I
nte
rna
t
io
na
l J
o
urna
l o
f
P
o
w
er
E
lect
ro
nics
a
nd
Driv
e
Sy
s
t
e
m
(
I
J
P
E
DS)
Vo
l.
9
,
No
.
1
,
Ma
r
ch
2
0
1
8
,
p
p
.
80
~
88
I
SS
N:
2
0
8
8
-
8
6
9
4
,
DOI
: 1
0
.
1
1
5
9
1
/i
j
p
ed
s
.
v
9
.
i
1
.
p
p
80
-
88
80
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
co
r
e
.
co
m/
jo
u
r
n
a
ls
/in
d
ex
.
p
h
p
/
I
J
P
E
DS
A Nov
el App
ro
a
ch in
Sc
heduling
O
f
the
Real
-
Ti
m
e
Ta
sk
s In
H
ete
ro
g
eneo
us M
ulticore
Process
o
r w
ith
Fu
zz
y
Lo
g
ic
Techniqu
e
F
o
r M
icro
-
g
rid P
o
w
er Ma
na
g
em
e
nt
L
a
v
a
ny
a
Dha
nes
h
1
,
P.
M
urug
esa
n
2
1
S
a
th
y
a
b
a
m
a
Un
iv
e
rsit
y
,
Ch
e
n
n
a
i
,
In
d
ia
2
S
.
A
.
En
g
in
e
e
rin
g
Co
ll
e
g
e
,
Ch
e
n
n
a
i,
In
d
ia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l 1
0
,
2
0
1
7
R
ev
i
s
ed
A
u
g
2
8
,
2
0
1
7
A
cc
ep
ted
J
an
1
,
2
0
1
8
S
c
h
e
d
u
l
in
g
o
f
tas
k
s
b
a
s
e
d
o
n
re
a
l
ti
m
e
r
e
q
u
irem
e
n
t
is
a
m
a
jo
r
i
ss
u
e
in
th
e
h
e
tero
g
e
n
e
o
u
s
m
u
lt
ico
re
sy
ste
m
s
f
o
r
m
icro
-
g
rid
p
o
w
e
r
m
a
n
a
g
e
m
e
n
t.
He
tero
g
e
n
e
o
u
s
m
u
lt
ico
re
p
ro
c
e
ss
o
r
sc
h
e
d
u
les
th
e
se
rial
tas
k
s
in
th
e
h
ig
h
p
e
rf
o
r
m
a
n
c
e
c
o
re
a
n
d
p
a
ra
ll
e
l
t
a
sk
s
a
re
e
x
e
c
u
ted
o
n
th
e
lo
w
p
e
rf
o
r
m
a
n
c
e
c
o
re
s.
T
h
e
a
i
m
o
f
th
is
p
a
p
e
r
is
t
o
im
p
le
m
e
n
t
a
sc
h
e
d
u
li
n
g
a
lg
o
rit
h
m
b
a
se
d
o
n
f
u
z
z
y
lo
g
ic
f
o
r
h
e
tero
g
e
n
e
o
u
s
m
u
lt
ico
re
p
r
o
c
e
ss
o
r
f
o
r
e
ff
e
c
ti
v
e
m
icro
-
g
rid
a
p
p
li
c
a
ti
o
n
.
Re
a
l
–
ti
m
e
tas
k
s
g
e
n
e
ra
ll
y
h
a
v
e
d
iffere
n
t
e
x
e
c
u
ti
o
n
ti
m
e
a
n
d
d
e
a
d
li
n
e
.
T
h
e
m
a
in
id
e
a
is
t
o
u
se
tw
o
f
u
z
z
y
lo
g
ic
b
a
se
d
sc
h
e
d
u
li
n
g
a
lg
o
rit
h
m
,
f
irst
is
to
a
ss
ig
n
p
ri
o
ri
ty
b
a
se
d
o
n
e
x
e
c
u
ti
o
n
ti
m
e
a
n
d
d
e
a
d
li
n
e
o
f
th
e
tas
k
.
S
e
c
o
n
d
,
t
h
e
tas
k
w
h
ich
h
a
s
a
ss
ig
n
e
d
h
ig
h
e
r
p
rio
ri
ty
g
e
t
a
ll
o
tt
e
d
f
o
r
e
x
e
c
u
ti
o
n
i
n
h
ig
h
p
e
rf
o
rm
a
n
c
e
c
o
re
a
n
d
re
m
a
in
in
g
tas
k
s
w
h
ich
a
re
a
ss
ig
n
e
d
lo
w
p
rio
rit
y
g
e
t
a
ll
o
tt
e
d
in
l
o
w
p
e
rf
o
r
m
a
n
c
e
c
o
re
s.
T
h
e
m
a
in
o
b
jec
ti
v
e
o
f
th
is
sc
h
e
d
u
li
n
g
a
lg
o
rit
h
m
is
to
i
n
c
re
a
se
th
e
th
ro
u
g
h
p
u
t
a
n
d
to
im
p
ro
v
e
CP
U
u
ti
li
z
a
ti
o
n
th
e
re
b
y
re
d
u
c
in
g
th
e
o
v
e
ra
ll
p
o
w
e
r
c
o
n
su
m
p
ti
o
n
o
f
th
e
m
icro
-
g
rid
p
o
w
e
r
m
a
n
a
g
e
m
e
n
t
s
y
ste
m
s.
T
e
st
c
a
se
s
w
it
h
d
if
f
e
re
n
t
tas
k
e
x
e
c
u
ti
o
n
ti
m
e
a
n
d
d
e
a
d
l
in
e
w
e
re
g
e
n
e
ra
ted
to
e
v
a
lu
a
te
th
e
a
lg
o
rit
h
m
s
u
sin
g
M
AT
LAB
so
f
t
w
a
re
.
K
ey
w
o
r
d
:
Dea
d
lin
e
E
x
ec
u
t
io
n
ti
m
e
Fu
zz
y
lo
g
ic
Mic
r
o
-
g
r
id
Mu
ltico
r
e
P
r
io
r
ity
R
ea
l
–
T
im
e
ta
s
k
s
Sch
ed
u
l
in
g
Co
p
y
rig
h
t
©
2
0
1
8
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
L
a
v
an
y
a
D
h
an
e
s
h
,
R
esear
ch
Sch
o
lar
,
Sat
h
y
ab
a
m
a
Un
i
v
er
s
it
y
,
C
h
e
n
n
ai,
I
n
d
ia.
E
m
ail: la
v
an
y
ad
h
a
n
e
s
h
2
0
0
7
@
g
m
a
il.c
o
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
ex
p
lo
s
iv
e
g
r
o
w
t
h
o
f
n
et
w
o
r
k
b
an
d
w
id
t
h
an
d
I
n
ter
n
et
h
ig
h
tr
affic
ap
p
licatio
n
s
,
s
u
ch
as
w
e
b
b
r
o
w
s
i
n
g
,
o
n
li
n
e
s
ea
r
ch
i
n
g
,
v
id
eo
s
tr
ea
m
i
n
g
,
an
d
g
a
m
in
g
,
r
eq
u
ir
e
o
r
d
e
r
s
o
f
m
a
g
n
itu
d
e
i
n
cr
ea
s
e
in
p
ac
k
et
p
r
o
ce
s
s
in
g
t
h
r
o
u
g
h
p
u
t.
T
h
e
ad
v
en
t
o
f
co
m
m
o
d
it
y
m
u
ltico
r
e
p
la
tf
o
r
m
s
,
s
u
c
h
as
C
a
v
i
u
m
s
O
C
T
E
ON[
1
]
,
C
is
co
s
A
O
N
,
an
d
I
B
Ms
B
lad
e
C
en
t
er
[
2
,
3
]
,
h
as
o
p
en
ed
a
n
e
w
er
a
o
f
co
m
p
u
ti
n
g
f
o
r
n
et
w
o
r
k
a
p
p
licatio
n
s
to
tak
e
ad
v
an
ta
g
e
o
f
th
e
s
e
lo
w
c
o
s
t
m
ac
h
i
n
es
d
u
e
to
t
h
eir
s
u
p
er
io
r
it
y
in
p
er
f
o
r
m
a
n
c
e,
av
ailab
ilit
y
a
n
d
p
r
o
g
r
am
m
ab
ilit
y
.
Mo
r
e
an
d
m
o
r
e
n
et
w
o
r
k
p
ac
k
et
p
r
o
ce
s
s
i
n
g
s
y
s
te
m
s
h
av
e
b
ee
n
d
e
v
elo
p
ed
o
n
s
u
c
h
p
lat
f
o
r
m
s
r
an
g
i
n
g
f
r
o
m
g
en
er
al
-
p
u
r
p
o
s
e
p
r
o
ce
s
s
o
r
s
(
e.
g
.
,
I
n
tel’
s
Xe
o
n
an
d
A
M
D’
s
Op
ter
o
n
)
to
n
et
w
o
r
k
p
r
o
ce
s
s
o
r
s
(
e.
g
.
,
I
n
tel
’
s
I
XP
p
latf
o
r
m
)
a
n
d
p
r
o
g
r
a
m
m
ab
le
lo
g
ic
d
ev
i
ce
s
(
e.
g
.
,
Net
FP
GA
)
[
4
,
5
]
.
T
o
ex
p
lo
it
a
v
ailab
le
p
ar
allelis
m
f
o
r
b
etter
th
r
o
u
g
h
p
u
t,
n
et
w
o
r
k
ap
p
licatio
n
s
r
u
n
n
in
g
o
n
m
u
l
tico
r
e
p
latf
o
r
m
s
u
s
u
all
y
ta
k
e
o
n
e
o
f
th
e
f
o
llo
w
in
g
t
h
r
ee
f
o
r
m
s
:
Sp
atial
p
ar
allelis
m
,
w
h
er
e
m
u
lti
p
le
co
n
cu
r
r
en
t
p
ac
k
ets
ar
e
p
r
o
ce
s
s
ed
in
d
iffer
e
n
t
p
r
o
ce
s
s
o
r
s
in
d
ep
en
d
en
tl
y
.
