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
.
5
,
Octo
ber
201
9
, pp.
4258
~
42
65
IS
S
N:
20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v
9
i
5
.
pp4258
-
42
65
4258
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
Senso
rs data c
oll
ec
ti
on fr
amewor
k usin
g mobi
le id
entific
ation
with sec
ure dat
a sharin
g model
Fula
yjan
Alan
az
i
1
, A
hmed
E
lhad
ad
2
, Sa
fw
at H
amad
3
and A
Gh
areeb
4
1,2
Depa
rtment
of
Com
pute
r
Sci
en
ce
and information
,
Facu
lty
of
S
ci
en
ce a
nd
Ar
t, J
ouf
Univer
sit
y
,
Saudi
Arabi
a
2
,4
Depa
rtment
of
Mathe
m
atics
an
d
Com
pute
r
Sci
e
nce
,
Facu
lty
o
f
S
ci
en
ce,
South V
a
ll
e
y
Un
ive
rsit
y
,
Eg
y
p
t
3
Depa
rtment of
Scie
nti
f
ic Com
puti
ng,
Facult
y
of
Com
pute
r and
I
nform
at
i
on
Sci
e
nce
s,
Ain
Sham
s Unive
rsit
y
,
Eg
ypt
4
Depa
rtment of
Mathe
m
at
i
cs,
Fa
cul
t
y
of
Sci
ence, Al
-
Baha Unive
r
sit
y
,
Saud
i
Arab
i
a
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Feb
7
, 2
01
9
Re
vised
A
pr
4
,
201
9
Accepte
d
Apr
25
, 201
9
Sensors
are
the
m
odule
s
or
el
ec
t
ronic
device
s
th
at
are
used
to
m
ea
sure
and
get
envi
ronm
ental
eve
n
ts
and
send
the
ca
ptur
ed
dat
a
to
oth
er
dev
ic
es,
usua
l
l
y
computer
proc
e
ss
ors
al
loc
at
ed
on
the
cl
oud.
One
of
the
m
ost
rec
en
t
cha
l
le
nges
is
to
prote
c
t
and
save
the
privac
y
issues
of
those
sensors
dat
a
on
the
cl
oud
shari
ng.
In
thi
s
pape
r,
sensors
dat
a
col
lect
ion
fra
m
ework
is
proposed
using
m
obil
e
ide
nt
ifi
c
at
ion
and
prox
y
re
-
enc
r
y
p
ti
on
m
odel
for
data
sharing.
The
pr
oposed
fra
m
ework
inc
lude
s:
id
e
nti
t
y
broke
r
ser
ver
,
sensors
m
ana
ging
and
m
onit
o
ring
applications,
m
essage
s
queui
ng
seve
r
and
dat
a
rep
ositor
y
serv
e
r.
Finall
y
,
the
e
xper
imental
res
ult
s
show
tha
t
t
he
proposed
prox
y
re
-
enc
r
y
pt
ion
m
odel
ca
n
work i
n
r
eal
t
ime.
Ke
yw
or
d
s
:
Data s
har
in
g
Pr
oxy
Re
-
enc
ryptio
n
Sensors
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
:
Fu
la
yja
n Alana
zi
,
Dep
a
rtm
ent o
f C
om
pu
te
r
Scie
nce a
nd in
form
at
ion
,
Faculty
of S
ci
e
nce a
nd Art,
A
lj
ouf
Unive
rsity
. A
l
Qurayy
at
, S
a
ud
i
A
rab
ia
.
Em
a
il
:
fu
la
yjan@j
u.
e
du.sa
1.
INTROD
U
CTION
Re
centl
y,
In
te
r
net
has
be
com
e
an
ind
is
pe
nsa
ble
com
m
un
ic
at
ion
ser
vice
in
our
li
fe
since
it
facil
i
ta
te
s
the
exch
a
nge
of
num
ero
us
ty
pes
of
dig
it
al
data
su
ch
as
te
xt,
i
m
ages,
aud
i
o
tracks
,
vi
deos
an
d
3D
grap
hical
obj
ect
s.
All
th
os
e
kinds
of
dig
it
al
data
are
st
or
e
d,
tran
sfe
rr
e
d
a
nd
s
har
e
d
a
m
on
g
di
ff
e
ren
t
us
e
rs
globall
y
us
i
ng
cl
oud
data
sh
a
rin
g
ser
vice
[1
-
7].
Cl
oud
data
sh
arin
g,
al
s
o
cal
le
d
on
li
ne
data
sh
ari
ng,
is
a
syst
e
m
in
wh
ic
h
us
ers
ca
n
sto
re
their
data
on
a
ser
ver
over
the
In
te
r
net
[8
]
.
F
ur
t
her
m
or
e,
cl
oud
data
s
ha
rin
g
al
lo
ws
use
rs
t
o
i
m
po
rt/e
xport
data b
et
wee
n
t
he web
ob
j
ect
stora
ge
a
nd m
ulti
ple d
evices
[
8].
Sensors
are
the
m
od
ules
a
nd
el
ect
ronic
dev
ic
es
that
a
re
use
d
to
m
easur
e
an
d
ge
t
data
f
ro
m
env
i
ronm
ental
even
ts
[
1].
T
he
proce
ss
of
recordi
ng
i
nfo
rm
ation
from
ph
ysi
cal
en
vir
on
m
ental
sens
or
s
on
t
o
cy
berworld is
cal
le
d
sens
ory
d
at
a colle
ct
io
n fr
am
ewo
r
k [9,
10
]
. Fu
rthe
rm
or
e, se
nsory
dat
a colle
ct
ion
is
on
e
of
the
th
ree
m
ai
n
co
nce
pts
that
gove
r
ns
t
he
c
onnecti
on
bet
ween
wi
reless
sens
or
netw
orks
(
WANs
)
a
nd
the
In
te
r
net
of Th
i
ng
s
(IoT
s)
[1
1, 12]
.
Ther
e
a
re
m
a
ny
pro
posed
s
ens
or
y
data
c
ollec
ti
on
te
ch
niques
[
12]
.
