T
E
L
K
O
M
NIKA
T
elec
o
mm
un
ica
t
io
n Co
m
pu
t
i
ng
E
lect
ro
nics
a
nd
Co
ntr
o
l
Vo
l.
24
,
No
.
2
,
A
p
r
il
20
26
,
p
p
.
452
~
465
I
SS
N:
1693
-
6
9
3
0
,
DOI
: 1
0
.
1
2
9
2
8
/
T
E
L
KOM
NI
K
A
.
v
24
i
2
.
27316
452
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
H
y
brid in
trusio
n
detec
tion in Io
T d
ev
ices: a dee
p lea
rning
a
ppro
a
ch using
Kitsune a
nd
quan
ti
z
e
d auto
enco
der
M
d.
Rif
a
t
E
No
o
r,
M
d
.
T
o
f
a
el
Ah
m
ed,
Du
la
l C
ha
k
ra
bo
rt
y
,
P
intu C
ha
nd
ra
P
a
ul
,
So
ha
na
No
w
a
r
,
Rej
w
a
n Ah
m
ed,
T
a
njina
A
kte
r
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
mat
i
o
n
a
n
d
C
o
m
mu
n
i
c
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
C
o
mi
l
l
a
U
n
i
v
e
r
si
t
y
,
C
u
mi
l
l
a
,
B
a
n
g
l
a
d
e
sh
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
2
4
,
2
0
2
5
R
ev
i
s
ed
Dec
8
,
2
0
2
5
A
cc
ep
ted
J
an
3
0
,
2
0
2
6
In
tern
e
t
o
f
th
i
n
g
s
(Io
T
)
h
a
s
b
e
e
n
tran
sf
o
rm
in
g
th
e
w
a
y
to
c
o
n
n
e
c
t
a
n
d
c
o
m
m
u
n
ica
te
in
sm
a
rt
h
o
m
e
s,
h
e
a
lt
h
c
a
re
,
a
n
d
b
u
si
n
e
ss
e
s
so
f
a
st
a
n
d
ra
p
id
ly
a
ro
u
n
d
t
h
e
w
o
rld
.
Bu
t
th
is
g
ro
w
th
h
a
s
c
o
m
p
li
c
a
ted
se
c
u
rit
y
,
b
e
c
a
u
se
Io
T
d
e
v
ice
s
a
re
m
o
re
li
k
e
l
y
to
b
e
h
a
c
k
e
d
a
s
th
e
y
’r
e
s
m
a
ll
e
r,
w
it
h
o
u
t
e
v
e
n
re
g
u
lar
se
c
u
rit
y
p
ra
c
ti
c
e
s,
a
n
d
u
n
d
e
r
a
tt
a
c
k
b
y
m
o
re
so
p
h
isti
c
a
ted
th
re
a
ts.
T
ra
d
it
io
n
a
l
i
n
tr
u
sio
n
d
e
tec
ti
o
n
sy
ste
m
s
(IDS)
a
re
n
o
t
f
u
n
c
ti
o
n
i
n
g
w
e
ll
in
Io
T
e
n
v
iro
n
m
e
n
ts
a
s
th
e
y
a
r
e
c
o
m
p
u
tatio
n
a
ll
y
e
x
p
e
n
siv
e
a
n
d
stru
g
g
le
to
acco
m
m
o
d
a
te
th
e
h
e
tero
g
e
n
e
o
u
s
n
a
tu
re
o
f
Io
T
n
e
tw
o
rk
s.
Th
is
p
a
p
e
r
in
tro
d
u
c
e
s
a
c
ro
ss
-
d
o
m
a
in
in
tr
u
si
o
n
d
e
tec
ti
o
n
b
a
se
d
o
n
a
d
a
p
ti
v
e
a
d
v
e
rsa
rial
train
in
g
u
si
n
g
Kitsu
n
e
a
n
d
q
u
a
n
ti
z
e
d
a
u
t
o
e
n
c
o
d
e
rs
(QA
E)
f
o
r
a
n
o
m
a
l
y
d
e
tec
ti
o
n
a
n
d
c
las
sif
ica
ti
o
n
.
T
h
e
m
o
d
e
l
is
c
a
p
a
b
le
o
f
c
a
p
tu
rin
g
d
iff
e
re
n
t
a
tt
a
c
k
in
g
tec
h
n
iq
u
e
s,
su
c
h
a
s
d
i
strib
u
te
d
d
e
n
ial
o
f
se
rv
ice
(DD
o
S
),
M
irai
b
o
t
n
e
t
a
tt
a
c
k
s,
a
d
d
re
ss
re
so
lu
ti
o
n
p
r
o
t
o
c
o
l
(A
RP
)
sp
o
o
f
in
g
,
a
n
d
d
a
ta
e
x
f
il
tratio
n
,
b
y
lev
e
ra
g
in
g
th
e
re
c
o
n
stru
c
ti
o
n
e
rro
r
g
e
n
e
ra
ted
b
y
Kitsu
n
e
a
u
to
e
n
c
o
d
e
rs.
T
h
e
d
e
g
re
e
-
b
a
se
d
c
las
sif
ica
ti
o
n
e
n
a
b
les
th
e
s
y
ste
m
to
d
y
n
a
m
ic
a
ll
y
c
a
te
g
o
rize
a
n
o
m
a
li
e
s
a
c
c
o
rd
in
g
to
t
h
e
ir
se
v
e
rit
y
,
re
n
d
e
rin
g
th
e
m
o
d
e
l
e
x
c
e
p
ti
o
n
a
ll
y
a
d
a
p
ti
v
e
to
v
a
rio
u
s
a
tt
a
c
k
s.
T
h
e
a
n
o
m
a
li
e
s
a
re
a
lso
c
las
si
f
ied
in
to
d
if
f
e
re
n
t
t
y
p
e
s
o
f
a
tt
a
c
k
s
(n
o
r
m
a
l,
su
sp
icio
u
s,
a
n
d
m
a
li
c
io
u
s)
b
a
se
d
o
n
b
in
a
rize
d
e
rro
r
v
a
lu
e
s.
T
h
e
a
p
p
ro
a
c
h
a
c
h
iev
e
s
a
h
ig
h
a
c
c
u
ra
c
y
w
it
h
a
n
F
1
sc
o
re
o
f
8
5
.
9
%
a
n
d
s
u
p
p
o
rts
re
a
l
-
ti
m
e
c
h
a
ra
c
teriz
a
ti
o
n
t
o
in
c
re
a
se
se
c
u
rit
y
in
Io
T
sc
e
n
a
rio
s.
.
