T
E
L
K
O
M
N
I
K
A
T
elec
o
m
m
un
ica
t
io
n Co
m
pu
t
ing
E
lect
ro
nics
a
nd
Co
ntr
o
l
Vo
l.
24
,
No
.
3
,
J
u
n
e
20
26
,
p
p
.
751
~
7
6
4
I
SS
N:
1
6
9
3
-
6
9
3
0
,
DOI
: 1
0
.
1
2
9
2
8
/TE
L
KOM
NI
KA.
v
24
i
3
.
2
7
6
0
9
751
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//jo
u
r
n
a
l.u
a
d
.
a
c.
id
/in
d
ex
.
p
h
p
/TELK
OM
N
I
K
A
Network
traf
fic
a
na
ly
sis
and ba
ndwidth
foreca
sting
for using
M
eta
’
s
P
ro
ph
et
:
a
ca
se study
Yus
uf
O
nim
is
i
I
s
a
a
c
1
,
Ay
o
dej
i J
a
mes
B
a
m
is
a
y
e
2
,
I
j
a
g
be
mi Adedo
t
un
1
,
T
heo
ph
il
us
O
lu
s
eg
un
Da
da
1
,
O
ny
em
ena
m
O
bia
j
ulu
J
o
hn
1
1
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
i
c
a
l
a
n
d
I
n
f
o
r
mat
i
o
n
E
n
g
i
n
e
e
r
i
n
g
,
C
o
l
l
e
g
e
o
f
E
n
g
i
n
e
e
r
i
n
g
,
La
n
d
mar
k
U
n
i
v
e
r
si
t
y
,
O
m
u
-
A
r
a
n
,
N
i
g
e
r
i
a
2
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
i
c
a
l
El
e
c
t
r
o
n
i
c
En
g
i
n
e
e
r
i
n
g
,
F
a
c
u
l
t
y
o
f
E
n
g
i
n
e
e
r
i
n
g
,
Tsh
w
a
n
e
U
n
i
v
e
r
si
t
y
o
f
Te
c
h
n
o
l
o
g
y
,
P
r
e
t
o
r
i
a
,
S
o
u
t
h
A
f
r
i
c
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Oct
16
,
2
0
2
5
R
ev
is
ed
Dec
1
7
,
2
0
2
5
Acc
ep
ted
J
an
30
,
2
0
2
6
Th
is
st
u
d
y
c
re
a
ted
a
f
o
rwa
rd
-
l
o
o
k
in
g
b
a
n
d
wid
t
h
p
re
d
icti
o
n
s
y
ste
m
fo
r
stu
d
e
n
ts
’
h
a
ll
s
o
f
re
sid
e
n
c
e
a
t
La
n
d
m
a
rk
Un
iv
e
rsity
.
T
h
e
sy
ste
m
u
se
s
M
e
ta
’
s
P
ro
p
h
e
t,
a
m
e
th
o
d
fo
r
a
n
a
l
y
z
in
g
p
a
tt
e
rn
s
in
d
a
ta
o
v
e
r
ti
m
e
,
a
n
d
w
a
s
train
e
d
o
n
p
a
st
i
n
tern
e
t
traffic
d
a
ta
fr
o
m
Oc
to
b
e
r
t
o
De
c
e
m
b
e
r
2
0
2
4
.
Th
e
sy
ste
m
wa
s
a
b
le
to
p
re
d
ict
fu
t
u
re
b
a
n
d
wid
th
u
sa
g
e
wit
h
o
v
e
r
9
0
%
a
c
c
u
ra
c
y
.
T
o
a
ss
e
ss
h
o
w
we
ll
t
h
e
s
y
ste
m
wo
rk
e
d
,
se
v
e
ra
l
c
o
m
m
o
n
m
e
tri
c
s
we
re
u
se
d
,
in
c
lu
d
in
g
m
e
a
n
a
b
so
l
u
te
e
rro
r
(
M
AE),
r
o
o
t
m
e
a
n
s
q
u
a
re
e
rro
r
(
RM
S
E),
a
n
d
m
e
a
n
a
b
s
o
lu
te
p
e
rc
e
n
tag
e
e
rro
r
(M
APE
)
.
T
h
e
M
AE
wa
s
c
a
lc
u
late
d
a
s
1
0
,
0
9
9
,
8
6
3
.
1
0
b
it
s
p
e
r
se
c
o
n
d
(
b
p
s)
,
a
n
d
t
h
e
RM
S
E
wa
s
1
3
,
5
7
0
,
9
5
9
.
5
8
b
p
s
.
Wh
il
e
th
e
m
e
a
n
sq
u
a
re
d
e
rro
r
(M
S
E)
a
p
p
e
a
rs
larg
e
n
u
m
e
rica
ll
y
,
th
is
is
a
n
ti
c
ip
a
ted
d
u
e
to
t
h
e
siz
e
o
f
th
e
b
a
n
d
wi
d
th
d
a
ta
in
v
o
l
v
e
d
in
it
s
c
a
lcu
latio
n
.
Im
p
o
rtan
tl
y
,
th
e
p
re
d
icti
o
n
e
rro
rs
a
re
c
o
n
si
d
e
re
d
re
a
so
n
a
b
le
wh
e
n
c
o
n
sid
e
re
d
in
re
latio
n
to
th
e
a
c
tu
a
l
p
e
a
k
b
a
n
d
wid
t
h
u
sa
g
e
,
w
h
ich
flu
c
tu
a
ted
b
e
twe
e
n
4
7
a
n
d
5
0
m
e
g
a
b
it
s
p
e
r
se
c
o
n
d
(M
b
p
s)
.
T
h
e
se
fin
d
i
n
g
s
s
u
g
g
e
st
t
h
a
t
m
a
c
h
in
e
lea
rn
in
g
c
a
n
b
e
a
v
a
lu
a
b
le
t
o
o
l
f
o
r
re
fi
n
in
g
n
e
two
r
k
in
fr
a
stru
c
tu
re
a
n
d
imp
ro
v
i
n
g
t
h
e
u
se
r
e
x
p
e
rien
c
e
q
u
a
li
ty
o
f
se
rv
ice
(Qo
S
)
in
e
n
v
i
ro
n
m
e
n
ts
with
m
a
n
y
u
se
rs,
su
c
h
a
s u
n
iv
e
rsi
ty
re
sid
e
n
c
e
s.
K
ey
w
o
r
d
s
:
Aca
d
em
ic
in
f
r
astru
ctu
r
e
B
an
d
wid
th
f
o
r
ec
asti
n
g
Ma
ch
in
e
lear
n
in
g
Me
ta
Pro
p
h
et
Netwo
r
k
tr
af
f
ic
a
n
aly
s
is
T
im
e
-
s
er
ies m
o
d
elin
g
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Yu
s
u
f
On
im
is
i I
s
aa
c
Dep
ar
tm
en
t o
f
E
lectr
ical
an
d
I
n
f
o
r
m
atio
n
E
n
g
in
ee
r
in
g
,
C
o
lleg
e
o
f
E
n
g
in
ee
r
i
n
g
,
L
a
n
d
m
ar
k
Un
iv
er
s
ity
Om
u
-
Ar
an
,
Nig
e
r
ia
E
m
ail:
y
u
s
u
f
.
is
aa
c@
lm
u
.
ed
u
.
n
g
1.
I
NT
RO
D
UCT
I
O
N
Netwo
r
k
tr
af
f
ic
an
al
y
s
is
en
tails
o
b
s
er
v
atio
n
,
q
u
an
tific
atio
n
,
an
d
in
ter
p
r
etatio
n
o
f
d
ata
p
a
ck
et
f
lo
w
in
s
id
e
a
co
m
m
u
n
icatio
n
n
etw
o
r
k
to
co
m
p
r
eh
en
d
u
s
ag
e
tr
en
d
s
,
en
h
an
ce
p
er
f
o
r
m
a
n
ce
,
an
d
id
en
tify
an
o
m
alies.
Acc
o
r
d
in
g
to
Olateju
[
1
]
,
th
e
b
asis
o
f
tr
af
f
ic
a
n
aly
s
is
is
th
e
h
an
d
lin
g
o
f
tr
af
f
ic
as
a
ti
m
e
s
er
ies,
u
tili
zin
g
h
is
to
r
ical
o
b
s
er
v
atio
n
s
to
f
o
r
ec
ast
f
u
tu
r
e
b
eh
av
io
r
[
2
]
.
T
h
i
s
s
tati
s
tical
p
er
s
p
ec
tiv
e
h
as
tr
ad
itio
n
ally
ass
is
ted
o
p
er
ato
r
s
in
r
ec
o
g
n
izin
g
tr
en
d
s
,
s
ea
s
o
n
ality
,
an
d
u
n
f
o
r
eseen
an
o
m
alies in
n
etwo
r
k
u
s
ag
e
[
3
]
.
T
h
e
s
ig
n
if
ican
ce
o
f
tr
af
f
ic
a
n
aly
s
is
h
as
escalate
d
co
n
s
id
er
ab
l
y
d
u
e
t
o
th
e
d
i
g
ital
r
ev
o
lu
tio
n
o
f
s
o
ciety
.
A
d
eb
ay
o
an
d
Och
ay
i
[
4
]
n
o
ted
th
at
telec
o
m
m
u
n
icatio
n
n
etwo
r
k
s
en
co
u
n
ter
h
ig
h
ly
ir
r
eg
u
lar
an
d
b
u
s
ty
tr
af
f
ic
r
esu
ltin
g
f
r
o
m
th
e
u
n
ex
p
ec
te
d
n
atu
r
e
o
f
i
n
ter
n
et
u
tili
za
tio
n
.
Netwo
r
k
ac
tiv
ities
s
u
ch
as
v
i
d
eo
s
tr
ea
m
in
g
,
o
n
lin
e
g
am
i
n
g
,
f
ile
d
o
wn
lo
a
d
s
,
clo
u
d
b
ac
k
u
p
s
,
v
id
e
o
co
n
f
er
e
n
cin
g
,
liv
e
s
tr
ea
m
in
g
,
p
ee
r
-
to
-
p
ee
r
(
P2
P)
f
ile
s
h
ar
in
g
ar
e
m
ajo
r
ca
u
s
e
o
f
s
u
d
d
en
s
p
ik
es
an
d
th
ey
n
ee
d
r
o
b
u
s
t
id
en
tific
atio
n
an
d
m
an
ag
em
en
t.
I
n
a
n
etwo
r
k
s
y
s
tem
v
o
id
o
f
m
o
n
ito
r
in
g
,
s
u
ch
ac
tiv
ities
g
o
u
n
id
en
tifie
d
s
u
b
s
eq
u
en
tly
ca
u
s
in
g
co
n
g
esti
o
n
,
lo
w
q
u
ality
o
f
s
er
v
ice
(
Qo
S)
,
an
d
in
th
e
wo
r
s
t
-
ca
s
e
s
ce
n
ar
io
,
to
tal
o
u
tag
e
lead
in
g
to
h
u
g
e
ec
o
n
o
m
ic
lo
s
s
.
I
s
s
a
et
a
l
.
[
5
]
u
s
in
g
Nig
er
ian
u
n
iv
er
s
ities
as
th
e
ca
s
e
s
tu
d
y
d
is
co
v
er
s
th
at
s
tu
d
y
in
g
th
e
tr
a
f
f
ic
p
atter
n
s
o
f
n
etwo
r
k
tr
af
f
i
c
p
atter
n
s
is
cr
itical
to
ef
f
ec
ti
v
e
allo
ca
tio
n
o
f
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
TEL
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
3
,
J
u
n
e
20
26
:
7
5
1
-
7
6
4
752
b
an
d
wid
th
d
u
r
in
g
p
ea
k
d
em
a
n
d
p
er
io
d
s
s
u
ch
as
r
eg
is
tr
atio
n
,
e
-
lear
n
in
g
ac
tiv
ities
,
ex
am
in
atio
n
p
er
io
d
.
T
h
is
s
tu
d
y
s
u
g
g
ests
th
e
n
ee
d
t
o
d
e
p
lo
y
a
r
ea
l
tim
e
m
o
n
ito
r
in
g
s
y
s
tem
with
a
p
r
e
d
ictiv
e
m
o
d
e
l
to
f
o
r
ec
ast
f
u
tu
r
e
tr
af
f
ic
p
atter
n
s
b
ased
o
n
th
e
p
ast
an
d
p
r
esen
t
d
ata.
Ak
id
i
et
a
l
.
[
6
]
in
th
eir
c
o
m
p
r
e
h
en
s
iv
e
r
ev
iew,
id
en
tifie
d
th
e
p
r
o
life
r
atio
n
o
f
in
ter
n
et
o
f
th
in
g
s
(
I
o
T
)
d
ev
ices
an
d
ar
t
if
icial
in
tellig
en
t
b
ased
ap
p
licatio
n
s
as
th
e
m
ajo
r
r
ea
s
o
n
b
e
h
in
d
t
h
e
co
m
p
lex
it
y
o
f
th
e
tr
a
f
f
ic
b
e
h
av
io
r
.
T
h
is
co
m
p
lex
ity
is
ch
ar
ac
ter
iz
ed
with
lo
n
g
-
ter
m
in
ter
d
ep
en
d
en
cy
a
n
d
n
o
n
lin
e
ar
d
y
n
a
m
ics
m
ak
in
g
t
h
e
co
n
v
en
tio
n
al
tec
h
n
iq
u
es
to
f
ail
in
p
r
ed
ictin
g
th
e
n
etwo
r
k
p
atter
n
.
As
a
r
esu
lt
o
f
th
is
,
n
etwo
r
k
p
atter
n
p
r
ed
ictio
n
p
r
o
b
lem
h
as
tilt
ed
to
war
d
s
d
ep
lo
y
m
en
t
o
f
r
o
b
u
s
t
alg
o
r
ith
m
s
lik
e
m
ac
h
in
e
lear
n
in
g
,
d
ee
p
lear
n
in
g
,
a
r
tific
ial
n
eu
r
al
n
etwo
r
k
s
(
ANN
s)
,
an
d
o
th
er
m
ac
h
in
e
lear
n
in
g
to
o
ls
f
o
r
ef
f
ec
tiv
e
r
e
p
r
esen
tatio
n
o
f
f
u
tu
r
e
n
etwo
r
k
tr
af
f
ic
tr
en
d
s
[
7
]
,
u
s
in
g
f
ac
eb
o
o
k
as
a
ca
s
e
s
tu
d
y
illu
s
tr
ated
th
at
ex
ten
s
iv
e
t
r
af
f
ic
an
aly
s
is
is
cr
u
cial
in
d
e
s
ig
n
in
g
a
s
ca
lab
le
d
ec
is
io
n
s
y
s
tem
.
Usi
n
g
th
e
Face
b
o
o
k
f
o
r
ec
asti
n
g
f
r
am
e
wo
r
k
lik
e
Pro
p
h
et,
n
etwo
r
k
m
ain
ten
an
ce
team
ca
n
p
r
e
d
i
ct
ch
an
g
es
i
n
u
s
er
en
g
ag
em
e
n
t a
n
d
tr
a
f
f
ic
d
e
m
an
d
th
er
eb
y
av
er
tin
g
s
er
v
ice
d
eli
v
er
y
b
o
ttlen
ec
k
s
.
