I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
15
,
No
.
5
,
Octo
b
er
2
0
2
6
,
p
p
.
4
6
4
5
~
4
6
5
5
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
15
.i
5
.
p
p
4
6
4
5
-
4
6
5
5
4645
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
O
ptimi
zing
cloud
energy
wit
h
deep
reinforceme
nt
l
ea
rning
:
a
n
a
rtif
icia
l be
e colo
ny
-
Q
-
l
ea
rning
hy
brid app
ro
a
ch
M
o
ha
m
m
ed
Z
ia
ur
Ra
hm
a
n
1
,
Ana
nd
a
ra
j
Sh
a
nthi P
icha
n
di
2
1
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
S
c
h
o
o
l
o
f
E
n
g
i
n
e
e
r
i
n
g
a
n
d
Te
c
h
n
o
l
o
g
y
,
S
a
p
t
h
a
g
i
r
i
N
P
S
U
n
i
v
e
r
si
t
y
,
B
e
n
g
a
l
u
r
u
,
I
n
d
i
a
2
S
c
h
o
o
l
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
P
r
e
s
i
d
e
n
c
y
U
n
i
v
e
r
si
t
y
,
B
e
n
g
a
l
u
r
u
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ma
y
2
6
,
2
0
2
5
R
ev
is
ed
J
u
l 2
7
,
2
0
2
6
Acc
ep
ted
Au
g
1
2
,
2
0
2
6
Lo
a
d
b
a
lan
c
in
g
in
t
h
e
c
lo
u
d
i
s
a
c
ru
c
ial
c
h
a
ra
c
teristic
fo
r
d
istri
b
u
t
in
g
re
so
u
rc
e
s
fro
m
d
a
ta
c
e
n
ters
.
Ho
we
v
e
r,
lo
a
d
b
a
la
n
c
in
g
i
n
a
c
lo
u
d
e
n
v
iro
n
m
e
n
t
re
m
a
in
s
a
c
h
a
ll
e
n
g
in
g
issu
e
d
u
e
to
t
h
e
in
fl
u
x
o
f
n
u
m
e
ro
u
s
in
c
o
m
in
g
tas
k
s,
p
a
rti
c
u
larl
y
i
n
a
n
in
fra
str
u
c
tu
re
a
s
a
se
rv
ice
(Ia
a
S
)
c
lo
u
d
e
n
v
iro
n
m
e
n
t.
To
a
d
d
re
ss
th
e
c
o
m
p
lex
it
y
a
n
d
h
i
g
h
d
ime
n
sio
n
a
li
ty
o
f
th
e
p
ro
c
e
ss
,
m
a
c
h
in
e
lea
rn
in
g
(M
L
)
tec
h
n
iq
u
e
s
c
o
m
b
in
e
d
wit
h
o
p
ti
m
iza
ti
o
n
a
lg
o
rit
h
m
s
a
re
u
se
d
,
w
h
ich
fa
ste
n
t
h
e
p
ro
c
e
ss
a
n
d
m
in
imiz
e
e
x
e
c
u
ti
o
n
ti
m
e
.
Th
e
d
e
e
p
re
in
f
o
rc
e
m
e
n
t
lea
rn
in
g
(DRL)
–
a
rti
ficia
l
b
e
e
c
o
lo
n
y
with
Q
-
lea
rn
in
g
(ABCQ
)
a
p
p
r
o
a
c
h
is
p
r
o
p
o
se
d
to
so
l
v
e
lo
a
d
b
a
lan
c
in
g
i
n
a
c
lo
u
d
e
n
v
iro
n
m
e
n
t.
Th
e
a
g
e
n
t
in
DRL
m
a
k
e
s
in
telli
g
e
n
t
d
e
c
isio
n
s
to
h
a
n
d
le
a
larg
e
n
u
m
b
e
r
o
f
tas
k
s
e
fficie
n
tl
y
a
t
les
se
r
ti
m
e
a
n
d
in
c
re
a
sin
g
sp
e
e
d
.
T
h
e
a
rti
ficia
l
b
e
e
c
o
lo
n
y
(
ABC
)
a
l
g
o
ri
th
m
u
se
s
a
g
lo
b
a
l
g
u
i
d
i
n
g
fa
c
to
r
to
in
it
ializ
e
p
o
siti
o
n
u
p
d
a
tes
,
e
n
h
a
n
c
in
g
t
h
e
se
a
rc
h
a
b
il
it
y
a
n
d
p
e
rfo
rm
a
n
c
e
o
f
th
e
a
lg
o
rit
h
m
.
Th
e
p
r
o
p
o
se
d
DRL
-
ABCQ
a
p
p
ro
a
c
h
o
p
ti
m
ize
s
sc
h
e
d
u
li
n
g
a
n
d
re
so
u
rc
e
u
ti
l
iza
ti
o
n
,
i
n
c
re
a
se
s
th
e
th
ro
u
g
h
p
u
t
o
f
v
irt
u
a
l
m
a
c
h
in
e
s
(
VMs),
a
n
d
b
a
lan
c
e
s
th
e
lo
a
d
b
e
twe
e
n
VM
s.
Th
e
DRL
-
ABCQ
a
p
p
ro
a
c
h
a
c
h
iev
e
s
a
m
a
k
e
sp
a
n
o
f
1
2
5
.
5
2
s,
e
n
e
rg
y
u
t
il
iza
ti
o
n
o
f
9
9
.
8
1
k
J,
a
n
d
a
t
h
ro
u
g
h
p
u
t
o
f
2
.
3
9
j
o
b
s
/s,
c
o
m
p
a
re
d
t
o
e
x
isti
n
g
a
p
p
ro
a
c
h
e
s.
K
ey
w
o
r
d
s
:
Ar
tific
ial
b
ee
co
lo
n
y
Dee
p
r
ein
f
o
r
ce
m
en
t le
ar
n
in
g
L
o
ad
b
alan
ci
n
g
Q
-
lear
n
in
g
Vir
tu
al
m
ac
h
in
es
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Mo
h
am
m
ed
Z
iau
r
R
ah
m
an
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
E
n
g
in
ee
r
in
g
,
Sch
o
o
l o
f
E
n
g
in
ee
r
in
g
a
n
d
T
ec
h
n
o
lo
g
y
Sap
th
ag
ir
i N
PS
Un
iv
er
s
ity
B
en
g
alu
r
u
,
I
n
d
ia
E
m
ail:
m
o
h
am
m
e
d
ziau
r
@
s
n
p
s
u
.
ed
u
.
in
1.
I
NT
RO
D
UCT
I
O
N
I
n
r
e
c
en
t
t
i
m
e
s
,
t
h
e
c
l
o
u
d
n
e
tw
o
r
k
h
a
s
b
e
co
m
e
a
p
o
p
u
l
ar
te
c
h
n
iq
u
e
t
h
a
t
o
f
f
er
s
v
ar
i
o
u
s
s
e
r
v
i
c
e
s
a
s
p
e
r
u
s
e
r
r
eq
u
e
s
t
s
a
cr
o
s
s
in
t
er
n
e
t
[
1
]
.
C
l
o
u
d
c
o
m
p
u
t
i
n
g
(
C
C
)
p
r
o
v
i
d
e
s
u
s
e
r
s
w
i
th
th
e
a
b
i
l
i
ty
t
o
a
c
c
e
s
s
c
o
m
p
u
t
in
g
r
e
s
o
u
r
ce
s
o
r
e
n
ab
l
e
s
t
h
e
m
t
o
p
u
r
ch
a
s
e
r
e
q
u
ir
e
d
c
l
o
u
d
s
e
r
v
i
c
e
s
o
n
-
d
em
an
d
v
i
a
t
h
e
i
n
te
r
n
e
t
,
b
a
s
e
d
o
n
ap
p
l
i
ca
t
i
o
n
s
[
2
]
–
[
4
]
.
T
h
e
C
C
s
y
s
t
e
m
h
a
n
d
l
e
s
v
a
r
i
o
u
s
s
e
r
v
i
ce
s
b
a
s
e
d
o
n
p
r
o
c
e
s
s
i
n
g
p
la
t
f
o
r
m
s
r
e
q
u
e
s
t
e
d
b
y
u
s
er
s
[
5
]
.
C
C
i
s
t
h
e
in
t
e
g
r
a
t
io
n
o
f
p
ar
a
l
l
e
l
a
n
d
d
i
s
t
r
ib
u
te
d
c
o
m
p
u
t
in
g
w
i
t
h
r
e
s
o
u
r
c
e
s
h
ar
i
n
g
,
s
u
c
h
a
s
h
ar
d
w
ar
e
a
n
d
s
o
f
t
w
ar
e
,
w
h
i
ch
o
p
e
r
a
t
e
s
o
n
a
“
p
a
y
-
as
-
y
o
u
-
g
o
”
m
o
d
e
l
[
6
]
.
I
n
f
r
a
s
t
r
u
c
t
u
r
e
a
s
a
s
e
r
v
i
ce
(
I
a
a
S
)
i
s
a
t
e
ch
n
iq
u
e
b
eh
i
n
d
s
e
r
v
e
r
m
a
n
ag
e
m
e
n
t
,
d
a
t
a
c
e
n
t
e
r
s
,
a
n
d
v
ir
t
u
a
l
m
a
ch
i
n
es
(
V
M
s
)
[
7
]
.
I
a
a
S
p
r
o
v
i
d
e
s
a
c
l
o
u
d
s
e
r
v
i
c
e
t
h
a
t
g
iv
e
s
u
s
e
r
s
a
c
c
e
s
s
to
s
er
v
er
s
o
r
c
lo
u
d
co
m
p
u
t
er
s
to
p
r
o
ce
s
s
a
n
d
s
t
o
r
e
d
a
t
a.
U
s
e
r
s
c
a
n
o
p
er
a
t
e
ap
p
l
ic
a
t
i
o
n
s
o
r
s
er
v
er
s
o
n
r
en
t
a
l
s
e
r
v
e
r
s
,
w
i
t
h
m
a
i
n
te
n
an
c
e
an
d
h
a
r
d
w
ar
e
co
s
t
s
[
8
]
.
C
l
o
u
d
i
n
f
r
a
s
tr
u
c
tu
r
e
p
r
o
c
e
s
s
es
o
n
V
M
s
,
a
n
d
I
a
a
S
o
f
f
er
s
b
en
e
f
i
t
s
i
n
c
lu
d
in
g
p
r
o
v
id
i
n
g
u
s
e
r
a
c
c
e
s
s
t
o
s
e
r
v
e
r
s
i
n
th
e
c
l
o
s
e
s
t
lo
c
a
t
io
n
s
.
B
e
ca
u
s
e
o
f
t
h
e
d
y
n
a
m
i
c
n
a
tu
r
e
o
f
t
h
e
c
l
o
u
d
en
v
ir
o
n
m
en
t
,
l
o
a
d
s
f
lu
c
tu
a
t
e
o
v
e
r
t
i
m
e
,
m
ak
i
n
g
i
t
ch
a
l
l
e
n
g
i
n
g
to
a
l
lo
c
a
t
e
th
e
a
p
p
r
o
p
r
ia
t
e
VM
[
9
]
.
T
h
i
s
i
s
s
u
e
l
e
ad
s
to
h
i
g
h
co
s
t
s
,
lo
w
e
r
p
r
o
f
i
t
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
15
,
No
.
5
,
Octo
b
er
2
0
2
6
:
4
6
4
5
-
4
6
5
5
4646
f
o
r
o
r
g
a
n
i
za
t
i
o
n
s
,
a
n
d
r
ed
u
c
ed
m
o
d
e
l
p
er
f
o
r
m
a
n
c
e
[
1
0
]
.
E
a
c
h
p
h
y
s
i
c
a
l
m
a
c
h
in
e
co
n
s
u
m
e
s
e
l
e
c
t
r
i
c
a
l
p
o
w
er
b
a
s
e
d
o
n
th
e
ta
s
k
s
i
t
p
e
r
f
o
r
m
s
[
1
1
]
,
an
d
c
o
n
s
u
m
p
t
i
o
n
in
c
r
e
a
s
e
s
a
s
t
h
e
s
er
v
er
h
an
d
l
e
s
n
u
m
e
r
o
u
s
wo
r
k
l
o
ad
s
,
c
a
u
s
i
n
g
h
i
g
h
e
n
e
r
g
y
u
s
ag
e
[
1
2
]
.
T
o
m
i
t
i
g
a
t
e
th
i
s
i
s
s
u
e
,
a
n
e
f
f
e
c
t
i
v
e
s
c
h
ed
u
l
in
g
a
lg
o
r
i
t
h
m
i
s
r
e
q
u
i
r
e
d
t
o
m
a
n
ag
e
l
o
ad
b
e
t
w
ee
n
V
M
s
,
c
o
m
p
u
t
e
e
ac
h
i
n
co
m
in
g
t
a
s
k
w
i
t
h
l
e
s
s
c
o
m
p
l
e
t
io
n
t
i
m
e
,
a
n
d
r
e
d
u
c
e
e
n
e
r
g
y
c
o
n
s
u
m
p
t
io
n
i
n
th
e
d
a
t
a
c
e
n
ter
[
1
3
]
.
I
n
a
c
lo
u
d
en
v
ir
o
n
m
en
t,
in
c
o
m
in
g
ta
s
k
s
ar
e
a
s
s
ig
n
ed
to
th
e
p
r
o
p
er
VM
to
a
ll
o
ca
te
n
ec
e
s
s
a
r
y
r
es
o
u
r
c
es
f
o
r
co
m
p
u
ta
ti
o
n
[
1
4
]
.
T
h
o
u
g
h
,
wh
en
s
u
it
ab
l
e
r
eso
u
r
ce
s
a
r
e
u
n
av
ai
la
b
le
o
r
s
er
v
er
i
s
h
ea
v
ily
lo
ad
ed
,
in
co
m
in
g
t
ask
s
r
eq
u
ir
e
m
o
r
e
t
im
e
f
o
r
co
m
p
le
ti
o
n
[
1
5
]
.
