I
nte
rna
t
io
na
l
J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
9
,
No
.
2
,
A
p
r
il
201
9
,
p
p
.
120
1
~
1
2
0
8
I
SS
N:
2088
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
9
i
2
.
pp
1
2
01
-
1
2
0
8
1201
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
co
r
e
.
co
m/
jo
u
r
n
a
ls
/in
d
ex
.
p
h
p
/
I
JE
C
E
Ela
stic neural n
et
w
o
rk
m
ethod
f
o
r
lo
a
d predict
io
n
in cloud
co
m
pu
ti
ng
grid
K
ef
a
y
a
S.
Q
a
dd
o
u
m
1
,
Na
m
e
er
N.
E
l E
m
a
m
2
,
M
o
s
leh.
A
Abua
lha
j
3
1
D
e
p
a
rtme
n
t
o
f
Co
m
p
u
ter In
f
o
rm
a
ti
o
n
S
y
ste
m
s
,
Hig
h
e
r
Co
ll
e
g
e
s o
f
T
e
c
h
n
o
lo
g
y
,
Un
it
e
d
A
ra
b
Em
ira
tes
2
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter S
c
ien
c
e
,
P
h
il
a
d
e
lp
h
ia Un
iv
e
rsity
,
Un
it
e
d
S
tate
3
De
p
a
rtme
n
t
o
f
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
,
Am
m
a
n
A
l
A
h
li
y
y
a
Un
iv
e
rsit
y
,
Yo
rd
a
n
ia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Oct
10
,
2
0
1
7
R
ev
i
s
ed
Sep
2
1
,
2
0
1
8
A
cc
ep
ted
Oct
1
1
,
2
0
1
8
Clo
u
d
c
o
m
p
u
ti
n
g
stil
l
h
a
s
n
o
st
a
n
d
a
rd
d
e
f
in
it
io
n
,
y
e
t
it
is
c
o
n
c
e
rn
e
d
w
it
h
In
tern
e
t
o
r
n
e
tw
o
rk
o
n
-
d
e
m
a
n
d
d
e
li
v
e
r
y
o
f
re
so
u
rc
e
s
a
n
d
se
rv
ic
e
s.
It
h
a
s
g
a
in
e
d
m
u
c
h
p
o
p
u
larity
in
las
t
f
e
w
y
e
a
rs
d
u
e
to
ra
p
id
g
ro
w
th
i
n
t
e
c
h
n
o
l
o
g
y
a
n
d
th
e
In
ter
n
e
t.
M
a
n
y
issu
e
s
y
e
t
to
b
e
tac
k
led
w
it
h
in
c
lo
u
d
c
o
m
p
u
ti
n
g
tec
h
n
ica
l
c
h
a
ll
e
n
g
e
s,
su
c
h
a
s
V
ir
tu
a
l
M
a
c
h
i
n
e
m
ig
ra
ti
o
n
,
se
rv
e
r
a
ss
o
c
iatio
n
,
f
a
u
lt
to
lera
n
c
e
,
sc
a
lab
il
it
y
,
a
n
d
a
v
a
il
a
b
il
it
y
.
T
h
e
m
o
st
we
a
re
c
o
n
c
e
rn
e
d
w
it
h
in
th
is
re
se
a
rc
h
is
b
a
lan
c
in
g
se
rv
e
rs
lo
a
d
;
th
e
w
a
y
o
f
sp
re
a
d
in
g
th
e
lo
a
d
b
e
tw
e
e
n
v
a
rio
u
s
n
o
d
e
s
e
x
ists
in
a
n
y
d
istri
b
u
ted
sy
ste
m
s
th
a
t
h
e
l
p
to
u
ti
li
z
e
re
so
u
rc
e
a
n
d
jo
b
re
sp
o
n
se
t
im
e
,
e
n
h
a
n
c
e
sc
a
lab
il
it
y
,
a
n
d
u
se
r
s
a
ti
sf
a
c
ti
o
n
.
L
o
a
d
re
b
a
lan
c
in
g
a
lg
o
rit
h
m
w
it
h
d
y
n
a
m
ic
r
e
so
u
rc
e
a
ll
o
c
a
ti
o
n
is
p
r
e
se
n
ted
to
a
d
a
p
t
w
it
h
c
h
a
n
g
in
g
n
e
e
d
s
o
f
a
c
lo
u
d
e
n
v
iro
n
m
e
n
t.
T
h
is
re
se
a
rc
h
p
re
se
n
ts
a
m
o
d
if
ied
e
la
stic
a
d
a
p
ti
v
e
n
e
u
ra
l
n
e
tw
o
rk
(E
A
N
N)
w
it
h
m
o
d
if
ie
d
a
d
a
p
ti
v
e
s
m
o
o
th
i
n
g
e
rro
rs,
to
b
u
il
d
a
n
e
v
o
lv
in
g
sy
ste
m
to
p
re
d
ict
Virtu
a
l
M
a
c
h
in
e
lo
a
d
.
T
o
e
v
a
lu
a
te
th
e
p
ro
p
o
se
d
b
a
lan
c
in
g
m
e
th
o
d
,
w
e
c
o
n
d
u
c
ted
a
se
ries
o
f
sim
u
latio
n
stu
d
ies
u
sin
g
c
lo
u
d
si
m
u
lato
r
a
n
d
m
a
d
e
c
o
m
p
a
riso
n
s
w
it
h
p
re
v
io
u
sly
su
g
g
e
st
e
d
a
p
p
r
o
a
c
h
e
s
in
th
e
p
re
v
io
u
s
w
o
rk
.
T
h
e
e
x
p
e
rim
e
n
tal
re
su
lt
s
sh
o
w
th
a
t
su
g
g
e
ste
d
m
e
th
o
d
b
e
tt
e
rs
p
re
se
n
t
a
p
p
r
o
a
c
h
e
s
sig
n
if
ica
n
tl
y
a
n
d
a
ll
t
h
e
se
a
p
p
r
o
a
c
h
e
s.
K
ey
w
o
r
d
s
:
C
lo
u
d
co
m
p
u
tin
g
L
o
ad
b
alan
cin
g
Neu
r
al
n
et
w
o
r
k
s
Vir
tu
al
m
ac
h
in
e
m
i
g
r
atio
n
Co
p
y
rig
h
t
©
2
0
1
9
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts
re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Kef
a
y
a
Qad
d
o
u
m
,
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
I
n
f
o
r
m
atio
n
S
y
s
te
m
s
,
Hig
h
er
C
o
lleg
e
s
o
f
T
ec
h
n
o
lo
g
y
,
T
r
af
f
ic
Dep
ar
t
m
en
t R
o
ad
,
A
l
A
i
n
,
U
n
ited
A
r
ab
E
m
ir
ates
.
E
m
ail: K
.
s
.
Qad
d
o
u
m
@
g
m
ai
l.c
o
m
1.
I
NT
RO
D
UCT
I
O
N
A
cc
o
r
d
in
g
to
th
e
Natio
n
al
I
n
s
tit
u
te
o
f
Stan
d
ar
d
s
a
n
d
T
ec
h
n
o
lo
g
y
(
NI
ST
)
:
"
C
lo
u
d
co
m
p
u
ti
n
g
i
s
a
m
o
d
el
f
o
r
en
ab
li
n
g
co
n
v
en
ie
n
t,
o
n
-
d
e
m
a
n
d
n
et
w
o
r
k
ac
ce
s
s
to
a
s
h
ar
ed
p
o
o
l
o
f
co
n
f
i
g
u
r
ab
le
co
m
p
u
ti
n
g
r
eso
u
r
ce
s
[
1
]
.
C
lo
u
d
i
s
a
h
u
g
e
s
ca
le
d
is
tr
ib
u
ted
co
m
p
u
t
in
g
,
tr
an
s
f
er
r
in
g
co
m
p
u
ter
o
p
er
atio
n
s
an
d
s
to
r
ag
e
f
r
o
m
d
esk
to
p
an
d
p
o
r
tab
le
P
C
s
to
a
h
u
g
e
h
o
s
ti
n
g
ce
n
ter
[
1
]
,
w
h
e
r
e
r
eso
u
r
ce
s
co
u
ld
b
e
ea
s
il
y
u
s
ed
r
e
m
o
tel
y
,
w
i
th
ef
f
ec
tiv
e
co
s
t
[
2
]
,
[
3
]
.
Data
ce
n
ter
s
u
s
e
Vir
t
u
al
Ma
c
h
i
n
e
(
VM
)
f
o
r
in
teg
r
atio
n
o
f
d
if
f
e
r
en
t
s
o
l
u
tio
n
s
[
4
]
.
I
n
f
r
as
tr
u
ct
u
r
e
-
as
-
a
-
Ser
v
ice
(
I
aa
S),
P
latf
o
r
m
-
as
-
a
-
S
er
v
ice
(
P
aa
S),
an
d
So
f
t
w
ar
e
-
a
-
a
-
Ser
v
ice
(
SaaS)
ar
e
th
e
o
f
f
er
ed
t
y
p
es o
f
m
ar
k
et
s
er
v
ic
es a
m
o
n
g
clo
u
d
.
Ma
n
y
m
ill
io
n
s
o
f
o
n
li
n
e
d
ev
i
ce
s
ar
e
co
n
n
ec
ted
th
r
o
u
g
h
t
h
e
I
n
ter
n
et,
w
h
er
e
u
s
i
n
g
clo
u
d
co
m
p
u
ti
n
g
s
er
v
ice
i
n
d
u
s
tr
ies
is
i
n
cr
ea
s
i
n
g
ev
er
y
d
a
y
[
4
]
.