T
h
e
s
u
r
g
e
o
f
lin
k
e
d
d
ev
ices;
th
e
escalatin
g
n
ee
d
f
o
r
r
ea
l
–
t
im
e
d
ig
ital
s
er
v
ices
ex
ac
er
b
at
ed
b
y
th
e
ex
p
o
n
e
n
tial
in
cr
ea
s
e
in
in
ter
n
et
tr
af
f
ic
h
av
e
in
cr
ea
s
ed
t
h
e
n
ee
d
f
o
r
b
an
d
wid
th
f
o
r
ec
asti
n
g
to
o
l
b
y
n
etwo
r
k
ad
m
in
is
tr
ato
r
s
.
B
o
x
et
a
l
.
[
8
]
s
tr
ess
ed
th
at
t
im
e
s
er
ies
f
o
r
ec
asti
n
g
em
p
h
asized
th
e
r
ein
f
o
r
ce
s
th
e
p
r
ed
ictio
n
o
f
r
eso
u
r
ce
s
u
tili
za
tio
n
,
h
en
ce
aim
in
g
at
p
r
ev
en
tin
g
n
etwo
r
k
c
o
n
g
esti
o
n
s
.
Fo
r
ec
asti
n
g
f
r
am
e
wo
r
k
lik
e
Pro
p
h
et
,
au
to
r
eg
r
ess
iv
e
in
te
g
r
ated
m
o
v
in
g
av
er
a
g
e
(
AR
I
MA
)
,
ANN
ca
n
ass
is
t
in
s
titu
tio
n
s
lik
e
Un
i
v
er
s
ities
,
Po
ly
tech
n
ics
to
m
an
ag
e
th
eir
n
etwo
r
k
v
ia
ef
f
icien
t
allo
ca
tio
n
o
f
b
a
n
d
wid
th
.
B
ajab
er
et
a
l.
[
9
]
elu
cid
ate
d
th
at
b
an
d
wid
th
r
eq
u
ir
em
en
ts
in
t
elec
o
m
m
u
n
icatio
n
n
etwo
r
k
s
ar
e
n
o
t
ju
s
t
s
ea
s
o
n
al
b
u
t
also
m
ar
k
ed
ly
b
u
s
ty
,
in
f
lu
en
ce
d
b
y
p
h
en
o
m
en
a
s
u
ch
as
v
ir
al
co
n
ten
t
p
r
o
p
ag
ati
o
n
o
r
ab
r
u
p
t
s
u
r
g
es
in
v
id
eo
s
tr
ea
m
in
g
d
em
an
d
.
Pre
cise
b
an
d
wid
th
p
r
ed
ictio
n
h
elp
s
tr
af
f
ic
o
p
er
ato
r
s
to
ex
ec
u
te
co
r
r
ec
t
tr
af
f
ic
en
g
i
n
e
er
in
g
m
ea
s
u
r
es
lik
e
d
y
n
am
ic
r
o
u
tin
g
to
u
p
h
o
ld
s
e
r
v
ice
q
u
ality
[
1
0
]
.
I
n
ab
ilit
y
t
o
ef
f
ec
tiv
ely
m
an
a
g
e
tr
af
f
ic
p
atter
n
s
v
ia
p
r
ec
is
e
b
an
d
wid
th
f
o
r
ec
asti
n
g
m
ay
le
ad
to
cu
s
to
m
er
d
is
co
n
ten
t,
b
r
e
ac
h
es
o
f
s
er
v
ice
lev
el
ag
r
ee
m
en
t
(
SLA)
r
esu
ltin
g
in
lo
s
s
o
f
in
co
m
e.
Fu
r
th
er
m
o
r
e
,
Hu
a
et
a
l
.
[
1
0
]
,
Gh
ee
wala
et
a
l
.
[
1
1
]
,
h
ig
h
lig
h
ted
th
at
d
ee
p
lear
n
in
g
-
en
a
b
led
ap
p
licatio
n
s
,
in
clu
d
in
g
v
id
e
o
co
n
f
er
en
cin
g
,
cl
o
u
d
g
am
i
n
g
,
an
d
ar
tific
ial
i
n
tellig
en
ce
(
AI
)
wo
r
k
lo
ad
s
,
ar
e
p
ar
ticu
lar
ly
s
u
s
ce
p
tib
le
to
v
ar
iatio
n
s
in
laten
cy
an
d
b
an
d
wi
d
th
d
u
e
to
th
e
p
r
o
life
r
atio
n
o
f
clo
u
d
co
m
p
u
tin
g
an
d
d
ata
-
d
r
iv
en
s
er
v
ices
[
1
1
]
.
W
ith
f
o
r
ec
asti
n
g
,
p
r
o
ac
tiv
e
a
d
ju
s
tm
en
t
o
f
n
etwo
r
k
r
eso
u
r
c
es
is
m
an
d
ato
r
y
f
o
r
ef
f
ec
tiv
e
m
a
n
ag
em
e
n
t
o
f
s
u
c
h
wo
r
k
lo
ad
s
.
His
to
r
ically
,
b
a
n
d
wid
th
f
o
r
ec
asti
n
g
h
as
d
ep
e
n
d
ed
o
n
s
tatis
tical
m
o
d
els,
m
ac
h
in
e
lear
n
in
g
,
a
n
d
,
m
o
r
e
r
ec
e
n
tly
,
d
ee
p
lear
n
in
g
m
eth
o
d
o
l
o
g
ies.
E
ac
h
te
ch
n
iq
u
e
p
o
s
s
ess
es
d
is
tin
ct
ad
v
an
tag
es
an
d
d
r
a
wb
ac
k
s
th
at
af
f
ec
t
th
eir
ap
p
r
o
p
r
iaten
ess
f
o
r
p
r
ac
tical
n
et
wo
r
k
m
an
ag
e
m
en
t.
C
o
n
v
en
tio
n
al
s
tatis
tical
m
eth
o
d
s
lik
e
AR
I
MA
an
d
s
ea
s
o
n
al
AR
I
MA
(
SA
R
I
MA
)
co
n
tin
u
e
to
b
e
u
s
ed
o
win
g
to
its
m
ath
em
atica
l
s
im
p
licity
an
d
r
o
b
u
s
t
b
asis
in
tim
e
s
er
ie
s
a
n
aly
s
is
.
Ji
[
1
2
]
elu
cid
ated
AR
I
MA
’
s
ef
f
icac
y
in
m
o
d
elin
g
s
h
o
r
t
-
ter
m
tem
p
o
r
al
r
elatio
n
s
h
ip
s
,
r
en
d
er
i
n
g
it
ap
p
r
o
p
r
iate
f
o
r
s
tead
y
tr
af
f
ic
co
n
d
itio
n
s
ch
ar
ac
ter
ized
b
y
c
o
n
s
tan
t
s
ea
s
o
n
al
o
r
d
aily
d
em
an
d
p
atter
n
s
.
No
n
eth
eless
,
AR
I
MA
h
as
d
if
f
icu
lties
in
h
ig
h
l
y
n
o
n
lin
ea
r
o
r
v
o
latile
en
v
ir
o
n
m
en
ts
,
s
u
ch
as
a
b
r
u
p
t
in
c
r
ea
s
es
in
v
id
e
o
s
tr
ea
m
in
g
o
r
u
n
f
o
r
eseen
u
s
er
s
u
r
g
es,
wh
ich
f
r
eq
u
en
tly
ar
is
e
in
c
o
n
t
em
p
o
r
ar
y
n
etwo
r
k
s
[
1
3
]
-
[
1
5
]
.
Ma
ch
in
e
lear
n
in
g
m
eth
o
d
o
lo
g
ies
h
av
e
b
ee
n
em
p
lo
y
ed
t
o
ad
d
r
ess
th
ese
d
ef
icien
cies.
C
h
en
et
a
l
.
[
1
6
]
u
tili
ze
d
ANNs
to
p
r
ed
ict
ac
tu
al
b
r
o
ad
b
a
n
d
tr
af
f
ic
f
r
o
m
C
h
in
a
T
elec
o
m
.
T
h
eir
m
o
d
e
l
attain
ed
s
u
p
er
io
r
ac
cu
r
ac
y
c
o
m
p
ar
e
d
to
AR
I
MA
,
esp
ec
ially
in
id
e
n
tify
in
g
n
o
n
lin
ea
r
c
o
r
r
elatio
n
s
b
etwe
en
h
is
to
r
ical
an
d
f
u
tu
r
e
tr
af
f
ic.
T
h
e
ANN
ad
ep
tly
ad
j
u
s
ted
to
ab
r
u
p
t
f
lu
ctu
atio
n
s
in
cu
s
to
m
er
d
em
an
d
,
but
it
n
ec
ess
itated
s
u
b
s
tan
tial
tr
ain
in
g
d
ata
an
d
m
eticu
lo
u
s
p
ar
am
eter
s
elec
tio
n
,
wh
ich
m
ay
b
e
u
n
f
ea
s
ib
le
f
o
r
s
m
aller
f
ir
m
s
with
co
n
s
tr
ain
ed
r
eso
u
r
ce
s
.
Su
p
p
o
r
t
v
ec
to
r
m
a
ch
in
es
(
SVMs)
h
av
e
b
ee
n
e
v
alu
ated
f
o
r
in
ter
n
et
tr
af
f
ic
p
r
e
d
ictio
n
.
C
h
en
et
a
l
.
[
1
7
]
r
e
v
ea
led
th
at
SVM
ex
ce
l
led
in
f
o
r
ec
asti
n
g
b
u
r
s
ty
an
d
ir
r
eg
u
lar
tr
af
f
ic
p
atter
n
s
,
in
w
h
ich
co
n
v
en
tio
n
al
lin
ea
r
m
o
d
els
f
r
e
q
u
en
tly
f
alter
.
T
h
eir
r
esear
ch
s
h
o
wn
th
at
S
VM
s
m
ay
g
en
er
alize
ac
r
o
s
s
d
i
v
er
s
e
d
ata
co
n
tex
ts
,
y
ield
in
g
p
r
ec
is
e
p
r
ed
ictio
n
s
ev
en
in
h
ig
h
l
y
d
y
n
a
m
ic
s
ettin
g
s
.
T
h
e
co
n
s
tr
ain
t,
h
o
we
v
er
,
r
esid
es
in
th
e
co
m
p
u
tatio
n
al
c
o
m
p
lex
ity
o
f
SVMs;
tr
ain
in
g
g
ets
p
r
o
g
r
ess
iv
ely
ex
p
e
n
s
iv
e
with
ex
ten
s
iv
e
d
atasets
,
p
o
s
in
g
a
ch
allen
g
e
i
n
in
ter
n
et
s
er
v
ice
p
r
o
v
id
er
(
I
SP
)
-
lev
el
o
p
er
atio
n
s
.
Hy
b
r
id
m
eth
o
d
o
lo
g
ies
h
av
e
ar
is
en
to
ca
p
italize
o
n
th
e
s
y
n
er
g
is
tic
ad
v
an
ta
g
es
o
f
m
an
y
tech
n
i
q
u
es.
Ad
e
k
itan
et
a
l
.
[
1
8
]
in
tr
o
d
u
ce
d
a
h
y
b
r
id
AR
I
MA
-
n
eu
r
a
l
n
etwo
r
k
m
o
d
el
f
o
r
p
r
ed
ictin
g
in
ter
n
et
tr
af
f
ic
in
Nig
er
ian
u
n
iv
er
s
ity
n
etwo
r
k
s
.
T
h
eir
f
i
n
d
in
g
s
in
d
icate
d
th
at
th
e
h
y
b
r
i
d
m
o
d
el
s
u
r
p
ass
ed
th
e
s
tan
d
alo
n
e
AR
I
MA
b
y
d
i
m
in
is
h
in
g
f
o
r
ec
asti
n
g
e
r
r
o
r
s
d
u
r
in
g
p
ea
k
ac
tiv
ity
in
ter
v
als,
s
u
ch
as e
x
am
in
atio
n
r
eg
is
tr
atio
n
.
T
h
i
s
m
e
t
h
o
d
’
s
d
r
a
w
b
a
c
k
w
as
t
h
e
a
d
d
e
d
c
o
m
p
l
e
x
i
t
y
o
f
i
n
t
e
g
r
a
t
i
n
g
t
w
o
m
o
d
e
l
i
n
g
m
e
t
h
o
d
o
l
o
g
i
e
s
,
n
e
c
e
s
s
i
t
at
i
n
g
b
o
t
h
s
t
a
ti
s
t
i
c
al
co
m
p
e
t
e
n
c
e
a
n
d
m
a
c
h
i
n
e
l
e
a
r
n
in
g
p
r
o
f
i
c
i
e
n
c
y
.
L
o
n
g
s
h
o
r
t
-
t
e
r
m
m
e
m
o
r
y
(
L
S
T
M
)
n
e
t
w
o
r
k
s
,
a
d
e
e
p
n
e
u
r
a
l
n
e
t
wo
r
k
(
D
N
N
)
p
o
s
s
e
s
s
es
e
n
o
r
m
o
u
s
p
o
t
e
n
t
i
a
l
i
n
t
r
a
f
f
i
c
p
r
e
d
i
ct
io
n
.
W
a
n
e
t
a
l
.
[
1
9
]
e
m
p
l
o
y
e
d
L
S
T
Ms
a
s
a
n
et
wo
r
k
t
r
a
f
f
i
c
f
o
r
e
c
a
s
ti
n
g
f
r
a
m
ew
o
r
k
f
o
r
I
S
P
i
n
C
h
i
n
a
.
T
h
e
r
e
s
u
l
t
o
b
t
ai
n
e
d
i
n
c
o
m
p
a
r
i
s
o
n
w
i
t
h
AR
I
MA
a
n
d
s
h
a
l
l
o
w
n
e
u
r
a
l
n
e
tw
o
r
k
s
w
as
o
u
t
s
t
a
n
d
i
n
g
.
T
h
e
a
b
i
l
it
y
o
f
L
S
T
M
t
o
e
f
f
e
c
ti
v
e
l
y
i
d
e
n
t
i
f
y
l
o
n
g
-
t
e
r
m
d
e
p
e
n
d
e
n
c
i
es
i
n
n
e
t
w
o
r
k
t
r
a
f
f
i
c
is
t
h
e
b
e
d
r
o
c
k
o
f
t
h
i
s
f
r
a
m
e
w
o
r
k
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
N
etw
o
r
k
t
r
a
ffic a
n
a
lysi
s
a
n
d
b
a
n
d
w
id
th
fo
r
ec
a
s
tin
g
f
o
r
u
s
in
g
Meta
’
s
P
r
o
p
h
et:
…
(
Yu
s
u
f
On
imis
i I
s
a
a
c
)
753
Ho
wev
er
,
th
e
ap
p
r
o
ac
h
is
co
m
p
u
tatio
n
ally
in
te
n
s
iv
e,
r
eq
u
ir
in
g
h
u
g
e
lab
el
d
atasets
th
er
ef
o
r
e
lim
itin
g
its
s
u
itab
ilit
y
to
lar
g
e
f
ir
m
s
w
ith
en
o
u
g
h
r
eso
u
r
ce
s
.