T
y
p
ical
ex
a
m
p
les
ca
n
b
e
f
o
u
n
d
i
n
w
o
r
k
f
o
r
T
C
P
(
T
r
an
s
m
i
s
s
io
n
C
o
n
tr
o
l
P
r
o
to
co
l)
p
ar
allelis
m
,
s
ca
lab
le
DP
I
(
Dee
p
P
ac
k
et
I
n
s
p
ec
tio
n
)
d
es
ig
n
,
fl
o
w
le
v
el
p
ac
k
et
p
r
o
ce
s
s
i
n
g
]
a
n
d
p
ar
allel
m
u
lti
m
ed
ia
tr
an
s
co
d
in
g
.
T
e
m
p
o
r
al
p
ar
allelis
m
(
p
ip
eli
n
in
g
)
,
w
h
er
e
m
u
ltip
le
p
r
o
ce
s
s
o
r
s
ar
e
s
ch
ed
u
led
in
to
a
p
ip
elin
e
to
o
v
er
lap
p
er
io
d
ic
ex
ec
u
t
io
n
s
f
r
o
m
d
iffer
en
t
th
r
ea
d
s
.
I
t
h
as
b
ee
n
w
id
el
y
a
d
o
p
ted
in
n
et
w
o
r
k
p
r
o
ce
s
s
o
r
s
,
in
clu
d
i
n
g
S
h
a
n
g
r
i
-
L
a,
au
to
-
p
ar
titi
o
n
i
n
g
,
s
tatis
t
ical
ap
p
r
o
ac
h
an
d
Gr
ee
d
y
.
H
y
b
r
id
p
ar
allelis
m
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
er
E
lectr
o
n
&
Dr
i
S
y
s
t
I
SS
N:
2
0
8
8
-
8694
A
N
o
ve
l A
p
p
r
o
a
ch
in
S
c
h
ed
u
li
n
g
Of
th
e
R
ea
l
-
Time
Ta
s
ks I
n
Hete
r
o
g
en
eo
u
s
Mu
ltico
r
e…
(
L
a
va
n
ya
Dh
a
n
esh
)
81
in
te
g
r
ates
b
o
th
s
p
atial
a
n
d
tem
p
o
r
al
p
ar
allelis
m
to
b
en
e
fi
t
f
r
o
m
th
e
ad
v
a
n
ta
g
es
o
f
b
o
th
s
id
es.
I
t
f
o
r
m
s
a
p
ar
allel
p
ip
elin
e
co
r
e
to
p
o
l
o
g
y
,
w
h
er
e
ea
c
h
s
ta
g
e
co
n
tai
n
s
m
u
ltip
le
p
ar
allel
co
r
es,
s
u
c
h
a
s
R
a
n
d
o
m
an
d
B
ip
ar
[
6
,
7
]
.
T
r
a
d
itio
n
al
task
s
c
h
ed
u
li
n
g
s
c
h
e
m
e
s
,
s
u
c
h
as
lis
t
-
b
ased
s
c
h
e
d
u
lin
g
an
d
clu
s
ter
in
g
b
ased
s
ch
ed
u
li
n
g
,
ar
e
ca
p
ab
le
o
f
r
ed
u
cin
g
p
r
o
g
r
a
m
late
n
c
y
b
y
ex
p
lo
iti
n
g
fin
e
g
r
ai
n
ed
tas
k
-
lev
el
p
ar
all
elis
m
[
8
,
9
,
1
0
]
.
T
h
e
P
er
f
o
r
m
a
n
ce
An
al
y
s
i
s
o
f
P
r
ee
m
p
tiv
e
B
ased
Un
ip
r
o
ce
s
s
o
r
Sch
ed
u
l
in
g
w
a
s
d
is
c
u
s
s
ed
f
o
r
th
e
r
ea
l
ti
m
e
ta
s
k
[
1
1
]
.
T
h
e
Pro
ce
s
s
o
r
Sp
ee
d
C
o
n
tr
o
l
f
o
r
th
e
R
ea
l
-
T
i
m
e
S
y
s
t
e
m
s
f
o
r
p
o
w
er
r
ed
u
ctio
n
w
as
an
al
y
s
ed
[
1
2
]
.
B
u
t
th
ese
s
c
h
ed
u
lin
g
s
c
h
e
m
es
d
o
n
o
t
ap
p
ly
p
ip
elin
i
n
g
p
r
o
ce
s
s
an
d
t
h
e
y
s
u
ffer
f
r
o
m
s
i
g
n
i
fican
t
th
r
o
u
g
h
p
u
t
d
eter
io
r
atio
n
w
h
e
n
e
x
ec
u
t
in
g
p
er
io
d
ic
p
ac
k
et
p
r
o
ce
s
s
in
g
t
ask
s
.
Ma
n
y
r
e
s
ea
r
ch
er
s
p
r
ese
n
ted
th
e
r
es
u
lt
s
o
n
r
ed
u
cin
g
p
r
o
to
co
l
laten
c
y
f
o
r
h
ig
h
-
s
p
ee
d
g
ate
w
a
y
s
a
n
d
telec
o
m
m
u
n
ica
tio
n
s
y
s
te
m
s
b
ased
o
n
h
y
b
r
id
p
ar
allelis
m
.
De
v
elo
p
in
g
a
p
ac
k
et
p
r
o
ce
s
s
in
g
s
y
s
te
m
t
h
at
co
n
s
id
er
s
b
o
th
late
n
c
y
a
n
d
th
r
o
u
g
h
p
u
t
f
o
r
m
u
ltico
r
e
ar
ch
itect
u
r
es
is
b
o
th
i
n
ter
esti
n
g
a
n
d
c
h
alle
n
g
i
n
g
[
1
3
,
1
4
]
.
T
h
u
s
,
w
e
p
r
ese
n
t
a
late
n
c
y
a
n
d
th
r
o
u
g
h
p
u
t
-
a
w
ar
e
s
ch
ed
u
lin
g
s
c
h
e
m
e
b
ased
o
n
p
ar
allel
-
p
ip
elin
e
to
p
o
lo
g
y
.
A
lo
n
g
w
it
h
i
n
cr
ea
s
ed
th
r
o
u
g
h
p
u
t
an
d
r
ed
u
ce
d
laten
cy
,
h
o
w
ev
er
,
co
m
es
i
n
cr
ea
s
ed
p
o
w
er
co
n
s
u
m
p
tio
n
f
o
r
n
et
w
o
r
k
ap
p
licatio
n
s
r
u
n
n
i
n
g
o
n
m
u
ltico
r
e
ar
ch
itectu
r
e.
C
o
llec
tiv
e
l
y
,
m
i
llio
n
s
o
f
s
er
v
er
s
in
th
e
g
lo
b
al
n
et
w
o
r
k
co
n
s
u
m
e
a
g
r
ea
t
d
ea
l
o
f
p
o
w
er
.
T
h
e
ch
ip
m
a
n
u
f
ac
tu
r
es
co
n
ti
n
u
e
to
in
cr
ea
s
e
b
o
th
t
h
e
n
u
m
b
er
o
f
co
r
es
a
n
d
th
eir
f
r
e
q
u
en
cie
s
,
s
u
b
s
ta
n
tiall
y
in
cr
ea
s
in
g
b
o
th
d
y
n
a
m
ic
an
d
s
tatic
p
o
w
er
co
n
s
u
m
p
tio
n
.
Hig
h
er
p
o
w
er
co
n
s
u
m
p
tio
n
i
n
cr
ea
s
es
co
s
ts
,
b
o
t
h
d
ir
ec
tl
y
an
d
in
d
ir
ec
tl
y
.
E
n
er
g
y
it
s
el
f
is
e
x
p
ec
ted
to
b
ec
o
m
e
m
o
r
e
e
x
p
en
s
i
v
e,
esp
ec
iall
y
if
e
n
v
ir
o
n
m
e
n
tal
i
m
p
ac
t
s
ar
e
f
ac
to
r
ed
in
to
co
n
s
u
m
p
ti
o
n
.
Hig
h
er
p
o
w
er
co
n
s
u
m
p
tio
n
also
i
n
cr
ea
s
e
s
co
r
e
te
m
p
er
atu
r
e,
w
h
ich
ex
p
o
n
en
tiall
y
i
n
cr
ea
s
e
s
t
h
e
co
s
t
o
f
co
o
lin
g
an
d
p
ac
k
ag
i
n
g
.
Hi
g
h
er
te
m
p
er
a
tu
r
es
al
s
o
i
n
cr
ea
s
e
i
n
d
ir
ec
t
an
d
li
f
e
-
c
y
c
le
co
s
t
s
d
u
e
t
o
r
ed
u
ce
d
s
y
s
te
m
p
er
f
o
r
m
a
n
ce
,
cir
c
u
it r
eliab
ili
t
y
,
an
d
ch
ip
l
i
f
e
-
t
i
m
e.
T
h
er
e
f
o
r
e,
p
o
w
er
m
a
n
a
g
e
m
en
t
is
a
fi
r
s
t
-
o
r
d
er
d
esig
n
i
s
s
u
e.
As
w
e
p
r
o
p
o
s
e
t
h
e
p
ar
allel
-
p
i
p
elin
e
s
c
h
ed
u
lin
g
o
n
ta
s
k
-
le
v
e
l,
w
e
r
ea
lize
t
h
at
th
er
e
h
a
s
b
ee
n
n
o
ex
is
ti
n
g
w
o
r
k
co
n
s
id
er
in
g
th
e
p
o
w
er
b
u
d
g
e
t
is
s
u
e
s
f
o
r
it.