On
e
of
th
os
e
m
at
he
m
at
ic
a
l
m
od
el
ing
te
ch
niques
is
pro
pose
d
in
[13],
t
his
m
et
ho
d
is
base
d
on
a
sp
a
ti
al
-
cor
relat
io
n
w
her
e
pa
r
ti
al
sens
or
data
is
colle
ct
ed
an
d
Mult
iv
ariat
e
Gau
ssia
n
m
od
el
est
i
m
at
es
the
non
-
tr
ansm
itted
data.
The
aut
hors
in
[14]
pro
po
se
d
a
prob
a
bili
sti
c
te
c
hn
i
qu
e
cal
le
d
an
ene
r
gy
eff
i
ci
ent
k
-
c
overa
ge
al
gorithm
wh
ic
h
buil
d
a
cov
e
ra
ge
netw
ork
by
e
xpect
ing
the
m
i
nim
u
m
no
des
.
Re
centl
y,
com
pr
ess
ed
se
ns
or
y
data
colle
ct
i
on
te
ch
niques
becam
e
com
m
on
for
da
ta
recordi
ng
[
15
-
17
]
.
In
ge
ne
ral,
the
c
om
pressi
on
m
et
ho
ds
colle
ct
the
da
ta
from
the
sen
so
rs
in
a tim
e interval
and se
ns
ors
dat
a are sy
nchr
on
i
zed
[12].
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
Sensors
dat
a
c
ollec
ti
on
fr
am
e
work
us
in
g m
obil
e iden
ti
fi
cation w
it
h
sec
ur
e
da
t
a
…
(
F
ula
yj
an Ala
nazi
)
4259
In
t
his
pa
per,
s
ens
o
rs
data
col
le
ct
ion
fr
am
ew
ork
pr
opos
e
d
usi
ng
m
ob
il
e
id
entifi
cat
ion
a
nd
pro
xy
re
-
encr
y
ption
m
od
el
suppo
rts
th
e
sensory
data sh
ari
ng.
The
re
st
of
this
pa
pe
r
organ
iz
e
d
as
the
f
ollow
i
ng
:
s
ect
ion
II
is
the
syst
e
m
ov
er
view
,
S
ect
ion
I
II
is
e
ntit
ie
s
enrolm
e
nt
an
d
c
omm
un
ic
at
ion
sc
he
m
es,
sect
ion
I
V
is
the
pro
xy r
e
-
enc
ryption m
od
el
fo
r data
sh
a
rin
g
s
ect
ion
V
is
res
ults an
d dissect
ion
.
2.
THE
SYSTE
M
O
VE
RV
IE
W
The
pr
opos
e
d
f
ram
ewo
rk
ca
n
be
di
vid
e
d
into
four
m
ai
n
par
ts:
the
first
par
t
sta
nd
s
for
ide
nt
it
y
br
ok
e
r
serv
e
r
w
hic
h
r
efers
to
i
den
ti
ty
and
acce
ss
m
anag
em
ent
syst
e
m
;
the
second
sta
nds
for
the
se
nsors
m
anag
i
ng
and
m
on
it
ori
ng
ap
plica
ti
on
s;
the
third
pa
rt
sta
nd
s
for
the
m
essages
queu
ing
se
ver
w
hich
will
be
c
on
c
erne
d
m
ai
nly
by
the
Adva
nced
Me
s
sage
Qu
e
uing
Pr
ot
oc
ol
(A
M
QP
)
.
Fi
nally
,
al
l
data
will
be
store
d
in
beare
r
-
on
ly
acce
ss
data
re
posit
or
y
se
rv
e
r.
The
co
ntri
bu
ti
on
a
nd
relat
ionship
am
on
g
t
he
m
ai
n
par
ts
of
the
fr
am
ewo
r
k
in
a
si
m
plifie
d
way
is
sh
ow
e
d
in
Fig
ur
e
1.
I
n
wh
at
fo
ll
ow
s
we
prov
i
de
a
m
or
e
detai
le
d
desc
riptio
n
of
th
e
functi
onal
role
play
ed by eac
h o
ne of
these a
r
chite
ct
ur
al
c
om
po
nen
ts.
Figure
1. O
verview
diag
ram
2
.
1.
Iden
tit
y broker
ser
ver
An
i
den
ti
ty
brok
e
r
ser
ve
r
is
a
kind
of
an
i
den
ti
ty
and
ac
cess
m
anag
e
m
ent
(Id
AM)
s
ol
ution
wh
i
c
h
aim
s
to
pr
ovid
e
a
us
er
-
cent
ric
and
ce
ntrali
z
ed
way
f
or
m
a
nag
i
ng
i
den
ti
ti
es
acro
s
s
dif
fe
ren
t
process
dom
ai
ns
.
In
a
ddit
ion
,
it
was
a
m
idd
le
w
are
ser
vice
to
c
onnect
sec
ur
el
y
var
io
us
fr
am
ewor
k
ser
vice
pro
vid
er
s
with
us
ers
.
The
ide
ntit
y
bro
ker
sup
ports
oth
e
r
feat
ur
es
su
c
h
as:
Sin
gle
-
Sig
n
on
us
i
ng
m
ob
il
e
ide
ntific
at
ion
,
Ac
count
Ma
nag
em
ent,
Stand
a
r
d
P
ro
t
oco
ls
an
d
Au
t
horizat
ion
to
c
reate
a
tr
us
t
r
el
at
ion
sh
i
p
between
the
fr
am
ewor
k
m
ai
n
par
ts.
M
or
e
over,
a
n
e
xisti
ng
acc
ount
can
be
li
nke
d
with
one
or
m
or
e
i
den
ti
ti
es
f
ro
m
diff
ere
nt
identi
ty
pro
vid
er
s
or
e
ve
n
create
d
bas
ed
on
the
ide
ntit
y
inform
at
ion
ob
ta
ine
d
from
oth
er
prov
i
ders
su
ch
as
Face
book,
Goo
gle or T
wi
tt
er.
2
.
2.
