I
A
E
S
I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
r
t
if
ic
ia
l
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
A
I
)
V
ol
.
10
, N
o.
1
,
M
a
r
c
h
2021
, pp.
244
~
252
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
10
.i
1
.pp
244
-
252
244
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
V
i
r
t
u
al
m
ac
h
i
n
e
m
i
gr
at
i
on
i
n
M
E
C
b
ase
d
ar
t
i
f
i
c
i
al
i
n
t
e
l
l
i
ge
n
c
e
t
e
c
h
n
i
q
u
e
A
li
O
U
A
C
H
A
1
,
M
oh
am
e
d
E
L
G
h
m
ar
y
2
1
Department of Compu
ter Science, Mohamed V Uni
versity, Facult
y of Sciences, Rabat,
Morocco
2
Department of Compu
ter Science, SidiM
ohamed Ben Abdellah Un
iversity, FS
DM, Fez, Morocco
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
O
c
t
2
7
, 20
20
R
e
vi
s
e
d
J
a
n
2
7
, 20
21
A
c
c
e
pt
e
d
F
eb
2
2
, 20
21
The
whole
world
is
inundated
with
smaller
devices
equipped
with
wireless
communi
cation i
nterfaces. At
the sam
e time,
the amou
nt
of data g
ener
ated by
these
devices
is
becoming
more
important.
The
smaller
size
of
th
ese
devices
has
the
disadvantage
of
being
s
hort
of
processing
and
storage
re
sources
(memory,
processes,
energy,...),
especially
when
it
needs
to
process
larger
amounts
of
data.
In
order
to
overcome
this
weakness
and
process
massive
data,
devices
must
help
each
other.
A
low
-
resource
node
can
deleg
at
e
the
execution
of
a
set
of
computi
only
heavy
tasks
to
another
machine
in
the
network
to
process
them
for
it.
The
machine
with
sufficient
comput
ational
resources
must
also
deposit
the
appropriate
environment
represented
by
the
adapted
virtual
machine.
Thus
,
in
this
paper,
in
order
to
migrate
the
virtual
machine to
an edge
server
in a mobile
edge co
mputing
environmen
t,
we have
proposed
an
approach
based
on
artificial
intelligence.
More
specifica
lly,
the
main
idea
of
this
paper
is
to
cut
a
virtual
machine
int
o
several
small
pieces
and
then
send
them
to
an
appropriate
target
node
(Edge
Server)
using
the
ant
colony
algorithm.
In
order
to
test
and
prove
the
effectiveness
of
our
approach,
several
simulat
ions
are
made
by
NS3.
The
obtained
result
s
show
that our a
ppr
oach is well adapted to mobile environments.
K
e
y
w
o
r
d
s
:
A
nt
c
ol
onny a
lg
or
it
hm
e
I
nt
e
ll
ig
e
nc
e
a
r
ti
f
ic
ia
l
M
ig
r
a
ti
on vir
tu
a
l
m
a
c
hi
ne
s
M
obi
le
e
dge
c
om
put
in
g
R
out
in
g pr
ot
oc
ol
e
This is an
open
acce
ss artic
le unde
r the
CC BY
-
SA
license.
C
or
r
e
s
pon
di
n
g A
u
th
or
:
A
li
O
U
A
C
H
A
I
nt
e
ll
ig
e
nt
P
r
oc
e
s
s
in
g S
ys
te
m
s
&
S
e
c
ur
it
y T
e
a
m
(
I
P
S
S
)
C
om
put
e
r
S
c
ie
nc
e
D
e
pa
r
tm
e
nt
M
oha
m
e
d V
U
ni
ve
r
s
it
y, F
a
c
ul
ty
of
S
c
ie
nc
e
s
4 A
ve
nue
I
bn B
a
tt
out
a
, B
.P
. 1014 R
P
, R
a
b
a
t,
M
or
oc
c
o
E
m
a
il
:
a
li
.oua
c
ha
@
um
5.a
c
.m
a
1.
I
N
T
R
O
D
U
C
T
I
O
N
T
he
gr
ow
in
g
de
ve
lo
pm
e
nt
in
f
ie
ld
s
of
c
om
put
in
g
a
nd
e
le
c
tr
oni
c
s
ha
s
gi
ve
n
r
is
e
to
s
m
a
ll
e
r
m
a
c
hi
ne
s
w
it
h
w
ir
e
le
s
s
c
om
m
uni
c
a
ti
on
in
te
r
f
a
c
e
s
a
nd
e
ne
r
gy
a
ut
onomy
.
T
he
in
te
r
c
onne
c
ti
on
of
th
e
s
e
m
a
c
hi
ne
s
,
th
e
a
bi
li
ty
to
m
ove
f
r
e
e
ly
a
nd
th
e
a
bi
li
ty
to
s
e
lf
-
or
ga
ni
z
e
c
on
s
ti
tu
t
e
s
a
ne
twor
k
known
by
m
obi
le
a
dhoc
ne
twor
k
a
bbr
e
vi
a
te
d
a
s
M
A
N
E
T
[
1
-
2
]
.
I
n
a
ddi
ti
on,
th
e
a
m
ount
s
of
da
ta
ge
ne
r
a
te
d
by
e
a
c
h
de
vi
c
e
a
nd
th
e
de
m
a
nds
of
pr
oc
e
s
s
in
g
r
e
s
our
c
e
s
ha
ve
a
ls
o
be
c
om
e
in
c
r
e
a
s
in
gl
y
la
r
ge
.
H
o
w
e
ve
r
,
th
e
s
m
a
ll
e
r
s
iz
e
of
th
e
s
e
s
ophi
s
ti
c
a
te
d
de
vi
c
e
s
po
s
e
s
th
e
c
on
s
tr
a
in
t
of
la
c
k
of
r
e
s
our
c
e
s
a
nd
m
or
e
p
a
r
ti
c
ul
a
r
ly
,
th
os
e
th
a
t
a
r
e
r
e
la
t
e
d
to
c
om
pu
ti
ng
a
nd
s
to
r
a
ge
.
O
ne
of
th
e
s
ol
ut
io
ns
c
ons
is
t
s
in
pl
a
c
in
g
a
m
or
e
pow
e
r
f
ul
m
a
c
hi
ne
(
E
dge
S
e
r
ve
r
)
,
in
te
r
m
s
of
pr
oc
e
s
s
in
g
r
e
s
our
c
e
s
,
in
th
e
pe
r
ip
he
r
y
of
th
e
ne
twor
k.
T
hi
s
s
ol
ut
io
n
f
a
ll
s
w
it
hi
n
th
e
c
onc
e
pt
of
m
obi
le
e
dge
c
om
put
in
g
(
M
E
C
)
[
3
-
6
]
,
il
lu
s
tr
a
te
d
by
F
ig
ur
e
1.
H
ow
e
ve
r
,
i
n
or
de
r
to
be
a
bl
e
to
pr
oc
e
s
s
in
f
or
m
a
ti
on
or
pe
r
f
or
m
a
s
e
t
of
ta
s
ks
c
om
in
g
f
r
om
a
ny
node
in
th
e
ne
twor
k,
th
e
e
dge
s
e
r
ve
r
(
E
S
)
m
us
t
ha
v
e
th
e
v
ir
tu
a
l
m
a
c
hi
ne
(
V
M
)
r
e
pr
e
s
e
nt
in
g
th
e
e
nvi
r
onm
e
nt
f
or
pr
oc
e
s
s
in
g
or
e
xe
c
ut
in
g
th
e
s
e
t
a
s
ks
.
T
hus
,
th
e
m
a
in
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
V
ir
tu
al
m
ac
hi
ne
m
ig
r
at
io
n i
n M
E
C
bas
e
d ar
ti
fi
c
ia
l
in
te
ll
ig
e
nc
e
te
c
hni
que
(
A
li
O
U
A
C
H
A
)
245
obj
e
c
ti
ve
of
t
hi
s
w
or
k
i
s
t
o
f
in
d a
t
e
c
hni
que
t
ha
t
a
ll
ow
s
node
s
i
n t
he
ne
twor
k t
o
s
e
nd t
he
ir
v
ir
tu
a
l
m
a
c
hi
ne
s
t
o
ot
he
r
r
e
s
our
c
e
-
r
ic
h
node
s
.
T
o
th
i
s
e
nd,
w
e
ha
v
e
pr
opos
e
d
a
n
a
ppr
oa
c
h
th
a
t
is
ba
s
e
d
on
a
n
a
r
ti
f
ic
ia
l
in
te
ll
ig
e
nc
e
(
A
I
)
te
c
hni
que
.
T
hus
,
by
a
ppl
yi
ng
th
e
a
nt
c
ol
ony
a
lg
or
it
hm
,
our
pr
in
c
ip
a
l
a
im
is
to
m
ig
r
a
te
a
vi
r
tu
a
l
m
a
c
hi
ne
,
w
hi
c
h
is
c
ons
id
e
r
e
d
a
s
la
r
ge
a
m
ount
of
da
ta
,
by
di
vi
di
ng
it
in
to
s
m
a
ll
pi
e
c
e
s
tr
a
ns
por
te
d
th
r
ough
s
e
ve
r
a
l
ot
he
r
node
s
to
a
de
s
ti
na
ti
on
node
(
e
dge
s
e
r
ve
r
)
.
T
o
im
pl
e
m
e
nt
th
is
a
ppr
oa
c
h,
w
e
w
il
l
a
ls
o
pr
of
it
f
r
om
th
e
a
dva
nt
a
ge
s
th
e
O
L
S
R
pr
ot
oc
ol
[
7
-
8
]
.
F
ol
lo
w
in
g
it
s
pr
oa
c
ti
ve
pr
in
c
ip
le
,
th
e
c
ont
r
ol
m
e
s
s
a
ge
s
e
xc
ha
nge
d be
twe
e
n t
he
node
s
s
e
r
ve
, on the
one
ha
nd, t
o di
s
s
e
m
in
a
te
i
nf
or
m
a
ti
on o
n
t
he
pr
oc
e
s
s
in
g c
a
pa
c
it
ie
s
a
nd, on the
ot
he
r
ha
nd, t
o de
c
id
e
on t
he
pa
th
s
a
dopt
e
d t
o s
e
nd t
he
c
om
pos
it
e
pi
e
c
e
s
of
t
he
vi
r
tu
a
l
m
a
c
hi
ne
.
T
he
r
e
s
t
of
th
e
pa
pe
r
i
s
or
ga
ni
z
e
d
a
s
:
s
e
c
ti
on
2
i
s
a
bout
a
li
te
r
a
t
ur
e
r
e
vi
e
w
. T
he
pr
obl
e
m
f
or
m
ul
a
ti
on
a
nd
it
s
s
ol
ut
io
n
a
r
e
di
s
c
u
s
s
e
d
in
s
e
c
ti
on
3.
T
h
e
s
im
ul
a
ti
on
r
e
s
ul
ts
a
r
e
di
s
c
u
s
s
e
d
in
s
e
c
ti
on
4.
F
in
a
ll
y,
th
e
c
onc
lu
s
io
n of
t
hi
s
pa
pe
r
i
s
in
s
e
c
ti
on
5.
F
ig
ur
e
1.
M
obi
le
e
dge
c
om
put
in
g (
M
E
C
)
2.
