou
nd
the
hom
e,
office,
war
e
hous
e
or
even
sc
hool
re
su
lt
ed
unsu
it
a
bl
e
to
be
c
onne
ct
ed
on
wire
d
-
base
d.
Lots
of
wirin
g
nee
d
t
o
util
iz
e
and
eve
n
desig
ning
the
wire
d
la
yout
is
an
ot
her
c
halle
nge
t
o
co
ns
i
der.
W
i
reless
te
ch
no
l
ogy
co
uld
be
a
ppli
ed
to
the
co
nnect
ed
thi
ngs
to
offe
r
the
fle
xib
i
li
ty
of
the
thin
gs
.
This
wi
reless
te
ch
nolo
gy
co
uld
be
Bl
ue
too
t
h
te
c
hnology,
Ra
dio
Fr
e
quency
(RF)
,
WI
F
I
or
even
with
3G
or
LTE
[
7]
[
9]
.
H
oweve
r,
the
desi
gn
is
base
d
on
the
locat
ion
,
di
sta
nce
and
en
erg
y
co
nsum
pt
ion
lim
it
a
ti
on
for
the
thin
gs
wh
e
n
co
nnect
ed
to
the
intern
et
.
RF
cou
l
d
be
t
he
m
ai
n
play
er
i
n
Io
T
as
m
entione
d
by
Mc
a
dam
s
[10]
a
nd
the
a
ut
hor
ad
vised
to
get
t
he
RF
fun
dam
ental
esp
eci
al
ly
in
desi
gn
i
ng the
wirel
ess to p
rovide
re
li
abili
ty
to
IoT n
et
works.
Re
centl
y,
ther
e
hav
e
bee
n
lots
of
rese
arch
done
on
protect
ing
t
he
I
oT
dev
ic
es
an
d
the
com
m
un
ic
at
ion
of
the
c
onne
ct
ed
dev
ic
es
.
A
heter
og
e
ne
ous
sig
ncr
ypti
on
schem
e
fo
r
the
onli
ne
an
d
offli
ne
app
li
cat
io
n
ha
s
been
s
uggest
ed
by
Li
and
Xiong
[
11]
.
The
auth
ors
pro
po
s
ed
sec
ur
e
d
channel
betw
een
the
sens
or
no
de
an
d
the
In
te
rn
et
wh
ic
h
was
al
s
o
able
to
reduc
e
the
sensor
no
des
cost
in
te
r
m
s
of
energy
and
tim
e
t
aken
wh
il
e
integ
rati
ng
sec
ur
it
y.
Sin
gh
e
t
al
.
[12]
have
desig
ne
d
an
d
im
ple
m
ented
the
sec
ur
e
MQT
T
protoc
ols
(S
M
QTT,
SM
QT
TSN)
with
ne
w
sec
ure
publish
c
omm
an
d
“
SP
ub
li
s
h
”.
SP
ubli
sh
pu
blishes
encr
y
pted
data
with
l
ig
htwei
gh
t
ECC
te
ch
ni
qu
es
t
hroug
h
op
ti
m
iz
ed
par
a
m
et
ers
and
c
om
pu
ta
ti
on
al
gorithm
s
ov
e
r
t
he
el
li
ptic
cu
rv
e
(EC
).
The
a
utho
rs
ha
ve
c
om
par
e
d
t
heir
sc
hem
e
an
d
the
re
su
lt
s
s
how
that
it
is
be
tt
er
in
te
rm
s
of
per
f
or
m
ance
against
Wang
et
al
.
[
13]
wh
ic
h
util
iz
ed
the
Ja
va
li
brary
f
or
a
n
asy
m
m
e
tric
encr
y
ption
schem
e
i
m
p
le
m
ented
on
a
n
and
r
oid
base
d
phone
cal
le
d
Attrib
ute
-
Ba
sed
Enc
ryptio
n
(A
BE
).
A
pa
rt
fr
om
MQTT
wh
ic
h
util
iz
ing
the
TCP
prot
oco
ls
,
there
is
al
s
o
Con
st
raine
d
Applic
at
ion
P
r
oto
c
ol
(CoAP
)
that
util
iz
ed
U
DP
pro
toc
ols
.
A
re
search
that
i
ntr
oduce
d
“
Secu
r
eSense”
wa
s
don
e
by
Ra
za
e
t
al
.
in
2017
[
14
]
by
integrati
ng
the
CoA
P
a
n
d
se
cur
it
y
protoc
ol
DLTS
directl
y
between
e
nd
to
end
(E
2E)
Io
T
dev
ic
es
.
They
te
ste
d
the
“Se
cur
e
Sens
e”
on
6L
owPAN
ne
tworks
side
a
nd
their
cl
oud
platfo
rm
side
nam
el
y
Sics
th
Sense
wh
ic
h
co
nclu
de
s
their
“Secu
r
eSense”
ti
m
ing
and
ene
rg
y
is
tolerable
f
or
I
oT
app
li
cat
ion
s
.