P
r
ev
io
u
s
p
o
w
er
-
a
w
ar
e
alg
o
r
ith
m
s
ei
t
h
er
h
av
e
n
o
t
co
n
s
id
er
ed
laten
c
y
,
o
r
h
a
v
e
n
o
t
ex
p
lo
r
ed
t
h
e
p
ar
allel
p
ip
eli
n
e
to
p
o
lo
g
y
f
o
r
task
s
c
h
ed
u
li
n
g
.
Si
n
ce
p
o
w
er
g
ati
n
g
ca
n
n
o
t
b
e
d
ir
ec
tl
y
ap
p
lied
to
task
s
c
h
ed
u
li
n
g
,
w
e
r
eso
r
t
to
DVF
S
to
in
te
g
r
ate
p
o
w
er
-
a
w
ar
en
e
s
s
i
n
t
o
p
ar
allel
p
ip
elin
e
s
ch
ed
u
lin
g
.
2.
M
E
T
H
O
DO
L
O
G
I
E
S
2
.
1
L
a
t
ency
And T
hro
ug
hp
ut
Aw
a
re
Schedu
lin
g
(
L
AT
A)
W
e
p
r
o
p
o
s
e
L
A
T
A
,
a
late
n
c
y
a
n
d
T
h
r
o
u
g
h
p
u
t
-
Aw
ar
e
p
ac
k
et
p
r
o
ce
s
s
i
n
g
s
y
s
te
m
f
o
r
m
u
ltico
r
e
ar
ch
itect
u
r
es.
I
t
ad
o
p
ts
h
y
b
r
i
d
p
ar
allelis
m
w
it
h
p
ar
allel
p
i
p
elin
e
co
r
e
to
p
o
lo
g
y
i
n
fin
e
-
g
r
ain
ed
ta
s
k
le
v
el
to
ac
h
iev
e
lo
w
late
n
c
y
an
d
h
i
g
h
th
r
o
u
g
h
p
u
t.
W
e
ac
co
m
p
li
s
h
t
h
e
ab
o
v
e
g
o
al
t
h
r
o
u
g
h
t
h
e
f
o
l
lo
w
i
n
g
t
h
r
ee
s
tep
s
.
First,
w
e
d
esi
g
n
a
lis
t
-
b
ased
p
ip
elin
e
s
ch
ed
u
li
n
g
alg
o
r
it
h
m
f
r
o
m
t
h
e
tas
k
g
r
ap
h
.
Sec
o
n
d
,
w
e
ap
p
ly
a
d
eter
m
in
i
s
tic
s
ea
r
ch
-
b
ased
r
e
fin
e
m
en
t
p
r
o
ce
s
s
to
r
ed
u
ce
laten
c
y
a
n
d
i
m
p
r
o
v
e
t
h
r
o
u
g
h
p
u
t
th
r
o
u
g
h
lo
ca
l
ad
j
u
s
t
m
e
n
t.
T
h
ir
d
,
w
e
d
ev
i
s
e
a
ca
ch
e
-
a
w
ar
e
r
eso
u
r
ce
m
ap
p
i
n
g
s
c
h
e
m
e
to
g
en
er
ate
a
p
r
ac
ti
ca
l
m
ap
p
in
g
o
n
to
a
r
ea
l
m
ac
h
i
n
e.
T
o
th
e
b
est
o
f
o
u
r
k
n
o
w
led
g
e,
L
A
T
A
i
s
t
h
e
fi
r
s
t
o
f
it
s
k
i
n
d
to
co
n
s
id
er
b
o
th
la
ten
c
y
a
n
d
t
h
r
o
u
g
h
p
u
t
in
p
ac
k
et
p
r
o
ce
s
s
i
n
g
s
y
s
te
m
s
.
W
e
i
m
p
le
m
e
n
t
L
A
T
A
o
n
an
I
n
tel
m
ac
h
i
n
e
w
i
th
t
w
o
Qu
ad
-
C
o
r
e
Xeo
n
E
5
3
3
5
p
r
o
ce
s
s
o
r
s
an
d
co
n
d
u
ct
e
x
te
n
s
iv
e
ex
p
er
i
m
en
ts
to
s
h
o
w
it
s
b
etter
p
er
f
o
r
m
a
n
ce
o
v
er
o
th
er
s
y
s
te
m
s
s
u
ch
as
P
ar
allel,
Gr
ee
d
y
,
R
an
d
o
m
an
d
B
ip
a
r
.
B
ased
o
n
s
ix
r
ea
l
p
ac
k
et
p
r
o
ce
s
s
in
g
ap
p
licatio
n
s
ch
o
s
en
f
r
o
m
Ne
t
B
en
ch
an
d
P
ac
k
et
B
en
ch
,
L
AT
A
ex
h
ib
it
s
an
av
er
a
g
e
o
f
3
6
.
5
%
r
ed
u
ctio
n
o
f
laten
c
y
ac
r
o
s
s
all
ap
p
licatio
n
s
w
it
h
o
u
t
s
u
b
s
tan
tiall
y
d
eg
r
ad
in
g
th
e
t
h
r
o
u
g
h
p
u
t.
I
t
s
h
o
w
s
a
m
ax
i
m
u
m
o
f
6
2
.
2
%
r
ed
u
ctio
n
o
f
laten
c
y
f
o
r
UR
L
ap
p
licatio
n
o
v
er
R
an
d
o
m
w
i
th
co
m
p
ar
ab
le
th
r
o
u
g
h
p
u
t p
er
f
o
r
m
an
ce
.
2
.
2
L
a
t
a
Sy
s
t
e
m
De
s
ig
n
I
n
t
h
e
L
A
T
A’
s
s
y
s
te
m
d
esi
g
n
,
w
e
fi
r
s
t
g
en
er
ate
its
co
r
r
esp
o
n
d
in
g
ta
s
k
g
r
ap
h
w
i
th
b
o
th
co
m
p
u
tat
io
n
an
d
co
m
m
u
n
icatio
n
i
n
f
o
r
m
ati
o
n
.
T
h
en
,
w
e
p
r
o
ce
ed
in
a
th
r
ee
-
s
tep
p
r
o
ce
d
u
r
e
to
s
ch
ed
u
l
e
an
d
m
ap
t
h
e
tas
k
g
r
ap
h
ac
co
r
d
in
g
to
o
u
r
n
o
v
el
d
esig
n
.
L
as
t,
w
e
d
ep
lo
y
th
e
p
r
o
g
r
a
m
o
n
to
a
r
ea
l
m
u
lt
ico
r
e
m
ac
h
in
e
to
o
b
tain
it
s
p
er
f
o
r
m
a
n
ce
r
es
u
lt.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8694
I
n
t J
P
o
w
er
E
lectr
o
n
&
Dr
i S
y
s
t
,
Vo
l.
9
,
No
.
1
,
Ma
r
ch
2
0
1
8
:
80
–
88
82
Fig
u
r
e
1
.
L
A
T
A
s
y
s
te
m
d
esi
g
n
fl
o
w
ch
ar
t.
2
.
3
Co
mm
un
ica
t
io
n
m
ea
s
ure
m
e
nt
T
h
e
co
m
m
u
n
icatio
n
ti
m
e
ca
n
n
o
t
b
e
ac
c
u
r
atel
y
m
ea
s
u
r
e
d
b
et
w
ee
n
t
w
o
co
r
es
in
a
m
u
ltico
r
e
ar
ch
itect
u
r
e
u
n
less
w
e
k
n
o
w
t
h
e
ex
ac
t
lo
ca
tio
n
o
f
t
h
e
co
r
es.
I
n
L
A
T
A
d
esig
n
,
w
e
u
s
e
t
h
e
av
er
a
g
e
co
m
m
u
n
icatio
n
co
s
t
b
ased
o
n
d
ata
ca
ch
e
ac
ce
s
s
ti
m
e,
as
g
i
v
en
in
E
q
u
a
tio
n
s
1
an
d
2
.
C
o
m
m
av
g
m
ea
n
s
t
h
e
av
er
ag
e
co
m
m
u
n
ica
tio
n
co
s
t
to
tr
an
s
f
er
a
u
n
it
d
ata
s
et,
w
h
ic
h
ca
n
b
e
ap
p
r
o
x
i
m
ated
b
y
s
y
s
te
m
m
e
m
o
r
y
laten
cie
s
(
L
1
,
L
2
an
d
m
ain
m
e
m
o
r
y
ac
ce
s
s
ti
m
e)
a
n
d
p
r
o
g
r
a
m
d
ata
ca
ch
e
p
er
f
o
r
m
a
n
ce
s
(
L
1
an
d
L
2
ca
c
h
e
h
it
r
ate)
.
Data
Size
r
ef
er
s
t
o
th
e
tr
an
s
f
er
r
ed
d
ata
s
et
s
ize
b
et
w
ee
n
t
w
o
co
m
m
u
n
icati
n
g
tas
k
s
.
C
omm
av
g
=
C
omm
av
g
.
Da
ta
Size
(
1
)
C
omm
av
g
=
(
(
TL
1
.
Hit
L
1
)
+
(
TL
2
.
(
1
−
Hit
L
1
)
)
.
HitL
2
+
T
m
em
.
(
(
1
−
Hit
L
1
)
.
(
1
−
Hit
L
2
)
)
)
(
2
)
As
w
e
k
n
o
w
,
th
e
t
h
r
o
u
g
h
p
u
t
ca
n
b
e
ca
lcu
lated
b
y
th
e
i
n
v
e
r
s
e
o
f
th
e
lo
n
g
e
s
t
s
ta
g
e
ti
m
e
1
/T
m
ax
i
n
p
ip
elin
in
g
.
T
h
u
s
,
w
e
f
o
r
m
o
u
r
o
b
j
ec
tiv
e
f
u
n
c
tio
n
i
n
E
q
u
atio
n
3
,
w
h
er
e
L
is
t
h
e
s
c
h
ed
u
led
la
ten
c
y
.
2
.
