I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
20
24
,
p
p
.
7
2
4
4
~
7
2
5
3
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
14
i
6
.
pp
7
2
4
4
-
7
2
5
3
7244
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
Enha
ncing
reso
ur
ce man
a
g
ement i
n f
o
g
-
clo
ud
in
tern
et
o
f
things
sy
stems with
deep
learning
-
ba
sed
tas
k allo
ca
tion
Vij
a
y
a
la
k
s
hm
i V
enk
a
t
esa
n
,
Sa
ra
v
a
na
n M
uru
gan
D
e
p
a
r
t
me
n
t
o
f
N
e
t
w
o
r
k
i
n
g
a
n
d
C
o
m
mu
n
i
c
a
t
i
o
n
s,
F
a
c
u
l
t
y
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
R
M
I
n
st
i
t
u
t
e
o
f
S
c
i
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
C
h
e
n
n
a
i
,
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
1
7
,
2
0
2
4
R
ev
is
ed
Au
g
9
,
2
0
2
4
Acc
ep
ted
Au
g
2
0
,
2
0
2
4
A
fo
g
-
c
lo
u
d
in
ter
n
e
t
o
f
t
h
i
n
g
s
(I
o
T)
s
y
ste
m
in
te
g
ra
tes
fo
g
c
o
m
p
u
ti
n
g
wi
th
c
lo
u
d
i
n
fra
stru
c
tu
re
t
o
e
fficie
n
tl
y
m
a
n
a
g
e
p
ro
c
e
ss
in
g
d
a
ta
c
lo
s
e
r
to
th
e
so
u
rc
e
,
re
d
u
c
i
n
g
late
n
c
y
a
n
d
b
a
n
d
wid
t
h
u
sa
g
e
.
Eff
icie
n
t
tas
k
sc
h
e
d
u
li
n
g
i
n
fo
g
-
c
lo
u
d
sy
ste
m
is
c
ru
c
ial
f
o
r
o
p
ti
m
izi
n
g
re
so
u
rc
e
u
ti
li
z
a
ti
o
n
a
n
d
m
in
imiz
in
g
e
n
e
r
g
y
c
o
n
su
m
p
ti
o
n
.
E
v
e
n
th
o
u
g
h
m
a
n
y
a
u
t
h
o
rs
p
ro
p
o
se
d
e
n
e
rg
y
e
fficie
n
t
a
l
g
o
rit
h
m
s,
fa
il
e
d
t
o
p
ro
v
id
e
e
fficie
n
t
m
e
th
o
d
t
o
d
e
c
id
e
t
h
e
tas
k
p
lac
e
m
e
n
t
b
e
twe
e
n
f
o
g
n
o
d
e
s
a
n
d
c
lo
u
d
n
o
d
e
s.
T
h
e
p
ro
p
o
s
e
d
h
y
b
ri
d
a
p
p
ro
a
c
h
is
u
se
d
t
o
d
isti
n
g
u
ish
t
h
e
tas
k
p
lac
e
m
e
n
t
b
e
twe
e
n
fo
g
a
n
d
c
lo
u
d
n
o
d
e
s.
T
h
e
h
y
b
rid
a
p
p
r
o
a
c
h
c
o
m
p
rise
s
th
e
p
a
ra
m
e
tri
c
tas
k
c
a
t
e
g
o
riza
ti
o
n
a
lg
o
rit
h
m
(P
TCA)
f
o
r
tas
k
c
a
teg
o
riza
ti
o
n
a
n
d
th
e
m
u
lt
i
m
e
tri
c
f
o
re
c
a
stin
g
m
o
d
e
l
(M
M
F
M
)
b
a
se
d
o
n
d
e
e
p
d
e
term
in
isti
c
p
o
l
icy
g
ra
d
ien
t
(DD
P
G
)
re
c
u
rre
n
t
n
e
u
ra
l
n
e
tw
o
rk
s
fo
r
sc
h
e
d
u
l
in
g
d
e
c
isio
n
s.
P
TCA
c
las
sifies
tas
k
s
b
a
se
d
o
n
p
rio
r
it
y
,
q
u
a
li
t
y
o
f
se
rv
ice
(Qo
S
)
d
e
m
a
n
d
s,
a
n
d
c
o
m
p
u
tati
o
n
a
l
n
e
e
d
s,
fa
c
il
it
a
ti
n
g
i
n
fo
rm
e
d
d
e
c
i
sio
n
s
o
n
tas
k
e
x
e
c
u
ti
o
n
lo
c
a
ti
o
n
s.
M
M
F
M
e
n
h
a
n
c
e
s
sc
h
e
d
u
li
n
g
b
y
o
p
ti
m
izi
n
g
e
n
e
r
g
y
e
fficie
n
c
y
a
n
d
tas
k
c
o
m
p
letio
n
ti
m
e
.
Th
e
e
x
p
e
rime
n
tal
e
v
a
l
u
a
ti
o
n
o
u
t
p
e
rfo
rm
s
th
e
e
x
isti
n
g
m
o
d
e
ls,
in
c
lu
d
in
g
ra
n
d
o
m
fo
re
st
(
R
F)
,
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
(
S
VM),
a
n
d
k
-
n
e
a
re
st
n
e
ig
h
b
o
rs
(KN
N)
.
Th
e
p
r
o
p
o
se
d
re
su
lt
sh
o
ws
a
n
a
c
c
u
ra
c
y
ra
te
o
f
8
9
%
,
a
n
d
e
n
e
rg
y
is
c
o
n
su
m
e
d
5
0
%
les
se
r
th
a
n
t
h
e
e
x
isti
n
g
m
o
d
e
ls.
T
h
e
p
r
o
p
o
se
d
re
se
a
rc
h
a
d
v
a
n
c
e
s
e
n
e
rg
y
-
e
ffici
e
n
t
tas
k
sc
h
e
d
u
li
n
g
,
e
n
a
b
l
in
g
in
telli
g
e
n
t
re
so
u
rc
e
m
a
n
a
g
e
m
e
n
t
in
f
o
g
-
c
l
o
u
d
Io
T
e
n
v
iro
n
m
e
n
ts.
K
ey
w
o
r
d
s
:
E
n
er
g
y
co
n
s
u
m
p
tio
n
Fo
g
co
m
p
u
tin
g
Qu
ality
o
f
s
er
v
ice
d
e
m
an
d
R
eso
u
r
ce
u
tili
za
tio
n
T
ask
s
ch
ed
u
lin
g
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
:
Vijay
alak
s
h
m
i
Ven
k
atesan
Dep
ar
tm
en
t o
f
Netwo
r
k
in
g
an
d
C
o
m
m
u
n
icatio
n
s
,
Facu
lty
o
f
E
n
g
in
ee
r
i
n
g
an
d
T
ec
h
n
o
lo
g
y
,
SR
M
I
n
s
titu
te
o
f
Scien
ce
an
d
T
ec
h
n
o
lo
g
y
Kattan
k
u
lath
u
r
-
6
0
3
2
0
3
,
C
h
en
n
ai,
T
am
il Na
d
u
,
I
n
d
ia
E
m
ail: v
ijay
alv
@
s
r
m
is
t.e
d
u
.
in
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
in
ter
n
et
o
f
th
in
g
s
(
I
o
T
)
a
n
d
f
o
g
co
m
p
u
tin
g
en
v
ir
o
n
m
e
n
ts
ar
e
r
ap
id
ly
d
ev
elo
p
in
g
tec
h
n
o
lo
g
ies
th
at
f
ac
ilit
ate
th
e
s
ea
m
less
co
m
m
u
n
icatio
n
a
n
d
ex
c
h
an
g
e
o
f
d
ata
b
etwe
en
d
ev
ices
[
1
]
.
Ho
wev
er
,
as
co
n
n
ec
te
d
d
ev
ices
g
r
o
w,
ef
f
icien
t
task
s
ch
ed
u
lin
g
b
ec
o
m
es
cr
u
cia
l
to
o
p
tim
ize
r
eso
u
r
ce
u
tili
za
tio
n
an
d
e
n
er
g
y
co
n
s
u
m
p
tio
n
.
E
n
er
g
y
-
ef
f
icien
t
task
s
ch
ed
u
lin
g
ad
d
r
ess
es
t
h
e
ch
allen
g
e
o
f
allo
ca
tin
g
task
s
to
ap
p
r
o
p
r
iate
r
eso
u
r
ce
s
with
in
a
n
I
o
T
-
f
o
g
e
n
v
ir
o
n
m
en
t
t
h
at
m
in
im
izes
e
n
er
g
y
c
o
n
s
u
m
p
tio
n
w
h
ile
m
ax
i
m
izin
g
th
e
o
v
er
all
s
y
s
tem
p
er
f
o
r
m
a
n
ce
[
2
]
.
T
h
is
is
esp
ec
ially
im
p
o
r
tan
t
as
m
a
n
y
d
e
v
ices
in
I
o
T
-
f
o
g
en
v
ir
o
n
m
en
ts
ar
e
b
atter
y
-
p
o
wer
ed
an
d
h
av
e
lim
ited
e
n
e
r
g
y
r
eso
u
r
ce
s
.
On
e
o
f
th
e
k
e
y
m
o
tiv
atio
n
s
f
o
r
e
n
er
g
y
-
ef
f
ici
en
t
task
s
ch
ed
u
lin
g
in
I
o
T
-
f
o
g
e
n
v
ir
o
n
m
en
ts
is
th
e
n
ee
d
to
ex
ten
d
th
e
life
tim
e
o
f
d
ev
ices
[
3
]
.
B
y
ef
f
icien
t
s
ch
e
d
u
lin
g
o
f
task
s
an
d
lev
er
ag
in
g
th
e
co
m
p
u
tatio
n
al
ca
p
ab
ilit
ies
o
f
f
o
g
n
o
d
es,
e
n
er
g
y
c
o
n
s
u
m
p
tio
n
ca
n
b
e
m
i
n
im
ized
,
u
ltima
tely
p
r
o
lo
n
g
in
g
t
h
e
life
o
f
d
ev
ices
an
d
r
ed
u
cin
g
t
h
e
n
ee
d
f
o
r
f
r
eq
u
en
t r
ep
lace
m
e
n
t.
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
-
8
7
0
8
E
n
h
a
n
cin
g
r
eso
u
r
ce
ma
n
a
g
e
men
t in
fo
g
-
clo
u
d
in
tern
et
o
f th
in
g
s
…
(
V
ija
ya
l
a
ksh
mi
V
en
k
a
tesa
n
)
7245
T
h
e
g
en
er
al
u
tili
za
tio
n
o
f
t
h
e
r
eso
u
r
ce
s
ca
n
b
e
o
p
tim
ized
b
y
p
lan
n
in
g
an
d
s
ch
e
d
u
lin
g
t
h
e
t
ask
b
ased
o
n
th
e
p
o
wer
th
e
y
co
n
s
u
m
e
a
n
d
th
e
ab
ilit
y
o
f
th
e
av
ailab
le
r
eso
u
r
ce
s
[
4
]
,
[
5
]
.
T
h
is
h
elp
s
in
ac
co
m
p
lis
h
in
g
task
s
with
o
u
t
m
u
ch
laten
cy
an
d
im
p
r
o
v
es
th
e
ex
p
er
ien
ce
o
f
u
s
er
s
o
r
is
b
en
ef
icial
to
u
s
e
r
in
ter
ac
tio
n
s
.
T
h
e
d
y
n
am
ic
n
atu
r
e
o
f
th
ese
e
n
v
i
r
o
n
m
en
ts
,
with
d
e
v
ices
jo
in
in
g
an
d
leav
in
g
t
h
e
n
etwo
r
k
f
r
eq
u
en
tly
,
m
ak
es
it
d
if
f
icu
lt to
p
r
e
d
ict
r
eso
u
r
ce
a
v
ailab
ilit
y
ac
cu
r
ately
.
T
h
is
u
n
c
er
tain
ty
ca
u
s
es a
ch
allen
g
e
wh
en
s
ch
ed
u
lin
g
task
s
an
d
m
ay
lead
to
s
u
b
o
p
tim
al
r
eso
u
r
ce
allo
ca
tio
n
.
Fu
r
th
e
r
m
o
r
e,
d
ec
id
in
g
wh
e
r
e
to
ex
ec
u
te
task
s
,
eith
er
at
th
e
f
o
g
n
o
d
es
o
r
in
th
e
clo
u
d
,
a
d
d
s
co
m
p
lex
ity
to
th
e
task
s
ch
ed
u
lin
g
p
r
o
ce
s
s
.