These
stud
ie
s
sh
ow
that a l
ot
of r
es
earche
rs ha
ve pu
t a
n
e
ffo
rt to
secu
re th
e
IoT
for
t
he wave
s
of IoT.
2.
AA
β
CR
Y
PTO
SY
STE
M
The
fi
ndin
gs
to
sec
ur
e
IoT
things
can
be
sy
m
m
e
tric
or
asym
m
e
tric
encr
y
ption
or
e
ven
hybri
d
schem
e.
Thu
s
,
this
m
otivate
s
this
resea
rch
t
o
ex
plo
re
the
c
om
petence
of
a
n
asy
m
m
et
ric
schem
e,
the
AA
-
Be
ta
(
AA
β
)
sch
em
e.
AA
β
is
an
al
te
rn
at
ive
asy
m
m
e
tric
crypto
sc
hem
e
that
was
i
ntrod
uc
ed
by
A
riff
i
n
et
al
.
in
20
12
[
15]
.
The
a
uthor
s
dem
on
strat
ed
e
m
pirically
that
AA
β
e
ncr
y
ptio
n
s
hows
faste
r
runtim
e
co
m
par
e
to
oth
e
r
asy
m
m
e
tric
encr
y
ptio
n
sc
hem
es,
su
ch
as
the
Ri
ve
st
Sh
am
ir
Ad
le
m
an
(RSA
)
and
Ell
ipti
c
curv
e
crypto
gr
a
phy (
ECC
).
Additi
on
al
ly
,
t
he
decr
y
ption
perform
ance
was
m
entione
d
as
fast
a
s
R
SA
a
nd
m
arg
inall
y
be
hind
ECC
.
I
n
their
assessm
ent
b
y
Ar
if
fin
et
al
.
[16]
ha
ve
s
hown
t
heir
sc
he
m
e
on
Ma
ple
13
us
i
ng
Wi
ndows
7
op
e
rati
ng
syst
em
and
dem
on
s
trat
ed
that
AA
β
encr
y
ption
s
pe
ed
is
faster
eve
n
f
or
la
r
ge
dat
a
set
s
with
la
rge
key
siz
es.
The
aut
hors
c
on
ti
nue
th
ei
r
researc
h
in
[17]
wh
ic
h
sho
ws
that
AA
β
c
ould
ha
ndle
la
rge
m
essage
siz
e
and
perform
s f
ast
er
than RS
A
a
nd
ECC
in
Ma
ple
13 envir
onm
ent.
The
AA
β
e
ncry
ption
sc
hem
e
al
go
rithm
that
rev
eal
ed
by
Ar
iffi
n
et
al
.
in
[16]
consi
st
of
key
gen
e
rati
on, e
nc
ryptio
n
a
nd d
e
crypti
on ca
n b
e d
esc
ribe
d:
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
11
, N
o.
2
,
A
ugust
2018
:
542
–
548
544
2.1
AA
β
Key
G
enera
tion
AA
β
key
gen
e
ra
ti
on
ste
ps
ca
n be
descr
i
bed in
A
lg
ori
thm
1
:
Algori
th
m
1:
Ke
y G
ener
atio
n
f
or
AA
β
1.
BE
GIN
2.
Choose r
an
dom primes
p an
d q wit
h
size
o
f
n
-
bit
un
ti
l
and
.
3.
Get a r
andom
integer
4.
Get intege
r e s
uch th
at e.
d ≡1 m
od pq
Ad
d mu
lt
iple
p.q until
(
if
n
ecessary)
5.
Get A
1
=
whe
re
A
1
6.
Set A
2
=
e
7.
Ret
ur
n (
n, A
1
, A
2
)
a
nd
(p
q, d)
8.
END
2.2
AA
β
Encr
yption
AA
β
e
ncr
y
ptio
n process
can
be
de
fine
d
in
A
l
gorithm
2
:
Algori
th
m
2:
AA
β
Encr
yp
tio
n
1.
BE
GIN
2.
Get p
l
ain
te
xt
M
3.
Divi
de plai
ntext
M to get
pla
i
ntext
pair
4.
Compu
te
cip
he
rte
xt
,
5.
Ret
ur
n ci
phert
ext
, C
6.
END
2.3
AA
β
Decr
yption
AA
β
dec
ryptio
n
proce
dure
is
s
how
n
in
A
l
gori
thm
3
:
Algori
th
m
3:
AA
β
Decryp
ti
on
1.
BE
GIN
2.
Rea
d
ci
phert
ex
t, C
3.
Com
pu
te
4.
Compu
te
5.
Compu
te
6.
Compu
te
7.
Compu
te
8.
Loop
Compu
te
9.
While
cou
nter
<4
10.
If
>
0,
p
ic
k th
e pair
11.
Set
12.
13.
END
The
AA
β
e
ncry
ption
schem
e
ov
er
view
is
descr
i
bed
in
F
igure
1.
Th
e
public
key
(
n,
A
1
,
A
2
)
are
publicl
y
avail
able
to
al
l
dev
i
ces.