K
ey
w
o
r
d
s
:
An
o
m
al
y
d
etec
tio
n
Dee
p
lear
n
in
g
I
n
ter
n
et
o
f
t
h
i
n
g
s
s
ec
u
r
it
y
I
n
tr
u
s
io
n
d
etec
tio
n
s
y
s
te
m
Kits
u
n
e
Qu
a
n
tized
au
to
en
co
d
er
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
:
Md
T
o
f
ae
l
Ah
m
ed
Dep
ar
t
m
en
t o
f
I
n
f
o
r
m
atio
n
an
d
C
o
m
m
u
n
icatio
n
T
ec
h
n
o
lo
g
y
,
C
o
m
illa U
n
i
v
er
s
it
y
Ko
tb
ar
i
-
3
5
0
6
,
C
u
m
illa,
B
an
g
l
ad
esh
E
m
ail:
to
f
ae
l@
co
u
.
ac
.
b
d
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
im
p
ac
t
o
f
th
e
i
n
ter
n
et
o
f
th
i
n
g
s
(
I
o
T
)
o
n
th
e
w
a
y
u
s
e
r
s
in
ter
ac
t
w
it
h
t
h
e
w
o
r
ld
,
b
e
it
s
m
ar
t
h
o
m
e
s
o
r
h
ea
lth
ca
r
e
s
y
s
te
m
s
,
in
d
u
s
tr
ial
au
to
m
atio
n
,
o
r
cr
itical
in
f
r
astru
c
tu
r
es,
h
as
b
ee
n
n
o
th
in
g
s
h
o
r
t
o
f
r
e
m
ar
k
ab
le.
T
h
er
e
w
ill
b
e
a
m
i
n
i
m
u
m
o
f
3
9
.
6
b
illi
o
n
I
o
T
d
ev
ices
i
n
u
s
e
b
y
2
0
3
0
,
ex
p
er
ts
h
a
v
e
s
aid
.
T
h
ese
d
ev
ices
w
ill
cr
ea
te
a
to
n
o
f
d
ata
an
d
f
o
s
ter
i
n
n
o
v
atio
n
i
n
m
u
ltip
le
in
d
u
s
tr
ies.
B
u
t
w
i
th
t
h
i
s
r
ap
id
g
r
o
w
t
h
also
co
m
e
n
e
w
a
n
d
d
an
g
er
o
u
s
s
e
cu
r
it
y
th
r
ea
t
s
.
I
t
is
e
f
f
o
r
tles
s
f
o
r
c
y
b
er
-
attac
k
s
to
b
e
d
ir
ec
ted
at
m
u
ltip
le
I
o
T
d
ev
ices
w
ith
t
h
e
co
n
s
tr
ai
n
ts
o
f
co
m
p
u
ti
n
g
ca
p
ab
ilit
y
,
ar
b
itr
ar
y
f
o
r
m
an
d
s
ize,
an
d
th
e
ab
s
en
ce
o
f
a
co
m
m
o
n
s
ec
u
r
e
m
ec
h
a
n
i
s
m
[
1
]
.
T
o
s
ee
h
o
w
co
m
p
lex
th
e
I
o
T
w
o
r
ld
is
,
j
u
s
t th
i
n
k
ab
o
u
t h
o
w
d
i
f
f
icu
l
t
it is
f
o
r
tr
ad
itio
n
al
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
te
m
s
(
I
DS)
to
d
ea
l
w
it
h
t
h
at
w
o
r
ld
.
Fo
r
ex
a
m
p
le,
an
I
DS
b
ased
o
n
s
ig
n
at
u
r
e
s
ca
n
id
en
ti
f
y
k
n
o
w
n
attac
k
s
,
b
u
t
it
do
es
n
o
t
p
er
f
o
r
m
e
f
f
ec
t
iv
el
y
f
o
r
n
e
w
,
u
n
h
ea
r
d
th
r
ea
t
s
.
An
o
m
al
y
-
b
a
s
ed
,
lik
e
I
DS,
th
e
y
ca
n
b
e
u
s
ed
to
d
etec
t
s
u
s
p
icio
u
s
b
eh
a
v
io
r
;
h
o
w
e
v
er
,
th
e
y
r
eq
u
ir
e
to
o
m
a
n
y
r
e
s
o
u
r
ce
s
,
an
d
t
h
e
y
ar
e
p
r
o
n
e
to
f
alse
p
o
s
itiv
e
s
;
th
i
s
is
n
’
t
a
g
o
o
d
r
is
k
to
r
u
n
o
n
a
b
u
n
ch
o
f
d
ev
ices
lik
e
all
I
o
T
d
ev
ices
ar
e.
Ov
er
6
0
%
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Hyb
r
id
in
tr
u
s
io
n
d
etec
tio
n
in
I
o
T d
ev
ices:
a
d
ee
p
lea
r
n
in
g
a
p
p
r
o
a
ch
u
s
in
g
…
(
Md
.
R
ifa
t E
N
o
o
r
)
453
o
f
I
o
T
n
et
w
o
r
k
s
g
et
h
it
b
y
a
s
ec
u
r
it
y
i
n
cid
en
t
ea
c
h
y
ea
r
,
f
r
o
m
d
i
s
tr
ib
u
ted
d
en
ial
o
f
s
er
v
ice
(
DDo
S)
to
th
ef
t
o
f
d
ata
an
d
b
o
tn
et
in
f
estatio
n
s
.
Ho
w
e
v
er
,
th
e
ex
is
ti
n
g
I
DS
s
i
n
I
o
T
n
et
w
o
r
k
s
h
a
v
e
s
ev
e
r
al
li
m
itatio
n
s
.
Fir
s
t,
it
is
c
h
alle
n
g
i
n
g
f
o
r
r
eg
u
lar
s
y
s
te
m
s
,
s
u
c
h
as
s
i
g
n
atu
r
e
-
b
ased
I
DS,
to
h
a
n
d
le
h
eter
o
g
en
eo
u
s
I
o
T
en
v
ir
o
n
m
en
ts
,
w
h
er
ei
n
d
iv
er
s
e
t
y
p
es
o
f
d
ev
ices
h
av
e
co
n
s
tr
a
in
ed
r
eso
u
r
ce
s
.
T
h
ey
ar
e
h
e
n
ce
n
o
t
s
ca
lab
le
to
h
an
d
le
r
ea
l
-
ti
m
e
an
al
y
s
is
in
a
lar
g
e
n
et
w
o
r
k
a
n
d
ar
e
co
m
p
u
t
atio
n
all
y
e
x
p
en
s
i
v
e.
I
n
ad
d
itio
n
,
m
a
n
y
o
f
th
e
s
e
s
y
s
te
m
s
d
o
n
o
t
d
etec
t
ze
r
o
-
d
ay
attac
k
s
,
w
h
ic
h
h
a
v
e
b
ee
n
i
n
cr
e
asin
g
i
n
th
e
d
y
n
a
m
ic
I
o
T
en
v
i
r
o
n
m
e
n
t.