C
lo
u
d
co
m
p
u
ti
n
g
i
n
v
o
lv
es
s
h
ar
i
n
g
s
u
p
p
li
es,
s
o
f
t
w
ar
e,
w
h
ich
s
u
p
p
o
s
ed
to
r
ed
u
ce
d
f
u
n
ctio
n
in
g
co
s
t,
r
e
g
ar
d
in
g
th
e
t
i
m
e
n
ee
d
ed
to
p
r
o
ce
s
s
d
ata,
ad
j
u
s
t
th
e
s
y
s
te
m
p
er
m
a
n
e
n
ce
ac
co
r
d
in
g
to
d
em
an
d
s
ch
a
n
g
es
i
n
t
h
e
s
y
s
te
m
as
it
g
o
es.
L
o
ad
b
alan
ce
r
s
ar
e
p
ar
t
o
f
clo
u
d
s
tr
u
ct
u
r
e;
it
s
g
o
al
i
s
to
f
o
r
w
a
r
d
in
co
m
i
n
g
r
eq
u
est
to
t
h
e
c
u
s
to
m
e
r
w
h
o
d
e
m
a
n
d
s
t
h
e
s
er
v
ice.
D
ep
en
d
i
n
g
o
n
av
ailab
ilit
y
a
n
d
e
x
is
t
in
g
lo
ad
,
th
e
lo
ad
b
ala
n
ce
r
tas
k
is
to
ap
p
ly
s
c
h
ed
u
l
in
g
a
lg
o
r
it
h
m
to
d
e
cid
e
th
e
b
e
s
t
s
er
v
er
th
at
co
u
ld
h
a
n
d
le
th
e
r
eq
u
e
s
t
[
5
]
.
I
t
g
ets
an
u
p
d
ate
o
n
clien
t
’
s
s
er
v
er
'
s
s
tat
u
s
an
d
cu
r
r
en
t
j
o
b
.
I
f
o
n
e
s
er
v
er
is
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
2
,
A
p
r
il
201
9
:
1
2
0
1
-
1
2
0
8
1202
o
v
er
lo
ad
ed
w
it
h
ta
s
k
s
,
t
h
e
n
i
t
is
r
ed
ir
ec
ted
to
an
o
t
h
er
s
er
v
e
r
to
k
ee
p
lo
ad
b
ala
n
ce
d
[
6
-
7
]
.
I
n
th
i
s
p
ap
er
,
th
e
o
b
j
ec
tiv
e
o
f
lo
ad
s
ch
ed
u
lin
g
i
s
m
i
n
i
m
izi
n
g
t
h
e
av
er
ag
e
s
h
i
f
t
ti
m
e
o
f
a
s
et
o
f
o
n
lin
e
lo
ad
s
,
w
h
er
e
s
h
i
f
ti
n
g
ti
m
e
r
ep
r
esen
ts
th
e
to
tal
elap
s
ed
tim
e
f
o
r
a
lo
ad
a
p
p
licatio
n
f
r
o
m
lo
ad
s
u
b
m
i
s
s
io
n
to
lo
ad
c
o
m
p
letio
n
,
i
n
clu
d
i
n
g
w
ait
in
g
ti
m
e
an
d
e
x
ec
u
t
io
n
ti
m
e.
As
elab
o
r
ated
in
[
8
]
,
lo
ad
co
m
p
u
ti
n
g
i
s
a
g
o
o
d
w
a
y
to
u
n
d
er
s
tan
d
j
o
b
p
ar
allelis
m
w
h
er
e
ea
ch
j
o
b
r
eq
u
ir
es
a
s
i
n
g
le
p
r
o
ce
s
s
o
r
f
o
r
ex
ec
u
tio
n
a
n
d
d
i
f
f
er
e
n
t
ta
s
k
s
m
i
g
h
t
r
u
n
co
n
cu
r
r
e
n
tl
y
o
r
f
o
llo
w
s
u
p
er
io
r
it
y
r
estrictio
n
s
.
O
n
t
h
e
o
th
er
s
i
d
e,
p
ar
allel
p
r
o
g
r
am
s
ar
e
w
r
itten
w
it
h
s
o
m
e
m
e
s
s
a
g
e
-
p
ass
i
n
g
l
ib
r
ar
y
,
e.
g
.
,
MP
I
[
9
]
,
ar
e
a
class
ic
w
a
y
to
a
cc
o
m
p
li
s
h
d
ata
p
ar
allelis
m
.
A
lo
ad
m
o
d
el
ca
n
b
e
u
s
ed
to
d
escr
ib
e
th
e
s
c
h
ed
u
lin
g
j
o
b
to
k
ee
p
b
alan
ce
a
m
o
n
g
t
h
e
cl
o
u
d
VM
s
.
As
th
e
h
i
g
h
-
p
er
f
o
r
m
a
n
ce
co
m
p
u
ti
n
g
p
latf
o
r
m
e
v
o
l
v
es
i
n
to
g
r
id
an
d
clo
u
d
en
v
ir
o
n
m
e
n
ts
,
t
h
e
u
n
d
er
l
y
i
n
g
s
p
ee
d
-
h
eter
o
g
e
n
eo
u
s
m
u
lti
-
cl
u
s
ter
d
esig
n
[
1
0
]
m
a
k
es
m
ix
ed
-
p
ar
allel
lo
ad
s
ch
ed
u
li
n
g
m
o
r
e
lab
o
r
io
u
s
an
d
d
if
f
ic
u
lt
t
h
an
tr
ad
itio
n
al
tas
k
-
p
ar
allel
lo
ad
s
s
ch
ed
u
li
n
g
b
ec
au
s
e
o
f
t
h
e
r
eso
u
r
ce
d
iv
is
io
n
i
s
s
u
e
[
1
1
]
ex
p
er
ien
ce
d
b
y
p
ar
allel
tas
k
d
is
tr
ib
u
tio
n
.
I
n
[
1
2
]
,
On
li
n
e
L
o
ad
Ma
n
ag
e
m
e
n
t
w
as
p
r
o
p
o
s
ed
as
th
e
f
ir
s
t
s
tep
to
tack
li
n
g
s
u
ch
p
r
o
b
le
m
s
.
AN
N
is
w
id
el
y
u
s
ed
in
d
i
f
f
er
en
t
asp
ec
ts
an
d
ap
p
licatio
n
s
[
1
7
]
,
in
th
i
s
w
o
r
k
,
w
e
p
r
o
p
o
s
e
an
E
last
ic
A
r
ti
f
icia
l
Neu
r
al
Net
w
o
r
k
(
E
A
NN)
f
r
a
m
e
w
o
r
k
,
w
h
ic
h
f
o
r
m
a
li
ze
s
th
e
o
n
li
n
e
-
lo
ad
s
ch
ed
u
lin
g
p
r
o
ce
s
s
i
n
to
j
o
b
a
ll
o
ca
tio
n
.
T
h
e
j
o
b
-
clu
s
ter
in
g
p
h
ase
m
an
a
g
es
th
e
j
o
b
lo
ad
w
it
h
i
n
a
s
er
v
er
.
T
h
e
j
o
b
r
ea
r
r
an
g
e
m
en
t
p
h
ase
allo
w
s
s
o
m
e
j
o
b
s
r
ea
llo
ca
tio
n
s
to
d
if
f
er
e
n
t
s
er
v
er
s
to
h
an
d
le
t
h
e
r
eq
u
est
to
in
cr
ea
s
e
r
eso
u
r
ce
u
tili
za
tio
n
o
n
o
t
h
er
s
er
v
er
s
.
T
h
e
jo
b
allo
ca
tio
n
p
h
ase
ass
i
g
n
s
a
p
r
o
p
er
s
et
o
f
r
eso
u
r
ce
s
to
a
j
o
b
.
W
e
liv
e
i
n
t
h
e
c
lo
u
d
er
a,
w
h
er
e
u
s
er
s
o
f
th
e
clo
u
d
ar
e
i
n
cr
ea
s
in
g
f
as
t,
t
h
at
ca
u
s
ed
h
ea
v
y
j
o
b
o
n
clo
u
d
s
er
v
er
s
w
h
ic
h
m
ad
e
t
h
e
lo
a
d
b
alan
cin
g
as
an
u
r
g
en
t
is
s
u
e
o
f
clo
u
d
co
m
p
u
ti
n
g
.
T
h
e
co
r
e
g
o
al
o
f
lo
ad
b
alan
cin
g
is
to
allo
ca
te
th
e
j
o
b
co
n
s
is
ten
tl
y
to
th
e
w
h
o
le
cl
o
u
d
to
g
u
ar
an
tee
n
o
o
v
er
lo
ad
ed
o
r
u
n
d
er
lo
ad
ed
s
er
v
er
s
e
x
is
ts
i
n
t
h
e
e
n
tire
n
et
w
o
r
k
[
1
3
]
.
L
o
ad
b
alan
ci
n
g
al
g
o
r
ith
m
s
'
g
o
al
i
s
to
m
o
n
ito
r
th
e
n
o
d
es
f
o
r
u
n
d
er
lo
ad
in
g
o
r
o
v
er
lo
ad
in
g
an
d
tak
e
ap
p
r
o
p
r
iate
ac
tio
n
to
n
o
r
m
alize
t
h
e
lo
ad
a
m
o
n
g
VM
s
.
So
th
at
u
tili
za
t
io
n
o
f
r
eso
u
r
ce
s
is
b
et
ter
.
Ma
n
y
lo
ad
co
n
tr
o
llin
g
te
ch
n
iq
u
es,
li
k
e
A
SK
AL
ON
[
1
4
]
,
DA
G
m
a
n
[
1
5
]
,
h
av
e
d
e
v
elo
p
ed
s
y
s
te
m
s
to
m
an
ag
e
lo
ad
u
s
es
in
b
o
th
p
ar
allel
an
d
d
is
tr
ib
u
ted
s
y
s
te
m
s
.