Alon
g
an
d
Bus
u
are
a
n
exam
ple
of
two
e
ntit
ie
s
to
est
ablish
a
secur
e
com
m
un
ic
at
ion
c
hannel.
Along ca
n be
d
e
sc
ribe
d
as t
he
se
r
ver,
wh
il
e Bu
s
u
is t
he
Cl
ie
nt.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
RF
Si
mu
l
atio
ns f
or
AAβ Cry
pt
os
yst
em, a
n A
symm
et
ric
…
(
Sy
ed
F
ar
id
S
ye
d Ad
nan
)
54
5
Figure
1
.
AA
β
Encr
y
ption
Fo
r
the
AA
β
de
crypti
on
sc
he
m
e,
the
proces
s
can
be
s
ummari
zed
as
i
n
Fi
gure
2.
T
he
pr
i
vate
key
(
pq,
d)
on
ly
av
ai
la
bl
e to A
l
ong
a
nd the
private
ke
y i
s u
se
d
to
dec
rypt Bu
su’s
m
essage.
Figure
2
.
AA
β
Decr
y
ption
3.
E
X
PERI
MEN
TAL SET
UP
The
e
xp
e
rim
ents
had
bee
n
c
onduct
ed
on
tw
o
ty
pes
of
en
vi
ronm
ent.
The
first
set
up
is
on
W
i
ndows
10
operati
ng
s
yst
e
m
(O
S)
,
w
hich
is
usual
ly
for
a
n
e
nd
use
r.
On
an
oth
e
r
hand,
the
sec
ond
en
vir
onm
ent
is
on
Ra
sp
bia
n
OS
.
The
Ra
s
pb
ia
n
OS
is
a
Li
nux
OS
that
has
be
en
e
nh
a
nce
d
f
or
Ra
s
pber
ry
P
i
,
em
bed
ded
de
vices
that
are
base
d on
the
ARM
process
or
.
The
pla
tfor
m
s ar
e
:
a)
Toshiba
L
745
No
te
book.
T
he
p
r
ocess
or
is
In
te
l
Co
re
i5
2450M
(
2
nd
ge
ner
at
io
n)
with
2.5
G
Hz
bas
e
fr
e
qu
e
ncy a
nd
4Gb of D
DR
3
1333Mhz
m
e
mo
ry.
Ru
nn
i
ng
with
W
i
ndo
ws
10.
b)
Ra
sp
be
rr
y
Pi
2,
a
So
C
that
is
based
on
32
-
bi
t
Ar
m
Cortex
-
A7
CP
U
at
900MHz
qua
d
-
c
ore
and
1Gb
of
RAM sha
re
d w
it
h
GPU.
Ru
nn
ing
on Ras
pb
ia
n
“Jessi
e”
v
e
rs
ion
.
On
t
he
c
omm
un
ic
at
ion
s
be
tween
de
vice
s
,
the
cod
e
s
are
wr
it
te
n
in
Pyt
hon
scriptin
g.
The
com
m
un
ic
at
ion
s
incl
ud
e
d
tra
ns
m
issi
on
(T
x)
and
receive
(
Rx)
wh
il
e
the
serial
sp
ee
d
ch
os
e
n
is
1152
00
ba
ud
rate.
On
the
oth
e
r
ha
nds,
t
he
co
de
s
wer
e
w
ritt
en
us
in
g
the
C
pro
gr
am
m
ing
la
ngua
ge
f
or
AA
β
enc
ryptio
n
process
on
Li
nux
a
nd
Ra
spbia
n
e
nviro
nme
nt
.
T
o
ha
ndl
e
m
ult
iple
pr
e
ci
sion
nu
m
bers
an
d
opti
m
ize
the
encr
y
ption
pro
cess,
the
co
de
s
we
re
wr
it
te
n
with
Bi
gnum
li
br
ary
(
GMP
).
GMP
li
brar
y
is
usual
ly
use
d
for
arb
it
ra
ry
pr
eci
sion
a
rithm
etic
that
su
pp
or
ts
the
cal
culat
ions
up
to
t
he
avai
la
ble
m
e
m
or
y
of
the
syst
em
rather
than ha
ving
a
lim
i
ta
ti
on
on t
he
ty
pes of
t
he v
ariable i
tse
lf
[18]
.
In these e
xperi
m
ents, th
e size
of
plainte
xt
pa
ir, m
1
and
m
2
a
re
based o
n
t
he
cond
it
io
n wh
e
re:
a)
b)
Wh
e
re
n
is
t
he
le
ng
t
h
of
pri
m
e
and
m
is
the
plainte
xt.
T
he
plainte
xt
ra
ng
e
use
d
is
ba
sed
on
A
rif
fin
et
al
.
def
i
niti
on
in
[
15
]
.
Me
a
nwhile
,
this
resea
rc
h
ass
um
ed
the
key
gen
e
rati
on
is
ge
ner
at
e
d
earli
er
an
d
the
public
key is
pr
es
har
e
d wit
h
the
cli
ent and t
he
se
rver.