4
P
ro
ble
m
Sta
t
e
m
e
nt
T
h
e
laten
c
y
ca
n
b
e
d
e
f
in
ed
as
t
h
e
s
ch
ed
u
le
le
n
g
th
o
f
a
p
r
o
g
r
am
an
d
th
r
o
u
g
h
p
u
t
as
t
h
e
s
y
s
te
m
th
r
o
u
g
h
p
u
t.
T
h
e
p
r
o
b
lem
s
tate
m
en
t
is
:
g
i
v
e
n
th
e
laten
c
y
co
n
s
tr
ain
t
L
0
,
s
ch
ed
u
le
a
p
ac
k
et
p
r
o
ce
s
s
in
g
p
r
o
g
r
a
m
in
p
ar
allel
p
ip
elin
e
co
r
e
to
p
o
l
o
g
y
s
o
as
to
m
a
x
i
m
ize
t
h
e
th
r
o
u
g
h
p
u
t
T
h
.
T
h
e
ai
m
is
to
r
ea
r
r
an
g
e
th
e
ta
s
k
s
i
n
to
th
e
p
ar
allel
p
ip
elin
e
ta
s
k
g
r
ap
h
as
s
h
o
w
n
i
n
Fi
g
u
r
e
3
,
s
o
t
h
at
th
e
to
tal
e
x
ec
u
tio
n
ti
m
e
T
1
+
T
2
+
T
3
+
T
4
is
m
i
n
i
m
ized
w
h
ile
m
ain
tain
i
n
g
th
e
t
h
r
o
u
g
h
p
u
t
as
h
i
g
h
a
s
p
o
s
s
ib
le.
As
w
e
k
n
o
w
,
t
h
e
t
h
r
o
u
g
h
p
u
t
ca
n
b
e
ca
lcu
lated
b
y
t
h
e
in
v
er
s
e
o
f
th
e
lo
n
g
est
s
tag
e
ti
m
e
1
/T
m
ax
i
n
p
ip
elin
in
g
.
T
h
u
s
,
w
e
f
o
r
m
o
u
r
o
b
j
ec
tiv
e
f
u
n
ctio
n
in
E
q
u
a
tio
n
3
,
w
h
er
e
L
is
t
h
e
s
c
h
ed
u
led
late
n
c
y
.
Ma
x
im
ize
Th
=
1
T
ma
x
(
s
.
t
.
at
L
≤
L
0
)
(3
)
3
P
RO
P
O
SE
D
L
A
T
E
NCY
R
E
DUCT
I
O
N
L
ate
n
c
y
ca
n
b
e
r
ed
u
ce
d
b
y
r
ed
u
cin
g
eit
h
er
co
m
p
u
tatio
n
ti
m
e
o
r
co
m
m
u
n
icatio
n
ti
m
e.
B
ec
au
s
e
co
m
p
u
tatio
n
d
o
m
i
n
ates
t
h
e
o
v
er
all
ex
ec
u
tio
n
ti
m
e
f
o
r
m
o
s
t
p
ac
k
et
p
r
o
ce
s
s
i
n
g
ap
p
licatio
n
s
r
u
n
n
i
n
g
o
n
m
u
ltico
r
e
ar
ch
itect
u
r
es,
w
e
p
r
io
r
itize
co
m
p
u
tatio
n
r
ed
u
ctio
n
in
d
esig
n
i
n
g
L
A
T
A
.
He
n
ce
,
L
A
T
A
fi
r
s
t
ap
p
lies
laten
c
y
h
id
i
n
g
to
r
ed
u
ce
co
m
p
u
tatio
n
t
i
m
e
.
T
h
en
,
C
C
P
eli
m
in
at
io
n
a
n
d
C
C
P
r
ed
u
ctio
n
a
r
e
u
s
ed
to
r
e
d
u
ce
co
m
m
u
n
ica
tio
n
ti
m
e.
C
o
m
p
u
tatio
n
r
ed
u
ctio
n
:
W
e
d
efin
ed
a
cr
itic
al
n
o
d
e
as
th
e
n
o
d
e
in
a
p
ip
eli
n
e
s
ta
g
e
w
h
ich
d
o
m
in
a
tes
t
h
e
co
m
p
u
tatio
n
ti
m
e.
T
h
en
,
L
a
ten
c
y
h
id
i
n
g
ca
n
b
e
d
efin
ed
as
a
tec
h
n
iq
u
e
t
h
at
p
l
ac
es
a
cr
itica
l
n
o
d
e
f
r
o
m
o
n
e
s
ta
g
e
to
o
n
e
o
f
it
s
ad
j
ac
en
t
s
tag
es
w
it
h
o
u
t
v
io
lati
n
g
d
ep
en
d
en
c
ies,
s
o
th
at
it
s
co
m
p
u
tat
io
n
ti
m
e
is
s
h
ad
o
w
ed
b
y
th
e
o
th
er
cr
itical
n
o
d
e
in
th
e
n
e
w
s
tag
e.
B
ac
k
w
ar
d
h
id
i
n
g
(
B
aH)
r
ef
er
s
to
p
lacin
g
a
cr
itical
n
o
d
e
in
to
i
ts
p
r
ec
ed
en
t
s
tag
e.
Fo
r
w
ar
d
h
id
i
n
g
(
Fo
H)
r
ef
er
s
t
o
p
lacin
g
a
cr
itical
n
o
d
e
in
to
it
s
f
o
llo
w
i
n
g
s
tag
e
w
h
ic
h
is
s
h
o
w
n
i
n
Fi
g
u
r
e
2
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
er
E
lectr
o
n
&
Dr
i
S
y
s
t
I
SS
N:
2
0
8
8
-
8694
A
N
o
ve
l A
p
p
r
o
a
ch
in
S
c
h
ed
u
li
n
g
Of
th
e
R
ea
l
-
Time
Ta
s
ks I
n
Hete
r
o
g
en
eo
u
s
Mu
ltico
r
e…
(
L
a
va
n
ya
Dh
a
n
esh
)
83
Fig
u
r
e
2
.
L
aten
c
y
h
id
in
g
o
n
n
o
d
e
E
.
4
P
E
RF
O
RM
ANCE E
VA
L
U
AT
I
O
N
T
h
e
laten
c
y
a
n
d
th
r
o
u
g
h
p
u
t
f
o
r
s
ix
ap
p
licatio
n
s
b
y
L
A
T
A
,
P
ar
allel
an
d
L
i
s
t a
r
e
s
h
o
w
n
i
n
t
h
e
F
ig
u
r
es
3
an
d
4
.
W
e
o
b
s
er
v
e
th
at
P
ar
al
lel
s
u
ffer
s
f
r
o
m
h
i
g
h
late
n
c
y
d
u
e
to
it
s
s
eq
u
e
n
tia
l e
x
ec
u
tio
n
o
f
tas
k
s
.
C
o
m
p
ar
ed
w
it
h
P
ar
allel,
L
A
T
A
r
ed
u
ce
s
t
h
e
laten
c
y
b
y
a
n
av
er
a
g
e
o
f
3
4
.
2
%.
Fig
u
r
e
3
.
L
aten
c
y
o
f
s
ix
ap
p
lic
atio
n
s
b
y
L
A
T
A
,
P
ar
allel
an
d
L
is
t
Fig
u
r
e
4
.
T
h
r
o
u
g
h
p
u
t o
f
s
i
x
ap
p
licatio
n
s
b
y
L
A
T
A
,
P
ar
allel
an
d
L
is
t
P
ar
ticu
lar
l
y
,
f
o
r
U
R
L
,
L
A
T
A
ac
h
ie
v
es
t
h
e
m
a
x
i
m
al
lat
en
c
y
r
ed
u
c
tio
n
o
f
6
2
.
2
%.
I
n
ad
d
itio
n
,
L
A
T
A’
s
t
h
r
o
u
g
h
p
u
t
is
clo
s
e
to
th
at
o
f
P
ar
allel
in
s
p
ite
o
f
t
h
e
7
5
%
laten
c
y
co
n
s
tr
ain
t.
T
h
is
is
b
ec
au
s
e
L
A
T
A
is
ca
p
ab
le
o
f
o
p
tim
iz
in
g
its
p
a
r
allel
p
ip
elin
e
co
r
e
to
p
o
l
o
g
y
t
o
p
r
o
d
u
ce
g
o
o
d
th
r
o
u
g
h
p
u
t.
W
ith
r
esp
ec
t
to
L
is
t,
w
h
ic
h
is
d
es
ig
n
ed
to
p
r
o
d
u
ce
th
e
lo
w
es
t
late
n
c
y
,
L
A
T
A
a
ctu
all
y
m
atch
e
s
it
s
late
n
c
y
p
e
r
f
o
r
m
an
ce
in
m
o
s
t
ca
s
es
b
y
ag
g
r
es
s
i
v
el
y
e
x
p
lo
iti
n
g
ta
s
k
-
l
e
v
el
p
ar
allelis
m
.
F
u
r
t
h
er
m
o
r
e,
L
A
T
A
o
u
tp
er
f
o
r
m
s
L
is
t
in
t
h
r
o
u
g
h
p
u
t
b
y
a
n
av
er
a
g
e
o
f
4
1
.
0
% a
n
d
a
m
ax
i
m
u
m
o
f
5
6
.
7
% f
o
r
R
o
u
te.
5.
P
O
WE
R
AWA
RE
P
ARA
L
L
E
L
P
I
P
E
L
I
N
E
SCH
E
DUL
I
NG
A
L
G
O
R
I
T
H
M
W
e
in
tr
o
d
u
ce
t
h
e
n
o
v
el
p
ar
all
el
-
p
ip
elin
e
s
c
h
ed
u
li
n
g
o
n
tas
k
-
le
v
el
f
o
r
n
et
w
o
r
k
ap
p
licatio
n
s
t
h
at
ca
n
attain
h
ig
h
th
r
o
u
g
h
p
u
t
u
n
d
er
g
iv
en
late
n
c
y
co
n
s
tr
ai
n
ts
.