W
h
ile
ex
ec
u
tin
g
task
s
at
th
e
f
o
g
n
o
d
e
ca
n
r
e
d
u
ce
late
n
cy
a
n
d
e
n
er
g
y
co
n
s
u
m
p
tio
n
,
it
m
ig
h
t
n
o
t
alwa
y
s
b
e
f
ea
s
ib
le
d
u
e
to
li
m
ited
r
eso
u
r
ce
s
o
r
s
p
ec
if
ic
ap
p
licatio
n
r
e
q
u
ir
em
e
n
ts
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
o
lo
g
y
u
s
es
th
e
p
ar
am
etr
ic
task
ca
t
eg
o
r
izatio
n
m
o
d
u
le
(
PTCM)
alg
o
r
ith
m
wh
ich
s
er
v
es
as
th
e
co
r
n
er
s
to
n
e
o
f
o
u
r
task
s
ch
ed
u
lin
g
p
latf
o
r
m
,
p
r
o
v
i
d
in
g
a
m
ec
h
a
n
is
m
f
o
r
class
if
y
in
g
task
s
b
ased
o
n
m
u
ltip
le
p
ar
am
eter
s
s
u
ch
as
p
r
io
r
ity
,
q
u
ality
o
f
s
er
v
ice
(
Q
o
S)
r
eq
u
ir
em
en
t
s
,
an
d
p
r
o
ce
s
s
in
g
ch
ar
ac
ter
is
tics
.
On
ce
task
s
ar
e
class
if
ied
b
y
th
e
p
ar
am
etr
ic
t
ask
ca
teg
o
r
izatio
n
alg
o
r
ith
m
(
PTCA)
alg
o
r
ith
m
,
th
e
d
ee
p
d
ete
r
m
in
is
tic
p
o
licy
g
r
ad
ien
t (
DDPG)
-
b
ased
m
u
lti
-
m
etr
ic
f
o
r
ec
asti
n
g
m
o
d
el
(
M
MFM
)
m
o
d
el
is
u
s
ed
to
o
p
tim
ize
task
s
ch
ed
u
lin
g
d
ec
is
io
n
s
.
DDP
G
is
a
Q
-
lear
n
i
n
g
b
ased
r
ein
f
o
r
ce
m
en
t
lear
n
in
g
alg
o
r
ith
m
th
a
t
tak
es
p
o
licy
g
r
ad
ien
t
lo
g
ic
i
n
to
ac
co
u
n
t,
b
u
ilt
f
r
o
m
ex
p
e
r
ien
ce
r
ep
lay
an
d
r
eq
u
ir
es
n
o
p
r
io
r
m
o
d
el.
I
t
is
co
m
p
o
s
ed
o
f
ac
to
r
an
d
a
cr
itic
wh
ich
m
ak
e
u
p
th
e
DDPG
m
eth
o
d
.
W
h
er
ea
s
ac
to
r
in
p
u
ts
s
tates
an
d
o
u
tp
u
ts
ac
tio
n
s
,
cr
itic
tak
es
s
tate
an
d
two
ac
ts
to
o
u
tp
u
t
th
e
Q
-
v
al
u
e
[
6
]
.
Fo
r
e
n
er
g
y
-
ef
f
icien
t
s
ch
ed
u
lin
g
o
f
task
s
DDPG
is
u
til
ized
in
f
o
g
-
cl
o
u
d
s
y
s
tem
s
an
d
o
f
f
e
r
s
ad
ap
tiv
e
r
ea
l
-
tim
e
ad
ju
s
tm
en
ts
,
en
s
u
r
i
n
g
o
p
tim
al
r
eso
u
r
ce
allo
ca
tio
n
to
m
ee
t
Qo
S
d
em
an
d
s
ef
f
icien
tly
[
7
]
.
I
ts
in
teg
r
atio
n
en
ab
les
en
h
a
n
ce
d
d
ec
is
io
n
-
m
ak
in
g
,
s
u
r
p
ass
in
g
co
n
v
en
tio
n
al
tech
n
iq
u
es b
y
d
y
n
am
ically
o
p
tim
izin
g
en
er
g
y
c
o
n
s
u
m
p
tio
n
an
d
task
c
o
m
p
leti
o
n
tim
e.
Qiu
et
a
l.
[
8
]
p
r
o
p
o
s
ed
a
n
o
v
e
l
en
er
g
y
m
an
ag
e
m
en
t
alg
o
r
ith
m
u
s
in
g
th
e
DDPG
m
eth
o
d
f
o
r
wir
eles
s
s
en
s
o
r
n
etwo
r
k
s
with
r
e
n
ewa
b
le
en
er
g
y
s
o
u
r
ce
s
d
e
m
o
n
s
tr
a
tes
s
u
p
er
io
r
l
o
n
g
-
ter
m
p
er
f
o
r
m
an
ce
an
d
e
f
f
ec
tiv
e
en
er
g
y
m
an
a
g
em
en
t
i
n
r
ea
l
-
wo
r
ld
s
ce
n
ar
i
o
s
.
T
h
e
p
r
o
p
o
s
e
d
w
o
r
k
b
y
L
i
et
a
l.
[
9
]
a
d
d
r
ess
es
e
n
e
r
g
y
co
n
s
u
m
p
t
io
n
i
n
E
d
g
e
o
f
t
h
i
n
g
s
c
o
m
p
u
ti
n
g
,
a
n
o
v
el
s
ca
li
n
g
m
ec
h
an
is
m
m
o
n
it
o
r
s
I
o
T
ap
p
l
icat
io
n
w
o
r
k
l
o
a
d
s
,
ca
t
eg
o
r
iz
es
th
em
b
y
i
n
te
n
s
it
y
,
an
d
d
y
n
a
m
i
ca
ll
y
a
d
j
u
s
ts
p
r
o
c
e
s
s
o
r
s
p
e
ed
t
o
b
ala
n
ce
Q
o
S
a
n
d
e
n
e
r
g
y
e
f
f
ic
ie
n
c
y
,
th
o
u
g
h
c
o
n
ti
n
u
o
u
s
m
o
n
it
o
r
in
g
m
a
y
s
o
m
e
ti
m
es
in
cr
ea
s
e
p
o
w
er
c
o
n
s
u
m
p
ti
o
n
a
n
d
d
eg
r
a
d
e
p
er
f
o
r
m
a
n
c
e.
B
ab
ar
an
d
Kh
an
[
1
0
]
ex
am
in
es
a
s
ca
lab
le
en
er
g
y
-
ef
f
icien
t
p
a
r
ad
ig
m
th
at
u
s
es
r
ec
u
r
s
iv
e
clu
s
ter
in
g
to
p
r
io
r
itize
h
ig
h
-
p
r
io
r
ity
task
s
o
n
f
o
g
s
er
v
er
s
,
em
p
lo
y
in
g
E
u
clid
ea
n
d
is
tan
ce
f
o
r
clu
s
ter
in
g
a
n
d
r
o
u
n
d
-
r
o
b
in
s
ch
ed
u
lin
g
f
o
r
clu
s
ter
lead
er
s
,
b
u
t
f
ac
es
ch
allen
g
es
lik
e
o
v
er
f
itti
n
g
an
d
in
cr
ea
s
ed
co
m
p
u
tatio
n
al
co
s
ts
.
Yu
et
a
l.
[
1
1
]
ex
p
lo
r
es
th
e
in
teg
r
atio
n
o
f
an
u
n
m
a
n
n
ed
ae
r
ial
v
eh
icle
in
to
a
wir
eless
I
o
T
n
etwo
r
k
,
em
p
lo
y
in
g
a
f
ly
-
h
o
v
er
-
co
m
m
u
n
icate
p
r
o
to
c
o
l
f
o
r
d
at
a
co
llectio
n
an
d
d
ev
ice
c
h
ar
g
in
g
.
I
t
f
o
r
m
u
lates
a
m
u
lti
-
o
b
j
ec
tiv
e
o
p
tim
izatio
n
p
r
o
b
lem
wh
ich
m
ax
im
izes
d
a
ta
r
ate
a
n
d
h
ar
v
ested
en
er
g
y
.
I
n
a
m
o
b
ile
ed
g
e
c
o
m
p
u
ti
n
g
s
y
s
tem
,
a
d
y
n
am
ic
o
f
f
lo
ad
i
n
g
m
ec
h
an
is
m
b
ased
o
n
d
ee
p
Q
n
etwo
r
k
(
DQN)
is
s
u
g
g
ested
in
[
1
2
]
.
L
o
n
g
s
h
o
r
t
-
ter
m
m
e
m
o
r
y
(
L
STM
)
an
d
Q
-
lear
n
in
g
f
o
r
o
f
f
lo
ad
in
g
d
esti
n
atio
n
s
elec
tio
n
wer
e
in
teg
r
ated
[
1
3
]
t
o
f
o
r
ec
a
s
t
th
e
ch
an
n
el
g
ain
in
th
e
n
etwo
r
k
.
T
h
e
d
ee
p
r
ei
n
f
o
r
ce
m
e
n
t
lear
n
i
n
g
(
DR
L
)
-
b
ased
alg
o
r
ith
m
ca
n
p
r
o
d
u
ce
h
ig
h
er
ac
c
u
r
ac
y
an
d
b
etter
s
elf
-
co
r
r
ec
tin
g
p
er
f
o
r
m
an
ce
as
co
m
p
ar
ed
t
o
ex
is
tin
g
tech
n
iq
u
es
[
1
4
]
.
As
I
o
T
s
y
s
tem
s
co
n
tin
u
e
to
ex
p
an
d
,
m
an
a
g
in
g
th
e
s
ca
lab
i
lity
an
d
c
o
m
p
lex
ity
o
f
task
s
ch
ed
u
lin
g
b
ec
o
m
es
in
cr
ea
s
in
g
l
y
ch
allen
g
i
n
g
[
1
5
]
.
DDPG
o
f
f
er
s
a
p
r
o
m
is
in
g
s
o
lu
tio
n
b
y
ad
ap
ti
n
g
to
ev
o
lv
i
n
g
s
y
s
tem
d
y
n
am
ics
an
d
o
p
t
im
izin
g
s
ch
ed
u
lin
g
d
ec
is
io
n
s
in
r
ea
l
-
tim
e.
T
h
e
m
ain
co
n
tr
ib
u
tio
n
s
o
f
th
e
p
r
o
p
o
s
ed
wo
r
k
ar
e
as f
o
llo
ws:
−
E
f
f
icien
t
task
s
ch
ed
u
lin
g
in
f
o
g
-
clo
u
d
I
o
T
s
y
s
tem
s
is
cr
u
cial
f
o
r
o
p
tim
izin
g
r
eso
u
r
ce
u
tili
za
tio
n
an
d
m
in
im
izin
g
e
n
er
g
y
co
n
s
u
m
p
ti
o
n
.
E
x
is
tin
g
en
e
r
g
y
-
ef
f
icien
t
alg
o
r
ith
m
s
s
tr
u
g
g
le
to
ef
f
ec
ti
v
ely
d
ete
r
m
in
e
task
p
lace
m
en
t b
etwe
en
f
o
g
a
n
d
clo
u
d
n
o
d
es,
co
m
p
licatin
g
th
e
d
ec
is
io
n
o
f
lo
ca
l v
er
s
u
s
cl
o
u
d
p
r
o
ce
s
s
in
g
.
−
T
h
e
p
r
o
p
o
s
ed
h
y
b
r
id
ap
p
r
o
ac
h
class
if
ies
task
s
b
ased
o
n
p
r
io
r
ity
,
Qo
S
d
em
a
n
d
s
,
an
d
co
m
p
u
tatio
n
al
n
ee
d
s
,
an
d
o
p
tim
izes
en
er
g
y
ef
f
icien
cy
an
d
task
co
m
p
letio
n
tim
e.
T
h
is
f
ac
ilit
ates
in
f
o
r
m
ed
task
p
lace
m
en
t
an
d
en
h
an
ce
s
o
v
e
r
all
s
y
s
tem
p
er
f
o
r
m
an
ce
.
−
T
h
e
p
r
o
p
o
s
ed
h
y
b
r
id
ap
p
r
o
ac
h
o
u
t
p
er
f
o
r
m
s
ex
is
tin
g
m
o
d
el
s
(
r
an
d
o
m
f
o
r
est
(
R
F)
,
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM
)
,
an
d
k
-
n
ea
r
est
n
eig
h
b
o
r
s
(
KNN
)
)
b
y
ac
h
iev
in
g
8
9
%
ac
cu
r
ac
y
in
task
s
ch
ed
u
lin
g
,
r
ed
u
cin
g
r
esp
o
n
s
e
tim
e
to
8
0
m
s
,
an
d
e
n
er
g
y
c
o
n
s
u
m
p
tio
n
b
y
5
0
%.
T
h
is
ad
v
an
ce
s
en
er
g
y
-
ef
f
icien
t
task
s
ch
ed
u
lin
g
an
d
en
a
b
les m
o
r
e
in
tellig
en
t
r
eso
u
r
ce
m
an
a
g
em
en
t in
f
og
-
clo
u
d
I
o
T
en
v
ir
o
n
m
en
ts
.
T
h
e
ar
ticle
is
o
r
g
an
ized
as
th
e
p
r
o
p
o
s
ed
en
e
r
g
y
-
ef
f
icien
t
MM
FM
-
DDPG
s
ch
ed
u
lin
g
f
r
am
ewo
r
k
in
s
ec
tio
n
2
.
S
ec
tio
n
3
s
h
o
ws
th
e
m
ath
em
atica
l
m
o
d
el
f
o
r
ca
teg
o
r
izatio
n
a
n
d
e
n
er
g
y
-
ef
f
icien
t
s
ch
ed
u
lin
g
.
S
ec
tio
n
4
r
ep
r
esen
ts
th
e
s
im
u
latio
n
an
d
r
esu
lt
.