T
r
ad
i
t
io
n
a
l
ap
p
r
o
a
ch
e
s
,
l
ik
e
r
o
u
n
d
-
r
o
b
in
an
d
we
ig
h
t
ed
l
o
ad
b
al
an
c
in
g
,
f
ai
l
to
ad
ap
t
to
r
ea
l
-
t
im
e
f
l
u
c
tu
at
io
n
s
in
r
e
s
o
u
r
ce
d
em
an
d
.
Ad
d
i
ti
o
n
a
lly
,
wh
il
e
h
eu
r
i
s
ti
c
-
b
a
s
ed
al
g
o
r
i
th
m
s
c
an
b
e
ef
f
ic
ien
t
in
c
er
t
ai
n
s
ce
n
ar
io
s
,
th
ey
l
ac
k
e
f
f
e
ct
iv
e
n
es
s
i
n
p
r
ed
ic
tin
g
wo
r
k
lo
ad
s
a
n
d
p
r
o
a
ct
iv
ely
a
d
ju
s
t
in
g
r
es
o
u
r
c
e
a
llo
ca
tio
n
.
T
h
e
s
e
li
m
i
ta
tio
n
s
r
e
s
u
lt
in
u
n
d
er
u
t
i
li
za
tio
n
o
f
r
es
o
u
r
c
es
,
in
cr
e
a
s
ed
r
esp
o
n
s
e
tim
e
s
,
an
d
d
eg
r
a
d
ed
q
u
al
ity
o
f
s
e
r
v
i
ce
(
Qo
S)
f
o
r
clo
u
d
u
s
e
r
s
.
T
o
m
it
ig
a
te
th
e
s
e
i
s
s
u
e
s
,
in
te
l
lig
en
t
lo
ad
b
ala
n
c
in
g
s
t
r
a
teg
ie
s
u
ti
l
iz
in
g
m
ac
h
in
e
le
ar
n
in
g
(
ML
)
an
d
d
ee
p
le
ar
n
in
g
(
DL
)
h
av
e
g
ain
ed
a
t
ten
tio
n
.
T
h
e
s
e
al
g
o
r
i
th
m
s
d
y
n
am
ic
al
ly
p
r
ed
i
ct
wo
r
k
lo
ad
p
a
tt
er
n
s
,
o
p
ti
m
i
ze
r
e
s
o
u
r
c
e
s
ch
ed
u
l
in
g
,
an
d
i
m
p
r
o
v
e
o
v
er
al
l
ef
f
ec
tiv
en
e
s
s
.
T
h
i
s
m
an
u
s
cr
ip
t
a
im
s
to
d
ev
e
lo
p
an
ad
ap
tiv
e
lo
ad
b
a
lan
cin
g
m
ec
h
an
i
s
m
th
at
co
m
b
in
e
s
ML
/
DL
a
lg
o
r
ith
m
s
to
ad
d
r
es
s
ex
i
s
t
in
g
li
m
i
ta
tio
n
s
an
d
en
h
a
n
ce
r
e
s
o
u
r
ce
d
i
s
tr
ib
u
tio
n
in
I
aa
S
-
en
ab
led
c
l
o
u
d
en
v
ir
o
n
m
en
t
s
.
J
en
a
et
a
l.
[
1
6
]
s
u
g
g
ested
th
e
h
y
b
r
id
izatio
n
o
f
m
o
d
if
ied
p
a
r
ticle
s
war
m
o
p
tim
izatio
n
(
MPSO)
an
d
an
en
h
an
ce
d
Q
-
lear
n
in
g
ap
p
r
o
ac
h
,
n
am
ed
QM
PS
O,
f
o
r
d
y
n
am
ic
lo
ad
b
alan
cin
g
b
e
twee
n
VM
s
.
T
h
e
h
y
b
r
id
izatio
n
p
r
o
ce
s
s
ad
ju
s
ted
th
e
MPSO
v
elo
city
th
r
o
u
g
h
p
b
est
an
d
g
b
est
v
al
u
es
o
f
b
ette
r
ac
tio
n
s
p
r
o
d
u
ce
d
b
y
en
h
a
n
ce
d
Q
-
lear
n
in
g
.
T
h
e
g
o
al
o
f
th
e
h
y
b
r
id
izatio
n
wa
s
to
en
h
an
ce
th
e
m
eth
o
d
’
s
p
e
r
f
o
r
m
a
n
ce
th
r
o
u
g
h
b
alan
cin
g
lo
ad
b
etwe
en
VM
s
,
in
cr
ea
s
in
g
VM
th
r
o
u
g
h
p
u
t
,
an
d
m
ain
tain
in
g
task
p
r
io
r
ity
b
alan
ce
th
r
o
u
g
h
o
p
tim
izin
g
task
waitin
g
tim
es.
Ho
wev
er
,
th
e
s
u
g
g
ested
m
et
h
o
d
co
n
s
u
m
ed
a
s
ig
n
if
ica
n
t
am
o
u
n
t
o
f
e
n
er
g
y
d
u
e
to
th
e
h
ea
v
y
lo
a
d
in
th
e
n
etwo
r
k
.
Pra
d
h
a
n
an
d
B
is
o
y
[
1
7
]
p
r
esen
ted
a
lo
ad
b
alan
cin
g
m
eth
o
d
ca
lled
m
o
d
i
f
ied
p
ar
ticle
s
war
m
o
p
tim
izatio
n
(
L
B
MPSO)
,
wh
ich
s
ch
ed
u
led
task
s
ac
r
o
s
s
ex
is
tin
g
clo
u
d
r
eso
u
r
ce
s
,
r
ed
u
ce
d
m
ak
esp
an
,
an
d
in
c
r
ea
s
ed
r
eso
u
r
ce
u
tili
za
tio
n
.
T
h
is
was
a
ch
iev
ed
th
r
o
u
g
h
ap
p
r
o
p
r
iate
m
atch
in
g
b
etwe
e
n
r
eso
u
r
ce
s
an
d
task
s
with
in
th
e
d
ata
ce
n
ter
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
p
e
r
f
o
r
m
ed
b
etter
in
m
in
im
izin
g
m
ak
esp
an
an
d
m
ax
im
izin
g
r
eso
u
r
ce
u
tili
za
tio
n
.
Ho
wev
er
,
it
co
n
s
u
m
e
d
ex
ce
s
s
iv
e
tim
e
d
u
e
to
h
ea
v
y
lo
ad
co
n
d
itio
n
s
.
Har
is
an
d
Z
u
b
air
[
1
8
]
em
p
lo
y
ed
lo
ad
b
alan
cin
g
ap
p
r
o
ac
h
d
e
p
en
d
in
g
o
n
h
y
b
r
id
a
p
p
r
o
ac
h
es
,
n
am
ely
m
an
tar
a
y
m
o
d
if
ied
m
u
lti
-
o
b
jectiv
e
Har
r
is
Haw
k
o
p
tim
izatio
n
(
MM
H
HO)
.
T
h
is
h
y
b
r
i
d
izatio
n
e
n
h
a
n
ce
d
Har
r
is
Haw
k
o
p
tim
izatio
n
(
HHO
)
s
ea
r
ch
t
h
r
o
u
g
h
in
teg
r
atin
g
th
e
m
an
ta
r
ay
f
o
r
ag
in
g
o
p
tim
izatio
n
(
MRF
O)
ap
p
r
o
ac
h
,
co
n
s
id
er
in
g
f
ac
to
r
s
lik
e
r
es
p
o
n
s
e
tim
e,
co
s
t,
an
d
r
eso
u
r
ce
u
s
ag
e.
T
h
e
d
ev
el
o
p
ed
m
eth
o
d
im
p
r
o
v
e
d
p
er
f
o
r
m
an
ce
b
y
im
p
r
o
v
i
n
g
th
e
th
r
o
u
g
h
p
u
t
o
f
VM
s
,
b
alan
ci
n
g
lo
ad
b
etwe
en
VM
s
,
an
d
m
ain
tain
in
g
b
alan
ce
b
etwe
en
task
p
r
io
r
ities
th
r
o
u
g
h
ad
ju
s
tin
g
task
waitin
g
tim
es.
T
h
o
u
g
h
,
th
e
m
eth
o
d
s
till
f
ac
ed
is
s
u
es
r
elate
d
to
lo
ad
b
alan
cin
g
an
d
s
u
b
o
p
tim
al
r
eso
u
r
ce
u
tili
za
tio
n
.
So
u
r
av
las
et
a
l.
[
1
9
]
im
p
le
m
e
n
ted
a
wo
r
k
lo
ad
-
p
er
-
VM
s
tr
ateg
y
f
o
r
task
lo
ad
b
alan
cin
g
,
w
h
ich
aim
ed
to
en
h
a
n
ce
m
ea
n
r
esp
o
n
s
e
tim
e
an
d
s
y
s
tem
m
ak
esp
an
i
n
th
e
clo
u
d
.
T
h
e
is
s
u
e
was
f
o
r
m
u
lated
as
a
f
i
n
ite
Ma
r
k
o
v
p
r
o
ce
s
s
,
r
ef
e
r
r
ed
t
o
as
th
e
b
alan
ce
eq
u
atio
n
f
o
r
e
ac
h
s
tate.
B
y
b
alan
cin
g
t
h
e
p
o
ten
tial
s
tates,
th
e
ex
p
ec
ted
u
s
ag
e
o
f
VM
s
co
u
ld
b
e
d
e
r
iv
ed
,
p
lay
in
g
an
ess
en
tial
r
o
le
in
th
e
task
allo
ca
tio
n
p
r
o
ce
s
s
.
T
h
e
lo
a
d
b
alan
ce
r
ac
te
d
as
th
e
ce
n
tr
al
s
er
v
er
,
u
tili
zin
g
f
air
task
allo
ca
tio
n
to
d
is
tr
ib
u
te
in
co
m
in
g
task
s
,
b
alan
ce
th
e
VM
s
,
an
d
ac
co
u
n
t
f
o
r
t
h
eir
c
u
r
r
en
t
s
tate
an
d
p
r
o
ce
s
s
ab
ilit
y
.
Ho
wev
er
,
th
e
im
p
lem
en
te
d
m
eth
o
d
h
a
d
h
ig
h
lo
ad
s
o
n
VM
s
b
ec
au
s
e
it
d
id
n
o
t
co
n
s
id
er
o
p
tim
al
s
o
lu
tio
n
s
.
L
ah
an
d
e
et
a
l.
[
2
0
]
in
tr
o
d
u
c
ed
a
lo
ad
b
alan
cin
g
m
ec
h
an
is
m
b
y
ex
p
er
im
en
tin
g
an
d
e
x
ec
u
tin
g
task
s
u
tili
zin
g
th
e
Sip
h
t
task
d
ataset.
T
h
e
R
S
ap
p
r
o
ac
h
es:
f
ir
s
t
co
m
e
f
ir
s
t
s
er
v
e
d
(
FC
FS
)
,
m
ax
im
u
m
-
m
in
im
u
m
(
m
ax
-
m
in
)
,
m
ax
im
u
m
co
m
p
letio
n
tim
e
(
MCT)
,
an
d
r
o
u
n
d
r
o
b
in
(
R
R
)
wer
e
co
n
s
id
er
e
d
f
o
r
b
alan
cin
g
ex
ec
u
tio
n
lo
ad
s
ac
r
o
s
s
VM
s
.
T
h
e
ex
p
er
im
en
t
was
p
er
f
o
r
m
ed
i
n
f
o
u
r
s
tag
es,
with
ea
ch
s
tag
e
in
clu
d
in
g
1
6
s
ce
n
ar
io
s
,
wh
er
e
e
ac
h
s
ce
n
ar
io
v
ar
ie
d
b
ased
o
n
t
h
e
n
u
m
b
er
o
f
VM
s
u
tili
ze
d
.
Ho
wev
er
,
th
e
in
tr
o
d
u
ce
d
m
eth
o
d
h
ad
h
i
g
h
en
er
g
y
co
n
s
u
m
p
tio
n
an
d
m
a
k
esp
an
d
u
e
to
h
ea
v
y
l
o
ad
co
n
d
itio
n
s
.
B
y
u
tili
zin
g
th
e
a
r
t
if
icial
b
ee
co
lo
n
y
with
Q
-
lea
r
n
in
g
(
AB
C
Q)
ap
p
r
o
ac
h
,
o
n
ce
a
b
etter
r
ewa
r
d
was
o
b
tain
ed
,
th
e
m
et
h
o
d
s
o
u
g
h
t
t
o
ac
h
iev
e
a
g
lo
b
al
o
p
tim
al
r
e
war
d
in
less
co
m
p
u
tatio
n
tim
e
b
y
ex
ch
a
n
g
in
g
d
ata
with
n
eig
h
b
o
r
in
g
n
o
d
es.
T
h
e
e
s
s
en
tial c
o
n
tr
ib
u
tio
n
s
o
f
th
is
r
esear
ch
ar
e
d
escr
ib
ed
as f
o
llo
ws:
T
h
e
d
ee
p
r
ein
f
o
r
ce
m
en
t
lear
n
in
g
(
DR
L
)
–
AB
C
Q
ap
p
r
o
ac
h
is
p
r
o
p
o
s
ed
to
b
alan
ce
lo
a
d
b
etwe
en
VM
s
in
clo
u
d
,
m
ax
im
izin
g
r
e
war
d
th
r
o
u
g
h
m
in
im
izin
g
m
a
k
esp
an
tim
e
an
d
en
er
g
y
c
o
n
s
u
m
p
tio
n
.