O
L
S
R
O
V
E
R
V
I
E
W
O
pt
im
iz
e
d
li
nk
s
ta
te
r
out
in
g
pr
ot
oc
ol
(
O
L
S
R
)
[
7
]
is
a
r
out
in
g
pr
ot
oc
ol
ge
ne
r
a
ll
y
us
e
d
in
M
A
N
E
T
[
1
]
ne
twor
ks
.
O
L
S
R
is
a
pr
oa
c
ti
ve
,
li
nk
-
s
ta
te
opt
im
iz
e
d
r
out
in
g
pr
ot
oc
ol
th
a
t
c
r
e
a
te
s
a
r
out
in
g
ta
bl
e
c
ont
a
in
in
g
r
out
e
s
to
a
ll
node
s
in
th
e
ne
two
r
k
a
he
a
d
of
ti
m
e
a
nd
e
ve
n
be
f
or
e
a
n
a
c
tu
a
l
da
ta
t
r
a
ns
m
is
s
io
n
r
e
que
s
t
oc
c
ur
s
.
I
t
us
e
s
c
ont
r
ol
m
e
s
s
a
ge
s
H
E
L
L
O
a
nd
to
pol
ogy
c
ont
r
ol
(
T
C
)
to
pr
opa
ga
te
in
f
or
m
a
ti
on
a
bout
node
s
a
nd
li
nk
s
ta
t
e
s
to
th
e
ne
ig
hbor
hood
a
nd
to
th
e
e
nt
ir
e
ne
twor
k.
T
ha
nks
to
th
is
e
xc
ha
ng
e
of
m
e
s
s
a
g
e
s
,
e
a
c
h
node
obt
a
in
s
in
f
or
m
a
ti
on
a
bout
to
pol
ogy
of
th
e
ne
twor
k
a
nd
f
e
e
ds
it
s
r
out
in
g
ta
bl
e
by
th
e
s
hor
te
s
t
p
a
th
s
to
a
ll
t
he
ot
he
r
node
s
.
T
he
O
L
S
R
pr
e
s
e
nt
s
a
n opti
m
iz
e
d ve
r
s
io
n of
t
he
ope
n s
hor
te
s
t
p
a
th
f
ir
s
t
(
O
S
P
F
)
[
9
]
li
nk
-
s
ta
te
r
out
in
g
pr
ot
oc
ol
.
I
nde
e
d,
in
s
te
a
d
of
us
in
g
th
e
f
lo
odi
ng
m
e
c
ha
ni
s
m
to
br
oa
dc
a
s
t
in
f
or
m
a
ti
on
a
bout
th
e
to
pol
ogy
of
th
e
ne
twor
k t
hr
ough T
C
m
e
s
s
a
g
e
s
, a
s
O
S
P
F
doe
s
, j
us
t
node
s
f
r
om
a
n e
le
c
te
d s
ub
s
e
t
a
m
ong ne
ig
hbor
s
of
t
he
node
w
is
hi
ng
to
br
oa
dc
a
s
t
or
r
e
br
oa
dc
a
s
t
th
e
s
e
m
e
s
s
a
ge
s
w
hi
c
h a
r
e
a
ut
hor
iz
e
d
to
r
e
tr
a
ns
m
it
th
e
m
. T
he
e
le
m
e
nt
s
of
th
is
s
ubs
e
t
a
r
e
c
a
ll
e
d
m
ul
ti
poi
nt
r
e
la
i
(
M
P
R
)
a
nd
e
a
c
h
node
of
th
e
ne
twor
k
s
e
le
c
ts
it
s
M
P
R
s
f
r
om
it
s
s
ym
m
e
tr
ic
one
hop
n
e
ig
hbor
s
s
o
a
s
to
c
on
s
tr
uc
t
th
e
m
o
s
t
m
i
ni
m
a
l
s
ubs
e
t
w
hi
c
h
c
ov
e
r
s
a
ll
of
th
e
two
-
hop
ne
ig
hbor
s
(
G
r
e
e
n
c
ol
or
node
s
in
F
ig
ur
e
2
(
d
)
)
.
I
t
s
houl
d
a
ls
o
be
not
e
d
th
a
t
H
E
L
L
O
m
e
s
s
a
ge
s
a
r
e
n
e
ve
r
r
e
tr
a
ns
m
it
te
d.
I
t
s
e
r
ve
s
f
or
di
s
c
ove
r
in
g
th
e
ne
ig
hb
or
hood
a
nd
a
ls
o
th
e
two
-
hop
ne
ig
hbor
hood.
I
n
f
a
c
t,
e
a
c
h
O
L
S
R
node
e
xt
r
a
c
ts
th
e
in
f
or
m
a
ti
on
on
th
e
node
s
of
th
e
f
ir
s
t
ne
ig
hbor
hood
a
nd
th
e
node
s
of
th
e
s
e
c
ond
ne
ig
hbor
hood
vi
a
r
e
c
e
iv
e
d
H
E
L
L
O
m
e
s
s
a
ge
.
I
n
a
ddi
ti
on,
th
e
node
w
hi
c
h
r
e
c
e
iv
e
d
th
e
H
E
L
L
O
m
e
s
s
a
ge
ca
lc
ul
a
te
s
th
e
node
w
hi
c
h
pr
ovi
de
s
th
e
be
s
t
pa
th
to
th
e
two
-
hop
node
s
a
m
ong
th
e
node
s
of
it
s
f
i
r
s
t
ne
ig
hbor
hood a
nd s
e
le
c
ts
i
t
a
s
t
he
M
P
R
. I
n t
hi
s
w
a
y, e
a
c
h node
bui
ld
s
a
s
e
t
of
M
P
R
s
.
I
n
or
de
r
to
f
u
r
th
e
r
il
lu
s
tr
a
te
th
e
O
L
S
R
pr
ot
oc
ol
ope
r
a
ti
ng,
th
e
F
ig
ur
e
2
gi
ve
s
th
e
s
tr
uc
tu
r
e
of
th
e
m
a
in
c
ont
r
ol
m
e
s
s
a
ge
s
us
e
d
w
he
n
a
ll
node
s
of
th
e
ne
twor
k
a
r
e
e
qui
pe
d
w
it
h
a
in
gl
e
in
te
r
f
a
c
e
a
nd
us
e
O
L
S
R
a
s
th
e
r
out
in
g
pr
ot
oc
ol
a
s
s
how
n
in
F
ig
ur
e
2
(
a
-
b
).
T
hi
s
f
ig
ur
e
a
ls
o
s
how
s
th
e
ti
m
e
in
te
r
va
ls
th
a
t
s
e
pa
r
a
t
e
th
e
s
e
ndi
ng
of
two
c
ons
e
c
ut
iv
e
m
e
s
s
a
ge
s
,
w
he
th
e
r
f
or
H
E
L
L
O
m
e
s
s
a
ge
s
or
f
or
T
C
a
s
s
how
n
in
F
ig
ur
e
2
(
c
)
m
e
s
s
a
ge
s
.
H
E
L
L
O
m
e
s
s
a
ge
s
a
r
e
s
e
nt
e
ve
r
y
two
s
e
c
onds
w
hi
le
T
C
m
e
s
s
a
ge
s
a
r
e
s
e
nt
e
ve
r
y
5
s
e
c
onds
.
I
t
a
ls
o
e
xpl
a
in
s
th
e
di
f
f
e
r
e
nt
c
a
te
gor
ie
s
of
node
s
a
s
w
e
ll
a
s
th
e
l
in
ks
t
ha
t
e
xi
s
t
be
twe
e
n
th
e
m
a
nd
th
e
ir
r
e
la
ti
ons
hi
p
w
it
h
th
e
node
(
c
ons
id
e
r
a
s
a
n
e
xa
m
pl
e
)
c
onc
e
r
ne
d
by
th
e
e
xe
c
ut
io
n
of
th
e
O
L
S
E
pr
ot
oc
ol
a
s
s
how
n
F
ig
ur
e
2
(
d
)
.
E
nde
d,
th
e
ope
r
a
ti
on
o
f
th
e
O
L
S
R
p
r
ot
oc
ol
is
ba
s
e
d
on
th
e
pe
r
io
d
ic
e
xc
ha
nge
of
c
ont
r
ol
m
e
s
s
a
ge
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
10
, N
o.
1
,
M
a
r
c
h
20
2
1
:
244
–
252
246
(
a
)
(
b)
(
c
)
(
d)
F
ig
ur
e
2
.
O
pt
im
iz
e
d l
in
k s
ta
te
r
out
in
g pr
ot
oc
ol
(
O
L
S
R
)
.
(
a
)
H
E
L
L
O
m
e
s
s
a
ge
s
tr
uc
tu
r
e
,
(
b
)
T
C
m
e
s
s
a
g
e
s
tr
uc
tu
r
e
,
(
c
)
H
E
L
L
O
a
nd T
C
m
e
s
s
a
ge
s
t
im
e
t
r
ns
m
is
s
io
n,
(
d
)
O
L
S
R
node
s
t
ype
3.
R
E
L
A
T
E
D
WO
R
K
S
I
n
th
is
s
e
c
ti
on,
w
e
w
il
l
pr
e
s
e
nt
s
om
e
pr
e
vi
ous
r
e
s
e
a
r
c
hs
on
th
e
m
ig
r
a
ti
on
of
vi
r
tu
a
l
m
a
c
hi
ne
s
(
)
to
a
n e
dge
s
e
r
ve
r
(
E
S
)
i
n a
m
obi
le
e
dge
c
om
put
in
g (
M
E
C
)
e
nvi
r
onm
e
nt
.
I
nde
e
d, t
he
a
ut
hor
s
of
[
10
]
pr
opos
e
a
m
e
th
odol
ogy
ba
s
e
d
on
tr
a
f
f
ic
a
nd
ge
ogr
a
phi
c
lo
c
a
ti
ons
of
da
t
a
c
e
nt
e
r
s
to
m
e
e
t
c
us
to
m
e
r
de
m
a
nds
f
or
VM
m
ig
r
a
ti
on
c
a
ll
e
d
v
ir
tu
a
l
m
a
c
hi
ne
m
ig
r
a
ti
on
a
ppr
oa
c
h.
T
he
a
lg
or
it
hm
us
e
d
he
r
e
i
s
r
un
a
t
r
e
gul
a
r
in
te
r
va
ls
to
c
he
c
k
f
or
ne
twor
k
tr
a
f
f
ic
.
I
t
a
ls
o
c
he
c
ks
th
e
di
s
ta
nc
e
of
th
e
da
t
a
c
e
nt
e
r
s
w
he
r
e
th
e
c
us
to
m
e
r
'
s
r
e
que
s
t
is
to
be
pl
a
c
e
d.
I
n
[
11
]
,
th
e
a
ut
hor
s
pr
opos
e
a
m
e
th
od
of
m
ig
r
a
ti
on
of
V
M
,
na
m
e
d
A
da
-
T
hi
ng
s
,
w
hi
c
h
is
de
te
r
m
in
e
d
by
it
s
w
or
kl
oa
d
c
ha
r
a
c
te
r
is
ti
c
s
.