T
h
ese
s
h
o
r
tco
m
in
g
s
m
ak
e
it c
lear
th
at
a
m
o
r
e
e
f
f
icie
n
t,
s
ca
lab
le,
a
n
d
f
le
x
ib
le
I
DS
i
s
n
ee
d
ed
,
ca
p
ab
le
o
f
m
a
n
a
g
in
g
t
h
ese
p
r
o
b
le
m
s
.
T
h
e
p
r
o
p
o
s
ed
h
y
b
r
id
f
r
a
m
e
w
o
r
k
d
ir
ec
tly
ta
ck
les
th
e
s
e
is
s
u
es
o
f
[
2
]
im
p
r
o
v
in
g
th
e
co
m
p
u
tatio
n
a
l
co
m
p
le
x
it
y
a
n
d
[
3
]
s
ca
lab
ilit
y
,
a
n
d
p
r
o
v
id
es
a
s
o
l
u
tio
n
t
h
at
is
ca
p
ab
le
o
f
d
etec
t
in
g
b
o
th
k
n
o
w
n
a
n
d
u
n
k
n
o
w
n
attac
k
s
in
a
r
ea
l
-
ti
m
e
m
a
n
n
er
o
n
I
o
T
n
et
w
o
r
k
s
.
T
o
ad
d
r
ess
th
e
s
e
p
r
o
b
le
m
s
,
in
t
h
i
s
w
o
r
k
,
a
n
e
w
h
y
b
r
id
in
tr
u
s
io
n
d
etec
tio
n
ap
p
r
o
ac
h
is
p
r
esen
ted
.
T
h
e
ap
p
r
o
ac
h
co
m
b
i
n
ed
t
w
o
p
o
w
er
f
u
l te
ch
n
iq
u
e
s
: t
h
e
q
u
a
n
tized
au
t
o
en
co
d
er
(
QA
E
)
,
a
co
m
p
u
tatio
n
al
m
o
d
el
th
at
ca
n
p
r
o
d
u
ce
ac
cu
r
ate
r
es
u
lt
s
w
i
th
m
i
n
i
m
u
m
co
m
p
u
tat
io
n
al
c
o
s
t,
an
d
Kit
s
u
n
e,
a
m
ac
h
in
e
lear
n
in
g
(
ML
)
b
ased
an
o
m
al
y
d
etec
tio
n
s
y
s
te
m
,
wh
ich
i
s
ca
p
ab
le
o
f
an
al
y
zin
g
I
o
T
tr
af
f
ic
in
r
ea
l
-
ti
m
e.
T
h
is
s
y
s
te
m
i
s
b
u
i
lt
u
p
o
n
v
ar
io
u
s
s
tate
-
of
-
t
h
e
-
ar
t
tech
n
iq
u
e
s
f
o
r
d
ata
m
a
n
ip
u
lat
io
n
.
T
h
ese
ar
e:
o
n
e
-
h
o
t
en
co
d
in
g
,
o
v
er
s
a
m
p
li
n
g
to
r
em
o
v
e
th
e
c
lass
i
m
b
ala
n
ce
,
a
n
d
p
r
in
cip
al
co
m
p
o
n
e
n
t
a
n
al
y
s
i
s
(
P
C
A
)
to
m
a
k
e
o
u
r
d
ata
less
co
m
p
lex
.
I
n
th
is
wo
r
k
,
w
e
h
a
v
e
esp
ec
iall
y
v
ali
d
ated
th
is
s
etu
p
o
n
t
w
o
o
f
th
e
w
e
ll
-
ad
ap
ted
I
o
T
d
atasets
:
b
o
tn
et
-
I
o
T
(
B
o
t
-
I
o
T
)
(
w
h
ic
h
co
n
s
i
s
ts
o
f
o
f
f
li
n
e
t
r
af
f
ic)
an
d
r
ea
l
-
t
i
m
e
I
o
T
(
RT
-
I
o
T
2
0
2
2
)
(
w
h
ich
in
v
o
l
v
es
r
ea
l
-
ti
m
e
s
t
r
ea
m
in
g
)
.
I
t
ac
h
iev
ed
b
etter
p
er
f
o
r
m
an
c
e
(
ac
cu
r
ac
y
=
8
8
.
7
%,
p
r
ec
is
io
n
=
8
5
.
8
%,
r
ec
all
=
8
5
.
6
%,
an
d
F1
s
co
r
e
=
8
5
.
9
%)
th
an
o
t
h
er
f
ea
s
ib
le
m
et
h
o
d
s
.
Sev
er
al
v
a
lu
ab
le
co
n
tr
ib
u
tio
n
s
w
er
e
m
ad
e
b
y
t
h
e
r
esear
ch
r
e
p
o
r
ted
h
er
e:
−
L
i
g
h
t
w
ei
g
h
t
h
y
b
r
id
ar
ch
itectu
r
e
:
th
e
p
r
o
p
o
s
ed
f
r
a
m
e
w
o
r
k
is
s
u
itab
le
f
o
r
I
o
T
ed
g
e
d
ev
ices
b
y
co
m
b
i
n
i
n
g
th
e
f
a
s
t sear
c
h
alg
o
r
it
h
m
o
f
Q
A
E
w
it
h
t
h
e
an
o
m
al
y
d
etec
tio
n
o
f
Kit
s
u
n
e.
−
E
n
er
g
y
ef
f
icie
n
c
y
:
t
h
e
s
y
s
te
m
is
m
o
r
e
ap
p
r
o
ac
h
ab
le
f
o
r
I
o
T
-
b
ased
w
ea
r
ab
le
an
d
b
atter
y
-
d
r
iv
en
g
ad
g
et
s
d
u
e
to
th
e
m
et
h
o
d
o
f
q
u
a
n
tizat
io
n
th
at
r
ed
u
ce
s
m
e
m
o
r
y
a
n
d
p
r
o
ce
s
s
in
g
co
s
ts
b
y
a
t
h
ir
d
.
−
No
v
el
h
y
b
r
id
ap
p
r
o
ac
h
:
th
is
p
ap
er
in
tr
o
d
u
ce
s
th
e
n
e
w
h
y
b
r
id
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
te
m
t
h
r
o
u
g
h
t
h
e
co
m
b
i
n
atio
n
o
f
k
its
u
n
e
w
i
th
Q
A
E
.
T
h
is
s
o
l
u
tio
n
i
s
s
p
ec
i
f
ical
l
y
cr
ea
ted
f
o
r
I
o
T
o
f
T
h
in
g
s
e
n
v
ir
o
n
m
e
n
t
s
.