T
h
e
ag
e
n
t in
DR
L
m
ak
es
in
tellig
en
t
d
ec
is
io
n
s
to
h
a
n
d
le
h
u
g
e
n
u
m
b
er
o
f
task
s
in
less
tim
e
an
d
at
h
ig
h
er
s
p
ee
d
.
T
h
e
ar
tific
ial
b
ee
co
l
o
n
y
(
AB
C
)
alg
o
r
ith
m
u
s
es
th
e
g
l
o
b
al
g
u
id
in
g
f
ac
to
r
in
th
e
p
o
s
itio
n
u
p
d
atin
g
p
h
ase,
wh
ic
h
im
p
r
o
v
es th
e
s
ea
r
c
h
ab
ilit
y
o
f
t
r
ad
itio
n
al
AB
C
in
f
in
d
in
g
o
p
ti
m
al
s
o
lu
tio
n
s
.
T
h
is
r
esear
ch
p
a
p
er
is
o
r
g
a
n
ized
as
f
o
llo
ws.
Sectio
n
2
p
r
esen
ts
th
e
d
etails
o
f
t
h
e
p
r
o
p
o
s
ed
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
.
Sectio
n
3
p
r
esen
ts
th
e
o
u
tco
m
es
an
d
d
is
cu
s
s
io
n
.
Fin
ally
,
s
ec
ti
o
n
4
p
r
o
v
i
d
es
th
e
co
n
clu
s
io
n
s
o
f
t
h
is
r
esear
ch
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Op
timiz
in
g
clo
u
d
en
erg
y
w
ith
d
ee
p
r
ein
fo
r
ce
men
t le
a
r
n
in
g
:
a
n
a
r
tifi
cia
l b
ee
…
(
Mo
h
a
mme
d
Zia
u
r
R
a
h
ma
n
)
4647
2.
M
E
T
H
O
D
T
h
is
p
ap
er
p
r
o
p
o
s
es
th
e
DR
L
-
AB
C
Q
s
ch
ed
u
lin
g
ap
p
r
o
ac
h
,
wh
e
r
e
th
e
ag
e
n
t
r
ec
ei
v
es
a
s
ep
ar
ate
r
ewa
r
d
b
y
c
o
n
s
id
er
in
g
th
e
ac
tio
n
tak
en
in
ev
er
y
s
tate
o
f
t
h
e
en
v
ir
o
n
m
e
n
t.
T
h
e
AB
C
Q
ap
p
r
o
ac
h
aim
s
to
ac
h
iev
e
a
g
lo
b
ally
b
etter
r
ewa
r
d
with
less
co
m
p
u
tatio
n
tim
e
b
y
ex
ch
an
g
in
g
d
ata
with
n
eig
h
b
o
r
in
g
n
o
d
es
af
ter
o
b
tain
in
g
b
etter
in
d
i
v
id
u
al
r
ewa
r
d
s
.
I
n
itially
,
th
e
ap
p
r
o
p
r
iate
ac
tio
n
is
ch
o
s
en
b
y
co
n
s
id
er
in
g
th
e
v
alu
e
f
u
n
ctio
n
o
f
s
tate
-
ac
tio
n
,
an
d
t
h
e
p
er
s
o
n
al
b
etter
r
ewa
r
d
is
t
h
en
o
b
tain
e
d
f
o
r
ev
er
y
s
er
v
er
,
wh
ich
o
p
tim
izes
lo
ad
an
d
its
p
ar
a
m
eter
s
.
Af
ter
ac
h
iev
in
g
p
e
r
s
o
n
al
b
etter
r
e
war
d
s
,
ev
er
y
s
er
v
er
id
en
tifie
s
g
lo
b
al
b
etter
v
alu
e
th
r
o
u
g
h
s
h
ar
in
g
th
eir
d
ata
b
et
wee
n
s
er
v
er
s
at
h
ig
h
s
p
ee
d
.
T
h
e
f
o
llo
win
g
s
ec
tio
n
ex
p
lain
s
th
e
r
ewa
r
d
f
u
n
ctio
n
,
DR
L
,
Q
-
lear
n
in
g
,
a
n
d
AB
C
alg
o
r
ith
m
s
.
Fig
u
r
e
1
r
ep
r
esen
ts
a
p
r
o
ce
s
s
o
f
lo
ad
b
alan
cin
g
in
th
e
clo
u
d
u
s
in
g
t
h
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
.
Fig
u
r
e
1
.
Pro
ce
s
s
o
f
lo
ad
b
alan
cin
g
in
cl
o
u
d
2
.
1
.
Rew
a
rd
f
un
ct
io
n
T
h
e
p
r
o
b
lem
o
f
lo
ad
b
ala
n
c
in
g
is
d
e
f
in
ed
as
Ma
r
k
o
v
d
ec
is
io
n
p
r
o
ce
s
s
(
MD
P)
b
ec
a
u
s
e
o
f
th
e
s
im
u
ltan
eo
u
s
n
atu
r
e
o
f
task
s
in
a
clo
u
d
en
v
ir
o
n
m
en
t.
I
n
a
d
ev
elo
p
e
d
ap
p
r
o
ac
h
,
d
ata
ce
n
ter
co
n
tr
o
ller
is
d
escr
ib
ed
as
an
ag
en
t,
an
d
d
at
a
ce
n
ter
its
elf
is
d
escr
ib
ed
as
t
h
e
en
v
ir
o
n
m
en
t,
wh
e
r
e
ag
en
t
allo
ca
tes
in
co
m
in
g
task
s
to
th
e
ap
p
r
o
p
r
iate
VM
in
ev
e
r
y
c
y
cle.
T
h
e
MD
P
in
clu
d
es
f
o
u
r
s
ig
n
i
f
ican
t
v
a
r
iab
les,
n
am
ely
s
tate
,
ac
tio
n
,
tr
an
s
itio
n
f
u
n
ctio
n
,
an
d
r
ewa
r
d
,
wh
ich
ar
e
ex
p
lain
ed
a
s
f
o
llo
ws
[
2
1
]
.
2
.
1
.
1
.
Sta
t
e
a
nd
a
ct
io
n sp
a
ce
I
n
th
e
s
tate
s
p
ac
e,
a
p
er
f
ec
t
m
o
v
e
is
m
ad
e
b
y
th
e
ag
en
t
d
ep
e
n
d
in
g
o
n
VM
d
ata
lik
e
am
o
u
n
t
o
f
VM
s
,
th
e
ex
is
ten
ce
o
f
m
illi
o
n
s
o
f
i
n
s
tr
u
ctio
n
s
p
er
s
ec
o
n
d
(
MI
PS
)
,
th
e
ce
n
tr
al
p
r
o
ce
s
s
in
g
u
n
it
(
C
PU)
,
b
an
d
wid
th
,
an
d
m
em
o
r
y
.
T
h
e
m
ath
em
atic
al
ex
p
r
ess
io
n
f
o
r
th
e
s
tate
s
p
ac
e
is
g
iv
en
in
(
1
)
.
I
n
th
e
ac
tio
n
s
p
ac
e,
ev
er
y
task
is
em
p
lo
y
ed
to
VM
f
o
r
p
r
o
ce
s
s
in
g
.
E
v
er
y
ac
tio
n
in
clu
d
es
d
if
f
er
en
t
p
iece
s
o
f
d
ata
lik
e
n
u
m
b
er
o
f
task
s
,
len
g
th
o
f
task
an
d
s
ize
o
f
th
e
f
ile.
T
h
e
m
ath
em
atica
l e
x
p
r
ess
io
n
f
o
r
th
e
ac
tio
n
s
p
ac
e
is
g
iv
en
in
(
2
)
.
=
{
,
,
,
,
}
(
1
)
=
{
,
,
}
(
2
)
2
.
1
.
2
.
T
ra
ns
it
io
n f
un
ct
io
n a
n
d a
ct
io
n select
io
n
T
h
is
p
h
ase
r
ep
r
esen
ts
wh
en
th
e
ag
en
t
co
n
s
id
er
s
th
e
ac
tio
n
in
th
e
p
r
esen
t
s
tate,
wh
ich
r
ea
ch
e
s
th
e
n
ew
s
tate.
T
h
e
ag
en
t
alwa
y
s
a
ttem
p
ts
to
s
elec
t
th
e
ap
p
r
o
p
r
ia
te
ac
tio
n
th
at
r
esu
lts
in
an
o
p
t
im
u
m
s
tate,
wh
ich
p
r
o
v
id
es
a
h
ig
h
est
r
ewa
r
d
.
T
h
e
tr
a
n
s
itio
n
f
u
n
ctio
n
is
d
escr
ib
ed
as
(
,
,
,
+
1
)
.
I
t
r
ep
r
esen
ts
th
e
p
o
s
s
ib
ilit
y
o
f
r
ea
c
h
in
g
th
e
f
o
llo
win
g
s
tate
+
1
an
d
o
b
tain
in
g
th
e
r
ewa
r
d
af
ter
ex
ec
u
tin
g
th
e
c
h
o
s
en
ac
tio
n
a
t
in
th
e
p
r
esen
t
s
tate
.
Du
e
to
t
h
e
clo
u
d
’
s
d
y
n
a
m
ic
n
atu
r
e,
ta
s
k
s
v
ar
y
with
r
esp
ec
t
to
th
e
le
n
g
th
,
tim
e
a
n
d
s
ize
o
f
a
f
ile
[
2
2
]
–
[
2
4
]
.
C
er
tai
n
VM
s
ar
e
o
v
er
lo
a
d
ed
,
wh
ile
s
o
m
e
ar
e
u
n
d
er
lo
ad
e
d
,
as
d
ef
i
n
ed
u
s
in
g
(
3
)
.
Fo
r
b
alan
cin
g
lo
a
d
b
etwe
en
VM
s
,
th
e
ag
en
t
c
o
n
s
id
er
s
an
ac
tio
n
to
tr
an
s
f
er
ex
t
r
a
task
s
f
r
o
m
an
o
v
er
lo
a
d
ed
VM
to
u
n
d
er
lo
a
d
ed
VM
at
h
i
g
h
s
p
ee
d
.
T
h
e
s
p
ee
d
o
f
task
tr
a
n
s
f
er
is
m
ea
s
u
r
ed
u
s
in
g
(
4
)
.
=
{
<
|
|
>
|
|
=
|
|
(
3
)
=
∑
∑
∑
(
1
−
,
,
)
,
,
=
1
=
1
=
1
(
4
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
15
,
No
.
5
,
Octo
b
er
2
0
2
6
:
4
6
4
5
-
4
6
5
5
4648
W
h
er
e
,
r
ep
r
esen
ts
th
e
ad
d
itio
n
al
task
th
at
is
tr
an
s
f
er
r
ed
b
etwe
en
to
VM
;
,
r
ep
r
esen
ts
th
e
r
eq
u
ir
ed
b
an
d
wid
th
b
etwe
en
an
d
VM
.
Her
e,
,
=
1
s
ig
n
if
ies
th
at
VM
is
o
v
er
lo
ad
ed
,
wh
er
e
,
=
1
r
ep
r
esen
ts
th
at
th
e
e
x
tr
a
task
i
s
tr
an
s
f
er
r
ed
f
r
o
m
to
VM
.
W
h
en
task
t
r
an
s
f
er
s
p
ee
d
is
in
c
r
ea
s
ed
,
task
co
m
p
letio
n
tim
e
also
r
e
d
u
ce
s
.
T
h
e
ch
o
s
en
ac
tio
n
in
iter
atio
n
is
m
ea
s
u
r
ed
b
y
(
5
)
.
=
(
)
(
5
)
2
.
1
.
3
.
Rew
a
rd
I
n
th
is
p
h
ase,
th
e
ag
e
n
t
r
ec
ei
v
es
a
r
ewa
r
d
f
o
r
m
ak
in
g
m
o
v
es
in
d
if
f
er
en
t
s
tates.
T
h
e
a
g
e
n
t
aim
s
to
s
elec
t
th
e
s
tate
th
at
y
ield
s
h
ig
h
er
r
ewa
r
d
s
,
th
er
e
b
y
in
cr
ea
s
in
g
ac
cu
r
ac
y
.
I
n
th
e
p
r
o
p
o
s
ed
m
eth
o
d
,
th
e
r
ewa
r
d
f
u
n
ctio
n
is
ch
ar
ac
ter
ize
d
th
r
o
u
g
h
a
r
ed
u
ctio
n
in
task
co
m
p
letio
n
tim
e
o
n
a
p
a
r
ticu
l
ar
VM
an
d
en
er
g
y
co
n
s
u
m
p
tio
n
.
T
h
e
m
ath
em
atic
al
ex
p
r
ess
io
n
d
escr
ib
in
g
th
e
r
e
war
d
is
g
iv
en
in
(
6
)
.
=
{
,
+
}
(
6
)
2
.
2
.
Q
-
lea
rning
a
pp
ro
a
ch
Q
-
lear
n
in
g
is
a
r
ein
f
o
r
ce
m
en
t le
ar
n
in
g
(
R
L
)
ap
p
r
o
ac
h
t
h
at
h
elp
s
th
e
ag
en
t
lear
n
h
o
w
to
ac
t
o
p
tim
al
ly
in
a
g
iv
en
en
v
ir
o
n
m
en
t.
Q
-
lea
r
n
in
g
aim
s
to
id
en
tify
th
e
o
p
tim
al
p
o
licy
f
o
r
ac
tio
n
s
elec
tio
n
th
at
m
in
im
izes
th
e
ex
p
ec
ted
s
u
m
o
f
r
ewa
r
d
s
f
o
r
th
e
ag
en
t.
C
o
n
s
id
er
a
s
et
o
f
s
tates
=
{
₁
,
₂
,
₃
,
.
.
.
,
ₙ
}
in
th
e
en
v
ir
o
n
m
en
t,
wh
er
e
ev
e
r
y
s
tate
co
n
tain
s
a
s
et
o
f
ac
tio
n
s
.
