T
E
L
KO
MNIK
A
, V
ol
.
1
7
,
No.
6
,
Dec
e
mb
er
201
9
, p
p.
2
80
3
~
2811
IS
S
N: 1
69
3
-
6
93
0
,
accr
ed
ited
F
irst
Gr
ad
e b
y K
em
en
r
istekdikti,
Decr
ee
No: 2
1/E/
K
P
T
/20
18
DOI:
10.12928/TE
LK
OM
N
IK
A
.v
1
7
i
6
.
10262
◼
28
03
Rec
ei
v
ed
J
un
e
7
, 2
01
8
; Re
v
i
s
ed
J
ul
y
2
7
, 2
01
9
; A
c
c
ep
t
ed
A
ug
us
t
10
,
20
1
9
Q
uali
ty
of
Ser
vi
ce
in
b
and
w
i
dt
h
adap
ted
hyb
ri
d
UMT
S/WL
A
N
i
n
t
erwo
rki
ng
net
w
ork
R.
S
h
an
ka
r
*
1
,
P.
D
anan
ja
y
an
2
1
Pro
fe
s
s
o
r,
J
EPPIAAR
SRR
E
n
g
i
n
e
e
ri
n
g
Col
l
e
g
e
,
Che
n
n
a
i
,
I
n
d
i
a
2
Pro
fe
s
s
o
r,
Po
n
d
i
c
h
e
rry
E
n
g
i
n
e
e
r
ing
Col
l
e
g
e
,
Po
n
d
i
c
h
e
rry
,
I
n
d
i
a
*C
o
rre
s
p
o
n
d
i
n
g
a
u
th
o
r,
e
-
m
a
i
l
:
s
h
a
n
k
a
r
_
p
e
c
@red
i
ff
m
a
i
l
.
c
o
m
1
,
d
r.s
h
a
n
k
a
rr@h
o
tm
a
i
l
.
c
o
m
1
Ab
strac
t
In
te
g
r
a
ti
o
n
of
U
n
i
v
e
rs
a
l
M
o
b
i
l
e
T
e
l
e
c
o
m
m
u
n
i
c
a
ti
o
n
s
S
y
s
t
e
m
(UM
TS)
and
Wi
r
e
l
e
s
s
L
o
c
a
l
Ar
e
a
Net
work
(W
L
AN)
re
s
u
l
t
in
u
b
i
q
u
i
to
u
s
c
o
n
n
e
c
t
i
o
n
fo
r
e
n
d
u
s
e
r
s
.
In
th
e
i
n
te
g
r
a
te
d
n
e
two
rk
,
e
n
s
u
ri
n
g
th
e
q
u
a
l
i
ty
of
s
e
rv
i
c
e
to
u
s
e
rs
a
n
d
e
n
h
a
n
c
i
n
g
c
a
p
a
c
i
t
y
of
n
e
two
rk
a
re
p
ro
m
i
n
e
n
t
i
s
s
u
e
s
.
Ba
n
d
wid
th
a
d
a
p
t
a
ti
o
n
t
e
c
h
n
i
q
u
e
is
o
n
e
of
th
e
s
o
l
u
t
i
o
n
s
to
o
v
e
r
c
o
m
e
th
e
s
e
i
s
s
u
e
s
.
B
a
n
d
wi
d
th
a
d
a
p
ta
t
i
o
n
b
a
s
e
d
on
per
fl
o
w
and
per
c
l
a
s
s
s
c
h
e
m
e
s
were
p
ro
p
o
s
e
d
fo
r
l
o
o
s
e
l
y
c
o
u
p
l
e
d
i
n
te
rwor
k
i
n
g
n
e
two
r
k
.
In
th
i
s
p
a
p
e
r
,
h
y
b
r
i
d
c
o
u
p
l
e
d
UM
TS
and
W
L
AN
i
n
t
e
rwork
i
n
g
n
e
tw
o
rk
is
a
n
a
l
y
z
e
d
wit
h
b
a
n
d
w
i
d
th
a
d
a
p
ta
ti
o
n
b
a
s
e
d
on
per
fl
o
w
and
per
c
l
a
s
s
s
c
h
e
m
e
s
and
th
e
p
e
rf
o
rm
a
n
c
e
s
h
a
v
e
b
e
e
n
c
o
m
p
a
r
e
d
.
S
i
m
u
l
a
ti
o
n
re
s
u
l
t
s
h
o
w
s
th
a
t
th
e
p
ro
p
o
s
e
d
h
y
b
ri
d
c
o
u
p
l
e
d
i
n
te
rwork
i
n
g
n
e
two
r
k
wit
h
b
a
n
d
wid
th
a
d
a
p
t
a
ti
o
n
b
a
s
e
d
on
p
e
r
c
l
a
s
s
s
c
h
e
m
e
p
e
rfo
rm
s
b
e
tt
e
r
wit
h
e
n
h
a
n
c
e
d
q
u
a
l
i
ty
of
s
e
rv
i
c
e
a
n
d
n
e
tw
o
rk
c
a
p
a
c
i
ty
.
Key
w
ords
:
c
a
l
l
b
l
o
c
k
i
n
g
p
ro
b
a
b
i
l
i
t
y
,
d
y
n
a
m
i
c
UM
TS
/
WL
AN
,
o
p
n
e
t,
Q
o
S
a
d
a
p
ti
v
e
n
e
two
r
k
,
th
ro
u
g
h
p
u
t
Copy
righ
t
©
2
0
1
9
Uni
v
e
rsi
t
a
s
Ahm
a
d
D
a
hl
a
n.
All
righ
t
s
r
e
s
e
rve
d
.
1.
Int
r
o
d
u
ctio
n
T
he
i
nc
r
ea
s
e
in
de
m
an
d
fo
r
data
c
o
mm
un
i
c
ati
on
t
hrou
gh
m
ob
i
l
e
ph
on
e
,
p
av
ed
the
way
for
the
mi
grat
i
on
of,
G
l
ob
al
S
y
s
tem
f
or
Mo
bi
l
e
(
G
S
M),
a
c
i
r
c
u
i
t
s
wi
tc
h
ed
s
y
s
t
em
to
G
en
eral
P
ac
k
et
R
ad
i
o
S
erv
i
c
e
(
G
P
RS
)
and
3
rd
G
en
era
ti
o
n
(
3G
)
ne
tw
ork
s
,
whi
c
h
ha
v
e
the
pa
c
k
et
s
wi
tc
he
d
fun
c
t
i
on
al
i
ty
.
Rec
en
t
tr
e
nd
is
that
ne
t
wor
k
s
l
i
k
e
U
MTS
and
W
LA
N
are
i
nt
eg
r
ate
d
to
prov
i
de
a
un
i
v
ers
al
a
l
l
I
nte
r
ne
t
P
r
oto
c
o
l
(
I
P
)
p
l
at
form
w
hi
c
h
c
an
prov
i
de
mu
c
h
us
ef
ul
ap
p
l
i
c
at
i
on
s
to
t
he
e
nd
us
e
r
s
.
T
he
al
l
IP
pl
atf
or
m
c
an
prov
i
de
e
nd
us
ers
wi
t
h
be
ne
fi
ts
s
uc
h
as
l
ower
c
os
t
of
tr
an
s
mi
s
s
i
o
n
and
hi
g
he
r
b
an
d
wi
d
th
w
i
th
r
oa
mi
ng
f
ea
t
ures
and
pe
r
v
as
i
v
e
as
pe
c
ts
.
W
L
A
N
s
y
s
tem
s
prov
i
de
h
i
gh
da
t
a
r
ate
s
at
a
r
el
a
ti
v
e
l
y
l
ow
c
os
t
c
om
pa
r
ed
to
U
MTS,
bu
t
it
has
a
l
i
mi
t
ed
c
ov
erag
e
area
a
nd
l
es
s
s
up
p
ort
for
hi
gh
s
pe
ed
mo
b
i
l
i
ty
.
W
LA
N
ha
v
e
be
e
n
wi
d
el
y
de
p
l
oy
ed
in
p
l
ac
es
l
i
k
e
ai
r
po
r
ts
,
ho
tel
s
an
d
c
am
pu
s
e
s
.
T
he
3
rd
g
en
era
ti
o
n
pa
r
t
ne
r
s
hi
p
proj
ec
t
(
3G
P
P
)
has
b
ee
n
wor
k
i
ng
on
s
ta
nd
ard
i
z
at
i
on
for
i
nt
e
grati
n
g
c
e
l
l
ul
ar
an
d
W
L
A
N
s
y
s
tem
s
[1,
2]
,
Network
arc
hi
tec
t
ures
for
c
on
v
ergi
ng
UMT
S
/
W
L
A
N
s
y
s
tem
s
c
an
be
group
ed
i
nto
two
c
at
eg
orie
s
ba
s
ed
on
t
he
a
uto
no
my
be
t
ween
two
ne
tw
ork
s
s
uc
h
as
,
ti
g
ht
c
ou
pl
i
ng
an
d
l
oo
s
e
c
ou
p
l
i
n
g
[3
].
In
l
oo
s
e
c
ou
pl
i
ng
,
th
e
two
n
etwo
r
k
s
are
i
nte
gra
ted
b
ey
on
d
Cor
e
Net
wor
k
(
CN)
of
UMTS
and
are
c
on
ne
c
te
d
t
hroug
h
ga
tew
ay
s
of
i
nt
ernet.