Sens
ors
mana
ge
r
an
d
mon
it
orin
g ap
pli
cat
io
ns
The
pro
pose
d
fr
am
ewo
r
k
is
inclu
ding
two
web
a
pp
li
cat
io
ns
:
the
Senso
r
s
Ma
nag
e
r
ap
pl
ic
at
ion
will
be
us
e
d
f
or
cre
at
ing
a
new
se
ns
or
by
ad
ding
the
required
i
nfor
m
at
ion
for
the
syst
e
m
to
acqu
i
re
the
dat
a
fr
om
this
sens
or.
Th
e
require
d
inf
orm
ation
de
pe
nds
on
the
ty
pe
of
the
se
nsor
and
t
he
co
nne
ct
ion
ty
pe
w
hi
le
the
syst
e
m
gen
erates
a
un
iq
ue
I
D
fo
r
eac
h
se
nsor
.
In
t
he
sam
e
con
te
xt,
the
sens
or
m
anag
e
r
ap
plica
ti
on
e
nab
le
s
the
us
er
to
m
od
ify
the
sens
or
inform
ation
in
the
case
of
pa
ram
et
ers
chan
gi
ng
su
c
h
as
the
connecti
on
ty
pe
or
data
port.
O
n
t
he
ot
her
ha
nd,
the
ap
plica
ti
on
enab
le
s
the
use
r
to
delet
e
the
sensor
in
form
at
ion
in
the
ca
se
of
no
t
us
in
g
t
his s
ens
or
a
ny m
or
e. Ho
wev
e
r,
T
he
Se
nsor
Mo
ni
toring a
pp
li
cat
ion
is t
he gra
phic
al
u
se
r
inte
r
face to
disp
la
y t
he
v
a
r
iou
s
sen
s
or
data colle
ct
ion.
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.
5
,
Oct
ober
201
9
:
4
2
5
8
-
4
2
6
5
4260
2
.
3
.
Me
ssage
queuin
g
ser
ve
r
The
m
essage
Qu
e
uing
se
rv
e
r
is
a
ge
ner
al
l
y
serv
e
r
us
i
ng
Adva
nced
Me
ssage
Qu
e
uing
P
ro
t
oc
ol
(A
M
QP
)
wh
ic
h
aim
s
to
rece
ive
al
l
the
senso
rs
data
in
asy
nchron
ou
s
m
at
te
r.
Si
m
ply,
t
he
Me
ssage
Q
ueu
i
ng
serv
e
r
c
onsist
s
of
pro
ducer,
Queu
e
and c
ons
um
er.
The pr
oduce
r
is t
he
fi
r
st st
at
ion
to
r
ec
e
ive the
data f
r
om
the
var
i
ou
s
se
ns
or
s
the
n
it
will
be
t
ran
s
ferre
d
to
the
que
ue
wh
ic
h
fina
ll
y
delivered
it
to
the
c
ons
um
er.
The
be
nef
it
s
of
us
in
g
the
M
essage
Q
ueu
i
ng
se
rv
e
r
a
rch
it
ect
ur
e
is
to
s
upport
receivi
ng
m
any
data
from
diff
e
re
nt senso
rs
in
the
sam
e t
i
m
e. I
n
a
ddit
ion
, it al
lo
ws get
sens
or
s
d
at
a e
ve
n
the
user
offl
ine.
2
.
4
.
Sens
ors
mana
ge
r
an
d
mon
it
orin
g ap
pli
cat
io
ns
Data
Re
posit
ory
Ser
ver
use
s
a
we
b
se
r
vice
ap
plica
ti
on
t
o
prov
i
des
al
l
requests
of
da
ta
from
the
sens
or
s
m
on
it
ori
ng
ap
plica
ti
on
.
It
al
so
sto
re
s
the
data
c
ollec
te
d
f
ro
m
the
consum
er
in
the
Me
ssa
ge
qu
eui
ng
serv
e
r.
T
he
da
ta
in
the
re
posit
or
y
ser
ver
i
s
secu
red
fro
m
any
un
a
uthorize
d
acce
ss.
On
ly
an
d
on
l
y
if
the
request
with
va
li
d
tok
e
n
c
an
Get
/Post d
at
a i
n t
he
re
posit
or
y
serv
e
r.
3.
ENTIT
IES E
NROL
ME
NT
AND CO
M
M
UN
I
C
ATIO
N SCHE
MES
In
t
his
f
ram
ewo
r
k,
we
will
focus
m
ai
nly
on
dev
el
op
i
ng
sec
ur
el
y
sen
sors
da
ta
colle
ct
ion
,
m
on
it
or
ing
and
s
ha
rin
g
wi
th
the
al
lowa
nc
e
of
the
us
e
r.
The
pr
opos
e
d
f
ram
ewo
rk
c
on
sist
s
of
re
gister
ing
a
new
us
er
us
in
g
m
ob
il
e
identific
at
ion
by
creat
ing
a
stron
g
ke
y;
this
key
wi
ll
be
ke
pt
in
t
he
m
ob
il
e
fo
r
si
ng
i
ng
in
proce
dure.
In
t
he
sam
e
tim
e,
the
us
e
r
in
form
ation
detai
ls
will
be
store
d
in
t
he
ide
nt
it
y
br
oker
ser
ve
r
with
kee
ping
use
r
pr
i
vacy d
ur
i
ng
com
m
un
ic
at
ion
betwee
n fr
a
m
ewo
r
k
a
ppli
c
at
ion
s.
Figure
2
il
lustr
at
es
reg
ist
erin
g
a
new
us
er
in
the
propose
d
f
ram
ewo
rk.
Th
e
reg
ist
rati
on
proces
s
will
be
s
pecified
in
the user,
m
ob
il
e an
d
the
ide
ntit
y brok
e
r
se
rv
e
r,
where t
he pr
ocess wil
l be a
s the
fo
ll
owin
g:
Re
gisterin
g
: T
he user
w
il
l fil
l t
he
basic
data
su
c
h
as
First
N
a
m
e, Fam
il
y Na
m
e o
r
Em
ai
l.
Ver
ify
in
g
: T
he
iden
ti
ty
broke
r
will
bu
il
d a
ne
w
sessi
on
of
use
r publi
c k
ey
and enc
ode it
a
s a Q
R
code.
Confirm
ing
:
T
he
User
will
use
the
m
ob
il
e
QR
c
od
e
sca
nner
to
c
onfi
rm
his
dig
it
al
id
entit
y
us
in
g
t
hi
s
m
ob
il
e p
hone
.