I
n
s
u
m
m
ar
y
,
alth
o
u
g
h
AR
I
MA
an
d
an
alo
g
o
u
s
s
tatis
tical
tech
n
iq
u
es
p
r
o
v
id
e
ea
s
e
o
f
u
s
e,
th
ey
ar
e
in
ad
eq
u
ate
in
n
o
n
lin
ea
r
an
d
er
r
atic
s
ce
n
ar
io
s
.
Ma
ch
in
e
lear
n
in
g
tech
n
iq
u
es
s
u
ch
as
ANN
a
n
d
SVM
en
h
an
ce
ac
cu
r
ac
y
b
u
t
n
ec
ess
itate
m
eticu
lo
u
s
tw
ea
k
in
g
an
d
m
o
r
e
p
r
o
ce
s
s
in
g
r
eso
u
r
ce
s
.
So
,
h
y
b
r
id
m
eth
o
d
s
ac
h
iev
e
eq
u
ili
b
r
iu
m
b
u
t
in
tr
o
d
u
ce
c
o
m
p
le
x
ity
,
wh
er
ea
s
d
ee
p
lear
n
in
g
tech
n
iq
u
es
lik
e
L
STM
d
eliv
er
s
u
p
e
r
io
r
p
er
f
o
r
m
an
ce
at
th
e
co
s
t
o
f
s
ca
lab
ilit
y
an
d
r
eso
u
r
ce
d
em
a
n
d
s
.
T
h
ese
co
n
s
tr
ain
ts
d
r
iv
e
th
e
p
u
r
s
u
it
o
f
alter
n
ativ
e
f
o
r
ec
asti
n
g
f
r
am
ewo
r
k
s
th
at
ar
e
p
r
ec
is
e,
co
m
p
r
eh
en
s
ib
le,
an
d
r
eso
u
r
ce
-
ef
f
icien
t
,
f
ac
ilit
atin
g
th
e
d
e
v
elo
p
m
e
n
t
o
f
to
o
ls
lik
e
Me
ta
’
s
Pro
p
h
et
[
2
0
]
.
Nu
m
er
o
u
s
ap
p
lied
s
tu
d
ies
h
av
e
ass
ess
ed
Pro
p
h
et
in
ac
tu
al
n
etwo
r
k
o
r
r
elate
d
s
itu
atio
n
s
,
ea
ch
illu
s
tr
atin
g
p
r
ac
tical
ad
v
an
tag
es
wh
ile
also
ex
p
o
s
in
g
lim
itatio
n
s
in
b
an
d
wid
th
p
r
ed
ictio
n
s
.
Owu
s
u
-
K
u
m
ih
et
a
l
.
[
2
1
]
ex
ec
u
ted
a
tar
g
ete
d
ca
s
e
s
tu
d
y
to
esti
m
ate
th
e
n
atio
n
al
in
ter
n
et
co
n
s
u
m
p
tio
n
an
d
r
ev
e
n
u
e
tr
e
n
d
s
in
Gh
a
n
a,
u
tili
zin
g
Pro
p
h
et
to
an
aly
ze
citizen
s
’
n
etwo
r
k
tr
af
f
ic
a
g
g
r
e
g
ated
o
n
a
d
aily
b
asis
.
T
h
e
s
tu
d
y
s
y
s
tem
atica
lly
co
n
tr
asted
Pro
p
h
et
p
r
o
jectio
n
s
with
AR
I
M
A
b
aselin
es
ac
r
o
s
s
s
tan
d
ar
d
n
atio
n
al
s
ea
s
o
n
ality
,
in
clu
d
in
g
d
aily
u
s
ag
e
p
atter
n
s
,
wee
k
d
ay
/wee
k
e
n
d
v
ar
iatio
n
s
,
an
d
y
ea
r
ly
ev
e
n
ts
.
T
h
e
s
tu
d
y
in
d
icate
d
th
at
Pro
p
h
et
m
o
r
e
p
r
ec
is
ely
id
en
tifie
d
r
ep
ea
tin
g
wee
k
ly
an
d
y
e
ar
ly
-
r
elate
d
cy
cles,
y
ield
in
g
n
ar
r
o
wer
p
r
e
d
ictio
n
i
n
ter
v
als
d
u
r
i
n
g
r
eg
u
lar
in
ter
v
a
ls
an
d
im
p
r
o
v
ed
co
r
r
elatio
n
w
ith
o
b
s
er
v
e
d
p
ea
k
s
ass
o
ciate
d
with
y
ea
r
ly
ev
en
ts
an
d
f
esti
v
ity
p
er
io
d
s
.
T
h
e
p
r
im
ar
y
d
r
awb
ac
k
s
ee
n
was
Pro
p
h
et
’
s
s
u
b
s
tan
tial
u
n
d
er
esti
m
ate
o
f
tr
an
s
ien
t
s
p
i
k
es,
s
u
ch
as
u
n
e
x
p
ec
ted
ev
en
t
-
d
r
iv
en
s
u
r
g
es,
as
th
e
m
o
d
el
p
r
io
r
itizes
r
ec
u
r
r
en
t
s
ea
s
o
n
ality
an
d
tr
en
d
s
u
n
less
ex
p
licit e
x
ter
n
al
r
e
g
r
ess
o
r
s
f
o
r
p
ar
ticu
lar
ev
e
n
ts
ar
e
p
r
o
v
id
ed
.
R
es
e
a
r
c
h
e
r
s
[2
2
]
,
[2
3
]
i
n
c
o
r
p
o
r
a
t
e
d
P
r
o
p
h
e
t
i
n
t
o
a
h
y
b
r
i
d
c
l
o
u
d
r
e
s
o
u
r
c
e
m
o
n
i
t
o
r
i
n
g
p
l
a
t
f
o
r
m
t
o
p
r
e
d
i
c
t
b
a
n
d
w
i
d
t
h
u
t
i
li
z
a
ti
o
n
f
o
r
o
r
c
h
e
s
t
r
a
t
i
o
n
a
n
d
a
u
t
o
s
c
al
in
g
.
T
h
e
i
r
d
e
p
l
o
y
m
e
n
t
-
f
o
c
u
s
e
d
r
e
s
e
a
r
c
h
i
n
t
e
g
r
at
e
d
P
r
o
p
h
e
t
f
o
r
e
c
a
s
t
s
w
it
h
s
y
s
t
e
m
-
l
e
v
e
l
g
u
i
d
e
li
n
e
s
f
o
r
s
c
al
i
n
g
d
e
ci
s
i
o
n
s
,
d
e
m
o
n
s
t
r
at
i
n
g
t
h
at
P
r
o
p
h
e
t
-
d
e
r
i
v
e
d
s
i
g
n
al
s
m
i
n
i
m
i
z
e
d
s
u
p
e
r
f
l
u
o
u
s
s
ca
l
e
-
u
p
s
a
n
d
d
e
c
r
e
a
s
e
d
c
o
s
ts
r
e
l
at
i
v
e
t
o
s
i
m
p
li
s
ti
c
t
h
r
es
h
o
l
d
i
n
g
.
T
h
e
au
th
o
r
s
in
d
icate
d
th
at
Pro
p
h
et
’
s
ad
d
itiv
e
s
tr
u
ctu
r
e
o
cc
asio
n
ally
f
alter
ed
d
u
r
in
g
s
u
d
d
e
n
wo
r
k
lo
a
d
tr
an
s
itio
n
s
b
etwe
en
o
n
-
p
r
em
is
es
an
d
clo
u
d
n
o
d
es;
th
ey
a
d
d
r
ess
ed
th
is
b
y
in
co
r
p
o
r
atin
g
ab
r
u
p
t
m
ig
r
atio
n
f
la
g
s
as
ex
ter
n
al
r
eg
r
ess
o
r
s
,
r
esu
ltin
g
in
h
eig
h
ten
e
d
o
p
er
atio
n
al
co
m
p
lex
ity
.
T
h
ese
ca
s
e
s
tu
d
ies
co
llectiv
ely
af
f
ir
m
Pro
p
h
et
’
s
p
r
ac
tical
ad
v
an
tag
e
s
:
in
ter
p
r
etab
ilit
y
,
m
an
ag
em
e
n
t
o
f
m
u
lti
-
s
ea
s
o
n
ality
,
r
esil
ie
n
ce
to
m
is
s
in
g
d
ata,
an
d
m
in
i
m
al
co
m
p
u
tatio
n
al
d
em
an
d
s
.
C
o
m
m
o
n
d
r
awb
ac
k
s
in
clu
d
e
ch
allen
g
es
i
n
m
an
ag
i
n
g
u
n
iq
u
e
,
u
n
p
r
ec
e
d
en
ted
s
p
ik
es
with
o
u
t
ex
ter
n
al
r
e
g
r
ess
o
r
s
,
s
en
s
itiv
it
y
to
ch
a
n
g
ep
o
in
t
an
d
s
ea
s
o
n
ality
co
n
f
ig
u
r
atio
n
s
,
an
d
d
ep
e
n
d
en
ce
o
n
p
r
ec
is
e
ev
en
t la
b
elin
g
f
o
r
aty
p
ical
in
ter
v
als.
T
h
ese
tr
ad
e
-
o
f
f
s
d
ictate
th
e
co
n
f
ig
u
r
atio
n
an
d
en
h
an
ce
m
e
n
t
o
f
Pro
p
h
et
with
i
n
an
o
p
er
atio
n
al
b
an
d
wid
th
-
f
o
r
ec
asti
n
g
p
ip
elin
e.
E
lem
ile
et
a
l
.
[2
4
]
f
o
cu
s
ed
o
n
v
er
if
y
in
g
n
o
is
e
lev
els
with
in
Om
u
-
Ar
an
t
o
wn
s
h
ip
b
y
em
p
l
o
y
in
g
an
ANN.
No
is
e
d
ata
was
g
ath
er
ed
f
r
o
m
2
1
ch
o
s
en
s
ites
in
th
e
m
o
r
n
in
g
,
m
id
d
ay
,
a
n
d
e
v
en
in
g
,
d
u
r
in
g
b
o
th
s
ch
o
o
l
s
ess
io
n
s
an
d
v
ac
a
tio
n
tim
es,
o
v
er
a
3
-
wee
k
p
er
io
d
u
s
in
g
a
SL4
0
1
0
s
o
u
n
d
lev
el
m
eter
.
T
h
e
ANN
m
o
d
el
was
s
tr
u
ctu
r
ed
to
u
tili
ze
1
9
n
o
is
e
-
r
elate
d
elem
en
ts
as
in
p
u
ts
;
th
ese
wer
e
co
llected
th
r
o
u
g
h
o
n
-
s
ite
o
b
s
er
v
atio
n
a
n
d
d
o
cu
m
e
n
tatio
n
.
A
to
tal
o
f
1
1
3
4
d
at
a
p
o
in
ts
wer
e
in
p
u
t
in
t
o
th
e
m
o
d
el.
T
h
r
o
u
g
h
a
r
an
d
o
m
s
elec
tio
n
p
r
o
ce
s
s
,
7
0
%
o
f
th
e
d
ata
was
u
s
ed
to
tr
ain
th
e
n
etwo
r
k
,
with
t
h
e
r
em
ain
i
n
g
3
0
%
allo
ca
ted
f
o
r
test
in
g
.
T
h
e
im
p
lem
en
tatio
n
o
f
alg
o
r
ith
m
s
an
d
th
e
d
eter
m
in
atio
n
o
f
t
h
e
ep
o
ch
v
a
lu
e
wer
e
g
u
id
ed
b
y
th
e
s
elec
tio
n
o
f
a
n
etwo
r
k
with
two
h
id
d
en
lay
er
s
.
A
k
an
d
e
et
a
l
.
[2
5
]
in
tr
o
d
u
ce
d
a
n
o
v
e
l
ap
p
r
o
ac
h
to
im
p
r
o
v
e
h
o
w
we
s
p
o
t
n
etwo
r
k
in
tr
u
s
io
n
s
.
I
t
u
s
es
a
co
m
b
in
ed
tech
n
iq
u
e
th
at
b
len
d
s
th
e
s
tr
en
g
th
s
o
f
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
s
(
C
NNs)
an
d
DNNs.
T
h
e
g
o
a
l
is
t
o
b
u
il
d
a
n
i
n
t
r
u
s
i
o
n
d
e
t
e
ct
i
o
n
s
y
s
te
m
(
I
D
S
)
t
h
at
c
a
n
a
n
a
l
y
z
e
n
e
t
w
o
r
k
t
r
a
f
f
i
c
a
n
d
a
c
c
u
r
at
e
l
y
l
a
b
e
l
i
t
a
s
ei
t
h
e
r
n
o
r
m
a
l
o
r
h
a
r
m
f
u
l
,
e
f
f
e
c
t
i
v
e
l
y
i
d
e
n
t
i
f
y
i
n
g
s
e
c
u
r
i
t
y
b
r
e
a
c
h
e
s
.
T
h
e
C
N
N
-
b
a
s
e
d
a
p
p
r
o
a
c
h
d
e
m
o
n
s
t
r
a
t
e
d
s
i
g
n
i
f
i
c
a
n
tl
y
b
e
t
t
e
r
a
c
c
u
r
a
c
y
c
o
m
p
a
r
e
d
t
o
o
t
h
e
r
m
e
t
h
o
d
s
,
r
e
a
c
h
i
n
g
a
n
i
m
p
r
e
s
s
i
v
e
9
9
.
1
8
%
.
W
e
a
s
s
e
s
s
e
d
t
h
e
e
f
f
ec
t
i
v
e
n
e
s
s
o
f
t
h
i
s
I
D
S
u
s
i
n
g
s
ta
n
d
a
r
d
m
e
t
r
i
cs
,
i
n
c
l
u
d
in
g
a
c
c
u
r
a
c
y
,
p
r
e
c
i
s
i
o
n
,
r
e
c
al
l
,
F
1
-
s
c
o
r
e
,
a
n
d
t
h
e
f
a
ls
e
p
o
s
i
t
i
v
e
r
a
t
e
(
FP
R
)
.
2.
M
E
T
H
O
D
T
h
is
s
ec
tio
n
o
u
tlin
es
th
e
p
r
o
ce
d
u
r
es,
to
o
ls
,
an
d
c
o
n
f
ig
u
r
atio
n
s
u
s
ed
f
o
r
f
o
r
ec
asti
n
g
n
etwo
r
k
b
an
d
wid
th
d
e
m
an
d
with
in
L
an
d
m
ar
k
Un
i
v
er
s
ity
h
alls
o
f
r
esid
en
ce
.
I
t
is
d
iv
id
ed
in
to
f
iv
e
k
ey
ar
ea
s
:
in
tr
o
d
u
ctio
n
,
r
esear
ch
d
esig
n
,
h
ar
d
war
e
an
d
s
o
f
twar
e
r
eq
u
ir
em
en
ts
,
d
ata
co
llectio
n
,
an
d
d
ata
p
r
e
p
r
o
ce
s
s
in
g
.
T
h
is
s
tu
d
y
ap
p
lies
d
ata
-
d
r
iv
en
tech
n
iq
u
es
to
esti
m
ate
an
d
m
an
ag
e
b
a
n
d
wid
th
in
th
e
s
tu
d
en
t
h
o
s
tels
.
I
t
f
o
cu
s
es
o
n
a
n
aly
zin
g
h
is
to
r
ical
tr
a
f
f
ic
p
atter
n
s
u
s
in
g
q
u
an
titativ
e
to
o
ls
lik
e
Paess
ler
R
o
u
ter
T
r
af
f
i
c
Gr
ap
h
e
r
(
PR
T
G
)
f
o
r
m
o
n
ito
r
in
g
an
d
Me
ta
’
s
Pro
p
h
et
f
o
r
f
o
r
ec
asti
n
g
.