Co
m
mu
n
i
c
ati
on
be
tw
ee
n
ne
twor
k
s
are
r
ea
l
i
s
e
d
throug
h
s
tan
d
ard
IP
and
mo
b
i
l
i
ty
of
mo
bi
l
e
s
tat
i
on
s
is
ma
na
ge
d
t
hroug
h
prot
oc
ol
s
s
uc
h
as
Mo
b
i
l
e
I
P
.
L
oo
s
e
c
ou
p
l
i
ng
arc
hi
tec
ture
en
ab
l
es
t
he
t
w
o
ne
twork
s
to
be
de
p
l
oy
e
d
s
ep
arat
el
y
as
s
ho
wn
in
F
i
g
ure
1,
b
ut
th
e
l
i
mi
tat
i
on
is
l
on
ge
r
de
l
ay
for
s
i
gn
a
l
i
n
g
an
d
v
erti
c
a
l
ha
nd
ov
ers
.
In
ti
g
ht
c
ou
pl
i
ng
arc
hi
t
ec
ture,
t
he
t
wo
ne
t
wor
k
s
are
i
nte
gr
ate
d
at
CN
of
U
MTS
as
s
ho
wn
in
F
i
g
ure
2,
whi
c
h
has
l
ow
er
d
el
ay
fo
r
s
i
gn
a
l
i
ng
an
d
v
ert
i
c
al
h
an
do
v
er
bu
t
h
as
m
ore
i
mp
l
em
e
nta
t
i
o
n
c
om
pl
ex
i
ty
.
To
ov
erc
o
me
the
s
e
d
i
s
ad
v
a
nta
g
es
hy
brid
c
ou
pl
ed
U
MTS/
W
L
A
N
i
nt
erw
ork
i
ng
arc
hi
tec
ture
w
as
pro
po
s
ed
[4,
5].
Co
mp
are
d
w
i
t
h
t
he
e
x
i
s
ti
ng
c
ou
pl
i
ng
s
c
he
m
es
,
hy
brid
c
ou
p
l
e
d
i
nte
r
w
ork
i
ng
n
etw
ork
c
an
di
s
tr
i
b
ute
t
he
s
i
g
na
l
i
n
g
an
d
data
l
oa
d
on
both
U
MT
S
CN
an
d
c
ore
i
nte
r
n
et.
It
off
ered
l
es
s
ha
n
do
ff
l
at
en
c
y
in
v
ert
i
c
al
ha
nd
off
proc
es
s
[6]
.
W
he
n
t
he
Mo
b
i
l
e
N
od
e
(
M
N)
is
pres
e
nt
in
UMT
S
n
et
wor
k
as
s
ho
w
n
in
F
i
gu
r
e
3,
it
is
di
r
ec
tl
y
c
on
ne
c
te
d
wi
t
h
th
e
B
as
e
S
tat
i
on
(
B
S
)
by
w
i
r
e
l
es
s
l
i
nk
“
a”.
If
t
he
MN
is
pr
es
en
t
in
W
LA
N
then
it
c
o
nn
ec
ts
w
i
th
r
em
o
te
ac
c
es
s
po
i
nt
(
A
P
)
by
wi
r
el
es
s
l
i
nk
“
b”
a
nd
in
turn
r
em
ote
AP
is
c
on
ne
c
te
d
to
W
LA
N
s
erv
er.
T
he
du
al
mo
de
Us
er
E
q
ui
pm
e
nt
(
U
E
)
h
ardwar
e
is
us
ed
for
c
on
v
erged
U
MTS
an
d
W
L
A
N
ne
t
wor
k
s
whi
c
h
ha
v
e
c
ap
ab
i
l
i
ty
to
ac
c
es
s
ei
the
r
ne
twork
[7,
8].
Many
d
i
ff
i
c
ul
t
i
es
em
erg
e
when
att
e
mp
t
i
n
g
to
prov
i
d
e
Q
ua
l
i
ty
of
S
erv
i
c
e
(
Q
o
S
)
s
ol
ut
i
on
s
f
or
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
MNIK
A
V
ol
.
1
7
,
No
.
6
,
D
ec
em
b
er
20
19
:
28
0
3
-
2811
2804
i
nte
grate
d
W
LA
N
and
c
e
l
l
ul
ar
ne
tw
ork
s
,
be
c
au
s
e
of
the
u
nb
a
l
a
nc
ed
c
ap
ac
i
ty
of
the
two
s
y
s
tem
s
,
i
s
s
ue
s
r
ai
s
ed
by
ha
nd
ov
er
and
t
he
u
nrel
i
ab
l
e
c
om
mu
ni
c
at
i
o
n
in
w
i
r
el
es
s
me
di
a
[
9].
To
en
ab
l
e
eff
i
c
i
en
t
us
e
of
th
e
s
c
arc
e
r
es
ou
r
c
es
prov
i
de
d
by
the
c
e
l
l
ul
ar
ne
tw
ork
s
whi
l
e
ma
i
nt
ai
ni
n
g
s
tr
on
g
Q
o
S
gu
arant
ee
s
,
a
ge
ne
r
i
c
r
es
erv
at
i
on
ba
s
ed
Q
o
S
m
od
e
l
h
as
be
e
n
propos
e
d
for
the
l
oo
s
e
l
y
c
ou
pl
ed
i
nt
eg
r
at
ed
c
el
l
u
l
ar
an
d
W
LA
N
ne
t
w
ork
[10].
B
an
dwi
dth
a
da
p
tat
i
on
ba
s
ed
on
per
f
l
ow
a
l
go
r
i
thm
a
dj
us
ts
the
Q
o
S
to
ma
i
nta
i
n
on
g
oi
n
g
s
es
s
i
o
ns
at
a
s
ati
s
f
a
c
tory
l
ev
e
l
w
i
th
eff
i
c
i
en
t
us
e
of
t
he
s
y
s
tem
r
es
ou
r
c
es
[11].
T
h
e
Q
o
S
ma
na
ge
me
nt
wi
t
h
ba
nd
w
i
d
th
ad
ap
ta
ti
o
n
ba
s
ed
on
per
c
l
as
s
al
go
r
i
t
h
m
prov
i
de
s
b
ett
er
fa
i
r
ne
s
s
than
ba
nd
w
i
dth
a
da
pt
ati
on
ba
s
ed
on
per
fl
ow
a
l
g
orit
h
m
[1
2].
By
a
da
pti
n
g
the
ba
n
dwi
dth
,
as
an
d
when
de
ma
n
d
aris
es
,
a
greate
r
n
um
b
er
of
us
ers
c
ou
l
d
be
s
erv
ed
wi
th
ac
c
ep
ta
bl
e
l
ev
el
Q
o
S
[1
3
-
15].
In
t
hi
s
p
ap
er,
hy
brid
c
ou
pl
ed
UMTS/
W
L
A
N
i
nt
erw
ork
i
ng
ne
twork
w
i
th
ba
n
dwi
dth
ad
ap
tat
i
o
n
ba
s
e
d
on
per
f
l
ow
and
per
c
l
as
s
al
g
orit
h
m
is
an
al
y
s
ed
,
th
ei
r
p
erfor
ma
nc
es
are
c
o
mp
are
d
w
i
th
t
he
ex
i
s
ti
ng
non
ad
ap
t
ati
v
e
s
c
he
me
v
i
a
s
i
mu
l
ati
on
.
T
he
r
es
t
of
th
i
s
paper
is
organ
i
s
ed
as
fo
l
l
ows
.
S
ec
ti
on
2
r
ev
i
ews
the
ad
ap
t
i
v
e
Q
o
S
fr
am
ewo
r
k
.
B
an
dw
i
th
ad
ap
t
ati
on
ba
s
ed
on
per
c
l
as
s
a
l
go
r
i
th
m
is
a
na
l
y
s
ed
in
s
ec
ti
on
3.
P
erfor
ma
nc
e
an
al
y
s
i
s
in
s
ec
t
i
on
4
d
es
c
r
i
b
es
the
s
i
mu
l
at
i
on
s
c
en
ari
o
an
d
di
s
c
us
s
es
the
s
i
mu
l
at
i
on
r
es
u
l
ts
.
F
i
na
l
l
y
,
s
ec
ti
o
n
5
s
u
mm
ar
i
s
es
th
i
s
s
tud
y
an
d
g
i
v
es
c
on
c
l
ud
i
ng
r
em
ark
s
.
F
i
gu
r
e
1.
L
oo
s
e
c
o
up
l
i
n
g
ar
c
hi
tec
ture
F
i
gu
r
e
2.
T
i
gh
t
c
o
up
l
i
ng
arc
hi
tec
t
ure
F
i
gu
r
e
3.
Hy
br
i
d
c
ou
pl
ed
U
MTS/
W
L
A
N
i
nte
r
w
ork
i
ng
ar
c
hi
tec
ture
2.
Ad
apt
iv
e
Q
o
S
F
r
amew
o
r
k
W
L
A
N
and
UMT
S
ne
t
wor
k
s
ha
v
e
di
ff
erent
tr
an
s
mi
s
s
i
on
c
a
pa
c
i
t
i
es
;
Cons
i
d
er
the
s
i
gn
i
fi
c
a
nt
di
f
ferenc
e
in
tr
an
s
mi
s
s
i
o
n
c
ap
ac
i
ty
b
etwe
e
n
t
he
two
s
y
s
tem
s
e
s
pe
c
i
al
l
y
w
he
n
us
er
ha
nd
o
ff
t
ak
es
p
l
ac
e
t
he
r
efo
r
e
,
h
an
d
off
be
twee
n
the
tw
o
s
y
s
tem
s
ma
k
es
it
v
ery
ha
r
d
to
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
MNIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
Q
ua
l
i
ty
of
S
erv
i
c
e
i
n
ba
n
dw
i
dth
ad
ap
te
d h
y
br
i
d U
MTS/
W
L
A
N
…
(
R.