Af
te
r
re
gistrati
on
ta
ke
place,
the u
ser
can
use
the
m
ob
il
e
ph
on
e
as
a
di
gital
identific
at
io
n; so
, no
us
er
nam
e
or
pass
w
ord
is
re
quire
d
to
sig
n
in.
C
onse
qu
e
ntly
,
the
sens
or
s
m
anag
em
ent
app
li
cat
ion
will
be
use
d
by
the
us
e
r
to
c
re
at
e,
edit
or
de
l
et
e
sens
or
s
i
nfor
m
at
ion
as
show
n
in
Fig
ur
e
3.
F
or
the
first
tim
e
of
loggin
g
in
t
he
sens
or
s
m
anag
er
ap
plica
ti
on,
the
us
e
r
is
a
ut
om
atical
ly
red
irect
ed
t
o
the
i
den
ti
ty
bro
ker
sever
f
or
getti
ng
the
log
i
n
in
f
or
m
ation
from
the mobile
phone.
Figure
2. Re
gi
s
te
r
ne
w use
r
Figure
3. Se
nsor
m
anag
ing
The
reg
ist
e
red
sens
or
,
se
nd
i
ng
the
data
of
t
he
c
urren
t
sta
te
thr
ough
t
he
s
el
ect
ed
cha
nne
l
directl
y
to
the
pro
ducer
of
t
he
m
essage
qu
e
uing
se
r
ver.
Wh
il
e
the
m
essage
que
uing
ser
ve
r
wi
ll
us
e
the
O
A
uth
2.0
protoc
ol
[
18
]
to
a
uthorize
a
nd
a
uth
e
ntica
te
the
se
ns
or
c
on
ta
ct
by
sen
sors
m
anag
er
a
pp
l
ic
at
ion
a
nd
ide
ntit
y
bro
ker
ser
ver
resp
ect
ively
.
Af
te
r
that,
the
m
essage
qu
e
ui
ng
ser
ver
c
onsu
m
er
delivers
the
data
to
the
data
reposit
ory
serv
er to be
sto
red.
Figure
4
s
ho
w
s
the
sens
or
s
m
on
it
or
ing
pro
cedure.
T
he
se
ns
ors
m
on
it
or
i
ng
a
ppli
cat
ion
requests
use
r
sign
in
us
i
ng
t
he
m
ob
il
e
iden
ti
ficat
ion
w
hile
the
identit
y
brok
e
r
ser
ve
r
aut
horizes
a
sin
gl
e
sign
on
duri
ng
the
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
Sensors
dat
a
c
ollec
ti
on
fr
am
e
work
us
in
g m
obil
e iden
ti
fi
cation w
it
h
sec
ur
e
da
t
a
…
(
F
ula
yj
an Ala
nazi
)
4261
browsi
ng
pr
oc
ess.
T
he
ide
ntit
y
br
oke
r
ser
ve
r
pr
ov
i
des
the
sens
or
m
on
it
or
ing
a
ppli
cat
ion
by
a
valid
to
ke
n
to
get the
re
qu
i
re
d data f
ro
m
the
d
at
a
reposit
or
y serv
e
r.
Figure
4. Se
nsors
m
on
it
or
in
g
The
act
ive
novel
scena
rio
i
n
the
pro
posed
fr
am
ewo
r
k
is
sh
ari
ng
sens
or
s
data
aut
horiz
at
ion
,
wh
ic
h
al
lows
t
he
own
er
of
t
he
data
to
s
har
e
them
with
a
re
quest
er
who
nee
ds
to
be
nef
it
of
t
his
da
ta
.
I
n
this
sc
e
nar
i
o,
the
requester
r
equ
e
sts
a
sp
eci
fic
us
er
se
nsor
/s
data
us
ing
t
he
sens
ors
m
o
nitor
i
ng
a
pp
li
c
at
ion
w
hich
re
directs
the
re
qu
e
ste
r
t
o
the
i
den
ti
ty
br
oker
se
r
ver
for
ide
nt
ific
at
ion.
A
fter
i
dent
ify
ing
the
requester,
the
ide
ntit
y
bro
ker
ser
ver
s
end
noti
ficat
io
n
to
the
owne
r’s
m
ob
il
e
to
re
quest
a
per
m
issio
n
f
or
ena
blin
g
the
requester
to
get
the
sens
or/
s
da
ta
.
The
owner
gr
a
nts
or
de
nie
s
this
request.
In
t
he
case
of
gr
a
ntin
g
the
re
qu
e
st,
the
own
er
ca
n
sel
ect
wh
ic
h
pr
ivil
eges
can
be
us
e
d.
Fi
nally
,
the
ide
ntit
y
br
oker
se
r
ver
a
uthorizes
t
he
re
quest
and
provide
s
th
e
sens
or
s
m
on
it
or
i
ng
ap
plica
ti
on
with
OAut
h
-
t
ok
e
n
t
o
get
the
require
d
data
f
r
om
the
data
re
posit
or
y
ser
ver.
Figure
5
il
lustr
at
es the e
ntit
ies
dia
gr
am
o
f
t
he
sh
a
rin
g data
auth
or
iz
at
io
n
s
cenari
o.
Figure
5. S
har
i
ng d
at
a a
utho
rizi
ng
4.
PRO
X
Y
R
E
-
ENC
RYPT
IO
N MO
DEL F
OR D
ATA S
HARIN
G
Hen
ce
,
we
pro
po
s
e
a
un
idi
rec
ti
on
al
pro
xy
re
-
enc
ryptio
n
m
od
el
.
T
he
pro
pose
d
m
od
el
al
lows
Alic
e
to
sh
are
en
crypte
d
plainte
xt
on
a
public
se
r
ver
with
Bo
b
us
in
g
B
ob’s
key
w
it
ho
ut
disclo
sing
Alic
e’s
e
nc
ryptio
n
key
to
Bo
b.
Fi
gure
6
s
hows
a
n
ove
rv
ie
w
of
the
pro
pose
d
m
od
el
fr
am
ewo
r
k.