A
l
o
n
g
Busu
G
ene
rate:
Publ
i
c
Key
(
n, A
1
, A
2
)
Pri
v
ate K
e
y
(pq
,
d)
Publ
i
c
Key
(
n, A
1
, A
2
)
ar
e av
ai
l
abl
e
p
ubl
i
c
l
y
Enc
r
y
pti
on b
y
Bu
s
u,
Pl
ai
ntext M pai
r
(m
1
,
m
2
)
Bus
u s
ends
Ci
phertex
t,
C to A
l
on
g
Busu
A
l
on
g
Bus
u s
ends
Ci
phertex
t,
C to A
l
on
g
Al
ong dec
r
y
pt t
h
e c
i
phertex
t,
C
w
i
th Pri
v
ate k
e
y
pai
r (
pq, d
) and
gets
the pl
ai
ntext
M pai
r, (
m
1
,
m
2
)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
11
, N
o.
2
,
A
ugust
2018
:
542
–
548
546
4.
RESU
LT
S
A
ND AN
ALYSIS
This
sect
io
n
pr
esents
the
re
sul
ts
of
the
e
xp
e
rim
ents
fr
om
t
he
pre
vious
se
ct
ion
.
T
he
re
s
ults
i
nclu
de
the RF sim
ulati
on
t
ran
sm
issi
on
a
nd r
ecei
ve r
un
ti
m
e w
hile t
r
ansm
itti
ng
the
ci
ph
e
rtext.
Table
1
sho
ws
the
si
m
ulati
o
n
of
AA
β
RF
transm
issi
on
in
Tosh
i
ba
N
oteboo
k
L7
45.
Wh
il
e
Table
2
sh
ows
the
RF
si
m
ulati
on
f
or
AA
β
on
Ra
spber
ry
Pi
2
platfo
rm
.
The
data
siz
e
is
re
ferr
ing
t
o
the
ci
pherte
xt
gen
e
rated fr
om t
he
encr
ypti
on
secti
on
. The da
ta
size i
s r
epr
esented in bit
s.
Both
res
ults sim
ula
te
the co
ndit
ion
wh
e
re
a
cl
ie
nt
transm
itti
ng
da
ta
to
the
se
r
ver.
This
co
uld
be
a
tran
sm
issi
on
from
a
sens
or
to
a
se
r
ver
,
w
here
low r
un
ti
m
e and
lo
w
e
ne
rg
y i
s
r
e
qu
ire
d
e
sp
ec
ia
ll
y fo
r
a l
ow
powe
red de
vic
e.
Table
1
. T
oshi
ba
L
745 N
oteb
ook
RF
T
ran
s
m
issi
on
Sim
ul
at
ion
R
un
ti
m
e
Key
Size
Data Siz
e (
b
its)
Clien
t
Server
Tr
an
s
m
it
(s)
Receiv
ed
(s)
1536
3587
0
.06
8
9
7
9
2
2
2
0
.99
9
0
8
2
2
6
3072
7172
0
.14
7
9
4
8
7
1
8
1
.03
7
3
7
4
7
6
5
6144
1
4
3
4
0
0
.30
4
9
4
8
2
5
3
1
.08
2
1
4
3
8
0
4
Table
2
. Ras
pb
err
y Pi
2
R
F T
r
ansm
issi
on
Si
m
ula
ti
on
Ru
ntim
e
Key
Size
Data Siz
e (
b
its)
Clien
t
Server
Tr
an
s
m
it
(s)
Receiv
ed
(s)
1536
3587
0
.00
1
2
3
4
6
6
7
0
.05
6
0
9
0
3
3
3
3072
7172
0
.00
1
3
4
9
6
6
7
0
.12
0
4
4
7
3
3
3
6144
1
4
3
4
0
0
.00
1
7
1
6
6
6
7
0
.23
9
1
3
7
3
3
3
Figure
3
s
ho
ws
the
c
om
par
iso
n
of
RF
s
i
m
ulati
on
on
Ra
sp
be
rr
y
Pi
2(
R
pi2)
an
d
T
os
hi
ba
L
745
No
te
book
wh
i
ch
s
hows
t
he
t
ran
sm
it
(TX
)
on
Rpi2
is
le
ss
than
0.0
5
se
conds
ac
r
os
s
a
ll
data
siz
e.
A
nd
th
e
runtim
e
r
ise
s
expo
nen
ti
al
ly
for
T
os
hi
ba
L
745
No
te
book
from
0.
069
s
econds
with
3587
-
bits
data
s
iz
e
and
highest at
0.305 s
ec
onds wit
h 1
4340
-
bits dat
a size
.
Fu
rt
her
m
or
e,
Figure
4
s
how
s
the
si
m
ulati
on
on
AA
β
RF
r
ecei
ve
(Rx).
It
is
cl
early
see
n
that
Rpi2
do
m
inate
s the lo
west R
x
r
unti
m
e even
for
la
r
ge
data siz
e at
14340
-
bits at
0
.239
s
eco
nds. While
T
os
hi
ba
L745
No
te
book
has
a
ste
ady
Rx
r
untim
e
at
1
secon
ds
ac
r
os
s
al
l
data
siz
e.