−
C
o
m
p
r
eh
e
n
s
i
v
e
an
a
l
y
s
is
:
w
e
d
em
o
n
s
tr
ate
t
h
e
ef
f
ec
ti
v
en
e
s
s
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
to
ac
cu
r
atel
y
d
etec
t
an
d
class
i
f
y
in
c
u
r
s
io
n
s
v
ia
t
wo
lar
g
e
-
s
ca
le
I
o
T
d
atasets
.
−
R
ea
l
-
w
o
r
ld
ap
p
licatio
n
s
:
w
e
p
r
o
v
id
e
p
r
ac
tical
r
ec
o
m
m
en
d
atio
n
s
f
o
r
b
alan
cin
g
cr
o
s
s
-
tr
an
s
ac
tio
n
co
m
p
u
tatio
n
s
p
ee
d
,
ac
cu
r
ac
y
,
an
d
r
eso
u
r
ce
co
n
s
u
m
p
tio
n
.
Descr
ip
tio
n
s
o
f
tec
h
n
iq
u
e
s
to
i
m
p
le
m
e
n
t
t
h
e
ar
ch
itect
u
r
e
in
r
ea
l
I
o
T
en
v
ir
o
n
m
e
n
t
s
ar
e
al
s
o
p
r
o
v
id
ed
.
T
h
is
w
o
r
k
t
h
e
n
f
i
n
all
y
i
m
p
r
o
v
e
s
t
h
e
s
ec
u
r
it
y
o
f
I
o
T
d
ev
ices
b
y
d
esi
g
n
i
n
g
an
elastic,
f
ea
s
ib
le,
an
d
lo
w
-
co
s
t
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
te
m
.
B
u
t
it
co
v
er
s
a
n
u
m
b
er
o
f
th
e
p
r
i
m
e
ch
al
len
g
es
t
h
e
I
o
T
in
d
u
s
tr
y
f
ac
es,
t
h
o
u
g
h
it
g
ets
b
ig
i
n
n
o
ti
m
e.
T
h
e
p
r
o
p
o
s
ed
s
tu
d
y
e
v
en
t
u
all
y
e
n
h
an
ce
s
t
h
e
s
ec
u
r
it
y
o
f
I
o
T
d
ev
ices
b
y
d
ev
elo
p
in
g
a
f
lex
i
b
le,
en
er
g
y
-
ef
f
ic
ien
t,
a
n
d
ap
p
licab
le
I
DS.
A
n
d
it a
d
d
r
ess
es so
m
e
o
f
t
h
e
g
r
ea
test
c
h
alle
n
g
es o
f
t
h
e
f
ast
-
e
x
p
an
d
in
g
I
o
T
w
o
r
ld
.
Kasin
a
th
a
n
et
a
l.
[
4
]
p
r
o
p
o
s
ed
a
n
e
w
s
o
lu
t
io
n
f
o
r
i
m
p
r
o
v
i
n
g
s
ec
u
r
it
y
i
n
I
o
T
an
d
w
ir
el
ess
s
e
n
s
o
r
n
et
w
o
r
k
s
(
W
SNs
)
.
T
o
en
h
an
c
e
s
ec
u
r
it
y
a
g
ai
n
s
t
w
ir
e
less
d
e
n
ial
-
of
-
s
er
v
ice
attac
k
s
a
n
d
to
s
p
ee
d
u
p
m
ess
a
g
e
d
is
s
e
m
in
at
io
n
,
t
h
eir
ar
ch
itect
u
r
e
co
n
n
ec
ts
th
e
I
D
S
n
o
d
e
to
t
h
e
p
ar
en
t
s
tatio
n
’
s
I
D
S.
Si
n
c
e
Su
r
icata
w
a
s
n
o
t
in
itial
l
y
b
u
il
t
f
o
r
h
an
d
li
n
g
h
a
n
d
le
n
o
n
-
i
n
ter
n
e
t
p
r
o
to
co
l
(
I
P
)
-
b
ased
n
et
w
o
r
k
s
,
th
e
d
e
s
ig
n
o
f
d
ec
o
d
er
s
r
e
q
u
ir
ed
f
o
r
it
to
u
n
d
er
s
tan
d
in
ter
n
et
p
r
o
to
co
l
v
er
s
io
n
6
(
I
P
v
6
)
is
a
m
aj
o
r
asp
ec
t
o
f
th
is
w
o
r
k
.
O
h
et
a
l.
[
5
]
p
r
o
p
o
s
ed
a
n
o
v
el
p
atter
n
m
atc
h
i
n
g
ap
p
r
o
ac
h
f
o
r
d
ev
ices
w
it
h
li
m
ite
d
r
eso
u
r
ce
s
,
w
h
ich
ap
p
lies
a
ttack
s
i
g
n
atu
r
e
s
o
f
C
la
m
AV
a
n
d
SNO
R
T
to
th
e
p
ac
k
et
s
tr
ea
m
.
So
m
e
cla
s
s
ic
I
D
Ss
,
li
k
e
S
u
r
icata
a
n
d
SNO
R
T
,
u
s
e
th
i
s
ap
p
r
o
a
ch
,
b
u
t
th
e
y
o
f
te
n
h
av
e
p
r
o
b
le
m
s
s
ca
lin
g
d
o
w
n
to
s
m
all
I
o
T
n
et
w
o
r
k
s
.
T
h
e
m
aj
o
r
is
s
u
e
i
s
t
h
at
t
h
e
r
u
le
s
e
ts
ar
e
p
r
etty
lar
g
e
a
n
d
r
eq
u
ir
e
a
h
u
g
e
a
m
o
u
n
t
o
f
co
m
p
u
ti
n
g
p
o
w
er
.
I
n
o
r
d
er
to
s
o
lv
e
t
h
is
is
s
u
e,
t
h
e
a
u
th
o
r
s
p
r
o
p
o
s
e
P
ass
b
an
,
a
p
o
r
tab
le
I
DS
th
at
d
is
co
v
e
r
s
n
e
w
th
r
ea
t
s
w
it
h
a
r
ed
u
ce
d
n
u
m
b
er
o
f
f
alse
p
o
s
i
tiv
es
v
ia
a
n
o
m
a
l
y
d
etec
tio
n
.
T
h
i
s
a
p
p
r
o
a
ch
h
o
l
d
s
p
o
t
e
n
t
i
al
,
b
u
t
a
d
d
i
t
i
o
n
al
r
es
e
a
r
c
h
i
s
n
e
e
d
e
d
t
o
a
d
a
p
t
i
t f
o
r
I
o
T
s
y
s
tem
s
.