T
h
e
ag
en
t
=
{
₁
,
₂
,
₃
,
.
.
.
,
ₘ
}
ch
o
o
s
es
an
ac
tio
n
ₜ
∈
at
tim
e
in
s
tate
ₜ
∈
to
tr
a
n
s
itio
n
to
th
e
n
ex
t
s
tate
ₜ₊
₁
∈
,
r
ec
eiv
in
g
th
e
r
ewa
r
d
ₜ₊
₁
f
r
o
m
th
e
en
v
ir
o
n
m
en
t.
Fo
r
task
p
r
o
ce
s
s
in
g
,
it
is
ess
en
tial
to
ch
o
o
s
e
r
e
lev
an
t
ac
tio
n
s
to
in
cr
ea
s
e
Q
-
v
alu
e
o
f
ev
er
y
s
tate,
th
at
is
cr
u
cial
f
o
r
id
en
tify
i
n
g
t
h
e
o
p
tim
u
m
p
o
licy
in
th
e
clo
u
d
en
v
ir
o
n
m
e
n
t.
T
h
e
Q
-
v
alu
e
f
u
n
ctio
n
is
b
ased
o
n
th
e
ac
tio
n
ch
o
s
en
in
a
s
tate.
Fig
u
r
e
2
illu
s
tr
ates
th
e
p
r
o
ce
s
s
o
f
Q
-
lear
n
in
g
.
Fig
u
r
e
2
.
Pro
ce
s
s
o
f
Q
-
lear
n
in
g
C
o
n
s
id
er
an
ag
en
t
in
s
tate
ₜ
an
d
th
e
ch
o
s
e
n
ac
tio
n
ₜ
.
T
h
e
p
r
o
c
ess
o
f
Q
-
v
alu
e
is
d
ir
ec
ted
to
w
ar
d
s
a
b
etter
s
tate
,
m
ax
im
izin
g
to
tal
r
ewa
r
d
in
t
h
e
en
v
ir
o
n
m
en
t.
T
h
e
m
ath
e
m
atica
l
ex
p
r
ess
io
n
f
o
r
th
e
Q
-
v
alu
e
is
g
iv
en
in
(
7
)
,
wh
er
e
r
ep
r
esen
t
s
lear
n
in
g
r
ate,
r
ep
r
esen
ts
th
e
d
is
co
u
n
t
f
ac
to
r
,
w
h
ich
in
f
lu
en
ce
s
s
u
cc
ess
iv
e
s
tates
b
ased
o
n
p
ast ac
tio
n
s
,
an
d
ₜ
r
ep
r
esen
ts
th
e
r
ewa
r
d
o
b
ta
in
ed
f
r
o
m
p
r
o
ce
s
s
in
g
th
e
ac
tio
n
in
s
tate
ₜ
.
(
,
)
=
(
1
−
)
(
,
)
+
[
+
+
1
(
+
1
,
+
1
)
]
(
7
)
T
h
e
Q
-
v
a
l
u
e
i
s
d
e
r
i
v
e
d
f
r
o
m
Q
-
t
a
b
l
e
,
w
h
i
c
h
s
t
o
r
es
a
ll
p
r
o
b
a
b
l
e
s
t
a
t
e
s
,
t
h
e
i
r
c
o
r
r
es
p
o
n
d
i
n
g
Q
-
v
a
l
u
e
s
,
a
n
d
t
h
e
c
o
r
r
e
c
t
a
c
t
i
o
n
s
.
T
h
e
Q
-
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
a
t
t
e
m
p
t
s
t
o
es
t
i
m
at
e
t
h
e
o
p
t
i
m
u
m
s
t
a
t
e
f
r
o
m
i
t
s
e
x
p
e
r
i
e
n
c
e
a
n
d
i
n
c
o
r
p
o
r
a
t
e
s
a
g
r
e
e
d
y
s
t
r
a
t
e
g
y
w
i
t
h
i
n
t
h
e
Q
-
l
e
a
r
n
i
n
g
p
r
o
c
e
s
s
.
T
h
e
m
a
t
h
e
m
a
t
i
ca
l
e
x
p
r
e
s
s
i
o
n
f
o
r
c
o
m
p
u
t
i
n
g
t
h
e
Q
-
v
a
l
u
e
i
s
g
i
v
e
n
i
n
(
8
)
,
w
h
e
r
e
r
e
p
r
e
s
e
n
t
s
t
h
e
l
e
a
r
n
i
n
g
r
a
t
e
a
n
d
r
e
p
r
e
s
e
n
t
s
t
h
e
t
r
a
n
s
i
ti
o
n
f
u
n
c
t
i
o
n
.
(
+
1
)
(
,
)
=
(
,
)
+
[
+
〖
〗
_
[
(
(
,
)
,
́
_
)
]
−
(
,
)
(
8
)
2
.
2
.
1
.
F
it
nes
s
f
un
ct
io
n
T
h
e
f
itn
ess
f
u
n
ctio
n
o
f
t
h
e
D
R
L
-
AB
C
Q
ap
p
r
o
ac
h
is
b
ased
o
n
m
a
k
esp
an
,
waitin
g
tim
e
,
a
n
d
r
eso
u
r
ce
u
tili
za
tio
n
.
T
h
e
m
ath
em
atica
l
ex
p
r
ess
io
n
f
o
r
t
h
e
f
itn
ess
f
u
n
c
tio
n
is
g
iv
en
in
(
9
)
.
Ma
k
esp
an
is
m
in
im
ized
wh
en
th
e
h
ig
h
est
n
u
m
b
er
o
f
task
s
ar
e
s
u
cc
ess
f
u
lly
p
r
o
ce
s
s
ed
in
th
e
d
ata
ce
n
ter
,
wh
ich
i
m
p
ac
ts
th
e
m
o
d
el’
s
th
r
o
u
g
h
p
u
t.
MS
r
ep
r
esen
ts
th
e
to
tal
co
m
p
letio
n
tim
e
f
o
r
ea
c
h
task
,
an
d
th
e
m
at
h
em
atica
l
ex
p
r
ess
io
n
is
g
iv
en
in
(
1
0
)
.
Ne
x
t,
ᵤ
is
m
ax
im
ized
,
an
d
its
m
ath
em
atica
l
ex
p
r
ess
io
n
is
g
iv
en
in
(
1
1
)
.
ᵤ
is
m
ea
s
u
r
ed
as
th
e
to
tal
co
m
p
letio
n
tim
e
o
f
th
e
task
s
d
iv
id
ed
b
y
t
h
e
h
ig
h
est
m
a
k
esp
an
o
r
ex
ec
u
tio
n
tim
e.
W
h
er
e
is
f
itn
ess
f
u
n
ctio
n
,
is
wai
tin
g
tim
e
,
is
m
ak
esp
an
,
an
d
ᵤ
is
r
eso
u
r
ce
u
tili
za
tio
n
.
T
h
e
alg
o
r
ith
m
f
o
r
Q
-
u
p
d
ate
is
g
iv
en
as
in
Alg
o
r
ith
m
1
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Op
timiz
in
g
clo
u
d
en
erg
y
w
ith
d
ee
p
r
ein
fo
r
ce
men
t le
a
r
n
in
g
:
a
n
a
r
tifi
cia
l b
ee
…
(
Mo
h
a
mme
d
Zia
u
r
R
a
h
ma
n
)
4649
=
+
+
(
9
)
=
{
}
(
1
0
)
(
)
=
(
1
1
)
Alg
o
r
ith
m
1
.
Fo
r
Q
-
v
al
u
e
u
p
d
atin
g
Initialize values to
α
(learning rate),
(discount factor) and
(reward)
Initialize the Q
-
values
for
=
1
for
=
1
(
,
)
=
0
end for
end for
choose action
a
I
form the agent
=
{
1
,
2
,
3
,
…
,
}
and compute that and to following state
+
1
Calculate learning rate
Compute reward
Update
+
1
using (8)
Repeat this phase for the new state till that converges
2
.
3
.
Art
if
ici
a
l
bee
co
lo
ny
wit
h g
lo
ba
l g
uid
ing
f
a
ct
o
r
T
h
e
AB
C
alg
o
r
ith
m
with
a
g
lo
b
al
g
u
id
i
n
g
f
ac
to
r
is
a
s
war
m
in
tellig
en
ce
tech
n
iq
u
e
wh
i
ch
m
im
ics
f
o
r
ag
in
g
b
eh
av
i
o
r
o
f
b
ee
s
.
T
h
e
alg
o
r
ith
m
is
d
iv
id
ed
to
th
r
ee
p
h
ases
:
lead
er
,
f
o
llo
wer
,
an
d
s
co
u
ter
.
T
h
e
lead
er
r
ep
r
esen
ts
th
e
f
o
o
d
s
o
u
r
ce
an
d
ca
r
r
ies
in
f
o
r
m
atio
n
ab
o
u
t
it.
I
n
AB
C
,
ag
en
ts
ar
e
class
if
ied
b
ased
o
n
th
e
r
o
les
o
f
b
ee
s
:
em
p
lo
y
ed
,
o
n
lo
o
k
er
,
an
d
s
co
u
t
b
ee
s
[
2
5
]
.
E
ac
h
b
ee
g
r
o
u
p
ex
h
ib
its
d
if
f
er
en
t
b
eh
av
io
r
s
r
elate
d
to
ex
p
lo
r
atio
n
an
d
e
x
p
lo
itatio
n
.
T
h
e
ex
p
lo
itatio
n
b
eh
a
v
io
r
in
v
o
lv
es
s
ea
r
ch
in
g
f
o
r
n
ew
f
o
o
d
s
o
u
r
ce
s
to
av
o
id
lo
ca
l
o
p
tim
a,
wh
ile
th
e
ex
p
l
o
r
atio
n
b
eh
a
v
io
r
f
o
cu
s
es
o
n
s
ea
r
ch
in
g
f
o
r
o
p
tim
al
f
o
o
d
s
o
u
r
c
es
n
ea
r
th
e
cu
r
r
en
t
o
n
e.
I
n
th
is
r
esear
ch
,
Q
-
lear
n
i
n
g
is
u
tili
ze
d
to
en
h
an
ce
th
e
s
o
lu
tio
n
s
an
d
o
b
tain
m
an
y
co
r
r
ec
t
s
o
lu
tio
n
s
to
th
e
p
r
o
b
lem
.
T
h
e
AB
C
alg
o
r
ith
m
is
ch
ar
ac
ter
ized
th
r
o
u
g
h
s
tr
o
n
g
e
x
p
lo
r
atio
n
b
eh
av
io
r
a
n
d
wea
k
er
ex
p
lo
itatio
n
b
eh
av
io
r
;
th
u
s
,
g
o
al
is
to
o
p
ti
m
ize
th
e
ex
p
lo
itatio
n
b
eh
av
io
r
.
Du
r
in
g
th
e
in
itializatio
n
s
tag
e,
AB
C
alg
o
r
ith
m
d
ef
in
es
a
p
o
s
itio
n
o
f
f
o
o
d
s
o
u
r
ce
with
p
o
s
s
ib
le
s
o
lu
tio
n
s
to
is
s
u
e.
First,
th
e
f
o
o
d
s
o
u
r
ce
is
r
an
d
o
m
l
y
p
r
o
d
u
ce
d
,
a
n
d
th
e
m
ath
em
atic
al
ex
p
r
ess
io
n
f
o
r
th
is
is
g
iv
en
i
n
(
1
2
)
.
,
0
=
+
(
0
,
1
)
×
(
−
)
(
1
2
)
W
h
er
e
d
ef
in
es
lo
wer
b
o
u
n
d
o
f
ℎ
o
p
tim
izatio
n
p
ar
am
ete
r
an
d
d
ef
in
es
th
e
u
p
p
er
b
o
u
n
d
o
f
ℎ
p
ar
am
eter
o
f
o
p
tim
izatio
n
.
Ad
d
itio
n
ally
,
th
e
in
itializatio
n
s
tag
e
AB
C
alg
o
r
ith
m
d
i
v
id
es
alg
o
r
ith
m
in
t
o
th
r
ee
s
u
b
-
s
tag
es
:
em
p
lo
y
ed
,
o
n
lo
o
k
er
,
an
d
s
co
u
t
b
ee
p
h
ases
.
T
h
e
ap
p
r
o
ac
h
r
e
p
ea
ts
th
e
th
r
ee
s
tag
es
u
n
til
th
e
h
ig
h
est
n
u
m
b
e
r
o
f
iter
atio
n
s
is
attain
ed
.
I
n
em
p
lo
y
ee
b
ee
s
tag
e
,
th
e
alg
o
r
ith
m
id
e
n
tifie
s
p
o
s
itio
n
o
f
n
eig
h
b
o
r
in
g
f
o
o
d
s
o
u
r
ce
r
ela
tiv
e
to
th
e
cu
r
r
en
t
f
o
o
d
s
o
u
r
ce
.
T
h
e
cu
r
r
en
t
lo
ca
tio
n
u
p
d
ate
ex
p
r
ess
io
n
h
as
s
tr
o
n
g
s
ea
r
ch
ca
p
a
b
ilit
ies
b
u
t
lack
s
s
u
f
f
icien
t
ex
p
lo
r
atio
n
ca
p
ab
ilit
ies
.
T
h
is
lim
itatio
n
o
f
ten
ca
u
s
es
th
e
alg
o
r
ith
m
to
f
all
in
to
lo
ca
l
o
p
tim
a
an
d
r
ed
u
ce
s
th
e
u
p
d
ate
s
p
ee
d
.
T
o
o
v
er
c
o
m
e
th
is
is
s
u
e,
th
e
p
o
s
itio
n
u
p
d
atin
g
f
o
r
m
u
la
in
c
o
r
p
o
r
ates
g
lo
b
al
g
u
id
in
g
f
ac
to
r
s
,
an
d
t
h
e
u
p
d
ated
m
ath
e
m
atica
l
ex
p
r
ess
io
n
f
o
r
p
o
s
itio
n
u
p
d
ate
is
g
iv
en
in
(
1
3
)
.