S
o
m
e
ex
i
s
ti
n
g
lo
ad
m
an
a
g
e
m
e
n
t
s
y
s
t
e
m
s
,
e.
g
.
,
DAG
m
a
n
a
n
d
[
1
6
]
,
m
ak
e
s
u
r
e
th
e
ex
ec
u
tio
n
o
f
a
lo
ad
d
o
es
n
o
t
v
io
late
t
h
e
p
r
ec
ed
en
ce
co
n
s
tr
ain
t
s
a
m
o
n
g
tas
k
s
b
u
t
p
a
y
litt
le
atte
n
tio
n
to
i
m
p
r
o
v
i
n
g
s
ch
ed
u
li
n
g
ti
m
e.
Oth
er
s
y
s
te
m
s
f
o
cu
s
o
n
co
m
p
leti
n
g
ta
s
k
s
ch
e
d
u
lin
g
w
it
h
i
n
lo
ad
s
to
en
h
a
n
ce
d
esire
d
f
in
al
ti
m
e,
e.
g
.
[
1
7
]
.
T
h
e
r
est
o
f
t
h
e
p
ap
er
co
v
er
s
.
Sectio
n
2
d
is
c
u
s
s
es
r
elate
d
w
o
r
k
o
f
lo
ad
b
alan
c
in
g
;
Sectio
n
3
d
is
cu
s
s
es
th
e
p
r
o
p
o
s
ed
s
y
s
te
m
,
in
cl
u
d
i
n
g
L
o
ad
b
alan
cin
g
,
s
u
g
g
e
s
te
d
f
r
a
m
e
w
o
r
k
,
alg
o
r
ith
m
to
p
r
ed
ict
clo
u
d
lo
a
d
,
s
er
v
er
’
s
class
if
ica
tio
n
u
s
ed
i
n
C
lo
u
d
co
m
p
u
ti
n
g
.
Sectio
n
4
s
h
o
w
s
s
i
m
u
latio
n
a
n
d
r
esu
lt
s
o
f
th
e
s
u
g
g
ested
s
y
s
te
m
.
Sectio
n
5
h
a
s
th
e
co
n
c
lu
s
io
n
w
it
h
s
o
m
e
id
ea
s
f
o
r
f
u
t
u
r
e
w
o
r
k
.
2.
RE
S
E
ARCH
M
E
T
H
O
D
O
L
O
G
Y
E
x
p
lain
i
n
g
r
esear
ch
c
h
r
o
n
o
lo
g
ical,
in
c
lu
d
i
n
g
r
esear
c
h
d
esi
g
n
,
r
esear
c
h
p
r
o
ce
d
u
r
e
(
in
th
e
f
o
r
m
o
f
alg
o
r
ith
m
s
,
P
s
e
u
d
o
co
d
e
o
r
o
th
er
)
,
h
o
w
to
test
an
d
d
ata
ac
q
u
i
s
itio
n
[
1
8
]
,
[
1
9
]
.
T
h
e
d
escr
ip
ti
o
n
o
f
th
e
co
u
r
s
e
o
f
r
esear
ch
s
h
o
u
ld
b
e
s
u
p
p
o
r
ted
r
ef
er
en
ce
s
s
o
th
at
t
h
e
ex
p
la
n
ati
o
n
ca
n
b
e
ac
ce
p
ted
s
cien
ti
f
ical
l
y
[
2
0
]
,
[
2
1
]
.
W
h
en
e
v
er
th
er
e
i
s
a
ce
r
tai
n
j
o
b
o
n
th
e
C
lo
u
d
,
th
e
lo
ad
b
alan
ce
r
s
e
n
d
s
t
h
e
j
o
b
r
eq
u
est
th
r
o
u
g
h
t
h
e
s
u
g
g
e
s
ted
al
g
o
r
ith
m
to
b
e
cl
u
s
ter
ed
to
w
h
ic
h
s
er
v
er
to
b
e
s
e
n
t
[
2
2
]
.
K
-
m
ea
n
s
cl
u
s
ter
in
g
i
s
o
n
e
o
f
t
h
e
s
i
m
p
lest
clu
s
ter
i
n
g
alg
o
r
it
h
m
s
u
s
ed
f
o
r
u
n
s
u
p
er
v
i
s
ed
lea
r
n
i
n
g
p
r
o
b
le
m
s
,
Ma
c
Qu
ee
n
,
1
9
6
7
in
tr
o
d
u
ce
d
it,
it
w
o
r
k
s
m
ai
n
l
y
o
n
m
in
i
m
izin
g
th
e
d
i
s
tan
ce
b
et
w
ee
n
t
h
e
cl
u
s
ter
ce
n
ter
a
n
d
t
h
e
d
ata
p
o
i
n
t.
K
-
m
ea
n
s
is
u
s
ed
i
n
t
h
i
s
r
esear
ch
ap
p
o
in
tin
g
j
o
b
s
as
d
ata
p
o
in
ts
o
n
th
e
s
er
v
er
an
d
b
ased
o
n
th
e
p
r
o
ce
s
s
o
r
ty
p
e
an
d
s
p
ee
d
to
clu
s
ter
th
ese
j
o
b
s
.
A
f
ter
d
eter
m
i
n
i
n
g
h
ar
d
d
is
k
ca
p
ac
it
y
,
a
n
o
th
er
cl
u
s
ter
i
n
g
is
d
o
n
e
b
ased
o
n
t
h
ei
r
co
s
t
p
er
h
o
u
r
.
A
n
o
p
er
ato
r
n
o
d
e
to
m
an
a
g
e
n
o
d
e
p
er
f
o
r
m
a
n
ce
ex
i
s
ts
,
it
s
j
o
b
is
to
s
u
p
er
v
is
e
t
h
e
n
o
d
e
th
e
n
r
ep
o
r
t
u
tili
za
tio
n
an
d
s
en
d
it
to
m
a
n
a
g
er
co
n
tr
o
ller
,
th
e
m
a
n
ag
er
i
s
a
cl
u
s
ter
m
a
n
ag
e
m
e
n
t
s
y
s
te
m
ai
m
ed
at
d
e
cr
ea
s
in
g
t
h
e
e
f
f
o
r
t
d
o
n
e
to
m
a
n
ag
e
clu
s
ter
r
eso
u
r
ce
s
w
h
i
le
r
ef
i
n
i
n
g
t
h
e
f
ac
ilit
y
o
f
cu
s
to
m
er
'
s
d
e
m
a
n
d
an
d
m
an
a
g
e
r
e
s
o
u
r
ce
s
.
T
h
e
m
a
n
ag
er
w
ait
s
u
n
til
th
e
w
h
o
le
clu
s
ter
is
e
m
p
t
y
b
ef
o
r
e
lo
o
k
in
g
f
o
r
w
o
r
k
f
r
o
m
o
th
er
clu
s
ter
s
.
Fig
u
r
e
1
[
7
]
s
h
o
w
s
s
y
s
te
m
m
o
d
el,
w
h
er
e
th
e
m
a
n
ag
er
co
n
tr
o
ller
s
en
d
s
a
r
e
p
o
r
t
as
in
p
u
t
w
h
ic
h
is
u
s
ed
in
p
r
ed
ictio
n
alg
o
r
ith
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
E
la
s
tic
n
eu
r
a
l n
etw
o
r
k
meth
o
d
fo
r
lo
a
d
p
r
ed
ictio
n
in
cl
o
u
d
co
mp
u
tin
g
g
r
id
(
K
efa
ya
S
.
Qa
d
d
o
u
m
)
1203
Fig
u
r
e
1
.
L
o
ad
b
alan
ce
r
2
.
1
.
E
ANN
des
ig
n f
o
r
lo
a
d pre
dict
io
n
T
o
f
o
r
ec
ast
ex
p
ec
ted
r
eso
u
r
ce
r
eq
u
ests
,
w
e
n
ee
d
to
lo
o
k
in
s
id
e
a
VM
an
d
an
al
y
ze
lo
g
s
o
f
p
en
d
in
g
r
eq
u
ests
.
T
o
d
o
th
i
s
,
VM
s
m
o
d
if
icatio
n
s
ar
e
r
eq
u
ir
ed
,
w
h
ic
h
m
i
g
h
t
b
e
d
i
f
f
icu
l
t.
T
h
e
p
r
ed
ictio
n
i
s
b
u
ilt
o
n
t
h
e
h
is
to
r
ical
lo
g
b
eh
a
v
io
r
s
o
f
V
Ms.
T
h
e
alg
o
r
ith
m
d
e
p
lo
y
s
a
n
ar
tif
icial
n
e
u
r
al
n
et
w
o
r
k
to
p
r
ed
ict
lo
ad
b
alan
cin
g
in
t
h
e
clo
u
d
;
th
i
s
is
ac
h
ie
v
ed
b
y
tr
ain
i
n
g
n
eu
r
al
n
et
w
o
r
k
o
n
a
b
ig
d
ataset
co
n
tai
n
i
n
g
d
i
f
f
e
r
en
t
lo
ad
s
it
u
atio
n
s
.
B
ac
k
P
r
o
p
ag
atio
n
lear
n
in
g
al
g
o
r
ith
m
w
a
s
u
s
ed
d
u
r
in
g
E
ANN
tr
ain
i
n
g
p
h
ase
s
o
th
a
t
it
c
an
m
an
a
g
e
i
n
co
m
i
n
g
j
o
b
s
to
th
e
clo
u
d
.