H
e
im
d
al
l
,
a
n
I
D
S
u
s
in
g
w
h
it
e
li
s
t
s
t
o
d
en
y
DD
o
S
at
t
a
ck
s
lik
e
th
e
M
ir
a
i
b
o
tn
e
t
,
w
a
s
in
t
r
o
d
u
c
e
d
b
y
H
a
b
i
b
i
e
t
a
l
.
[
6
]
.
Vir
u
s
T
o
tal
lo
o
k
s
at
a
n
y
UR
L
o
r
DNS
r
esp
o
n
s
e
a
n
d
d
eter
m
in
e
s
it
s
s
a
f
et
y
s
tat
u
s
o
r
p
o
ten
tial
h
ar
m
.
T
h
e
g
ate
w
a
y
a
llo
w
s
tr
a
f
f
ic
o
n
l
y
f
r
o
m
s
o
u
r
ce
s
t
h
at
h
av
e
b
ee
n
a
u
th
o
r
ized
.
B
u
t
Hei
m
d
all
d
ep
en
d
s
a
g
r
ea
t
d
ea
l
o
n
Vir
u
s
T
o
tal
’
s
s
ec
u
r
it
y
,
s
o
it
is
at
r
is
k
o
f
ze
r
o
-
d
a
y
attac
k
s
b
ec
au
s
e
t
h
e
y
h
av
e
n
’
t
b
ee
n
an
al
y
ze
d
b
y
Vir
u
s
T
o
tal.
Hittin
g
a
r
e
m
o
te
e
n
d
p
o
in
t f
o
r
t
r
af
f
ic
a
n
al
y
s
i
s
also
s
lo
w
s
q
u
ic
k
th
e
s
y
s
te
m
ca
n
r
esp
o
n
se
.
W
allg
r
en
et
a
l.
[
7
]
in
v
est
ig
ate
d
th
e
v
u
l
n
er
ab
ilit
y
o
f
r
o
u
t
in
g
p
r
o
to
co
ls
to
s
elec
tiv
e
f
o
r
w
ar
d
i
n
g
a
ttack
s
in
lo
s
s
-
a
w
ar
e
a
n
d
lo
s
s
-
i
n
d
if
f
e
r
en
t
n
et
w
o
r
k
s
(
r
o
u
ti
n
g
p
r
o
to
co
l
f
o
r
lo
w
-
p
o
w
er
an
d
lo
s
s
y
n
et
w
o
r
k
s
/
R
P
L
)
,
an
d
also
lo
w
-
p
o
w
er
n
et
w
o
r
k
s
at
t
h
e
n
e
t
w
o
r
k
la
y
er
.
I
n
s
u
c
h
atta
ck
s
,
a
r
o
g
u
e
n
o
d
e
ad
v
er
tis
es
i
ts
elf
as
h
av
in
g
t
h
e
s
h
o
r
test
p
at
h
,
lead
in
g
tr
af
f
ic
t
o
b
e
r
er
o
u
ted
an
d
p
ac
k
ets
to
b
e
d
r
o
p
p
ed
.
Su
ch
an
attac
k
w
o
u
ld
d
estro
y
n
et
w
o
r
k
p
er
f
o
r
m
a
n
ce
.
Am
ar
al
et
a
l.
[
8
]
p
r
esen
t
a
t
r
af
f
ic
s
i
g
n
at
u
r
es
-
b
ased
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
te
m
f
o
r
W
SNs
.
T
h
eir
p
r
o
p
o
s
ed
s
o
lu
tio
n
m
et
h
o
d
is
d
iv
id
ed
in
to
th
r
ee
co
m
p
o
n
en
t
s
,
in
cl
u
d
in
g
a
p
ac
k
e
t
m
o
n
ito
r
in
g
co
m
p
o
n
e
n
t,
an
an
o
m
al
y
d
etec
tio
n
w
it
h
s
o
m
e
p
r
ed
eter
m
i
n
ed
r
u
les
,
a
n
d
a
n
a
ttack
n
o
tice
to
t
h
e
ad
m
i
n
i
s
tr
at
o
r
f
o
r
th
e
d
etec
ted
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
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
.
2
,
A
p
r
il
20
26
:
4
5
2
-
465
454
attac
k
.
J
u
n
a
n
d
C
h
i
[
9
]
in
tr
o
d
u
ce
d
an
ev
en
t
p
r
o
ce
s
s
in
g
en
g
in
e
to
d
etec
t
an
an
o
m
alo
u
s
p
atter
n
o
f
tr
af
f
ic
i
n
r
ea
l
-
ti
m
e
i
n
t
h
e
s
co
p
e
o
f
I
o
T
.
T
h
eir
r
u
les
-
b
a
s
ed
I
DS
is
lig
h
t
-
w
ei
g
h
t
in
ter
m
s
o
f
m
e
m
o
r
y
u
s
a
g
e,
b
u
t
u
s
e
s
a
s
ig
n
i
f
ic
a
n
t
a
m
o
u
n
t
o
f
ce
n
tr
a
l
p
r
o
ce
s
s
in
g
u
n
it
(
C
P
U)
r
eso
u
r
ce
s
an
al
y
zi
n
g
th
e
d
ata.
R
ieck
er
et
a
l.
[
1
0
]
co
n
ce
n
tr
ated
o
n
th
e
en
er
g
y
c
o
n
s
u
m
p
tio
n
o
f
I
DSs
f
o
r
W
S
Ns.
L
in
d
a
e
t
a
l.
[
1
1
]
p
r
o
p
o
s
ed
an
a
n
o
m
al
y
-
b
a
s
ed
I
DS
ap
p
licab
le
to
p
h
y
s
ical
i
n
f
r
a
s
tr
u
ct
u
r
es
s
u
c
h
as
w
ater
s
u
p
p
l
y
f
ac
ilit
ies
a
n
d
p
o
w
er
p
lan
ts
.
D
u
r
in
g
t
h
e
lear
n
in
g
p
er
io
d
,
th
e
s
y
s
te
m
co
n
s
tr
u
cts
a
r
e
f
er
en
ce
m
o
d
el
f
r
o
m
n
et
w
o
r
k
d
ata
a
n
d
th
e
n
c
h
ec
k
s
i
n
co
m
in
g
tr
a
f
f
ic
ag
ain
s
t
th
is
m
o
d
el.
T
h
ey
e
m
p
lo
y
ed
an
ar
ti
f
icial
n
e
u
r
al
n
et
wo
r
k
(
A
NN
)
th
at
is
u
s
ed
as
a
tr
af
f
ic
p
r
o
f
ilin
g
to
o
l
to
ass
is
t
a
n
o
m
a
l
y
d
etec
tio
n
s
y
s
te
m
s
i
n
d
is
ti
n
g
u
i
s
h
in
g
b
et
w
ee
n
t
h
e
n
o
r
m
al
tr
af
f
ic
an
d
th
e
attac
k
tr
af
f
ic.