This
m
igh
t
be
due
t
o
wi
ndows
lim
it
at
ion
on
sim
ulati
on
us
in
g
serial
co
m
m
un
ic
at
ion
s
com
par
ed
to
Ra
sp
be
rr
y
Pi
wh
e
re
Ra
spbia
n,
a
Lin
ux
OS
wer
e
instal
le
d.
Eve
n
tho
ug
h
the
sa
m
e
con
fig
urat
ion
wer
e
c
onfi
gure
d
su
c
h
as
baud
rate
at
1152
00
baud
on
both
platfo
rm
.
Figure
5
su
m
m
ariz
es
the
total
RF
trans
m
issi
on
on
both
R
pi2
a
nd
To
sh
i
ba
L
745
N
oteb
ook.
At
the
highest
key
si
ze
of
61
44
-
bit
an
d
25
85
-
bits
of
data,
t
he
Rpi2
do
m
inate
d
T
os
hi
ba
L
74
5
N
ote
book
on
total
transm
issi
on
tim
e
at
5.
8
tim
e
s,
resu
lt
in
g
re
duced
runtim
e
a
nd
e
nergy
consum
ption
to
secur
e
a
plaint
e
xt
an
d
transm
it
u
sing
so
le
ly
the
AA
β
encr
y
ption sch
e
m
e.
Table
3
s
how
s
the
AA
β
e
nc
ryptio
n
sc
hem
e
proces
s
on
Rpi2.
This
inc
lud
es
f
ro
m
en
crypti
ng
the
pl
ai
ntext,
sim
ulati
on
to
tra
ns
m
it
the
ci
ph
erte
xt,
sim
ul
at
ion
to
recei
ve
the
ci
phert
ext
and
decr
y
pted
the
ci
ph
e
rtext to
gai
n
the
ori
gin
al
p
la
intext
.
T
hi
s pro
ce
ss ca
n b
e su
m
m
arized in
Figure
6
for
AA
β
cry
pto
syst
e
m
wh
il
e
the
public
key
is
consi
der
e
d
pre
-
sh
a
re
d
earli
er
.
Th
e
total
runtim
e
exp
on
entia
ll
y
increased
f
ro
m
key
si
ze
1536
-
bits
to
6144
-
bits
and
worst
tim
e
fo
r
total
ru
ntim
e
s
how
s
le
ss
than
1
sec
onds
t
o
for
la
r
gest
key
siz
e,
6144
-
bits
an
d
la
rg
est
plainte
xt
siz
e
at
12
28
7
-
bits.
O
n
the
oth
e
r
hand,
the p
la
in
te
xt
siz
e
was
i
ncr
ease
d
ar
ound
14.
2%
to
14.
3%
to
ci
ph
e
rte
xt
wh
e
n
enc
rypted
to
AA
β
for
al
l
key
siz
e.
Figure
3
.
AA
β
RF Tra
ns
m
it
(
Tx) sim
ulati
on
com
par
ison o
f R
aspb
e
rr
y
Pi 2
(Rpi2) a
nd T
oshiba
Figure
4
.
AA
β
RF Re
cei
ve (Rx)
sim
ulatio
n
com
par
ison o
f R
aspb
e
rr
y
Pi 2
(Rpi2) a
nd T
oshiba
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
RF
Si
mu
l
atio
ns f
or
AAβ Cry
pt
os
yst
em, a
n A
symm
et
ric
…
(
Sy
ed
F
ar
id
S
ye
d Ad
nan
)
547
L7
45 Noteb
oo
k
L7
45 Noteb
oo
k
Figure
5
.
AA
β
Com
par
ison o
n Total
of
R
F
Transm
issi
on
run
ti
m
e si
m
ulatio
n from
Cl
ie
nt to
Ser
ver
on Ras
pber
ry Pi
2 (Rpi
2) an
d
T
oshiba
L74
5
No
te
book
platf
or
m
Figure
6
.
AA
β
Total
Ru
ntim
e
for
E
ncr
y
ption,
Decr
y
ption an
d
T
ransm
issi
on
on Ras
pb
e
rr
y
Pi 2
Table
3
.