S
pe
c
if
ic
a
ll
y,
ba
s
e
d
on
th
e
va
r
ia
ti
on
in
th
e
c
ur
r
e
nt
r
a
te
of
th
ir
ty
pa
ge
s
of
m
e
m
or
y
in
I
oT
a
pp
li
c
a
ti
ons
,
A
da
-
T
hi
ngs
c
a
n
a
da
pt
iv
e
ly
s
e
le
c
t
th
e
m
os
t
s
ui
ta
bl
e
m
ig
r
a
ti
on
m
e
th
od
to
c
opy
th
e
m
e
m
or
y
pa
ge
s
.
T
he
a
ut
or
s
of
[
12
]
pr
opos
e
d
a
n
a
d
a
pt
iv
e
pr
e
-
pa
gi
ng
to
e
li
m
in
a
te
dupl
ic
a
te
p
a
ge
tr
a
ns
m
is
s
io
n
a
nd
dyna
m
ic
s
e
lf
-
ba
ll
ooni
ng
to
a
voi
d
th
e
tr
a
ns
f
e
r
of
f
r
e
e
m
e
m
or
y
pa
ge
s
.
T
he
a
ut
hor
s
of
[
13
]
s
ugge
s
t
a
f
r
a
m
e
w
or
k
f
or
th
e
a
ll
oc
a
ti
on
a
nd
m
ig
r
a
ti
on
of
vi
r
tu
a
l
m
a
c
hi
ne
s
th
a
t
e
xpl
oi
ts
p
e
r
f
or
m
a
nc
e
-
to
-
pow
e
r
r
a
ti
os
(
P
P
R
s
)
f
or
d
if
f
e
r
e
nt
ty
pe
s
of
hos
ts
.
B
y
a
c
hi
e
vi
ng
th
e
opt
im
um
ba
la
nc
e
be
twe
e
n
hos
t
us
a
ge
a
nd
pow
e
r
c
ons
um
pt
io
n,
th
is
f
r
a
m
e
w
or
k
is
a
bl
e
to
e
ns
ur
e
th
a
t
ho
s
ts
a
r
e
ope
r
a
ti
ng
a
t
th
e
m
o
s
t
e
ne
r
gy
e
f
f
ic
ie
nt
us
a
g
e
le
ve
ls
.
I
n
th
e
r
e
s
e
a
r
c
h
m
a
de
by
[
14
]
,
th
e
a
ut
hor
s
pr
opos
e
a
vi
r
tu
a
l
m
a
c
hi
ne
m
ig
r
a
ti
on
s
tr
a
te
gy
w
hi
c
h
a
im
s
to
r
e
duc
e
ne
twor
k
c
onge
s
ti
on
in
a
M
E
C
e
nvi
r
onm
e
nt
a
nd
th
u
s
i
m
pr
ove
Q
oS
.
I
t
opt
s
f
or
ba
ndw
id
th
s
e
ns
it
iv
e
a
ppl
ic
a
ti
ons
to
b
e
of
f
lo
a
de
d
to
e
dge
s
e
r
ve
r
s
to
be
e
x
e
c
ut
e
d
by
V
M
on
th
o
s
e
e
dg
e
s
e
r
ve
r
s
w
it
h
m
in
im
a
l
c
os
t
s
to
s
ur
r
ounding us
e
r
s
. I
n
or
de
r
to
i
m
pr
ove
t
he
Q
oS
s
uc
h a
s
T
C
P
t
hr
oughput, t
he
a
ut
hor
s
of
[
15
]
pr
opos
e
a
V
M
m
ig
r
a
ti
on
m
e
th
od
th
a
t
r
out
e
s
a
V
M
f
r
om
one
c
onge
s
te
d
node
to
a
not
he
r
node
of
a
m
obi
le
de
vi
c
e
.
I
n
th
is
m
e
th
od,
us
e
r
s
c
a
n
c
hoos
e
a
le
s
s
c
onge
s
te
d
node
e
ve
n
if
it
is
f
a
r
a
w
a
y
in
s
te
a
d
of
a
c
lo
s
e
r
but
c
onge
s
te
d
one
.
T
he
c
hoi
c
e
i
s
m
a
d
e
a
c
c
or
di
ng
to
a
n
e
xpe
c
te
d
T
C
P
s
pe
e
d.
T
he
a
ut
hor
s
of
[
16
]
pr
opos
e
a
s
e
a
m
le
s
s
li
ve
V
M
m
ig
r
a
ti
on
be
twe
e
n
ne
ig
hbor
in
g
c
lo
udl
e
ts
w
it
h
th
e
goa
l
of
e
li
m
in
a
ti
ng
th
e
de
la
y
c
a
us
e
d
by
s
e
r
vi
c
e
in
it
ia
ti
on
ti
m
e
a
f
te
r
m
ovi
ng
a
w
a
y
f
r
om
th
e
c
lo
udl
e
t.
A
s
e
a
m
le
s
s
li
ve
V
M
m
ig
r
a
ti
on
is
a
c
hi
e
ve
d
w
it
h
th
e
pr
io
r
knowle
dge
of
th
e
I
P
a
ddr
e
s
s
of
th
e
V
M
be
in
g
m
ig
r
a
te
d
in
th
e
d
e
s
ti
na
ti
on
c
lo
udl
e
t
a
nd
m
or
e
im
por
ta
nt
ly
w
it
h
m
ul
ti
pa
th
T
C
P
.
F
or
th
e
pur
pos
e
of
opt
im
iz
in
g
ne
two
r
k
a
nd
c
om
put
in
g
r
e
s
our
c
e
s
,
s
e
ve
r
a
l
s
tu
di
e
s
us
e
a
r
ti
f
ic
ia
l
i
nt
e
ll
ig
e
nc
e
te
c
hni
que
s
[
17
-
19
]
to
im
pl
e
m
e
nt
V
M
m
ig
r
a
ti
on
a
p
pr
oa
c
he
s
.
I
nde
e
d,
th
e
a
ut
hor
s
of
[
17
]
pr
e
s
e
nt
a
m
ul
ti
-
obj
e
c
ti
ve
vi
r
tu
a
l
m
a
c
hi
ne
(
V
M
)
pl
a
c
e
m
e
nt
s
c
he
m
e
f
or
E
C
D
C
e
dge
c
lo
ud
da
ta
c
e
nt
e
r
s
w
hi
c
h
a
im
s
to
m
in
im
iz
e
ne
twor
k
tr
a
f
f
ic
of
in
te
r
a
c
ti
ng
V
M
s
.
T
he
pr
opos
e
d
a
ppr
oa
c
h
is
ba
s
e
d
on
th
e
a
r
ti
f
ic
ia
l
be
e
c
ol
ony
opt
im
iz
a
ti
on
a
lg
or
it
hm
[
20
-
21
]
to
a
f
f
e
c
t
V
M
s
on
phys
ic
a
l
m
a
c
hi
ne
s
.
T
he
a
ut
hor
s
of
[
18
]
pr
opos
e
a
m
a
c
hi
ne
le
a
r
ni
ng
-
ba
s
e
d
V
M
m
ig
r
a
ti
on
a
ppr
oa
c
h
[
22
]
.
T
he
m
a
in
c
ont
r
i
but
io
n
of
th
is
a
r
ti
c
le
is
to
us
e
a
n
a
da
pt
iv
e
li
ve
m
ig
r
a
ti
on
a
ppr
oa
c
h
ba
s
e
d
on
pr
e
di
c
ti
ve
m
e
c
ha
ni
s
m
s
th
a
t
r
e
du
c
e
dow
nt
im
e
dur
in
g
li
ve
m
ig
r
a
ti
on
ove
r
w
id
e
a
r
e
a
ne
twor
ks
f
or
s
ta
nda
r
d
w
or
kl
oa
ds
.
T
he
a
u
th
or
s
of
[
19
]
a
s
s
um
e
th
a
t
th
e
V
M
pl
a
c
e
m
e
nt
pr
obl
e
m
c
a
n
be
f
or
m
a
li
z
e
d
a
s
a
bi
n
pa
c
ki
ng
p
r
obl
e
m
,
w
hi
c
h
tu
r
ns
out
to
b
e
N
P
-
ha
r
d.
T
o
s
ol
ve
it
,
th
e
y
us
e
d
a
ge
ne
ti
c
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
V
ir
tu
al
m
ac
hi
ne
m
ig
r
at
io
n i
n M
E
C
bas
e
d ar
ti
fi
c
ia
l
in
te
ll
ig
e
nc
e
te
c
hni
que
(
A
li
O
U
A
C
H
A
)
247
a
lg
or
it
hm
[
23
]
to
de
ve
lo
p
a
n
a
lg
or
it
hm
ba
s
e
d
on
c
lu
s
te
r
s
a
nd
w
hi
c
h
pr
oduc
e
s
a
n
a
ppr
oxi
m
a
ti
on
r
e
s
ul
t
of
th
e
bi
n
pa
c
k
pr
obl
e
m
.
I
nde
e
d,
th
e
pr
opos
e
d
a
lg
or
it
hm
gr
oups
th
e
p
opul
a
ti
on
of
th
e
c
ur
r
e
nt
ge
ne
r
a
ti
on
a
nd
s
e
l
e
c
ts
in
di
vi
dua
ls
f
r
om
di
f
f
e
r
e
nt
g
r
oups
w
it
h r
e
duc
e
d c
r
os
s
br
e
e
di
ng o
pe
r
a
ti
ons
.
U
nl
ik
e
th
e
s
e
r
e
s
e
a
r
c
h
e
s
,
our
r
e
s
e
a
r
c
h
s
e
e
ks
to
ta
ke
a
dva
nt
a
ge
,
on
th
e
one
ha
nd,
of
th
e
c
ont
r
ol
in
f
or
m
a
ti
on
e
xc
ha
nge
d
be
twe
e
n
th
e
node
s
of
th
e
ne
twor
k;
a
nd
on
th
e
ot
he
r
ha
nd,
to
us
e
a
n
a
r
ti
f
ic
ia
l
in
te
ll
ig
e
nc
e
t
e
c
hni
que
t
ha
t
c
a
n be
s
um
m
e
d up in t
he
t
he
or
y of
t
he
a
nt
c
ol
ony in or
de
r
t
o f
in
d
th
e
m
os
t
s
ui
t
a
bl
e
pa
th
s
by c
ons
id
e
r
in
g a
s
a
phe
r
om
one
, a
m
e
tr
ic
c
a
l
c
ul
a
te
d f
r
om
t
he
c
ont
r
ol
m
e
s
s
a
ge
s
.
4.
P
R
O
B
L
E
M
F
O
R
M
U
L
A
T
I
O
N
A
N
D
R
E
S
O
L
U
T
I
O
N
4
.1
.
P
r
ob
le
m
c
on
t
e
xt
B
e
f
or
e
we
be
gi
n
to
of
f
lo
a
d
a
ny
in
f
or
m
a
ti
on
or
s
e
t
of
ta
s
ks
to
be
tr
e
a
te
d
,
w
e
ne
e
d
to
be
c
e
r
ta
in
th
a
t
th
e
node
f
or
w
h
ic
h
th
os
e
ta
s
ks
a
r
e
of
f
lo
a
de
d
c
ont
e
nt
s
th
e
e
xe
c
ut
io
n
e
nvi
r
onm
e
nt
w
hi
c
h
is
r
e
pr
e
s
e
nt
e
d
by
th
e
a
ppr
opr
ia
te
to
th
e
e
xe
c
ut
io
n
of
th
o
s
e
ta
s
ks
.