In
Alic
e’s
sta
ti
on
,
Alic
e
create
s
the
p
la
inte
xt
a
nd
enc
rypts
it
us
in
g
Ke
y
1
.
T
hen,
s
he
se
nds
the
ci
pherte
xt
to
the
public
serv
e
r
via
the
pro
xy
sta
ti
on
.
T
her
e
f
or
e
,
Eve
-
the
eaves
droppe
r
-
cannot
recove
r
the
plainte
xt.
In
Bo
b’
s
sta
ti
on,
Bo
b
is
authorize
d
to
recove
r
Alic
e’s
plainte
xt,
s
o
he
reques
ts
t
he
ci
pherte
xt
f
ro
m
the
public
serv
e
r
via
the
pro
xy
sta
ti
on
a
nd
t
he
pro
xy
sta
ti
on
r
e
-
enc
rypts
the
ci
ph
e
rtext
us
i
ng
the
Re
key.
Finall
y,
Bob
de
crypts
the
Re
ci
ph
e
rtext
us
i
ng
his
own key(
Ke
y
2
)
. H
e
nce,
the
pr
ox
y ca
nnot
dec
rypt the
p
la
inte
xt du
rin
g
the
r
e
-
enc
ryptio
n process.
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.
5
,
Oct
ober
201
9
:
4
2
5
8
-
4
2
6
5
4262
Figure
6. The
pro
xy r
e
-
enc
ryption m
od
el
The
key
ge
ne
rator
sta
ti
on
pe
rfor
m
s
two
di
ff
ere
nt
kind
of
pr
ocedu
res
fo
r
gen
e
rati
ng
the
keys.
The
fi
rst
proce
dure
gen
e
rates
us
ers
’
keys
w
hi
ch
are
us
e
d
by
Alic
e
and
B
ob.
S
o,
the
f
ollow
i
ng
ste
ps
a
r
e
us
e
d
to g
e
ne
rate any
u
se
r key
:
Algori
th
m
1:
Key g
e
ne
rati
on
Step1
:
each
key co
ns
i
sts of uni
qu
e
64 bina
ry b
l
ocks each
b
l
ock in
cl
ud
es
6
-
bits.
Step2
:
each
blo
c
k divi
ded into
3 pa
rt
s each
p
a
rt is
one
of this set {
00, 01,
10, 11}
.
The
s
eco
nd
procedu
re
gen
e
r
at
es
the
pro
xy
Re
key
w
hich
i
s
de
pe
nd
i
ng
on
Key1
an
d
K
ey
2
of
Alic
e
a
nd
Bo
b
keys r
e
sp
ect
ive
ly
as the
f
ollo
wing:
Algori
th
m
2:
Re
-
Key
gen
e
ra
ti
on
Step1:
Fo
r
eac
h bloc
k i
n Key1:
ReKe
y
=
bl
oc
k
in
dex
(
Ke
y1
)
–
bl
oc
k
in
dex
(
Ke
y2
(
bl
oc
k
K
ey
1
)
)
Step2:
Convert
i
ntege
r
in
de
xes
to
b
i
nar
y
us
in
g 8
-
bi
t represe
ntati
on
.
So
,
Alic
e/
Bob
has
a
key
of
384
bi
nar
y
bits
and
Re
key
has
51
2
bi
nary
bits.
The
tri
ple
keys
ar
e
transm
itted
secur
el
y
to
Alic
e,
Bo
b
an
d
prox
y
sta
ti
ons.
The
Un
i
directi
on
al
c
rypto
syst
e
m
te
chn
iqu
e
incl
ud
e
three
proce
du
res:
E
ncr
ypti
on,
Decr
y
ptio
n
a
nd
Re
-
e
nc
ryptio
n.
The
Enc
ryptio
n
proce
dure
is
so
m
e
com
pr
ehe
ns
ive
ste
ps
of
c
od
i
ng
the
plai
ntext
and
enc
rypts
them
us
ing
Ali
ce’s
Key.
In
ad
diti
on
,
it
encl
ose
s
the
Cip
her
te
xt
a
nd
Plai
nte
xt
ar
e
the
sam
e
le
ng
t
h
of
strin
g.
T
he
f
ollo
wing
ste
ps
de
sc
ribes
t
he
e
ncry
ption
proce
dure:
Algori
th
m
3:
Encr
y
ption
Step1:
Re
ad
the
Plai
nt
ext
Step2:
C
onve
rt the P
la
inte
xt t
o bina
ry u
si
ng
8
-
bit re
pr
ese
nt
at
ion
.
Step3:
Divid
e
the
bi
na
ry Plai
ntext
s
equ
e
nce i
nto 6
bits bl
ock
s
.
Step4:
Fo
r
eac
h bloc
k o
f
the
b
i
nar
y
Plai
ntext calc
ul
at
e:
_
=
(
⨁
_
=
0
)
−
1
Step5:
Convert t
he
i
nt_
in
dex
vecto
r f
r
om
integer
t
o bina
ry u
si
ng 6
-
bit re
pr
ese
ntati
on.
Step6:
Convert t
he bi
nar
y se
quence
to strin
g usi
ng
8
-
bit re
pr
ese
nt
at
ion
to
g
et
the
Ci
ph
e
rtext.
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
Sensors
dat
a
c
ollec
ti
on
fr
am
e
work
us
in
g m
obil
e iden
ti
fi
cation w
it
h
sec
ur
e
da
t
a
…
(
F
ula
yj
an Ala
nazi
)
4263
In
t
he
sam
e
con
te
xt,
t
he
decry
ption
proce
dure
is
the
i
nv
e
r
se
of
the
e
ncry
ption
process
us
in
g
t
he
sam
e
ke
y
.
The follo
wing
ste
ps
descr
i
bes
the
decr
y
ption p
ro
ce
dure:
Algori
th
m
4:
Decr
y
ption
Step1:
Re
ad
the
Ci
ph
ertext
Step2:
Convert t
he
Ci
ph
e
rtext t
o bin
ary us
i
ng 8
-
bit represe
ntati
on
.
Step3:
Divid
e
the
bi
na
ry Ci
ph
e
rtext
seq
uen
ce
into
6 bit
s b
l
ock
s
.