AA
β
E
ncr
y
ption,
Dec
ryptio
n
a
nd Tra
ns
m
issi
on
Ru
nti
m
e o
n
Ra
s
pber
ry Pi
2
Key
Size
Plain
tex
t Size
Encry
p
tio
n
(
s)
Decr
y
p
tio
n
(
s)
Clien
t
Server
Tr
an
s
m
it
(s)
Receiv
ed
(s)
1536
3073
0
.00
0
9
5
8
7
0
.01
8
6
1
6
1
0
.00
1
2
3
4
6
6
7
0
.05
6
0
9
0
3
3
3
3072
6143
0
.00
1
7
5
5
0
.11
5
3
9
6
1
0
.00
1
3
4
9
6
6
7
0
.12
0
4
4
7
3
3
3
6144
1
2
2
8
7
0
.00
3
9
2
1
7
0
.72
3
4
3
7
3
0
.00
1
7
1
6
6
6
7
0
.23
9
1
3
7
3
3
3
Give
n
the
sce
na
rio
a
bove
, giv
en
the
m
easur
e
d
c
urren
t
on
R
pi2
is
0.25
8 A
m
p
at
5
Volt
, t
he
e
nergy
can
be
cal
c
ulate
d from
the total runti
m
e abo
ve wh
e
re e
nerg
y i
s:
Th
us
,
for
e
ncry
pting
a
m
essage,
tra
ns
m
it
thro
ug
h
RF
sim
ulator
a
nd
dec
ry
pt
the
ci
ph
e
rtext
bac
k
with
a
key
siz
e
of
6144
-
bit,
t
he
e
ne
rg
y
is
ta
ke
n
4.8
4
J
oule
s.
Let
say
a
2200m
a
H
batte
ry
pac
k
is
co
nnect
ed
t
o
the
Rpi2,
the
R
pi2
can
sta
nd
up
to
30,
697.2
hour
s
if
t
he
encr
y
ption
is
perform
ed
cont
inu
ously
.
Whi
ch
is
equ
i
valent t
o 1
,279.05
days.
5.
CONCL
US
I
O
N
This
resear
ch
pr
ese
nted
a
n
inv
e
sti
gation
of
an
asym
m
et
ri
c
encr
ypti
on
s
chem
e,
the
AA
β
encr
ypti
on
schem
e
in
te
rm
s
of
enc
rypti
on
a
nd
dec
rypt
ion
ti
m
e
and
in
ad
diti
on,
the
RF
transm
issio
n
runtim
e
si
m
ula
ti
on
.
The
sim
ulati
on
on
Ra
s
pb
e
rr
y
Pi
2
run
ning
on
Ra
spbia
n
OS
has
res
ulted
in
82
.
6%
RF
sim
ulati
on
tra
ns
m
i
ssion
runtim
e
enh
an
ce
m
ent
com
par
ed
to
T
oshiba
L7
45
N
oteb
oo
k
with
W
i
ndows
10
O
S.
The
reas
on
f
or
slo
wer
RF
transm
issi
on
ru
ntim
e
si
m
u
la
t
ion
in
To
sh
i
ba
L745
N
oteb
ook
m
igh
t
du
e
to
def
a
ult
confi
gurati
on
on
windows
OS
us
e
d
e
ve
n
with
sam
e
ba
ud
rate
s
peed
f
or
bo
t
h
platfo
rm
s.
Nev
e
rtheless
,
this
dem
on
str
at
ed
that
it
is
vi
able
to
e
xp
l
or
e
the
AA
β
a
bili
ty
in
real
RF
co
m
m
un
ic
at
ion
s
for
IoT
du
e
to
faster
e
ncr
ypt
ion
tim
e
achieved
i
n
e
m
bed
de
d
de
vi
ces
an
d
only
14.2%
ci
ph
e
rtext
siz
e
increased
after
plain
te
xt
encr
ypte
d.
This
RF
si
m
u
la
ti
on
hav
e
zer
o
e
r
ror
as
it
is
in
a
c
ontr
olled
e
nv
i
ronm
ent
and
doe
s
not hav
e
d
at
a
loss or
delay
e
ven
up
f
or
20
0
tim
es
data
transm
issio
n
as
s
uggeste
d
by
Pu
tra
et
al
.
in
transm
i
tt
i
ng
the
data
in
wireless
co
ndit
ion
[
19]
.
In
a
r
eal
Io
T
dev
ic
e,
the
R
F
com
m
un
ic
at
ion
bet
ween
dev
ic
es
m
igh
t
ha
ve
c
halle
nges
i
n
tra
ns
m
issi
on
co
nf
i
gurati
on.
Nonetheless
,
t
his
sc
hem
e
may
be
be
ne
fici
al
to
low
po
w
ered
de
vices
t
o
e
ncr
ypt
t
he
data
with
a
sym
m
e
tric
encr
y
ption
s
ch
e
m
e
of
fe
rin
g
si
m
ple
m
at
h
cal
culat
ion
on
encr
y
ption,
a
dd
it
io
n,
m
ultip
li
cat
i
on
a
nd
sq
ua
r
e
com
par
ed
t
o
R
SA
wh
ic
h
requ
ired
m
od
ul
us
operati
on,
res
ulti
ng
a
ddit
ion
al
ov
e
r
head
i
n
e
nc
ryptin
g
a
m
essage.
The
f
uture
work
s
on
this
res
earch
ca
n
be
te
sti
ng
of
AA
β
encr
y
ption
sc
hem
e
on
real
R
F
transm
it
te
r,
W
i
fi
or
Bl
uetoo
t
h
co
nnect
ion,
de
pe
ndin
g
on
the
e
nvir
on
m
ent
for
the
I
oT
ta
king
place.