I
n
t
h
i
s
ch
ap
ter
,
w
e
ad
d
r
ess
th
e
p
o
w
er
b
u
d
g
et
is
s
u
e
f
o
r
th
is
s
c
h
ed
u
li
n
g
p
ar
ad
ig
m
f
o
r
n
et
w
o
r
k
p
ac
k
e
t
p
r
o
ce
s
s
in
g
.
W
e
ai
m
at
o
p
ti
m
iz
in
g
b
o
th
t
h
r
o
u
g
h
p
u
t
a
n
d
laten
c
y
u
n
d
er
g
i
v
e
n
p
o
w
er
b
u
d
g
et
b
y
ap
p
r
o
p
r
iately
ap
p
l
y
i
n
g
p
er
-
c
o
r
e
DVFS.
W
e
p
r
o
p
o
s
e
a
th
r
ee
-
s
tep
s
o
l
u
tio
n
to
ac
h
iev
e
o
u
r
g
o
al.
5
.
1
A
t
hree
-
s
t
ep
re
cursiv
e
a
lg
o
rit
h
m
ST
E
P
1
:
I
n
th
e
fi
r
s
t
s
tep
,
w
e
r
ed
u
ce
th
e
p
o
w
er
w
it
h
o
u
t
c
o
m
p
r
o
m
is
i
n
g
t
h
r
o
u
g
h
p
u
t
o
r
laten
c
y
b
y
k
ee
p
in
g
th
e
p
ip
e
lin
e
s
ta
g
e
ti
m
e
T
i,
i
=
1
,
2
.
.
.
S
u
n
ch
an
g
ed
.
W
e
d
efi
n
e
a
cr
itical
n
o
d
e
as
th
e
n
o
d
e
in
a
p
i
p
elin
e
s
tag
e
t
h
at
d
o
m
in
ate
s
th
e
co
m
p
u
tatio
n
ti
m
e.
T
h
er
ef
o
r
e,
th
e
co
m
p
u
tatio
n
ti
m
e
o
f
a
cr
itical
n
o
d
e
is
eq
u
al
to
th
e
p
ip
elin
e
s
ta
g
e
ti
m
e
(
ti
=
T
i)
.
Fo
r
ea
ch
s
ta
g
e
Si,
w
e
i
n
cr
ea
s
e
t
h
e
co
m
p
u
tatio
n
ti
m
e
o
f
n
o
n
-
cr
itical
n
o
d
es
i
n
t
h
a
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8694
I
n
t J
P
o
w
er
E
lectr
o
n
&
Dr
i S
y
s
t
,
Vo
l.
9
,
No
.
1
,
Ma
r
ch
2
0
1
8
:
80
–
88
84
s
tag
e
to
th
e
le
n
g
th
o
f
T
i.
Sin
c
e
all
s
tag
e
ti
m
es
r
e
m
ain
t
h
e
s
a
m
e,
th
e
t
h
r
o
u
g
h
p
u
t
an
d
th
e
lat
en
c
y
w
i
ll
also
k
ee
p
u
n
c
h
a
n
g
ed
d
u
r
i
n
g
t
h
i
s
s
tep
w
h
ich
is
d
ep
icted
in
t
h
e
Fi
g
u
r
e
5
.
Fig
u
r
e
5
.
I
llu
s
tr
atio
n
o
f
th
e
fi
r
s
t step
o
f
t
h
e
alg
o
r
it
h
m
Fig
u
r
e
6
.
I
llu
s
tr
atio
n
o
f
th
e
s
e
co
n
d
s
tep
o
f
th
e
alg
o
r
ith
m
Fig
u
r
e
7
.
I
llu
s
tr
atio
n
o
f
th
e
t
h
i
r
d
s
tep
o
f
th
e
alg
o
r
ith
m
ST
E
P
2
:
I
n
th
e
s
ec
o
n
d
s
tep
,
w
e
r
ed
u
ce
t
h
e
p
o
w
er
w
it
h
th
r
o
u
g
h
p
u
t
u
n
c
h
a
n
g
ed
an
d
m
i
n
i
m
al
late
n
c
y
in
cr
ea
s
e
w
h
ic
h
is
s
h
o
w
n
in
F
ig
u
r
e
6
.
T
h
is
is
ac
h
ie
v
ed
b
y
k
ee
p
in
g
t
h
e
lo
n
g
est
s
tag
e
ti
m
e
T
m
ax
u
n
ch
a
n
g
ed
w
h
ile
w
e
i
n
cr
ea
s
e
t
h
e
s
tag
e
t
i
m
e
o
f
o
t
h
er
s
ta
g
es.
W
e
d
en
o
t
e
th
e
s
ta
g
e
w
it
h
T
m
ax
a
s
t
h
e
b
o
ttlen
ec
k
s
tag
e
i
n
th
e
p
ip
elin
e.
T
h
u
s
,
all
o
th
er
s
tag
es
ar
e
n
o
n
-
b
o
ttlen
ec
k
s
ta
g
es.
W
e
d
efi
n
e
∆T
as
th
e
s
h
o
r
test
ti
m
e
p
er
io
d
b
y
w
h
ic
h
w
e
ca
n
in
cr
ea
s
e
th
e
la
t
en
c
y
.
T
o
m
in
i
m
ize
th
e
la
ten
c
y
in
cr
ea
s
e,
w
e
iter
ati
v
el
y
in
cr
ea
s
e
th
e
la
ten
c
y
b
y
∆T
u
n
til
t
h
e
p
o
w
er
b
u
d
g
et
i
s
s
atis
fied
o
r
all
th
e
s
tag
e
s
r
ea
c
h
T
m
a
x
.
I
f
th
e
f
o
r
m
er
co
m
es
tr
u
e,
th
e
a
lg
o
r
it
h
m
r
etu
r
n
s
a
n
d
th
e
r
es
u
lti
n
g
s
c
h
e
d
u
lin
g
g
u
ar
a
n
tees
t
h
e
m
in
i
m
a
l
laten
c
y
in
cr
ea
s
e,
w
h
ic
h
w
ill
b
e
p
r
o
v
ed
s
h
o
r
tl
y
.
Oth
er
w
i
s
e,
if
t
h
e
latter
co
m
e
s
tr
u
e,
w
e
p
r
o
ce
ed
to
s
tep
th
r
ee
.
ST
E
P
3
:
I
n
th
e
t
h
ir
d
s
tep
,
w
e
r
ed
u
ce
th
e
p
o
w
er
b
y
m
in
i
m
izi
n
g
b
o
t
h
th
e
th
r
o
u
g
h
p
u
t
a
n
d
t
h
e
laten
c
y
p
er
f
o
r
m
a
n
ce
lo
s
s
.
Af
ter
s
tep
t
w
o
,
e
v
er
y
s
tag
e
h
as
th
e
s
a
m
e
s
tag
e
t
i
m
e
T
m
a
x
.
Fo
llo
w
i
n
g
t
h
e
s
a
m
e
r
u
le
o
f
ch
o
o
s
in
g
a
ca
n
d
id
ate
s
ta
g
e
i
n
s
tep
t
w
o
,
w
e
o
p
ti
m
all
y
c
h
o
o
s
e
a
s
ta
g
e
to
f
u
r
t
h
er
i
n
cr
ea
s
e
it
s
s
ta
g
e
ti
m
e
b
y
∆T
.
Sin
ce
th
e
o
r
ig
i
n
al
T
m
a
x
i
s
i
n
cr
ea
s
ed
,
t
h
e
t
h
r
o
u
g
h
p
u
t
i
s
c
o
m
p
r
o
m
is
ed
ac
co
r
d
in
g
l
y
w
h
i
ch
i
s
s
h
o
w
n
i
n
th
e
Fig
u
r
e
7
.
6.
P
O
WE
R
M
O
DE
L
C
o
n
s
id
er
th
at
tas
k
T
co
n
s
is
t
s
o
f
C
clo
ck
c
y
cle
s
o
n
p
r
o
ce
s
s
o
r
P
,
w
h
ic
h
r
u
n
s
at
v
o
lta
g
e
V
an
d
f
r
eq
u
en
c
y
f
.
W
e
ass
u
m
e
th
at
C
d
o
es
n
o
t
c
h
an
g
e
w
it
h
d
iffer
en
t
V
an
d
f
.
Fo
r
a
g
i
v
en
v
o
lta
g
e
V
,
p
r
o
ce
s
s
o
r
P
h
as
a
n
a
v
er
ag
e
p
o
w
er
co
n
s
u
m
p
tio
n
P
o
w
.
I
t
is
k
n
o
w
n
th
a
t
p
r
o
ce
s
s
o
r
p
o
w
er
co
n
s
u
m
p
t
i
o
n
is
d
o
m
in
ated
b
y
d
y
n
a
m
ic
p
o
w
er
d
is
s
ip
atio
n
g
i
v
en
b
y
:
Pow
=
K
a
.
f
.
V
2
(
4
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
er
E
lectr
o
n
&
Dr
i
S
y
s
t
I
SS
N:
2
0
8
8
-
8694
A
N
o
ve
l A
p
p
r
o
a
ch
in
S
c
h
ed
u
li
n
g
Of
th
e
R
ea
l
-
Time
Ta
s
ks I
n
Hete
r
o
g
en
eo
u
s
Mu
ltico
r
e…
(
L
a
va
n
ya
Dh
a
n
esh
)
85
W
h
er
e
Ka
is
a
task
/p
r
o
ce
s
s
o
r
d
ep
en
d
en
t f
ac
to
r
d
eter
m
i
n
ed
b
y
th
e
s
w
itc
h
ed
ca
p
ac
itan
ce
.
T
h
e
en
er
g
y
co
n
s
u
m
ed
b
y
e
x
ec
u
ti
n
g
ta
s
k
T
o
n
p
r
o
ce
s
s
o
r
P
is
co
m
p
u
ted
as:
E
=
C
.