,
=
,
+
,
(
,
−
,
)
+
(
,
−
,
)
(
1
3
)
W
h
er
e
,
r
ep
r
esen
ts
th
e
n
ew
p
o
s
itio
n
p
r
o
d
u
ce
d
clo
s
er
to
,
,
,
r
ep
r
esen
t
r
an
d
o
m
n
u
m
b
er
s
p
r
o
d
u
ce
d
b
y
r
an
d
o
m
f
o
r
m
u
las
,
an
d
,
r
ep
r
esen
ts
th
e
b
est
v
alu
e.
T
h
e
g
lo
b
al
g
u
id
i
n
g
f
ac
to
r
(
,
−
,
)
is
in
clu
d
ed
to
en
ab
le
th
e
b
ee
s
ea
r
ch
to
h
av
e
a
h
ig
h
er
d
ir
ec
tio
n
,
an
d
th
e
in
f
lu
en
c
in
g
f
ac
to
r
i
s
in
clu
d
ed
as
th
e
g
lo
b
al
f
ac
to
r
f
o
r
co
n
s
tr
ain
in
g
th
e
s
ea
r
ch
am
p
litu
d
e.
At
th
is
s
tag
e,
th
e
v
alu
e
o
f
a
Q
-
ta
b
le
i
s
u
p
d
ated
b
y
u
s
in
g
(
8
)
.
I
f
a
f
o
o
d
s
o
u
r
ce
p
r
o
v
id
es
g
o
o
d
f
it
n
ess
v
alu
e,
em
p
lo
y
ee
b
ee
d
o
es
n
o
t
r
e
p
lace
cu
r
r
en
t
f
o
o
d
s
o
u
r
ce
with
a
n
ew
s
o
u
r
ce
.
I
n
s
tead
,
it
u
p
d
ates
Q
-
v
alu
e
th
r
o
u
g
h
r
ewa
r
d
in
g
n
ewly
ch
o
s
en
f
o
o
d
s
o
u
r
ce
a
n
d
p
en
alizin
g
c
u
r
r
e
n
t
f
o
o
d
s
o
u
r
ce
.
Similar
ly
,
if
a
n
e
w
f
o
o
d
s
o
u
r
ce
d
o
es n
o
t p
r
o
v
id
e
g
o
o
d
f
itn
ess
v
alu
e,
it is
p
en
alize
d
,
wh
ile
cu
r
r
en
t
f
o
o
d
s
o
u
r
ce
is
r
ewa
r
d
ed
.
T
h
e
Q
-
tab
le
is
u
p
d
ated
ev
er
y
tim
e
th
e
em
p
l
o
y
ee
b
ee
id
en
tifie
s
a
n
ap
p
r
o
p
r
iate
f
o
o
d
s
o
u
r
ce
.
Hen
ce
,
n
u
m
b
er
o
f
em
p
lo
y
ee
b
ee
s
is
d
en
o
ted
as
,
an
d
Q
-
tab
le
is
u
p
d
ate
d
o
n
ce
f
o
r
ea
ch
em
p
lo
y
e
e
b
ee
.
I
n
an
o
n
l
o
o
k
e
r
s
tag
e,
th
e
o
n
lo
o
k
er
b
ee
s
elec
ts
a
f
o
o
d
s
o
u
r
ce
f
r
o
m
Q
-
tab
le
b
ased
o
n
Q
-
v
al
u
es
.
T
h
e
o
n
lo
o
k
er
b
ee
s
ea
r
ch
e
s
f
o
r
a
n
ew
f
o
o
d
s
o
u
r
ce
,
a
n
d
m
ath
em
a
tical
ex
p
r
ess
io
n
f
o
r
th
is
is
g
iv
en
in
(
1
4
)
.
I
f
a
n
ew
f
o
o
d
s
o
u
r
ce
h
as
h
i
g
h
f
itn
ess
v
alu
e,
it r
ep
lace
s
p
r
esen
t
f
o
o
d
s
o
u
r
ce
,
a
n
d
Q
-
v
al
u
e
is
u
p
d
ate
d
ac
co
r
d
in
g
l
y
.
,
=
,
+
∅
,
∙
∙
(
,
−
,
)
(
1
4
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
15
,
No
.
5
,
Octo
b
er
2
0
2
6
:
4
6
4
5
-
4
6
5
5
4650
W
h
er
e
,
r
ep
r
esen
ts
th
e
p
ar
am
e
ter
f
o
r
o
p
tim
izatio
n
in
th
e
n
ei
g
h
b
o
r
in
g
f
o
o
d
s
o
u
r
ce
,
wh
er
e
∅
∈
[
−
1
,
1
]
,
an
d
,
r
ep
r
esen
t
p
ar
am
eter
f
o
r
o
p
tim
izatio
n
in
t
h
e
o
p
tim
u
m
f
o
o
d
s
o
u
r
ce
,
wh
ich
is
s
elec
ted
th
r
o
u
g
h
r
a
n
d
o
m
s
elec
tio
n
.
Usi
n
g
th
e
(
1
4
)
im
p
r
o
v
es
ex
p
lo
itatio
n
,
r
ath
e
r
th
an
j
u
s
t
u
p
d
atin
g
th
e
d
im
en
s
io
n
v
a
lu
es.
T
h
e
o
n
lo
o
k
e
r
b
ee
s
ex
p
lo
it
th
e
cu
r
r
en
t
f
o
o
d
s
o
u
r
ce
an
d
u
p
d
ate
all
d
im
en
s
io
n
s
with
v
ar
y
in
g
weig
h
t
v
alu
es
,
an
d
th
e
Q
-
v
alu
es
ar
e
s
u
b
s
eq
u
en
tly
u
p
d
ated
.
I
n
t
h
e
s
co
u
t
b
ee
s
tag
e,
if
th
e
r
e
h
a
v
e
b
ee
n
s
ev
er
al
s
u
cc
ess
f
u
l
att
em
p
ts
at
id
en
tify
in
g
th
e
b
est
n
eig
h
b
o
r
s
,
th
e
c
u
r
r
e
n
t
f
o
o
d
s
o
u
r
ce
is
d
is
ca
r
d
ed
,
an
d
th
e
s
co
u
t
b
ee
r
an
d
o
m
ly
s
ea
r
ch
es
f
o
r
n
ew
f
o
o
d
s
o
u
r
ce
.
T
h
e
al
g
o
r
ith
m
f
o
r
AB
C
with
a
g
lo
b
al
g
u
id
i
n
g
f
ac
t
o
r
is
d
escr
ib
ed
as in
Alg
o
r
ith
m
2
.
Alg
o
r
ith
m
2
.
AB
C
with
a
g
lo
b
al
g
u
id
in
g
f
ac
to
r
Initialization
Initialize the population and measure the fitness value
Set parameters such as best solution, highest number of iterations, population size
Identify good solution
Employee Bees stage
for every position do
Update employee bee position through (13)
Evaluate new position
if fitness value is good replace the present location with a new
location
end if
Measure the possibility and update the Q
-
table to choose a location in
onlooker bee stage
end for
Onlooker Bees stage
for every onlooker do
choose the location depended on Q
-
value and probability
Update the location of the onlooker bee through (14)
Evaluate new location
if fitness value is good replace the present location with a new
location
end if
Update Q
-
table by (8)
end for
Identify a better solution (Q
-
table)
Scout Bees stage
for every position do
Abandon solutions which are not updated and produce new solutions randomly
Update Q
-
table by (8)
end for
end while
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
a
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
is
s
im
u
lated
in
a
C
lo
u
d
Sim
en
v
ir
o
n
m
en
t
with
f
o
llo
win
g
co
n
f
ig
u
r
atio
n
s
:
an
i5
p
r
o
ce
s
s
o
r
,
W
in
d
o
ws
1
0
OS,
an
d
8
GB
o
f
R
A
M.
T
h
e
s
im
u
latio
n
in
v
o
lv
es
two
d
if
f
er
en
t
ex
p
er
im
en
ts
to
ass
es
s
th
e
alg
o
r
ith
m
'
s
ef
f
ec
tiv
en
ess
.
T
h
e
in
itial
ex
p
er
im
en
t
u
s
es
two
s
ets
o
f
task
s
to
ev
alu
ate
r
ewa
r
d
p
er
ce
n
tag
e
a
n
d
c
o
m
p
ar
es
en
er
g
y
c
o
n
s
u
m
p
tio
n
a
n
d
co
m
p
u
tatio
n
tim
e
o
f
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
with
ex
is
tin
g
ap
p
r
o
ac
h
es.
T
h
e
s
ec
o
n
d
e
x
p
er
im
en
t
f
o
c
u
s
es
o
n
m
ea
s
u
r
in
g
t
h
e
ac
cu
r
ac
y
o
f
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
.
Per
f
o
r
m
an
ce
m
etr
ics
s
u
ch
as
m
ak
esp
an
,
lo
ad
b
alan
cin
g
in
d
ex
,
th
r
o
u
g
h
p
u
t,
e
n
er
g
y
co
n
s
u
m
p
tio
n
,
co
m
p
letio
n
ti
m
e,
an
d
r
ewa
r
d
ar
e
co
n
s
id
er
ed
to
ev
alu
ate
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
'
s
p
er
f
o
r
m
an
ce
.
Key
p
ar
am
eter
s
o
f
th
e
p
r
o
p
o
s
ed
DR
L
-
AB
C
Q
alg
o
r
ith
m
in
cl
u
d
e
a
lear
n
in
g
r
ate
o
f
0
.
0
1
,
wh
ich
co
n
tr
o
ls
th
e
m
eth
o
d
’
s
ad
ap
tatio
n
s
p
ee
d
an
d
b
ala
n
ce
s
co
n
v
er
g
en
ce
with
s
tab
ilit
y
.
T
h
e
ex
p
lo
r
atio
n
-
e
x
p
lo
itatio
n
tr
ad
eo
f
f
e
n
s
u
r
es
ad
ap
ta
b
ilit
y
wh
ile
p
r
ev
en
tin
g
s
u
b
-
o
p
tim
al
d
ec
is
io
n
s
.
T
h
e
b
atch
s
ize
o
f
6
4
im
p
ac
ts
th
e
tr
ain
in
g
s
p
ee
d
a
n
d
s
tab
ilit
y
,
wh
ile
th
e
d
is
co
u
n
t
f
ac
to
r
(
γ
=
0
.
9
)
s
tr
ik
es
a
b
ala
n
ce
b
etw
ee
n
s
h
o
r
t
-
ter
m
a
n
d
lo
n
g
-
ter
m
r
ewa
r
d
s
.
Ad
d
itio
n
ally
,
p
ar
am
eter
s
o
f
th
e
AB
C
alg
o
r
ith
m
ar
e
s
et
to
5
0
b
ee
s
an
d
a
m
ax
im
u
m
iter
atio
n
o
f
1
0
0
,
wh
ich
en
h
a
n
ce
s
o
p
tim
izatio
n
with
o
u
t
e
x
ce
s
s
iv
e
co
m
p
u
tatio
n
al
o
v
e
r
h
ea
d
.
Fin
e
-
tu
n
i
n
g
t
h
ese
p
ar
am
eter
s
en
s
u
r
es
a
b
ala
n
c
e
b
etwe
en
en
er
g
y
ef
f
icien
cy
an
d
o
v
er
all
p
er
f
o
r
m
a
n
ce
.
T
ab
le
1
p
r
esen
ts
a
s
im
u
latio
n
p
ar
am
eter
s
o
f
a
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
.
T
ab
le
2
p
r
esen
ts
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
ev
alu
ate
d
o
n
d
if
f
er
e
n
t
t
ask
s
2
,
000
an
d
4
,
0
0
0
.
T
h
e
p
ar
ticle
s
war
m
o
p
tim
izatio
n
(
PS
O)
,
g
r
e
y
wo
lf
o
p
tim
izatio
n
(
GW
O)
,
wh
ale
o
p
tim
izatio
n
alg
o
r
ith
m
(
W
OA)
,
an
d
d
war
f
m
o
n
g
o
o
s
e
o
p
tim
izatio
n
(
DM
O)
ar
e
th
e
ex
is
tin
g
alg
o
r
ith
m
s
u
tili
ze
d
to
v
alid
ate
p
er
f
o
r
m
an
ce
o
f
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
.
I
n
T
a
b
le
2
,
th
e
m
et
r
ics
co
n
s
id
er
ed
in
clu
d
e
m
ak
e
s
p
an
,
in
d
ex
o
f
lo
ad
b
alan
cin
g
,
th
r
o
u
g
h
p
u
t,
en
e
r
g
y
co
n
s
u
m
p
tio
n
,
co
m
p
letio
n
tim
e
,
an
d
r
ewa
r
d
.
T
h
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
o
b
tai
n
s
1
2
5
.
2
5
s
f
o
r
m
a
k
esp
an
,
0
.
4
2
f
o
r
th
e
in
d
e
x
o
f
lo
a
d
b
alan
ci
n
g
,
2
.
3
9
jo
b
s
/s
f
o
r
th
r
o
u
g
h
p
u
t,
3
6
7
,
8
2
5
.
9
W
f
o
r
en
er
g
y
c
o
n
s
u
m
p
tio
n
,
1
5
.
3
2
s
f
o
r
co
m
p
letio
n
tim
e,
an
d
4
2
.
7
5
%
f
o
r
r
ewa
r
d
with
2
0
0
0
task
s
.
Fo
r
4
0
0
0
task
s
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Op
timiz
in
g
clo
u
d
en
erg
y
w
ith
d
ee
p
r
ein
fo
r
ce
men
t le
a
r
n
in
g
:
a
n
a
r
tifi
cia
l b
ee
…
(
Mo
h
a
mme
d
Zia
u
r
R
a
h
ma
n
)
4651
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
o
b
tain
s
1
5
0
.