T
o
s
e
nd
it
s
,
th
e
n
ode
ne
e
ds
a
t
f
ir
s
t
de
c
id
e
w
hi
c
h
node
in
th
e
pe
r
ip
he
r
y
of
th
e
ne
twor
k
is
pr
iv
il
e
ge
d
to
hos
t
th
e
or
th
e
.
I
nde
e
d,
w
e
w
il
l
ta
ke
a
dva
nt
a
ge
of
th
e
pr
oa
c
ti
ve
c
onc
e
pt
of
th
e
O
L
S
R
r
out
in
g
pr
ot
oc
ol
[
24
]
s
o
a
s
to
in
s
e
r
t
in
th
e
c
ont
r
ol
m
e
s
s
a
ge
s
,
pe
r
io
di
c
a
ll
y
e
xc
ha
nge
d
be
twe
e
n
th
e
node
s
,
in
f
or
m
a
ti
on
c
onc
e
r
ni
ng
th
e
pr
o
c
e
s
s
in
g
c
a
p
a
c
it
ie
s
of
th
e
s
e
r
ve
r
node
or
of
th
e
s
e
r
ve
r
node
s
of
th
e
ne
twor
k
a
nd
num
be
r
of
ta
s
ks
th
a
t
a
r
e
in
pr
ogr
e
s
s
or
w
a
it
in
g
to
be
pr
oc
e
s
s
e
d.
E
a
c
h
node
c
a
n
e
xt
r
a
c
t
f
r
om
th
e
s
e
m
e
s
s
a
ge
s
in
f
or
m
a
ti
on
a
bout
th
e
s
ta
te
of
th
e
li
nks
be
twe
e
n
node
s
.
O
nc
e
th
e
ta
r
ge
t
th
a
t
w
il
l
hos
t
th
e
is
known,
th
e
vi
r
tu
a
l
m
a
c
hi
ne
m
ig
r
a
ti
on
pr
oc
e
s
s
w
il
l
be
tr
ig
ge
r
e
d.
T
he
te
c
hni
que
us
e
d
in
th
is
s
tu
dy
is
in
s
p
ir
e
d
f
r
om
th
e
a
nt
li
f
e
s
ty
le
.
I
nde
e
d,
e
ve
r
yone
knows
th
a
t
a
nt
s
w
or
k
c
ol
le
c
ti
ve
ly
a
nd
ti
r
e
le
s
s
ly
to
s
to
r
e
e
nough
f
ood
to
ge
t
th
r
ough
th
e
w
in
te
r
.
A
nd
in
th
e
ga
th
e
r
in
g
of
th
is
f
ood,
th
e
a
nt
s
of
a
c
ol
ony
a
lwa
ys
m
a
na
ge
to
f
in
d
th
e
s
hor
te
s
t
r
out
e
e
ve
n
unde
r
th
e
c
ons
tr
a
in
ts
of
c
ha
ngi
ng
e
nvi
r
onm
e
nt
[
25
]
.
T
he
ke
y
a
nd
f
unda
m
e
nt
a
l
c
onc
e
pt
t
o be
dr
a
w
n he
r
e
i
s
be
in
g a
bl
e
t
o m
ov
e
a
m
ount
a
in
of
w
he
a
t
w
hi
le
w
or
ki
ng
to
ge
th
e
r
a
nd
c
a
r
r
yi
ng only one gr
a
in
of
w
he
a
t
pe
r
a
nt
. T
hus
, by c
or
r
e
s
ponde
nc
e
t
o t
he
a
nt
, t
he
ge
ne
r
a
l
id
e
a
of
t
hi
s
a
r
ti
c
le
i
s
to
tr
a
ns
f
e
r
a
ve
r
y
la
r
ge
(
m
ount
a
in
of
w
he
a
t)
w
hi
le
s
ubdi
vi
di
n
g
it
in
to
pi
e
c
e
s
(
gr
a
in
of
w
he
a
t)
of
s
m
a
ll
e
r
s
iz
e
s
o a
s
not
t
o di
s
tu
r
b t
he
f
unc
ti
on of
t
he
e
nt
ir
e
ne
two
r
k.
4.2.
S
e
r
ve
r
e
d
ge
s
e
le
c
t
io
n
(
d
e
s
t
in
at
io
n
n
od
e
s
e
le
c
t
io
n
)
A
M
A
N
E
T
[
1
]
ne
twor
k
is
m
a
d
e
up
of
a
s
e
t
of
m
a
c
hi
ne
s
of
a
he
te
r
oge
ne
ous
na
tu
r
e
.
T
o
be
a
bl
e
to
br
oa
dc
a
s
t
it
s
pr
oc
e
s
s
in
g
c
a
p
a
c
it
ie
s
,
e
a
c
h
node
in
je
c
t
s
in
to
c
ont
r
ol
m
e
s
s
a
ge
s
of
th
e
O
L
S
R
pr
ot
oc
ol
[
7
]
a
ll
th
e
phys
ic
a
l
c
ha
r
a
c
te
r
is
ti
c
s
th
a
t
a
r
e
r
e
la
te
d
to
th
e
e
xe
c
ut
io
n
of
ta
s
ks
.
T
hu
s
,
a
s
il
lu
s
tr
a
te
d
by
F
ig
ur
e
3,
be
f
or
e
s
e
ndi
ng
it
s
f
ir
s
t
c
ont
r
ol
m
e
s
s
a
ge
s
(
H
e
ll
o
or
T
C
)
th
e
node
in
que
s
ti
on
in
s
e
r
ts
in
th
e
s
e
m
e
s
s
a
g
e
s
th
e
num
be
r
of
pr
oc
e
s
s
or
s
i
t
ha
s
a
t
it
s
di
s
pos
a
l,
t
he
numbe
r
of
c
or
e
s
of
e
a
c
h pr
oc
e
s
s
or
, t
he
f
r
e
que
nc
y of
e
xe
c
ut
io
n, t
he
s
iz
e
of
th
e
pr
oc
e
s
s
c
a
c
he
m
e
m
or
y,
th
e
c
a
pa
c
it
y
of
th
e
m
a
in
m
e
m
or
y,
th
e
a
c
c
e
s
s
s
pe
e
d
of
th
is
m
e
m
or
y,
a
nd
th
e
bu
s
f
r
e
que
nc
y.
T
hi
s
c
a
n
be
im
pl
e
m
e
nt
e
d
by
m
odi
f
yi
ng
th
e
s
tr
uc
tu
r
e
of
th
e
H
E
L
L
O
a
nd
T
C
m
a
s
s
a
ge
s
,
a
s
s
how
n
in
F
ig
ur
e
2
(
a
-
b
)
by
a
ddi
ng
a
num
be
r
of
32
byt
e
le
nt
h
li
ne
s
(
one
li
ne
or
two
or
th
r
e
e
de
p
e
ndi
ng
on
th
e
a
m
ount
of
in
f
or
m
a
ti
on
to
be
t
r
a
ns
m
it
te
d)
be
f
or
e
th
e
f
ir
s
t
“
N
e
ig
hbor
fi
e
ld
I
nt
e
r
fa
c
e
A
ddr
e
s
s
”
in
th
e
H
E
L
L
O
m
e
s
s
a
ge
or
be
f
or
e
t
he
f
i
r
s
t
“
A
dv
e
r
ti
s
e
d N
e
ig
hbor
M
ai
n A
ddr
e
s
s
”
f
ie
ld
i
n t
he
T
C
m
e
s
s
a
ge
. A
f
te
r
t
ha
t,
onc
e
t
he
m
e
s
s
a
g
e
is
r
e
c
e
iv
e
d, e
a
c
h node
i
n M
A
N
E
T
s
to
r
e
s
t
he
s
e
pe
r
f
or
m
a
nc
e
s
i
n a
l
oc
a
l
ta
bl
e
w
hos
e
ke
y i
s
t
he
i
de
nt
if
ie
r
of
t
he
s
e
ndi
ng
node
.
B
a
s
e
d
on
s
to
r
e
d
in
f
or
m
a
ti
on,
e
a
c
h
node
of
th
e
n
e
twor
k,
w
he
r
e
th
e
c
om
put
a
ti
on
pe
r
f
o
r
m
a
nc
e
s
a
r
e
l
im
it
e
d, de
c
id
e
s
a
bout
t
he
node
w
hi
c
h w
il
l
be
t
he
t
a
r
ge
t
of
i
ts
vi
r
tu
a
l
m
a
c
hi
ne
.
F
ig
ur
e
3
.
S
e
r
ve
r
e
dge
s
e
le
c
ti
on i
n M
E
C
e
nvi
r
onm
e
nt
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
10
, N
o.
1
,
M
a
r
c
h
20
2
1
:
244
–
252
248
4.3.
V
ir
t
u
al
m
ac
h
in
e
s
p
li
t
t
in
g
a
n
d
m
ig
r
at
io
n
F
ir
s
t,
th
e
vi
r
tu
a
l
m
a
c
hi
ne
is
di
vi
de
d
in
to
e
qua
l
s
iz
e
pi
e
c
e
s
ℎ
=
1
,
2
,
…
,
.
T
h
e
s
iz
e
of
e
a
c
h
pi
e
c
e
i
s
c
ho
s
e
n
s
o
a
s
to
c
or
r
e
s
pond
to
th
e
gr
e
a
te
s
t
p
os
s
ib
le
v
a
lu
e
of
th
e
pa
c
ke
ts
w
hi
c
h
c
a
n
b
e
tr
a
ns
por
te
d
by
th
e
ne
twor
k.
S
in
c
e
w
e
a
r
e
onl
y
in
te
r
e
s
te
d
in
th
e
pr
io
r
s
e
ndi
ng
of
,
r
out
in
g,
pa
r
ti
ti
oni
n
g
a
nd
r
e
a
s
s
e
m
bl
in
g
ti
m
e
s
of
a
r
e
ne
gl
e
c
te
d.
O
n
c
e
th
e
ta
r
ge
t
n
ode
is
w
e
ll
de
f
in
e
d
a
nd
th
e
is
w
e
ll
pa
r
ti
ti
one
d,
a
ll
th
a
t
r
e
m
a
in
s
is
to
f
in
d
th
e
pa
th
or
pa
th
s
to
f
o
ll
ow
to
s
e
nd
th
e
pi
e
c
e
s
f
r
om
th
e
to
th
e
de
s
ti
na
ti
on node
. T
o do thi
s
,
a
n a
da
pt
e
d v
e
r
s
io
n of
t
he
a
nt
c
ol
o
ny opti
m
iz
a
ti
on a
lg
or
it
hm
[
25
]
is
a
dopt
e
d.
4.4.
P
r
ovi
d
e
d
d
at
a
I
n
th
is
s
e
c
ti
on,
w
e
w
il
l
de
f
in
e
a
ll
th
e
c
onc
e
pt
s
th
a
t
w
il
l
be
us
e
f
ul
la
te
r
.
T
hus
,
w
e
de
f
in
e
by
th
e
f
in
it
e
s
e
t
r
e
pr
e
s
e
n
ti
ng
th
e
nod
e
s
of
th
e
n
e
twor
k.