S
h
an
k
ar
)
2805
ma
i
nta
i
n
th
e
en
ou
g
h
Q
oS
f
or
on
go
i
ng
c
o
nn
ec
t
i
on
s
.
A
l
s
o,
W
L
A
N
op
erates
on
IS
M
band
a
nd
h
as
a
l
ot
of
u
nc
on
tr
o
l
l
ab
l
e
i
n
terfer
en
c
e.
T
h
e
s
pre
ad
s
pe
c
tr
um
tec
h
ni
q
ue
is
us
ed
to
r
ed
uc
e
the
i
nte
r
fere
nc
e.
To
s
up
po
r
t
Q
oS
,
in
pa
c
k
et
s
wi
tc
hi
ng
ne
twork
s
the
ne
tw
ork
l
oa
d
mu
s
t
be
k
ep
t
under
a
t
hres
ho
l
d.
T
h
e
th
i
r
d
prob
l
em
is
that
t
he
ac
h
i
e
v
ab
l
e
Q
o
S
l
ev
el
s
in
W
LA
N
and
UMT
S
do
not
ma
tc
h
e
ac
h
oth
er.
UM
T
S
is
v
ery
we
l
l
d
es
i
gn
ed
wi
th
c
arefu
l
ne
tw
ork
p
l
an
ni
ng
and
m
atu
r
e
ad
m
i
s
s
i
on
c
on
tr
ol
a
l
go
r
i
th
ms
.
T
he
r
efo
r
e,
t
he
ac
h
i
ev
a
bl
e
Q
o
S
l
ev
e
l
is
r
el
ati
v
e
l
y
hi
g
h
in
U
MTS,
whi
l
e
80
2
.11
e
W
L
A
N
wor
k
s
wi
th
i
n
a
l
es
s
c
on
tr
o
l
l
e
d
en
v
i
r
on
me
n
t
a
nd
it
is
d
i
ffi
c
ul
t
to
ac
hi
ev
e
ha
r
d
Q
oS
.
T
he
Q
o
S
fr
am
e
wor
k
ba
s
ed
on
an
a
da
pt
i
v
e
ap
pro
ac
h
[9]
f
or
i
nt
eg
r
at
ed
W
LA
N
and
UMTS
ad
dres
s
es
the
ab
o
v
e
i
s
s
ue
s
a
nd
prov
i
d
es
pr
ac
ti
c
al
an
d
us
er
-
s
at
i
s
fi
e
d
Q
oS
.
It
a
da
p
ts
the
l
o
ng
du
r
at
i
o
n
c
al
l
s
in
the
s
y
s
tem
ba
s
e
d
on
the
i
r
s
tat
e
i
nf
ormat
i
o
n,
be
c
au
s
e
they
ha
v
e
a
gre
ate
r
proba
bi
l
i
ty
of
qu
i
t
ti
n
g
t
he
s
y
s
tem
th
us
l
e
av
i
n
g
fe
wer
ad
a
pte
d
c
on
ne
c
ti
on
s
in
the
s
y
s
tem
.
Us
e
of
a
da
p
tat
i
on
prof
i
l
e
c
an
gu
arant
ee
th
e
s
ati
s
fi
ed
Q
o
S
l
ev
e
l
to
t
he
end
us
er
an
d
ad
ap
ti
n
g
the
l
o
ng
du
r
a
ti
on
c
al
l
s
c
an
m
i
n
i
m
i
s
e
the
Q
o
S
r
ed
uc
t
i
on
of
the
e
nti
r
e
s
y
s
tem
.
F
i
g
ure
4
de
s
c
r
i
be
s
th
e
b
an
dw
i
dt
h
a
d
ap
tat
i
o
n
a
l
go
r
i
thm
ba
s
e
d
on
per
f
l
ow
s
c
h
em
e.
F
i
gu
r
e
4.
P
s
eu
d
o
c
od
e
of
p
er
fl
ow
ad
ap
t
ati
on
a
l
g
orit
h
m
,
wh
ere
Bi
r
e
pres
en
ts
the
r
eq
ui
r
e
d
b
an
d
wi
d
th,
Di
de
no
tes
th
e
m
i
n
i
mu
m
b
an
d
wi
dt
h
r
eq
ue
s
t
d
efi
ne
d
in
t
he
c
on
ne
c
ti
on
de
gra
da
t
i
on
prof
i
l
e,
R
B
W
is
r
e
qu
es
te
d
ba
nd
w
i
dth
,
S
A
B
is
s
y
s
tem
av
ai
l
ab
l
e
ba
nd
wi
dth
,
an
d
GB
is
gu
ard
b
an
dw
i
dt
h
3.
A
D
ynamic
Band
w
idt
h
Ma
n
agem
ent
S
ch
eme
In
w
i
r
el
es
s
c
om
mu
n
i
c
ati
on
ne
t
wor
k
s
,
ne
go
t
i
at
i
o
n
b
et
ween
d
i
ffe
r
e
nt
ne
t
wor
k
f
l
o
ws
is
an
eff
ec
t
i
v
e
way
to
i
mp
r
ov
e
ov
eral
l
s
y
s
tem
pe
r
f
ormanc
e.
T
he
ba
nd
w
i
dth
ad
a
pta
t
i
o
n
al
g
orit
h
m
as
the
k
ey
fac
tor
of
t
he
ad
ap
t
i
v
e
Q
oS
fr
am
ework
de
c
i
de
s
how
to
a
dj
us
t
the
Q
o
S
c
on
ne
c
ti
o
ns
,
s
i
nc
e
mo
b
i
l
e
us
ers
s
ho
ul
d
be
ab
l
e
to
s
ea
m
l
es
s
l
y
m
ai
n
tai
n
t
he
i
r
o
ng
o
i
ng
s
es
s
i
on
s
at
a
s
ati
s
fi
e
d
l
ev
el
.
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
MNIK
A
V
ol
.
1
7
,
No
.
6
,
D
ec
em
b
er
20
19
:
28
0
3
-
2811
2806
T
hi
s
c
an
be
do
n
e
in
two
w
ay
s
.
On
e
is
to
a
da
p
t
a
fl
ow
to
i
ts
s
am
e
l
ev
e
l
,
w
he
n
it
mi
gra
tes
fr
om
one
s
y
s
tem
i
nto
a
no
th
er.
I
de
a
l
l
y
,
ea
c
h
c
al
l
in
the
s
y
s
tem
s
ho
ul
d
be
al
l
oc
at
ed
the
ma
x
i
mu
m
al
l
owa
bl
e
ba
n
dw
i
dth
.
Ho
wev
er,
W
L
A
N
and
U
MT
S
ha
v
e
di
ff
erent
tr
an
s
m
i
s
s
i
on
c
ap
ac
i
ty
;
a
s
es
s
i
on
th
at
c
o
ns
um
es
a
mo
de
r
ate
am
ou
nt
of
b
an
dw
i
dth
in
W
L
A
N
s
y
s
tem
c
an
be
gre
ed
y
a
nd
the
r
efo
r
e
c
ou
l
d
be
r
ej
ec
te
d
in
the
UMT
S
.
A
c
on
ne
c
ti
o
n
s
wi
tc
h
ed
fr
o
m
c
e
l
l
ul
ar
ne
t
wor
k
to
W
L
A
N
needs
to
l
ev
e
l
up
i
ts
ba
nd
wi
dt
h
o
the
r
w
i
s
e
it
wi
l
l
l
os
e
the
me
a
ni
ng
of
the
i
nte
grat
i
on
.
T
he
ot
he
r
way
is
to
ad
ap
t
f
l
ows
to
us
e
the
s
y
s
tem
r
es
ou
r
c
e
mo
r
e
eff
i
c
i
e
ntl
y
in
a
s
i
n
gl
e
s
y
s
tem
or
both
to
ac
c
om
mo
da
te
mo
r
e
n
ew
ar
r
i
v
al
s
an
d
h
an
d
off
c
a
l
l
s
.
T
h
e
ba
nd
w
i
dt
h
a
da
pt
ati
on
ba
s
ed
on
p
er
fl
ow
me
th
od
ad
ap
ts
t
he
l
on
g
d
urati
o
n
c
a
l
l
s
in
th
e
s
y
s
tem
b
a
s
ed
on
t
he
i
r
s
tat
e
i
nfo
r
ma
t
i
on
w
i
th
a
g
oo
d
hope
th
at
t
ho
s
e
fl
ows
h
av
e
a
bi
gg
er
pro
ba
b
i
l
i
ty
to
q
ui
t
the
s
y
s
tem
and
l
ea
v
e
f
ewer
a
da
p
ted
c
on
ne
c
ti
on
s
in
th
e
s
y
s
tem
[10
].