Si
m
i
lar
l
y
,
Ho
d
o
et
a
l.
[
1
2
]
e
m
p
lo
y
ed
ANNs
f
o
r
Do
S/DDo
S
attac
k
d
etec
tio
n
w
it
h
in
I
o
T
s
.
Yet
th
eir
m
et
h
o
d
s
u
f
f
er
s
f
r
o
m
h
i
g
h
f
al
s
e
alar
m
s
w
h
e
n
ap
p
l
y
i
n
g
p
r
ed
ef
i
n
e
d
m
o
d
els.
I
t
is
t
h
er
e
f
o
r
e
r
elativ
el
y
u
s
eles
s
f
o
r
id
en
ti
f
y
i
n
g
n
e
w
o
r
u
n
f
a
m
iliar
th
r
ea
ts
.
L
ee
et
a
l.
[
1
3
]
p
r
esen
ted
a
li
g
h
t
w
eig
h
t
I
DS
w
h
er
e
d
if
f
er
en
t
en
er
g
y
co
n
s
u
m
p
tio
n
ca
n
b
e
tak
en
i
n
to
ac
co
u
n
t
i
n
in
tr
u
s
io
n
d
etec
tio
n
.
T
h
ey
co
n
s
id
er
ea
ch
n
o
d
e
in
d
iv
id
u
all
y
,
en
er
g
y
co
n
s
u
m
p
tio
n
b
ein
g
a
m
aj
o
r
cr
iter
io
n
.
Su
ch
a
tech
n
iq
u
e
f
u
n
ctio
n
s
w
ell
in
t
h
e
ca
s
e
o
f
h
o
m
o
g
e
n
eo
u
s
n
et
w
o
r
k
s
s
u
c
h
as
W
SNs
,
b
u
t
p
er
f
o
r
m
s
p
o
o
r
ly
f
o
r
I
o
T
n
et
w
o
r
k
s
w
h
e
r
e
n
o
d
es
ex
p
er
ien
ce
w
id
el
y
v
ar
y
in
g
p
o
w
er
co
n
s
u
m
p
tio
n
b
eh
av
io
r
.
Kr
i
m
m
l
in
g
an
d
P
eter
[
1
4
]
p
r
o
p
o
s
ed
a
c
o
n
s
tr
ain
ed
ap
p
licatio
n
p
r
o
to
co
l
(
C
o
A
P
)
m
o
d
u
lar
d
etec
tio
n
f
r
a
m
e
w
o
r
k
o
n
t
h
e
ap
p
licatio
n
lev
el
f
o
r
th
e
I
o
T
n
et
w
o
r
k
s
in
s
m
ar
t
cities.
A
s
lig
h
t
w
ei
g
h
t
as
th
eir
m
eth
o
d
is
,
it
o
n
ly
d
ef
e
n
d
s
ag
ain
s
t
ce
r
tai
n
attac
k
s
,
s
u
c
h
as
r
o
u
tin
g
attac
k
s
.
T
h
e
i
m
p
le
m
en
ta
tio
n
o
f
in
te
g
r
ated
,
s
ig
n
a
tu
r
e
-
b
ased
,
an
d
an
o
n
y
m
i
ze
d
I
DS
m
a
y
lead
t
o
in
cr
ea
s
ed
d
etec
tio
n
ca
p
ab
ilit
ies,
th
e
y
s
u
g
g
est.
C
er
v
an
te
s
et
a
l.
[
1
5
]
p
r
esen
ted
in
tr
u
s
io
n
d
etec
tio
n
o
f
s
i
n
k
h
o
l
e
attac
k
s
in
I
o
T
(
I
N
T
I
)
,
an
I
D
S
ad
d
r
ess
in
g
t
h
e
d
etec
tio
n
o
f
s
in
k
h
o
le
attac
k
s
i
n
I
P
v
6
o
v
er
lo
w
-
p
o
w
er
w
ir
ele
s
s
p
er
s
o
n
al
ar
ea
n
et
w
o
r
k
s
(
6
L
o
W
P
A
N
)
-
b
ased
I
o
T
n
et
w
o
r
k
s
.
T
h
e
y
ad
o
p
t
a
r
ep
u
tatio
n
-
d
r
iv
e
n
ap
p
r
o
ac
h
;
th
at
is
,
n
o
d
es
o
b
s
er
v
e
n
o
d
es
’
tr
a
f
f
ic
an
d
g
o
s
s
ip
in
th
e
n
et
w
o
r
k
to
in
f
o
r
m
o
th
er
s
i
f
t
h
e
y
d
etec
t
a
m
alicio
u
s
n
o
d
e.
Ho
w
e
v
er
,
th
eir
s
y
s
te
m
d
o
es
n
o
t
co
v
er
h
o
w
it
af
f
ec
t
s
lo
w
-
ca
p
ac
it
y
n
o
d
es,
w
h
ic
h
co
u
ld
b
e
s
ca
r
ce
in
a
r
eso
u
r
ce
-
co
n
s
tr
ain
ed
en
v
ir
o
n
m
e
n
t.
Mid
i
et
a
l.
[
1
6
]
in
tr
o
d
u
ce
d
Kalis,
an
I
D
S
th
a
t
in
te
g
r
ates
an
o
m
al
y
-
b
ased
an
d
s
ig
n
a
tu
r
e
-
b
ased
d
etec
tio
n
m
e
th
o
d
s
.
Kalis
s
t
u
d
ies
n
et
w
o
r
k
to
p
o
lo
g
y
an
d
tr
af
f
ic
to
d
ef
en
d
a
g
ai
n
s
t
Do
S
attac
k
s
.
B
u
t
it
i
s
r
estricte
d
to
r
o
u
tin
g
attac
k
s
an
d
r
eq
u
ir
es
o
f
f
-
r
ac
k
d
etec
tio
n
m
o
d
u
les
f
o
r
d
if
f
er
en
t
t
y
p
es
o
f
attac
k
p
atter
n
s
,
w
h
i
ch
ca
n
co
m
p
li
ca
te
th
e
s
y
s
te
m
.
E
lr
a
w
y
et
a
l.
[
1
7
]
,
as
w
ell
as
C
h
aa
b
o
u
n
i
et
a
l.
[
1
8
]
,
p
r
o
v
id
ed
a
co
m
p
r
eh
e
n
s
i
v
e
r
ev
ie
w
o
f
I
DS
tech
n
o
lo
g
ie
s
o
n
I
o
T
an
d
th
e
d
if
f
er
en
t
s
ec
u
r
it
y
i
s
s
u
es
f
ac
ed
b
y
I
o
T
n
et
w
o
r
k
s
.