F
ur
t
herm
or
e,
the
sche
m
e
can
be
c
om
par
ed w
it
h
an
oth
e
r
asy
m
m
e
tric
sch
em
e su
c
h
as
RS
A and EC
C
on RF co
m
m
un
ic
at
ion
s
.
ACKN
OWLE
DGE
MENT
T
he
a
ut
ho
r
s
w
ou
l
d
l
i
ke
t
o
a
ck
n
ow
l
e
dg
e
t
he
M
i
ni
st
r
y
of
H
i
gh
e
r
E
du
c
a
t
i
on
(
M
O
H
E
)
M
a
l
ay
s
i
a
f
or
pr
o
vi
di
ng
t
he
g
r
a
nt
60
0
-
R
M
I
/
N
R
G
S
5/
3
(
5/
2
01
3)
a
nd
U
ni
v
e
r
s
i
t
i
T
e
kn
ol
og
i
M
A
R
A
(
U
i
T
M
)
f
or
s
u
pp
or
t
i
ng
t
hi
s
r
e
s
e
a
r
c
h
w
or
k.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
11
, N
o.
2
,
A
ugust
2018
:
542
–
548
548
REFERE
NCE
S
[1]
Hewle
tt Pac
k
ard
Enterpri
se
,
“
Int
ern
et of Things
Resea
rc
h
Stud
y
2015
Report,” n
o.
Jul
y
,
p.
4,
201
5.
[2]
Marga
re
t
Rous
e
and
Iv
y
W
igmore,
“
Inte
rne
t
of
Thi
n
gs
(IoT
),
”
Te
ch
Tar
get
.
[Online
]
.
Availab
le
:
htt
p://int
ern
e
tofthingsagenda.tec
hta
rge
t.
com/d
efiniti
on/In
te
rn
et
-
of
-
Thi
ngs
-
IoT. [
A
cc
essed:
29
-
Feb
-
2016]
.
[3]
K.
Rose,
S. E
ldridge,
and
.
L
y
m
an,
“
Th
e
in
te
rn
et of
th
ings: an
ov
erv
ie
w,
”
In
te
rne
t
Soc
.
,
no.
Octob
er,
p
.
53
,
20
15.
[4]
K.
J.
P.
Or
ti
z
,
J.
P.
O.
D
ava
lo
s,
E.
S.
Eusebio,
and
D.
M.
T
ucay
,
“
IoT:
El
e
ct
roc
ard
iogr
am
(E G)
m
onit
ori
ng
s
y
stem,”
Indone
s.
J
.
E
lectr.
Eng. Comput.
S
ci.
,
v
ol.
10
,
no
.
2
,
pp
.
480
–
489,
2018
.
[5]
M.
S.
M,
S.
Da
s,
S.
Heble,
U.
Raj
,
and
R.
Kar
thi
k,
“
Inte
rn
et
o
f
Thi
ngs
base
d
W
ire
le
ss
Plant
Sensor
for
Sm
ar
t
Farm
ing,
”
Indon
es.
J
.
Elec
tr.
En
g.
Comput.
Sc
i.
,
vol.
10
,
no
.
2
,
pp
.
456
–
468
,
2018
.
[6]
K.
N.
Q.
Raj
a
W
ase
em
Anwa
r,
Maj
id
Bakh
ti
ari,
Anaz
ida
Za
in
al
,
“
Secur
i
t
y
in
W
irele
ss
Sensor
Network :
Approac
hes and
Iss
ues,
”
TEL
KOMNIKA
Indone
s
.
J. Ele
c
tr.
Eng.
,
vol.
15
,
no
.
3
,
pp
.
584
–
590
,
2015
.
[7]
R.
Mahm
oud,
T.
Yous
uf,
F.
Aloul,
and
I
.
Zu
al
ke
rna
n,
“Interne
t
of
thi
ng
s
(
IoT
)
s
ec
urit
y:
Curr
ent
status,
challeng
es
and
pros
pec
tive
measur
es,
”
2015
10th
In
te
rna
ti
o
nal
Conf
ere
nc
e
f
or
Inte
rn
et
Tech
nolog
y
and
Secu
re
d
Tr
ansa
c
ti
ons
(ICIT
ST)
.
p
p.
33
6
–
341,
2015
.
[8]
J.
Mac
aul
a
y
,
L. Buc
kalew,
and
G.
hung,
“
Inte
r
net
of
Th
ings
in
Logi
stic
s
,
”
DH
L
Tr
end
Re
s
.
,
vol.
1,
no
.
1
,
pp
.
1
–
2
7
,
2015.
[9]
F.
K.
Santoso
and
N.
C.
H.
Vun,
“Securing
IoT
for
sm
a
rt
home
s
yste
m,”
Proc.
In
t.
S
y
m
p.
Consum
.
El
ec
t
ron.
ISCE
,
vol.
2015
–
Augu
s,
pp.
5
–
6,
2015.
[10]
B.
Mca
dams
,
“
RF
FU
N
DAMENT
ALS
for
the
INTERNET
of
THING
S,”
2017.