T
h
e
s
u
g
g
e
s
t
ed
E
A
NN
co
n
s
i
s
t
o
f
t
h
r
ee
tier
s
;
t
h
e
f
ir
s
t
tier
is
i
n
p
u
t,
p
r
ese
n
tin
g
t
h
e
cu
r
r
e
n
t
j
o
b
f
o
r
N
n
o
d
es.
T
h
e
s
ec
o
n
d
is
t
h
e
h
id
d
en
tier
,
w
h
ile
t
h
e
th
ir
d
is
o
u
tp
u
t
tier
,
w
h
ic
h
s
ig
n
i
f
ies
t
h
e
b
alan
ce
d
j
o
b
f
o
r
N
n
o
d
es.
E
ac
h
n
o
d
e
in
t
h
e
in
p
u
t
tier
ac
t
f
o
r
eit
h
er
th
e
c
u
r
r
en
t
s
er
v
er
's
j
o
b
o
r
th
e
cu
r
r
en
t
av
er
ag
e
j
o
b
o
f
a
clu
s
ter
o
f
s
er
v
er
s
w
h
er
e
its
g
iv
e
n
a
n
u
m
b
er
.
On
t
h
e
o
th
er
h
an
d
,
th
e
eq
u
iv
a
len
t
n
o
d
e
in
th
e
o
u
tp
u
t
tier
r
ep
r
esen
ts
eit
h
er
s
er
v
er
’
s
j
o
b
o
r
clu
s
ter
’
s
r
eg
u
lar
j
o
b
af
ter
b
al
an
cin
g
co
r
r
esp
o
n
d
in
g
l
y
.
T
h
e
s
y
s
te
m
p
r
o
p
o
s
ed
is
a
n
i
n
t
ellig
e
n
t
tech
n
iq
u
e,
w
h
er
e
th
e
v
alu
e
s
o
f
t
h
e
n
e
u
r
al
n
et
w
o
r
k
co
n
n
ec
tio
n
w
ei
g
h
ts
ar
e
m
o
d
i
f
ied
w
ith
in
tr
ain
in
g
p
h
a
s
e,
u
s
i
n
g
a
r
ef
o
r
m
ed
o
p
ti
m
izat
io
n
m
eth
o
d
o
lo
g
y
;
it
co
n
s
i
s
ts
o
f
b
u
ild
in
g
an
elast
ic
ar
ti
f
icial
n
eu
r
al
n
et
w
o
r
k
(
E
A
NN)
b
ased
o
n
m
o
d
if
ied
ad
ap
tiv
e
s
m
o
o
th
i
n
g
er
r
o
r
s
(
MA
SE)
.
T
h
en
to
ad
j
u
s
t
th
e
n
eu
r
al
n
et
w
o
r
k
co
n
n
ec
tio
n
w
eig
h
t
s
d
u
r
i
n
g
t
h
e
tr
ain
i
n
g
p
h
ase.
T
h
e
alg
o
r
ith
m
u
s
es
E
A
N
N
tech
n
iq
u
e
w
o
r
k
o
n
r
ed
u
ci
n
g
er
r
o
r
s
.
First,
w
e
n
ee
d
to
d
eter
m
in
e
a
v
er
ag
e
w
i
th
w
ei
g
h
ts
b
y
a
p
r
ed
eter
m
in
ed
s
ch
e
m
e,
w
h
er
e
t
h
e
ex
p
ec
ted
an
d
d
etec
ted
lo
ad
at
a
g
iv
en
p
er
io
d
p
as seen
in
(
1
)
.
D(
p
)
=
c*
D(
p
-
1
)
+
(
1
-
c)
*
O(
p
)
; 0
<
c
<
1
(
1
)
as
c
in
d
icate
s
,
t
h
e
b
alan
ce
a
m
o
n
g
s
tead
i
n
es
s
an
d
ap
p
r
o
ac
h
ab
ilit
y
.
E
A
NN
is
u
s
ed
to
p
r
ed
ict
th
e
l
o
ad
o
f
C
P
U
o
n
a
d
ef
i
n
ed
s
er
v
er
.
W
e
g
r
o
u
p
th
e
ti
m
e
c
=0
.
5
an
d
u
s
e
it t
o
p
r
ed
ict
th
e
n
ex
t
m
i
n
u
te
lo
ad
an
d
co
m
p
ar
e
it
w
i
th
t
h
e
cu
r
r
e
n
t
m
in
u
te.
T
h
e
n
ex
t
s
tep
is
m
o
n
ito
r
i
n
g
d
ec
r
ea
s
ed
u
s
ag
e;
w
e
n
ee
d
to
w
o
r
k
o
n
t
h
e
o
b
s
er
v
ed
v
al
u
es
w
h
ich
ar
e
h
ig
h
er
b
y
6
9
% n
o
r
m
a
ll
y
.
I
f
b
alan
ce
v
a
lu
e
s
g
r
o
u
p
s
w
it
h
in
[
0
to
1
]
,
th
en
th
e
b
ala
n
ce
i
s
w
it
h
in
t
h
e
d
esire
d
r
an
g
e.
W
e
g
iv
e
c
-
1
r
ate,
s
o
th
at
v
al
u
e
w
e
g
et
f
r
o
m
eq
u
atio
n
(
1
)
is
co
n
v
er
ted
as (
2
)
h
er
e.
D(
p
)
=
-
|
c
|
*
D(
p
-
1
)
+
(
1
+
|
c
|
)
*
O(
p
)
;
-
1
<
c
<
0
=
O(
p
)
+
|
c
|
*
(
O(
p
)
-
D(
p
–
1
*
c)
)
(
2
)
T
h
e
p
r
e
d
icted
v
alu
e
is
co
n
cl
u
d
ed
b
ased
o
n
h
is
to
r
ical
b
eh
av
io
r
s
o
n
th
e
clo
u
d
.
2
.
2
.
E
ANN
a
rc
hite
ct
ure
T
h
e
s
u
g
g
e
s
ted
s
y
s
te
m
i
n
cl
u
d
es
th
r
ee
tier
s
,
n
eu
r
o
n
s
i
n
t
h
e
f
ir
s
t,
s
ec
o
n
d
,
a
n
d
th
ir
d
tier
s
.
E
A
N
N
as
s
h
o
w
n
in
F
ig
u
r
e
2
g
o
es a
s
f
o
ll
o
w
s
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
2
,
A
p
r
il
201
9
:
1
2
0
1
-
1
2
0
8
1204
Fig
u
r
e
2
.
E
last
ic
A
NN
ar
ch
ite
ctu
r
e
P
h
ase
1
:
R
eg
r
o
u
p
n
eu
r
o
n
a
n
d
allo
ca
te
n
eu
r
o
n
s
w
i
th
d
i
v
er
s
e
r
o
u
tes lo
c
atio
n
s
.
P
h
ase
2
:
g
r
o
u
p
a
lo
ca
tio
n
r
o
u
te
Xs
o
f
s
er
v
er
n
o
d
e
n
as
a
n
in
p
u
t.
Si
n
c
e
th
er
e
ar
e
th
r
ee
n
e
u
r
o
n
s
i
n
t
h
e
in
p
u
t tier
,
Xs
s
h
o
u
ld
b
e
th
e
in
p
u
t o
f
d
if
f
er
en
t
n
e
u
r
o
n
o
f
i
n
p
u
t tier
s
ea
ch
ti
m
e.
P
h
ase
3
: Cre
ate
a
g
r
o
u
p
B
r
elate
s
s
er
v
er
n
o
d
e
s
co
r
r
esp
o
n
d
s
to
th
e
n
eu
r
o
n
n
.
S =
{i
|
Xi
∈
Sq
∩
k
Xs −
Xi
k
≤
r
i}
(
3
)
w
h
er
e
Xi
is
t
h
e
lo
ca
tio
n
r
o
u
te
o
f
n
e
u
r
o
n
n
a
n
d
th
e
as
s
o
ciatio
n
s
r
elate
d
to
Sq
ca
n
s
h
ield
s
er
v
er
n
o
d
e
n
.
P
h
ase
4
:
I
f
g
r
o
u
p
S is
u
n
f
i
lled
,
iter
ate
th
is
p
r
o
ce
d
u
r
e
b
y
p
ic
k
i
n
g
a
n
o
t
h
er
r
an
d
o
m
s
er
v
er
n
o
d
e.
Xn
∗
=
m
i
n
k
X
s
−X
n
k
.
(
4
)
n
∈
Sq
P
h
ase
5
:
alter
th
e
ch
o
s
e
n
s
i
tes o
f
n
eu
r
o
n
an
d
n
e
x
t o
n
e
s
ac
co
r
d
in
g
to
:
Xn
∗
=X
n
∗
+
α
(
Xs−X
n
)
(
5
)
E
ls
e,
g
o
b
ac
k
to
s
tep
(
2
)
.
T
h
e
m
ea
n
s
q
u
ar
ed
er
r
o
r
(
MSE
)
b
etw
ee
n
th
e
p
r
ed
icted
an
d
th
e
r
eq
u
ir
ed
r
esu
lt
r
ep
r
esen
ts
th
e
f
it
n
e
s
s
o
f
th
e
i
n
p
u
t.
(
)
(
)
(
6
)
z
=
(
t
z
−O
z
)
(
(
)
)
(
7
)
(
)
(
8
)
T
h
e
f
ac
to
r
δ
z
an
d
th
e
ac
tiv
atio
n
h
t
o
f
th
e
h
id
d
en
n
eu
r
o
n
H
t
,
d
eter
m
in
e
s
ad
j
u
s
t
m
e
n
t
m
ad
e
to
th
e
v
alu
e
o
f
V
st
an
d
er
r
o
r
f
ac
to
r
s
b
e
o
u
tlin
ed
in
(
8
)
.