As
h
ar
d
w
ar
e
r
eso
u
r
ce
s
o
n
I
o
T
d
ev
ices
ar
e
li
m
ited
,
L
i
e
t
a
l.
[
1
9
]
s
tu
d
ied
th
e
ap
p
licab
ilit
y
o
f
u
s
in
g
s
tatis
tical
m
e
th
o
d
s
f
o
r
I
o
T
I
DSs
an
d
e
m
p
h
a
s
ized
th
e
n
ec
ess
it
y
o
f
ca
r
ef
u
ll
y
ch
o
o
s
in
g
t
h
e
s
tati
s
tical
m
et
h
o
d
s
.
2.
M
E
T
H
O
D
I
n
th
i
s
s
ec
t
io
n
,
w
e
d
escr
ib
e
a
s
o
lid
th
eo
r
etica
l
b
ase
b
u
il
t
o
n
th
eo
r
etica
l
co
n
ce
p
t
s
in
to
p
r
ac
tical
i
m
p
le
m
en
ta
tio
n
s
c
h
e
m
e
s
to
b
u
ild
th
e
h
y
b
r
id
I
DS.
T
h
e
m
et
h
o
d
o
lo
g
y
p
r
esen
ted
in
Fi
g
u
r
e
1
tr
ies
to
tack
le
th
e
s
p
ec
if
ic
s
ec
u
r
it
y
ch
al
len
g
es
o
f
th
e
I
o
T
ar
ch
itectu
r
e
in
a
m
an
n
er
t
h
at
is
r
ep
ea
tab
le
w
it
h
in
n
o
v
ati
v
e
ch
o
ice
s
ab
o
u
t
ar
ch
itectu
r
es
a
n
d
o
p
tim
izatio
n
m
et
h
o
d
s
.
T
h
e
p
r
o
p
o
s
ed
h
y
b
r
id
I
DS
co
m
b
i
n
es
Q
A
E
f
o
r
clas
s
if
icatio
n
w
it
h
Kits
u
n
e
to
id
e
n
ti
f
y
an
o
m
alie
s
.
T
h
e
w
o
r
k
f
lo
w
co
n
s
is
t
s
o
f
f
o
u
r
m
a
in
s
ta
g
es
a
s
d
ep
i
cted
in
F
ig
u
r
e
1
:
(
i)
d
ata
co
llectio
n
an
d
p
r
e
p
ar
atio
n
,
(
ii)
f
ea
tu
r
e
en
g
in
ee
r
i
n
g
a
n
d
d
i
m
en
s
io
n
al
it
y
r
ed
u
ctio
n
,
(
iii)
m
o
d
el
ar
ch
itec
tu
r
e
d
esig
n
(
Kits
u
n
e
+
Q
A
E
)
,
a
n
d
(
iv
)
tr
ain
i
n
g
a
n
d
ev
al
u
atio
n
.
T
w
o
w
id
el
y
u
s
ed
d
ataset
s
,
n
a
m
e
l
y
“
B
o
t
-
I
o
T
”
an
d
“
RT
-
I
o
T
(
2
0
2
2
)
”
,
w
er
e
e
m
p
l
o
y
ed
i
n
th
is
s
tu
d
y
.
T
h
e
d
ata
s
ets
ar
e
w
id
el
y
u
s
ed
b
ec
au
s
e
th
e
y
co
n
s
i
s
t
o
f
a
lab
eled
lar
g
e
-
s
ca
le
co
llectio
n
w
it
h
d
iv
er
s
e
attac
k
t
y
p
es
,
i
n
cl
u
d
in
g
DDo
S,
d
ata
ex
f
iltra
tio
n
,
an
d
k
e
y
lo
g
g
in
g
.
2
.
1
.
Da
t
a
c
o
llect
io
n
a
nd
prepro
ce
s
s
ing
T
h
e
B
o
t
-
I
o
T
d
a
t
as
e
t
w
as
d
ev
e
lo
p
e
d
b
y
t
h
e
Un
iv
e
r
s
ity
o
f
New
S
o
u
th
W
al
e
s
(
U
NS
W
)
t
o
b
r
i
d
g
e
th
e
g
a
p
o
f
r
e
a
l
-
l
ik
e
I
o
T
b
o
t
n
e
t
a
t
t
a
ck
d
a
t
a
f
o
r
in
t
r
u
s
i
o
n
d
e
t
e
ct
i
o
n
im
p
r
o
v
em
en
ts
.
T
h
e
2
0
2
2
m
o
d
e
l
i
s
a
m
o
r
e
s
o
p
h
is
ti
c
a
t
ed
v
e
r
s
i
o
n
,
w
h
i
ch
in
c
lu
d
e
s
f
u
r
t
h
e
r
a
t
t
a
ck
s
im
u
la
t
i
o
n
s
an
d
im
p
r
o
v
e
d
f
e
atu
r
e
en
g
in
e
e
r
in
g
[
2
0
]
.
T
ab
le
1
s
u
m
m
ar
ize
s
th
e
k
e
y
d
i
f
f
er
e
n
ce
s
b
et
w
ee
n
t
h
e
B
o
t
-
I
o
T
an
d
R
T
-
I
o
T
d
atasets
in
ter
m
s
o
f
p
u
r
p
o
s
e,
d
ata
ch
ar
ac
ter
is
tics
,
attac
k
t
y
p
es,
a
n
d
d
ep
lo
y
m
e
n
t
s
u
itab
i
lit
y
.
T
h
is
co
m
p
ar
is
o
n
c
lar
if
ie
s
th
e
r
atio
n
ale
f
o
r
u
s
i
n
g
b
o
th
d
atasets
to
e
v
al
u
ate
th
e
p
r
o
p
o
s
ed
h
y
b
r
id
Kits
u
n
e
–
QA
E
i
n
tr
u
s
io
n
d
etec
tio
n
s
y
s
te
m
u
n
d
er
o
f
f
lin
e
a
n
d
r
ea
l
-
ti
m
e
s
ce
n
ar
io
s
.
T
h
e
RT
-
I
o
T
d
ataset
is
d
esig
n
ed
f
o
r
r
ea
l
-
ti
m
e
in
tr
u
s
io
n
d
etec
tio
n
o
f
I
o
T
n
et
w
o
r
k
s
.
RT
-
I
o
T
also
h
ig
h
li
g
h
ts
t
h
e
s
ig
n
i
f
ica
n
ce
o
f
s
tr
ea
m
i
n
g
d
ata
,
w
h
er
e
th
e
n
u
ll
v
alu
e
s
w
o
u
ld
d
eg
r
ad
e
t
h
e
p
er
f
o
r
m
a
n
ce
a
n
d
ef
f
icien
c
y
o
f
a
n
y
ML
m
o
d
el.