Fin
ally
,
s
ec
tio
n
5
d
escr
ib
es th
e
co
n
cl
u
s
io
n
.
2.
P
RO
P
O
SE
D
M
E
T
H
O
D
I
n
t
e
g
r
a
t
i
n
g
f
o
g
a
n
d
c
l
o
u
d
c
o
m
p
u
t
i
n
g
i
n
I
o
T
p
r
e
s
e
n
t
s
u
n
p
r
e
c
e
d
e
n
t
e
d
o
p
p
o
r
t
u
n
i
t
i
e
s
f
o
r
a
d
v
a
n
c
i
n
g
s
e
r
v
i
c
es
a
n
d
i
m
p
r
o
v
i
n
g
o
u
t
c
o
m
e
s
.
E
f
f
i
c
i
e
n
t
t
a
s
k
s
c
h
e
d
u
l
i
n
g
,
f
a
c
i
l
i
t
at
e
d
b
y
D
D
PG
,
e
m
e
r
g
e
s
a
s
a
c
r
i
t
ic
a
l
e
n
a
b
l
e
r
i
n
r
e
a
l
iz
i
n
g
t
h
es
e
b
e
n
e
f
i
ts
.
H
o
w
e
v
e
r
,
s
e
v
e
r
a
l
c
h
al
l
e
n
g
es
r
e
la
te
d
t
o
s
c
a
l
a
b
il
i
t
y
,
c
o
m
p
l
e
x
i
t
y
,
an
d
Q
o
S
r
e
q
u
i
r
e
m
e
n
t
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
e
r
20
24
:
7
2
4
4
-
7
2
5
3
7246
p
e
r
s
i
s
t
,
w
a
r
r
a
n
ti
n
g
f
u
r
t
h
e
r
r
e
s
ea
r
c
h
a
n
d
i
n
n
o
v
a
t
i
o
n
i
n
t
h
i
s
d
o
m
a
i
n
.
A
s
s
h
o
w
n
i
n
F
i
g
u
r
e
1
,
t
h
e
p
r
o
p
o
s
e
d
s
y
s
t
e
m
is
d
e
s
i
g
n
e
d
t
o
b
e
s
ca
l
a
b
l
e
,
a
d
a
p
t
iv
e
,
a
n
d
m
o
d
u
l
a
r
t
o
m
e
e
t
t
h
e
v
ar
i
e
d
r
e
q
u
i
r
e
m
e
n
t
s
a
n
d
c
o
n
s
t
r
a
in
t
s
.
Fig
u
r
e
1
.
Sy
s
tem
ar
c
h
itectu
r
e
o
f
task
o
f
f
l
o
ad
in
g
ap
p
r
o
ac
h
T
ask
s
ch
ed
u
lin
g
in
d
is
tr
ib
u
ted
co
m
p
u
tin
g
en
v
ir
o
n
m
en
ts
n
ec
ess
itates
a
s
y
s
tem
atic
a
p
p
r
o
ac
h
to
allo
ca
te
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
ef
f
icien
tly
.
T
h
e
PTCM
u
t
ilizes
a
d
ec
is
io
n
tr
ee
-
b
ased
a
p
p
r
o
ac
h
to
class
if
y
task
s
in
to
d
is
tin
ct
ca
teg
o
r
ies
b
ased
o
n
v
ar
io
u
s
p
ar
am
ete
r
s
s
u
ch
as
p
r
io
r
ity
,
Qo
S
r
eq
u
ir
em
e
n
ts
,
an
d
p
r
o
ce
s
s
in
g
ch
ar
ac
ter
is
tics
.
I
t
co
n
s
tr
u
cts
a
d
ec
is
io
n
tr
ee
iter
ativ
ely
b
y
s
elec
tin
g
th
e
m
o
s
t
in
f
o
r
m
ativ
e
f
e
atu
r
es
an
d
s
p
litt
in
g
th
e
d
ataset
b
ased
o
n
th
o
s
e
f
e
atu
r
es
u
n
til
th
e
en
tire
d
ataset
is
co
r
r
ec
tly
class
if
ied
.
E
ac
h
d
ec
is
io
n
tr
ee
n
o
d
e
r
ep
r
esen
ts
a
f
ea
tu
r
e,
an
d
ea
c
h
b
r
an
c
h
r
ep
r
esen
ts
a
p
o
s
s
ib
le
f
ea
tu
r
e
v
alu
e.
T
h
e
leaf
n
o
d
es
co
n
tain
th
e
class
lab
els
ass
ig
n
ed
to
th
e
task
s
b
ased
o
n
th
e
s
elec
ted
f
ea
t
u
r
es.
T
h
is
ca
teg
o
r
izatio
n
en
s
u
r
es
th
at
task
s
ar
e
p
r
io
r
itized
an
d
m
a
n
ag
ed
ef
f
ec
tiv
ely
,
en
ab
lin
g
th
e
s
y
s
tem
to
allo
ca
te
r
eso
u
r
ce
s
th
at
m
ax
im
izes
p
er
f
o
r
m
an
ce
an
d
m
ee
t u
s
er
e
x
p
ec
tatio
n
s
.
E
f
f
ec
tiv
e
task
s
ch
ed
u
lin
g
also
r
eq
u
ir
es
an
ab
ilit
y
to
p
r
ed
i
ct
f
u
tu
r
e
wo
r
k
lo
ad
s
an
d
ad
a
p
t
r
eso
u
r
ce
allo
ca
tio
n
ac
co
r
d
in
g
ly
.
T
h
e
MM
FM
in
teg
r
ates
ad
v
an
ce
d
m
ac
h
in
e
lear
n
in
g
tech
n
i
q
u
es
to
p
r
o
v
id
e
th
is
p
r
ed
ictiv
e
ca
p
ab
ilit
y
.
Utilizin
g
a
r
ein
f
o
r
ce
m
e
n
t
lear
n
in
g
f
r
a
m
ewo
r
k
,
MM
FM
an
aly
ze
s
h
is
to
r
ical
task
d
ata
to
f
o
r
ec
ast
tr
en
d
s
a
n
d
p
atter
n
s
in
wo
r
k
lo
a
d
d
e
m
an
d
s
.
T
h
is
f
o
r
e
s
ig
h
t
en
ab
les
th
e
s
y
s
tem
t
o
p
r
o
ac
tiv
ely
ad
ju
s
t
its
s
ch
ed
u
lin
g
s
tr
ateg
y
,
en
s
u
r
in
g
th
at
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
ar
e
o
p
tim
ally
u
tili
ze
d
,
r
esp
o
n
s
e
tim
es
ar
e
m
in
im
ized
,
a
n
d
Qo
S
cr
iter
ia
ar
e
co
n
s
is
ten
tly
m
et.
DDPG
is
o
n
e
ty
p
e
o
f
r
ei
n
f
o
r
ce
m
en
t
l
ea
r
n
in
g
wh
er
e
t
h
e
p
o
licy
is
lear
n
ed
d
ir
ec
tly
with
o
u
t
t
h
e
u
s
e
o
f
a
p
ar
am
et
r
ic
m
ax
im
u
m
s
tr
ateg
y
t
h
r
o
u
g
h
r
e
test
in
g
th
e
m
e
m
o
r
y
an
d
co
n
s
tan
tly
lear
n
in
g
th
r
o
u
g
h
v
alu
ed
id
ea
s
o
f
Q
-
l
ea
r
n
in
g
a
n
d
p
o
lic
y
g
r
a
d
ien
t lo
g
ics.
T
h
e
DDPG
alg
o
r
ith
m
is
im
p
l
em
en
ted
b
y
th
e
ac
to
r
a
n
d
th
e
cr
itic.
T
h
e
cr
itic
tak
es
a
s
ta
te
an
d
ac
ts
to
g
iv
e
th
e
Q
-
v
alu
e,
wh
ile
th
e
ac
to
r
tak
es
th
e
s
tate
s
an
d
g
iv
es
ac
tio
n
s
.
B
y
iter
ativ
ely
ad
ju
s
ti
n
g
its
p
ar
am
eter
s
in
r
esp
o
n
s
e
to
o
b
s
er
v
ed
r
ewa
r
d
s
an
d
s
y
s
tem
v
ar
iab
les,
th
e
DD
PG
-
b
ased
MM
FM
m
o
d
el
lear
n
s
th
is
p
o
licy
.
T
h
is
en
ab
les
it
to
a
d
ju
s
t
to
c
h
an
g
i
n
g
en
v
ir
o
n
m
en
tal
c
o
n
d
itio
n
s
an
d
g
r
ad
u
ally
o
p
tim
ize
task
s
ch
ed
u
lin
g
d
ec
is
io
n
s
[
1
6
]
.
On
ce
th
e
task
s
h
av
e
b
ee
n
ca
teg
o
r
ized
an
d
s
ch
ed
u
led
,
th
e
s
y
s
tem
'
s
o
v
er
all
en
er
g
y
co
n
s
u
m
p
tio
n
m
ay
b
e
co
m
p
u
ted
b
y
ad
d
in
g
u
p
th
e
e
n
er
g
y
u
s
ed
b
y
ea
ch
task
o
n
ea
c
h
d
ev
ice.
DDPG
is
u
s
ed
in
f
o
g
-
clo
u
d
s
y
s
tem
s
f
o
r
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
-
8
7
0
8
E
n
h
a
n
cin
g
r
eso
u
r
ce
ma
n
a
g
e
men
t in
fo
g
-
clo
u
d
in
tern
et
o
f th
in
g
s
…
(
V
ija
ya
l
a
ksh
mi
V
en
k
a
tesa
n
)
7247
task
s
ch
ed
u
lin
g
th
at
is
en
er
g
y
-
ef
f
icien
t
an
d
p
r
o
v
id
es
ad
a
p
tiv
e
r
ea
l
-
tim
e
m
o
d
if
icatio
n
s
to
en
s
u
r
e
o
p
tim
al
r
eso
u
r
ce
allo
ca
tio
n
to
f
u
lf
il
Qo
S
d
em
an
d
s
ef
f
ec
tiv
ely
[
1
7
]
.
T
h
r
o
u
g
h
its
in
teg
r
atio
n
,
d
ec
is
io
n
-
m
ak
in
g
is
im
p
r
o
v
e
d
,
a
n
d
it
o
u
tp
e
r
f
o
r
m
s
tr
ad
itio
n
al
m
eth
o
d
s
b
y
d
y
n
a
m
ically
o
p
tim
izin
g
en
er
g
y
co
n
s
u
m
p
tio
n
an
d
task
co
m
p
letio
n
tim
e.
B
y
in
co
r
p
o
r
atin
g
th
e
MM
FM,
o
u
r
s
y
s
te
m
is
eq
u
ip
p
e
d
to
f
o
r
ec
ast
f
l
u
ctu
atio
n
s
in
task
wo
r
k
lo
ad
s
an
d
ad
a
p
t
th
e
s
ch
ed
u
lin
g
s
tr
ateg
y
ac
co
r
d
in
g
ly
.
T
h
is
p
r
ed
ictiv
e
f
u
n
ctio
n
ality
f
ac
ilit
ates
p
r
o
ac
tiv
e
r
eso
u
r
ce
allo
ca
tio
n
,
en
s
u
r
in
g
o
p
tim
al
u
s
e
o
f
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
,
m
in
im
izin
g
r
e
s
p
o
n
s
e
tim
es,
an
d
m
ee
tin
g
Qo
S c
r
iter
ia.
T
h
e
p
r
o
p
o
s
ed
MM
FM
-
DDP
G
alg
o
r
ith
m
s
h
o
wn
in
Alg
o
r
ith
m
1
lear
n
s
f
r
o
m
h
is
to
r
ical
task
d
ata,
wh
ich
in
clu
d
es
p
r
ev
io
u
s
wo
r
k
lo
ad
p
atter
n
s
,
r
eso
u
r
ce
u
tili
za
tio
n
,
an
d
en
v
ir
o
n
m
en
tal
c
o
n
d
itio
n
s
.
T
h
r
o
u
g
h
iter
ativ
e
tr
ain
in
g
an
d
o
p
tim
iz
atio
n
,
th
e
m
o
d
el
g
ain
s
a
d
ee
p
u
n
d
er
s
tan
d
i
n
g
o
f
th
e
u
n
d
er
l
y
in
g
d
y
n
am
ics
th
at
in
f
lu
en
ce
task
s
ch
ed
u
lin
g
with
in
th
e
s
y
s
tem
.
T
h
e
DDPG
alg
o
r
ith
m
u
s
ed
in
th
e
f
o
r
ec
asti
n
g
m
o
d
el
f
o
r
e
f
f
icien
t
s
ch
ed
u
lin
g
in
f
o
g
-
clo
u
d
b
ase
d
s
y
s
tem
s
s
h
o
wn
in
Fig
u
r
e
2
aim
s
to
o
p
tim
ize
th
e
allo
ca
tio
n
o
f
task
s
b
etwe
en
f
o
g
n
o
d
es a
n
d
clo
u
d
s
er
v
er
s
to
m
in
im
ize
en
er
g
y
co
n
s
u
m
p
tio
n
wh
ile
m
ain
tain
in
g
th
e
Qo
S.
Alg
o
r
ith
m
1
.