3.
RE
SU
L
T
S
A
ND
AN
AL
Y
SI
S
I
n
th
i
s
s
ec
tio
n
,
it
is
e
x
p
lai
n
ed
th
e
r
esu
lts
o
f
r
esear
ch
an
d
at
th
e
s
a
m
e
ti
m
e
is
g
iv
e
n
t
h
e
co
m
p
r
e
h
en
s
iv
e
d
i
s
cu
s
s
io
n
.
T
h
e
ex
p
er
i
m
en
t
s
w
er
e
co
n
d
u
ct
ed
o
n
a
clo
u
d
s
i
m
u
lato
r
.
I
n
c
lo
u
d
s
i
m
u
lato
r
,
th
e
clo
u
d
s
i
m
,
t
h
at
aid
s
m
o
d
eli
n
g
,
an
d
s
i
m
u
latio
n
o
f
C
lo
u
d
co
m
p
u
ti
n
g
s
y
s
te
m
s
an
d
ap
p
licatio
n
s
etti
n
g
s
.
I
t
p
r
o
v
is
io
n
s
b
o
th
s
y
s
te
m
a
n
d
b
eh
av
io
r
m
o
d
elin
g
o
f
C
lo
u
d
s
y
s
te
m
co
m
p
o
n
en
t
s
s
u
ch
as
d
ata
ce
n
ter
s
,
v
ir
t
u
al
m
ac
h
in
e
s
(
VM
s
)
a
n
d
r
eso
u
r
ce
p
r
o
v
is
io
n
i
n
g
p
o
licies.
Us
in
g
t
h
e
lo
ad
b
ala
n
cin
g
p
ar
a
m
eter
s
o
f
v
ir
tu
al
m
ac
h
i
n
e
an
d
d
ata
ce
n
ter
,
w
e
h
a
v
e
ex
a
m
in
ed
o
n
clo
u
d
s
i
m
-
3
.
0
a
clo
u
d
-
co
m
p
u
tin
g
s
i
m
u
lato
r
an
d
ev
alu
ate
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
s
u
g
g
es
ted
m
et
h
o
d
u
s
i
n
g
j
o
b
ass
ig
n
i
n
g
t
o
th
e
clo
u
d
s
u
b
s
et.
Fig
u
r
e
4
r
ep
r
esen
ts
t
h
e
lo
ad
b
alan
cin
g
i
n
ea
c
h
clo
u
d
p
ar
titi
o
n
,
a
n
d
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
es
s
u
ch
as
E
ANN
b
as
ed
lo
ad
b
alan
cin
g
alg
o
r
ith
m
s
h
o
w
s
i
m
p
r
o
v
ed
r
es
u
lts
t
h
an
t
h
e
ex
i
s
tin
g
tech
n
iq
u
es.
I
n
th
e
b
eg
in
n
i
n
g
s
i
m
u
lato
r
f
ir
s
t,
g
e
n
er
ates
t
h
e
in
p
u
t
w
o
r
k
lo
ad
co
n
s
i
s
ti
n
g
o
f
a
s
et
o
f
w
o
r
k
f
lo
w
s
ar
r
iv
i
n
g
at
d
if
f
er
en
t
ti
m
e
in
s
tan
ts
i
n
an
o
n
li
n
e
m
a
n
n
er
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
E
la
s
tic
n
eu
r
a
l n
etw
o
r
k
meth
o
d
fo
r
lo
a
d
p
r
ed
ictio
n
in
cl
o
u
d
co
mp
u
tin
g
g
r
id
(
K
efa
ya
S
.
Qa
d
d
o
u
m
)
1205
Fo
r
ea
ch
task
o
f
a
w
o
r
k
f
lo
w
,
th
e
n
u
m
b
er
o
f
p
r
o
ce
s
s
o
r
s
r
eq
u
ir
ed
to
ex
ec
u
te
it
is
s
et
b
y
p
ick
in
g
u
p
a
r
an
d
o
m
n
u
m
b
er
w
it
h
in
a
p
r
e
-
co
n
f
i
g
u
r
e
d
r
an
g
e
ac
co
r
d
in
g
to
a
s
p
ec
if
ie
d
p
r
o
b
ab
ilit
y
d
is
tr
ib
u
tio
n
.
E
ac
h
e
x
p
er
i
m
e
n
t
i
n
v
o
k
es
2
0
r
u
n
s
,
o
f
w
h
ich
ea
c
h
s
i
m
u
late
s
1
0
0
o
n
lin
e
w
o
r
k
f
lo
w
s
i
n
a
m
u
lti
-
cl
u
s
te
r
en
v
ir
o
n
m
e
n
t
co
m
p
o
s
ed
o
f
5
c
lu
s
ter
s
ea
ch
co
n
tain
in
g
5
0
-
7
0
p
r
o
ce
s
s
o
r
s
.
E
A
NN
al
g
o
r
ith
m
i
s
u
s
ed
to
g
iv
e
b
es
t
pos
s
ib
le
allo
ca
tio
n
r
esu
lt
s
.
Set
p
=
1
6
s
f
o
r
ex
a
m
p
le
an
d
a
s
et
o
f
d
e
m
an
d
s
o
n
s
er
v
ice
is
D,
wh
er
e
clu
s
ter
ed
s
et
s
r
ep
r
esen
ted
b
y
C
1
,
C
2
,
C
3
.
T
h
e
o
u
tco
m
e
o
f
9
2
it
er
atio
n
s
was
b
alan
ce
d
an
d
r
ea
s
o
n
ab
le
n
o
d
e
allo
ca
tio
n
.
T
ask
allo
ca
tio
n
u
s
in
g
E
A
NN
g
o
es
as
Ass
o
ciatio
n
C
1
=
{S7
,
S1
5
,
S2
0
},
th
at
is
,
s
er
v
er
s
S7
,
S1
5
,
an
d
S1
5
ar
e
elec
ted
t
o
tr
ac
k
d
e
m
a
n
d
D1
.
C
o
ali
tio
n
C
2
=
{S2
,
S6
,
S2
2
},
th
at
is
,
s
er
v
er
s
S2
,
S6
,
an
d
S2
2
ar
e
elec
t
ed
to
tr
ac
k
d
e
m
a
n
d
D2
.
C
o
alitio
n
C
3
=
{S8
,
S1
7
,
S1
8
},
th
at
is
,
s
er
v
er
s
S8
,
S1
7
,
an
d
S1
8
ar
e
elec
ted
to
tr
ac
k
d
e
m
an
d
D3
.
R
e
s
u
lt
s
ar
e
s
h
o
w
n
i
n
T
ab
le
1
.
T
ab
le
1
.
T
im
e
o
f
E
x
ec
u
tio
n
o
f
t
h
e
Selecte
d
C
u
s
to
m
er
'
s
C
lo
u
d
T
ask
s
S
e
t
-
up
R
e
t
o
r
t
P
r
o
c
e
ssi
n
g
t
i
me
F
i
r
st
D
a
t
a
c
e
n
t
e
r
w
i
t
h
3
0
V
M
s
2
4
4
.
6
8
S
e
c
o
n
d
Tw
o
d
a
t
a
c
e
n
t
e
r
s w
i
t
h
1
5
V
M
s e
a
c
h
2
6
0
.
7
7
T
h
i
r
d
Tw
o
d
a
t
a
c
e
n
t
e
r
s w
i
t
h
3
0
V
M
s e
a
c
h
1
7
4
.
9
2
T
h
e
p
r
ed
icted
v
alu
e,
w
h
ic
h
i
s
d
o
n
e
w
ith
th
e
h
elp
o
f
p
ast
i
n
p
u
t
v
al
u
e
i
s
clea
r
l
y
s
h
o
w
n
in
Fig
u
r
e
3
.
T
h
e
p
r
ed
icted
v
alu
e
p
r
o
v
id
es
a
h
i
g
h
er
e
f
f
icie
n
c
y
o
f
o
u
t
p
u
t.
Sin
ce
t
h
e
p
r
ed
icted
v
al
u
e
p
r
o
v
id
es
g
r
ea
ter
ef
f
icien
c
y
,
t
h
e
m
ea
n
s
q
u
ar
e
e
r
r
o
r
is
s
lig
h
t
l
y
h
i
g
h
.
A
s
Fig
u
r
e
3
illu
s
tr
ates
o
u
tp
u
t
i
f
t
h
e
r
ate=
0
.
7
,
an
d
th
is
is
s
elec
ted
o
n
ac
co
u
n
t
o
f
leas
t
e
r
r
o
r
ca
lcu
latio
n
f
o
r
o
u
r
j
o
b
s
p
r
ed
ictio
n
o
n
VM
,
w
h
er
e
t
h
is
r
esu
lt
en
h
a
n
ce
s
t
h
e
p
r
ed
ictio
n
ac
cu
r
ac
y
an
d
w
o
r
k
f
o
r
th
e
b
en
e
f
it o
f
b
etter
s
c
h
ed
u
li
n
g
i
n
t
h
e
n
e
x
t step
.
Fig
u
r
e
3
.
E
r
r
o
r
r
ate
f
o
r
b
est p
e
r
f
o
r
m
an
ce
T
h
e
lo
ad
m
ig
r
atio
n
b
ef
o
r
e
a
n
d
af
ter
i
s
g
i
v
en
in
Fi
g
u
r
e
4
,
a
s
w
e
ca
n
s
ee
t
h
at
o
v
er
a
p
er
io
d
th
e
n
u
m
b
er
o
f
j
o
b
s
ex
ec
u
ted
co
m
p
ar
in
g
to
av
ailab
le
VM
s
is
b
ec
o
m
in
g
r
ea
s
o
n
ab
le
af
ter
m
i
g
r
ati
n
g
j
o
b
s
ar
o
u
n
d
.