T
h
e
ma
n
ag
e
me
nt
of
per
fl
ow
al
go
r
i
t
hm
is
v
er
y
c
om
pl
ex
as
the
C
al
l
A
dm
i
s
s
i
on
Cont
r
ol
(
CA
C)
m
us
t
s
ea
r
c
h
for
t
he
l
on
ge
s
t
d
urat
i
on
c
a
l
l
in
th
e
s
y
s
tem
ea
c
h
ti
me
.
It
c
an
c
r
e
ate
un
f
ai
r
n
es
s
be
c
au
s
e
if
the
s
y
s
tem
s
tat
e
do
es
not
c
ha
ng
e
fr
e
qu
en
tl
y
,
it
c
a
n
ha
v
e
s
om
e
c
on
ne
c
ti
on
s
ad
ap
te
d
for
l
on
g
t
i
me
whi
l
e
ot
he
r
s
are
g
oi
n
g
at
fu
l
l
or
at
h
i
gh
s
p
ee
d
.
In
th
e
ba
nd
w
i
dt
h
a
da
pt
ati
on
s
c
he
m
e,
w
hi
c
h
is
ba
s
e
d
on
per
c
l
as
s
de
gra
da
t
i
o
n
[12
],
when
a
new
us
er
arr
i
v
es
a
nd
t
he
r
e
are
no
en
ou
g
h
r
e
s
ou
r
c
es
in
the
s
y
s
tem
,
the
s
y
s
tem
s
ho
u
l
d
as
k
al
l
c
on
n
ec
ti
o
ns
in
th
e
l
o
wes
t
pri
orit
y
l
ev
el
to
ad
ap
t
un
ti
l
t
he
y
r
ea
c
h
th
ei
r
mi
n
i
m
um
ac
c
ep
tab
l
e
l
ev
el
an
d
if
the
r
es
ou
r
c
es
is
s
ti
l
l
not
en
o
ug
h
the
proc
es
s
c
on
ti
n
ue
s
wi
th
th
e
n
ex
t
prio
r
i
ty
l
ev
e
l
(
c
l
as
s
)
.
If
al
l
c
on
n
ec
ti
o
ns
that
ha
v
e
l
ow
er
or
eq
u
al
pri
orit
y
to
the
new
c
on
ne
c
ti
on
ha
v
e
ad
a
pte
d
to
t
he
i
r
m
i
n
i
mu
m
an
d
s
t
i
l
l
the
r
e
is
no
en
ou
gh
r
es
ou
r
c
e,
th
i
s
n
ew
c
on
n
ec
ti
o
n
wi
l
l
be
r
ej
ec
ted
.
T
he
ad
a
pta
t
i
on
a
l
g
orit
h
m
b
as
ed
on
per
c
l
as
s
s
c
he
me
is
de
s
c
r
i
be
d
in
F
i
g
ure
5.
By
us
i
n
g
thi
s
per
c
l
as
s
me
th
od
,
be
t
ter
fa
i
r
ne
s
s
is
en
s
ured
be
t
wee
n
th
e
c
on
ne
c
ti
on
s
of
th
e
s
am
e
c
l
as
s
.
T
he
a
bo
v
e
i
de
a
a
l
s
o
c
an
i
m
prov
e
th
e
s
c
al
ab
i
l
i
ty
of
the
s
c
he
me
an
d
r
ed
uc
e
the
c
om
p
uta
t
i
o
n
c
om
p
l
ex
i
ty
.
F
i
gu
r
e
5.
P
s
eu
d
o
c
od
e
of
p
er
c
l
as
s
ad
a
pta
t
i
o
n
a
l
go
r
i
th
m
4.
P
er
f
o
r
m
ance
An
aly
sis
T
he
s
i
m
ul
a
ti
o
ns
are
c
arr
i
e
d
out
in
O
P
NE
T
14.5
[16]
wi
th
t
he
s
i
mu
l
at
i
on
pa
r
a
me
ters
me
nt
i
on
ed
in
T
ab
l
e
1.
S
i
mu
l
ati
on
n
etwo
r
k
c
r
ea
te
d
for
hy
brid
c
ou
pl
e
d
W
LA
N
and
U
MTS
i
nte
r
w
ork
i
ng
n
etwo
r
k
is
s
ho
wn
in
F
i
gu
r
e
6.
AP
of
W
L
A
N
is
c
on
ne
c
ted
to
Int
ern
et
s
erv
er
throug
h
IP
ba
c
k
bo
ne
wh
i
c
h
is
r
ou
t
ed
throu
gh
a
r
ou
ter.
Nod
e
B
of
UMTS
is
c
on
ne
c
te
d
to
the
s
erv
er
throug
h
RNC,
S
erv
i
ng
G
a
tewa
y
S
u
pp
ort
N
od
e
(
S
G
S
N)
and
G
P
RS
G
at
eway
S
up
po
r
t
Nod
e
(
G
G
S
N)
of
UMTS
ne
twork
that
is
c
on
n
ec
ted
to
IP
ba
c
k
bo
ne
,
w
hi
c
h
is
pa
th
of
a
l
oo
s
el
y
c
ou
p
l
e
d
arc
hi
tec
ture
.
A
l
s
o,
W
LA
N
AP
is
c
on
n
ec
ted
to
s
erv
er
th
r
ou
gh
G
G
S
N
of
U
MTS
,
wh
i
c
h
is
th
e
path
of
t
i
gh
t
c
o
up
l
ed
arc
hi
tec
tur
e.
T
h
us
,
in
hy
bri
d
c
o
up
l
ed
arc
hi
tec
ture
t
he
W
L
A
N
AP
is
c
on
ne
c
ted
to
Int
erne
t
s
erv
er
throu
gh
t
he
paths
pres
e
nt
in
l
oo
s
e
c
ou
p
l
ed
as
we
l
l
as
ti
g
ht
c
ou
pl
ed
s
c
en
ario
s
.
T
ab
l
e
1.
S
i
mu
l
at
i
on
P
ar
a
m
e
t
ers
S
l.
No
P
a
r
a
met
e
r
V
a
lue
S
l.
No
P
a
r
a
met
e
r
V
a
lue
1
U
MT
S
C
a
p
a
c
i
t
y
(U)
2
Mb/
s
4
W
L
A
N
to
U
MT
S
H
a
n
d
o
f
f
0
.
0
1
2
W
L
A
N
C
a
p
a
c
i
t
y
(W)
1
1
Mb/
s
5
S
e
s
s
ion
T
ime
180s
3
U
MT
S
to
W
L
A
N
H
a
n
d
o
f
f
0
.
0
5
6
S
im
u
la
t
ion
T
im
e
180s
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
MNIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
Q
ua
l
i
ty
of
S
erv
i
c
e
i
n
ba
n
dw
i
dth
ad
ap
te
d h
y
br
i
d U
MTS/
W
L
A
N
…
(
R.
S
h
an
k
ar
)
2807
F
i
gu
r
e
6.
Hy
br
i
d
c
ou
pl
ed
U
MTS
a
nd
W
L
A
N
i
nt
eg
r
at
ed
ne
twork
Cal
l
arr
i
v
a
l
s
in
th
e
s
i
mu
l
ati
on
fo
l
l
ow
an
i
nd
ep
en
d
en
t
P
oi
s
s
on
proc
es
s
a
nd
t
he
s
es
s
i
on
ti
me
of
e
ac
h
c
on
ne
c
ti
on
is
ex
p
on
e
nti
al
l
y
d
i
s
tr
i
b
ut
ed
.
It
is
we
l
l
k
no
w
n
th
a
t
drop
pi
ng
an
es
tab
l
i
s
he
d
c
om
m
un
i
c
a
ti
o
n
is
wor
s
e
than
r
ej
ec
ti
ng
a
new
c
al
l
.
T
he
r
ef
ore,
UMTS
r
es
erv
e
a
gu
ard
ba
n
dwi
dth
f
or
the
h
an
do
ff
c
al
l
s
to
r
ed
uc
e
the
h
an
do
ff
drop
pi
ng
pro
ba
b
i
l
i
ty
.
T
he
r
es
erv
ed
gu
ard
b
an
d
wi
d
th
c
an
be
ei
th
er
s
tat
i
c
or
dy
na
m
i
c
[17,
18
]
.
T
he
dy
n
am
i
c
ap
pro
ac
h
oft
e
n
ou
tp
erfor
ms
th
e
s
tat
i
c
at
th
e
ex
pe
ns
e
of
ge
ne
r
at
i
n
g
m
ore
c
on
tr
o
l
ov
er
he
a
ds
[19
-
22].
H
owev
er,
the
s
tat
i
c
a
pp
r
oa
c
h
is
often
att
r
ac
ti
v
e
in
prac
ti
c
e
ow
i
n
g
to
i
ts
de
s
i
g
n
s
i
mp
l
i
c
i
ty
.
In
t
he
s
i
mu
l
at
i
o
n,
a
s
tat
i
c
g
ua
r
d
ba
nd
w
i
dt
h,
5%
of
t
he
s
y
s
tem
c
ap
ac
i
ty
is
em
p
l
oy
e
d
to
de
a
l
wi
t
h
ha
nd
off
c
al
l
s
.
T
he
i
n
teg
r
at
ed
n
etwo
r
k
in
the
s
i
m
ul
a
ti
o
n
c
on
s
i
s
ts
of
one
UMTS
and
one
W
L
A
N
ho
ts
po
t.