Pro
p
o
s
ed
m
u
lti
-
m
etr
ic
f
o
r
ec
asti
n
g
m
o
d
el
with
d
ee
p
d
eter
m
i
n
is
tic
p
o
licy
g
r
a
d
ien
t
(
MM
FM
-
DDPG)
In
pu
t:
Cl
as
si
fi
ed
ta
sk
s,
Cr
it
ic
an
d
Ac
to
r
Networks
Output:
Optimized
task
sc
heduling
for
energy
efficiency and QoS fulfillment
Initialize MMFM Lists:
Cr
ea
te
li
st
s
fo
r
sc
he
du
le
d
ta
sk
s
)
an
d
high
-
priority tasks (
ℎ
).
Initialize DDPG:
Initialize
critic
and
actor
networks
for
task scheduling decisions.
Initialize
a
replay
buffer
to
store
experience tuples
(
,
,
,
+
1
)
.
MMFM Task Evaluation and Scheduling:
For each task T in the classified task list:
Initialize
=
None
For each task T in the list:
If
>
=
(
)
:
=
El
se
if
Pr
io
ri
ty
(T
)=
hi
gh
an
d
us
ag
e(
T)
<
MMFM(T):
Append the task
ℎ
=
ℎ
∪
{T}
If
∑
(
)
>
=
(
)
∈
L
ℎ
T
=
max
∈
L
ℎ
(
)
End inner loop.
If
≠ None:
ℎ
=
ℎ
∪
{
}
Else:
Task list
= Task list
∪
{T}
If
≠
Null and
=
None:
=
∪
{T}
If
≠ None:
=
∪
{T}
DDPG for Energy Efficiency and QoS:
For each episode from 1 to M:
Initialize the starting state s
0
.
While the state is not a terminal state:
Select an action a
t
:
=
{
ℎ
(
|
)
ℎ
Execute
action
a
t
an
d
ob
se
rv
e
re
wa
rd
r
t
and next state s
t+1
.
St
or
e
th
e
ex
pe
ri
en
ce
tu
pl
e
(
,
,
,
+
1
)
in the replay buffer.
Update the current state to
+
1
.
Sa
mp
le
a
ra
nd
om
mi
ni
-
ba
tc
h
fr
om
th
e
replay buffer.
Set target value
=
+
(
+
1
|
)
|
Up
da
te
th
e
cr
it
ic
ne
t
wo
rk
by
mi
ni
mi
zi
ng
the loss L
c
:
=
−
1
∑
(
−
(
,
∣
)
)
∧
2
=
1
Up
da
te
th
e
ac
to
r
n
et
wo
rk
us
in
g
th
e
gradient:
≅
≅
−
1
∑
=
1
(
,
∣
)
(
∣
)
P
e
r
fo
r
m
s
o
f
t
u
p
da
t
es
o
n
t
h
e
t
a
rg
e
t
networks
:
=
+
(
1
−
)
,
=
+
(
1
−
)
I
n
th
is
alg
o
r
ith
m
,
ac
to
r
an
d
cr
itic
n
eu
r
al
n
etwo
r
k
s
ar
e
in
itial
ized
to
m
ak
e
task
ass
ig
n
m
en
t
d
ec
is
io
n
s
b
ased
o
n
th
e
cu
r
r
en
t
s
y
s
tem
s
tate,
wh
ich
in
clu
d
es
in
f
o
r
m
a
tio
n
ab
o
u
t
task
s
,
f
o
g
n
o
d
es,
an
d
clo
u
d
s
er
v
er
s
.
Du
r
in
g
ea
ch
ep
is
o
d
e,
th
e
s
y
s
tem
s
tate
i
s
r
eset,
an
d
f
o
r
ea
ch
t
ask
,
an
ac
tio
n
(
task
ass
ig
n
m
en
t)
is
s
elec
ted
u
s
in
g
th
e
cu
r
r
en
t
p
o
licy
an
d
ex
p
l
o
r
atio
n
s
tr
ateg
y
.
T
h
e
ac
tio
n
'
s
im
p
ac
t
is
s
im
u
lated
in
t
h
e
f
o
g
-
clo
u
d
I
o
T
en
v
ir
o
n
m
en
t
to
co
m
p
u
te
a
r
ewa
r
d
b
ased
o
n
en
er
g
y
ef
f
icie
n
cy
an
d
Qo
S
f
u
lf
illme
n
t.
T
h
e
ex
p
er
ien
ce
tu
p
les
co
m
p
r
is
in
g
th
e
s
tate,
ac
tio
n
,
r
ewa
r
d
,
n
ex
t
s
tate,
an
d
ter
m
in
a
tio
n
f
lag
ar
e
s
to
r
ed
in
a
r
ep
la
y
b
u
f
f
er
.
T
h
e
ac
t
o
r
an
d
cr
itic
n
etwo
r
k
s
ar
e
th
en
u
p
d
ated
u
s
in
g
th
e
s
am
p
lin
g
p
o
licy
g
r
ad
ie
n
t
an
d
l
o
s
s
f
u
n
cti
o
n
,
wh
ile
th
e
tar
g
et
n
etwo
r
k
s
ar
e
u
p
d
ate
d
u
s
in
g
s
o
f
t
u
p
d
ates.
T
h
is
p
r
o
ce
s
s
iter
ates
o
v
er
m
u
ltip
le
ep
is
o
d
es
to
tr
ain
th
e
DDPG
m
o
d
el
to
m
a
k
e
ef
f
icien
t ta
s
k
s
ch
ed
u
lin
g
d
ec
is
io
n
s
tailo
r
ed
f
o
r
f
o
g
-
clo
u
d
I
o
T
s
y
s
tem
s
.
On
ce
tr
ain
ed
,
t
h
e
MM
FM
co
n
tin
u
ally
a
n
aly
ze
s
in
co
m
i
n
g
d
ata
to
p
r
o
d
u
ce
r
ea
l
-
tim
e
f
o
r
ec
asts
o
f
f
u
tu
r
e
wo
r
k
lo
ad
tr
en
d
s
.
T
h
ese
f
o
r
ec
asts
ar
e
th
en
in
co
r
p
o
r
at
ed
in
to
th
e
task
-
s
ch
ed
u
lin
g
p
r
o
ce
s
s
,
en
ab
lin
g
th
e
s
y
s
tem
to
allo
ca
te
r
eso
u
r
ce
s
p
r
o
ac
tiv
ely
b
ased
o
n
an
ticip
a
ted
d
em
an
d
.
T
h
e
in
teg
r
atio
n
o
f
MM
FM
in
to
th
e
task
-
s
ch
ed
u
lin
g
f
r
a
m
ewo
r
k
p
r
o
v
id
es
s
ev
er
al
k
ey
ad
v
an
tag
es
.
Firstl
y
,
it
b
o
o
s
ts
th
e
s
y
s
tem
's
r
esp
o
n
s
iv
en
ess
b
y
f
ac
ilit
atin
g
p
r
o
ac
tiv
e
r
eso
u
r
c
e
allo
ca
tio
n
in
an
ticip
atio
n
o
f
p
o
ten
tial
wo
r
k
lo
a
d
ch
an
g
es.
T
h
is
p
r
o
ac
tiv
e
ap
p
r
o
ac
h
g
u
ar
a
n
tees
s
m
o
o
th
o
p
er
atio
n
ev
en
in
an
en
v
ir
o
n
m
en
t
o
f
v
ar
y
in
g
wo
r
k
lo
ad
s
a
n
d
h
elp
s
to
r
ed
u
ce
p
o
ten
tial p
er
f
o
r
m
an
c
e
is
s
u
es.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
e
r
20
24
:
7
2
4
4
-
7
2
5
3
7248
Fig
u
r
e
2
.
DDPG
u
s
ed
in
MM
F
M
3.
M
AT
H
E
M
AT
I
CA
L
M
O
D
E
L
L
I
NG
T
h
e
p
ar
am
etr
ic
task
ca
teg
o
r
i
za
tio
n
alg
o
r
ith
m
(
PTCA)
em
p
lo
y
s
a
d
ec
is
io
n
tr
ee
-
b
ased
a
p
p
r
o
ac
h
to
class
if
y
task
s
in
to
d
is
tin
ct
ca
teg
o
r
ies
b
ased
o
n
p
r
e
d
ef
i
n
ed
cr
iter
ia.
T
h
is
alg
o
r
ith
m
u
s
es
en
tr
o
p
y
an
d
in
f
o
r
m
atio
n
g
ain
wh
ich
is
u
s
ed
to
b
u
ild
a
d
ec
is
io
n
tr
ee
f
o
r
t
ask
ca
teg
o
r
izatio
n
.
T
h
e
f
ea
tu
r
e
s
p
ac
e,
r
ep
r
esen
ted
by
,
in
clu
d
es
task
ch
ar
ac
ter
is
tics
,
wh
ile
Y
r
ep
r
esen
ts
th
e
s
e
t
o
f
p
o
s
s
ib
le
task
ca
teg
o
r
ies.
T
h
e
d
ec
is
io
n
tr
ee
lear
n
s
a
m
ap
p
in
g
:
→
th
at
p
ar
titi
o
n
s
th
e
f
ea
tu
r
e
s
p
ac
e
in
to
d
is
jo
in
t
r
eg
io
n
s
,
ea
c
h
co
r
r
esp
o
n
d
in
g
to
a
s
p
ec
if
ic
task
ca
teg
o
r
y
.
T
h
is
m
ap
p
in
g
is
d
er
iv
ed
th
r
o
u
g
h
an
i
ter
ativ
e
p
r
o
ce
s
s
th
at
m
a
x
im
iz
es
in
f
o
r
m
atio
n
g
ain
at
ea
ch
n
o
d
e,
r
esu
ltin
g
i
n
an
e
f
f
ec
tiv
e
class
if
icatio
n
m
o
d
el
f
o
r
task
s
in
I
o
T
en
v
ir
o
n
m
en
ts
.
E
n
tr
o
p
y
is
u
s
ed
to
m
ea
s
u
r
e
th
e
u
n
c
er
tain
ty
in
a
d
ataset
b
ased
o
n
th
e
task
ca
teg
o
r
ies
s
u
ch
as
p
r
io
r
ity
,
Qo
S
d
em
an
d
s
an
d
co
m
p
u
tatio
n
al
n
ee
d
s
an
d
is
ca
lcu
lated
as:
(
)
=
−
∑
l
og
2
(
)
=
1
(
1
)
wh
er
e
r
ep
r
esen
ts
th
e
p
r
o
p
o
r
ti
o
n
o
f
task
s
b
elo
n
g
i
n
g
to
class
,
in
th
e
d
ataset
,
an
d
th
e
n
u
m
b
er
o
f
d
if
f
e
r
en
t
class
es i
s
r
ep
r
esen
ted
as
.
3
.
1
.
I
nfo
rm
a
t
io
n
g
a
in
I
n
f
o
r
m
atio
n
g
ain
is
th
e
q
u
an
ti
ty
o
f
en
tr
o
p
y
th
at
m
a
y
b
e
s
av
ed
wh
en
a
s
p
ec
if
ic
cr
iter
io
n
is
u
s
ed
as
a
s
p
litt
in
g
f
ea
tu
r
e
to
o
p
tim
ize
a
d
ataset's
ca
teg
o
r
izatio
n
.
B
ased
o
n
th
e
"
"
th
e
p
r
o
v
id
e
d
d
ata
s
et,
d
esig
n
ated
as
,
is
f
u
r
th
er
s
ep
a
r
ated
in
to
s
u
b
-
s
ets
1
,
2
,
…
,
an
d
.
T
h
e
weig
h
ted
s
u
m
o
f
th
e
en
tr
o
p
ies
o
f
th
ese
s
u
b
s
ets
is
th
e
en
tr
o
p
y
f
o
llo
win
g
th
e
s
p
lit.
T
h
e
in
f
o
r
m
atio
n
g
ain
is
th
e
d
if
f
er
e
n
ce
b
etwe
en
th
e
p
r
e
-
s
p
lit
an
d
p
o
s
t
-
s
p
lit en
tr
o
p
ies.
I
n
f
o
r
m
atio
n
g
ain
is
ca
lcu
lated
as
(
2
)
.
(
,
)
=
(
)
−
∑
|
|
|
|
∗
(
)
(
2
)
T
h
e
f
ea
tu
r
e
with
t
h
e
h
ig
h
est in
f
o
r
m
atio
n
g
ain
is
s
elec
ted
as th
e
n
o
d
e
f
o
r
th
e
cu
r
r
en
t sp
lit.
3
.
2
.
E
nerg
y
c
o
ns
um
ptio
n
T
h
e
to
tal
en
er
g
y
c
o
n
s
u
m
p
ti
o
n
ca
n
b
e
e
x
p
r
ess
ed
as
th
e
s
u
m
o
f
th
e
en
er
g
y
co
n
s
u
m
e
d
b
y
e
ac
h
task
ac
r
o
s
s
all
s
tep
s
.
=
∑
∑
,
=
1
=
1
(
3
)
wh
er
e
,
is
th
e
am
o
u
n
t
o
f
en
er
g
y
u
s
ed
b
y
task
at
tim
e
s
tep
,
an
d
is
th
e
to
tal
n
u
m
b
er
o
f
task
s
.