T
h
e
r
esu
lt
o
f
p
r
ed
icted
jo
b
s
is
k
ee
p
in
g
t
h
e
b
alan
ce
o
f
d
i
f
f
er
e
n
t
s
er
v
er
s
as
eq
u
al
as
p
o
s
s
ib
le.
Fi
g
u
r
e
5
g
iv
e
s
t
h
e
g
r
ap
h
ical
d
em
o
n
s
tr
atio
n
o
f
t
h
e
lo
ad
b
al
an
cin
g
in
ea
ch
clo
u
d
p
ar
titi
o
n
a
n
d
t
h
e
s
u
g
g
e
s
ted
ap
p
r
o
ac
h
E
ANN
b
ased
lo
ad
b
alan
cin
g
al
g
o
r
ith
m
s
h
o
w
s
t
h
e
b
etter
r
esu
lts
t
h
an
t
h
e
o
th
er
te
ch
n
iq
u
es li
k
e
An
t c
o
lo
n
y
a
n
d
Ho
n
e
y
b
eh
a
v
io
r
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
2
,
A
p
r
il
201
9
:
1
2
0
1
-
1
2
0
8
1206
Fig
u
r
e
4
.
L
o
ad
m
ig
r
at
io
n
o
n
cl
o
u
d
Fig
u
r
e
5
.
L
o
ad
b
alan
cin
g
co
m
p
ar
is
o
n
4.
CO
NCLU
SI
O
N
L
o
ad
p
r
ed
ictio
n
is
n
ee
d
ed
f
o
r
allo
ca
tin
g
d
ata
ce
n
ter
r
e
s
o
u
r
ce
s
d
ep
en
d
in
g
o
n
ap
p
licatio
n
s
d
e
m
an
d
s
.
T
h
e
f
lu
ct
u
atio
n
o
f
d
e
m
a
n
d
w
i
ll
b
e
less
d
etec
ted
u
s
i
n
g
o
u
r
ap
p
r
o
ac
h
.
T
h
e
s
y
s
te
m
is
m
a
n
a
g
in
g
r
eso
u
r
ce
s
an
d
av
o
id
s
o
v
er
lo
ad
.
T
h
e
s
u
g
g
e
s
t
ed
tech
n
iq
u
e
u
s
es
s
o
m
e
o
f
t
h
e
d
esig
n
p
r
o
ce
s
s
es
o
f
a
lo
ad
b
alan
cin
g
.
E
A
NN
f
o
r
ec
ast
t
h
e
r
eq
u
est
t
h
e
n
ac
c
o
r
d
in
g
l
y
a
s
s
i
g
n
r
eso
u
r
ce
s
ac
c
o
r
d
in
g
to
th
e
r
eq
u
e
s
t.
T
h
at
m
ai
n
tai
n
s
t
h
e
ac
ti
v
e
s
er
v
er
s
ac
co
r
d
in
g
to
c
u
r
r
en
t
d
e
m
an
d
.
T
h
is
p
ap
er
,
s
u
g
g
es
ted
a
n
e
w
e
f
f
ec
ti
v
e
lo
ad
b
alan
ci
n
g
tec
h
n
iq
u
e
-
u
s
i
n
g
E
A
NN
b
ased
lo
ad
-
b
alan
c
in
g
tech
n
iq
u
e
to
ac
h
ie
v
e
d
etec
t
ab
le
en
h
a
n
ce
m
e
n
ts
i
n
r
e
s
o
u
r
ce
d
ep
lo
y
m
e
n
t
a
n
d
av
ailab
ilit
y
a
m
o
n
g
clo
u
d
-
co
m
p
u
tin
g
at
m
o
s
p
h
er
e.
Fo
r
f
u
tu
r
e,
to
f
u
r
th
er
i
m
p
r
o
v
e
t
h
e
b
ala
n
cin
g
p
er
f
o
r
m
an
c
e
o
f
lo
ad
i
n
clo
u
d
e
n
v
ir
o
n
m
en
ts
,
f
o
r
in
s
ta
n
ce
,
t
h
e
p
r
o
ac
tiv
e
j
o
b
a
llo
ca
tio
n
m
et
h
o
d
co
u
ld
b
e
ex
ten
d
ed
to
co
n
s
id
er
m
u
lt
ip
le
r
u
n
n
in
g
ta
s
k
s
f
o
r
s
i
m
u
lta
n
eo
u
s
p
r
ee
m
p
tio
n
a
n
d
w
o
u
ld
in
cr
ea
s
e
th
e
p
r
o
b
ab
ilit
y
f
o
r
h
ig
h
-
p
r
io
r
it
y
tas
k
s
to
s
ta
r
t
ex
ec
u
t
io
n
ea
r
lier
an
d
t
h
u
s
i
m
p
r
o
v
e
th
e
o
v
er
all
s
y
s
te
m
p
er
f
o
r
m
a
n
ce
.
Fo
r
th
e
s
h
o
r
test
-
w
o
r
k
f
lo
w
-
f
ir
s
t
p
o
lic
y
,
o
th
er
m
etr
ics
f
o
r
p
r
io
r
itizin
g
w
o
r
k
f
lo
w
s
co
u
ld
b
e
in
v
e
s
ti
g
ated
in
ad
d
itio
n
to
th
e
r
em
a
in
i
n
g
ex
ec
u
tio
n
t
i
m
e
u
s
ed
in
th
i
s
p
ap
er
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
E
la
s
tic
n
eu
r
a
l n
etw
o
r
k
meth
o
d
fo
r
lo
a
d
p
r
ed
ictio
n
in
cl
o
u
d
co
mp
u
tin
g
g
r
id
(
K
efa
ya
S
.
Qa
d
d
o
u
m
)
1207
RE
F
E
R
E
NC
E
S
[1
]
R.
P
.
M
a
h
o
w
a
ld
,
W
o
rld
w
id
e
S
o
ftw
a
r
e
A
s
A
S
e
r
v
ice
2
0
1
0
–
2
0
1
4
F
o
re
c
a
st:
S
o
f
t
w
a
re
W
il
l
N
e
v
e
r
B
e
S
a
m
e
,
In
,
IDC,
2
0
1
0
.
[2
]
Nu
ñ
e
z
D,
F
e
rn
a
n
d
e
z
–
G
a
g
o
C,
P
e
a
rso
n
S
,
F
e
li
c
i
M
.
“
A
M
e
ta
m
o
d
e
l
f
o
r
M
e
a
su
rin
g
Ac
c
o
u
n
t
a
b
il
it
y
A
tt
rib
u
tes
in
th
e
Clo
u
d
,
”
In
:
Pro
c
e
e
d
i
n
g
s
o
f
th
e
2
0
1
3
IEE
E
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
Cl
o
u
d
C
o
mp
u
ti
n
g
T
e
c
h
n
o
l
o
g
y
a
n
d
S
c
ien
c
e
(
Cl
o
u
d
Co
m
2
0
1
3
),
IEE
E
Pro
c
e
e
d
in
g
s
o
f
th
e
th
ird
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Clo
u
d
Co
mp
u
ti
n
g
.
IEE
E,
M
iam
i,
pp
2
9
7
-
3
0
4
,
2
0
1
3
.
[3
]
Ch
a
z
a
let
A
.
,
“
S
e
rv
ic
e
Lev
e
l
Ch
e
c
k
in
g
in
th
e
Cl
o
u
d
Co
m
p
u
ti
n
g
Co
n
tex
t
,”
In
:
W
ied
e
r
P
,
Bu
tl
e
r
J
M
,
T
h
e
il
m
a
n
n
W
,
Ya
h
y
a
p
o
u
r
R
(e
d
s)
(2
0
1
1
)
S
e
rv
ice
L
e
v
e
l
Ag
re
e
me
n
ts
fo
r Clo
u
d
Co
mp
u
ti
n
g
.
S
p
r
in
g
e
r,
2
0
1
0
.
[4
]
C.
L
i
m
,
S
.
L
u
,
A
.
,
Ch
e
b
o
tk
o
,
a
n
d
F
.
F
o
to
u
h
i
,
“
S
to
ri
n
g
,
Re
a
so
n
in
g
,
a
n
d
Qu
e
ry
in
g
Op
m
c
o
m
p
li
a
n
t
S
c
ien
ti
f
ic
L
o
a
d
P
r
o
v
e
n
a
n
c
e
Us
in
g
Re
latio
n
a
l
Da
t
a
b
a
se
s
,
"
Fu
tu
re
Ge
n
e
ra
ti
o
n
Co
m
p
u
ter
S
y
ste
ms
2
7
,
p
p
.
7
8
1
–
7
8
9
,
2
0
1
1
.
[5
]
H.
T
o
p
c
u
o
g
lu
,
S
.
Ha
riri
a
n
d
M
.
W
u
,
“
P
e
rf
o
r
m
a
n
c
e
-
e
ffe
c
ti
v
e
a
n
d
L
o
w
-
Co
m
p
lex
it
y
T
a
s
k
S
c
h
e
d
u
li
n
g
f
o
r
He
tero
g
e
n
e
o
u
s Co
m
p
u
ti
n
g
,
”
IEE
E
T
ra
n
s.
o
n
P
a
ra
ll
e
l
a
n
d
Distri
b
u
ted
S
y
ste
ms
,
v
o
l.
1
3
,
p
p
.
2
6
0
-
2
7
4
,
2
0
0
2
.
[6
]
J.
D.
Ul
lm
a
n
,
“
N
P
-
c
o
m
p
lete
S
c
h
e
d
u
li
n
g
P
r
o
b
lem
s,”
J
o
u
rn
a
l
o
f
C
o
mp
u
ter
a
n
d
S
y
ste
m
S
c
ien
c
e
s
,
Vo
l.