T
h
e
s
e
to
o
ls
h
elp
r
ev
ea
l
u
s
ag
e
p
atter
n
s
an
d
s
u
p
p
o
r
t
in
f
o
r
m
e
d
n
etwo
r
k
o
p
tim
izatio
n
.
T
o
b
u
il
d
a
r
eliab
le
f
o
r
ec
asti
n
g
m
o
d
el,
t
h
e
s
tu
d
y
in
v
o
lv
ed
o
n
g
o
in
g
m
o
n
ito
r
in
g
o
f
in
ter
n
et
ac
tiv
ity
o
v
er
tim
e.
T
h
e
ap
p
r
o
ac
h
aim
s
to
cr
ea
te
a
s
ca
lab
le
f
r
am
ewo
r
k
f
o
r
u
n
d
e
r
s
tan
d
in
g
b
an
d
wid
th
u
s
ag
e
tr
en
d
s
ac
r
o
s
s
th
e
u
n
iv
er
s
ity
.
T
h
is
s
tu
d
y
f
o
cu
s
ed
o
n
u
n
d
e
r
s
tan
d
in
g
an
d
q
u
a
n
tify
in
g
b
an
d
w
id
th
n
o
t
as
a
f
ix
ed
lim
it,
b
u
t
a
s
th
e
r
ea
l
-
tim
e
am
o
u
n
t
o
f
d
ata
m
o
v
in
g
ac
r
o
s
s
a
n
etwo
r
k
,
o
r
th
r
o
u
g
h
p
u
t.
T
o
p
r
ed
ict
f
u
tu
r
e
d
ata
u
s
ag
e,
th
e
a
u
th
o
r
s
em
p
lo
y
ed
Me
ta
’
s
Pro
p
h
et
m
o
d
el,
tr
ain
in
g
it
o
n
p
ast
tr
af
f
ic
d
ata
to
f
o
r
ec
ast
b
a
n
d
wid
th
u
s
ag
e,
r
ep
r
esen
te
d
as
(
)
.
W
e
ev
alu
ated
th
e
m
o
d
el
’
s
p
er
f
o
r
m
an
ce
b
ased
o
n
ac
tu
al
d
ata
f
lo
w,
u
s
in
g
m
etr
ics
lik
e
m
ea
n
ab
s
o
lu
te
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
TEL
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
3
,
J
u
n
e
20
26
:
7
5
1
-
7
6
4
754
er
r
o
r
(
MA
E
)
an
d
r
o
o
t
m
ea
n
s
q
u
ar
e
er
r
o
r
(
R
MSE
)
,
m
ea
s
u
r
e
d
in
b
its
p
er
s
ec
o
n
d
(
b
p
s
)
.
T
h
e
p
r
ed
ictio
n
e
r
r
o
r
s
wer
e
co
n
s
id
er
ed
ac
ce
p
tab
le
w
h
en
co
m
p
ar
ed
to
th
e
o
b
s
er
v
e
d
p
ea
k
b
an
d
wid
th
u
s
ag
e,
wh
ic
h
v
ar
ied
b
etwe
en
4
7
an
d
5
0
m
eg
a
b
its
p
er
s
ec
o
n
d
(
Mb
p
s
)
.
Data
was
co
llected
u
s
in
g
PR
T
G
m
o
n
ito
r
in
g
s
o
f
tw
ar
e,
wh
ich
tr
ac
k
ed
m
etr
ics
s
u
ch
as
b
an
d
wid
th
u
s
ag
e
alo
n
g
with
n
etwo
r
k
laten
cy
an
d
u
p
tim
e.
I
n
ess
en
ce
,
th
is
ap
p
r
o
ac
h
em
p
h
asized
th
e
f
l
u
ctu
atin
g
v
o
lu
m
e
o
f
d
ata
tr
an
s
f
er
r
ed
ac
r
o
s
s
th
e
n
etwo
r
k
o
v
er
a
p
er
io
d
o
f
tim
e,
r
ef
lectin
g
th
r
o
u
g
h
p
u
t
an
d
u
tili
za
tio
n
.
T
h
is
s
tu
d
y
p
r
ed
ictio
n
f
o
cu
s
ed
o
n
h
o
w
well
th
e
m
o
d
els
p
er
f
o
r
m
an
d
h
o
w
th
e
p
r
ed
ictio
n
s
wer
e
ch
ec
k
e
d
f
o
r
ac
cu
r
ac
y
,
b
u
t
d
o
es
n
o
t
g
o
in
to
g
r
ea
t
d
ep
th
a
b
o
u
t
f
i
n
e
-
t
u
n
in
g
th
e
m
o
d
el
’
s
s
ettin
g
s
.
T
h
e
r
esear
ch
u
s
ed
c
o
m
m
o
n
m
etr
ics
f
o
r
tim
e
-
b
ased
d
ata
to
m
ea
s
u
r
e
h
o
w
well
th
e
Pro
p
h
et
m
o
d
el
p
r
ed
icted
tr
af
f
ic
co
m
p
a
r
ed
t
o
r
ea
l
-
wo
r
ld
d
ata.
T
h
e
p
r
ed
ictio
n
er
r
o
r
s
wer
e
c
o
n
s
id
er
e
d
r
ea
s
o
n
ab
le
wh
e
n
co
m
p
ar
ed
to
t
h
e
p
ea
k
o
b
s
er
v
e
d
b
an
d
wid
th
u
s
ag
e
o
f
ap
p
r
o
x
im
ately
4
7
–
5
0
Mb
p
s
.
I
n
s
tead
o
f
a
s
tan
d
ar
d
cr
o
s
s
-
v
alid
atio
n
ap
p
r
o
ac
h
,
th
e
s
tu
d
y
u
s
ed
a
s
in
g
le
s
p
lit
o
f
t
h
e
d
ata
.
T
h
e
d
ata
was
d
iv
id
e
d
in
to
8
0
%
f
o
r
t
r
ain
in
g
th
e
m
o
d
el
an
d
2
0
%
f
o
r
test
in
g
its
p
r
ed
ictio
n
s
.
T
h
is
s
p
lit
was
d
esig
n
ed
to
m
im
ic
r
ea
l
-
wo
r
ld
p
r
ed
ictio
n
s
ce
n
a
r
io
s
b
y
p
r
eser
v
i
n
g
th
e
o
r
d
er
o
f
th
e
d
ata
o
v
er
tim
e.
Mo
r
eo
v
e
r
,
t
h
e
tr
ain
in
g
an
d
v
alid
atio
n
p
r
o
ce
s
s
wer
e
r
ep
ea
ted
m
u
ltip
le
tim
es.
W
h
ile
th
e
p
r
e
cise
s
tep
s
f
o
r
ad
ju
s
tin
g
th
e
m
o
d
el
’
s
s
ettin
g
s
ar
e
n
o
t
f
u
lly
d
escr
ib
ed
,
th
e
s
tu
d
y
r
ec
o
g
n
ized
th
e
im
p
o
r
ta
n
ce
o
f
th
ese
s
ettin
g
s
alth
o
u
g
h
.
T
h
e
Pro
p
h
et
m
o
d
el
was
n
o
ted
to
b
e
s
en
s
itiv
e
to
its
co
n
f
ig
u
r
atio
n
,
p
ar
ticu
lar
ly
co
n
ce
r
n
in
g
h
o
w
it
id
en
tifie
d
p
o
in
ts
wh
er
e
th
e
d
ata
ch
an
g
ed
an
d
h
o
w
it
h
a
n
d
led
r
ep
ea
tin
g
p
atter
n
s
(
s
ea
s
o
n
ality
)
.
Po
o
r
ch
o
ices
in
th
ese
s
ettin
g
s
ca
n
lead
to
p
r
o
b
lem
s
lik
e
th
e
m
o
d
el
b
ein
g
to
o
clo
s
ely
tailo
r
ed
to
th
e
tr
ai
n
in
g
d
ata
(
o
v
er
-
f
itti
n
g
)
o
r
s
m
o
o
th
in
g
o
u
t
im
p
o
r
tan
t
d
etails.
T
h
e
s
y
s
tem
also
in
clu
d
ed
d
aily
a
n
d
wee
k
ly
r
e
p
ea
tin
g
p
atter
n
s
,
wh
ich
wer
e
ty
p
ically
ad
ju
s
ted
th
r
o
u
g
h
t
h
e
Pro
p
h
et
m
o
d
el
’
s
co
n
f
ig
u
r
atio
n
p
ar
a
m
eter
s
.
Fig
u
r
e
1
d
ep
icts
th
e
f
lo
wch
ar
t
o
f
th
e
s
y
s
tem
p
r
o
ce
s
s
es
o
f
th
e
s
tu
d
y
f
r
o
m
th
e
b
eg
in
n
in
g
to
th
e
en
d
.
Fig
u
r
e
1
.
Flo
wch
ar
t
o
f
th
e
s
y
s
tem
p
r
o
ce
d
u
r
es
2
.
1
.
Resea
rc
h
des
ig
n
T
h
is
s
tu
d
y
ad
o
p
ts
a
ca
s
e
s
tu
d
y
ap
p
r
o
ac
h
ce
n
ter
ed
o
n
L
an
d
m
ar
k
Un
iv
er
s
ity
’
s
h
o
s
tel
n
etwo
r
k
s
,
u
s
in
g
a
q
u
an
titativ
e
m
eth
o
d
o
l
o
g
y
.
Netwo
r
k
p
er
f
o
r
m
a
n
ce
was
ev
alu
ated
u
n
d
er
b
o
th
ty
p
i
ca
l
an
d
h
ig
h
-
u
s
ag
e
co
n
d
itio
n
s
u
s
in
g
h
is
to
r
ical
d
at
a.
Key
v
ar
iab
les s
u
ch
as tr
af
f
ic
s
p
ee
d
an
d
v
o
lu
m
e
(
in
b
o
u
n
d
an
d
o
u
t
b
o
u
n
d
)
wer
e
co
llected
v
ia
PR
T
G.
Pro
p
h
et,
an
o
p
en
-
s
o
u
r
ce
f
o
r
ec
asti
n
g
m
o
d
el,
was
em
p
lo
y
ed
to
a
n
aly
z
e
tim
e
-
s
er
ies
tr
en
d
s
an
d
p
r
ed
ict
f
u
tu
r
e
b
a
n
d
wid
th
d
em
an
d
s
.
T
h
e
d
esig
n
is
iter
ativ
e,
en
ab
lin
g
ad
j
u
s
tm
en
ts
b
ased
o
n
f
o
r
ec
asti
n
g
ac
cu
r
ac
y
an
d
d
ata
q
u
ality
.
T
h
is
ad
ap
tab
ilit
y
im
p
r
o
v
es
lo
n
g
-
ter
m
m
o
d
el
r
elia
b
ilit
y
an
d
s
u
p
p
o
r
ts
ex
p
a
n
s
io
n
t
o
o
th
er
ar
ea
s
with
in
th
e
ca
m
p
u
s
.
2
.
2
.
H
a
rdwa
re
a
nd
s
o
f
t
wa
re
re
qu
irem
ent
s
T
h
is
s
ec
tio
n
o
u
tlin
es
th
e
es
s
e
n
tial
h
ar
d
war
e
an
d
s
o
f
twar
e
c
o
m
p
o
n
en
ts
u
s
ed
to
co
llect,
p
r
o
ce
s
s
,
an
d
an
aly
ze
n
etwo
r
k
tr
a
f
f
ic
d
ata.
T
h
e
s
elec
ted
to
o
ls
an
d
d
ev
i
ce
s
en
s
u
r
e
s
y
s
tem
s
tab
ilit
y
,
d
ata
ac
cu
r
ac
y
,
an
d
co
m
p
atib
ilit
y
with
f
o
r
ec
asti
n
g
alg
o
r
ith
m
s
r
eq
u
ir
ed
f
o
r
b
a
n
d
wid
th
p
r
ed
ictio
n
.
T
h
e
s
o
f
twar
e
to
o
ls
u
s
ed
in
th
is
p
r
o
ject
en
a
b
le
th
e
m
o
n
ito
r
in
g
,
p
r
ep
r
o
ce
s
s
in
g
,
a
n
aly
s
is
,
an
d
f
o
r
ec
asti
n
g
o
f
n
etwo
r
k
d
ata.
E
ac
h
was
ch
o
s
en
f
o
r
its
s
ca
lab
ilit
y
,
ea
s
e
o
f
in
teg
r
atio
n
,
an
d
s
u
itab
ilit
y
f
o
r
h
an
d
lin
g
tim
e
-
s
er
ies
tr
af
f
ic
d
ata
with
in
a
ca
m
p
u
s
en
v
ir
o
n
m
en
t.
−
PR
T
G
n
etwo
r
k
m
o
n
ito
r
:
ca
p
t
u
r
es
r
ea
l
-
tim
e/h
is
to
r
ical
tr
af
f
ic
d
ata
an
d
v
is
u
aliza
tio
n
s
.
Fig
u
r
e
2
d
ep
icts
a
v
is
u
aliza
tio
n
o
f
t
h
e
PR
T
G
I
n
ter
f
ac
e
o
f
v
ar
io
u
s
h
alls
o
f
r
esid
en
ce
f
o
r
th
e
ca
s
e
s
tu
d
y
,
n
a
m
ely
;
Ab
ig
ail
h
all,
Sar
ah
h
all,
Deb
o
r
ah
h
all,
I
s
aa
c
h
all,
J
o
s
ep
h
h
all,
Dan
iel
h
all
,
an
d
Do
r
ca
s
h
all.
−
Me
ta
Pro
p
h
et:
tim
e
-
s
er
ies f
o
r
e
ca
s
tin
g
to
o
l w
ith
s
u
p
p
o
r
t f
o
r
tr
en
d
s
an
d
s
ea
s
o
n
ality
.
−
Py
th
o
n
: f
o
r
d
ata
a
n
aly
s
is
,
u
s
in
g
Pan
d
as,
Nu
m
Py
,
Scik
it
-
lear
n
,
Ma
tp
lo
tlib
,
an
d
Seab
o
r
n
.
−
J
u
p
y
ter
No
teb
o
o
k
: in
ter
ac
tiv
e
en
v
ir
o
n
m
en
t f
o
r
co
d
in
g
a
n
d
m
o
d
el
v
is
u
aliza
tio
n
.
−
E
x
ce
l: u
s
ed
in
ea
r
ly
-
s
tag
e
d
ata
in
s
p
ec
tio
n
an
d
f
o
r
m
at
co
n
v
er
s
io
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
N
etw
o
r
k
t
r
a
ffic a
n
a
lysi
s
a
n
d
b
a
n
d
w
id
th
fo
r
ec
a
s
tin
g
f
o
r
u
s
in
g
Meta
’
s
P
r
o
p
h
et:
…
(
Yu
s
u
f
On
imis
i I
s
a
a
c
)
755
Fig
u
r
e
2
.