T
h
is
co
m
p
u
tatio
n
p
r
o
v
id
es
a
n
ac
c
u
r
ate
ass
ess
m
en
t
o
f
th
e
s
y
s
tem
'
s
to
tal
en
er
g
y
ef
f
icien
cy
b
y
ac
co
u
n
tin
g
f
o
r
v
ar
iab
les
in
clu
d
i
n
g
task
ex
ec
u
tio
n
tim
e,
d
ev
ice
p
o
wer
co
n
s
u
m
p
tio
n
r
ates,
an
d
an
y
a
d
d
itio
n
al
o
v
er
h
ea
d
r
elate
d
to
task
ex
ec
u
tio
n
[
1
8
]
.
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
-
8
7
0
8
E
n
h
a
n
cin
g
r
eso
u
r
ce
ma
n
a
g
e
men
t in
fo
g
-
clo
u
d
in
tern
et
o
f th
in
g
s
…
(
V
ija
ya
l
a
ksh
mi
V
en
k
a
tesa
n
)
7249
W
ith
in
th
e
r
ea
lm
o
f
f
o
g
c
o
m
p
u
tin
g
,
th
e
task
s
ch
ed
u
lin
g
d
ilem
m
a
r
ev
o
lv
es
ar
o
u
n
d
ass
ig
n
in
g
I
o
T
task
s
to
ap
p
r
o
p
r
iate
f
o
g
n
o
d
es
f
r
o
m
a
p
o
o
l
o
f
p
o
ten
tial
ca
n
d
id
ates
to
en
h
an
ce
Qo
S.
T
h
is
r
esear
ch
s
p
ec
if
ically
co
n
s
id
er
s
Qo
S
p
ar
am
ete
r
s
s
u
ch
as
laten
cy
,
e
n
er
g
y
co
n
s
u
m
p
t
io
n
,
an
d
n
etwo
r
k
u
tili
za
tio
n
.
T
h
e
to
tal
n
u
m
b
er
o
f
task
s
is
d
en
o
ted
b
y
=
{
1
,
2
,
…
,
}
an
d
f
o
g
n
o
d
es a
s
=
{
1
,
2
,
…
,
}
.
T
h
e
o
v
e
r
all
tr
an
s
m
is
s
io
n
d
u
r
at
io
n
o
f
task
c
o
n
s
id
er
in
g
b
o
t
h
f
o
g
an
d
cl
o
u
d
n
o
d
es is
g
iv
en
b
y
(
4
)
.
,
=
∗
(
)
+
∗
(
)
(
4
)
wh
er
e
(
)
is
th
e
tim
e
tak
en
to
tr
an
s
m
it
a
task
to
a
f
o
g
n
o
d
e
a
n
d
(
)
is
th
e
tim
e
tak
en
to
tr
an
s
m
it
a
task
to
a
clo
u
d
n
o
d
e
.
an
d
ar
e
th
e
p
r
o
b
ab
ilit
y
o
f
ta
s
k
s
b
ein
g
tr
an
s
m
itted
to
f
o
g
a
n
d
clo
u
d
.
T
h
e
tim
e
tak
en
t
o
ex
ec
u
te
task
o
n
f
o
g
n
o
d
e
an
d
t
o
th
e
clo
u
d
is
d
eter
m
in
ed
as
(
5
)
,
(
6
)
.
(
)
=
(
)
+
(
)
(
5
)
(
)
=
(
)
+
(
)
(
6
)
T
h
e
task
’
s
r
esp
o
n
s
e
tim
e
o
f
t
h
e
task
is
f
u
r
th
er
p
r
o
ce
s
s
ed
ac
r
o
s
s
s
ev
er
al
lay
er
s
,
in
clu
d
in
g
t
h
e
ex
ec
u
tio
n
tim
e
o
f
th
e
r
ec
eiv
er
n
o
d
e
an
d
th
e
tim
e
tak
en
b
y
th
e
task
to
b
e
t
r
an
s
m
itted
f
r
o
m
th
e
s
en
d
er
n
o
d
e
to
t
h
e
r
ec
eiv
e
r
n
o
d
e.
T
h
e
f
o
r
m
u
la
(
7
)
a
n
d
(
8
)
is
em
p
lo
y
ed
to
ca
lc
u
late
th
e
r
e
s
p
o
n
s
e
tim
e
o
f
task
h
an
d
led
a
t f
o
g
n
o
d
e
.
(
)
=
(
)
+
(
)
(
7
)
(
)
=
(
)
+
(
)
(
8
)
T
o
ca
lcu
late
th
e
en
er
g
y
n
ee
d
ed
to
tr
an
s
m
it
task
t
i
to
f
o
g
n
o
d
e
f
j
,
t
h
e
tr
an
s
m
is
s
io
n
tim
e
is
m
u
ltip
lied
b
y
a
co
n
s
tan
t c
o
ef
f
icien
t a
s
(
9
)
an
d
(
1
0
)
:
(
)
=
∗
(
)
∗
(
1
+
µ
)
(
9
)
(
)
=
∗
(
)
∗
(
1
+
µ
)
(
1
0
)
wh
er
e
(
)
is
th
e
t
o
tal
en
er
g
y
c
o
n
s
u
m
p
tio
n
f
o
r
ex
ec
u
tin
g
task
o
n
a
f
o
g
n
o
d
e.
(
)
t
o
tal
en
er
g
y
c
o
n
s
u
m
p
tio
n
f
o
r
ex
ec
u
tin
g
task
o
n
a
clo
u
d
n
o
d
e.
λ
is
th
e
t
r
an
s
m
is
s
io
n
en
er
g
y
co
ef
f
icien
t
,
r
ep
r
esen
tin
g
th
e
en
er
g
y
r
eq
u
ir
ed
p
er
u
n
it
o
f
d
ata
tr
an
s
m
is
s
io
n
.
(
)
t
o
tal
tr
an
s
m
is
s
io
n
tim
e
f
o
r
task
j
o
n
a
f
o
g
n
o
d
e
(
)
to
tal
tr
an
s
m
is
s
io
n
tim
e
f
o
r
ex
ec
u
tin
g
task
o
n
a
clo
u
d
n
o
d
e.
is
th
e
p
r
o
ce
s
s
in
g
en
er
g
y
co
e
f
f
icie
n
t,
r
ep
r
esen
tin
g
th
e
en
e
r
g
y
r
eq
u
ir
ed
f
o
r
p
r
o
ce
s
s
in
g
ta
s
k
s
r
elativ
e
to
th
e
tr
an
s
m
is
s
io
n
en
er
g
y
.
4.
SI
M
UL
A
T
I
O
N
A
ND
RE
SU
L
T
Simu
latio
n
s
ar
e
ca
r
r
ie
d
o
u
t
in
th
is
s
ec
tio
n
to
en
s
u
r
e
th
at
th
e
s
u
g
g
ested
alg
o
r
ith
m
wo
r
k
s
p
r
o
p
er
ly
.
Fo
r
im
p
lem
en
tin
g
th
e
MM
FM
u
s
in
g
DDPG
in
a
f
o
g
co
m
p
u
tin
g
en
v
ir
o
n
m
en
t,
s
ev
er
al
s
im
u
lat
io
n
p
ar
am
ete
r
s
an
d
s
etu
p
co
n
f
ig
u
r
atio
n
s
ar
e
co
n
s
i
d
er
ed
.
T
h
e
s
im
u
latio
n
e
n
v
ir
o
n
m
en
t
was
d
esig
n
ed
to
em
u
la
te
a
f
o
g
-
clo
u
d
I
o
T
s
y
s
tem
[
1
9
]
,
in
co
r
p
o
r
atin
g
m
u
ltip
le
f
o
g
n
o
d
es
an
d
clo
u
d
s
er
v
er
s
.
Key
p
ar
am
eter
s
in
clu
d
ed
th
e
n
u
m
b
er
o
f
f
o
g
n
o
d
es,
wh
ic
h
wer
e
v
ar
ied
b
et
wee
n
5
t
o
2
0
,
a
n
d
clo
u
d
s
er
v
er
s
s
et
to
3
f
o
r
m
a
n
ag
in
g
o
f
f
lo
ad
ed
task
s
.
T
ask
g
en
er
atio
n
was
s
im
u
lated
u
s
in
g
a
Po
is
s
o
n
p
r
o
ce
s
s
to
m
im
i
c
r
ea
l
-
tim
e
task
ar
r
i
v
als,
with
task
attr
ib
u
tes
s
u
ch
as
p
r
io
r
ity
,
Qo
S
d
em
a
n
d
s
,
an
d
co
m
p
u
tatio
n
al
r
eq
u
ir
e
m
en
ts
b
ein
g
r
an
d
o
m
ly
ass
ig
n
ed
b
a
s
ed
o
n
p
r
ed
e
f
in
e
d
d
is
tr
ib
u
tio
n
s
[
2
0
]
,
[
2
1
]
.
Netwo
r
k
b
a
n
d
wid
th
an
d
laten
cy
wer
e
c
o
n
f
i
g
u
r
ed
to
r
ef
lect
r
ea
lis
tic
co
n
d
itio
n
s
,
with
av
er
ag
e
laten
cy
r
an
g
in
g
b
etwe
en
5
to
2
0
m
s
f
o
r
f
o
g
n
o
d
es
an
d
h
ig
h
er
laten
cies
f
o
r
clo
u
d
in
ter
ac
tio
n
s
.
T
h
e
DDPG
ag
en
t
was
tr
ain
ed
o
v
er
1
0
,
0
0
0
ep
is
o
d
es
,
with
ea
ch
ep
is
o
d
e
c
o
n
s
is
tin
g
o
f
5
0
task
s
ch
ed
u
lin
g
iter
a
tio
n
s
.
T
h
e
r
ewa
r
d
f
u
n
ctio
n
was
d
esig
n
e
d
to
b
al
an
ce
en
er
g
y
ef
f
icien
c
y
an
d
t
ask
co
m
p
letio
n
tim
e,
e
n
co
u
r
ag
in
g
o
p
tim
al
task
allo
ca
tio
n
b
etwe
en
f
o
g
n
o
d
es
an
d
cl
o
u
d
s
er
v
er
s
.
Hy
p
er
p
ar
a
m
eter
s
f
o
r
th
e
DDPG
m
o
d
el
i
n
clu
d
ed
a
lear
n
in
g
r
ate
o
f
0
.
0
0
1
,
a
d
is
co
u
n
t
f
ac
to
r
o
f
0
.
9
9
,
an
d
a
b
atch
s
ize
o
f
6
4
.
T
h
e
s
im
u
latio
n
was
ex
ec
u
ted
u
s
in
g
a
co
m
b
in
atio
n
o
f
Py
th
o
n
lib
r
a
r
ies,
s
u
ch
as
T
en
s
o
r
Flo
w
f
o
r
th
e
DDPG
im
p
lem
en
tatio
n
an
d
SimPy
f
o
r
th
e
d
is
cr
ete
-
ev
en
t
s
im
u
latio
n
o
f
th
e
f
o
g
co
m
p
u
tin
g
en
v
ir
o
n
m
en
t
[
2
2
]
.
T
h
is
s
etu
p
e
n
s
u
r
ed
a
co
m
p
r
eh
e
n
s
iv
e
ev
alu
atio
n
o
f
t
h
e
MM
FM'
s
p
er
f
o
r
m
a
n
ce
in
d
y
n
am
ically
a
d
ju
s
tin
g
task
allo
ca
tio
n
s
t
o
m
ee
t
Qo
S
r
eq
u
ir
em
en
ts
an
d
en
h
an
ce
en
e
r
g
y
e
f
f
icien
cy
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
e
r
20
24
:
7
2
4
4
-
7
2
5
3
7250
T
h
e
p
er
f
o
r
m
an
ce
an
al
y
s
is
o
f
alg
o
r
ith
m
s
an
d
m
o
d
els
in
v
o
lv
es
ev
alu
atin
g
th
eir
ac
cu
r
ac
y
,
an
d
co
m
p
u
tatio
n
al
e
f
f
icien
cy
.
Acc
u
r
ac
y
m
ea
s
u
r
es
th
e
co
r
r
ec
tn
es
s
o
f
p
r
ed
ictio
n
s
,
wh
ile
co
m
p
u
tatio
n
al
ef
f
icien
cy
ass
es
s
es
r
eso
u
r
ce
u
s
ag
e.
Gen
er
aliza
tio
n
ex
am
in
es
h
o
w
well
m
o
d
els
ap
p
ly
to
u
n
s
ee
n
d
ata
[
2
3
]
,
[
2
4
]
.
C
o
m
p
ar
ativ
e
a
n
aly
s
is
ag
ain
s
t
b
aselin
es
an
d
s
tate
-
of
-
th
e
-
ar
t
m
eth
o
d
s
b
en
c
h
m
ar
k
s
p
er
f
o
r
m
an
ce
.
E
v
alu
atin
g
th
ese
asp
ec
ts
en
s
u
r
es
m
o
d
el
s
m
ee
t
th
e
d
esire
d
cr
iter
ia
f
o
r
r
ea
l
-
wo
r
ld
a
p
p
licatio
n
s
,
b
alan
cin
g
p
r
ed
ictiv
e
p
o
wer
,
r
eso
u
r
ce
co
n
s
u
m
p
tio
n
,
an
d
in
ter
p
r
etab
ilit
y
.