1
0
,
Iss
.
3
,
p
p
.
384
–
3
9
3
,
1
9
7
5
.
[4
]
E
.
K.
By
u
n
,
Y.
S
.
Ke
e
,
J.
S
.
Kim
,
a
n
d
S
.
M
a
e
n
g
,
“
Co
st
o
p
ti
m
ize
d
p
ro
v
isi
o
n
i
n
g
o
f
e
las
ti
c
re
so
u
rc
e
s f
o
r
a
p
p
li
c
a
ti
o
n
lo
a
d
s”
,
Fu
tu
re
Ge
n
e
ra
ti
o
n
Co
m
p
u
ter
S
y
st
e
ms
2
7
,
p
p
.
1
0
1
1
-
1
0
2
6
,
2
0
1
1
.
[7
]
T
.
N'
ta
k
p
e
'
a
n
d
F
.
S
u
ter,
“
A
Co
m
p
a
riso
n
o
f
S
c
h
e
d
u
li
n
g
Ap
p
r
o
a
c
h
e
s
f
o
r
M
ix
e
d
-
P
a
ra
ll
e
l
A
p
p
li
c
a
ti
o
n
s
o
n
He
tero
g
e
n
e
o
u
s
P
latf
o
rm
s
,”
Pro
c
.
th
e
6
t
h
In
ter
n
a
ti
o
n
a
l
S
y
mp
o
si
u
m
o
n
P
a
ra
l
lel
a
n
d
Distrib
u
t
e
d
Co
mp
u
ti
n
g
(
IS
-
PDC)
,
Ha
g
e
n
b
e
rg
,
A
u
stria,
Ju
ly
2
0
0
7
.
[8
]
M
e
ss
a
g
e
P
a
ss
in
g
In
terf
a
c
e
,
h
tt
p
:
//
ww
w
.
m
p
i
-
f
o
ru
m
.
o
rg
/,
(2
0
1
5
.
3
)
[9
]
M
.
Ba
rre
to
,
R.
A
v
il
a
a
n
d
P
.
N
a
v
a
u
x
,
“
T
h
e
M
u
lt
iclu
ste
r
M
o
d
e
l
to
th
e
In
teg
ra
ted
Us
e
o
f
M
u
lt
i
p
le
W
o
rk
sta
ti
o
n
Clu
ste
rs
,”
3
rd
W
o
rk
sh
o
p
o
n
Per
so
n
a
l
Co
m
p
u
ter
b
a
se
d
Ne
two
rk
s
o
f
W
o
rk
sta
ti
o
n
s
,
p
p
.
7
1
–
8
0
,
2
0
0
0
.
[1
0
]
K.
C.
Hu
a
n
g
,
“
On
Ef
f
e
c
ts
o
f
Re
s
o
u
rc
e
F
ra
g
m
e
n
tatio
n
o
n
Jo
b
S
c
h
e
d
u
li
n
g
P
e
rf
o
rm
a
n
c
e
in
Co
m
p
u
ti
n
g
G
rid
s
”
,
2
0
0
9
1
0
t
h
In
ter
n
a
ti
o
n
a
l
S
y
mp
o
si
u
m o
n
Per
v
a
siv
e
S
y
ste
ms
,
Al
g
o
rit
h
ms
,
a
n
d
Ne
tw
o
rk
s
,
p
p
.
7
0
1
-
7
0
5
,
2
0
0
9
.
[1
1
]
C.
C.
Hs
u
,
K.C.
H
u
a
n
g
a
n
d
F
.
J.
W
a
n
g
,
“
On
li
n
e
S
c
h
e
d
u
li
n
g
o
f
Lo
a
d
A
p
p
li
c
a
ti
o
n
s
i
n
G
rid
E
n
v
iro
n
m
e
n
ts
,”
Fu
tu
re
Ge
n
e
ra
ti
o
n
Co
mp
u
ter
S
y
ste
ms
2
7
,
p
p
.
8
6
0
–
8
7
0
,
2
0
1
1
.
[1
2
]
S
a
n
tan
u
Da
m
,
G
o
p
a
M
a
n
d
a
l,
Ko
u
sik
Da
sg
u
p
ta
a
n
d
P
a
ra
m
a
rth
a
Du
tt
a
,
“
G
e
n
e
ti
c
A
lg
o
rit
h
m
a
n
d
G
ra
v
it
a
ti
o
n
a
l
Em
u
latio
n
Ba
se
d
Hy
b
rid
L
o
a
d
B
a
lan
c
in
g
S
trate
g
y
in
Cl
o
u
d
C
o
m
p
u
ti
n
g
”
,
i
n
p
ro
c
.
th
ir
d
I
n
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
Co
mp
u
ter
,
C
o
mm
u
n
ica
ti
o
n
,
Co
n
t
ro
l
a
n
d
I
n
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
(
C3
IT
),
I
EE
E
,
p
p
.
1
-
7
,
F
e
b
ru
a
ry
2
0
1
5
.
[1
3
]
A
S
K
AL
ON
,
h
tt
p
:/
/w
ww
.
d
p
s.u
ib
k
.
a
c
.
a
t/
p
ro
jec
ts/t
e
u
ta/
(
2
0
1
5
.
3
)
[1
4
]
DAG
m
a
n
,
h
tt
p
:/
/res
e
a
rc
h
.
c
s.w
isc
.
e
d
u
/
h
tco
n
d
o
r
/d
a
g
m
a
n
/d
a
g
m
a
n
.
h
t
m
l
(2
0
1
5
.
3
)
[1
5
]
M
.
W
iec
z
o
re
k
,
M
.
S
id
d
iq
u
i,
A
.
V
il
laz
ó
n
,
R.
P
r
o
d
a
n
,
T
.
F
a
h
rin
g
e
r,
“
A
p
p
ly
in
g
A
d
v
a
n
c
e
Re
se
r
v
a
t
io
n
t
o
In
c
re
a
se
P
re
d
icta
b
il
it
y
o
f
W
o
rk
f
lo
w
E
x
e
c
u
ti
o
n
o
n
th
e
G
rid
”
,
P
r
o
c
e
e
d
in
g
s
o
f
2
n
d
IEE
E
I
n
tern
a
ti
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
e
-
S
c
ien
c
e
a
n
d
G
rid
Co
m
p
u
ti
n
g
,
p
p
.
8
2
,
De
c
e
m
b
e
r
4
-
6
,
A
m
ste
rd
a
m
,
N
e
th
e
rlan
d
s,
2
0
0
6
.
[1
6
]
Zeb
a
Kh
a
n
,
M
a
h
f
o
o
z
A
lam
,
Ra
z
a
A
b
b
a
s
Ha
id
ri
“
Eff
e
c
ti
v
e
L
o
a
d
Ba
lan
c
e
S
c
h
e
d
u
li
n
g
S
c
h
e
m
e
s
f
o
r
He
tero
g
e
n
e
o
u
s
Distrib
u
te
d
S
y
s
te
m
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
El
e
c
trica
l
a
n
d
Co
m
p
u
ter
E
n
g
i
n
e
e
rin
g
(
IJ
ECE
)
,
Vo
l.
7
,
N
o
.
5
,
Oc
to
b
e
r
2
0
1
7
,
p
p
.
2
7
5
7
~
2
7
6
5
,
IS
S
N:
2
0
8
8
-
8
7
0
8
,
2
0
1
7
.
[1
7
]
R.
P
r
o
d
a
n
,
T
.
F
a
h
r
in
g
e
r,
“
Dy
n
a
m
ic
S
c
h
e
d
u
li
n
g
o
f
S
c
ien
ti
f
ic
W
o
rk
f
lo
w
A
p
p
li
c
a
ti
o
n
s
o
n
t
h
e
G
rid
:
A
Ca
se
S
tu
d
y
”
,
Pro
c
e
e
d
in
g
s
o
f
t
h
e
2
0
th
S
y
mp
o
siu
m o
n
Ap
p
li
e
d
C
o
mp
u
ti
n
g
(
S
AC
2
0
0
5
)
,
p
p
.
6
8
7
-
6
9
4
,
2
0
0
5
.
[1
8
]
Ke
fa
y
a
Qa
d
d
o
u
m
,
Ev
o
r
Hin
e
s,
a
n
d
Da
c
ian
a
Ili
e
sc
u
,
“
Yie
l
d
P
re
d
i
c
ti
o
n
T
e
c
h
n
i
q
u
e
Us
in
g
Hy
b
rid
A
d
a
p
ti
v
e
Ne
u
ra
l
G
e
n
e
ti
c
Ne
t
w
o
rk
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Co
mp
u
t
a
ti
o
n
a
l
In
telli
g
e
n
c
e
a
n
d
Ap
p
li
c
a
ti
o
n
s
,
V
o
l.
1
1
,
No
.
0
4
,
1
2
5
0
0
2
1
,
2
0
1
2
,
DO
I:
h
tt
p
:/
/
d
x
.
d
o
i.
o
rg
/1
0
.
1
1
4
2
/S
1
4
6
9
0
2
6
8
1
2
5
0
0
2
1
6
.
[1
9
]
T.
S
a
sid
h
a
r,
V.
Ha
v
ish
a
,
S.
Ko
u
sh
ik
,
M.
De
e
p
,
V
K.
Re
d
d
y
,
"
L
o
a
d
Ba
lan
c
in
g
T
e
c
h
n
iq
u
e
s
f
o
r
Ef
f
ici
e
n
t
T
ra
ff
i
c
M
a
n
a
g
e
m
e
n
t
in
Clo
u
d
En
v
ir
o
n
m
e
n
t,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
El
e
c
trica
l
a
n
d
Co
m
p
u
ter
En
g
i
n
e
e
rin
g
(
IJ
ECE
)
,
v
o
l.