A
v
is
u
aliza
tio
n
o
f
th
e
PR
T
G
i
n
ter
f
ac
e
T
h
e
h
ar
d
war
e
s
etu
p
f
o
r
m
s
th
e
f
o
u
n
d
atio
n
f
o
r
c
o
llectin
g
an
d
m
an
ag
in
g
n
etwo
r
k
tr
af
f
ic
d
at
a.
Dev
ices
wer
e
s
elec
ted
f
o
r
th
eir
ab
ili
ty
to
s
u
p
p
o
r
t
co
n
tin
u
o
u
s
m
o
n
ito
r
in
g
,
r
eliab
le
d
ata
tr
an
s
m
is
s
io
n
,
an
d
h
ig
h
p
r
o
ce
s
s
in
g
ca
p
ac
ity
,
w
h
ich
ar
e
cr
itical
f
o
r
ac
cu
r
ate
b
an
d
wid
th
an
aly
s
is
an
d
f
o
r
ec
asti
n
g
.
−
R
o
u
ter
s
(
d
u
al
-
b
an
d
(
2
.
4
GHz
/
5
GHz
)
,
1
Gb
p
s
)
:
t
h
e
r
o
u
ter
s
s
er
v
e
as
th
e
p
r
im
ar
y
d
ata
c
o
llectio
n
p
o
in
ts
,
ca
p
tu
r
in
g
r
ea
l
-
tim
e
tr
af
f
ic
in
f
o
r
m
atio
n
s
u
ch
as
b
an
d
wid
t
h
u
s
ag
e,
p
ac
k
et
tr
an
s
f
er
,
an
d
laten
cy
.
T
h
e
F
ig
u
r
e
3
s
h
o
ws th
e
r
o
u
ter
u
s
ed
to
ca
p
tu
r
e
r
ea
l
-
tim
e
tr
a
f
f
ic
i
n
f
o
r
m
atio
n
f
o
r
th
is
s
tu
d
y
.
−
Switch
es:
Sw
itch
es
ar
e
u
s
ed
f
o
r
ag
g
r
eg
atin
g
d
ata
tr
a
f
f
ic
f
r
o
m
m
u
ltip
le
d
ev
ices,
en
a
b
lin
g
s
ea
m
less
m
o
n
ito
r
in
g
o
f
n
etwo
r
k
p
er
f
o
r
m
an
ce
in
th
e
h
alls
o
f
r
esid
en
c
e.
C
is
co
s
witch
th
at
was
u
s
ed
f
o
r
th
is
s
tu
d
y
is
s
h
o
wn
in
F
ig
u
r
e
4
.
−
Ser
v
er
s
:
th
e
u
n
iv
er
s
ity
s
er
v
er
u
s
ed
f
o
r
th
is
s
tu
d
y
is
s
h
o
wn
in
F
ig
u
r
e
5
.
I
ts
s
p
ec
if
icatio
n
s
ar
e
I
n
tel
Xeo
n
Go
ld
5
2
1
8
,
6
4
GB
R
AM
,
4
T
B
HDD,
1
T
B
SS
D.
−
Oth
er
p
er
i
p
h
er
als:
u
n
in
ter
r
u
p
tib
le
p
o
wer
s
u
p
p
l
y
(
UPS
)
u
n
its
,
E
th
er
n
et
ca
b
les,
ex
te
r
n
al
s
to
r
ag
e
f
o
r
b
ac
k
u
p
.
−
Op
er
atin
g
s
y
s
tem
s
: Wi
n
d
o
ws Ser
v
er
2
0
1
9
an
d
2
0
2
2
s
u
p
p
o
r
ted
th
e
an
aly
tical
to
o
ls
.
Fig
u
r
e
3
.
A
r
o
u
ter
Fig
u
r
e
4
.
A
C
is
co
s
witch
Fig
u
r
e
5
.
Ser
v
e
r
2
.
3
.
Da
t
a
c
o
llect
io
n
T
h
is
s
ec
tio
n
d
etails
h
o
w
h
is
to
r
ical
n
etwo
r
k
t
r
af
f
ic
d
ata
w
as
g
ath
er
ed
f
r
o
m
L
a
n
d
m
ar
k
Un
iv
er
s
ity
h
o
s
tels
.
I
t d
escr
ib
es th
e
s
etu
p
o
f
m
o
n
ito
r
i
n
g
d
ev
ices a
n
d
th
e
u
s
e
o
f
PR
T
G
to
lo
g
m
etr
ics
li
k
e
b
an
d
wid
th
u
s
ag
e
an
d
laten
cy
,
p
r
o
v
id
in
g
th
e
r
aw
in
p
u
t n
ee
d
ed
f
o
r
ac
c
u
r
ate
an
a
ly
s
is
an
d
f
o
r
ec
asti
n
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
TEL
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
3
,
J
u
n
e
20
26
:
7
5
1
-
7
6
4
756
−
T
r
af
f
ic
m
o
n
ito
r
i
n
g
s
etu
p
:
s
im
p
le
n
etwo
r
k
m
an
ag
em
e
n
t p
r
o
t
o
co
l
(
SNMP
)
a
p
p
licatio
n
lay
e
r
p
r
o
to
co
l
u
s
ed
f
o
r
m
an
ag
i
n
g
an
d
m
o
n
ito
r
in
g
n
etwo
r
k
d
ev
ices
,
en
ab
le
d
d
ev
ices
in
eig
h
t
h
alls
o
f
r
esi
d
en
ce
(
Dan
iel
,
Ab
r
ah
am
,
J
o
s
ep
h
,
Do
r
ca
s
,
I
s
a
ac
,
Sar
ah
,
an
d
Ab
ig
ail)
r
ec
o
r
d
ed
tr
af
f
ic
m
etr
ics
s
u
ch
as
b
a
n
d
wid
th
u
s
ag
e,
laten
cy
,
an
d
u
p
tim
e.
E
ac
h
d
ata
p
o
in
t w
as tim
e
s
tam
p
ed
an
d
h
all
-
s
p
ec
if
ic
f
o
r
lo
ca
tio
n
-
awa
r
e
an
aly
s
is
.
−
PR
T
G
d
ata
co
llectio
n
:
b
etwe
en
Octo
b
er
1
a
n
d
Dec
em
b
er
1
5
,
2
0
2
4
,
d
ata
was
co
llected
h
o
u
r
ly
u
s
in
g
PR
T
G.
T
h
e
to
o
l
p
r
o
v
id
e
d
ex
p
o
r
ts
in
v
ar
io
u
s
f
o
r
m
ats
(
C
SV,
PDF,
an
d
XM
L
)
an
d
av
er
a
g
in
g
in
ter
v
als.
I
t
s
u
p
p
o
r
ted
s
elec
tiv
e
ex
p
o
r
t
o
f
s
p
ec
if
ic
tr
af
f
ic
ty
p
es
(
in
b
o
u
n
d
,
o
u
tb
o
u
n
d
,
a
n
d
to
tal)
.
Dev
ice
s
lo
g
g
ed
o
v
e
r
9
9
.
5
% u
p
tim
e,
en
s
u
r
in
g
d
ata
r
eliab
ilit
y
.
2
.
4
.
Da
t
a
p
re
pro
ce
s
s
ing
B
ef
o
r
e
an
aly
s
is
an
d
f
o
r
ec
asti
n
g
,
r
aw
d
ata
m
u
s
t
b
e
clea
n
e
d
an
d
tr
an
s
f
o
r
m
ed
in
to
a
u
s
a
b
le
f
o
r
m
at.
T
h
is
s
ec
tio
n
d
escr
ib
es
h
o
w
th
e
d
ataset
was
co
n
v
er
te
d
,
f
ilter
ed
,
an
d
s
tr
u
ctu
r
ed
to
eli
m
in
ate
er
r
o
r
s
an
d
in
co
n
s
is
ten
cies,
en
s
u
r
in
g
th
e
m
o
d
el
is
tr
ain
ed
with
h
ig
h
-
q
u
ality
in
p
u
ts
.
B
ef
o
r
e
we
co
u
ld
u
s
e
Me
ta
’
s
Pro
p
h
et
to
o
l
to
an
aly
ze
n
etwo
r
k
tr
af
f
ic
an
d
p
r
ed
ict
b
an
d
wid
th
n
ee
d
s
,
th
er
e
is
n
ee
d
to
p
u
t
t
o
g
eth
er
a
th
o
r
o
u
g
h
p
r
ep
ar
atio
n
p
r
o
ce
s
s
to
m
a
k
e
s
u
r
e
th
e
d
ata
was
g
o
o
d
q
u
ality
,
co
n
s
is
ten
t
o
v
e
r
tim
e,
a
n
d
wo
u
ld
wo
r
k
well
with
th
e
to
o
l.
First,
th
e
r
aw
n
etwo
r
k
tr
af
f
ic
d
ata
g
ath
e
r
ed
f
r
o
m
P
R
T
G
was
co
m
b
in
ed
in
to
co
n
s
is
ten
t
tim
e
ch
u
n
k
s
,
lik
e
ev
er
y
h
o
u
r
.
T
h
en
,
f
o
r
m
att
ed
it
in
a
way
th
at
Pro
p
h
et
c
o
u
ld
u
n
d
er
s
tan
d
,
u
s
in
g
d
ate
a
n
d
tim
e
lab
els.
Of
ten
,
s
o
m
e
d
ata
was
m
is
s
in
g
d
u
e
to
eq
u
ip
m
e
n
t
p
r
o
b
lem
s
o
r
lo
g
g
in
g
f
ailu
r
es.
T
o
f
ix
th
is
,
a
f
e
w
d
if
f
er
en
t
m
eth
o
d
s
wer
e
ap
p
lied
.
Fo
r
s
h
o
r
t
g
a
p
s
,
a
s
im
p
le
f
ill
-
in
-
t
h
e
-
b
lan
k
s
a
p
p
r
o
ac
h
was
em
p
lo
y
ed
.
Fo
r
l
o
n
g
er
g
ap
s
,
a
m
o
r
e
ad
v
an
ce
d
tec
h
n
iq
u
e
t
h
at
lo
o
k
e
d
at
s
ea
s
o
n
al
p
atter
n
s
in
th
e
d
a
ta
(
lik
e
d
aily
o
r
wee
k
ly
tr
e
n
d
s
)
to
m
ak
e
ed
u
ca
ted
g
u
ess
es
was
ad
o
p
ted
.
Als
o
f
illed
f
o
r
war
d
m
is
s
in
g
d
ata
in
s
tab
le
s
ec
tio
n
s
o
f
th
e
tr
af
f
ic
d
ata.
Ou
tlier
s
wer
e
d
ea
lt
with
,
wh
ich
ar
e
u
n
u
s
u
al
d
ata
p
o
in
ts
o
f
ten
ca
u
s
ed
b
y
attac
k
s
,
m
is
tak
es
in
co
n
f
ig
u
r
atio
n
,
o
r
lar
g
e
f
ile
tr
an
s
f
er
s
.
T
h
ese
o
u
tlier
s
wer
e
tr
ea
ted
u
s
in
g
s
tatis
tical
r
u
les
an
d
a
m
eth
o
d
ca
lled
t
h
e
in
ter
q
u
ar
tile
r
an
g
e
(
I
QR
)
.
T
h
en
,
r
ep
lace
d
th
em
with
s
m
o
o
th
ed
-
o
u
t
v
alu
es,
ch
ec
k
in
g
s
y
s
tem
lo
g
s
to
m
ak
e
s
u
r
e
n
o
ac
cid
en
tally
r
em
o
v
a
l
o
f
r
ea
l
ev
e
n
ts
.
T
h
is
h
elp
ed
Pr
o
p
h
et
w
o
r
k
b
etter
an
d
allo
we
d
th
e
au
th
o
r
s
to
ea
s
ily
co
m
p
a
r
e
d
if
f
er
en
t
n
etwo
r
k
co
n
n
ec
tio
n
s
.
A
f
ter
m
ak
i
n
g
p
r
e
d
ictio
n
s
,
we
co
n
v
er
ted
th
e
s
ca
led
v
alu
es
b
ac
k
to
th
eir
o
r
ig
in
al
u
n
its
,
lik
e
Mb
p
s
.
Nex
t,
th
e
au
th
o
r
s
m
ad
e
s
u
r
e
th
e
tim
e
s
er
ies
d
ata
h
ad
p
er
f
ec
tl
y
r
eg
u
la
r
tim
e
in
ter
v
als,
r
em
o
v
in
g
an
y
d
u
p
licates
an
d
co
m
b
in
i
n
g
an
y
o
v
er
lap
p
in
g
d
ata.
T
h
is
en
s
u
r
es
th
at
t
h
e
d
ata
m
ee
ts
Pro
p
h
et
’
s
r
eq
u
ir
em
en
t
o
f
ev
e
n
ly
s
p
ac
ed
tim
estam
p
s
.
Fin
ally
,
th
e
d
ata
s
p
lit
in
to
two
s
e
ts
:
a
tr
ain
in
g
s
et
(
8
0
%)
to
teac
h
Pro
p
h
et
an
d
a
test
in
g
s
et
(
2
0
%)
to
e
v
alu
ate
its
p
r
e
d
ictio
n
s
.
T
h
e
a
u
th
o
r
s
m
a
d
e
s
u
r
e
n
o
t
to
s
h
u
f
f
le
t
h
e
d
ata
r
an
d
o
m
ly
,
s
o
we
c
o
u
l
d
s
im
u
late
r
ea
l
-
wo
r
ld
f
o
r
ec
asti
n
g
s
ce
n
ar
io
s
.
2
.
4
.
1
.
Da
t
a
c
o
nv
er
s
io
n
PR
T
G
’
s
ex
p
o
r
ted
PDFs
wer
e
co
n
v
er
ted
to
C
SV
u
s
in
g
iL
o
v
ePDF
.
T
h
is
co
n
v
er
s
io
n
en
s
u
r
e
d
th
at
th
e
d
ata
b
ec
am
e
s
tr
u
ctu
r
e
d
an
d
m
ac
h
in
e
-
r
ea
d
a
b
le,
m
ak
i
n
g
it
ea
s
ier
to
wo
r
k
with
in
an
aly
tical
en
v
ir
o
n
m
en
ts
.
T
h
e
C
SV
f
o
r
m
at
also
allo
wed
f
o
r
s
ea
m
less
im
p
o
r
t
in
to
J
u
p
y
ter
No
teb
o
o
k
an
d
en
s
u
r
ed
c
o
m
p
a
tib
ilit
y
with
v
ar
io
u
s
Py
th
o
n
lib
r
a
r
ies u
s
ed
f
o
r
d
ata
p
r
o
ce
s
s
in
g
an
d
an
aly
s
is
.
2
.
4
.
2
.
Da
t
a
clea
nin
g
a
nd
a
g
g
re
g
a
t
io
n
Py
th
o
n
s
cr
ip
ts
wer
e
u
s
ed
to
f
ilter
,
clea
n
,
an
d
f
o
r
m
at
th
e
d
ata.
I
n
v
alid
e
n
tr
ies,
d
u
p
licates,
an
d
m
is
s
in
g
tim
estam
p
s
wer
e
ad
d
r
ess
ed
u
s
in
g
Pan
d
as.
T
im
e
-
b
ased
a
g
g
r
eg
atio
n
h
elp
ed
s
m
o
o
th
o
u
t
n
o
is
e
an
d
r
ev
ea
l
u
n
d
er
ly
i
n
g
tr
en
d
s
.