As
W
LA
N
has
h
i
g
he
r
c
a
pa
c
i
ty
an
d
is
c
he
ap
er
th
an
U
MTS,
ha
nd
o
ff
proba
bi
l
i
ty
fr
om
UMTS
to
W
L
A
N
is
5
ti
me
s
as
mu
c
h
as
th
at
fr
om
W
LA
N
to
U
MTS.
T
he
s
i
m
ul
a
ti
o
n
r
es
u
l
ts
de
m
on
s
tr
at
e
the
eff
ec
ti
v
en
es
s
of
the
ba
nd
wi
dth
a
da
p
tat
i
on
in
th
e
i
nte
r
wor
k
i
n
g
ne
t
wor
k
.
T
he
hy
brid
c
ou
p
l
e
d
i
nte
r
w
ork
i
ng
ne
t
wor
k
tha
t
wor
k
s
under
s
a
me
tr
a
ffi
c
l
oa
d
and
no
r
m
al
op
erati
on
c
on
di
ti
o
n
w
i
th
non
-
a
da
pt
i
v
e,
per
fl
o
w
a
nd
per
c
l
as
s
fr
am
ework
r
es
p
ec
ti
v
el
y
wer
e
s
i
mu
l
at
ed
.
E
ac
h
s
i
m
ul
ati
on
was
ex
ec
ute
d
u
nti
l
the
s
y
s
tem
r
ea
c
he
s
i
ts
s
tab
l
e
s
tat
e.
To
me
as
ure
t
he
s
y
s
tem
pe
r
forma
nc
e
me
r
i
ts
,
t
hroug
hp
ut
a
nd
d
el
ay
[23
-
27
]
wer
e
ob
s
erv
ed
.
Q
oS
p
aram
ete
r
s
s
uc
h
as
c
al
l
bl
oc
k
i
ng
proba
bi
l
i
t
i
es
a
nd
d
ata
bl
oc
k
i
ng
prob
ab
i
l
i
ti
es
w
ere
a
l
s
o
ex
am
i
ne
d
.
4.1
.
T
h
r
o
u
g
h
p
u
t
T
he
thro
ug
hp
ut
of
ex
i
s
ti
n
g
hy
bri
d
c
ou
pl
e
d
i
nte
r
w
ork
i
ng
n
etw
ork
wi
th
no
n
-
ad
ap
ti
v
e
ba
nd
wi
d
th
an
d
t
he
ne
t
wor
k
wi
th
pro
po
s
ed
a
da
p
ti
v
e
ba
nd
wi
d
th
s
c
h
em
es
is
s
h
o
wn
in
F
i
g
ure
7.
It
is
ob
s
erv
e
d
fr
o
m
the
f
i
g
ure
t
ha
t
th
e
hy
brid
arc
hi
tec
ture
w
i
th
p
er
c
l
as
s
ad
ap
t
at
i
on
off
ers
h
i
gh
throug
hp
u
t
c
om
p
ared
to
non
-
ad
a
pti
v
e
and
per
f
l
ow
c
o
un
terp
arts
.
It
is
du
e
to
the
tr
aff
i
c
of
us
ers
fl
o
w
t
hroug
h
two
pa
ths
pres
en
t
be
tw
ee
n
th
e
us
ers
and
i
nte
r
n
et.
M
oreov
er,
th
e
pe
r
c
l
as
s
hy
brid
c
ou
pl
e
d
arc
hi
tec
t
ure
off
er
ed
tr
aff
i
c
,
for
c
on
s
i
de
r
ab
l
y
mo
r
e
us
ers
w
i
th
hi
gh
er
b
a
nd
wi
dth
th
an
per
fl
ow
and
non
-
a
da
pt
i
v
e
s
c
h
em
es
,
it
is
ob
s
erv
ed
th
at
the
t
hroug
hp
u
t
in
per
c
l
as
s
is
5%
an
d
7%
hi
gh
er
t
ha
n
non
-
a
da
p
ti
v
e
and
per
fl
ow
ad
ap
t
i
v
e
s
c
he
me
s
r
es
pe
c
ti
v
e
l
y
.
T
he
throug
hp
u
t
is
mo
d
erate
in
n
on
-
ad
ap
t
i
v
e
s
c
he
me
as
it
a
c
c
om
mo
d
ate
s
fewe
r
us
ers
whos
e
ba
nd
w
i
dth
is
no
t
m
od
i
fi
ed
t
ha
n
p
er
fl
o
w
a
nd
per
c
l
as
s
c
ou
nt
er
pa
r
ts
.
It
is
d
ue
to
i
nc
r
ea
s
ed
us
ers
wi
th
l
es
s
ba
nd
w
i
dt
h
in
th
e
n
etwo
r
k
,
per
fl
ow
s
c
he
me
h
as
l
e
as
t
throug
hp
u
t.
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
MNIK
A
V
ol
.
1
7
,
No
.
6
,
D
ec
em
b
er
20
19
:
28
0
3
-
2811
2808
F
i
gu
r
e
7.
T
hro
ug
h
pu
t
c
o
mp
aris
on
4.2.
D
ela
y
T
he
en
d
to
end
de
l
ay
of
ex
i
s
ti
ng
hy
br
i
d
c
o
up
l
ed
i
nt
erw
ork
i
ng
n
etwo
r
k
w
i
th
non
-
a
d
ap
ti
v
e
ba
nd
wi
d
th
an
d
t
he
ne
t
wor
k
wi
th
pro
po
s
ed
a
da
p
ti
v
e
ba
nd
wi
d
th
s
c
h
em
es
is
s
h
o
wn
in
F
i
g
ure
8,
it
is
ev
i
d
en
t
fr
o
m
th
e
fi
g
ure
that
th
e
de
l
ay
is
mo
de
r
at
e
in
per
c
l
as
s
s
c
he
me
wh
i
c
h
is
1.4%
l
es
s
er
and
2.8%
hi
gh
er
tha
n
per
fl
ow
and
no
n
-
ad
a
pti
v
e
s
c
he
me
s
r
es
p
ec
ti
v
el
y
,
as
it
tak
es
c
om
pa
r
at
i
v
el
y
l
es
s
er
t
i
m
e
for
ad
ap
ta
ti
o
n
in
per
c
l
as
s
th
an
per
f
l
ow
s
c
he
me
an
d
in
non
-
a
da
pt
ati
on
hy
bri
d
c
o
up
l
i
n
g
t
i
m
e
is
n
ot
r
eq
u
i
r
ed
.
It
is
al
s
o
ev
i
de
nt
th
at
de
l
ay
is
4.2%
hi
gh
er
in
per
f
l
ow
th
an
no
n
-
ad
a
pti
v
e
s
c
he
me
as
in
p
er
f
l
ow
m
ore
c
om
pl
ex
a
l
go
r
i
th
m
is
i
nv
ol
v
ed
wh
i
c
h
s
ea
r
c
h
for
th
e
fl
ow
t
ha
t
are
pres
en
t
in
t
he
s
y
s
tem
f
or
l
o
ng
er
d
urati
on
.
F
i
gu
r
e
8.
D
el
ay
c
o
mp
ar
i
s
on
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
MNIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
Q
ua
l
i
ty
of
S
erv
i
c
e
i
n
ba
n
dw
i
dth
ad
ap
te
d h
y
br
i
d U
MTS/
W
L
A
N
…
(
R.
S
h
an
k
ar
)
2809
4.3
.
C
all
B
locking
P
r
o
b
abi
lity
(
CBP)
B
l
oc
k
i
ng
prob
ab
i
l
i
ty
r
ep
r
es
en
ts
th
e
am
o
un
t
of
tr
aff
i
c
bl
oc
k
ed
by
t
he
ne
t
wor
k
,
wh
e
n
us
ers
att
em
pt
to
get
th
e
s
erv
i
c
e
f
r
om
th
e
ne
tw
ork
.
Rat
i
o
of
t
ota
l
v
oi
c
e
c
al
l
s
s
erv
ed
to
t
ota
l
v
oi
c
e
c
al
l
s
att
em
pte
d,
in
t
he
n
etwo
r
k
is
k
no
wn
as
Cal
l
B
l
oc
k
i
ng
P
r
o
ba
b
i
l
i
ty
(
CB
P
)
.
F
i
g
ure
9
s
ho
ws
the
ev
al
ua
t
ed
CB
P
wi
t
h
r
es
pe
c
t
to
us
ers
,
b
as
ed
on
E
r
l
an
g
B
for
12
00
us
ers
.
F
r
o
m
t
he
fi
gu
r
e,
it
is
ev
i
d
en
t
th
at
as
ma
ny
as
800
us
ers
c
an
be
ac
c
om
mo
da
ted
in
th
e
ne
tw
ork
wi
t
h
no
n
-
ad
ap
t
i
v
e
s
c
he
me
wi
tho
ut
any
b
l
oc
k
i
ng
wher
ea
s
it
is
po
s
s
i
b
l
e
to
ac
c
om
mo
da
t
e
10
00
us
er
s
in
the
c
as
e
of
per
f
l
ow
me
tho
d
w
i
th
z
ero
bl
oc
k
i
ng
pro
ba
b
i
l
i
ty
,
al
s
o
it
is
ob
s
erv
ed
tha
t
mo
r
e
th
an
10
0
0
us
ers
c
an
be
ac
c
om
mo
da
t
ed
in
per
c
l
as
s
s
c
he
me
w
i
th
v
ery
l
es
s
b
l
oc
k
i
n
g
proba
bi
l
i
ty
.