W
e
a
ls
o
de
f
in
e
by
U
=
{
(
,
)
ℎ
,
j
∈
X
}
th
e
f
in
it
e
s
e
t
of
li
nks
c
onne
c
ti
ng
node
s
of
.
T
he
va
r
ia
bl
e
s
d
i
,
j
r
e
pr
e
s
e
nt
in
g
th
e
qua
l
it
y
(
c
os
t,
m
e
tr
ic
...)
of
e
a
c
h
li
nk
(
i
,
j
)
∈
U
a
r
e
e
s
ti
m
a
te
d
by
th
e
(
1
)
a
s
a
f
unc
ti
on
of
th
e
m
ob
il
it
y
of
th
e
no
de
s
us
in
g
th
e
a
s
a
m
e
tr
ic
.
N
ot
e
th
a
t
th
is
m
e
tr
ic
e
s
ti
m
a
te
s
th
e
ti
m
e
r
e
m
a
in
in
g
f
or
a
node
to
le
a
ve
th
e
ne
i
ghbor
hood
of
a
not
he
r
node
.
I
t
is
d
e
duc
e
d
f
r
om
th
e
c
oor
di
na
te
s
e
xc
ha
ng
e
d t
hr
ough the
H
E
L
L
O
m
e
s
s
a
ge
s
[
26
]
.
(
)
=
,
(
)
=
,
(
)
(
1)
W
e
a
im
th
a
t
pi
e
c
e
s
of
th
e
pa
s
s
on
c
e
a
nd
onl
y
onc
e
th
r
ough
th
e
node
s
of
a
s
ub
s
e
t
of
(
P
a
r
ti
c
ul
a
r
ly
th
e
node
s
f
r
om
th
e
pa
th
to
th
e
ta
r
ge
t
node
)
.
B
y
us
in
g
a
s
a
phe
r
om
one
,
w
e
th
e
n
w
a
nt
to
de
te
r
m
in
e
t
he
pa
th
s
t
ha
t
w
il
l
be
t
r
a
ve
le
d by the
di
f
f
e
r
e
nt
p
ie
c
e
s
of
t
he
a
nd t
ha
t
w
il
l
m
in
im
iz
e
pa
c
ke
ts
l
os
s
(
in
c
r
e
a
s
e
th
e
s
ta
bi
li
ty
of
li
nks
)
.
T
he
r
e
f
or
e
,
w
h
e
n
a
n
a
nt
m
ove
s
f
r
om
one
node
to
a
not
he
r
,
th
e
qua
nt
it
y
of
phe
r
om
one
de
pos
it
e
d
a
t
ti
m
e
is
de
not
e
d
by
τ
i
,
j
(
t
)
.
T
he
va
lu
e
of
τ
i
,
j
(
t
)
is
c
a
lc
ul
a
te
d
a
nd
upda
te
d
a
t
th
e
e
nd
of
e
a
c
h pa
th
a
c
c
or
di
ng t
o
(
2
)
. N
ot
e
t
ha
t
e
a
c
h a
nt
ha
s
t
r
a
ve
r
s
e
d t
he
node
s
t
ha
t
m
a
ke
t
hi
s
pa
th
.
,
(
+
)
=
.
,
(
)
+
Δ
,
(
)
(
2)
W
he
r
e
ρ
∈
[
0
,
1
[
is
a
c
oe
f
f
ic
ie
nt
w
hi
c
h
w
il
l
de
f
in
e
th
e
s
pe
e
d
of
e
v
a
por
a
ti
on
of
phe
r
om
one
s
on
th
e
li
nks
be
twe
e
n
ti
m
e
a
nd
ti
m
e
+
,
a
nd
∆
τ
i
,
j
(
t
)
r
e
pr
e
s
e
nt
s
th
e
qua
nt
it
y
of
phe
r
om
one
s
de
po
s
it
e
d
by
th
e
a
nt
s
on t
he
l
in
k
(
i
,
j
)
in
t
hi
s
s
a
m
e
t
im
e
i
nt
e
r
va
l.
I
t
is
de
f
in
e
d (
c
a
lc
ul
a
te
d)
by
(
3
).
∆
,
(
)
=
∑
Δ
,
(
)
=
1
(
3)
I
n
th
is
e
qua
ti
on,
T
k
(
t
)
=
(
u
k
1
,
.
.
.
,
u
k
q
)
is
th
e
pa
th
tr
a
ve
le
d
by
th
e
ℎ
a
nt
in
th
e
ti
m
e
in
te
r
va
l
[
t
,
t
+
n
k
]
,
a
nd
L
k
(
t
)
it
s
le
ngt
h.
T
k
(
t
)
(
th
e
r
e
f
or
e
L
k
(
t
)
)
is
obt
a
in
e
d
by
a
na
ly
z
in
g
th
e
m
e
m
or
y
of
th
e
a
nt
.
∆
τ
i
,
j
k
(
t
)
,
c
a
lc
ul
a
te
d
by
(
4
)
,
is
th
e
a
m
ount
of
phe
r
om
one
de
po
s
it
e
d
by
t
hi
s
a
nt
on
li
nk
(
,
)
in
th
is
s
a
m
e
ti
m
e
in
te
r
va
l.
Δ
,
(
)
=
{
(
)
(
∈
(
)
=
(
,
)
)
0
ℎ
(
4)
W
he
r
e
Q
is
a
c
ons
ta
nt
.
T
he
le
ngt
h
L
k
(
t
)
of
a
pa
th
is
th
e
s
um
of
t
he
le
ngt
hs
of
th
e
li
nks
th
a
t
c
om
pos
e
i
t,
t
ha
t
is
:
(
)
=
∑
,
+
1
−
1
=
1
(
5)
4.5.
A
n
t
s
ys
t
e
m
N
ow
it
is
im
por
ta
nt
to
c
la
r
if
y
th
e
be
ha
vi
or
of
th
e
e
nt
ir
e
c
ol
ony.
A
t
a
ny
ti
m
e
,
e
a
c
h
a
nt
(
pa
c
ke
t
c
a
r
r
yi
ng
a
pi
e
c
e
of
th
e
)
c
hoos
e
s
a
de
s
ti
na
ti
on
node
f
r
om
it
s
ne
ig
hbor
hood
a
c
c
or
di
ng
to
a
de
f
in
e
d
c
hoi
c
e
.
A
ll
a
nt
s
m
ove
a
t
ti
m
e
+
1
in
a
not
he
r
node
of
th
e
ir
c
hoi
c
e
.
W
e
c
a
ll
a
n
it
e
r
a
ti
on
of
th
e
A
nt
S
y
s
te
m
(
A
S
)
a
lg
or
it
hm
,
th
e
s
e
t
o
f
di
s
pl
a
c
e
m
e
nt
s
of
th
e
e
nt
i
r
e
c
ol
ony
be
t
w
e
e
n
ti
m
e
a
nd
ti
m
e
+
1
.
T
hus
,
a
f
te
r
ite
r
a
ti
ons
e
a
c
h,
a
nt
w
il
l
ha
ve
tr
a
ve
le
d
a
pa
th
a
nd
r
e
a
c
he
d
a
n
e
nd
node
(
th
e
node
w
hi
c
h
w
il
l
c
ont
a
in
th
e
or
a
not
he
r
node
r
e
pr
e
s
e
nt
in
g a
n e
nd pa
th
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
V
ir
tu
al
m
ac
hi
ne
m
ig
r
at
io
n i
n M
E
C
bas
e
d ar
ti
fi
c
ia
l
in
te
ll
ig
e
nc
e
te
c
hni
que
(
A
li
O
U
A
C
H
A
)
249
4.6.
T
r
an
s
it
io
n
s
c
h
oi
c
e
A
n
a
nt
,
pl
a
c
e
d
on
node
,
a
t
in
s
t
a
nt
w
il
l
c
hoos
e
it
s
n
e
xt
hop
(
ne
ig
hbor
in
g
node
)
a
c
c
or
di
ng
to
th
e
qua
nt
it
y of
phe
r
om
one
s
τ
i
,
j
(
t
)
de
pos
it
e
d on the
l
in
k c
onne
c
ti
ng t
h
e
s
e
t
w
o node
s
. T
hi
s
c
hoi
c
e
w
il
l
be
m
a
de
r
a
ndoml
y, w
it
h a
pr
oba
bi
li
ty
of
c
hoos
in
g t
he
node
gi
ve
n by:
,
(
)
=
,
(
)
∑
,
(
)
∈
(
)
(
6)
W
he
r
e
N
i
k
de
f
in
e
s
th
e
s
e
t
of
node
s
be
lo
ngi
ng
to
th
e
ne
ig
hbor
hood of
node
a
nd
w
hi
c
h
a
nt
,
pl
a
c
e
d
on node
, ha
s
not
ye
t
vi
s
it
e
d
a
t
ti
m
e
in
t
he
c
ur
r
e
nt
pa
th
.
4.7.
G
lo
b
al
a
lg
or
it
h
m
−
D
iv
id
e
th
e
in
to
of
s
m
a
ll
pi
e
c
e
s
of
e
qua
l
s
iz
e
;
th
us
,
=
∑
=
1
.
e
a
c
h
pi
e
c
e
w
il
l
be
w
r
a
ppe
d
in
a
pa
c
ka
ge
w
hi
c
h
w
il
l
be
t
he
e
qui
va
le
nt
of
a
a
nt
.
−
I
ni
ti
a
li
z
e
t
he
va
lu
e
of
t
he
I
ni
ti
a
l
phe
r
om
one
qua
nt
it
y
τ
i
,
j
(
t
)
by a
c
ons
ta
nt
va
lu
e
.
−
F
or
e
a
c
h
a
nt
,
w
e
c
a
lc
ul
a
te
th
e
pr
oba
bi
li
ty
,
(
)
of
th
e
c
hoi
c
e
of
t
he
ne
xt
ne
ig
hbor
in
g
node
a
c
c
or
di
ng
to
(
6
)
. W
e
w
il
l
c
hoos
e
a
s
t
he
ne
xt
hop th
e
node
w
hos
e
pr
oba
bi
li
ty
i
s
gr
e
a
te
r
.
−
W
e
upda
te
t
he
a
m
ount
of
phe
r
om
one
a
c
c
or
di
ng t
o
(
2
).
−
W
e
r
e
pe
a
t
s
te
p
s
3 a
nd 4 unti
l
th
e
a
nt
k a
r
r
iv
e
s
a
t
th
e
d
e
s
ti
na
ti
on
node
or
t
o a
n unr
e
a
c
he
be
l
node
.
5.
R
E
S
U
L
T
S
A
ND
D
I
S
C
U
S
S
I
O
N
I
n
th
is
s
e
c
ti
on,
th
e
s
im
ul
a
ti
ons
a
r
e
us
e
d
to
pr
ove
th
e
e
f
f
ic
ie
nc
y
of
our
a
ppr
oa
c
h
of
s
e
ndi
ng
V
M
to
e
dge
s
e
r
ve
r
ba
s
e
d
on
th
e
a
nt
c
ol
ony
a
lg
or
i
th
m
on
th
e
M
E
C
.