Step4:
Convert eac
h b
inary
blo
c
k
int
o
inte
ger.
Step5:
Fo
r
eac
h
inte
ge
r value
of the
vecto
r
cal
culat
e:
_
=
(
_
+
1
)
Step6:
Convert t
he
bi
n_
s
eq
f
ro
m
b
in
ary to st
rin
g us
ing
8
-
bit re
pres
entat
ion
t
o get
the Plai
nte
xt.
Finall
y,
the
Re
-
enc
ryptio
n
pr
oc
ed
ur
e
t
akes
place
wh
e
n
Bo
b
is
delegated
by
Alic
e
to
decr
y
pt
the
Ci
ph
e
rtex
t.
So
,
the
prox
y
re
-
e
ncr
y
pt
s
the
Ci
ph
e
r
te
xt
us
in
g
t
he
corres
ponding
Re
key
a
nd
the
fo
ll
owin
g proc
edure:
Algori
th
m
5:
Re
-
Enc
ryptio
n
Step1:
Re
ad
the
Ci
ph
ertext.
Step2:
Convert t
he
Ci
ph
e
rtext t
o bin
ary us
i
ng 8
-
bit represe
ntati
on
.
Step3:
Divid
e
the
bi
na
ry Ci
ph
e
rtext
seq
uen
ce
into
6 bit
s b
l
ock
s
.
Step4:
Convert eac
h b
inary
blo
c
k
int
o
inte
ger.
Step5:
Fo
r
eac
h
inte
ge
r value
of the
vecto
r
cal
culat
e:
_
ℎ
=
_
ℎ
−
(
_
ℎ
+
1
)
Step6:
Convert t
he
i
nt_
Reci
pher
vect
or from
integer
to bina
ry u
si
ng
6
-
bit re
pr
ese
ntati
on
.
Step7:
Convert t
he bi
nar
y se
quence
to strin
g usi
ng
8
-
bit re
pr
ese
nt
at
ion
to
g
et
the
Re
Ci
ph
erte
xt.
The
unidirect
i
on
al
e
nc
ryptio
n
prox
y
m
od
e
l
is
def
i
ned
as
a
tu
ple
E=
(
K
eyGen
,
E
nc
,
Dec
,
ReEnc
)
the
key
ge
ne
ra
ti
on
pr
ocedure
Ke
yGen
ge
nerat
es
keys
f
or
A
li
ce
and
Bo
b.
The
n,
it
ge
ner
a
te
s
on
e
m
or
e
ke
y
fo
r
the
pro
xy.
Alic
e
encr
ypts
the
plainte
xts
us
i
ng
the
Enc
pr
oc
edure
an
d
dec
r
ypts
them
us
ing
the
Dec
proc
edure.
Wh
e
ne
ver
Bo
b
wa
nts
to
de
crypt
the
ci
phertext,
He
as
ks
the
prox
y
f
or
hel
p,
the
pro
xy
us
es
the
ReEnc
proce
dure
to
t
ran
s
f
or
m
the
ci
ph
e
rtext
into
diff
e
re
nt
ci
ph
e
rtext
an
d
se
nd
s
it
to
Bob.
Bob
a
ppli
es
the
Dec
proce
dure
to
the
receive
d
ci
ph
e
rtext
an
d
ge
ts
the
or
igi
nal
plainte
xt.
Th
e
fo
ll
owin
g
eq
ua
ti
on
re
flect
s
the
way
of the
pro
pose
d un
i
directi
ona
l
m
od
el
wo
rk for
the
re
-
e
ncr
y
ption p
r
ocess:
ℎ
=
1
(
)
→
ℎ
=
(
ℎ
)
→
2
(
ℎ
)
=
5.
RESU
LT
S
AND DI
SSE
CTION
The
pr
opos
e
d
fr
am
ewo
r
k
was
ap
pointe
d
a
seve
ral
op
e
n
s
ource
li
br
aries
an
d
app
li
cat
io
ns
.
The
I
den
ti
ty
Broker
Se
r
v
er
was
us
e
d
a
Ke
yc
loak
ser
ver
-
Ver
si
on
:
3.0
.0
[
19
]
.
The
Me
ssage
Q
ue
uing
Se
rv
e
r
was
us
e
d
Ra
bbit
MQ
serv
e
r
-
Ver
si
on
:
3.6
.9
[
20
]
.
T
he
pro
pose
d
pro
xy
re
-
encr
y
ption
m
od
el
was
i
m
plem
ented
to
eval
uate
th
e
pe
rfor
m
ance
of
the
pr
opose
d
m
et
ho
d.
F
ur
t
her
m
or
e,
t
he
plainte
xt
co
m
po
sed
of
ra
ndom
ly
sel
ect
ed
data
t
o
produce
va
riou
s
file
siz
e
st
arti
ng
from
1
KB
ti
ll
100
K
B.
I
n
ad
diti
on,
the
sim
ulati
o
n
was
i
m
ple
m
ented
usi
ng
In
te
l(R
)
Core
(TM)
i7
-
6700MQ
CP
U
,
3.4
0
G
Hz,
64
-
bit
wind
ow
s
10
operati
ng
syst
e
m
with
16GB
of
RAM an
d
M
A
TLAB
versi
on: 9.0.
0.3
41360 (
R201
6a)
.
Tw
o
seeds
we
r
e
us
ed
to
ge
ne
rate
rando
m
keys
fo
r
Alic
e
an
d
Bob.
The
pr
ox
y
Re
key
wa
s
gen
e
rated
b
y
usi
ng
Alic
e
an
d
B
ob
keys.
T
he
e
xec
utio
n
ti
m
e
per
f
orm
ances
are
0.
0003
sec
an
d
0.000
2
sec
to
gen
e
rate
Alic
e
an
d
Bo
b
keys
res
pecti
vely
.
H
oweve
r
,
the
re
-
e
ncr
y
ption
key
generati
on
e
xec
ution
ti
m
e
is
0.
01
sec.
Figure
7
sho
w
s
the
exec
utio
n
tim
e
of
the
pr
opos
e
d
E
ncr
y
pt
ion
pr
ocedure
for
va
rio
us
plainte
xts
us
i
ng
A
li
ce’s
key.