It
is
ob
s
erv
ed
that
the
per
c
l
as
s
hy
br
i
d
c
ou
p
l
i
n
g
has
2
7.2
%
l
es
s
er
CB
P
th
an
no
n
-
ad
a
pti
v
e
n
e
twork
an
d
it
is
12.9%
l
es
s
tha
n
per
f
l
ow
h
y
brid
c
o
up
l
i
n
g.
T
he
be
tt
er
p
erfor
ma
nc
e
of
p
er
c
l
as
s
hy
brid
c
ou
p
l
i
ng
is
due
to
th
e
of
fered
c
ap
ac
i
t
y
is
l
arg
e
as
c
om
pa
r
e
d
to
per
fl
ow
and
no
n
-
ad
a
pti
v
e
arc
hi
t
ec
tures
.
In
per
f
l
ow
hy
br
i
d
c
ou
pl
i
n
g,
the
C
B
P
is
14
.
3%
l
es
s
c
o
mp
are
d
to
n
on
-
a
da
pt
i
v
e
s
c
he
me
of
hy
br
i
d
ne
twork
,
thus
it
c
an
ac
c
om
mo
da
te
l
es
s
er
nu
mb
er
of
us
ers
t
ha
n
p
er
c
l
as
s
hy
brid
c
o
up
l
ed
ne
twork
a
nd
hi
g
he
r
n
um
b
er
of
us
er
t
ha
n
non
-
ad
a
pti
v
e
hy
brid
n
etw
ork
.
F
i
gu
r
e
9.
C
B
P
4.4
.
D
ata
Blo
ck
ing
P
r
o
b
a
b
ility
(
DBP)
DB
P
is
t
he
r
a
ti
o
of
tot
a
l
data
s
erv
ed
by
t
he
ne
tw
o
r
k
to
the
tot
al
da
t
a
at
tem
pte
d
in
the
ne
tw
ork
.
F
i
g
ure
10
s
ho
ws
the
an
a
l
y
z
ed
DB
P
a
ga
i
ns
t
nu
mb
er
of
us
ers
.
F
r
om
the
f
i
g
ure,
it
is
ob
s
erv
ed
t
ha
t
t
he
per
c
l
as
s
hy
brid
c
ou
pl
ed
n
etw
ork
off
ered
l
es
s
er
DB
P
du
e
to
i
ts
i
nc
r
ea
s
ed
c
ap
ac
i
ty
c
o
mp
are
d
to
per
fl
ow
and
no
n
-
ad
ap
t
i
v
e
hy
brid
arc
hi
t
ec
tures
.
F
r
o
m
t
he
f
i
g
ure,
it
is
ev
i
de
nt
th
at
mo
r
e
t
ha
n
85
0
data
us
ers
c
an
be
ac
c
om
mo
da
ted
in
th
e
ne
twork
w
i
t
h
no
n
-
ad
a
pti
v
e
s
c
he
me
wi
tho
ut
any
b
l
oc
k
i
ng
wher
ea
s
it
is
po
s
s
i
b
l
e
to
ac
c
om
mo
da
t
e
10
00
us
er
s
in
th
e
c
as
e
of
per
f
l
ow
me
tho
d
w
i
th
z
ero
bl
oc
k
i
ng
pro
ba
b
i
l
i
ty
,
al
s
o
it
is
ob
s
erv
ed
tha
t
mo
r
e
th
an
10
0
0
us
ers
c
an
be
ac
c
om
m
od
a
ted
in
per
c
l
as
s
s
c
he
me
wi
t
h
v
ery
l
es
s
b
l
oc
k
i
ng
pro
ba
b
i
l
i
ty
.
T
hu
s
,
the
n
etwo
r
k
w
i
th
per
c
l
as
s
s
c
he
me
h
as
18
.
2%
l
es
s
er
DB
P
th
an
non
-
a
da
p
ti
v
e
ne
tw
ork
and
9%
l
es
s
th
an
per
f
l
ow
hy
bri
d
c
ou
p
l
i
ng
.
In
p
er
fl
ow
a
da
pta
t
i
on
,
the
DB
P
is
9%
l
es
s
c
om
pa
r
ed
to
non
-
a
da
pt
i
v
e
hy
bri
d
c
ou
p
l
i
ng
,
as
it
c
an
ac
c
om
mo
da
t
e
l
es
s
er
nu
m
be
r
of
us
ers
than
p
er
c
l
as
s
hy
brid
c
o
up
l
i
n
g
ne
twork
an
d
hi
gh
er
nu
m
be
r
of
us
er
th
an
n
on
-
ad
ap
t
i
v
e
n
etwo
r
k
.
0
200
400
600
800
1000
1200
0
0
.
0
1
0
.
0
2
0
.
0
3
0
.
0
4
0
.
0
5
0
.
0
6
0
.
0
7
N
u
m
b
e
r
o
f
u
s
e
r
s
C
a
l
l
b
l
o
c
k
i
n
g
p
r
o
b
a
b
i
l
i
t
y
H
y
b
r
i
d
P
C
H
y
b
r
i
d
P
F
H
y
b
r
i
d
N
o
n
a
d
a
p
t
i
v
e
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
MNIK
A
V
ol
.
1
7
,
No
.
6
,
D
ec
em
b
er
20
19
:
28
0
3
-
2811
2810
F
i
gu
r
e
10.
DB
P
5.
Co
n
clus
ion
Int
eg
r
a
ti
n
g
W
L
A
N
an
d
U
M
T
S
n
etwo
r
k
s
ef
fi
c
i
en
t
l
y
wi
t
h
Q
o
S
s
up
po
r
t
is
a
c
ha
l
l
e
ng
i
ng
tas
k
.
P
r
op
os
e
d
hy
bri
d
c
ou
pl
e
d
i
n
terwor
k
i
n
g
n
etwo
r
k
is
eff
i
c
i
en
t
c
o
mp
ared
to
l
o
os
l
y
and
t
i
g
ht
c
ou
pl
ed
ne
tw
ork
.
T
he
ge
n
e
r
i
c
r
es
erv
ati
on
ba
s
e
d
Q
o
S
mo
de
l
f
or
th
e
l
oo
s
l
y
c
o
up
l
e
d
c
el
l
u
l
ar
a
nd
W
L
A
N
i
nt
erw
ork
i
ng
n
etwo
r
k
us
es
ba
n
dwi
dth
ad
ap
t
ati
o
n
b
as
ed
on
per
fl
ow
and
per
c
l
as
s
me
c
ha
n
i
s
m
for
e
ns
urin
g
en
ha
nc
ed
Q
o
S
.
In
thi
s
p
ap
er,
hy
brid
c
ou
p
l
e
d
UMT
S
an
d
W
LA
N
ba
s
e
d
i
nte
r
w
ork
i
ng
n
etwo
r
k
wi
t
h
ba
nd
wi
d
th
ad
ap
t
ati
on
ba
s
e
d
on
per
fl
ow
an
d
per
c
l
as
s
al
go
r
i
t
hm
are
an
a
l
y
s
ed
and
th
ei
r
pe
r
f
orm
an
c
e
is
c
o
mp
are
d.
T
he
r
es
ul
ts
s
h
ow
that
p
er
c
l
as
s
me
tho
d
a
da
pts
a
l
l
the
f
l
ows
in
c
ertai
n
c
l
as
s
and
o
ffe
r
ed
be
tt
er
pe
r
for
ma
nc
e
t
ha
n
per
fl
ow.
It
is
ev
i
de
nt
fr
om
the
hi
gh
er
thro
ug
h
pu
t
wh
i
c
h
is
5
an
d
7
p
erc
en
t
hi
gh
er,
tha
n
per
f
l
o
w
a
nd
no
n
-
ad
a
pti
v
e
s
c
he
me
s
.
It
al
s
o
prov
i
de
s
c
om
pu
ta
ti
o
n
s
i
mp
l
i
c
i
ty
and
ex
ec
u
tes
f
as
ter,
whi
c
h
is
i
m
po
r
ta
nt
fo
r
an
ad
m
i
s
s
i
on
c
on
tr
ol
c
om
p
on
e
nt
that
c
a
n
be
u
nd
ers
to
od
fr
om
the
mo
de
r
ate
en
d
to
e
nd
de
l
ay
pres
en
t
th
at
is
1.4
pe
r
c
en
t
l
es
s
er
de
l
ay
th
an
in
p
er
fl
ow
s
c
he
m
e.
P
er
c
l
as
s
method
us
es
the
s
y
s
tem
r
es
ou
r
c
es
mo
r
e
eff
i
c
i
e
nt
l
y
than
t
he
per
fl
ow
me
t
ho
d
s
i
nc
e
it
ad
ap
te
d
th
e
c
on
ne
c
t
i
on
s
ba
s
ed
on
r
eq
ui
r
e
me
nt
of
th
e
new
c
o
nn
ec
ti
on
,
w
hi
c
h
is
ob
s
erv
e
d
fr
om
th
e
l
e
as
t
c
al
l
and
data
b
l
oc
k
i
n
g
proba
bi
l
i
ty
wh
i
c
h
are
27
.
2,
12.9
pe
r
c
en
t
and
it
is
18
.
2,
9
pe
r
c
en
t
l
es
s
er
th
an
n
on
-
ad
ap
t
i
v
e
a
nd
ba
nd
wi
d
th
a
da
pta
t
i
on
ba
s
e
d
on
per
f
l
ow
s
c
he
me
r
es
p
ec
ti
v
el
y
.