F
ir
s
t,
th
e
s
im
ul
a
ti
ons
e
nvi
r
onm
e
nt
a
nd
pa
r
a
m
e
te
r
s
,
ge
ne
r
a
te
d
pa
c
ke
t
s
a
s
w
e
ll
a
s
th
e
ir
r
out
in
g
a
r
e
e
xp
la
in
e
d.
N
e
xt
,
th
e
obt
a
in
e
d
r
e
s
ul
ts
pr
e
s
e
nt
th
e
s
e
ndi
ng
of
di
f
f
e
r
e
nt
pa
r
ts
of
a
V
M
a
r
e
a
na
ly
z
e
d,
in
c
lu
di
ng
pe
r
c
e
nt
a
ge
of
s
e
nt
,
r
e
a
c
he
d,
unr
e
a
c
he
d
a
nd
lo
s
t
pa
c
ke
ts
.
5
.1.
S
im
u
la
t
io
n
e
n
vi
r
on
m
e
n
t
T
o
te
s
t
th
e
e
f
f
ic
ie
nc
y
of
our
a
ppr
oa
c
h,
s
e
v
e
r
a
l
s
im
ul
a
ti
ons
a
r
e
done
by
N
S
3
[
27
]
. T
he
y
to
ok
pl
a
c
e
in
two
e
nvi
r
onm
e
nt
s
w
hi
c
h
di
f
f
e
r
in
te
r
m
s
of
m
obi
li
ty
of
th
e
nod
e
s
.
I
n
th
e
f
ir
s
t
one
(
s
ta
ti
c
c
a
s
e
)
,
th
e
node
s
a
r
e
di
s
pe
r
s
e
d a
c
c
or
di
ng t
o a
gr
id
of
f
iv
e
node
s
i
n he
ig
ht
a
nd f
iv
e
n
ode
s
i
n w
id
th
(
5
×
5
)
s
pa
c
e
d by a
di
s
ta
nc
e
of
500
m
e
te
r
s
.
C
ons
e
qu
e
nt
ly
,
e
a
c
h
node
ha
s
two,
th
r
e
e
or
f
our
f
ix
e
d
ne
ig
hbor
s
de
pe
ndi
ng
on
it
s
po
s
it
io
n
in
th
e
gr
id
(
c
or
ne
r
,
bor
de
r
or
m
id
dl
e
node
)
.
O
n
th
e
ot
he
r
ha
nd,
in
th
e
s
e
c
ond
e
nvi
r
onm
e
nt
(
c
a
s
e
w
it
h
m
obi
li
ty
)
,
th
e
s
im
ul
a
ti
ons
c
o
ns
is
t
of
a
ne
twor
k
of
25
m
obi
le
node
s
m
ovi
ng
a
c
c
or
di
ng
to
th
e
r
a
ndom
w
a
ypoi
nt
(
R
W
P
)
m
obi
li
ty
m
ode
l
[
2
]
in
a
n a
r
e
a
of
1500
×
1500
2
.
I
n
bot
h
c
a
s
e
s
,
a
ll
node
s
ha
v
e
a
s
in
gl
e
w
ir
e
le
s
s
c
om
m
uni
c
a
ti
o
n
in
te
r
f
a
c
e
th
a
t
a
ls
o
us
e
s
O
L
S
R
a
s
a
ro
ut
in
g
pr
ot
oc
ol
.
T
he
ot
he
r
pa
r
a
m
e
te
r
s
of
th
e
s
im
ul
a
ti
on
e
n
vi
r
onm
e
nt
s
a
r
e
s
how
n
in
T
a
bl
e
1
. T
o
s
im
ul
a
te
th
e
s
e
ndi
ng
of
VM
k
pa
r
ts
of
a
to
th
e
de
s
ti
na
ti
on
nod
e
r
e
pr
e
s
e
nt
in
g
th
e
M
E
C
s
e
r
ve
r
,
dur
in
g
e
a
c
h
ti
m
e
in
te
r
va
l
(
r
ound)
la
s
ti
ng
of
one
s
e
c
ond,
a
r
a
ndom
node
is
c
hos
e
n
to
s
e
n
d
c
ons
e
c
ut
iv
e
ly
a
s
e
r
ie
s
of
pa
c
k
e
ts
w
ho
s
e
s
iz
e
is
1000
B
yt
e
.
T
he
r
out
in
g
of
VM
k
is
don
e
f
r
om
node
to
node
u
s
in
g
th
e
a
nt
th
e
or
y
a
lg
or
it
hm
de
f
in
e
d
in
s
e
c
ti
on
3. T
hus
,
e
a
c
h
pa
c
ke
t
w
il
l
be
c
ons
id
e
r
e
d a
s
a
n
a
nt
c
a
r
r
yi
ng
a
lo
a
d
to
be
r
out
e
d
to
a
s
p
e
c
if
ic
de
s
ti
na
ti
on
(
E
dge
S
e
r
ve
r
)
.
T
o
in
c
r
e
a
s
e
th
e
pr
oba
bi
li
ty
th
a
t
a
ll
a
nt
s
w
il
l
p
ur
s
ue
di
f
f
e
r
e
nt
pa
th
s
a
t
th
e
be
gi
nni
ng
of
th
e
s
e
nd,
th
e
num
be
r
of
a
ll
VM
k
pa
r
ts
is
di
vi
de
d
e
qua
ll
y
on
node
s
in
t
he
f
ir
s
t
ne
ig
hbor
hood
of
th
e
s
our
c
e
node
.
T
he
ne
xt
hop
is
d
e
f
in
e
d a
c
c
or
di
ng t
o t
he
a
lg
or
it
hm
1.
T
a
bl
e
1.
T
he
s
im
ul
a
ti
ons
e
nvi
r
onm
e
nt
s
P
a
r
a
m
e
t
e
r
V
a
l
ue
N
e
t
w
or
k S
i
m
ul
a
t
or
V
e
r
t
i
on
ns
-
3.29
P
r
ot
oc
ol
O
L
S
R
/
A
nt
C
ol
onny
S
i
m
ul
a
t
i
on T
i
m
e
100 s
S
i
m
ul
a
t
i
on A
r
e
a
1500
×1500 m
2
N
um
be
r
of
N
ode
s
25
T
r
a
ns
m
i
s
s
i
on R
a
nge
500 m
M
obi
l
i
t
y M
ode
l
R
a
ndom
W
a
yP
oi
nt
M
a
x S
pe
e
d
20 m
/
s
P
a
us
e
T
i
m
e
0 s
P
a
r
t
of
t
he
V
M
(
)
S
i
z
e
1000 B
yt
e
W
i
F
i
M
a
c
P
r
ot
oc
ol
802.11b
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
10
, N
o.
1
,
M
a
r
c
h
20
2
1
:
244
–
252
250
5
.2.
R
e
s
u
lt
s
a
nd
d
is
c
u
s
s
io
n
F
ig
ur
e
4
r
e
pr
e
s
e
nt
s
th
e
p
e
r
c
e
nt
a
ge
of
th
e
num
be
r
of
s
uc
c
e
s
s
f
ul
ly
s
e
nt
a
nd
r
e
c
e
iv
e
d
pa
c
k
e
ts
in
bot
h
s
ta
ti
c
a
nd
m
obi
le
c
a
s
e
a
s
a
f
unc
ti
on
of
th
e
r
ound
num
be
r
.
F
ig
u
r
e
4
(
a
)
s
how
s
a
n
im
pr
ove
m
e
nt
in
th
e
num
be
r
of
pa
c
ke
ts
r
e
c
e
iv
e
d
in
s
ta
ti
c
c
a
s
e
ov
e
r
ti
m
e
.
I
n
ot
he
r
w
or
ds
,
e
a
c
h
r
ound
h
a
s
a
m
or
e
im
pr
ove
d
num
b
e
r
of
r
e
c
e
iv
e
d
pa
c
ke
ts
th
a
n
th
e
pr
e
vi
ous
r
ound.
T
hi
s
s
how
s
th
e
e
f
f
e
c
ti
ve
ne
s
s
of
our
a
ppr
oa
c
h
in
te
r
m
s
of
le
a
r
ni
n
g
a
bout
f
in
di
ng
pa
th
s
to
th
e
ta
r
ge
t
node
.
I
f
th
e
im
pr
ove
m
e
nt
in
th
e
s
u
c
c
e
s
s
r
a
te
f
or
f
in
di
ng
th
e
pa
th
is
gua
r
a
nt
e
e
d
in
th
e
s
ta
ti
c
c
a
s
e
,
th
is
is
onl
y
pe
r
io
di
c
in
th
e
m
obi
le
c
a
s
e
,
a
s
s
how
n
in
F
ig
ur
e
4
(
b)
.
I
nde
e
d,
be
c
a
us
e
of
th
e
m
obi
li
ty
of
node
s
,
th
e
to
pol
ogy
of
th
e
ne
twor
k
c
ha
nge
s
ove
r
ti
m
e
.
T
hu
s
,
th
e
pa
th
s
c
a
n
c
ha
nge
f
r
om
r
ound
to
r
ound.
T
he
r
e
f
or
e
,
th
e
a
ppr
oa
c
h
us
e
d
to
f
in
d
th
e
pa
th
s
to
th
e
de
s
ti
na
ti
on
n
ode
m
us
t
b
e
a
da
pt
e
d
to
th
e
ne
w
to
pol
ogy
a
f
te
r
c
ha
nge
.
T
hi
s
ju
s
ti
f
ie
s
th
e
pe
r
io
di
c
im
pr
ove
m
e
nt
in
th
e
num
be
r
of
s
uc
c
e
s
s
f
ul
ly
r
e
c
e
iv
e
d pa
c
ke
t
s
.
(
a
)
(
b)
F
ig
ur
e
4
.
P
e
r
c
e
nt
a
ge
of
r
e
a
c
he
d pa
c
ke
t
s
s
e
nt
ba
s
e
d on the
r
oun
d numbe
r
i
n both s
ta
ti
c
a
nd mobi
le
e
nvi
r
one
m
e
nt
s
.
(
a
)
R
e
a
c
he
d p
a
c
ke
ts
i
n
s
ta
ti
c
c
a
s
e
,
(
b
)
R
e
a
c
he
d
pa
c
ke
ts
i
n m
obi
le
c
a
s
e
I
n
th
e
c
a
s
e
of
our
s
tu
dy,
vi
r
tu
a
l
m
a
c
hi
ne
s
a
r
e
known
by
la
r
ge
s
iz
e
s
.
S
o
to
b
e
a
bl
e
to
de
c
id
e
on
th
e
e
f
f
e
c
ti
ve
ne
s
s
of
our
a
ppr
oa
c
h, w
e
s
ta
r
te
d by s
pl
it
ti
ng e
a
c
h V
M
in
to
a
s
e
t
of
pa
c
ka
ge
s
of
s
im
il
a
r
s
iz
e
. T
he
n w
e
ha
ve
to
s
tu
dy
th
e
be
h
a
vi
or
of
ne
twor
k
to
w
a
r
ds
th
is
la
r
ge
qua
nt
i
ty
of
da
ta
.