6
2
s
f
o
r
m
a
k
esp
an
,
0
.
4
3
f
o
r
th
e
i
n
d
ex
o
f
lo
ad
b
al
an
cin
g
,
2
.
8
7
jo
b
s
/s
f
o
r
th
r
o
u
g
h
p
u
t,
4
4
1
,
3
9
1
.
0
8
w
f
o
r
en
e
r
g
y
c
o
n
s
u
m
p
ti
o
n
,
1
8
.
3
8
s
f
o
r
c
o
m
p
letio
n
tim
e,
an
d
4
2
.
7
5
% f
o
r
r
ewa
r
d
.
T
ab
le
1
.
Simu
latio
n
p
ar
am
eter
s
o
f
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
P
a
r
a
me
t
e
r
s
V
a
l
u
e
S
c
h
e
d
u
l
i
n
g
i
n
t
e
r
v
a
l
0
.
0
1
-
1
se
c
o
n
d
s
W
o
r
k
e
r
n
o
d
e
sp
e
e
d
2
0
0
-
4
0
0
i
n
st
r
u
c
t
i
o
n
s/
s
e
c
o
n
d
W
o
r
k
e
r
n
o
d
e
mem
o
r
y
si
z
e
2
-
8
G
B
W
o
r
k
e
r
n
o
d
e
sw
a
p
p
i
n
g
c
o
s
t
2
-
1
0
i
n
st
r
u
c
t
i
o
n
s/
G
B
W
o
r
k
e
r
q
u
a
n
t
u
m
0
.
0
1
-
0
.
5
se
c
o
n
d
s
W
o
r
k
e
r
n
o
d
e
st
a
r
t
u
p
t
i
m
e
1
2
0
-
6
0
0
se
c
o
n
d
s
W
o
r
k
e
r
n
o
d
e
sc
h
e
d
u
l
e
r
n
o
t
i
f
i
c
a
t
i
o
n
t
i
me
1
-
5
i
n
st
r
u
c
t
i
o
n
s
S
i
mu
l
a
t
i
o
n
t
i
me
10
,
000
Ta
sk
r
a
n
g
e
1
,
0
0
0
–
5
,
0
0
0
Le
n
g
t
h
1
,
0
0
0
–
2
,
0
0
0
F
i
l
e
si
z
e
3
0
0
-
4
0
0
T
ab
le
2
.
Per
f
o
r
m
an
ce
o
f
d
if
f
er
en
t
o
p
tim
izatio
n
s
with
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
f
o
r
d
if
f
e
r
e
n
t ta
s
k
s
Ta
sk
s
M
e
t
h
o
d
s
M
a
k
e
s
p
a
n
(
s)
I
n
d
e
x
o
f
l
o
a
d
b
a
l
a
n
c
i
n
g
Th
r
o
u
g
h
p
u
t
(
jo
b
s/
s)
En
e
r
g
y
c
o
n
su
m
p
t
i
o
n
(
w
a
t
t
)
C
o
m
p
l
e
t
i
o
n
t
i
m
e
(
s)
R
e
w
a
r
d
(
%)
2
,
0
0
0
PSO
1
4
7
.
8
0
0
.
8
5
1
.
3
2
3
8
4
2
1
7
.
8
3
2
.
2
3
3
8
.
5
0
G
W
O
1
3
3
.
2
4
0
.
7
7
1
.
6
9
4
1
2
3
0
9
.
6
7
3
8
.
1
7
3
1
.
6
5
W
O
A
1
5
9
.
3
7
0
.
9
2
1
.
8
7
4
9
5
6
2
3
.
7
4
2
9
.
5
4
2
9
.
6
0
D
M
O
1
7
8
.
1
7
0
.
7
6
1
.
9
1
5
1
7
8
9
0
.
4
3
3
3
.
5
4
3
9
.
5
3
D
R
L
–
A
B
C
Q
A
1
2
5
.
5
2
0
.
4
2
2
.
3
9
3
6
7
8
2
5
.
9
1
5
.
3
2
4
2
.
7
5
4
,
0
0
0
PSO
1
7
7
.
3
6
0
.
8
6
1
.
5
8
4
6
1
0
6
1
.
3
8
3
8
.
6
8
3
8
.
2
0
G
W
O
1
6
0
.
0
0
0
.
7
8
2
.
0
3
4
9
4
7
7
1
.
6
4
5
.
8
3
2
.
0
0
W
O
A
1
9
1
.
2
4
0
.
9
3
2
.
2
4
5
9
4
7
4
8
.
4
9
3
5
.
4
5
3
0
.
0
0
D
M
O
2
1
3
.
8
0
.
7
7
2
.
2
9
6
2
1
4
6
8
.
5
2
4
0
.
2
4
3
9
.
6
0
D
R
L
-
A
B
C
Q
A
1
5
0
.
6
2
0
.
4
3
2
.
8
7
4
4
1
3
9
1
.
0
8
1
8
.
3
8
4
2
.
7
5
T
ab
le
3
p
r
esen
ts
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
ev
alu
ated
with
d
if
f
e
r
en
t
n
u
m
b
er
s
o
f
VM
s
,
s
p
ec
if
ically
1
0
an
d
5
0
.
T
h
e
ex
is
tin
g
alg
o
r
ith
m
s
u
tili
ze
d
to
v
alid
ate
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
in
cl
u
d
e
PS
O,
GW
O,
W
OA,
an
d
DM
O.
I
n
T
a
b
le
3
,
th
e
m
etr
ics
co
n
s
id
er
ed
ar
e
m
ak
esp
an
,
in
d
ex
o
f
lo
ad
b
alan
cin
g
,
th
r
o
u
g
h
p
u
t,
e
n
er
g
y
co
n
s
u
m
p
tio
n
,
co
m
p
leti
o
n
tim
e,
an
d
r
ewa
r
d
.
T
h
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
ac
h
iev
es
1
4
0
s
f
o
r
m
ak
esp
a
n
,
0
.
4
5
f
o
r
th
e
in
d
ex
o
f
lo
a
d
b
ala
n
cin
g
,
2
.
5
0
jo
b
s
/s
f
o
r
th
r
o
u
g
h
p
u
t
,
3
8
0
,
0
0
0
W
f
o
r
en
er
g
y
c
o
n
s
u
m
p
tio
n
,
2
0
.
0
0
s
f
o
r
co
m
p
letio
n
tim
e
,
an
d
4
5
.
0
0
%
f
o
r
r
ewa
r
d
with
1
0
VM
s
.
Fo
r
5
0
VM
s
,
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
ac
h
iev
es
1
1
5
.
0
0
s
f
o
r
m
ak
esp
an
,
0
.
4
0
f
o
r
t
h
e
in
d
ex
o
f
lo
a
d
b
alan
c
in
g
,
3
.
0
0
jo
b
s
/s
f
o
r
th
r
o
u
g
h
p
u
t
,
3
6
0
,
0
0
0
W
f
o
r
e
n
er
g
y
co
n
s
u
m
p
tio
n
,
2
2
.
0
0
s
f
o
r
co
m
p
letio
n
tim
e
,
a
n
d
4
7
.
0
0
% f
o
r
r
ewa
r
d
.
T
ab
le
3
.
Per
f
o
r
m
an
ce
o
f
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
with
d
if
f
e
r
e
n
t V
Ms
V
M
s
M
e
t
h
o
d
s
M
a
k
e
s
p
a
n
(
s)
I
n
d
e
x
o
f
l
o
a
d
b
a
l
a
n
c
i
n
g
Th
r
o
u
g
h
p
u
t
(
j
o
b
s/
s)
En
e
r
g
y
c
o
n
su
m
p
t
i
o
n
(
w
a
t
t
)
C
o
m
p
l
e
t
i
o
n
t
i
m
e
(
s)
R
e
w
a
r
d
(
%)
10
PSO
1
6
0
.
0
0
0
.
8
8
1
.
2
5
4
2
0
.
0
0
0
3
5
.
0
0
3
5
.
0
0
G
W
O
1
5
0
.
0
0
0
.
8
0
1
.
5
5
4
3
0
.
0
0
0
4
0
.
0
0
3
2
.
0
0
W
O
A
1
7
0
.
0
0
0
.
9
5
1
.
8
5
4
8
0
.
0
0
0
3
0
.
0
0
3
0
.
0
0
D
M
O
1
8
5
.
0
0
0
.
7
8
1
.
9
5
5
0
0
.
0
0
0
3
3
.
0
0
3
8
.
0
0
D
R
L
–
A
B
C
Q
A
1
4
0
.
0
0
0
.
4
5
2
.
5
0
3
8
0
.
0
0
0
2
0
.
0
0
4
5
.
0
0
50
PSO
1
3
0
.
0
0
0
.
8
0
1
.
4
0
3
7
0
.
0
0
0
2
8
.
0
0
4
0
.
0
0
G
W
O
1
2
0
.
0
0
0
.
7
5
1
.
7
5
3
8
0
.
0
0
0
3
2
.
0
0
3
5
.
0
0
W
O
A
1
3
5
.
0
0
0
.
8
5
2
.
0
0
4
0
0
.
0
0
0
2
7
.
0
0
3
3
.
0
0
D
M
O
1
4
5
.
0
0
0
.
7
4
2
.
0
5
4
1
0
.
0
0
0
3
0
.
0
0
4
0
.
0
0
D
R
L
–
A
B
C
Q
A
1
1
5
.
0
0
0
.
4
0
3
.
0
0
3
6
0
.
0
0
0
2
2
.
0
0
4
7
.
0
0
T
ab
le
4
p
r
esen
ts
th
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
ev
alu
ated
with
d
if
f
e
r
en
t
t
ask
s
izes,
n
am
ely
2
0
0
0
an
d
4
0
0
0
,
u
s
in
g
ex
is
tin
g
lo
ad
b
alan
cin
g
alg
o
r
ith
m
s
.
T
h
e
ex
is
tin
g
alg
o
r
ith
m
s
f
o
r
co
m
p
ar
is
o
n
in
clu
d
e
R
R
,
least
co
n
n
ec
tio
n
s
(
L
C
)
,
g
lo
b
al
s
er
v
e
r
lo
a
d
b
alan
cin
g
(
GSLB),
an
d
r
an
d
o
m
f
o
r
est
(
R
F).
I
n
T
ab
le
4
,
th
e
f
o
llo
win
g
m
etr
ics
ar
e
p
r
esen
ted
:
m
ak
esp
an
,
in
d
ex
o
f
lo
ad
b
alan
cin
g
,
t
h
r
o
u
g
h
p
u
t,
en
er
g
y
co
n
s
u
m
p
tio
n
,
co
m
p
letio
n
tim
e,
an
d
r
ewa
r
d
.
T
h
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
ac
h
iev
es
1
2
5
.
2
5
s
f
o
r
m
ak
esp
an
,
0
.
4
2
f
o
r
t
h
e
in
d
e
x
o
f
l
o
ad
b
alan
cin
g
,
2
.
3
9
(
jo
b
s
/s
)
f
o
r
th
r
o
u
g
h
p
u
t,
3
6
7
,
8
2
5
.
9
W
f
o
r
e
n
er
g
y
c
o
n
s
u
m
p
tio
n
,
1
5
.
3
2
s
f
o
r
c
o
m
p
letio
n
tim
e,
an
d
4
2
.
7
5
%
f
o
r
r
ewa
r
d
wit
h
2
0
0
0
task
s
.
Fo
r
4
0
0
0
task
s
,
th
e
DR
L
-
A
B
C
Q
ap
p
r
o
ac
h
ac
h
iev
es
1
5
0
.
6
2
s
f
o
r
m
ak
esp
a
n
,
0
.
4
3
f
o
r
th
e
in
d
e
x
o
f
lo
a
d
b
alan
cin
g
,
2
.
8
7
(
jo
b
s
/s
)
f
o
r
t
h
r
o
u
g
h
p
u
t,
4
4
1
,
3
9
1
.
0
8
W
f
o
r
en
er
g
y
co
n
s
u
m
p
tio
n
,
1
8
.
3
8
s
f
o
r
co
m
p
letio
n
tim
e,
an
d
4
2
.
7
5
% f
o
r
r
ewa
r
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
15
,
No
.
5
,
Octo
b
er
2
0
2
6
:
4
6
4
5
-
4
6
5
5
4652
T
ab
le
4
.
Per
f
o
r
m
an
ce
o
f
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
with
d
if
f
e
r
e
n
t ta
s
k
s
Ta
sk
s
M
e
t
h
o
d
s
M
a
k
e
s
p
a
n
(
s)
I
n
d
e
x
o
f
l
o
a
d
b
a
l
a
n
c
i
n
g
Th
r
o
u
g
h
p
u
t
(
j
o
b
s/
s)
En
e
r
g
y
c
o
n
su
m
p
t
i
o
n
(
w
a
t
t
)
C
o
m
p
l
e
t
i
o
n
t
i
m
e
(
s)
R
e
w
a
r
d
(
%)
2
,
0
0
0
RR
1
8
2
.
3
7
0
.
6
8
1
.
9
2
1
9
2
3
8
5
.
7
2
2
.
9
3
2
9
.
8
5
LC
1
7
6
.
8
4
0
.
6
2
.
0
7
1
8
1
4
7
2
.
3
2
1
.
6
5
3
1
.
2
0
G
S
LB
1
7
1
.
9
2
0
.
5
7
2
.
1
8
1
7
6
9
2
5
.
6
2
0
.
7
5
3
2
.
5
0
RF
1
8
6
.
4
5
0
.
7
1
1
.
8
8
1
9
6
7
3
4
.
9
2
3
.
2
7
2
8
.
7
5
D
R
L
–
A
B
C
Q
A
1
2
5
.
5
2
0
.
4
2
2
.