Pow
f
(
5)
W
e
ca
n
r
e
w
r
ite
it a
s
:
E
=
C
.
E
f
(
6)
V
=
C
.
K
a
.
V
2
(
7)
W
h
er
e
E
f
,
V
is
th
e
av
er
a
g
e
c
y
cle
en
er
g
y
.
Fro
m
th
i
s
w
e
ca
n
s
ee
t
h
at
lo
w
er
in
g
t
h
e
v
o
lta
g
e
w
o
u
l
d
y
ield
a
d
r
asti
c
d
ec
r
ea
s
e
in
en
er
g
y
co
n
s
u
m
p
tio
n
.
T
h
e
f
r
eq
u
en
c
y
f
is
al
m
o
s
t l
in
ea
r
l
y
r
elate
d
to
th
e
v
o
ltag
e:
f
=
K
b
.
(
V
−
V
T
)
(
V
−
V
T
)
2
.
V
(
8)
W
h
er
e
VT
is
t
h
e
t
h
r
es
h
o
ld
v
o
ltag
e
a
n
d
Kb
i
s
a
co
n
s
ta
n
t.
Fo
r
a
s
u
ffi
cie
n
tl
y
s
m
all
t
h
r
es
h
o
l
d
v
o
ltag
e,
th
e
f
r
eq
u
e
n
c
y
is
ap
p
r
o
x
i
m
a
ted
to
Kb
.
Ho
w
e
v
er
,
o
u
r
alg
o
r
ith
m
i
s
ab
le
to
g
u
ar
an
tee
a
m
i
n
i
m
al
p
er
f
o
r
m
a
n
ce
lo
s
s
i
n
t
h
is
s
ce
n
ar
io
.
T
h
e
p
r
o
o
f
o
f
o
p
ti
m
alit
y
is
i
n
li
n
e
w
it
h
th
at
in
s
tep
t
w
o
,
w
h
er
e
t
h
e
m
in
i
m
al
ti
m
e
p
er
io
d
in
cr
e
m
e
n
t
g
u
ar
an
tees
t
h
at
w
h
e
n
w
e
s
a
tis
f
y
th
e
p
o
w
er
b
u
d
g
et
c
o
n
s
tr
ain
t,
th
e
p
er
f
o
r
m
a
n
ce
lo
s
s
is
m
in
i
m
al
w
h
ic
h
is
s
h
o
w
n
i
n
th
e
Fig
u
r
e
8
.
Fig
u
r
e
8
.
T
h
e
p
o
w
er
-
a
w
ar
e
p
a
r
allel
-
p
ip
elin
e
s
ch
ed
u
li
n
g
al
g
o
r
ith
m
.
I
n
o
u
r
s
y
s
te
m
,
s
i
n
ce
en
er
g
y
in
cr
ea
s
es
w
h
en
p
er
f
o
r
m
an
ce
d
eg
r
ad
es
(
i.e
.
,
a
lo
n
g
er
e
x
ec
u
tio
n
ti
m
e)
,
w
e
ass
i
g
n
a
h
i
g
h
er
w
e
ig
h
t
to
p
er
f
o
r
m
a
n
ce
b
y
s
etti
n
g
α
c
=
0
.
1
5
α
m
.
Sp
ec
i
fi
ca
ll
y
,
th
e
lo
s
t
f
ac
to
r
s
f
o
r
GP
U
c
o
r
es
an
d
m
e
m
o
r
y
ar
e
ca
lcu
lated
as:
l
_
C
i
t
=
α
c
.
C
ie
t
+
(
1
−
α
c
)
(
l
_
C
ip
t
)
(
9
)
T
h
en
w
e
co
m
b
i
n
ed
co
r
e
an
d
m
e
m
o
r
y
lo
s
t
f
u
n
c
tio
n
s
to
g
eth
e
r
b
y
a
f
ac
to
r
,
w
h
ich
b
alan
ce
s
co
r
e
im
p
ac
t
an
d
m
e
m
o
r
y
i
m
p
ac
t in
i
n
flu
e
n
cin
g
s
y
s
te
m
p
er
f
o
r
m
a
n
ce
an
d
en
er
g
y
.
To
ta
l
L
oss
ij
t
=
∅
.
l
_
C
i
t
+
(
1
−
∅
)
.
l
_
m
j
t
(
10)
Sh
o
w
s
h
o
w
to
tal
lo
s
t
f
u
n
c
t
io
n
is
o
b
tain
ed
.
Fo
r
d
if
f
er
en
t
C
P
U
-
GP
U
s
y
s
te
m
s
,
b
y
t
u
n
in
g
φ
v
alu
e
t
h
e
s
y
s
te
m
ca
n
ac
h
iev
e
b
alan
ce
b
et
w
ee
n
co
r
e
a
n
d
m
e
m
o
r
y
i
n
flu
en
ce
.
I
n
o
u
r
h
ar
d
w
ar
e
test
ed
,
0
.
3
3
is
th
e
v
a
lu
e
r
efl
ec
t
s
s
y
s
te
m
c
h
ar
ac
ter
is
tic
d
er
iv
ed
f
r
o
m
ex
p
er
i
m
e
n
t
s
.
B
ased
o
n
th
e
to
tal
lo
s
s
,
t
h
e
w
ei
g
h
ts
u
s
ed
i
n
th
e
f
r
eq
u
en
c
y
s
ca
lin
g
al
g
o
r
ith
m
c
an
b
e
u
p
d
ated
as f
o
llo
w
s
.
we
ig
ht
ij
(
t
+
1
)
=
(
(
we
ight
ij
t
)
(
1
−
(
1
−
β
)
.
To
ta
l
L
oss
ij
t
)
(
11)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8694
I
n
t J
P
o
w
er
E
lectr
o
n
&
Dr
i S
y
s
t
,
Vo
l.
9
,
No
.
1
,
Ma
r
ch
2
0
1
8
:
80
–
88
86
T
h
e
alg
o
r
ith
m
w
i
ll
b
e
m
o
r
e
r
o
b
u
s
t
to
s
y
s
te
m
n
o
i
s
e.
A
s
m
all
er
β
g
iv
e
s
m
o
r
e
w
ei
g
h
t
o
n
lo
s
s
f
ac
to
r
o
f
cu
r
r
en
t
t
i
m
e
in
ter
v
al.
T
h
e
al
g
o
r
ith
m
w
ill
r
e
s
p
o
n
d
to
w
o
r
k
lo
ad
ch
an
g
e
i
n
a
s
h
o
r
t
t
i
m
e.
I
n
o
u
r
ex
p
er
i
m
e
n
t,
w
e
s
elec
t
β
=
0
.
2
t
o
fi
lter
o
u
t
s
y
s
te
m
n
o
is
e
w
it
h
q
u
ick
w
o
r
k
l
o
ad
ch
an
g
e
r
esp
o
n
s
e.
A
l
m
o
n
d
th
e
en
tire
N
×
M
w
ei
g
h
ts
(
a
s
s
u
m
e
w
e
h
a
v
e
N
co
r
e
f
r
eq
u
en
c
y
le
v
els
a
n
d
M
m
e
m
o
r
y
f
r
eq
u
en
c
y
le
v
el
s
)
,
th
e
h
ig
h
es
t
o
n
e
i
s
s
elec
ted
an
d
its
co
r
r
esp
o
n
d
in
g
co
r
e
an
d
m
e
m
o
r
y
f
r
eq
u
en
cie
s
ar
e
en
f
o
r
ce
d
in
th
e
n
ex
t p
er
io
d
.
6.
E
XP
E
R
I
M
E
NT
A
L
RE
SUL
T
S
T
h
e
p
o
w
er
an
d
t
h
er
m
al
-
a
w
a
r
e
s
ch
ed
u
li
n
g
r
es
u
lts
f
o
r
d
if
f
er
en
t
b
e
n
ch
m
ar
k
s
f
r
o
m
t
h
e
e
m
b
ed
d
ed
s
y
s
te
m
s
y
n
t
h
es
is
b
en
c
h
m
ar
k
s
s
u
ite
o
r
g
e
n
er
ated
u
s
in
g
t
h
e
MA
T
L
A
B
to
o
l.
T
h
e
p
r
o
ce
s
s
o
f
T
ask
s
c
h
ed
u
li
n
g
is
s
h
o
w
n
i
n
t
h
e
F
ig
u
r
e
9
.
T
h
e
I
L
P
s
o
lu
tio
n
s
w
er
e
g
en
er
ated
u
s
i
n
g
OP
E
NM
P
.
T
h
e
s
i
m
u
lati
o
n
s
w
er
e
p
er
f
o
r
m
ed
o
n
a
d
u
al
co
r
e
s
y
s
te
m
o
f
3
GHZ
p
r
o
ce
s
s
o
r
.
T
h
e
o
u
tp
u
t
ex
ec
u
t
io
n
an
d
th
e
r
es
u
lt
s
u
s
i
n
g
t
h
e
f
u
zz
y
lo
g
ic
tech
n
iq
u
es
ar
e
s
h
o
w
n
i
n
t
h
e
F
ig
u
r
e
s
1
0
,
1
1
an
d
1
2
.
T
h
e
ex
ec
u
tio
n
ti
m
e
a
n
d
th
e
r
elati
v
e
d
ea
d
lin
e
ar
e
ad
j
u
s
ted
to
g
et
t
h
e
d
esire
d
p
r
io
r
ity
o
f
t
h
e
task
s
.
T
h
e
m
o
d
if
ied
p
r
io
r
it
y
o
r
d
er
b
ased
o
n
th
e
f
u
zz
y
lo
g
ic
o
u
tp
u
t
i
s
s
h
o
w
n
i
n
th
e
Fi
g
u
r
e
1
3
.
T
h
e
co
r
r
esp
o
n
d
in
g
w
a
v
e
f
o
r
m
s
ar
e
d
ep
icted
i
n
th
e
F
ig
u
r
e
1
4
.