T
hus
,
in
th
e
F
ig
ur
e
5,
w
e
r
e
pr
e
s
e
nt
th
e
a
ve
r
a
ge
s
of
th
e
p
e
r
c
e
nt
a
ge
of
pa
c
ke
ts
s
e
nt
a
nd
lo
s
t
c
om
pa
r
e
d
to
th
e
num
be
r
of
pa
c
ke
ts
g
e
ne
r
a
te
d
in
s
ta
ti
c
a
nd
m
obi
le
e
nvi
r
onm
e
nt
s
.
I
n
f
a
c
t,
F
ig
u
r
e
5
(
a
)
r
e
pr
e
s
e
nt
s
th
e
pa
c
ke
ts
s
e
nt
a
nd
F
ig
ur
e
5(
b)
r
e
pr
e
s
e
nt
s
th
e
lo
s
t
pa
c
ke
ts
.
S
o,
f
or
m
or
e
de
ta
il
s
,
th
e
F
ig
ur
e
5
(
a
)
r
e
pr
e
s
e
nt
s
th
e
a
ve
r
a
ge
of
th
e
num
be
r
of
pa
c
k
e
ts
th
a
t
a
r
e
s
e
nt
w
he
th
e
r
in
th
e
s
ta
ti
c
or
in
th
e
m
obi
le
c
a
s
e
.
I
n
th
e
two
c
a
s
e
s
,
it
is
f
ound
th
a
t
th
e
pe
r
c
e
nt
a
ge
of
th
e
num
be
r
of
pa
c
ke
ts
th
a
t
a
r
e
s
uc
c
e
s
s
f
ul
ly
s
e
nt
d
e
c
r
e
a
s
e
s
a
s
th
e
num
be
r
o
f
ge
ne
r
a
te
d
pa
c
ke
ts
in
c
r
e
a
s
e
s
.
T
hi
s
r
e
duc
ti
on
a
ppl
ie
s
f
or
bot
h c
a
s
e
s
:
s
ta
ti
c
a
nd mobi
le
.
T
hus
, t
he
pe
r
c
e
nt
a
g
e
of
pa
c
ke
ts
s
e
nt
a
ppr
oa
c
he
s
a
lm
os
t
one
hundr
e
d
pe
r
c
e
nt
w
he
n
th
e
num
be
r
of
pa
c
k
e
ts
ge
n
e
r
a
te
d
is
s
m
a
ll
e
r
.
O
n
t
he
ot
he
r
ha
nd,
th
is
num
be
r
is
onl
y
35%
w
h
e
n
th
e
num
be
r
of
pa
c
k
e
ts
ge
ne
r
a
te
d
e
xc
e
e
ds
40
pa
c
ke
t
s
a
nd
te
nds
to
s
ta
bi
li
z
e
w
he
n
th
e
num
be
r
of
pa
c
ke
ts
s
e
nt
e
xc
e
e
ds
50. I
nde
e
d, s
e
ndi
ng of
l
a
r
ge
qua
nt
it
ie
s
of
pa
c
ka
ge
s
i
s
a
ls
o c
onf
r
ont
e
d w
it
h a
s
e
t
of
c
ons
tr
a
in
ts
r
e
la
te
d
to
th
e
node
s
of
th
e
ne
twor
k.
T
he
r
e
f
or
e
,
w
he
n
th
e
te
m
por
a
r
y
m
e
m
or
ie
s
(
que
ue
)
,
in
te
nde
d
to
c
ont
a
in
pa
c
ke
ts
w
a
it
in
g t
o be
s
e
nt
, a
r
e
f
ul
l,
t
he
ot
he
r
pa
c
ke
ts
t
ha
t
a
r
r
iv
e
a
r
e
a
ut
om
a
ti
c
a
ll
y r
e
je
c
te
d.
I
n
a
ddi
ti
on,
a
nd
s
in
c
e
th
e
ope
r
a
ti
ng
of
th
e
V
M
e
s
s
e
nt
ia
ll
y
d
e
pe
nds
on
th
e
to
ta
l
r
e
c
e
pt
io
n
of
a
ll
pi
e
c
e
s
,
w
e
s
tu
di
e
d
th
e
num
b
e
r
of
r
e
a
c
he
d
a
nd
unr
e
a
c
he
d
pa
c
k
e
ts
in
th
e
ne
twor
k
f
or
th
e
m
odi
f
ie
r
ve
r
s
io
n
of
th
e
O
L
S
R
pr
ot
oc
ol
in
two
ty
pe
s
of
e
nvi
r
on
m
e
nt
s
.
I
nde
e
d,
th
e
F
ig
ur
e
6
r
e
pr
e
s
e
nt
s
th
e
a
ve
r
a
ge
of
th
e
pe
r
c
e
nt
a
ge
of
r
e
a
c
he
d
pa
c
k
e
ts
a
nd
unr
e
a
c
he
d
pa
c
k
e
ts
c
om
pa
r
e
d
to
th
e
num
b
e
r
of
ge
ne
r
a
te
d
pa
c
ke
ts
in
s
ta
ti
c
a
nd
m
obi
le
e
nvi
r
onm
e
nt
s
.
I
n
one
ha
nd,
th
e
F
ig
u
r
e
6
(
a
)
c
onc
e
r
ns
th
e
unr
e
a
c
he
d
pa
c
ke
t
s
.
O
n
th
e
ot
he
r
ha
nd,
th
e
F
ig
u
r
e
6(
b)
c
onc
e
r
ns
th
e
unr
e
a
c
he
d
p
a
c
ke
ts
.
S
o,
r
e
ga
r
dl
e
s
s
t
he
pe
r
c
e
nt
a
ge
of
pa
c
k
e
ts
th
a
t
a
r
e
s
uc
c
e
s
s
f
ul
ly
r
e
c
e
iv
e
d
ve
r
s
us
th
e
to
ta
l
num
be
r
of
pa
c
ke
ts
g
e
ne
r
a
te
d,
he
r
e
to
o,
w
e
s
e
e
th
a
t
th
is
num
be
r
de
c
r
e
a
s
e
s
w
he
n
th
e
num
be
r
of
pa
c
ke
ts
ge
ne
r
a
te
d
in
c
r
e
a
s
e
s
.
T
hi
s
i
s
ju
s
ti
f
ie
d
by
th
e
pr
e
s
e
nc
e
of
c
ont
r
ol
tr
a
f
f
ic
.
S
o
be
c
a
us
e
th
e
ne
twor
k a
ls
o us
e
s
t
he
O
L
S
R
a
s
a
r
out
in
g pr
ot
oc
ol
, a
nd due
t
o i
t
s
pr
oa
c
ti
ve
na
tu
r
e
, c
ont
r
ol
m
e
s
s
a
ge
s
(
H
E
L
L
O
,
T
C
a
nd
ot
he
r
s
)
a
r
e
g
e
ne
r
a
te
d
p
e
r
io
di
c
a
ll
y.
P
a
c
ke
ts
c
ont
a
in
in
g
H
E
L
L
O
m
e
s
s
a
ge
s
a
r
e
ge
ne
r
a
te
d
by
e
a
c
h
node
in
t
he
ne
twor
k
e
ve
r
y t
w
o s
e
c
onds
a
nd f
ur
th
e
r
pa
c
ke
ts
c
ont
a
in
in
g T
C
m
e
s
s
a
ge
s
a
r
e
ge
ne
r
a
te
d
by a
ll
node
s
a
nd
r
e
tr
a
ns
m
it
te
d
by
a
s
ubs
e
t
of
node
s
e
ve
r
y
f
iv
e
s
e
c
onds
.
W
he
n
th
e
s
e
ndi
ng
of
th
e
pa
c
ke
t
s
th
a
t
r
e
pr
e
s
e
nt
th
e
pa
r
ts
of
th
e
V
M
c
oi
nc
id
e
s
w
it
h
th
e
s
e
ndi
ng
of
th
e
c
ont
r
ol
pa
c
ke
ts
,
it
c
a
u
s
e
s
c
ol
li
s
io
n
s
a
nd
a
ls
o
que
ue
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
V
ir
tu
al
m
ac
hi
ne
m
ig
r
at
io
n i
n M
E
C
bas
e
d ar
ti
fi
c
ia
l
in
te
ll
ig
e
nc
e
te
c
hni
que
(
A
li
O
U
A
C
H
A
)
251
c
onge
s
ti
on,
r
e
s
ul
ti
ng
in
th
e
l
os
s
of
a
s
e
t
of
pa
c
ke
ts
.
I
n
a
ddi
ti
on,
s
e
ndi
ng
la
r
ge
qua
nt
it
ie
s
of
pa
c
ke
t
s
a
ls
o
f
a
c
e
s
a
s
e
t
of
c
ons
tr
a
in
ts
r
e
la
te
d t
o l
in
ks
b
e
twe
e
n ne
twor
k node
s
a
s
w
e
ll
a
s
ne
twor
k pr
ot
oc
ol
s
.
(
a
)
(
b)
F
ig
ur
e
5
.
A
ve
r
a
ge
s
of
t
he
pe
r
c
e
nt
a
ge
of
s
e
nt
a
nd l
os
t
pa
c
k
e
ts
ve
r
s
us
t
he
numbe
r
of
ge
ne
r
a
te
d pa
c
ke
t
s
i
n both
s
ta
ti
c
a
nd mobi
le
e
nvi
r
one
m
e
nt
s
.
(
a
)
S
e
nt
pa
c
ke
t
s
, (
b)
L
os
t
pa
c
ke
ts
(
a
)
(
b)
F
ig
ur
e
6
.
A
ve
r
a
ge
s
of
t
he
pe
r
c
e
nt
a
ge
of
r
e
a
c
he
d
a
nd unr
e
a
c
he
d
pa
c
ke
ts
ve
r
s
u
s
t
he
numbe
r
of
ge
ne
r
a
te
d
pa
c
ke
ts
i
n bouth s
ta
ti
c
a
nd mobi
le
e
nvi
r
one
m
e
nt
s
.
(
a
)
R
e
a
c
he
d
p
a
c
ke
ts
, (
b)
U
nr
e
a
c
he
d
p
a
c
ke
ts
6.
C
O
N
C
L
U
S
I
O
N
T
he
obj
e
c
ti
ve
of
th
is
a
r
ti
c
le
is
to
s
how
th
a
t
th
e
c
a
s
e
of
a
node
w
hos
e
c
om
put
a
ti
ona
l
pe
r
f
or
m
a
nc
e
is
ve
r
y
li
m
it
e
d.
T
o
f
unc
ti
on
w
e
ll
,
th
e
node
m
us
t
d
e
le
ga
te
a
s
e
t
of
ta
s
ks
to
th
e
e
gde
s
e
r
ve
r
.
P
e
r
f
or
m
in
g
th
e
s
e
ta
s
ks
r
e
qui
r
e
s
th
e
pr
e
s
e
nc
e
of
th
e
V
M
r
e
pr
e
s
e
nt
in
g
th
e
s
im
ul
a
ti
on
e
nvi
r
onm
e
nt
of
th
e
nod
e
in
qu
e
s
ti
on. T
hus
,
a
ne
w
a
ppr
oa
c
h
ha
s
be
e
n
pr
opos
e
d
in
th
is
r
e
s
e
a
r
c
h.
I
t
is
ba
s
e
d
on
a
n
a
r
ti
f
ic
ia
l
in
te
ll
ig
e
nc
e
te
c
hni
que
to
s
e
nd
th
e
w
hol
e
V
M
to
th
e
e
dge
s
e
r
ve
r
.