Ad
d
itio
n
a
l
s
tep
s
in
clu
d
ed
co
n
v
er
tin
g
all
m
etr
ics
to
Mb
p
s
,
g
en
er
a
tin
g
en
co
d
e
d
tim
e
f
ea
tu
r
es
(
h
o
u
r
/d
a
y
/m
o
n
t
h
)
,
a
n
d
p
r
e
p
ar
in
g
tr
ain
in
g
an
d
test
in
g
d
atasets
.
Hig
h
-
q
u
ality
p
r
e
p
r
o
ce
s
s
in
g
en
s
u
r
ed
th
at
th
e
m
o
d
el
r
ec
eiv
ed
s
tr
u
ct
u
r
ed
in
p
u
t,
wh
ich
d
ir
ec
tly
im
p
r
o
v
ed
f
o
r
ec
asti
n
g
ac
c
u
r
ac
y
.
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
is
s
ec
tio
n
p
r
esen
ts
th
e
r
esu
lts
o
b
tain
ed
f
r
o
m
th
e
d
ata
co
l
lecte
d
,
th
e
m
o
d
elin
g
p
r
o
ce
s
s
,
ev
alu
atio
n
m
etr
ics,
an
d
a
d
is
cu
s
s
io
n
o
f
f
in
d
in
g
s
.
T
o
d
ev
el
o
p
an
ef
f
ec
ti
v
e
m
o
d
el
f
o
r
f
o
r
ec
asti
n
g
th
e
r
eq
u
ir
ed
b
an
d
wid
th
in
th
e
h
alls
o
f
r
esid
en
ce
,
s
e
v
er
al
ac
tiv
ities
n
ee
d
ed
to
b
e
co
n
d
u
cted
o
v
er
a
p
er
io
d
,
in
clu
d
in
g
co
n
tin
u
o
u
s
m
o
n
ito
r
in
g
o
f
in
te
r
n
et
c
o
n
n
ec
tiv
ity
.
T
h
is
a
p
p
r
o
ac
h
s
er
v
es
a
s
a
tem
p
late
f
o
r
f
u
r
th
er
s
tu
d
ies
o
n
i
n
ter
n
et
u
s
ag
e
p
atter
n
s
ac
r
o
s
s
th
e
en
tire
s
ch
o
o
l.
T
h
e
Face
b
o
o
k
Pro
p
h
et
m
o
d
el
(
)
,
is
an
a
d
d
itiv
e
r
eg
r
ess
io
n
m
o
d
el
d
ef
in
ed
by
(
1
)
:
(
)
=
(
)
+
ℎ
(
)
+
(
)
+
(
1
)
W
h
er
e:
−
(
)
:
r
ep
r
esen
ts
th
e
tr
en
d
co
m
p
o
n
e
n
t,
ca
p
tu
r
in
g
lo
n
g
-
ter
m
i
n
cr
ea
s
e
o
r
d
ec
r
ea
s
e
in
b
an
d
wid
th
u
s
ag
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
N
etw
o
r
k
t
r
a
ffic a
n
a
lysi
s
a
n
d
b
a
n
d
w
id
th
fo
r
ec
a
s
tin
g
f
o
r
u
s
in
g
Meta
’
s
P
r
o
p
h
et:
…
(
Yu
s
u
f
On
imis
i I
s
a
a
c
)
757
−
ℎ
(
)
:
r
ep
r
esen
ts
h
o
lid
ay
ef
f
ec
ts
,
ac
co
u
n
tin
g
f
o
r
s
p
ec
ial
d
ates o
r
e
v
en
ts
th
at
in
f
lu
en
ce
tr
af
f
ic.
−
(
)
:
r
ep
r
esen
ts
s
ea
s
o
n
ality
,
m
o
d
el
in
g
r
ec
u
r
r
in
g
p
atter
n
s
s
u
ch
as
d
aily
,
wee
k
ly
,
o
r
m
o
n
th
ly
c
y
cl
es.
−
ₜ
:
i
s
th
e
er
r
o
r
ter
m
,
ca
p
tu
r
in
g
u
n
p
r
ed
ictab
le
f
lu
ctu
atio
n
s
o
r
n
o
is
e
in
th
e
d
ata.
−
(
)
:
Face
b
o
o
k
Pro
p
h
et
m
o
d
el
p
r
e
d
icted
v
alu
e
−
(
)
)
̂
: Fac
eb
o
o
k
Pro
p
h
et
m
ea
n
p
r
e
d
icted
v
alu
e
−
:
d
u
r
atio
n
o
f
tim
e
o
v
er
wh
ic
h
p
r
ed
ictio
n
is
m
ad
e
T
h
e
u
s
e
o
f
Pro
p
h
et
allo
ws
t
h
e
s
e
co
m
p
o
n
en
ts
to
b
e
f
le
x
ib
ly
m
o
d
eled
a
n
d
co
m
b
i
n
ed
t
o
p
r
o
d
u
ce
h
i
g
h
-
q
u
ality
f
o
r
ec
asts
,
ev
en
in
t
h
e
p
r
esen
ce
o
f
ir
r
e
g
u
lar
ities
o
r
m
is
s
in
g
d
ata.
T
o
ass
ess
th
e
ac
cu
r
ac
y
an
d
r
eliab
ilit
y
o
f
th
e
p
r
ed
ictio
n
s
,
th
r
ee
s
tan
d
ar
d
er
r
o
r
m
etr
ics
wer
e
ca
lcu
la
ted
:
MA
E
,
m
ea
n
s
q
u
ar
e
er
r
o
r
(
MSE
)
,
an
d
R
MSE
.
T
h
ese
m
etr
ics
p
r
o
v
id
e
in
s
ig
h
t
in
to
th
e
m
o
d
el
’
s
p
r
e
d
ictiv
e
ca
p
ab
ilit
y
an
d
h
o
w
clo
s
ely
th
e
f
o
r
ec
asts
alig
n
with
ac
tu
al
o
b
s
er
v
ed
v
alu
es.
a.
MA
E
=
(
1
/
)
∗
∑
|
(
(
)
−
(
)
)
̂
|
(
2
)
=
10099863
.
10
b
p
s
;
t
h
is
m
ea
s
u
r
es th
e
av
er
a
g
e
m
ag
n
itu
d
e
o
f
e
r
r
o
r
s
in
a
s
et
o
f
p
r
e
d
ictio
n
s
.
b.
MSE
=
(
1
/
)
∗
∑
(
(
)
−
(
)
)
̂
2
(
3
)
=
184
,
170
,
943
,
484
,
006
.
78
b
p
s
;
t
h
is
in
d
icate
s
/m
ea
s
u
r
es
l
ar
g
er
er
r
o
r
s
m
o
r
e
th
an
MA
E
,
wh
ich
g
iv
es
a
b
etter
s
en
s
e
o
f
th
e
s
p
r
ea
d
.
c.
R
MSE
=
√
(
1
/
)
∗
∑
(
(
)
−
(
)
)
̂
2
(
4
)
=
√
=
13
,
570
,
959
.
58
b
p
s
;
T
h
is
r
e
f
lects
th
e
av
er
a
g
e
s
q
u
ar
e
d
if
f
er
en
ce
b
etwe
en
o
b
s
er
v
e
d
an
d
p
r
ed
icted
v
al
u
es.
T
h
e
m
o
d
el
e
v
alu
atio
n
r
esu
lts
d
em
o
n
s
tr
ate
th
at
th
e
Pro
p
h
et
f
o
r
ec
asti
n
g
m
o
d
el
p
er
f
o
r
m
s
well
u
n
d
er
th
e
cu
r
r
en
t
d
ata
co
n
d
itio
n
s
.
T
h
e
MA
E
an
d
R
MSE
v
alu
es,
wh
en
co
m
p
a
r
ed
to
th
e
e
x
p
ec
t
ed
b
an
d
wid
t
h
r
an
g
es,
co
n
f
ir
m
th
at
th
e
p
r
ed
ictio
n
s
ar
e
s
u
f
f
icien
tly
ac
cu
r
ate
f
o
r
p
r
ac
tical
u
s
e
in
n
etwo
r
k
p
la
n
n
in
g
an
d
ca
p
ac
ity
f
o
r
ec
asti
n
g
.
T
h
e
lar
g
e
MSE
v
a
lu
e,
wh
ile
m
ath
em
atica
lly
co
r
r
ec
t,
is
an
ex
p
ec
ted
o
u
tco
m
e
g
i
v
en
th
e
m
ag
n
itu
d
e
o
f
th
e
d
ata
a
n
d
th
e
s
q
u
ar
ed
n
a
tu
r
e
o
f
th
e
m
etr
ic.
T
h
e
c
o
n
s
is
ten
cy
b
etwe
en
th
e
MA
E
a
n
d
R
MSE
v
alu
es.
T
h
ese
r
esu
lts
in
d
icate
ac
ce
p
tab
le
er
r
o
r
m
ar
g
i
n
s
co
n
s
id
er
in
g
th
e
o
b
s
er
v
ed
p
ea
k
u
s
ag
e
r
an
g
ed
b
etwe
en
4
7
Mb
p
s
an
d
5
0
Mb
p
s
.
T
h
e
Pro
p
h
et
m
o
d
el
p
r
o
v
id
e
d
r
ich
v
is
u
aliza
tio
n
s
in
clu
d
in
g
:
(
i
)
f
o
r
e
c
a
s
t
p
l
o
t
s
h
o
w
i
n
g
p
r
e
d
i
c
t
e
d
b
a
n
d
w
i
d
t
h
u
s
a
g
e
w
it
h
u
p
p
e
r
an
d
l
o
w
e
r
c
o
n
f
i
d
e
n
c
e
i
n
t
e
r
v
a
ls
;
(
i
i
)
t
r
e
n
d
c
o
m
p
o
n
e
n
t
h
i
g
h
li
g
h
t
in
g
g
r
a
d
u
a
l
i
n
c
r
e
a
s
e
o
r
d
e
c
r
e
a
s
e
i
n
u
s
a
g
e
o
v
e
r
ti
m
e
;
a
n
d
(
i
i
i
)
s
e
a
s
o
n
al
i
t
y
p
l
o
t
s
h
o
w
i
n
g
d
a
i
l
y
u
s
a
g
e
p
e
a
k
s
(
t
y
p
ic
a
l
l
y
b
e
t
we
e
n
7
P
M
an
d
1
1
PM)
a
n
d
ea
r
ly
m
o
r
n
in
g
d
ip
s
.
Ad
d
itio
n
ally
,
s
p
ik
es
wer
e
o
b
s
er
v
e
d
d
u
r
in
g
s
tu
d
en
t
r
eg
is
tr
atio
n
wee
k
s
,
s
u
g
g
esti
n
g
b
an
d
wid
th
s
tr
ess
d
u
r
in
g
th
o
s
e
p
er
io
d
s
.
T
h
e
lin
ea
r
g
r
a
p
h
ab
o
v
e
v
is
u
ali
ze
s
th
e
b
an
d
wid
th
u
s
ag
e
o
v
er
a
s
er
ies
o
f
tim
e
in
ter
v
als
f
r
o
m
th
e
1
st
o
f
Octo
b
er
2
0
2
4
to
1
5
th
o
f
Dec
em
b
er
2
0
2
4
is
d
e
p
icted
in
F
i
g
u
r
e
6
.
T
h
e
X
-
a
x
is
r
ep
r
esen
t
s
th
e
tim
e
d
u
r
atio
n
s
h
o
win
g
wh
en
th
e
b
an
d
wid
t
h
m
ea
s
u
r
em
e
n
ts
wer
e
tak
en
.
T
h
is
Y
-
ax
is
r
ep
r
esen
ts
th
e
am
o
u
n
t
o
f
n
etwo
r
k
b
an
d
wid
th
u
s
ed
at
ea
ch
tim
e
p
o
in
t.
T
h
e
b
an
d
wid
th
r
an
g
e
s
f
r
o
m
n
ea
r
0
k
ilo
b
y
te
(
k
B
)
(
wh
en
th
er
e
was
d
o
wn
tim
e)
to
p
e
ak
s
ex
ce
ed
in
g
8
m
illi
o
n
k
B
.
T
o
e
n
h
an
ce
t
h
e
v
is
u
al
r
ep
r
esen
tatio
n
o
f
th
e
f
o
r
ec
asti
n
g
r
esu
lts
,
po
wer
b
u
s
in
ess
in
tellig
en
ce
(
BI
)
was
em
p
lo
y
ed
d
u
e
to
its
in
tu
itiv
e
d
ash
b
o
ar
d
in
g
ca
p
ab
i
liti
es
an
d
ef
f
ec
tiv
e
co
m
p
ar
ativ
e
tr
e
n
d
p
r
esen
tatio
n
.
T
h
e
Fig
u
r
e
7
is
a
b
o
x
p
lo
t
titl
ed
“
n
etwo
r
k
b
a
n
d
wid
th
u
s
ag
e
b
y
d
a
y
o
f
wee
k
,
”
s
h
o
win
g
th
e
d
is
tr
ib
u
tio
n
o
f
b
a
n
d
wid
th
u
s
ag
e
(
in
k
B
)
ac
r
o
s
s
ea
ch
d
ay
o
f
th
e
wee
k
f
o
r
a
s
p
ec
if
ic
d
at
aset
lab
eled
“
alp
h
a
s
em
ester
.
”
T
h
e
X
-
ax
is
r
ep
r
esen
ts
th
e
in
d
ep
en
d
en
t
v
a
r
iab
le,
s
h
o
win
g
th
e
d
ay
s
o
v
er
wh
ich
b
an
d
wid
th
u
s
ag
e
is
m
ea
s
u
r
ed
.
w
h
ile
th
e
Y
-
ax
is
r
ep
r
esen
ts
th
e
d
ep
en
d
en
t
v
ar
i
ab
le,
s
h
o
win
g
th
e
am
o
u
n
t
o
f
n
etwo
r
k
b
an
d
wid
t
h
u
s
ed
.
T
h
e
h
o
r
izo
n
tal
lin
e
in
s
id
e
th
e
b
o
x
r
ep
r
esen
ts
th
e
m
ed
ia
n
(
5
0
th
p
er
ce
n
tile)
o
f
th
e
b
an
d
wid
th
u
s
ag
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
TEL
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
3
,
J
u
n
e
20
26
:
7
5
1
-
7
6
4
758
Fig
u
r
e
6
.
B
an
d
wid
th
co
n
s
u
m
p
tio
n
o
v
er
tim
e
Fig
u
r
e
7
.
B
o
x
p
lo
t sh
o
win
g
n
e
two
r
k
b
an
d
wid
th
u
s
ag
e
b
y
d
ay
s
o
f
th
e
wee
k
T
h
is
r
esear
ch
in
v
esti
g
ated
th
e
ap
p
licatio
n
o
f
Me
ta’
s
Pro
p
h
et,
an
o
p
en
-
s
o
u
r
ce
f
o
r
ec
asti
n
g
to
o
l
d
ev
elo
p
e
d
b
y
Face
b
o
o
k
(
n
o
w
Me
ta)
,
to
m
o
d
el
a
n
d
p
r
ed
ict
h
o
u
r
ly
n
etwo
r
k
b
an
d
wid
th
u
s
ag
e
(
in
b
p
s
)
with
in
a
r
ea
l
-
wo
r
ld
in
s
titu
tio
n
al
n
etwo
r
k
.