A
s
t
h
e
p
o
w
er
in
f
o
r
m
atio
n
o
f
th
e
p
r
o
ce
s
s
in
g
e
le
m
e
n
t
s
in
th
e
s
e
b
en
ch
m
ar
k
s
is
n
o
t
av
ai
lab
le,
ap
p
r
o
x
im
atio
n
v
a
lu
e
s
b
as
ed
o
n
th
e
i
n
ter
n
al
s
tr
u
ct
u
r
e
o
f
ea
ch
co
r
e
i
s
u
s
ed
.
B
ased
o
n
t
h
ese
ap
p
r
o
x
i
m
ate
d
v
alu
e
s
,
t
h
e
p
o
w
er
co
n
s
u
m
p
t
io
n
o
f
ea
c
h
co
r
e
i
s
ad
j
u
s
ted
b
y
r
ep
laci
n
g
b
y
a
n
es
ti
m
ated
n
u
m
b
er
o
f
g
ate
s
i
n
t
h
e
m
o
d
u
le.
T
h
e
h
ar
d
w
ar
e
p
r
o
to
t
y
p
e
m
o
d
el
f
o
r
th
e
R
ea
l
-
T
i
m
e
ta
s
k
s
ch
ed
u
li
n
g
in
Hete
r
o
g
e
n
eo
u
s
Mu
l
tico
r
e
P
r
o
ce
s
s
o
r
f
o
r
Mic
r
o
g
r
id
P
o
w
er
Ma
n
ag
e
m
e
n
t
i
s
s
h
o
w
n
in
t
h
e
Fi
g
u
r
e1
5
.
Fig
u
r
e
9
.
T
ask
s
ch
ed
u
li
n
g
p
r
o
ce
s
s
Fig
u
r
e
1
0
.
E
x
ec
u
ti
n
g
o
u
tp
u
t a
cc
o
r
d
in
g
to
R
u
le
s
Fig
u
r
e
1
1
.
I
n
p
u
t W
av
ef
o
r
m
T
a
s
k
T
o
k
e
n
G
e
n
e
r
a
to
r
s
T
a
s
k
T
o
k
e
n
Ma
n
a
g
e
r
T
a
s
k
s
in
V
a
l
u
e
o
f
'
A
'
T
a
s
k
1
N
e
x
t
t
a
s
k
1
T
a
s
k
2
T
a
s
k
3
N
e
x
t
t
a
s
k
3
T
a
s
k
4
T
a
s
k
I
n
T
a
s
k
O
u
t
T
a
s
k
D
i
s
p
a
t
ch
e
r
I
S
R
T
o
k
e
n
i
n
T
a
s
k
T
o
k
e
n
i
n
T
a
s
k
T
o
k
e
n
O
u
t
T
a
s
k
D
e
l
a
y
f
un
c
t
ion
()
Ou
t
1
T
a
s
k
4
(
+
1
)
f
un
c
t
ion
()
Ou
t
1
T
a
s
k
3
(
+
2
)
f
un
c
t
ion
()
Ou
t
1
T
a
s
k
2
(
-
3
)
f
un
c
t
ion
()
Ou
t
1
T
a
s
k
1
(
+
1
0
)
T
a
s
k
C
o
m
p
l
e
t
e
Pl
o
t
s
f
un
c
t
ion
()
Ou
t
1
M
o
n
i
t
o
r
H
P
T
ok
en
I
n
t
e
r
r
u
p
t
-
P
r
i
o
r
i
t
y
T
a
s
k
T
o
k
e
n
G
e
n
e
r
a
t
o
r
T
a
s
k
I
D
1
f
un
c
t
ion
()
Ou
t
1
I
n
i
t
i
a
l
V
a
l
u
e
f
(
)
I
n
i
t
i
a
l
E
ve
n
t
A
D
a
t
a
S
t
o
r
e
M
e
m
o
r
y
LP
T
ok
en
A
p
p
l
i
ca
t
i
o
n
-
P
r
i
o
r
i
t
y
T
a
s
k
T
o
k
e
n
G
e
n
e
r
a
t
o
r
T
a
s
k
I
D
3
St
art
LP
T
ok
en
A
p
p
l
i
ca
t
i
o
n
-
P
r
i
o
r
i
t
y
T
o
k
e
n
G
e
n
e
r
a
t
o
r
T
a
s
k
I
D
4
St
art
LP
T
ok
en
A
p
p
l
i
ca
t
i
o
n
-
P
r
i
o
r
i
t
y
T
o
k
e
n
G
e
n
e
r
a
t
o
r
T
a
s
k
I
D
2
I
N
1
I
N
2
I
N
3
O
U
T
A
p
p
l
i
ca
t
i
o
n
T
a
s
k
C
o
m
b
i
n
e
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
er
E
lectr
o
n
&
Dr
i
S
y
s
t
I
SS
N:
2
0
8
8
-
8694
A
N
o
ve
l A
p
p
r
o
a
ch
in
S
c
h
ed
u
li
n
g
Of
th
e
R
ea
l
-
Time
Ta
s
ks I
n
Hete
r
o
g
en
eo
u
s
Mu
ltico
r
e…
(
L
a
va
n
ya
Dh
a
n
esh
)
87
Fig
u
r
e
1
2
.
Ou
tp
u
t W
av
e
f
o
r
m
s
Fig
u
r
e
1
3
.
A
d
j
u
s
ti
n
g
E
x
ec
u
t
io
n
an
d
Dea
d
lin
e
f
o
r
Gettin
g
P
r
io
r
ity
Fig
u
r
e
1
4
.
Ou
tp
u
t W
av
e
f
o
r
m
Fig
u
r
e
1
5
.
P
r
o
to
ty
p
e
Har
d
w
ar
e
m
o
d
el
7.
CO
NCLU
SI
O
N
C
u
r
r
en
t
r
esear
ch
o
n
GP
U
-
C
P
U
s
y
s
te
m
s
f
o
cu
s
es
m
ai
n
l
y
o
n
th
e
p
er
f
o
r
m
a
n
ce
asp
ec
ts
,
w
h
ile
t
h
e
en
er
g
y
e
f
f
icie
n
c
y
o
f
s
u
ch
s
y
s
te
m
s
r
ec
eiv
e
s
m
u
ch
les
s
at
ten
tio
n
.
T
h
er
e
ar
e
f
e
w
e
x
is
tin
g
s
tu
d
ies
th
at
s
tar
t
to
lo
w
er
th
e
e
n
er
g
y
co
n
s
u
m
p
tio
n
o
f
GP
U
-
C
P
U
ar
ch
itect
u
r
es,
b
u
t
t
h
e
y
ad
d
r
ess
eit
h
er
GP
U
o
r
C
P
U
in
an
i
s
o
lated
m
an
n
er
an
d
th
u
s
ca
n
n
o
t
ac
h
ie
v
e
m
a
x
i
m
ized
en
er
g
y
s
a
v
i
n
g
s
.
I
n
th
is
p
ap
er
,
w
e
h
av
e
p
r
ese
n
ted
Gr
ee
n
GP
U,
a
h
o
lis
tic
e
n
er
g
y
m
an
a
g
e
m
e
n
t
f
r
a
m
e
w
o
r
k
f
o
r
GP
U
-
C
P
U
h
ete
r
o
g
en
eo
u
s
ar
ch
itect
u
r
es.
O
u
r
s
o
lu
tio
n
f
ea
t
u
r
es
a
t
w
o
-
tier
d
es
ig
n
.
I
n
t
h
e
first
ti
er
,
Gr
ee
n
GP
U
d
y
n
a
m
ical
l
y
s
p
lits
a
n
d
d
is
tr
ib
u
tes
w
o
r
k
lo
ad
s
to
GP
U
a
n
d
C
P
U
b
ased
o
n
th
e
w
o
r
k
lo
ad
ch
ar
ac
ter
is
tics
,
s
u
ch
t
h
at
b
o
th
s
id
e
s
ca
n
fi
n
i
s
h
ap
p
r
o
x
i
m
a
tel
y
at
th
e
s
a
m
e
ti
m
e.
A
s
a
r
esu
lt,
t
h
e
en
er
g
y
w
asted
o
n
s
t
a
y
in
g
id
le
a
n
d
w
a
iti
n
g
f
o
r
th
e
s
lo
w
er
s
id
e
to
fin
i
s
h
i
s
m
in
i
m
i
ze
d
.
I
n
th
e
s
ec
o
n
d
tier
,
Gr
ee
n
GP
U
d
y
n
a
m
icall
y
th
r
o
ttles
th
e
f
r
eq
u
e
n
cie
s
o
f
GP
U
co
r
es
an
d
m
e
m
o
r
y
i
n
a
co
o
r
d
in
ated
m
an
n
er
,
b
ased
o
n
t
h
eir
u
tili
za
t
io
n
s
,
f
o
r
m
ax
i
m
ized
en
er
g
y
s
a
v
i
n
g
s
w
it
h
o
n
l
y
m
ar
g
i
n
al
p
er
f
o
r
m
an
ce
d
eg
r
ad
atio
n
.
L
i
k
e
w
is
e,
th
e
f
r
eq
u
e
n
c
y
an
d
v
o
ltag
e
o
f
t
h
e
C
P
U
ar
e
s
ca
le
d
s
i
m
i
lar
l
y
.
W
e
i
m
p
le
m
e
n
t
G
r
ee
n
GP
U
u
s
in
g
t
h
e
C
UD
A
f
r
a
m
e
w
o
r
k
o
n
a
r
ea
l
p
h
y
s
ical
tes
ted
w
it
h
Nv
id
ia
GeFo
r
ce
GP
Us
an
d
A
MD
P
h
en
o
m
I
I
C
P
Us.