I
nde
e
d,
it
is
de
c
om
pos
e
d
in
to
a
s
e
t
of
s
m
a
ll
pa
r
ts
th
a
t
a
r
e
e
nc
a
ps
ul
a
te
d
in
pa
c
ke
ts
e
a
c
h
of
w
hi
c
h
is
c
ons
id
e
r
e
d
a
n
a
nt
.
T
h
e
r
e
f
or
e
,
w
e
us
e
d
th
e
a
nt
th
e
or
y
a
lg
or
it
hm
to
r
out
e
th
e
s
e
di
f
f
e
r
e
nt
pi
e
c
e
s
to
a
de
s
ti
na
ti
on
nod
e
.
C
ons
id
e
r
in
g
th
e
R
T
T
Q
m
e
tr
ic
a
s
a
phe
r
om
one
,
th
e
m
or
e
s
ta
bl
e
pa
th
s
a
r
e
f
ounde
d.
T
he
r
e
s
ul
t
s
of
th
e
s
im
ul
a
ti
ons
pe
r
f
or
m
e
d
by
N
S
3
s
how
th
a
t
th
e
w
hol
e
pa
c
ke
ts
f
in
d
s
ta
bl
e
p
a
th
s
to
th
e
de
s
ti
na
ti
on
node
.
H
ow
e
ve
r
,
th
e
s
e
r
e
s
ul
t
s
s
how
th
a
t
th
e
pe
r
c
e
nt
a
ge
of
pa
c
ke
ts
th
a
t
a
r
e
r
e
c
e
iv
e
d
a
l
s
o
de
pe
nds
on
th
e
pe
r
f
or
m
a
nc
e
of
th
e
node
s
a
nd
th
e
ne
twor
k. T
hu
s
,
a
s
a
p
e
r
s
pe
c
ti
ve
, a
f
ut
ur
e
s
tu
dy
a
im
s
to
a
l
s
o
ta
ke
in
to
a
c
c
o
unt
th
e
w
a
it
in
g
que
ue
of
th
e
M
A
C
la
ye
r
a
nd
th
e
ba
ndw
id
th
of
th
e
li
nks
.
A
not
he
r
pe
r
s
pe
c
ti
ve
a
im
s
to
f
in
d
a
pe
r
f
e
c
t
s
ync
hr
oni
z
a
ti
on
be
twe
e
n
th
e
pa
r
ts
a
nd
th
e
c
ont
r
ol
pa
c
ke
ts
.
I
n
a
ddi
ti
on,
a
f
in
a
l
pe
r
s
pe
c
ti
ve
i
s
to
m
a
k
e
th
e
pa
r
ts
s
iz
e
va
r
ia
bl
e
.
I
t
w
il
l
b
e
s
e
t
dyna
m
ic
a
ll
y
de
p
e
ndi
ng
on
th
e
ne
twor
k
c
ondi
ti
on.
R
E
F
E
R
E
N
C
E
S
[1]
S.
Corson
and
J.
Macker,
“
Mobile
Ad
hoc
Networking
(MANET)
:
Routing
Protocol
Performance
Issues
and
Evaluation Considerations
,”
RFC2501, MANET Performance Issues
, 1999.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
10
, N
o.
1
,
M
a
r
c
h
20
2
1
:
244
–
252
252
[2]
R. R. Roy,
“
Handbook of Mobil
e Ad Hoc Networks for Mobi
lity Models
,
”
1 ed.: Springer US, 2011.
[3]
Q. Z. Jie Cao, Weis
ong Shi,
“
Edge Computing: A Pri
mer
,
”
1st ed. ed.: Springer, Cham, 2018.
[4]
S.
Nunna
and
K.
Ganesan,
“
Mobile
Edge
Computing,
”
in
Healt
h
4.0:
How
Virtualization
and
Big
Data
are
Revolutionizing
Healthcare
,
C.
Thuemmler
and
C.
Bai,
Eds.,
ed
Cha
m:
Springer
International
Publishing,
pp.
187
-
203
,
2017
.
[5]
R.
Torre,
T.
Doan,
and
H.
Salah,
“
Chapter
4
-
Mobile
edge
cloud,
”
in
C
omputing
in
Communi
cation
Networks
,
F.
H.
P. Fitze
k, F. Gr
anelli,
and P.
Seelin
g, Eds.
, ed: A
cade
mic Pre
ss, pp.
77
-
91
, 2020
.
[6]
N.
Abbas,
Y.
Zhang,
A.
Taherkordi,
and
T.
Skeie,
“
Mobile
Edge
Computing:
A
Survey,
”
IEEE
Internet
of
Things
Journal
,
vol. 5, pp. 450
-
465, 2018.
[7]
T. Clausen and P. Jacquet,
“
Optimized Link
State Routi
ng Protocol (OLS
R)
,”
vol. RFC3626: RFC Editor, 2003.
[8]
C.
Dearlove
and
T.
Clausen,
“
Routing
multipoint
relay
optimization
for
the
optimi
zed
link
state
routing
protocol
version 2 (olsrv2),
”
RFC 7187 (Proposed Standard), T
he Internet Engineering Task Force (IETF),
2014.
[9]
J. Moy,
“
OSPF Version
2,
”
IETF Tool
s,
p. April 1998, 1998.
[10]
N. H. Shahapure and
P. Jayarekha,
“
Distance and Traffic
Based Vi
rtu
al Mach
ine Migra
tion for
Scalab
ility in
Cloud
Computing,
”
Procedia Computer Science,
vol. 132, pp. 728
-
737, 2018
, doi:
10.1016/j.procs.2018.05.083
.
[11]
Z.
Wang,
D.
Sun,
G.
Xue,
S.
Qian,
G.
Li,
and
M.
Li,
“
Ada
-
Things:
An
adaptive
virtual
machine
monitoring
an
d
migration stra
tegy for in
ternet
of things
application
s,
”
Journal
of
Para
llel an
d Dist
ributed
Comput
ing,
vol. 132,
pp.
164
-
176, 2019
, doi:
10.1016/j.
jpdc.2018.06.
009
.
[12]
K.
Z.
Ibrahim,
S.
Hofmeyr,
C.
Iancu,
and
E.
Roman,
“
Optimized
pr
e
-
copy
live
migration
for
memory
intensive
applicati
ons,
”
presented
at
the
Proceedings
of
2011
International
C
onference
for
High
Performance
Computing
,
Networking, St
ora
ge and Analysis, Seattle, Washington, 2011
,
DOI: 10.1145/206
3384.2063437
.
[13]
X.
Ruan,
H.
Chen,
Y.
Tian,
and
S.
Yin,
“
Virtual
machine
allocation
a
nd
migration
based
on
performance
-
to
-
power
ratio
in
energy
-
efficient
clouds,
”
Future
Generation
Computer
Syst
ems,
vol.
100,
pp.
380
-
394,
2019
,
doi
:
10.1016/j.future.2019.05.036
.
[14]
B.
Yang,
J.
Li,
and
Y.
Li,
“
Resear
ch
on
QoS
-
Oriented
Virtual
M
achine
Migration
Strategy
in
Mobile
Edg
e
Computing,
”
in
2019
12th
International
Conference
on
Intelligent
Computation
Tec
hnology
and
Automation
(ICICTA)
, pp. 227
-
231
, 2019,
doi: 10.1109/ICICTA49267.2019.00055
.
[15]
J.
Kikuchi,
C.
Wu,
Y.
Ji,
and
T.
Murase,
“
Mobile
edge
computing
ba
sed
VM
migration
for
QoS
improvement,
”
in
2017
IEEE
6th
Global
Conference
on
Consumer
Electronic
s
(GCCE)
,
pp.
1
-
5
,
2017,
doi
:
10.1109/GCCE.2017.8229344
.
[16]
F.
Teka,
C.
-
H.
Lung,
and
S.
A.
Ajila,
“
Nearby
live
virtual
machine
m
igration
using
cloudlets
and
multipath
TCP,
”
Journal
of Cl
oud Com
puting
,
vol. 5,
no.
12, 2016
,
DOI 10.1186/s136
77
-
016
-
0061
-
0
.
[17]
S.
S.
Nabavi,
S.
S.
Gill,
M.
Xu,
M.
Masdari,
and
P.
Garraghan,
“
T
RACTOR:
Traffic
-
aware
and
power
-
efficient
virtual
machine
placement
in
edge
-
cloud
data
centers
using
artific
ial
bee
colony
optimization,"
Internati
onal
Journal
of Co
mmunicat
ion Syst
ems,
vol.
34,
n
o. 5,
2021,
DOI: 10.1002/dac.474
7
.
[18]
M.
Arif,
A.
K.
Kiani,
and
J.
Qadir,
“
Machine
learning
based
optimiz
ed
live
virtual
machine
migration
over
WAN
links,
”
Teleco
mmunic
ation Sy
stems,
vol. 64,
no. 2,
pp. 245
-
257, 2017
,
DOI:
10.1007/s11235
-
016
-
0173
-
3
.
[19]
B.
Zhang,
X.
Wang,
and
H.
Wang,
“
Virtual
machine
placement
stra
tegy
using
cluster
-
based
genetic
algorithm,
”
Neurocom
puting,
vol. 428, pp. 310
-
316, 2021
, doi:
10.1016/j.neucom.2020.06.120
.
[20]
G.
Lindfield
and
J.
Penny,
“
Cha
pter
7
-
Artificial
Bee
and
Ant
Colon
y
Optimization,
”
in
Introducti
on
to
Nature
-
Inspired Op
timizat
ion
, G. Lindfield and J. Penny, Eds., ed Boston: Academic Press, pp. 119
-
140
, 2017
.
[21]
D. Karaboga,
“
An idea based on honey b
ee swarm for numerical opti
mizati
on,
”
Citeseer
2005.
[22]
J. R. Anderson,
“
Machine lear
ning: An artificial inte
lligence appro
ach
,”
vol. 3: Morgan Kaufmann, 1990.
[23]
X.
-
S.
Yang,
“
Chapter
5
-
Genetic
Algorithms,
”
in
Nature
-
Inspired
O
ptimizat
ion
Algorit
hms
,
X.
-
S.
Yang,
Ed.,
ed
Oxford: Elsevie
r, pp. 77
-
87
, 2014
.
[24]
T. Clausen and P. Jacquet,
“
Optimized Link
State Routi
ng Protocol (OLS
R)
,”
RFC3626:
RFC Editor, 2
003.
[25]
M. Dorigo and T. Stützle
,
“
Ant Colony Op
timization
,”
Bradfor
d Company,
2004.
[26]
A.
Ouacha,
N.
Lakki,
J.
E.
Abbadi,
A.
Habbani,
B.
Bouamoud,
and
M
.
Elkoutbi,
“
Reliable
MPR
selectio
n
based
on
link
lifetime
-
prediction
method,
”
in
2013
10th
IEEE
Internati
onal
Conference
On
Networki
ng,
Sensing
a
nd
Control
(ICNSC)
, 2013, pp. 11
-
16.
[27]
nsn
am.
ns
-
3 Manual
.
Evaluation Warning : The document was created with Spire.PDF for Python.