Ref
er
en
ce
s
[1
]
3
G
PP
TK
2
2
.9
3
4
.
F
e
a
s
i
b
i
l
i
t
y
s
tu
d
y
on
3
G
PP
s
y
s
te
m
to
wir
e
l
e
s
s
l
o
c
a
l
a
r
e
a
n
e
two
r
k
(
WL
AN
)
i
n
te
rwor
k
i
n
g
v
.
I
.
0
.0
.
2
0
0
2
.
[2
]
3
G
PP
TS
2
3
.2
3
4
.
G
ro
u
p
Se
rv
i
c
e
s
a
n
d
Sy
s
te
m
As
p
e
c
t
s
.
3
G
PP
s
y
s
t
e
m
s
to
wir
e
l
e
s
s
l
o
c
a
l
a
re
a
n
e
two
rk
i
n
t
e
rwork
i
n
g
S
y
s
t
e
m
Des
c
ri
p
ti
o
n
v.
6
.
2
.0
.
2
0
0
4
.
[3
]
Ah
m
a
v
a
a
ra
K
,
Ha
v
e
ri
n
e
n
H
,
Pi
c
h
n
a
R
.
In
te
g
r
a
ti
o
n
of
w
i
re
l
e
s
s
L
AN
a
n
d
3G
wir
e
l
e
s
s
-
I
n
te
rwork
i
n
g
a
rc
h
i
te
c
tu
re
b
e
twe
e
n
3
G
PP
a
n
d
WL
AN
s
y
s
te
m
s
.
Com
m
u
n
i
c
a
ti
o
n
s
M
a
g
a
z
i
n
e
.
2
0
0
3
;
4
1
(1
1
)
:
74
-
81.
[4
]
So
n
g
JY
,
L
e
e
SW
,
Cho
DH
.
Hy
b
ri
d
c
o
u
p
l
i
n
g
s
c
h
e
m
e
fo
r
UM
TS
and
wir
e
l
e
s
s
L
AN
i
n
te
rwork
i
n
g
s
.
Pro
c
e
e
d
i
n
g
s
of
IEEE
Ve
h
i
c
u
l
a
r
Te
c
h
n
o
l
o
g
y
Con
fe
r
e
n
c
e
.
O
rl
a
n
d
o
.
2
0
0
3
;
4:
2
2
4
7
-
2
2
5
1
.
[5
]
Liu
C
,
Zh
o
u
C
.
HC
RAS:
A
n
o
v
e
l
h
y
b
ri
d
i
n
te
rn
e
two
rk
i
n
g
a
rc
h
i
t
e
c
t
u
re
b
e
twe
e
n
WL
AN
a
n
d
UM
TS
c
e
l
l
u
l
a
r
n
e
two
r
k
s
.
Pro
c
e
e
d
i
n
g
s
of
IEEE
Con
s
u
m
e
r
Com
m
u
n
i
c
a
ti
o
n
s
a
n
d
Net
wor
k
i
n
g
Con
fe
re
n
c
e
.
Nev
a
d
a
.
2
0
0
5
;
374
-
3
7
9
.
0
200
400
600
800
1000
1200
0
0
.
0
2
0
.
0
4
0
.
0
6
0
.
0
8
0
.
1
0
.
1
2
N
u
m
b
e
r
o
f
u
s
e
r
s
D
a
t
a
b
l
o
c
k
i
n
g
p
r
o
b
a
b
i
l
i
t
y
H
y
b
r
i
d
P
C
H
y
b
r
i
d
P
F
H
y
b
r
i
d
N
o
n
a
d
a
p
t
i
v
e
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
MNIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
Q
ua
l
i
ty
of
S
erv
i
c
e
i
n
ba
n
dw
i
dth
ad
ap
te
d h
y
br
i
d U
MTS/
W
L
A
N
…
(
R.
S
h
an
k
ar
)
2811
[6
]
So
n
g
J
Y,
Lee
HJ
,
Lee
SH,
L
e
e
SW,
Cho
DH
.
Hy
b
r
i
d
c
o
u
p
l
i
n
g
s
c
h
e
m
e
fo
r
U
M
TS
a
n
d
wir
e
l
e
s
s
L
AN
i
n
te
rwor
k
i
n
g
.
AEU
In
te
rn
a
ti
o
n
a
l
J
o
u
rn
a
l
of
E
l
e
c
tro
n
i
c
s
and
Co
m
m
u
n
i
c
a
ti
o
n
s
.
2
0
0
7
;
6
1
(5
):
3
2
9
–
3
3
6
.
[7
]
Zh
o
u
Y,
Ron
g
Y,
Cho
i
HA,
Ki
m
J
H,
So
h
n
J,
Cho
i
HI
.
Uti
l
i
t
y
-
b
a
s
e
d
l
o
a
d
b
a
l
a
n
c
i
n
g
in
WL
AN/UM
TS
i
n
te
rn
e
two
rk
i
n
g
s
y
s
te
m
s
.
IEEE
Rad
i
o
and
Wi
re
l
e
s
s
Sy
m
p
o
s
i
u
m
.
O
rl
a
n
d
o
.
2008
:
5
8
7
–
5
9
0
.
[8
]
Al
-
Hel
a
l
i
AA,
M
a
h
m
o
u
d
A,
Al
-
Kh
a
ro
b
i
T,
Sh
e
l
ta
m
i
T
.
S
i
m
u
l
a
ti
o
n
of
a
Nov
e
l
Dua
l
-
M
o
d
e
Us
e
r
Eq
u
i
p
m
e
n
t
De
s
i
g
n
fo
r
B3
G
Net
work
s
U
s
i
n
g
O
PNET
.
Th
i
rd
In
te
rn
a
ti
o
n
a
l
Con
fe
re
n
c
e
o
n
M
o
d
e
l
i
n
g
.
Si
m
u
l
a
ti
o
n
a
n
d
Ap
p
l
i
e
d
O
p
ti
m
i
z
a
ti
o
n
.
Sh
a
rj
a
h
.
2
0
0
9
:
20
-
2
2
.
[9
]
Hui
SY,
Ye
u
n
g
KH
.
Cha
l
l
e
n
g
e
s
in
th
e
M
i
g
r
a
ti
o
n
to
4G
M
o
b
i
l
e
Sy
s
te
m
s
.
IEEE
Com
m
u
n
.
M
a
g
a
z
i
n
e
.
2003;
4
1
(1
2
):
54
-
59.
[1
0
]
Wa
n
g
XG
,
M
i
n
G,
M
e
l
l
o
r
JE
.
Ad
a
p
t
i
v
e
Q
o
S
c
o
n
tro
l
in
c
e
l
l
u
l
a
r
and
WL
AN
i
n
t
e
rwork
i
n
g
n
e
two
rk
s
.
Pro
c
e
e
d
i
n
g
s
of
Se
c
o
n
d
i
n
te
r
n
a
ti
o
n
a
l
wor
k
i
n
g
c
o
n
fe
r
e
n
c
e
on
Pe
rfo
rm
a
n
c
e
m
o
d
e
l
l
i
n
g
a
n
d
e
v
a
l
u
a
t
i
o
n
of
h
e
te
ro
g
e
n
e
o
u
s
n
e
tw
o
rk
s
(
HET
-
NETs
)
.
We
s
t
Yo
r
k
s
h
i
re
.
2
0
0
4
:
1
-
9.
[1
1
]
Wa
n
g
XG
,
M
i
n
G,
M
e
l
l
o
r
J
E,
Al
-
Be
g
a
i
n
K,
G
u
a
n
L
.
An
a
d
a
p
ti
v
e
Q
o
S
fra
m
e
wor
k
fo
r
i
n
te
g
ra
t
e
d
c
e
l
l
u
l
a
r
and
W
L
AN
n
e
two
r
k
s
.
El
s
e
v
i
e
r
J
o
u
rn
a
l
on
Com
p
u
t
e
r
Net
wor
k
s
.
2
0
0
5
;
4
7
(2
):
167
-
183.
[1
2
]
Wa
n
g
XG
,
M
i
n
G,
M
e
l
l
o
r
J,
Al
-
Be
g
a
i
n
K
.
A
Q
o
S
b
a
s
e
d
b
a
n
d
wi
d
th
m
a
n
a
g
e
m
e
n
t
s
c
h
e
m
e
in
h
e
te
ro
g
e
n
e
o
u
s
wir
e
l
e
s
s
n
e
two
rk
s
.
I
n
te
r
n
a
ti
o
n
a
l
J
o
u
rn
a
l
of
s
i
m
u
l
a
ti
o
n
.
2
0
0
4
;
5
(1
-
2
):
9
-
1
7
.
[1
3
]
Wa
n
g
X,
Sc
h
u
l
z
r
i
n
n
e
H
.
In
t
e
g
ra
te
d
re
s
o
u
r
c
e
n
e
g
o
t
i
a
t
i
o
n
,
p
ri
c
i
n
g
&
Q
o
S
a
d
a
p
ta
ti
o
n
fra
m
e
wor
k
fo
r
m
u
l
t
i
m
e
d
i
a
a
p
p
l
i
c
a
t
i
o
n
s
.
IEEE
J
o
u
rn
a
l
on
Se
l
e
c
t
e
d
Are
a
s
in
Com
m
u
n
i
c
a
ti
o
n
s
.