T
h
e
f
in
d
in
g
s
,
b
ased
o
n
em
p
ir
ical
d
ata
co
llected
o
v
er
a
2
8
-
h
o
u
r
p
e
r
io
d
f
r
o
m
1
to
3
Dec
em
b
er
2
0
2
4
,
r
e
v
ea
l
s
ev
er
al
k
ey
in
s
ig
h
ts
r
e
g
ar
d
i
n
g
Pro
p
h
et’
s
p
er
f
o
r
m
a
n
ce
,
ac
cu
r
ac
y
,
an
d
s
u
itab
ilit
y
f
o
r
n
etwo
r
k
tr
af
f
ic
f
o
r
ec
asti
n
g
in
en
v
ir
o
n
m
en
ts
with
s
tr
o
n
g
d
iu
r
n
al
p
atter
n
s
.
a.
Pro
p
h
et
s
u
cc
ess
f
u
lly
ca
p
tu
r
e
d
d
iu
r
n
al
s
ea
s
o
n
ality
in
n
et
wo
r
k
tr
af
f
ic:
th
e
o
b
s
er
v
ed
n
etwo
r
k
tr
af
f
ic
ex
h
ib
ited
a
clea
r
d
aily
c
y
clica
l
p
atter
n
:
u
s
ag
e
was
lo
west
d
u
r
in
g
late
-
n
ig
h
t
a
n
d
ea
r
ly
-
m
o
r
n
in
g
h
o
u
r
s
(
e.
g
.
,
1
0
.
9
–
1
5
.
5
Mb
p
s
b
etwe
en
2
to
3
AM
)
an
d
p
ea
k
ed
d
u
r
in
g
ev
en
in
g
h
o
u
r
s
(
e.
g
.
,
4
1
.
6
Mb
p
s
at
9
PM)
.
Pro
p
h
et’
s
f
o
r
ec
asts
clo
s
ely
m
ir
r
o
r
ed
t
h
is
p
atter
n
,
d
em
o
n
s
tr
atin
g
its
ab
ilit
y
to
m
o
d
el
s
tr
o
n
g
in
tr
ad
ay
s
ea
s
o
n
ality
a
k
n
o
wn
s
tr
en
g
t
h
o
f
th
e
alg
o
r
ith
m
d
u
e
to
its
b
u
ilt
-
in
Fo
u
r
ier
s
er
ies
co
m
p
o
n
e
n
ts
f
o
r
p
er
io
d
ic
ef
f
ec
ts
.
T
h
is
co
n
f
ir
m
s
th
at
Pro
p
h
et
ca
n
e
f
f
ec
tiv
ely
lear
n
an
d
r
e
p
licate
r
o
u
tin
e
h
u
m
an
-
d
r
iv
en
n
etwo
r
k
b
eh
av
io
r
,
s
u
ch
as
in
cr
ea
s
ed
u
s
ag
e
d
u
r
in
g
ac
ad
em
ic
o
r
ad
m
in
is
tr
ativ
e
wo
r
k
in
g
h
o
u
r
s
an
d
r
ed
u
ce
d
ac
tiv
ity
o
v
er
n
ig
h
t
[2
6
]
.
b.
Fo
r
ec
ast
ac
cu
r
ac
y
v
ar
ie
d
ac
r
o
s
s
tim
e
o
f
d
ay
:
w
h
ile
th
e
o
v
e
r
all
tr
en
d
was
well
-
p
r
ed
icted
,
f
o
r
ec
ast
ac
cu
r
ac
y
was n
o
t u
n
if
o
r
m
.
−
Un
d
er
esti
m
atio
n
d
u
r
i
n
g
ea
r
ly
ev
en
in
g
d
ec
lin
e
(
Dec
1
,
1
0
PM
–
2
AM
)
:
Pro
p
h
et
co
n
s
is
ten
tly
u
n
d
e
r
-
p
r
ed
icted
tr
af
f
ic
as
u
s
ag
e
d
r
o
p
p
ed
f
r
o
m
4
2
.
1
Mb
p
s
to
1
9
.
1
Mb
p
s
.
Fo
r
ex
am
p
le:
a
t
1
1
PM
(
Dec
1
)
:
=
38
.
1
Mb
p
s
,
=
32
.
8
Mb
p
s
(
1
4
%
u
n
d
er
esti
m
atio
n
)
,
a
t
1
AM
(
Dec
2
)
:
=
25
.
5
Mb
p
s
,
an
d
=
18
.
7
Mb
p
s
(
2
7
%
u
n
d
er
esti
m
atio
n
)
.
T
h
is
s
u
g
g
ests
Pro
p
h
et’
s
tr
en
d
co
m
p
o
n
en
t
was
to
o
a
g
g
r
ess
iv
e
in
d
ec
lin
e,
p
o
s
s
ib
ly
d
u
e
t
o
li
m
ited
h
is
to
r
ical
d
ata
o
r
ab
r
u
p
t
b
eh
a
v
io
r
al
s
h
if
ts
n
o
t f
u
lly
ca
p
t
u
r
ed
d
u
r
in
g
m
o
d
el
tr
ai
n
in
g
.
−
Ov
er
esti
m
atio
n
d
u
r
in
g
d
ee
p
n
i
g
h
t
tr
o
u
g
h
(
Dec
3
,
3
AM
)
:
a
t
3
AM
o
n
Dec
em
b
er
3
,
ac
tu
al
t
r
af
f
ic
was
1
0
.
9
5
Mb
p
s
,
b
u
t
Pro
p
h
et
f
o
r
e
ca
s
t
1
4
.
5
2
Mb
p
s
a
3
3
%
o
v
er
e
s
tim
atio
n
.
T
h
is
m
ay
in
d
icate
t
h
e
m
o
d
el
r
ev
er
ted
to
war
d
a
s
ea
s
o
n
al
m
ea
n
r
ath
er
th
an
a
d
ap
tin
g
to
u
n
u
s
u
ally
lo
w
ac
tiv
ity
,
a
k
n
o
wn
lim
itatio
n
wh
en
an
o
m
alies o
cc
u
r
[
27
].
−
Hig
h
ac
cu
r
ac
y
d
u
r
in
g
s
tab
le
d
ay
tim
e
an
d
p
ea
k
e
v
en
in
g
h
o
u
r
s
(
7
AM
–
1
1
PM)
:
f
r
o
m
m
o
r
n
i
n
g
th
r
o
u
g
h
late
ev
en
in
g
o
n
Dec
em
b
er
2
,
f
o
r
ec
asts
alig
n
ed
cl
o
s
ely
with
ac
tu
als:
At
7
AM
:
=
+
0
.
6%
,
at
8
PM:
=
+
1
.
5%
,
at
1
0
PM:
=
−
0
.
6%
.
T
h
is
d
em
o
n
s
tr
ates
Pro
p
h
et’
s
r
o
b
u
s
tn
ess
in
s
tead
y
-
s
tate
co
n
d
itio
n
s
wh
er
e
h
is
to
r
ical
p
atter
n
s
ar
e
co
n
s
is
ten
t.
c.
Mo
d
el
ex
h
ib
ited
m
in
o
r
p
h
ase
lag
an
d
s
m
o
o
th
in
g
a
r
tifa
cts
:
a
s
lig
h
t
p
h
ase
lag
was
o
b
s
er
v
ed
:
Pro
p
h
et’
s
f
o
r
ec
ast
cu
r
v
e
ap
p
ea
r
ed
s
m
o
o
th
er
an
d
s
lig
h
tly
d
elay
ed
co
m
p
ar
ed
to
s
h
ar
p
r
ea
l
-
wo
r
ld
tr
an
s
itio
n
s
(
e.
g
.
,
th
e
r
ap
id
d
r
o
p
af
te
r
1
0
PM
o
n
D
ec
1
)
.
T
h
is
is
e
x
p
ec
ted
,
as
Pr
o
p
h
et
u
s
es
p
iece
wis
e
lin
ea
r
o
r
lo
g
is
tic
g
r
o
wth
tr
en
d
s
with
ch
an
g
e
p
o
in
t
r
e
g
u
lar
izatio
n
,
wh
ich
p
r
io
r
itizes
s
m
o
o
th
n
ess
o
v
e
r
ca
p
t
u
r
in
g
ab
r
u
p
t
ch
an
g
es
[
7
]
.
W
h
ile
b
en
ef
icial
f
o
r
n
o
is
e
r
ed
u
ctio
n
,
th
is
ca
n
r
ed
u
ce
r
esp
o
n
s
iv
en
ess
to
s
u
d
d
en
s
h
if
ts
in
n
e
two
r
k
b
eh
a
v
io
r
(
e.
g
.
,
u
n
ex
p
ec
ted
o
u
tag
es,
b
u
r
s
ts
,
o
r
p
o
licy
ch
a
n
g
es).
d.
Data
q
u
ality
co
n
s
id
er
atio
n
s
:
o
n
e
d
ata
p
o
in
t,
th
at
is
o
n
D
ec
em
b
er
2
,
1
2
:0
0
PM
ap
p
ea
r
s
to
co
n
tain
a
f
o
r
m
attin
g
er
r
o
r
:
“2
3
,
7
3
6
,
4
2
2
,
2
7
”
lik
ely
s
h
o
u
ld
b
e
2
3
,
7
3
6
,
4
2
2
.
2
7
(
u
s
in
g
a
p
e
r
io
d
as
d
ec
im
al
s
ep
ar
ato
r
)
.
Su
ch
in
co
n
s
is
ten
cies,
if
u
n
co
r
r
ec
ted
,
ca
n
d
is
to
r
t e
r
r
o
r
ca
lcu
l
atio
n
s
an
d
m
o
d
el
ev
alu
atio
n
.
T
h
is
u
n
d
er
s
co
r
es
th
e
im
p
o
r
tan
ce
o
f
r
ig
o
r
o
u
s
d
a
ta
p
r
ep
r
o
ce
s
s
in
g
in
tim
e
s
er
ies
f
o
r
ec
asti
n
g
,
as
em
p
h
asized
in
b
est
p
r
ac
tices
f
o
r
o
p
er
atio
n
al
m
ac
h
in
e
lear
n
in
g
[
27
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
N
etw
o
r
k
t
r
a
ffic a
n
a
lysi
s
a
n
d
b
a
n
d
w
id
th
fo
r
ec
a
s
tin
g
f
o
r
u
s
in
g
Meta
’
s
P
r
o
p
h
et:
…
(
Yu
s
u
f
On
imis
i I
s
a
a
c
)
759
e.
Pra
ctica
l
im
p
licatio
n
s
f
o
r
n
etwo
r
k
m
an
ag
em
e
n
t:
th
e
r
esu
lts
s
u
g
g
est
th
at
Pro
p
h
et
is
a
v
iab
le
to
o
l
f
o
r
s
h
o
r
t
-
ter
m
b
an
d
wid
th
f
o
r
ec
asti
n
g
i
n
ca
m
p
u
s
o
r
e
n
ter
p
r
is
e
n
etw
o
r
k
s
with
p
r
ed
ictab
le
u
s
ag
e
cy
cles.
Acc
u
r
ate
f
o
r
ec
asts
en
ab
le
p
r
o
ac
tiv
e
b
a
n
d
wid
th
allo
ca
tio
n
a
n
d
Q
o
S
o
p
tim
izatio
n
,
ca
p
ac
ity
p
lan
n
i
n
g
f
o
r
p
ea
k
-
h
o
u
r
in
f
r
astru
ctu
r
e
,
an
d
d
em
a
n
d
s
,
E
ar
ly
d
etec
tio
n
o
f
a
n
o
m
alies
(
e.
g
.
,
wh
en
ac
tu
al
tr
af
f
ic
d
ev
ia
tes
s
ig
n
if
ican
tly
f
r
o
m
f
o
r
ec
ast).
Ho
wev
er
,
f
o
r
en
v
ir
o
n
m
en
ts
with
h
ig
h
v
o
latilit
y
o
r
f
r
eq
u
en
t
an
o
m
alies,
Pro
p
h
et
s
h
o
u
ld
b
e
co
m
p
lem
en
ted
with
an
o
m
al
y
d
etec
tio
n
s
y
s
tem
s
o
r
h
y
b
r
id
m
o
d
els
(
e.
g
.
,
Pro
p
h
et
+
L
S
T
M)
to
im
p
r
o
v
e
r
esp
o
n
s
iv
en
ess
[
28
]
.
T
h
e
F
ig
u
r
e
8
s
h
o
ws
th
e
g
r
ap
h
ical
r
ep
r
esen
tatio
n
o
f
th
e
ac
tu
al
d
ata
an
d
f
o
r
ec
asted
d
ata
o
f
b
an
d
wid
th
u
s
ag
e
in
Ab
r
ah
am
Hall
o
v
er
a
tim
e
in
ter
v
al
o
f
f
i
v
e
d
ay
s
;
wh
ich
is
f
r
o
m
th
e
1
to
5
Dec
em
b
er
2
0
2
4
.
T
a
b
le
1
s
h
o
ws
th
e
ex
tr
ac
ts
a
n
d
c
o
m
p
ar
es
th
e
v
al
u
es
o
f
th
e
ac
tu
al
d
ata
an
d
f
o
r
ec
asted
d
ata
u
s
ed
f
o
r
th
e
s
tu
d
y
,
f
r
o
m
F
ig
u
r
e
8
b
u
t
f
o
r
o
v
er
th
r
ee
d
ay
s
(
f
r
o
m
1
to
3
Dec
em
b
er
2
0
2
4
)
.
T
er
m
i
n
o
lo
g
ies
a
b
o
v
e:
(
i)
ac
tu
al
d
ata:
th
is
co
m
p
r
is
es
o
f
th
e
r
ea
l
v
al
u
es
o
f
d
ata
r
ec
o
r
d
ed
f
r
o
m
th
e
s
y
s
tem
;
an
d
(
ii)
f
o
r
ec
asted
d
ata:
th
is
co
m
p
r
is
es
o
f
th
e
p
r
ed
icted
v
al
u
e
o
f
d
ata
d
e
r
iv
e
d
f
r
o
m
o
u
r
f
o
r
ec
asti
n
g
m
o
d
el.
T
h
e
av
er
ag
e
ab
s
o
lu
te
d
if
f
er
en
ce
b
etwe
en
th
e
f
o
r
ec
asted
an
d
ac
tu
al
b
an
d
wid
th
,
k
n
o
wn
as
th
e
MA
E
,
was
ab
o
u
t
1
0
.
1
Mb
p
s
.
T
h
is
was
d
er
iv
ed
f
r
o
m
an
e
q
u
iv
ale
n
t
v
alu
e
o
f
1
0
,
0
9
9
,
8
6
3
.
1
0
bps
.
T
h
e
R
MSE
,
an
o
th
er
m
ea
s
u
r
e
o
f
p
r
e
d
ictio
n
ac
cu
r
ac
y
,
was
r
o
u
g
h
l
y
1
3
.
6
Mb
p
s
,
ca
l
cu
lated
f
r
o
m
1
3
,
5
7
0
,
9
5
9
.
5
8
b
p
s
.
T
h
e
s
tu
d
y
its
elf
p
r
o
v
id
e
d
c
o
n
tex
t
f
o
r
u
n
d
er
s
ta
n
d
in
g
th
ese
er
r
o
r
v
alu
es.