6
,
n
o
.
3
,
p
p
.
9
6
3
-
9
7
3
,
2
0
1
6
.
[2
0
]
P
.
Qia
o
,
“
On
th
e
S
e
c
u
ri
ty
o
f
a
D
y
n
a
m
i
c
a
n
d
S
e
c
u
re
Ke
y
M
a
n
a
g
e
m
e
n
t
M
o
d
e
l
F
o
r
h
iera
rc
h
ica
l
He
tero
g
e
n
e
o
u
s
S
e
n
s
o
r
Ne
tw
o
rk
s
,”
T
EL
KOM
NIKA
In
d
o
n
e
sia
n
J
o
u
rn
a
l
of
El
e
c
-
trica
l
En
g
in
e
e
rin
g
,
v
o
l.
1
2
,
n
o
.
1
0
,
p
p
.
7
4
5
9
–
7
4
6
2
,
2
0
1
4
.
h
tt
p
:
//
iae
sjo
u
r
n
a
l.
c
o
m
/o
n
li
n
e
/i
n
d
e
x
.
p
h
p
/T
EL
KO
M
NIK
A
/article
/v
ie
w
/5
5
7
9
[2
1
]
Ca
o
,
X
.
,
Z
h
o
n
g
,
Y.,
Zh
o
u
,
Y.,
W
a
n
g
,
J.,
Zh
u
,
C.
,
&
Zh
a
n
g
,
W
.
“
In
tera
c
ti
v
e
T
e
m
p
o
ra
l
Re
c
u
rre
n
t
Co
n
v
o
lu
t
io
n
Ne
tw
o
rk
f
o
r
T
ra
ff
i
c
P
re
d
icti
o
n
in
Da
ta Ce
n
ters
,
”
IEE
E
Acc
e
ss
,
6
,
5
2
7
6
-
5
2
8
9
,
2
0
1
8
.
[2
2
]
P
e
n
g
,
G
.
,
W
a
n
g
,
H.
,
Zh
a
n
g
,
H.,
&
Do
n
g
,
J.
“
Kn
o
w
led
g
e
-
Ba
s
e
d
Re
so
u
rc
e
A
ll
o
c
a
ti
o
n
f
o
r
Co
ll
a
b
o
r
a
ti
v
e
S
im
u
latio
n
De
v
e
lo
p
m
e
n
t
in
a
M
u
lt
i
-
T
e
n
a
n
t
Clo
u
d
Co
m
p
u
ti
n
g
En
v
ir
o
n
m
e
n
t
,”
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
S
e
rv
ice
s
Co
mp
u
t
in
g
,
2
0
1
6
.
DO
I:
1
0
.
1
1
0
9
/T
S
C.
2
0
1
6
.
2
5
1
8
1
6
1
BI
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
Dr
.
K
e
fa
y
a
Q
a
d
d
o
u
m
is
a
n
As
sista
n
t
p
ro
f
e
ss
o
r
a
t
Hig
h
e
r
Co
ll
e
g
e
s
o
f
Tec
h
n
o
lo
g
y
U
A
E,
w
a
s
a
n
A
s
sista
n
t
p
ro
f
e
ss
o
r
in
A
l
-
A
h
li
y
y
a
Am
m
a
n
Un
iv
e
rsit
y
.
S
h
e
re
c
e
i
v
e
d
h
e
r
f
irst
d
e
g
re
e
in
Co
m
p
u
ter
S
c
ien
c
e
f
ro
m
P
h
i
lad
e
lp
h
ia
Un
iv
e
rsity
,
Jo
rd
a
n
,
in
Ju
ly
2
0
0
3
,
m
a
ste
r
d
e
g
re
e
in
Co
m
p
u
ter
In
f
o
rm
a
ti
o
n
S
y
st
e
m
f
ro
m
P
h
il
a
d
e
lp
h
ia
Un
iv
e
rsit
y
Ju
l
y
2
0
0
8
,
a
n
d
a
d
o
c
to
ra
te
d
e
g
re
e
in
A
.
I
f
ro
m
W
a
r
w
ic
k
Un
iv
e
rsit
y
,
th
e
UK
in
2
0
1
3
.
He
r
re
se
a
rc
h
a
r
e
a
o
f
in
tere
st
in
c
lu
d
e
s
A
.
I,
M
a
c
h
in
e
Lea
rn
in
g
,
Da
t
a
M
in
i
n
g
,
a
n
d
Ne
tw
o
rk
in
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
2
,
A
p
r
il
201
9
:
1
2
0
1
-
1
2
0
8
1208
Dr
.
M
o
sle
h
M
.
Ab
u
-
Alh
a
j
is
a
se
n
i
o
r
lec
tu
re
r
in
A
l
-
A
h
li
y
y
a
Am
m
a
n
Un
iv
e
rsit
y
.
He
re
c
e
iv
e
d
h
is
f
irst
d
e
g
re
e
in
Co
m
p
u
ter
S
c
ien
c
e
f
ro
m
P
h
il
a
d
e
lp
h
ia
Un
iv
e
rsit
y
,
Jo
rd
a
n
,
in
Ju
ly
2
0
0
4
,
m
a
ste
r
d
e
g
re
e
in
Co
m
p
u
ter
In
f
o
rm
a
ti
o
n
S
y
ste
m
f
ro
m
th
e
A
ra
b
Ac
a
d
e
m
y
f
o
r
Ba
n
k
in
g
a
n
d
F
in
a
n
c
ial
S
c
ien
c
e
s,
Jo
rd
a
n
i
n
Ju
ly
2
0
0
7
,
a
n
d
a
d
o
c
to
ra
te
in
M
u
lt
im
e
d
ia
Ne
tw
o
rk
s
P
r
o
to
c
o
ls
f
ro
m
Un
iv
e
rsiti
S
a
in
s
M
a
la
y
sia
in
2
0
1
1
.
His
re
se
a
rc
h
a
re
a
o
f
in
tere
st
in
c
lu
d
e
s
V
o
IP
,
M
u
l
ti
m
e
d
ia
Ne
t
w
o
rk
in
g
,
a
n
d
Co
n
g
e
stio
n
Co
n
tro
l.
A
p
a
rt
f
ro
m
re
se
a
rc
h
,
Dr.
M
o
sle
h
M
.
A
b
u
-
A
l
h
a
j
a
lso
d
o
e
s
c
o
n
su
lt
a
n
c
y
s
e
rv
ice
s
in
t
h
e
a
b
o
v
e
re
se
a
rc
h
a
re
a
s an
d
d
i
re
c
ts
th
e
Cisc
o
a
c
a
d
e
m
y
tea
m
a
t
A
l
-
A
h
li
y
y
a
Am
m
a
n
Un
iv
e
rsity
.
Na
m
e
e
r
N.
EL
-
E
m
a
m
:
He
c
o
m
p
lete
d
h
is
P
h
.
D
.
in
Co
m
p
u
ter
S
c
ie
n
c
e
w
it
h
h
o
n
o
r
i
n
1
9
9
7
.
He
w
o
rk
s
a
s
a
n
a
ss
ist
a
n
t
p
ro
f
e
ss
o
r
in
th
e
Co
m
p
u
ter
S
c
ien
c
e
De
p
a
rt
m
e
n
t,
Ba
sra
Un
iv
e
rsit
y
.
In
1
9
9
8
,
h
e
j
o
in
e
d
th
e
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
,
P
h
il
a
d
e
lp
h
ia
Un
iv
e
rsity
,
a
s
a
n
A
s
sista
n
c
e
P
ro
f
e
ss
o
r,
a
n
d
th
e
n
h
e
g
o
t
th
e
ra
n
k
"
As
so
c
iate
P
ro
f
e
s
so
r"
in
2
0
1
0
.
No
w
h
e
is
a
F
u
ll
P
r
o
f
e
ss
o
r
a
t
th
e
sa
m
e
u
n
iv
e
rsit
y
.
H
e
w
o
rk
s
a
s
a
c
h
a
ir
o
f
Co
m
p
u
ter
S
c
ien
c
e
De
p
a
rt
m
e
n
t
a
n
d
th
e
d
e
p
u
ty
d
e
a
n
o
f
th
e
f
a
c
u
lt
y
o
f
In
f
o
r
m
a
ti
o
n
T
e
c
h
n
o
lo
g
y
,
P
h
il
a
d
e
lp
h
ia
Un
iv
e
rsity
.
His
re
s
e
a
rc
h
in
tere
st
in
c
lu
d
e
s
Co
m
p
u
ter
S
im
u
latio
n
w
it
h
t
h
e
in
telli
g
e
n
t
sy
ste
m
,
P
a
ra
ll
e
l
A
lg
o
ri
th
m
s,
a
n
d
S
o
f
t
c
o
m
p
u
ti
n
g
u
si
n
g
Co
n
v
o
l
u
ti
o
n
Ne
u
ra
l
Ne
tw
o
rk
,
GA
,
A
CO,
a
n
d
P
S
O f
o
r
m
a
n
y
k
in
d
s o
f
a
p
p
li
c
a
ti
o
n
s li
k
e
I
m
a
g
e
P
ro
c
e
ss
in
g
,
S
o
u
n
d
P
r
o
c
e
ss
in
g
,
F
lu
i
d
F
l
o
w,
a
n
d
Co
m
p
u
ter S
e
c
u
r
it
y
(S
teg
a
n
o
g
ra
p
h
y
).
Evaluation Warning : The document was created with Spire.PDF for Python.