T
h
is
an
aly
s
is
g
u
id
es
iter
ativ
e
r
ef
in
em
en
t,
en
h
a
n
cin
g
m
o
d
el
r
eliab
ilit
y
,
ef
f
icien
cy
,
a
n
d
ef
f
ec
tiv
e
n
ess
in
s
o
lv
in
g
tar
g
et
p
r
o
b
lem
s
[
2
5
]
.
MM
FM
h
as
estab
lis
h
ed
its
elf
as
a
to
p
p
e
r
f
o
r
m
er
am
o
n
g
p
r
e
d
ictiv
e
alg
o
r
ith
m
s
,
o
f
f
e
r
in
g
cl
o
s
e
to
th
e
b
est
s
o
lu
tio
n
f
o
r
th
e
p
r
o
b
lem
with
an
o
u
t
p
u
t
clo
s
e
t
o
0
.
8
9
.
T
h
e
d
if
f
e
r
en
ce
i
n
th
e
ac
h
ie
v
ed
s
co
r
e
r
is
es
to
a
n
o
tab
ly
m
o
r
e
im
p
r
ess
iv
e
lev
el
co
m
p
ar
ed
to
r
esu
lts
o
b
tai
n
ed
b
y
o
th
e
r
p
r
o
m
in
en
t
r
ein
f
o
r
ce
m
en
t
lear
n
i
n
g
alg
o
r
ith
m
s
,
n
am
ely
r
an
d
o
m
f
o
r
est
(
RF
)
,
s
u
p
p
o
r
t
v
ec
t
o
r
m
ac
h
in
e
(
SVM
)
,
a
n
d
k
-
n
ea
r
est
n
eig
h
b
o
r
s
(
KNN
)
.
T
h
e
ev
alu
atio
n
p
r
o
ce
s
s
was
c
o
n
d
u
cted
in
a
d
etailed
m
a
n
n
er
,
b
y
co
llectin
g
s
ev
er
al
m
etr
ics
o
f
p
er
f
o
r
m
an
ce
wh
ich
in
clu
d
ed
F1
s
co
r
es,
R
MSE
,
an
d
MSE
.
T
h
r
o
u
g
h
th
is
,
it wa
s
d
em
o
n
s
tr
ated
th
at
MM
FM
n
o
t o
n
ly
h
ad
v
e
r
y
g
o
o
d
ac
cu
r
ac
y
a
n
d
p
r
ed
ictio
n
b
u
t
also
s
h
o
wed
a
h
ig
h
v
ar
iab
le
o
f
co
n
s
is
ten
cy
an
d
r
eliab
ilit
y
in
ca
s
e
th
e
v
ar
ied
s
ce
n
ar
io
s
an
d
d
atasets
wer
e
co
n
s
id
er
ed
.
T
h
e
h
o
r
iz
o
n
tal
b
ar
g
r
ap
h
illu
s
tr
ated
in
Fig
u
r
e
3
s
h
o
ws
th
e
ac
cu
r
ac
y
co
m
p
ar
is
o
n
am
o
n
g
f
o
u
r
m
o
d
els:
R
F,
SVM,
K
NN,
an
d
MM
FM.
T
h
e
x
-
ax
is
o
f
th
e
g
r
a
p
h
r
ep
r
esen
ts
th
e
ac
cu
r
ac
y
le
v
el,
p
r
o
p
o
s
ed
MM
FM
p
er
f
o
r
m
s
b
etter
co
m
p
a
r
ed
t
o
o
th
er
e
x
is
tin
g
m
o
d
els.
T
h
e
a
r
ea
ch
ar
t
as
illu
s
tr
ated
in
Fig
u
r
e
4
s
ig
n
if
ies
th
e
en
er
g
y
co
n
s
u
m
p
tio
n
o
f
f
o
u
r
m
o
d
els
wh
ich
in
clu
d
es
th
e
R
F,
SVM,
KNN
an
d
th
e
p
r
o
p
o
s
ed
MM
FM
in
th
e
x
-
ax
is
.
I
n
t
h
e
y
-
a
x
is
r
ep
r
esen
tin
g
en
er
g
y
co
n
s
u
m
p
tio
n
T
h
e
SVM
is
th
e
m
o
s
t
en
er
g
etic
m
o
d
el
co
n
s
u
m
in
g
a
q
u
an
tity
o
f
a
b
o
u
t
8
0
k
W
h
,
w
h
er
ea
s
th
e
KNN
m
o
d
el
is
th
e
s
ec
o
n
d
m
o
s
t
en
er
g
etic
co
n
s
u
m
in
g
a
q
u
a
n
tity
o
f
ab
o
u
t 7
5
k
W
h
.
Fig
u
r
e
3
.
Sh
o
ws th
e
a
cc
u
r
ac
y
co
m
p
ar
is
o
n
Fig
u
r
e
4
.
E
n
er
g
y
c
o
n
s
u
m
p
tio
n
f
o
r
d
i
f
f
er
en
t
m
o
d
els
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
-
8
7
0
8
E
n
h
a
n
cin
g
r
eso
u
r
ce
ma
n
a
g
e
men
t in
fo
g
-
clo
u
d
in
tern
et
o
f th
in
g
s
…
(
V
ija
ya
l
a
ksh
mi
V
en
k
a
tesa
n
)
7251
MM
FM,
o
n
th
e
o
th
e
r
h
an
d
,
u
s
es
less
th
an
5
0
u
n
its
o
f
en
e
r
g
y
r
esu
ltin
g
in
m
o
n
u
m
en
tal
e
f
f
icien
cy
.
T
o
ass
ess
th
e
ch
an
g
es
in
th
e
to
tal
en
er
g
y
co
n
s
u
m
p
tio
n
f
o
r
v
ar
io
u
s
s
izes
o
f
task
s
u
n
d
er
v
ar
io
u
s
cir
cu
m
s
tan
ce
s
,
as
well
as
to
co
m
p
ar
e
th
e
e
x
p
en
d
itu
r
es
with
th
e
p
r
ev
io
u
s
m
eth
o
d
s
,
Fig
u
r
e
5
is
p
r
o
v
i
d
ed
.
Ou
r
p
r
o
p
o
s
ed
alg
o
r
ith
m
’
s
en
er
g
y
co
n
s
u
m
p
ti
o
n
is
alwa
y
s
les
s
th
an
th
at
in
tr
ad
itio
n
al
lo
ca
l
co
m
p
u
tin
g
as
well
a
s
th
e
f
u
ll
o
f
f
lo
ad
i
n
g
.
I
n
Fig
u
r
e
6
,
th
e
a
d
v
en
tu
r
e
c
o
n
s
is
ts
o
f
an
ac
cu
m
u
latio
n
o
f
r
ewa
r
d
s
b
y
s
h
o
wi
n
g
th
e
f
o
u
r
clim
b
in
g
p
ass
ag
es.
T
h
e
x
-
ax
is
s
tan
d
s
f
o
r
th
e
ep
is
o
d
es
h
av
in
g
0
-
3
v
alu
es,
wh
ile
th
e
y
-
ax
is
d
em
o
n
s
tr
ates
th
e
d
esire
d
r
ewa
r
d
f
r
o
m
0
to
1
5
.
T
h
e
lin
e
in
g
r
ap
h
p
atter
n
is
aim
ed
at
d
em
o
n
s
tr
atin
g
a
p
o
s
itiv
e
c
o
r
r
elatio
n
,
wh
ich
is
an
in
cr
ea
s
e
in
to
tal
r
ewa
r
d
s
as
th
e
ep
is
o
d
es
g
et
f
o
r
war
d
.
Firstl
y
,
th
e
r
ewa
r
d
m
ay
s
tar
t
f
r
o
m
ab
o
u
t
3
a
n
d
q
u
ick
ly
in
cr
ea
s
e
to
ap
p
r
o
x
im
ately
1
5
as
th
e
ep
is
o
d
es
m
o
v
e
to
f
o
u
r
.
T
h
er
ef
o
r
e,
th
is
k
in
d
o
f
g
r
o
wt
h
s
h
o
ws
th
e
s
y
s
tem
p
er
f
o
r
m
an
ce
o
f
th
e
m
o
d
el
to
b
e
u
p
g
r
a
d
ed
o
v
er
tim
e
w
h
ich
s
h
o
u
ld
b
r
in
g
h
ig
h
e
r
r
ewa
r
d
s
o
v
er
ep
is
o
d
es.
Fig
u
r
e
6
.
T
o
tal
r
ewa
r
d
s
p
e
r
ep
is
o
d
e
Fig
u
r
e
5
.
T
o
tal
en
er
g
y
co
n
s
u
m
p
tio
n
with
d
if
f
er
en
t ta
s
k
s
ize
s
5.
CO
NCLU
SI
O
N
T
h
e
m
ain
g
o
al
is
to
p
r
eser
v
e
l
o
ca
l
en
er
g
y
r
eso
u
r
ce
s
b
y
p
r
io
r
itizin
g
cr
itical
task
s
f
o
r
m
ig
r
at
io
n
to
th
e
clo
u
d
.
T
h
e
im
p
lem
e
n
tatio
n
o
f
th
e
PTCA
en
ab
led
ef
f
icien
t
task
tr
an
s
f
er
to
th
e
Am
az
o
n
W
eb
Ser
v
ices
(
AW
S)
c
lo
u
d
in
f
r
astru
ctu
r
e,
lead
in
g
to
ef
f
ec
tiv
e
r
eso
u
r
ce
d
is
tr
ib
u
tio
n
a
n
d
o
p
tim
izatio
n
.
Fu
r
th
er
m
o
r
e
,
th
e
tas
k
s
ch
ed
u
ler
,
em
p
o
wer
e
d
b
y
th
e
p
r
ed
ictiv
e
f
e
atu
r
es
o
f
th
e
MM
FM,
ad
ep
tly
o
r
ch
estra
ted
task
ass
ig
n
m
en
ts
b
ased
o
n
f
o
r
th
c
o
m
in
g
p
r
ed
ictio
n
s
.
T
h
is
h
o
lis
tic
s
tr
ateg
y
en
s
u
r
ed
th
at
v
ital
task
s
wer
e
p
r
io
r
itized
,
m
ax
im
izin
g
lo
ca
l
en
er
g
y
co
n
s
er
v
atio
n
wh
ile
s
u
s
tain
in
g
s
y
s
tem
ef
f
icien
cy
.
T
h
r
o
u
g
h
o
u
t
th
e
s
tu
d
y
,
th
e
e
f
f
ec
tiv
e
task
m
ig
r
atio
n
to
th
e
AW
S
clo
u
d
v
ia
th
e
PTCM
alg
o
r
ith
m
h
ig
h
lig
h
te
d
th
e
v
iab
ilit
y
an
d
ef
f
icien
cy
o
f
f
o
g
-
clo
u
d
-
b
ased
ap
p
r
o
ac
h
es
f
o
r
r
eso
u
r
ce
-
d
e
m
an
d
in
g
ap
p
licatio
n
s
.
B
y
t
ap
p
in
g
i
n
to
th
e
s
ca
lab
ilit
y
an
d
co
m
p
u
tatio
n
al
ca
p
ab
ilit
ies
o
f
th
e
clo
u
d
,
th
e
s
y
s
tem
ex
h
ib
ited
im
p
r
o
v
ed
ef
f
icien
cy
an
d
ad
ap
tab
ilit
y
i
n
m
an
ag
in
g
v
ar
ied
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
e
r
20
24
:
7
2
4
4
-
7
2
5
3
7252
wo
r
k
lo
ad
s
.
Ad
d
itio
n
all
y
,
in
co
r
p
o
r
atin
g
th
e
MM
FM
m
o
d
el
in
to
th
e
task
-
s
ch
ed
u
lin
g
f
r
am
e
wo
r
k
was
p
iv
o
tal
in
r
ef
in
in
g
r
eso
u
r
ce
m
an
a
g
em
e
n
t
an
d
b
o
o
s
tin
g
s
y
s
tem
ef
f
i
cien
cy
.
T
h
r
o
u
g
h
p
r
ed
ictiv
e
an
aly
tics
,
th
e
task
s
ch
ed
u
ler
was
ab
le
to
f
o
r
esee
f
u
tu
r
e
wo
r
k
l
o
ad
r
eq
u
ir
em
e
n
ts
,
f
ac
ilit
atin
g
p
r
o
ac
tiv
e
d
e
cisi
o
n
-
m
ak
in
g
an
d
o
p
tim
al
r
eso
u
r
ce
d
is
tr
ib
u
tio
n
.
T
h
is
m
eth
o
d
o
l
o
g
y
n
o
t
o
n
ly
e
n
h
an
ce
d
task
co
m
p
letio
n
r
ates
b
u
t
also
p
r
o
m
o
te
d
en
er
g
y
c
o
n
s
er
v
atio
n
b
y
m
ig
r
at
in
g
n
o
n
-
ess
en
tial ta
s
k
s
d
u
r
in
g
p
ea
k
d
em
a
n
d
p
er
io
d
s
.
RE
F
E
R
E
NC
E
S
[
1
]
R
.