As
no
te
d,
the
com
pu
te
d
tim
e
increases
sign
ific
a
ntly
by
increasin
g
t
he
file
siz
e.
T
hus,
to
e
ncr
y
pt
a
1K
B,
10KB
an
d
100KB
file
s
ta
ke
about
0.0
6,
0.59
a
nd
5.4
1
se
conds
resp
e
ct
ively
.
H
ow
e
ve
r
,
Fig
ur
e
8
s
hows
the
pro
po
se
d
Re
-
Encr
y
ption
pr
ocedu
re
exec
ution
ti
m
e
c
os
t
f
or
t
he
corres
pondin
g
plainte
xts
file
siz
e.
As
s
how
n,
th
e
re
-
e
ncr
y
ption
process
require
s
a
li
tt
le
t
i
m
e
e
xecu
ti
on
w
hile
it
ta
ke
about
1.0
6,
4.65
an
d
45.
70
m
illi
secon
ds
to
enc
rypt
a
1K
B,
10KB
and
100KB
r
especti
vely
w
hich
is
com
patible
with
pro
xy
re
-
e
ncr
y
ption.
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.
5
,
Oct
ober
201
9
:
4
2
5
8
-
4
2
6
5
4264
Figure
7. Enc
r
ypti
om
p
ro
ce
dure
ex
ec
utio
n
t
i
m
e
Figure
8. Re
-
e
ncr
y
ption p
ro
c
edure e
xec
utio
n
ti
m
e
Hen
ce
,
the
De
crypti
on
proce
dure
will
be
a
ppli
ed
for
bo
t
h
t
he
Ci
pher
a
nd
ReC
iph
er
us
i
ng
Alic
e
an
d
Bob
Keys
re
spe
ct
ively
.
Figu
r
e
9
sh
ows
the
decr
y
ption
pro
cedure
exec
ution
ti
m
e
fo
r
Ali
ce
and
Bo
b
w
hi
ch
are
the
sam
e
fo
r
the
va
rio
us
file
s
siz
e.
Th
us
,
it
will
ta
ke
1.
70
,
9.6
2
an
d
93.
63
m
il
l
ise
cond
s
for
Alic
e
to
decr
y
pt
a 1KB,
10KB
a
nd 10
0k
B
m
e
ssage fi
le
s whi
le
it
w
il
l t
ake 0.96, 6.
62 a
nd 79.00 m
i
ll
ise
con
ds f
or Bo
b.
Figure
9
.
D
ec
r
ypti
om
p
ro
ce
dure
ex
ec
utio
n
t
i
m
e
6.
CONCL
US
I
O
N
In
this
pa
per,
we
pr
opos
e
sens
or
s
data
colle
ct
ion
f
r
a
m
ewo
r
k
usi
ng
m
ob
il
e
identific
at
ion
.
The
pr
opos
e
d
f
ram
ewo
rk
s
up
ports
data
sh
a
ri
ng
us
in
g
a
pro
xy
re
-
e
ncr
y
ption
m
od
el
.
T
he
execu
ti
on
ti
m
e
of
the
pro
po
se
d
m
od
el
to
encr
ypt
a
1K
B,
10K
B
and
100KB
file
s
ta
ke
abo
ut
5.0
6,
0.5
9
an
d
5.41
s
econd
s
resp
ect
ively
.
Wh
il
e,
the
re
-
e
ncr
y
ption
proc
ess
requires
a
li
ttle
tim
e
exec
ution
wh
il
e
it
t
ake
ab
ou
t
1.0
6,
4.65
and
45.
70
m
i
llisec
onds
t
o
en
crypt
a
1KB,
10KB
a
nd
10
0KB
resp
ect
ively
wh
ic
h
is
com
patible
with
pr
oxy
re
-
encr
y
ption.
ACKN
OWLE
DGE
MENTS
The
sec
ond
au
thor
w
ould
li
ke
to
e
xpress
hi
s
sincere
tha
nks
to
prof
es
sors
Luís
An
t
unes
and
Ma
nuel
Correi
a, CRA
CS &
INESC
TEC
-
Porto
, Por
t
ug
al
for hel
p
a
nd s
upport.
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
Sensors
dat
a
c
ollec
ti
on
fr
am
e
work
us
in
g m
obil
e iden
ti
fi
cation w
it
h
sec
ur
e
da
t
a
…
(
F
ula
yj
an Ala
nazi
)
4265
REFERE
NCE
S
[1]
M.
Arm
brust,
A.
Fox,
R.
Grif
fith,
A.
D.
Jos
eph
,
R.
Ka
tz,
A.
K
onwinski
,
et
a
l.
,
"A
vie
w
of
cl
o
ud
computing,
"
Comm
unic
ati
ons of
the A
CM,
vol
.
53
,
pp
.
50
-
58
,
2010.
[2]
M.
Arm
bru
st,
A
.
Fox,
R.
Griffi
th,
A.
D.
Jos
eph,
R.
H.
Katz
,
A.
Konw
inski
,
et
al.
,
"A
bove
the
clouds
:
A
ber
kele
y
vie
w
of
c
loud
c
om
puti
ng,
"
Tec
hnic
a
l
Report
UCB/EE
CS
-
2009
-
28,
EE
CS
Dep
a
rtment,
Univ
ersi
t
y
o
f
Californi
a,
Berke
l
e
y
2009
.
[3]
A.
Fu,
S.
Yu,
Y
.
Zha
ng
,
H.
W
an
g,
and
C
.
Huang
,
"N
PP
:
A
New
Privacy
-
Aw
are
Public
Audit
ing
Scheme
for
Clou
d
Data
Shar
ing
wi
t
h
Group Users,"
IEE
E
Tr
ansacti
o
ns on
Bi
g
Data,
2017.
[4]
M.
M.
Hass
an,
K.
Li
n
,
X.
Yue
,
and
J.
W
an,
"A
m
ult
imedia
he
al
t
hca
re
d
ata
shari
ng
a
pproa
ch
thr
ough
cl
oud
-
base
d
bod
y
area
n
et
wo
rk,
"
Fu
ture
Gen
e
ration
Computer
Syste
ms
,
vol. 66
,
pp
.