E
x
p
er
i
m
e
n
t
r
es
u
lt
s
w
i
th
s
ta
n
d
ar
d
R
o
d
in
ia
b
en
c
h
m
ar
k
s
s
h
o
w
t
h
at
Gr
ee
n
GP
U
ac
h
ie
v
e
s
2
1
.
0
4
%
av
er
ag
e
en
er
g
y
s
a
v
in
g
s
an
d
o
u
tp
er
f
o
r
m
s
e
v
er
al
w
ell
-
d
esi
g
n
ed
b
asel
in
es.
RE
F
E
R
E
NC
E
S
[1
]
Ca
v
iu
m
o
c
teo
n
p
ro
c
e
ss
o
r
f
a
m
il
y
.
h
tt
p
:
//
w
ww
.
c
a
v
iu
m
n
e
t
w
o
rk
s.c
o
m
/OCT
EONMI
P
S
6
4
.
h
tm
l.
[2
]
IBM
Blad
e
Ce
n
ter S
y
ste
m
.
h
tt
p
:/
/
ww
w
-
0
3
.
ib
m
.
c
o
m
/s
y
ste
m
s/b
lad
e
c
e
n
ter.
[3
]
IBM
P
OW
ER7
S
y
ste
m
s.
h
tt
p
:/
/w
ww
-
0
3
.
ib
m
.
c
o
m
/s
y
ste
m
s/p
o
w
e
r
.
[4
]
In
tel
ix
p
2
x
x
x
p
ro
d
u
c
t
l
in
e
o
f
n
e
tw
o
rk
p
ro
c
e
ss
o
rs.
h
tt
p
:
//
in
tel.
c
o
m
/d
e
sig
n
/n
e
tw
o
rk
/p
ro
d
u
c
ts/n
p
f
a
m
il
y
/i
n
d
e
x
.
h
tm
l
.
[5
]
In
tel
x
e
o
n
m
a
c
h
in
e
.
h
t
tp
:
//
w
ww
.
I
n
tel.
c
o
m
/X
e
o
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8694
I
n
t J
P
o
w
er
E
lectr
o
n
&
Dr
i S
y
s
t
,
Vo
l.
9
,
No
.
1
,
Ma
r
ch
2
0
1
8
:
80
–
88
88
[6
]
Ch
e
n
,
D.
Zh
a
n
g
a
n
d
Z
.
W
a
n
g
,
“
Re
se
a
rc
h
o
f
th
e
h
e
tero
g
e
n
e
o
u
s m
u
lt
i
-
Co
re
p
r
o
c
e
ss
o
r
a
rc
h
it
e
c
tu
re
d
e
s
ig
n
[
J
]
”
,
Co
mp
u
ter
E
n
g
in
e
e
rin
g
a
n
d
S
c
ien
c
e
,
v
o
l.
3
3
,
n
o
.
1
2
,
(
2
0
1
1
),
p
p
.
2
7
-
36.
[7
]
S
K.
Ba
ru
a
h
,
“
Pa
rtit
io
n
in
g
re
a
l
-
t
i
me
ta
sk
s a
mo
n
g
h
e
ter
o
g
e
n
e
o
u
s m
u
lt
ip
r
o
c
e
ss
o
rs
”
,
In
:
P
r
o
c
.
o
f
th
e
2
0
0
4
In
tern
a
ti
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
P
a
ra
l
lel
P
ro
c
e
ss
in
g
,
IC
P
P
,
T
o
ro
n
to
,
Ca
n
a
d
a
,
(2
0
0
4
),
p
p
.
4
6
7
-
4
7
4
.
[8
]
R.
L
i,
Y.
L
iu
a
n
d
X
.
Ch
e
n
g
,
“
A
S
u
rv
e
y
o
f
tas
k
sc
h
e
d
u
li
n
g
re
se
a
rc
h
p
ro
g
re
ss
o
n
m
u
lt
ip
ro
c
e
ss
o
r”
,
J
o
u
r
n
a
l
o
f
Co
mp
u
ter
Res
e
a
rc
h
a
n
d
De
v
e
lo
p
me
n
t
,
v
o
l.
4
5
,
n
o
.
9
,
(
2
0
0
8
)
,
p
p
.
1
6
2
0
-
1
6
2
9
.
[9
]
J.
L
i
a
n
d
S
.
Jin
,
“
Re
se
a
rc
h
o
n
sta
t
ic t
a
sk
sc
h
e
d
u
li
n
g
stra
teg
y
b
a
se
d
o
n
h
e
tero
g
e
n
e
o
u
s m
u
lt
i
-
c
o
re
p
ro
c
e
ss
o
rs [
J
]
”
,
Co
mp
u
ter
E
n
g
in
e
e
rin
g
a
n
d
De
sig
n
,
v
o
l.
3
4
,
n
o
.
1
,
(2
0
1
3
),
p
p
.
1
7
8
-
1
8
4
.
[1
0
]
J.
Jia
n
g
,
“
Re
se
a
rc
h
o
n
e
m
b
e
d
d
e
d
so
f
t
w
a
re
k
e
y
issu
e
s o
f
h
e
tero
g
e
n
e
o
u
s m
u
lt
i
-
c
o
re
p
r
o
c
e
ss
o
r
[
D
]
”
,
Ch
o
n
g
Qin
g
Un
iv
e
rsit
y
,
(2
0
1
1
).
[1
1
]
M
S
h
a
n
m
u
g
a
su
n
d
a
ra
m
,
R.
Ku
m
a
r
a
n
d
Ha
rish
M
Kitt
u
r
“
P
e
rf
o
rm
a
n
c
e
A
n
a
l
y
sis o
f
P
re
e
m
p
ti
v
e
B
a
se
d
Un
ip
r
o
c
e
ss
o
r
S
c
h
e
d
u
l
in
g
”
in
th
e
I
n
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
El
e
c
trica
l
a
n
d
C
o
mp
u
t
e
r E
n
g
i
n
e
e
rin
g
(
IJ
ECE
)
Vo
l.
6
,
No
.
4
,
A
u
g
u
st
2
0
1
6
,
p
p
.
1
4
8
9
-
1
4
9
8
[1
2
]
M
e
d
h
a
t
H A
w
a
d
a
ll
a
“
P
ro
c
e
ss
o
r
S
p
e
e
d
Co
n
tro
l
f
o
r
P
o
w
e
r
Re
d
u
c
ti
o
n
o
f
Re
a
l
-
T
i
m
e
S
y
ste
m
s”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
El
e
c
trica
l
a
n
d
C
o
mp
u
ter
En
g
in
e
e
rin
g
(
IJ
ECE
)
V
o
l
.
5
,
No
.
4
,
A
u
g
u
st 2
0
1
5
,
p
p
.
7
0
1
-
7
1
3
.
[1
3
]
S
.
A
lb
e
rs,
“
En
e
rg
y
-
e
ff
icie
n
t
a
lg
o
rit
h
m
s,”
Co
m
m
u
n
.
A
CM
,
v
o
l.
5
3
,
n
o
.
5
,
p
p
.
8
6
–
9
6
,
M
a
y
2
0
1
0
.
[
O
n
l
in
e
].
Av
a
il
a
b
le:
h
tt
p
:/
/d
o
i.
a
c
m
.
o
rg
/1
0
.
1
1
4
5
/1
7
3
5
2
2
3
.
1
7
3
5
2
4
5
.
[1
4
]
P. C
ich
o
w
sk
i,
J.
Ke
ll
e
r,
a
n
d
C.
K
e
ss
ler,
“
En
e
rg
y
-
e
ff
icie
n
t
ma
p
p
i
n
g
o
f
t
a
sk
c
o
ll
e
c
ti
o
n
s o
n
t
o
ma
n
y
c
o
re
p
ro
c
e
ss
o
rs
,
”
in
P
r
o
c
.
5
th
S
w
e
d
ish
W
o
rk
sh
o
p
o
n
M
u
li
tco
re
C
o
m
p
u
ti
n
g
(M
CC
2
0
1
2
),
2
0
1
2
.
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
M
rs L
a
v
a
n
y
a
Dh
a
n
e
sh
re
c
e
iv
e
d
h
e
r
B.
E.
d
e
g
re
e
in
E
lec
tri
c
a
l
a
n
d
E
lec
tro
n
ics
En
g
i
n
e
e
rin
g
f
ro
m
Bh
a
ra
th
iy
a
r
Un
iv
e
rsit
y
a
t
2
0
0
2
.
S
h
e
c
o
m
p
lete
d
h
e
r
M
.
E.
i
n
Em
b
e
d
d
e
d
S
y
ste
m
De
si
g
n
f
ro
m
A
n
n
a
Un
iv
e
rsit
y
,
Ch
e
n
n
a
i
a
t
2
0
0
9
.
S
h
e
is a Res
e
a
rc
h
S
c
h
o
lar at
S
a
th
y
a
b
a
m
a
Un
iv
e
rsit
y
a
n
d
c
u
rre
n
tl
y
w
o
rk
in
g
a
t
P
a
n
im
a
lar In
stit
u
te o
f
T
e
c
h
n
o
lo
g
y
,
Ch
e
n
n
a
i
Dr
P
.
Mu
r
u
g
e
s
an
h
a
s
d
o
n
e
h
is
s
p
ec
ializatio
n
in
P
o
w
er
S
y
s
te
m
s
.
He
is
cu
r
r
en
t
l
y
w
o
r
k
i
n
g
as
a
p
r
o
f
es
s
o
r
/EE
E
at
S.
A
.
E
n
g
i
n
e
er
i
n
g
co
lleg
e,
C
h
e
n
n
ai.
He
g
i
v
es
r
esear
ch
id
ea
s
a
n
d
m
u
c
h
i
n
ter
ested
i
n
th
e
f
ield
o
f
lo
w
-
p
o
w
er
co
m
p
u
ti
n
g
,
f
a
u
lt
to
ler
an
ce
a
n
d
r
ea
l
-
ti
m
e
e
m
b
ed
d
ed
s
y
s
te
m
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.