3
9
3
6
7
8
2
5
.
9
1
5
.
3
2
4
2
.
7
5
4
,
0
0
0
RR
2
0
3
.
4
8
0
.
6
9
2
.
3
3
3
7
2
5
4
8
.
9
2
7
.
6
3
2
8
.
4
0
LC
1
9
8
.
7
6
0
.
6
2
2
.
4
7
3
6
2
3
9
5
.
4
2
6
.
5
7
3
0
.
1
0
G
S
LB
1
9
1
.
5
9
0
.
5
9
2
.
6
2
3
5
6
7
8
5
.
1
2
5
.
4
2
3
1
.
5
0
RF
2
0
7
.
3
2
0
.
7
3
2
.
2
4
3
8
2
7
4
6
.
5
2
8
.
8
9
2
7
.
3
0
D
R
L
-
A
B
C
Q
A
1
5
0
.
6
2
0
.
4
3
2
.
8
7
4
4
1
3
9
1
.
0
8
1
8
.
3
8
4
2
.
7
5
Fig
u
r
es
3
a
n
d
4
illu
s
tr
ate
th
e
r
ewa
r
d
o
b
tain
e
d
f
r
o
m
t
h
e
en
v
i
r
o
n
m
e
n
t
f
o
r
r
ed
u
cin
g
en
er
g
y
co
n
s
u
m
p
tio
n
an
d
c
o
m
p
letio
n
tim
e.
T
h
ese
f
ig
u
r
es
co
m
p
ar
e
th
e
r
ewa
r
d
v
al
u
es
o
f
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
with
ex
is
tin
g
alg
o
r
ith
m
s
s
u
ch
as
PS
O,
GW
O,
W
OA,
an
d
DM
O
f
o
r
task
s
ets
o
f
2
0
0
0
a
n
d
4
0
0
0
.
Fig
u
r
e
3
s
h
o
ws
th
e
to
tal
r
ewa
r
d
p
er
ite
r
atio
n
,
f
r
o
m
1
0
to
1
0
0
iter
atio
n
s
,
f
o
r
th
e
2
0
0
0
-
task
s
et.
Fig
u
r
e
4
s
h
o
ws
th
e
to
tal
r
ewa
r
d
p
er
iter
atio
n
,
f
r
o
m
1
0
t
o
1
0
0
iter
atio
n
s
,
f
o
r
th
e
4
0
0
0
-
t
ask
s
et.
Fro
m
b
o
th
f
ig
u
r
es,
th
e
p
r
o
p
o
s
ed
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
o
u
t
p
er
f
o
r
m
s
th
e
ex
is
tin
g
alg
o
r
ith
m
s
.
As
th
e
n
u
m
b
er
o
f
task
s
in
cr
e
ases
f
r
o
m
2
0
0
0
to
4
0
0
0
,
r
ewa
r
d
p
e
r
ce
n
tag
e
also
s
u
r
p
ass
es
th
e
ex
is
tin
g
alg
o
r
ith
m
s
.
Fig
u
r
e
5
r
e
p
r
esen
ts
ac
cu
r
ac
y
o
f
a
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
,
cl
ea
r
ly
s
h
o
win
g
th
at
th
e
ac
cu
r
ac
y
in
cr
ea
s
es
s
tead
ily
as
th
e
iter
atio
n
co
u
n
t
p
r
o
g
r
e
s
s
es
f
r
o
m
1
0
to
1
0
0
.
T
h
e
f
i
n
al
o
u
tco
m
e
o
f
tr
ain
i
n
g
th
e
n
eu
r
al
n
etwo
r
k
is
im
p
r
o
v
e
d
ac
cu
r
ac
y
at
th
e
e
n
d
o
f
ea
ch
iter
atio
n
.
T
h
is
r
esu
lt
r
ef
lects
en
h
an
ce
d
ac
cu
r
ac
y
in
d
y
n
am
ic
en
v
ir
o
n
m
en
ts
.
Fig
u
r
e
6
s
h
o
ws
th
e
p
er
f
o
r
m
an
ce
o
f
d
if
f
er
e
n
t
alg
o
r
ith
m
s
with
2
0
0
0
to
4
0
0
0
task
s
allo
ca
ted
to
ex
is
tin
g
VM
s
.
Fro
m
Fig
u
r
e
6
,
it
is
ev
id
en
t
th
at
th
e
D
R
L
-
AB
C
Q
ap
p
r
o
ac
h
co
n
s
u
m
es
le
s
s
co
m
p
letio
n
tim
e
th
an
th
e
ex
is
tin
g
alg
o
r
ith
m
s
,
s
u
ch
as
PS
O,
GW
O,
W
OA,
an
d
DM
O.
Fig
u
r
e
7
r
ep
r
esen
ts
th
e
to
tal
en
er
g
y
co
n
s
u
m
p
tio
n
ac
r
o
s
s
v
ar
io
u
s
V
Ms
in
th
e
d
ata
ce
n
ter
.
Fr
o
m
F
ig
u
r
e
7
,
it
is
clea
r
th
at
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
s
av
es
m
o
r
e
en
er
g
y
th
an
th
e
e
x
is
tin
g
alg
o
r
ith
m
s
,
in
clu
d
in
g
PS
O,
G
W
O,
W
O
A,
an
d
DM
O.
T
h
e
n
u
m
b
er
o
f
VM
s
r
an
g
es f
r
o
m
1
0
to
5
0
to
ev
alu
ate
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
.
Fig
u
r
e
3
.
R
ewa
r
d
p
er
ce
n
tag
e
f
o
r
2
,
0
0
0
task
s
Fig
u
r
e
4
.
R
ewa
r
d
p
er
ce
n
tag
e
f
o
r
4
,
0
0
0
task
s
Fig
u
r
e
5
.
I
te
r
atio
n
v
s
ac
c
u
r
ac
y
Fig
u
r
e
6
.
Nu
m
b
er
o
f
task
s
v
s
co
m
p
letio
n
tim
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Op
timiz
in
g
clo
u
d
en
erg
y
w
ith
d
ee
p
r
ein
fo
r
ce
men
t le
a
r
n
in
g
:
a
n
a
r
tifi
cia
l b
ee
…
(
Mo
h
a
mme
d
Zia
u
r
R
a
h
ma
n
)
4653
Fig
u
r
e
7
.
Nu
m
b
er
o
f
VM
s
v
s
en
er
g
y
c
o
n
s
u
m
p
ti
o
n
3
.
1
.
Co
m
pa
ra
t
iv
e
a
na
ly
s
is
T
h
e
p
er
f
o
r
m
an
ce
o
f
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
is
co
m
p
ar
e
d
with
ex
is
tin
g
tech
n
iq
u
es
lik
e
QM
PS
O,
L
B
MPSO
,
an
d
MM
HHO,
u
s
i
n
g
m
etr
ics
lik
e
m
a
k
esp
an
,
e
n
e
r
g
y
u
tili
za
tio
n
,
an
d
th
r
o
u
g
h
p
u
t.
T
h
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
ac
h
iev
es a
m
a
k
esp
a
n
o
f
1
2
5
.
5
2
s
,
en
e
r
g
y
u
tili
za
tio
n
o
f
9
9
.
8
1
k
J
,
an
d
th
r
o
u
g
h
p
u
t
o
f
2
.
3
9
jo
b
s
/s
wh
en
co
m
p
ar
ed
to
th
e
e
x
is
tin
g
ap
p
r
o
ac
h
es.
T
ab
le
5
p
r
esen
ts
th
e
c
o
m
p
ar
ativ
e
r
esu
lts
o
f
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
.
T
ab
le
5
.
C
o
m
p
a
r
ativ
e
tab
le
o
f
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
M
e
t
h
o
d
s
M
a
k
e
s
p
a
n
(
s)
En
e
r
g
y
u
t
i
l
i
z
a
t
i
o
n
(
k
J)
Th
r
o
u
g
h
p
u
t
(
jo
b
s
/
s)
Q
M
P
S
O
[
1
6
]
NA
1
7
3
.
8
7
5
.
4
6
LB
M
P
S
O
[
1
7
]
1
0
8
.
8
7
NA
NA
M
M
H
H
O
[
1
8
]
NA
5
5
.
0
1
4
.
5
2
P
r
o
p
o
se
d
D
R
L
-
A
B
C
Q
a
p
p
r
o
a
c
h
1
2
5
.
5
2
9
9
.
8
1
2
.
3
9
3
.
2
.
Dis
cus
s
io
n
T
h
e
s
im
u
latio
n
o
f
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
is
p
er
f
o
r
m
e
d
in
a
Py
th
o
n
en
v
ir
o
n
m
en
t,
an
d
th
r
ee
d
if
f
er
en
t
ex
p
e
r
im
en
ts
ar
e
co
n
d
u
cted
t
o
v
alid
ate
th
e
r
ewa
r
d
p
er
ce
n
tag
e
an
d
ac
cu
r
ac
y
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
is
co
m
p
ar
ed
with
v
ar
io
u
s
o
p
tim
izatio
n
alg
o
r
ith
m
s
lik
e
PS
O,
GW
O,
W
OA,
an
d
DM
O.
Ad
d
itio
n
ally
,
DR
L
-
AB
C
Q
a
p
p
r
o
ac
h
is
co
m
p
ar
e
d
with
e
x
is
tin
g
alg
o
r
ith
m
s
lik
e
QM
PS
O,
L
B
MP
SO,
an
d
MM
HHO.
T
o
ad
d
r
ess
th
e
is
s
u
es
o
f
c
o
m
p
lex
ity
an
d
h
ig
h
d
im
e
n
s
io
n
ality
,
t
h
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
is
em
p
lo
y
ed
,
wh
ich
ac
ce
ler
ates
th
e
p
r
o
ce
s
s
an
d
m
in
im
izes
th
e
ex
ec
u
tio
n
tim
e.
T
h
e
a
g
en
t
in
DR
L
m
ak
es
in
tellig
en
t
d
ec
is
io
n
s
to
h
an
d
le
h
u
g
e
n
u
m
b
e
r
o
f
task
s
in
less
tim
e
an
d
at
h
ig
h
er
s
p
ee
d
s
.
T
h
e
AB
C
alg
o
r
ith
m
u
tili
ze
s
th
e
g
lo
b
al
g
u
id
in
g
f
ac
to
r
f
o
r
in
itializin
g
p
o
s
itio
n
u
p
d
ates,
wh
ich
en
h
a
n
ce
s
th
e
s
ea
r
ch
ab
ilit
y
p
er
f
o
r
m
an
ce
o
f
th
e
AB
C
alg
o
r
ith
m
.
T
h
e
p
r
o
p
o
s
ed
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
o
p
tim
izes
s
ch
ed
u
lin
g
an
d
r
eso
u
r
c
e
u
s
ag
e,
in
cr
ea
s
es
VM
th
r
o
u
g
h
p
u
t,
an
d
b
ala
n
ce
s
th
e
lo
a
d
b
et
wee
n
VM
s
.
T
h
e
ex
p
er
im
en
tal
o
u
tco
m
es
c
o
n
f
ir
m
th
at
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
d
e
liv
er
s
th
e
b
est
r
ewa
r
d
s
wh
en
l
ar
g
e
n
u
m
b
e
r
o
f
task
s
ar
e
allo
c
ated
to
d
ata
ce
n
ter
.
T
h
e
ex
p
er
im
e
n
tal
r
esu
lts
ac
q
u
ir
ed
b
y
p
r
o
p
o
s
ed
DR
L
-
AB
C
Q
m
o
d
el
ar
e
co
n
s
is
ten
t
wh
ich
em
p
lo
y
RL
an
d
h
y
b
r
id
o
p
tim
izatio
n
alg
o
r
it
h
m
s
.
Prio
r
r
esear
ch
es
h
a
v
e
s
h
o
w
n
th
at
DR
L
-
b
ased
s
ch
ed
u
ler
s
ef
f
icien
tly
ad
a
p
t
to
d
y
n
am
ic
wo
r
k
lo
ad
s
an
d
m
i
n
im
ize
m
ak
esp
an
an
d
e
x
ec
u
tio
n
tim
e
th
r
o
u
g
h
lear
n
i
n
g
o
p
tim
u
m
r
eso
u
r
c
e
a
l
l
o
c
a
ti
o
n
p
o
l
i
c
i
es
[
2
0
]
–
[
2
2
]
.
Si
m
i
l
a
r
l
y
,
h
y
b
r
i
d
m
et
a
-
h
e
u
r
i
s
t
i
c
a
l
g
o
r
i
t
h
m
s
l
i
k
e
Q
MP
S
O
[
1
6
]
a
n
d
M
M
H
HO
[
1
8
]
en
h
an
ce
d
co
n
v
er
g
e
n
ce
b
e
h
a
v
io
r
s
h
o
ws
h
ig
h
en
e
r
g
y
co
n
s
u
m
p
tio
n
u
n
d
e
r
h
ea
v
y
lo
a
d
co
n
d
itio
n
s
.
T
h
e
co
m
b
in
atio
n
o
f
a
g
lo
b
al
g
u
id
in
g
f
ac
to
r
in
AB
C
m
o
d
el
in
p
r
o
p
o
s
ed
DR
L
-
AB
C
Q
a
lg
o
r
ith
m
im
p
r
o
v
es
ex
p
lo
itatio
n
wh
e
n
a
v
o
id
in
g
p
r
em
atu
r
e
co
n
v
er
g
en
ce
,
r
esu
ltin
g
in
l
o
w
en
er
g
y
co
n
s
u
m
p
tio
n
an
d
en
h
an
ce
d
lo
ad
b
alan
cin
g
.