48
-
58
,
201
7.
[5]
J.
Shen,
T.
Zho
u,
D.
He,
Y.
Zh
ang,
X.
Sun,
an
d
Y.
Xiang,
"Bl
ock
design
-
base
d
ke
y
agr
ee
m
en
t
for
group
dat
a
sharing
in
c
loud com
puti
ng,
"
I
EEE
Tr
ansa
ct
ions
on
Depe
ndabl
e and Se
cure
Com
puti
ng,
2017.
[6]
M.
Sun,
C.
Ge,
L.
Fang,
and
J.
W
ang,
"A
prox
y
broa
dca
st
r
e
-
en
cr
y
p
ti
on
for
c
lou
d
dat
a
sharing
,
"
Mult
imedi
a
Tool
s
and
Applications,
pp
.
1
-
15
,
2017
.
[7]
J.
Li
,
Y
.
Zha
ng
,
X.
Chen,
and
Y.
Xiang,
"S
ec
u
re
attribute
-
bas
e
d
dat
a
shar
ing
f
or
resourc
e
-
l
imit
ed
users
in
cl
ou
d
computing,
"
Co
mputers
&
Sec
urity,
vol
.
72
,
pp
.
1
-
12,
2018
.
[8]
T.
Gali
bus,
V.
V.
Krasnopros
hin,
R.
de
Olive
ir
a
Albuquer
que,
and
E.
P.
de
Freitas,
El
eme
n
ts
of
cl
oud
storage
sec
urity
:
con
ce
p
ts,
designs
and
o
pti
mized
prac
ti
c
es
:
Springer
,
201
6.
[9]
C.
Kons
ta
ntopo
ulos,
G.
Pan
tz
io
u,
D.
Gav
alas,
A.
Mpitziopoulo
s,
and
B
.
Mam
alis,
"A
ren
de
zvo
us
-
base
d
appr
oa
ch
ena
bl
ing
ene
rg
y
-
eff
icient
senso
r
y
da
ta
co
llecti
on
with
m
obil
e
sinks,"
IEE
E
Tr
ansacti
ons
o
n
Parall
el
and
Distrib
ute
d
S
yst
ems,
vol
.
23
,
pp
.
809
-
817,
2012.
[10]
P.
Madhum
at
h
y
and
D.
Siv
akumar,
"En
abling
en
erg
y
eff
ic
i
ent
se
nsor
y
da
ta
colle
ct
ion
using
m
ultiple
m
obile
sink
,
"
China
Comm
unic
ati
ons,
vol
.
11
,
pp.
29
-
37
,
2014
.
[11]
A.
Al
-
Fuqaha,
M.
Guiza
ni
,
M.
Moham
m
adi
,
M.
Aledha
r
i,
a
n
d
M.
A
y
y
ash
,
"
Inte
rne
t
of
th
in
gs:
A
surve
y
o
n
ena
bl
ing
te
chno
logi
es,
proto
cols
,
and
appl
i
ca
t
ions,"
IEE
E
C
omm
unic
ati
ons
Surve
ys
&
Tut
orials,
vol.
1
7
,
pp.
2347
-
2376
,
2015.
[12]
S.
Cheng, Z.
C
ai,
and
J.
Li,
"A
pp
roximate
Sensor
y
D
at
a
Coll
ectio
n:
A Surve
y
,
"
Se
nsor
s,
vol.
17
,
p
.
564,
2017.
[13]
B.
Gedik
,
L
.
L
iu
,
and
S.
Y.
Phil
i
p,
"A
SA
P:
An
ada
pti
v
e
sam
pli
n
g
appr
oa
ch
to
d
a
ta
co
ll
e
ction
in
s
ensor
net
works
,
"
IEE
E
Tr
ansacti
o
ns on
Parallel
a
nd
distribut
ed
sy
stems,
vol
.
18
,
p
p.
1766
-
1783
,
2
007.
[14]
C.
Li,
Z.
Sun,
H.
W
ang,
and
H.
S
ong,
"A
novel
e
ner
g
y
-
eff
i
cient
k
-
Covera
ge
al
gor
i
thm
base
d
on
pr
obabi
lit
y
driv
en
m
ec
hani
sm
of
wire
le
ss
sensor
net
works
,
"
Int
ernati
onal
Journal
of
Distribu
te
d
Sensor
Ne
t
works,
vol.
1
2
,
p
p
.
1
-
9
,
2016
.
[15]
S.
Li
,
L.
Da
Xu,
and
X.
W
ang,
"
Com
pre
ss
ed
sensing
signal
and
dat
a
ac
quisi
ti
on
in
wire
l
ess
sensor
net
works
and
int
ern
et
of
th
ing
s,"
IEEE
Tr
ansacti
ons on
Industrial
In
formatic
s,
vol.
9
,
pp
.
2177
-
2186,
2013
.
[16]
Y.
Ta
ng
,
B
.
Zh
ang,
T.
Jing,
D.
W
u,
and
X
.
Cheng,
"R
obust
compress
ive
d
at
a
gat
h
eri
ng
in
wire
l
ess
sensor
net
works
,
"
IE
EE Tr
ansacti
ons on
Wireless
Comm
unic
ati
ons
,
vol
.
12,
pp
.
2754
-
27
61,
2013
.
[17]
M.
T.
Ngu
y
en
a
nd
K.
A.
Teague
,
"Com
pre
ss
ive
sensing
base
d
r
a
ndom
walk
routing
in
wire
le
ss
se
nsor
net
works
,
"
Ad
Hoc
N
e
twork
s,
vol. 54, pp. 99
-
110,
2017
.
[18]
D.
Hardt
,
"Th
e OA
uth
2.
0
aut
ho
riz
a
ti
on
fr
amework,"
2012.
[19]
Key
c
loak
.
Avai
l
abl
e
:
ht
tp:
/
/www
.
ke
y
c
loa
k
.
org/i
n
dex.
html
[20]
Rabbi
tMQ
-
Me
s
saging
that just wor
ks
.
Availabl
e
:
htt
ps:
//
ww
w.rabbitmq.c
om
/
Evaluation Warning : The document was created with Spire.PDF for Python.