2
0
0
0
;
1
8
(1
2
):
2
5
1
4
–
2
5
2
9
.
[1
4
]
M
i
rh
a
k
k
a
k
M,
Sc
h
u
l
t
N,
Th
o
m
s
o
n
D
.
Dy
n
a
m
i
c
b
a
n
d
wi
d
th
m
a
n
a
g
e
m
e
n
t
a
n
d
a
d
a
p
ti
v
e
a
p
p
l
i
c
a
ti
o
n
s
fo
r
a
v
a
ri
a
b
l
e
b
a
n
d
wi
d
th
wir
e
l
e
s
s
e
n
v
i
ro
n
m
e
n
t.
IEEE
J
o
u
rn
a
l
on
Se
l
e
c
te
d
Are
a
s
in
Com
m
u
n
i
c
a
ti
o
n
s
.
2001;
1
9
(1
0
):
1
9
8
4
–
1
9
9
7
.
[1
5
]
M
a
ri
a
p
p
a
n
B,
Pa
n
d
i
a
n
S
.
A
F
ra
m
e
wor
k
fo
r
i
m
p
ro
v
i
s
i
n
g
Se
r
v
i
c
e
of
M
u
l
ti
m
e
d
i
a
Ap
p
l
i
c
a
t
i
o
n
s
in
M
o
b
i
l
e
Net
work
s
.
Th
e
In
t
e
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
of
M
u
l
t
i
m
e
d
i
a
&
Its
Ap
p
l
i
c
a
ti
o
n
s
.
2
0
1
1
;
3
(
2
):
82
-
1
0
0
.
[1
6
]
Nare
j
o
S,
Ku
l
s
o
o
m
F,
J
a
w
a
i
d
MM
.
P
e
rfo
r
m
a
n
c
e
E
v
a
l
u
a
ti
o
n
of
In
te
ra
c
ti
v
e
Vi
d
e
o
Str
e
a
m
i
n
g
o
v
e
r
Wi
M
AX
Net
w
o
rk
.
In
te
r
n
a
ti
o
n
a
l
J
o
u
rn
a
l
of
El
e
c
tri
c
a
l
a
n
d
Com
p
u
te
r
En
g
i
n
e
e
ri
n
g
.
2
0
1
7
;
7
(1
):
3
4
4
-
3
5
2
.
[1
7
]
Ram
a
n
a
th
a
n
P,
Si
v
a
l
i
n
g
a
m
K
M
,
Ag
ra
wa
l
P,
Ki
s
h
o
re
S
.
D
y
n
a
m
i
c
re
s
o
u
rc
e
a
l
l
o
c
a
ti
o
n
s
c
h
e
m
e
s
d
u
ri
n
g
h
a
n
d
o
ff
fo
r
m
o
b
i
l
e
m
u
l
ti
m
e
d
i
a
wir
e
l
e
s
s
n
e
two
rk
s
.
IEEE
J
SAC.
1999;
1
7
(
7
):
1
2
7
0
-
1
2
8
3
.
[1
8
]
O
l
i
v
e
i
r
a
C,
Ki
m
J
B,
S
u
d
a
T
.
Ad
a
p
ti
v
e
b
a
n
d
w
i
d
th
r
e
s
e
r
v
a
t
i
o
n
s
c
h
e
m
e
fo
r
h
i
g
h
s
p
e
e
d
m
u
l
t
i
m
e
d
i
a
wir
e
l
e
s
s
n
e
two
r
k
s
.
IEEE
J
o
u
rn
a
l
on
s
e
l
e
c
te
d
a
re
a
s
in
Com
m
u
n
i
c
a
ti
o
n
s
.
1
9
9
8
;
1
6
(6
):
858
-
8
7
4
.
[1
9
]
Wi
s
e
l
y
D,
E
a
rd
l
e
y
P
,
Bu
rn
e
s
s
L
.
IP
fo
r
3G
.
J
o
h
n
Wi
l
e
y
.
2
0
0
2
:
21
-
70.
[2
0
]
Ka
ri
m
i
O
B,
Fa
t
h
y
M
.
Ad
a
p
t
i
v
e
end
-
to
-
e
n
d
Q
o
S
fo
r
m
u
l
t
i
m
e
d
i
a
o
v
e
r
h
e
t
e
ro
g
e
n
e
o
u
s
n
e
two
rk
s
.
Com
p
u
te
rs
&
E
l
e
c
tri
c
a
l
E
n
g
i
n
e
e
ri
n
g
.
2
0
1
0
;
3
1
(1
):
45
-
5
5
.
[2
1
]
Sh
u
m
i
n
o
s
k
i
T,
J
a
n
e
v
s
k
i
T
.
Nov
e
l
A
d
a
p
t
i
v
e
Q
o
S
Fra
m
e
work
fo
r
I
n
te
g
r
a
te
d
UM
TS/
WL
AN
En
v
i
ro
n
m
e
n
t.
T
e
l
f
o
r
J
o
u
rn
a
l
.
2
0
1
3
;
5
(
1
):
14
–
1
9
.
[2
2
]
Fe
i
Y,
W
o
n
g
VW
,
L
e
u
n
g
V
.
Eff
i
c
i
e
n
t
Q
o
S
Pro
v
i
s
i
o
n
i
n
g
fo
r
Ad
a
p
ti
v
e
M
u
l
t
i
m
e
d
i
a
in
M
o
b
i
l
e
Com
m
u
n
i
c
a
ti
o
n
Net
wor
k
s
by
Rei
n
fo
rc
e
m
e
n
t
L
e
a
rn
i
n
g
.
M
o
b
i
l
e
Net
work
s
and
Ap
p
l
i
c
a
ti
o
n
s
.
2
0
0
6
;
1
1
(1
):
101
–
110.
[2
3
]
Sh
a
n
k
a
r
R,
Dha
n
a
n
j
a
y
a
n
P
.
An
a
d
a
p
ti
v
e
c
a
l
l
a
d
m
i
s
s
i
o
n
c
o
n
t
ro
l
s
c
h
e
m
e
wit
h
l
o
a
d
b
a
l
a
n
c
i
n
g
fo
r
Q
o
S
e
n
h
a
n
c
e
m
e
n
t
in
RoF
b
a
s
e
d
c
o
n
v
e
rg
e
d
UM
TS/
WL
AN.
I
n
te
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
of
Com
p
u
te
r
s
and
Com
m
u
n
i
c
a
ti
o
n
s
.
2
0
1
1
;
5
(4
):
271
-
2
7
8
.
[2
4
]
Sh
a
n
k
a
r
R,
Dan
a
n
j
a
y
a
n
P
.
An
o
s
p
f
b
a
s
e
d
u
m
ts
/wl
a
n
ra
d
i
o
o
v
e
r
f
i
b
e
r
c
o
n
v
e
rg
e
n
c
e
n
e
two
rk
fo
r
e
n
h
a
n
c
e
d
q
u
a
l
i
t
y
of
s
e
r
v
i
c
e
.
I
n
te
rn
a
t
i
o
n
a
l
J
o
u
r
n
a
l
of
Com
p
u
te
r
Sc
i
e
n
c
e
I
s
s
u
e
s
.
2
0
1
2
;
9
(
2
):
355
-
3
6
1
.
[2
5
]
Red
d
y
NR,
Pa
k
a
n
a
ti
CR,
Pa
d
m
a
v
a
th
a
m
m
a
M
.
P
e
rfo
r
m
a
n
c
e
En
h
a
n
c
e
m
e
n
t
of
TCP
Fri
e
n
d
l
y
Rat
e
Con
tro
l
Pr
o
to
c
o
l
o
v
e
r
Wi
r
e
d
N
e
two
rk
s
.
In
te
rn
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
El
e
c
tri
c
a
l
a
n
d
Com
p
u
te
r
En
g
i
n
e
e
r
i
n
g
.
2016;
6
(6
):
2
9
4
9
-
2
9
5
4
.
[2
6
]
Ib
ra
h
i
m
SK,
Kh
a
m
i
s
s
NN
.
O
p
ti
m
i
z
a
t
i
o
n
a
S
c
h
e
d
u
l
i
n
g
Al
g
o
ri
th
m
of
CA
in
L
TE
ADV.
TEL
KO
M
NIKA
Te
l
e
c
o
m
m
u
n
i
c
a
ti
o
n
Com
p
u
t
i
n
g
El
e
c
tro
n
i
c
s
a
n
d
Con
tr
o
l
.
2
0
1
8
;
1
6
(6
):
2
5
1
4
-
2
5
2
1
.
[2
7
]
Pra
m
u
k
a
n
t
o
ro
ES,
L
u
c
k
i
e
s
M,
Ba
k
h
ti
a
r
FA
.
Br
i
d
g
i
n
g
I
o
T
i
n
fra
s
tru
c
tu
r
e
a
n
d
c
l
o
u
d
a
p
p
l
i
c
a
ti
o
n
u
s
i
n
g
c
e
l
l
u
l
a
r
-
b
a
s
e
d
i
n
te
rn
e
t
g
a
te
w
a
y
d
e
v
i
c
e
.
TEL
KO
M
NIKA
Te
l
e
c
o
m
m
u
n
i
c
a
ti
o
n
Com
p
u
ti
n
g
El
e
c
tr
o
n
i
c
s
and
Con
tro
l
.
2
0
1
9
;
1
7
(3
):
1439
-
1446.
Evaluation Warning : The document was created with Spire.PDF for Python.