[Online
]
.
Available:
htt
ps://
o
le
um
tec
h.
com/wp
-
conte
nt/
uploa
ds/down
loa
ds/ref
ere
n
ce
s/
rf
-
funda
m
ent
al
s
-
for
-
iot
-
i
iot
-
ol
eu
m
te
ch.
pdf
.
[11]
F.
Li
and
P
.
Xi
ong,
“
Prac
tical
s
ec
ure
comm
unic
at
ion
for
int
eg
ra
ti
ng
wire
le
ss
sensor
net
works
int
o
the
int
ern
e
t
of
thi
ngs,”
IE
EE
S
e
ns.
J
.
,
vol. 13, n
o.
10
,
pp
.
3677
–
3684,
2013
.
[12]
M.
Singh,
M.
A.
Raj
an
,
V.
L
.
Sh
ivra
j
,
and
P.
B
alam
ura
li
dhar
,
“S
ec
ure
MQT
T
for
Inte
rne
t
of
Things
(
IoT
)
,
”
Proc.
-
2015
5th
In
t. Co
nf.
Com
m
un.
S
y
st.
Netw
.
T
ec
hn
ol.
CS
NT
2015
,
pp.
746
–
751
,
20
15.
[13]
X.
W
ang,
J.
Zhang,
E.
M.
Schoo
le
r,
and
M.
Ion,
“Performance
e
val
uati
on
o
f
Attribute
-
Based
Encr
ypt
ion:
Tow
ar
d
data
priva
cy i
n
t
he
IoT,”
2014
I
EE
E
Int
.
Conf
.
Com
m
un.
ICC 2
014
,
pp
.
725
–
73
0,
2014
.
[14]
S.
Raz
a
,
T.
Helga
son,
P.
Pa
padi
m
it
ra
tos,
a
nd
T.
Voigt,
“
Secur
eSense
:
E
nd
-
to
-
end
sec
ur
e
comm
unic
at
i
on
arc
hi
te
c
ture for t
he
c
loud
-
conn
ecte
d
In
te
rn
et
of
T
hings,
”
Fut
ur.
Gene
r.
Comput
.
S
yst.
,
vol
.
77
,
pp
.
40
–
51,
2017
.
[15]
M.
R.
K.
Ariff
in
,
M.
A.
As
bullah
,
N.
A.
Abu,
and
Z.
Mah
ad,
“
A
New
Eff
ic
i
ent
A
s
y
m
m
et
ric
r
y
pt
os
y
stem
Based
o
n
the
Int
ege
r
Fa
ctoriz
a
ti
on
Proble
m
of
N=P^
2
.
q,
”
Malay
sian
J.
Math.
Sc
i.
7(
S)
19
-
37
Spec.
Iss
ue
3rd
Int.
Con
f
.
Cryptol
.
Comput
.
Se
cur.
201
2
,
vo
l.
7
,
pp
.
1
–
6
,
201
2.
[16]
Z.
Mah
ad
and
M.
R.
K.
Ariff
i
n,
“AA
Be
ta
pub
li
c
key
cry
ptosy
stem
-
A
new
practical
asymm
e
t
ric
impl
eme
ntati
on
based
on
th
e
sq
uare
root
proble
m,”
in
Com
puting
and
Conv
erg
enc
e
T
ec
hnolog
y
(ICCCT),
201
2
7th
In
te
rn
at
io
nal
Confer
ence
on
,
2012,
pp
.
584
–
5
88.
[17]
M.
R.
K.
Ariff
i
and
Z
.
M,
“
AAβ
Publ
i
c
Ke
y
C
ryptosyste
m
–
A
Compar
ati
ve
A
naly
sis
Agai
nst
RSA
and
ECC
,
”
Com
put.
Conver
g.
T
ec
hno
l. (ICC
CT),
2012
7th
In
t.
Conf
.
,
pp.
589
–
594,
2012
.
[18]
The
GN
U
M
P
Bignum
Li
bra
r
y
,
“
The
GN
U
M
ult
iple
Prec
ision
Arithmeti
c
Li
b
ra
r
y
(GM
P)
Li
bra
r
y
.
”
[On
li
n
e]
.
Avail
ab
le
:
htt
ps:
//
gm
pli
b.
org
/. [Ac
ce
ss
ed: 25
-
Mar
-
2016]
.
[19]
G.
D.
Putra
,
A.
R.
Prat
ama,
A.
L
az
ovik
,
and
M.
Aiel
lo
,
“C
ompar
ison
of
ene
rgy
c
onsum
pti
on
in
W
i
-
Fi
and
bluetoot
h
communic
ati
on
in
a
Smar
t
Buil
ding,
”
in
2017
IEE
E
7
th
Annu
al
Com
puti
ng
a
nd
C
om
m
unic
at
ion
W
orkshop
and
Confer
ence
(CC
W
C)
,
2017,
pp
.
1
–
6.
Evaluation Warning : The document was created with Spire.PDF for Python.