Mo
r
eo
v
er
,
a
n
in
c
r
ea
s
e
in
th
r
o
u
g
h
p
u
t
with
h
i
g
h
VM
co
u
n
ts
d
eter
m
in
es
th
e
s
ca
lab
ilit
y
an
d
r
o
b
u
s
tn
ess
o
f
th
e
p
r
o
p
o
s
ed
m
o
d
el
in
lar
g
e
-
s
ca
le
clo
u
d
en
v
ir
o
n
m
en
ts
.
4.
CO
NCLU
SI
O
N
T
h
is
r
esear
ch
p
r
o
p
o
s
es
an
ef
f
ec
tiv
e
ap
p
r
o
ac
h
f
o
r
lo
a
d
b
alan
cin
g
in
a
clo
u
d
en
v
i
r
o
n
m
en
t.
T
h
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
is
p
r
o
p
o
s
ed
to
b
alan
ce
lo
ad
in
cl
o
u
d
n
etwo
r
k
.
T
h
is
ap
p
r
o
ac
h
c
o
m
b
in
es
b
o
th
DR
L
alg
o
r
ith
m
an
d
AB
C
with
th
e
Q
-
lear
n
in
g
ap
p
r
o
ac
h
.
B
y
u
s
in
g
DR
L
,
th
e
n
eu
r
al
n
etwo
r
k
is
tr
ain
ed
to
ac
h
iev
e
o
p
tim
al
r
ewa
r
d
s
.
T
h
r
o
u
g
h
th
e
AB
C
Q
ap
p
r
o
ac
h
,
o
v
er
all
p
r
o
ce
s
s
in
g
tim
e
o
f
in
co
m
in
g
lo
a
d
is
m
in
im
ized
.
T
h
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
im
p
r
o
v
es
v
ar
io
u
s
l
o
ad
b
alan
cin
g
p
ar
am
eter
s
in
less
tim
e
c
o
m
p
ar
ed
to
ex
is
tin
g
alg
o
r
ith
m
s
in
th
e
clo
u
d
n
et
wo
r
k
.
T
h
e
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
ac
h
iev
es
a
m
ak
esp
an
o
f
1
2
5
.
5
2
s
,
en
er
g
y
u
tili
za
tio
n
o
f
9
9
.
8
1
kJ
,
an
d
th
r
o
u
g
h
p
u
t
o
f
2
.
3
9
jo
b
s
/s
wh
en
co
m
p
ar
ed
to
e
x
is
tin
g
ap
p
r
o
ac
h
es.
T
h
e
th
r
o
u
g
h
p
u
t
o
f
th
e
p
r
o
p
o
s
ed
DR
L
-
AB
C
Q
ap
p
r
o
ac
h
is
s
lig
h
tly
lo
wer
th
a
n
th
at
o
f
d
if
f
er
en
t
b
e
n
ch
m
ar
k
alg
o
r
ith
m
s
,
wh
ich
is
an
ess
en
tial
tr
ad
e
-
o
f
f
f
o
r
ac
h
iev
in
g
im
p
r
o
v
ed
en
er
g
y
ef
f
ic
ien
cy
.
T
h
e
in
tellig
en
t
s
ch
ed
u
lin
g
m
ec
h
an
is
m
in
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
15
,
No
.
5
,
Octo
b
er
2
0
2
6
:
4
6
4
5
-
4
6
5
5
4654
DR
L
-
AB
C
Q
m
in
im
izes
u
n
n
ec
ess
ar
y
task
m
ig
r
atio
n
s
,
an
d
o
p
tim
ized
C
PU
u
tili
za
t
io
n
lead
s
to
an
o
v
er
all
en
h
an
ce
m
e
n
t
in
en
er
g
y
ef
f
icien
cy
.
Fu
tu
r
e
w
o
r
k
f
o
cu
s
es
o
n
o
p
tim
izin
g
th
e
tr
a
d
e
-
o
f
f
b
etw
ee
n
th
r
o
u
g
h
p
u
t
an
d
en
er
g
y
co
n
s
u
m
p
tio
n
b
y
in
clu
d
in
g
an
ad
a
p
tiv
e
lear
n
in
g
m
e
ch
an
is
m
th
at
d
y
n
am
ically
ad
j
u
s
ts
th
e
s
ch
ed
u
lin
g
p
ar
am
eter
s
b
ased
o
n
r
ea
l
-
tim
e
wo
r
k
lo
ad
v
ar
iatio
n
s
.
ACK
NO
WL
E
DG
M
E
N
T
S
T
h
e
au
t
h
o
r
s
wo
u
ld
lik
e
to
t
h
an
k
th
e
Sch
o
o
l
o
f
C
o
m
p
u
t
er
Scien
ce
a
n
d
E
n
g
i
n
ee
r
in
g
,
Pre
s
id
en
cy
Un
iv
er
s
ity
,
B
en
g
alu
r
u
,
f
o
r
p
r
o
v
id
in
g
th
e
n
ec
ess
ar
y
s
u
p
p
o
r
t a
n
d
f
ac
ilit
ies to
ca
r
r
y
o
u
t th
is
r
e
s
ea
r
ch
wo
r
k
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
Au
th
o
r
s
s
tate
n
o
f
u
n
d
in
g
in
v
o
lv
ed
.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
Mo
h
am
m
ed
Z
iau
r
R
ah
m
an
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
An
an
d
ar
aj
Sh
an
t
h
i
Pich
an
d
i
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
t
er
est.
DATA AV
AI
L
AB
I
L
I
T
Y
Data
a
v
aila
b
i
lit
y
is
n
o
t
a
p
p
li
ca
b
le
t
o
t
h
is
p
ap
er
as
n
o
n
e
w
d
a
t
a
w
er
e
cr
ea
te
d
o
r
an
al
y
z
e
d
i
n
t
h
is
s
t
u
d
y
.
RE
F
E
R
E
NC
E
S
[
1
]
A
.
Ja
n
g
r
a
a
n
d
N
.
M
a
n
g
l
a
,
“
A
n
e
f
f
i
c
i
e
n
t
l
o
a
d
b
a
l
a
n
c
i
n
g
f
r
a
mew
o
r
k
f
o
r
d
e
p
l
o
y
i
n
g
r
e
s
o
u
r
c
e
sc
h
e
d
u
l
i
n
g
i
n
c
l
o
u
d
b
a
se
d
c
o
mm
u
n
i
c
a
t
i
o
n
i
n
h
e
a
l
t
h
c
a
r
e
,
”
Me
a
s
u
remen
t
:
S
e
n
so
r
s
,
v
o
l
.
2
5
,
F
e
b
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
me
a
se
n
.
2
0
2
2
.
1
0
0
5
8
4
.
[
2
]
P
.
U
d
a
y
a
sa
n
k
a
r
a
n
a
n
d
S
.
J.
J.
T
h
a
n
g
a
r
a
j
,
“
E
n
e
r
g
y
e
f
f
i
c
i
e
n
t
r
e
s
o
u
r
c
e
u
t
i
l
i
z
a
t
i
o
n
a
n
d
l
o
a
d
b
a
l
a
n
c
i
n
g
i
n
v
i
r
t
u
a
l
ma
c
h
i
n
e
s
u
s
i
n
g
p
r
e
d
i
c
t
i
o
n
a
l
g
o
r
i
t
h
ms,”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
C
o
g
n
i
t
i
v
e
C
o
m
p
u
t
i
n
g
i
n
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
4
,
p
p
.
1
2
7
–
1
3
4
,
J
u
n
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
j
c
c
e
.
2
0
2
3
.
0
2
.
0
0
5
.
[
3
]
D
.
T
e
n
n
a
k
o
o
n
,
M
.
C
h
o
w
d
h
u
r
y
,
a
n
d
T.
H
.
Lu
a
n
,
“
C
l
o
u
d
-
b
a
se
d
l
o
a
d
b
a
l
a
n
c
i
n
g
u
si
n
g
d
o
u
b
l
e
Q
-
l
e
a
r
n
i
n
g
f
o
r
i
m
p
r
o
v
e
d
q
u
a
l
i
t
y
o
f
se
r
v
i
c
e
,
”
Wi
r
e
l
e
ss
N
e
t
w
o
rks
,
v
o
l
.
2
9
,
n
o
.
3
,
p
p
.
1
0
4
3
–
1
0
5
0
,
A
p
r
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
2
7
6
-
0
1
8
-
1
8
8
8
-
8.
[
4
]
M
.
S
.
A
l
R
e
sh
a
n
e
t
a
l
.
,
“
A
f
a
st
c
o
n
v
e
r
g
i
n
g
a
n
d
g
l
o
b
a
l
l
y
o
p
t
i
m
i
z
e
d
a
p
p
r
o
a
c
h
f
o
r
l
o
a
d
b
a
l
a
n
c
i
n
g
i
n
c
l
o
u
d
c
o
mp
u
t
i
n
g
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
1
1
,
p
p
.
1
1
3
9
0
–
1
1
4
0
4
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
3
.
3
2
4
1
2
7
9
.
[
5
]
B
.
X
u
a
n
d
M
.
L
u
,
“
A
g
e
n
t
-
b
a
s
e
d
v
i
r
t
u
a
l
ma
c
h
i
n
e
mi
g
r
a
t
i
o
n
f
o
r
l
o
a
d
b
a
l
a
n
c
i
n
g
a
n
d
c
o
-
r
e
si
d
e
n
t
a
t
t
a
c
k
i
n
c
l
o
u
d
c
o
mp
u
t
i
n
g
,
”
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
3
,
n
o
.
6
,
M
a
r
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
3
0
6
3
7
0
3
.
[
6
]
S
.
S
e
f
a
t
i
,
M
.
M
o
u
s
a
v
i
n
a
sa
b
,
a
n
d
R
.
Z
.
F
a
r
k
h
a
d
y
,
“
L
o
a
d
b
a
l
a
n
c
i
n
g
i
n
c
l
o
u
d
c
o
mp
u
t
i
n
g
e
n
v
i
r
o
n
m
e
n
t
u
s
i
n
g
t
h
e
g
r
e
y
w
o
l
f
o
p
t
i
m
i
z
a
t
i
o
n
a
l
g
o
r
i
t
h
m
b
a
se
d
o
n
t
h
e
r
e
l
i
a
b
i
l
i
t
y
:
p
e
r
f
o
r
ma
n
c
e
e
v
a
l
u
a
t
i
o
n
,
”
T
h
e
J
o
u
r
n
a
l
o
f
S
u
p
e
r
c
o
m
p
u
t
i
n
g
,
v
o
l
.
7
8
,
n
o
.
1
,
p
p
.
1
8
–
4
2
,
Ja
n
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
1
2
2
7
-
0
2
1
-
0
3
8
1
0
-
8.
[
7
]
K
.
J.
R
a
j
a
sh
e
k
a
r
,
C
h
a
n
n
a
k
r
i
s
h
n
a
r
a
j
u
,
P
.
C
.
G
o
w
d
a
,
a
n
d
A
.
B
.
J
a
y
a
c
h
a
n
d
r
a
,
“
S
C
EH
O
-
I
P
S
O
:
a
n
a
t
u
r
e
-
i
n
s
p
i
r
e
d
m
e
t
a
h
e
u
r
i
s
t
i
c
o
p
t
i
m
i
z
a
t
i
o
n
f
o
r
t
a
s
k
-
sc
h
e
d
u
l
i
n
g
p
o
l
i
c
y
i
n
c
l
o
u
d
c
o
mp
u
t
i
n
g
,
”
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
3
,
n
o
.
1
9
,
S
e
p
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
3
1
9
1
0
8
5
0
.
[
8
]
W
.
A
l
mu
s
e
e
l
e
m
,
“
E
n
e
r
g
y
-
e
f
f
i
c
i
e
n
t
a
n
d
sec
u
r
i
t
y
-
a
w
a
r
e
t
a
s
k
o
f
f
l
o
a
d
i
n
g
f
o
r
m
u
l
t
i
-
t
i
e
r
e
d
g
e
-
c
l
o
u
d
c
o
m
p
u
t
i
n
g
s
y
st
e
ms,”
I
E
EE
Ac
c
e
ss
,
v
o
l
.
1
1
,
p
p
.
6
6
4
2
8
–
6
6
4
3
9
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
3
.
3
2
9
0
1
3
9
.
[
9
]
M
.
A
d
i
l
,
S
.
N
a
b
i
,
M
.
A
l
e
e
m,
V
.
G
.
D
i
a
z
,
a
n
d
J.
C
.
L
i
n
,
“
C
A
‐
M
LB
S
:
c
o
n
t
e
n
t
‐
a
w
a
r
e
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
b
a
s
e
d
l
o
a
d
b
a
l
a
n
c
i
n
g
sch
e
d
u
l
e
r
i
n
t
h
e
c
l
o
u
d
e
n
v
i
r
o
n
me
n
t
,
”
Ex
p
e
rt
S
y
st
e
m
s
,
v
o
l
.
4
0
,
n
o
.
4
,
M
a
y
2
0
2
3
,
d
o
i
:
1
0
.
1
1
1
1
/
e
x
s
y
.
1
3
1
5
0
.
[
1
0
]
H
.
M
i
k
r
a
m,
S
.
E
l
K
a
f
h
a
l
i
,
a
n
d
Y
.
S
a
a
d
i
,
“
H
EPG
A
:
a
n
e
w
e
f
f
e
c
t
i
v
e
h
y
b
r
i
d
a
l
g
o
r
i
t
h
m
f
o
r
s
c
i
e
n
t
i
f
i
c
w
o
r
k
f
l
o
w
s
c
h
e
d
u
l
i
n
g
i
n
c
l
o
u
d
c
o
m
p
u
t
i
n
g
e
n
v
i
r
o
n
me
n
t
,
”
S
i
m
u
l
a
t
i
o
n
Mo
d
e
l
l
i
n
g
Pra
c
t
i
c
e
a
n
d
T
h
e
o
ry
,
v
o
l
.
1
3
0
,
Ja
n
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
s
i
mp
a
t
.
2
0
2
3
.
1
0
2
8
6
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.