B
u
y
y
a
,
“
M
o
d
e
l
l
i
n
g
a
n
d
s
i
m
u
l
a
t
i
o
n
o
f
f
o
g
a
n
d
e
d
g
e
c
o
m
p
u
t
i
n
g
e
n
v
i
r
o
n
me
n
t
s
u
si
n
g
i
F
o
g
S
i
m
t
o
o
l
k
i
t
,
”
i
n
F
o
g
a
n
d
e
d
g
e
c
o
m
p
u
t
i
n
g
:
Pri
n
c
i
p
l
e
s
a
n
d
p
a
ra
d
i
g
m
s
,
n
o
.
A
p
r
i
l
,
2
0
1
8
,
p
p
.
1
–
3
5
.
[
2
]
A
.
G
.
J
a
k
w
a
,
A
.
Y
.
G
i
t
a
l
,
S
.
B
o
u
k
a
r
i
,
a
n
d
F
.
U
.
Za
mb
u
k
,
“
P
e
r
f
o
r
m
a
n
c
e
e
v
a
l
u
a
t
i
o
n
o
f
h
y
b
r
i
d
me
t
a
-
h
e
u
r
i
st
i
c
s
-
b
a
se
d
t
a
s
k
sch
e
d
u
l
i
n
g
a
l
g
o
r
i
t
h
m
f
o
r
e
n
e
r
g
y
e
f
f
i
c
i
e
n
c
y
i
n
f
o
g
c
o
mp
u
t
i
n
g
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
C
l
o
u
d
A
p
p
l
i
c
a
t
i
o
n
s
a
n
d
C
o
m
p
u
t
i
n
g
,
v
o
l
.
1
3
,
n
o
.
1
,
2
0
2
3
,
d
o
i
:
1
0
.
4
0
1
8
/
I
JC
A
C
.
3
2
4
7
5
8
.
[
3
]
V
.
S
i
n
d
h
u
,
M
.
P
r
a
k
a
s
h
,
a
n
d
P
.
M
o
h
a
n
K
u
mar
,
“
E
n
e
r
g
y
-
e
f
f
i
c
i
e
n
t
t
a
sk
s
c
h
e
d
u
l
i
n
g
a
n
d
r
e
s
o
u
r
c
e
a
l
l
o
c
a
t
i
o
n
f
o
r
i
mp
r
o
v
i
n
g
t
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
a
c
l
o
u
d
–
f
o
g
e
n
v
i
r
o
n
m
e
n
t
,
”
S
y
m
m
e
t
r
y
,
v
o
l
.
1
4
,
n
o
.
1
1
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
sy
m1
4
1
1
2
3
4
0
.
[
4
]
M
.
H
u
ss
a
i
n
,
L.
F
.
W
e
i
,
A
.
La
k
h
a
n
,
S
.
W
a
l
i
,
S
.
A
l
i
,
a
n
d
A
.
H
u
ss
a
i
n
,
“
En
e
r
g
y
a
n
d
p
e
r
f
o
r
m
a
n
c
e
-
e
f
f
i
c
i
e
n
t
t
a
sk
s
c
h
e
d
u
l
i
n
g
i
n
h
e
t
e
r
o
g
e
n
e
o
u
s
v
i
r
t
u
a
l
i
z
e
d
c
l
o
u
d
c
o
m
p
u
t
i
n
g
,
”
S
u
s
t
a
i
n
a
b
l
e
C
o
m
p
u
t
i
n
g
:
I
n
f
o
r
m
a
t
i
c
s
a
n
d
S
y
st
e
m
s
,
2
0
2
1
.
[
5
]
P
.
S
i
n
g
h
,
M
.
D
u
t
t
a
,
a
n
d
N
.
A
g
g
a
r
w
a
l
,
“
A
r
e
v
i
e
w
o
f
t
a
s
k
s
c
h
e
d
u
l
i
n
g
b
a
s
e
d
o
n
me
t
a
-
h
e
u
r
i
s
t
i
c
s
a
p
p
r
o
a
c
h
i
n
c
l
o
u
d
c
o
m
p
u
t
i
n
g
,
”
K
n
o
w
l
e
d
g
e
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
y
st
e
m
s
,
v
o
l
.
5
2
,
n
o
.
1
,
p
p
.
1
–
5
1
,
2
0
1
7
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
1
1
5
-
0
1
7
-
1
0
4
4
-
2.
[
6
]
Y
.
H
o
u
,
L
.
Li
u
,
Q
.
W
e
i
,
X
.
X
u
,
a
n
d
C
.
C
h
e
n
,
“
A
n
o
v
e
l
D
D
P
G
me
t
h
o
d
w
i
t
h
p
r
i
o
r
i
t
i
z
e
d
e
x
p
e
r
i
e
n
c
e
r
e
p
l
a
y
,
”
i
n
2
0
1
7
I
EE
E
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
S
y
st
e
m
s,
M
a
n
,
a
n
d
C
y
b
e
rn
e
t
i
c
s
,
S
MC
2
0
1
7
,
2
0
1
7
,
v
o
l
.
2
0
1
7
-
Ja
n
u
a
ry
,
p
p
.
3
1
6
–
3
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
S
M
C
.
2
0
1
7
.
8
1
2
2
6
2
2
.
[
7
]
Y
.
D
o
n
g
a
n
d
X
.
Z
o
u
,
“
M
o
b
i
l
e
r
o
b
o
t
p
a
t
h
p
l
a
n
n
i
n
g
b
a
s
e
d
o
n
i
m
p
r
o
v
e
d
D
D
P
G
r
e
i
n
f
o
r
c
e
m
e
n
t
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m,”
i
n
2
0
2
0
I
E
EE
1
1
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
S
o
f
t
w
a
re
E
n
g
i
n
e
e
ri
n
g
a
n
d
S
e
rvi
c
e
S
c
i
e
n
c
e
(
I
C
S
E
S
S
)
,
O
c
t
.
2
0
2
0
,
p
p
.
5
2
–
5
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
S
ESS
4
9
9
3
8
.
2
0
2
0
.
9
2
3
7
6
4
1
.
[
8
]
C
.
Q
i
u
,
Y
.
H
u
,
Y
.
C
h
e
n
,
a
n
d
B
.
Ze
n
g
,
“
D
e
e
p
d
e
t
e
r
mi
n
i
st
i
c
p
o
l
i
c
y
g
r
a
d
i
e
n
t
(
D
D
P
G
)
-
b
a
se
d
e
n
e
r
g
y
h
a
r
v
e
st
i
n
g
w
i
r
e
l
e
ss
c
o
mm
u
n
i
c
a
t
i
o
n
s
,
”
I
E
EE
I
n
t
e
rn
e
t
o
f
T
h
i
n
g
s J
o
u
rn
a
l
,
v
o
l
.
6
,
n
o
.
5
,
p
p
.
8
5
7
7
–
8
5
8
8
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
JI
O
T.
2
0
1
9
.
2
9
2
1
1
5
9
.
[
9
]
G
.
Li
e
t
a
l
.
,
“
E
n
e
r
g
y
e
f
f
i
c
i
e
n
t
d
a
t
a
c
o
l
l
e
c
t
i
o
n
i
n
l
a
r
g
e
-
sc
a
l
e
i
n
t
e
r
n
e
t
o
f
t
h
i
n
g
s v
i
a
c
o
m
p
u
t
a
t
i
o
n
o
f
f
l
o
a
d
i
n
g
,
”
I
E
EE
I
n
t
e
rn
e
t
o
f
T
h
i
n
g
s
J
o
u
rn
a
l
,
v
o
l
.
6
,
n
o
.
3
,
p
p
.
4
1
7
6
–
4
1
8
7
,
Ju
n
.
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
JI
O
T.
2
0
1
8
.
2
8
7
5
2
4
4
.
[
1
0
]
M
.
B
a
b
a
r
a
n
d
M
.
S
o
h
a
i
l
K
h
a
n
,
“
S
c
a
l
Ed
g
e
:
a
f
r
a
mew
o
r
k
f
o
r
sca
l
a
b
l
e
e
d
g
e
c
o
m
p
u
t
i
n
g
i
n
I
n
t
e
r
n
e
t
o
f
t
h
i
n
g
s
–
b
a
se
d
sm
a
r
t
sy
s
t
e
ms
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
D
i
s
t
ri
b
u
t
e
d
S
e
n
s
o
r
N
e
t
w
o
rks
,
v
o
l
.
1
7
,
n
o
.
7
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
7
7
/
1
5
5
0
1
4
7
7
2
1
1
0
3
5
3
3
2
.
[
1
1
]
Y
.
Y
u
,
J.
Ta
n
g
,
J.
H
u
a
n
g
,
X
.
Z
h
a
n
g
,
D
.
K
.
C
.
S
o
,
a
n
d
K
.
-
K
.
W
o
n
g
,
“
M
u
l
t
i
-
o
b
j
e
c
t
i
v
e
o
p
t
i
m
i
z
a
t
i
o
n
f
o
r
U
A
V
-
A
ssi
st
e
d
w
i
r
e
l
e
ss
p
o
w
e
r
e
d
I
o
T
n
e
t
w
o
r
k
s
b
a
se
d
o
n
e
x
t
e
n
d
e
d
D
D
P
G
a
l
g
o
r
i
t
h
m,”
I
EE
E
T
ra
n
s
a
c
t
i
o
n
s
o
n
C
o
m
m
u
n
i
c
a
t
i
o
n
s
,
v
o
l
.
6
9
,
n
o
.
9
,
p
p
.
6
3
6
1
–
6
3
7
4
,
S
e
p
.
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
O
M
M
.
2
0
2
1
.
3
0
8
9
4
7
6
.
[
1
2
]
J.
B
a
e
k
a
n
d
G
.
K
a
d
d
o
u
m,
“
H
e
t
e
r
o
g
e
n
e
o
u
s
t
a
sk
o
f
f
l
o
a
d
i
n
g
a
n
d
r
e
so
u
r
c
e
a
l
l
o
c
a
t
i
o
n
s
v
i
a
d
e
e
p
r
e
c
u
r
r
e
n
t
r
e
i
n
f
o
r
c
e
men
t
l
e
a
r
n
i
n
g
i
n
p
a
r
t
i
a
l
o
b
ser
v
a
b
l
e
m
u
l
t
i
f
o
g
n
e
t
w
o
r
k
s,”
I
EEE
I
n
t
e
rn
e
t
o
f
T
h
i
n
g
s
J
o
u
r
n
a
l
,
v
o
l
.
8
,
n
o
.
2
,
p
p
.
1
0
4
1
–
1
0
5
6
,
Jan
.
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
JI
O
T.
2
0
2
0
.
3
0
0
9
5
4
0
.
[
1
3
]
M
.
T
a
n
g
a
n
d
V
.
W
.
S
.
W
o
n
g
,
“
D
e
e
p
r
e
i
n
f
o
r
c
e
me
n
t
l
e
a
r
n
i
n
g
f
o
r
t
a
s
k
o
f
f
l
o
a
d
i
n
g
i
n
m
o
b
i
l
e
e
d
g
e
c
o
m
p
u
t
i
n
g
sy
st
e
ms,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
M
o
b
i
l
e
C
o
m
p
u
t
i
n
g
,
v
o
l
.
2
1
,
n
o
.
6
,
p
p
.
1
9
8
5
–
1
9
9
7
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
T
M
C
.
2
0
2
0
.
3
0
3
6
8
7
1
.
[
1
4
]
Q
.
F
a
n
,
J.
B
a
i
,
H
.
Z
h
a
n
g
,
Y
.
Y
i
,
a
n
d
L.
L
i
u
,
“
D
e
l
a
y
-
a
w
a
r
e
r
e
s
o
u
r
c
e
a
l
l
o
c
a
t
i
o
n
i
n
f
o
g
-
a
ssi
s
t
e
d
I
o
T
n
e
t
w
o
r
k
s t
h
r
o
u
g
h
r
e
i
n
f
o
r
c
e
me
n
t
l
e
a
r
n
i
n
g
,
”
I
EEE
I
n
t
e
r
n
e
t
o
f
T
h
i
n
g
s
J
o
u
rn
a
l
,
v
o
l
.
9
,
n
o
.
7
,
p
p
.
5
1
8
9
–
5
1
9
9
,
A
p
r
.
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
JI
O
T.
2
0
2
1
.
3
1
1
1
0
7
9
.
[
1
5
]
T.
Q
i
u
,
J.
C
h
i
,
X
.
Z
h
o
u
,
Z.
N
i
n
g
,
M
.
A
t
i
q
u
z
z
a
m
a
n
,
a
n
d
D
.
O
.
W
u
,
“
E
d
g
e
c
o
m
p
u
t
i
n
g
i
n
i
n
d
u
s
t
r
i
a
l
i
n
t
e
r
n
e
t
o
f
t
h
i
n
g
s
:
a
r
c
h
i
t
e
c
t
u
r
e
,
a
d
v
a
n
c
e
s
a
n
d
c
h
a
l
l
e
n
g
e
s
,
”
I
EE
E
C
o
m
m
u
n
i
c
a
t
i
o
n
s
S
u
r
v
e
y
s
a
n
d
T
u
t
o
ri
a
l
s
,
v
o
l
.
2
2
,
n
o
.