I
t
s
t
ated
th
at
t
h
e
MA
E
a
n
d
R
MSE
wer
e
c
o
n
s
id
er
ed
“
r
ea
s
o
n
ab
le
”
in
lig
h
t
o
f
th
e
a
ctu
al
h
ig
h
est
b
an
d
wid
t
h
u
s
ag
e,
wh
ich
v
ar
ied
b
etwe
en
4
7
Mb
p
s
an
d
5
0
Mb
p
s
.
T
h
is
im
p
lies
th
at
wh
ile
th
e
er
r
o
r
am
o
u
n
ts
ar
e
n
u
m
er
ically
l
ar
g
e,
r
ef
lectin
g
th
e
s
u
b
s
tan
tial
s
ca
le
o
f
th
e
d
ata,
th
ey
r
e
p
r
esen
t
an
ac
ce
p
tab
le
l
ev
el
o
f
in
ac
cu
r
ac
y
r
elativ
e
to
th
e
s
y
s
tem
’
s
m
ax
im
u
m
ca
p
ac
ity
.
Un
f
o
r
tu
n
ately
,
th
e
av
ailab
le
i
n
f
o
r
m
atio
n
d
o
es
n
o
t
in
clu
d
e
s
p
ec
if
ic,
n
u
m
er
ical
MA
E
o
r
R
MSE
v
al
u
es
f
r
o
m
p
r
e
v
io
u
s
f
o
r
ec
asti
n
g
m
eth
o
d
s
,
s
u
ch
as
AR
I
MA
o
r
L
STM
,
wh
en
ap
p
l
ied
to
th
e
s
am
e
d
ataset.
T
h
e
s
tu
d
y
o
n
ly
m
e
n
tio
n
s
th
at
AR
I
MA
an
d
L
STM
ar
e
b
ein
g
co
n
s
id
er
ed
f
o
r
f
u
t
u
r
e
ass
ess
m
en
t.
Ho
wev
er
,
b
a
s
ed
o
n
q
u
alitativ
e
o
b
s
er
v
atio
n
s
,
t
h
e
s
tu
d
y
p
r
o
p
o
s
es
th
at
Pro
p
h
et
o
f
f
e
r
s
b
e
n
ef
its
in
ter
m
s
o
f
ea
s
e
o
f
u
n
d
er
s
tan
d
in
g
an
d
ef
f
icie
n
t
u
s
e
o
f
r
eso
u
r
ce
s
wh
en
co
m
p
a
r
ed
to
th
ese
m
o
r
e
in
tr
icate
m
eth
o
d
s
.
As
a
r
esu
lt,
th
e
au
th
o
r
s
d
id
n
o
t
m
ak
e
a
d
ir
ec
t,
q
u
a
n
titativ
e
co
m
p
ar
is
o
n
o
f
t
h
e
er
r
o
r
s
izes
ag
ain
s
t
th
ese
alter
n
ativ
e
ap
p
r
o
ac
h
es
at
th
is
tim
e.
I
t
is
r
ec
o
m
m
en
d
ed
f
o
r
f
u
tu
r
e
s
tu
d
y
.
T
ab
le
1
: A
T
ab
le
s
h
o
win
g
th
e
ac
tu
al
an
d
f
o
r
ec
asted
d
ata
D
a
t
e
A
c
t
u
a
l
d
a
t
a
F
o
r
e
c
a
st
e
d
d
a
t
a
1
2
/
1
/
2
0
2
4
1
0
:
0
0
:
0
0
P
M
4
2
1
3
7
7
6
2
.
7
8
3
8
2
2
4
1
4
9
.
0
6
1
2
/
1
/
2
0
2
4
1
1
:
0
0
:
0
0
P
M
3
8
0
7
8
2
6
8
.
0
2
3
2
8
3
0
5
7
9
.
8
9
1
2
/
2
/
2
0
2
4
1
2
:
0
0
:
0
0
A
M
3
2
2
1
4
7
4
3
.
1
3
2
5
6
9
3
0
0
1
.
5
2
1
2
/
2
/
2
0
2
4
1
:
0
0
:
0
0
A
M
2
5
4
5
4
9
6
0
.
0
9
1
8
6
5
1
6
4
1
.
2
4
1
2
/
2
/
2
0
2
4
2
:
0
0
:
0
0
A
M
1
9
1
3
7
9
7
2
.
9
2
1
3
7
9
5
1
5
2
.
1
2
1
2
/
2
/
2
0
2
4
3
:
0
0
:
0
0
A
M
1
4
8
4
0
5
6
2
.
1
5
1
2
4
6
2
7
6
1
.
0
5
1
2
/
2
/
2
0
2
4
4
:
0
0
:
0
0
A
M
1
3
7
5
2
2
2
5
.
9
5
1
4
5
1
4
5
7
0
.
0
1
1
2
/
2
/
2
0
2
4
5
:
0
0
:
0
0
A
M
1
5
9
1
6
4
0
5
.
2
4
1
8
3
9
5
2
8
1
1
2
/
2
/
2
0
2
4
6
:
0
0
:
0
0
A
M
1
9
9
7
2
4
9
6
.
4
5
2
2
0
2
4
2
1
8
.
3
9
1
2
/
2
/
2
0
2
4
7
:
0
0
:
0
0
A
M
2
3
7
9
8
8
8
2
.
2
8
2
3
9
4
9
1
7
9
.
6
1
2
/
2
/
2
0
2
4
8
:
0
0
:
0
0
A
M
2
5
7
3
0
8
3
7
.
1
9
2
4
0
3
1
0
8
3
.
9
4
1
2
/
2
/
2
0
2
4
9
:
0
0
:
0
0
A
M
2
5
5
0
5
1
3
5
.
1
2
3
2
4
3
5
1
1
.
1
2
1
2
/
2
/
2
0
2
4
1
0
:
0
0
:
0
0
A
M
2
4
2
1
6
6
5
8
.
1
1
2
2
8
3
5
9
3
0
.
1
2
1
2
/
2
/
2
0
2
4
1
1
:
0
0
:
0
0
A
M
2
3
3
5
7
3
3
8
.
3
3
2
3
5
0
8
3
5
6
.
8
6
1
2
/
2
/
2
0
2
4
1
2
:
0
0
:
0
0
P
M
2
3
7
3
6
4
2
2
.
2
7
2
5
1
4
3
0
1
5
.
8
4
1
2
/
2
/
2
0
2
4
1
:
0
0
:
0
0
P
M
2
5
1
2
3
6
5
8
.
5
9
2
7
1
4
9
3
6
5
.
8
1
1
2
/
2
/
2
0
2
4
2
:
0
0
:
0
0
P
M
2
6
7
8
5
2
7
9
.
2
9
2
9
0
3
1
6
9
5
.
8
1
1
2
/
2
/
2
0
2
4
3
:
0
0
:
0
0
P
M
2
8
3
1
5
3
9
1
.
2
3
0
7
2
6
9
1
0
.
4
7
1
2
/
2
/
2
0
2
4
4
:
0
0
:
0
0
P
M
2
9
9
8
3
4
3
9
.
9
9
3
2
5
2
4
9
0
2
.
9
1
1
2
/
2
/
2
0
2
4
5
:
0
0
:
0
0
P
M
3
2
3
3
8
3
5
8
.
9
9
3
4
3
7
5
6
1
6
.
1
1
1
2
/
2
/
2
0
2
4
6
:
0
0
:
0
0
P
M
3
5
5
0
5
7
1
4
.
9
8
3
7
3
7
5
0
6
6
.
6
7
1
2
/
2
/
2
0
2
4
7
:
0
0
:
0
0
P
M
3
8
8
4
2
4
9
5
.
8
4
0
0
1
5
8
0
7
1
2
/
2
/
2
0
2
4
8
:
0
0
:
0
0
P
M
4
1
2
2
0
6
8
3
.
0
7
4
1
8
4
7
9
5
8
.
4
6
1
2
/
2
/
2
0
2
4
9
:
0
0
:
0
0
P
M
4
1
6
3
3
1
9
5
.
6
2
4
1
8
9
0
9
1
0
.
9
6
1
2
/
2
/
2
0
2
4
1
0
:
0
0
:
0
0
P
M
3
9
6
1
4
0
2
0
.
0
2
3
9
3
6
8
4
5
6
.
3
5
1
2
/
2
/
2
0
2
4
1
1
:
0
0
:
0
0
P
M
3
5
2
6
7
2
3
8
.
2
3
3
4
1
7
8
1
8
4
.
4
1
2
/
3
/
2
0
2
4
1
2
:
0
0
:
0
0
A
M
2
9
1
1
9
2
4
0
.
9
4
2
7
2
3
5
2
2
7
.
3
4
1
2
/
3
/
2
0
2
4
1
:
0
0
:
0
0
A
M
2
2
0
8
1
7
7
3
.
2
3
2
0
9
4
7
3
1
0
.
6
4
1
2
/
3
/
2
0
2
4
2
:
0
0
:
0
0
A
M
1
5
4
9
7
8
8
0
.
9
4
1
5
6
9
4
3
0
7
.
7
2
1
2
/
3
/
2
0
2
4
3
:
0
0
:
0
0
A
M
1
0
9
4
8
3
0
4
.
1
2
1
4
5
2
1
4
3
0
.
0
5
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
TEL
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
3
,
J
u
n
e
20
26
:
7
5
1
-
7
6
4
760
Fig
u
r
e
8
.
Gr
a
p
h
ical
illu
s
tr
atio
n
o
f
th
e
ac
tu
al
d
ata
a
n
d
f
o
r
ec
a
s
ted
d
ata
o
f
b
a
n
d
wid
th
u
s
ag
e
Ab
r
ah
am
Hall
o
v
e
r
a
tim
e
in
ter
v
al
4.
L
I
M
I
T
AT
I
O
NS A
N
D
RE
C
O
M
M
E
NDA
T
I
O
N
S F
O
R
F
UT
UR
E
WO
RK
T
h
is
s
tu
d
y
’
s
lim
itatio
n
s
,
w
h
ich
p
o
in
t
to
war
d
s
im
p
o
r
tan
t
a
r
e
as
f
o
r
f
u
tu
r
e
r
esear
c
h
an
d
a
r
e
cr
u
cial
f
o
r
p
u
b
licatio
n
,
p
r
im
a
r
ily
c
o
n
ce
r
n
h
o
w
well
th
e
m
o
d
el
ca
n
b
e
ap
p
lied
to
d
if
f
e
r
en
t
s
itu
atio
n
s
a
n
d
h
o
w
ea
s
ily
it
ca
n
b
e
in
co
r
p
o
r
ated
in
to
ex
is
tin
g
o
p
er
atio
n
s
.
Sp
ec
if
ically
:
−
R
estricte
d
tim
e
p
er
io
d
:
th
e
m
o
d
el
was
d
ev
el
o
p
ed
u
s
in
g
d
at
a
o
n
ly
f
r
o
m
Octo
b
e
r
to
Dec
e
m
b
er
2
0
2
4
.
T
h
is
n
ar
r
o
w
f
o
cu
s
m
ig
h
t
h
in
d
e
r
its
ab
ilit
y
to
ac
cu
r
ately
r
ef
lect
wid
er
s
ea
s
o
n
al
ch
a
n
g
es,
s
u
ch
as
h
ig
h
d
em
a
n
d
d
u
r
in
g
ex
a
m
p
e
r
io
d
s
o
r
l
o
w
u
s
ag
e
d
u
r
in
g
h
o
lid
ay
p
e
r
io
d
s
.
T
h
er
e
f
o
r
e,
f
u
t
u
r
e
wo
r
k
s
h
o
u
ld
in
v
o
lv
e
ex
p
an
d
i
n
g
th
e
d
ataset
to
co
v
er
an
en
tire
ac
ad
e
m
ic
y
ea
r
.
−
Ma
n
ag
in
g
u
n
ex
p
ec
te
d
in
cr
ea
s
es:
wh
ile
Pro
p
h
et
is
g
o
o
d
at
m
o
d
elin
g
estab
lis
h
ed
p
atter
n
s
an
d
s
ea
s
o
n
al
ch
an
g
es,
it
ten
d
s
to
u
n
d
er
esti
m
ate
s
u
d
d
en
,
u
n
u
s
u
al
s
p
ik
e
s
in
n
etwo
r
k
tr
af
f
ic
u
n
less
s
p
ec
if
ic
ex
ter
n
al
f
ac
to
r
s
ar
e
in
clu
d
ed
.
T
h
e
c
u
r
r
en
t
s
y
s
tem
lack
s
b
u
ilt
-
in
to
o
ls
f
o
r
id
en
tify
in
g
an
d
p
r
o
ac
ti
v
ely
m
an
ag
in
g
th
ese
u
n
p
r
e
d
ictab
le
ev
en
ts
.
−
Ab
s
en
ce
o
f
d
ir
ec
t
m
o
d
el
co
m
p
ar
is
o
n
:
th
e
c
u
r
r
en
t
v
alid
atio
n
is
b
ased
o
n
s
u
b
jectiv
e
ass
ess
m
en
ts
co
m
p
ar
e
d
to
o
th
er
c
o
m
p
lex
f
o
r
ec
asti
n
g
m
o
d
els.
A
v
ital
n
e
x
t
s
tep
i
s
a
d
ir
ec
t
co
m
p
a
r
is
o
n
u
s
in
g
m
o
d
els
s
u
ch
as
AR
I
MA
an
d
L
STM
o
n
th
e
s
am
e
d
ata.
T
h
is
wo
u
l
d
r
i
g
o
r
o
u
s
ly
ass
ess
th
e
b
alan
ce
b
et
wee
n
Pro
p
h
et
’
s
r
eso
u
r
ce
ef
f
icien
c
y
an
d
its
f
o
r
ec
asti
n
g
ac
cu
r
ac
y
,
esp
ec
ially
f
o
r
ex
tr
em
e
o
u
tlier
s
.
−
Fo
cu
s
o
n
o
v
er
all
b
a
n
d
wid
th
:
t
h
e
cu
r
r
en
t
m
et
h
o
d
an
al
y
ze
s
to
tal
b
an
d
wid
th
u
s
ag
e.
Fu
tu
r
e
r
esear
ch
s
h
o
u
ld
in
co
r
p
o
r
ate
d
ee
p
p
ac
k
et
in
s
p
ec
tio
n
(
DPI
)
to
an
aly
ze
tr
af
f
ic
b
ased
o
n
th
e
s
p
ec
if
ic
ap
p
licatio
n
s
b
ein
g
u
s
ed
.
T
h
is
wo
u
ld
en
ab
le
m
o
r
e
r
ef
in
e
d
tr
af
f
ic
m
a
n
ag
em
en
t
an
d
p
r
io
r
itizatio
n
s
tr
ateg
ies.
T
o
en
h
an
ce
t
h
e
o
v
e
r
all
p
er
f
o
r
m
an
ce
,
s
ca
lab
ilit
y
,
an
d
u
tili
ty
o
f
th
e
n
etwo
r
k
b
a
n
d
wid
th
f
o
r
ec
asti
n
g
s
y
s
tem
,
th
e
f
o
llo
win
g
r
ec
o
m
m
en
d
atio
n
s
ar
e
p
r
o
p
o
s
ed
:
Evaluation Warning : The document was created with Spire.PDF for Python.