4
,
p
p
.
2
4
6
2
–
2
4
8
8
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
C
O
M
S
T.
2
0
2
0
.
3
0
0
9
1
0
3
.
[
1
6
]
P
.
Za
n
g
,
“
A
p
p
l
i
c
a
t
i
o
n
o
f
I
D
3
d
e
c
i
s
i
o
n
t
r
e
e
c
l
a
ssi
f
i
c
a
t
i
o
n
a
l
g
o
r
i
t
h
m
i
n
mat
h
e
m
a
t
i
c
a
l
d
a
t
a
a
n
a
l
y
s
i
s,”
i
n
2
0
2
3
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
I
n
t
e
g
r
a
t
e
d
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
S
y
st
e
m
s
(
I
C
I
I
C
S
)
,
N
o
v
.
2
0
2
3
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
I
I
C
S
5
9
9
9
3
.
2
0
2
3
.
1
0
4
2
1
1
3
7
.
[
1
7
]
X
.
H
e
,
Y
.
K
u
a
n
g
,
N
.
S
o
n
g
,
a
n
d
F
.
L
i
u
,
“
I
n
t
e
l
l
i
g
e
n
t
n
a
v
i
g
a
t
i
o
n
o
f
i
n
d
o
o
r
r
o
b
o
t
b
a
s
e
d
o
n
i
mp
r
o
v
e
d
D
D
P
G
a
l
g
o
r
i
t
h
m,
”
Ma
t
h
e
m
a
t
i
c
a
l
Pr
o
b
l
e
m
s i
n
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
2
0
2
3
,
p
p
.
1
–
1
1
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
5
5
/
2
0
2
3
/
6
5
4
4
0
2
9
.
[
1
8
]
K
.
A
l
a
t
o
u
n
,
K
.
M
a
t
r
o
u
k
,
M
.
A
.
M
o
h
a
mm
e
d
,
J.
N
e
d
o
ma,
R
.
M
a
r
t
i
n
e
k
,
a
n
d
P
.
Zmi
j
,
“
A
n
o
v
e
l
l
o
w
-
l
a
t
e
n
c
y
a
n
d
e
n
e
r
g
y
-
e
f
f
i
c
i
e
n
t
t
a
s
k
sch
e
d
u
l
i
n
g
f
r
a
mew
o
r
k
f
o
r
i
n
t
e
r
n
e
t
o
f
m
e
d
i
c
a
l
t
h
i
n
g
s
i
n
a
n
e
d
g
e
f
o
g
c
l
o
u
d
sy
s
t
e
m
,
”
S
e
n
s
o
rs
,
v
o
l
.
2
2
,
n
o
.
1
4
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
2
1
4
5
3
2
7
.
[
1
9
]
B
.
K
r
i
s
h
n
a
m
u
r
t
h
y
a
n
d
S
.
G
.
S
h
i
v
a
,
“
E
n
e
r
g
y
e
f
f
i
c
i
e
n
t
d
o
u
b
l
e
c
r
i
t
i
c
d
e
e
p
d
e
t
e
r
m
i
n
i
st
i
c
p
o
l
i
c
y
g
r
a
d
i
e
n
t
f
r
a
mew
o
r
k
f
o
r
f
o
g
c
o
m
p
u
t
i
n
g
,
”
i
n
2
0
2
2
I
E
EE
Wo
r
l
d
AI
I
o
T
C
o
n
g
r
e
ss (AI
I
o
T
)
,
J
u
n
.
2
0
2
2
,
p
p
.
5
2
1
–
5
2
7
,
d
o
i
:
1
0
.
1
1
0
9
/
A
I
I
o
T5
4
5
0
4
.
2
0
2
2
.
9
8
1
7
1
5
7
.
[
2
0
]
X
.
C
a
i
,
S
.
G
e
n
g
,
D
.
I
.
W
u
,
J.
C
a
i
,
a
n
d
J.
C
h
e
n
,
“
A
mu
l
t
i
c
l
o
u
d
-
mo
d
e
l
-
b
a
se
d
man
y
-
o
b
j
e
c
t
i
v
e
i
n
t
e
l
l
i
g
e
n
t
a
l
g
o
r
i
t
h
m
f
o
r
e
f
f
i
c
i
e
n
t
t
a
s
k
sch
e
d
u
l
i
n
g
i
n
i
n
t
e
r
n
e
t
o
f
t
h
i
n
g
s,”
I
EEE
I
n
t
e
r
n
e
t
o
f
T
h
i
n
g
s
J
o
u
r
n
a
l
,
v
o
l
.
8
,
n
o
.
1
2
,
p
p
.
9
6
4
5
–
9
6
5
3
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
JI
O
T.
2
0
2
0
.
3
0
4
0
0
1
9
.
[
2
1
]
B
.
J
a
mi
l
,
H
.
I
j
a
z
,
M
.
S
h
o
j
a
f
a
r
,
K
.
M
u
n
i
r
,
a
n
d
R
.
B
u
y
y
a
,
“
R
e
so
u
r
c
e
a
l
l
o
c
a
t
i
o
n
a
n
d
t
a
sk
s
c
h
e
d
u
l
i
n
g
i
n
f
o
g
c
o
mp
u
t
i
n
g
a
n
d
i
n
t
e
r
n
e
t
o
f
e
v
e
r
y
t
h
i
n
g
e
n
v
i
r
o
n
m
e
n
t
s:
a
t
a
x
o
n
o
m
y
,
r
e
v
i
e
w
,
a
n
d
f
u
t
u
r
e
d
i
r
e
c
t
i
o
n
s,
”
A
C
M
C
o
m
p
u
t
i
n
g
S
u
rv
e
y
s
,
v
o
l
.
5
4
,
n
o
.
1
1
s,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
4
5
/
3
5
1
3
0
0
2
.
[
2
2
]
S
.
C
h
e
n
,
X
.
G
e
,
Q
.
W
a
n
g
,
Y
.
M
i
a
o
,
a
n
d
X
.
R
u
a
n
,
“
D
D
P
G
-
b
a
se
d
i
n
t
e
l
l
i
g
e
n
t
r
e
c
h
a
r
g
e
a
b
l
e
f
o
g
c
o
mp
u
t
a
t
i
o
n
o
f
f
l
o
a
d
i
n
g
f
o
r
I
o
T,
”
Wi
re
l
e
ss
N
e
t
w
o
r
k
s
,
v
o
l
.
2
8
,
n
o
.
7
,
p
p
.
3
2
9
3
–
3
3
0
4
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
2
7
6
-
0
2
2
-
0
3
0
5
4
-
1.
[
2
3
]
S
.
T
u
l
i
e
t
a
l
.
,
“
A
I
a
u
g
m
e
n
t
e
d
e
d
g
e
a
n
d
f
o
g
c
o
mp
u
t
i
n
g
:
t
r
e
n
d
s
a
n
d
c
h
a
l
l
e
n
g
e
s,
”
J
o
u
r
n
a
l
o
f
N
e
t
w
o
rk
a
n
d
C
o
m
p
u
t
e
r
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
2
1
6
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
n
c
a
.
2
0
2
3
.
1
0
3
6
4
8
.
[
2
4
]
M
.
A
b
d
E
l
a
z
i
z
,
L
.
A
b
u
a
l
i
g
a
h
,
a
n
d
I
.
A
t
t
i
y
a
,
“
A
d
v
a
n
c
e
d
o
p
t
i
mi
z
a
t
i
o
n
t
e
c
h
n
i
q
u
e
f
o
r
sc
h
e
d
u
l
i
n
g
I
o
T
t
a
s
k
s
i
n
c
l
o
u
d
-
f
o
g
c
o
mp
u
t
i
n
g
e
n
v
i
r
o
n
m
e
n
t
s,
”
F
u
t
u
r
e
G
e
n
e
r
a
t
i
o
n
C
o
m
p
u
t
e
r
S
y
st
e
m
s
,
v
o
l
.
1
2
4
,
p
p
.
1
4
2
–
1
5
4
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
f
u
t
u
r
e
.
2
0
2
1
.
0
5
.
0
2
6
.
[
2
5
]
J.
B
a
e
k
,
G
.
K
a
d
d
o
u
m,
S
.
G
a
r
g
,
K
.
K
a
u
r
,
a
n
d
V
.
G
r
a
v
e
l
,
“
M
a
n
a
g
i
n
g
f
o
g
n
e
t
w
o
r
k
s
u
s
i
n
g
r
e
i
n
f
o
r
c
e
me
n
t
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
a
l
g
o
r
i
t
h
m
,
”
i
n
2
0
1
9
I
EEE
Wi
re
l
e
ss
C
o
m
m
u
n
i
c
a
t
i
o
n
s
a
n
d
N
e
t
w
o
rk
i
n
g
C
o
n
f
e
r
e
n
c
e
(
W
C
N
C
)
,
A
p
r
.
2
0
1
9
,
p
p
.
1
–
7
,
d
o
i
:
1
0
.
1
1
0
9
/
W
C
N
C
.
2
0
1
9
.
8
8
8
5
7
4
5
.
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
-
8
7
0
8
E
n
h
a
n
cin
g
r
eso
u
r
ce
ma
n
a
g
e
men
t in
fo
g
-
clo
u
d
in
tern
et
o
f th
in
g
s
…
(
V
ija
ya
l
a
ksh
mi
V
en
k
a
tesa
n
)
7253
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Vija
y
a
la
k
s
h
m
i
Ve
n
k
a
te
sa
n
wo
rk
i
n
g
a
s
a
ss
istan
t
p
ro
fe
ss
o
r
i
n
th
e
De
p
a
rtme
n
t
o
f
Ne
two
rk
in
g
a
n
d
C
o
m
m
u
n
ica
ti
o
n
s,
S
c
h
o
o
l
o
f
C
o
m
p
u
ti
n
g
,
C
o
l
leg
e
o
f
En
g
in
e
e
rin
g
a
n
d
Tec
h
n
o
l
o
g
y
,
S
RM
I
n
stit
u
te
o
f
S
c
ien
c
e
a
n
d
Tec
h
n
o
l
o
g
y
,
Ka
tt
a
n
k
u
lath
u
r.
S
h
e
h
a
s
re
c
e
iv
e
d
a
b
a
c
h
e
lo
r’s
d
e
g
re
e
in
c
o
m
p
u
ter
s
c
ien
c
e
a
n
d
e
n
g
i
n
e
e
rin
g
fro
m
A
n
n
a
Un
i
v
e
rsity
,
m
a
ste
r’s
in
in
fo
rm
a
ti
o
n
tec
h
n
o
lo
g
y
fr
o
m
Bh
a
ra
th
Un
iv
e
rsity
,
Ch
e
n
n
a
i.
No
w
sh
e
is
p
u
rsu
i
n
g
P
h
.
D
d
e
g
re
e
in
c
o
m
p
u
ter
sc
ien
c
e
a
n
d
e
n
g
in
e
e
r
in
g
at
S
RM
I
n
stit
u
te
o
f
S
c
ien
c
e
a
n
d
Tec
h
n
o
lo
g
y
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
v
i
jay
a
lv
@s
rm
ist.
e
d
u
.
i
n
.
S
a
r
a
v
a
n
a
n
Mu
r
u
g
a
n
wo
rk
i
n
g
a
s
a
p
r
o
fe
ss
o
r
in
th
e
De
p
a
rtme
n
t
o
f
Ne
two
rk
i
n
g
a
n
d
Co
m
m
u
n
ica
ti
o
n
s,
S
c
h
o
o
l
o
f
Co
m
p
u
ti
n
g
,
Co
ll
e
g
e
o
f
En
g
in
e
e
r
in
g
a
n
d
Tec
h
n
o
lo
g
y
,
S
R
M
In
stit
u
te o
f
S
c
ien
c
e
a
n
d
Tec
h
n
o
lo
g
y
,
Ka
tt
a
n
k
u
lath
u
r
,
Tam
il
Na
d
u
,
In
d
ia.
He
re
c
e
iv
e
d
t
h
e
M
.
E.
d
e
g
re
e
in
c
o
m
p
u
ter
s
c
ien
c
e
a
n
d
e
n
g
in
e
e
rin
g
fro
m
t
h
e
G
o
v
e
rn
m
e
n
t
Co
ll
e
g
e
o
f
Tec
h
n
o
lo
g
y
,
Co
imb
a
to
re
,
Tam
il
Na
d
u
,
In
d
ia,
a
n
d
th
e
P
h
.
D.
d
e
g
re
e
in
c
o
m
p
u
t
e
r
s
c
ien
c
e
a
n
d
e
n
g
in
e
e
rin
g
fro
m
P
e
riy
a
r
M
a
n
iam
m
a
i
Un
iv
e
rsity
,
T
h
a
n
ja
v
u
r,
Tam
il
Na
d
u
,
I
n
d
ia.
His
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
c
lo
u
d
c
o
m
p
u
ti
n
g
,
f
o
g
c
o
m
p
u
ti
n
g
,
c
o
m
p
u
ter
v
isi
o
n
i
n
e
d
g
e
c
o
m
p
u
t
in
g
a
n
d
c
lo
u
d
se
c
u
rit
y
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
sa
ra
v
a
n
m
7
@s
rm
ist.
e
d
u
.
i
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.