T
h
u
s
,
a
k
e
y
co
m
p
o
n
en
t
o
f
ML
p
r
ep
r
o
ce
s
s
in
g
is
a
n
i
m
p
u
ter
to
r
e
m
o
v
e
n
u
l
l
v
alu
e
s
f
r
o
m
th
e
d
ataset.
T
h
e
i
n
d
iv
id
u
al
d
a
tasets
ar
e
w
ell
ela
b
o
r
ated
in
b
o
th
t
w
o
d
ataset
B
o
t
-
I
o
T
an
d
R
T
-
I
o
T
d
ata
s
ets.
T
h
u
s
,
i
t
co
n
tai
n
s
b
ar
el
y
an
y
n
u
ll
v
alu
e
s
.
B
u
t
t
h
e
p
r
ev
io
u
s
o
n
e
h
ad
t
h
e
n
u
ll.
I
t
is
s
u
itab
le
f
o
r
ed
g
e
co
m
p
u
ti
n
g
a
n
d
r
ea
l
-
ti
m
e
an
o
m
al
y
d
etec
tio
n
[
2
1
]
.
T
h
ese
n
o
t
-
a
-
n
u
m
b
er
(
NaN
)
v
alu
e
s
d
is
t
u
r
b
th
e
ac
cu
r
ac
y
an
d
ef
f
icien
c
y
o
f
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
an
y
ML
al
g
o
r
ith
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Hyb
r
id
in
tr
u
s
io
n
d
etec
tio
n
in
I
o
T d
ev
ices:
a
d
ee
p
lea
r
n
in
g
a
p
p
r
o
a
ch
u
s
in
g
…
(
Md
.
R
ifa
t E
N
o
o
r
)
455
Fig
u
r
e
1
.
R
esear
ch
m
et
h
o
d
o
lo
g
y
T
ab
le
1
.
Data
s
et
co
m
p
ar
is
o
n
s
A
sp
e
c
t
B
o
t
-
I
o
T
RT
-
I
o
T
P
u
r
p
o
se
O
f
f
l
i
n
e
M
L
t
r
a
i
n
i
n
g
R
e
a
l
-
t
i
me
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
D
a
t
a
v
o
l
u
me
L
a
r
g
e
(
7
2
M
+
r
e
c
o
r
d
s)
S
mal
l
e
r
,
o
p
t
i
mi
z
e
d
f
o
r
st
r
e
a
mi
n
g
A
t
t
a
c
k
t
y
p
e
s
D
D
o
S
,
k
e
y
l
o
g
g
i
n
g
,
d
a
t
a
t
h
e
f
t
M
i
r
a
i
,
b
r
u
t
e
-
f
o
r
c
e
,
sp
o
o
f
i
n
g
F
o
r
mat
C
S
V
/
P
C
A
P
K
a
f
k
a
,
M
Q
T
T
,
JS
O
N
B
e
st
f
o
r
R
e
se
a
r
c
h
,
mo
d
e
l
b
e
n
c
h
mar
k
i
n
g
Ed
g
e
A
I
,
l
i
v
e
mo
n
i
t
o
r
i
n
g
Hen
ce
,
o
n
e
o
f
th
e
m
o
s
t
i
m
p
o
r
t
an
t
p
r
e
-
r
ea
d
in
g
p
r
o
ce
d
u
r
es
in
ML
ap
p
ea
r
s
to
b
e
m
is
s
in
g
n
es
s
.
B
o
th
th
e
B
o
t
-
I
o
T
an
d
R
T
-
I
o
T
d
atasets
ar
e
w
ell
-
p
r
ep
ar
ed
d
atasets
.
T
h
er
ef
o
r
e,
th
er
e
ar
e
to
o
f
e
w
m
is
s
i
n
g
v
al
u
es.
T
h
e
is
s
u
es
r
aised
i
n
t
h
e
i
n
tr
o
d
u
ctio
n
ar
e
d
ir
ec
tl
y
f
ac
ed
b
y
t
h
e
p
r
ep
r
o
ce
s
s
in
g
tech
n
iq
u
e
s
,
s
u
c
h
a
s
P
C
A
an
d
o
n
e
-
h
o
t
en
co
d
in
g
(
e.
g
.
,
p
r
o
to
co
l
ty
p
e,
s
er
v
ice,
s
tate
,
attac
k
ca
teg
o
r
y
)
,
an
d
t
h
e
r
esp
o
n
s
e
o
f
t
h
e
n
et
w
o
r
k
to
th
o
s
e
alter
n
ati
v
es
is
r
ep
o
r
ted
.
P
C
A
r
ed
u
ce
s
t
h
e
d
i
m
e
n
s
io
n
ali
t
y
o
f
t
h
e
d
ataset
w
h
i
le
m
ain
tai
n
i
n
g
cr
itica
l
ch
ar
ac
ter
is
tic
s
th
er
eo
f
,
w
h
ic
h
is
ess
en
tial
f
o
r
I
o
T
-
co
n
s
tr
ain
e
d
d
ev
ices.
Fu
r
th
er
m
o
r
e,
o
n
e
-
h
o
t
en
co
d
in
g
en
s
u
r
es
an
ac
c
u
r
ate
a
n
d
co
m
p
lete
r
ep
r
esen
tatio
n
o
f
t
h
e
ca
te
g
o
r
ical
in
f
o
r
m
atio
n
,
w
h
ic
h
ca
n
al
s
o
h
elp
t
h
e
d
etec
tio
n
m
o
d
el
to
ca
p
tu
r
e
d
if
f
er
e
n
t
t
y
p
es
o
f
attac
k
s
,
in
c
lu
d
i
n
g
t
h
o
s
e
th
at
ar
e
ze
r
o
-
d
ay
.
T
h
ese
ar
e
n
ee
d
ed
s
tatem
e
n
t
s
to
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
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
.
2
,
A
p
r
il
20
26
:
4
5
2
-
465
456
en
s
u
r
e
th
at
t
h
e
h
y
b
r
id
K
it
s
u
n
e
-
Q
A
E
ar
ch
itect
u
r
e
ca
n
s
ca
le
o
u
t
ac
r
o
s
s
m
u
ltip
le
I
o
T
n
et
w
o
r
k
s
as
w
ell
a
s
p
r
o
v
id
e
r
ea
l
-
ti
m
e
an
al
y
s
is
.
H
o
w
e
v
er
,
th
e
p
r
ev
io
u
s
v
er
s
io
n
co
n
tain
ed
n
u
ll
v
alu
e
s
.
A
n
d
s
o
m
a
n
y
a
ttack
s
s
h
o
w
in
T
ab
le
s
2
an
d
3
,
in
th
e
b
alan
ce
s
et
,
an
d
w
ith
o
u
t
t
h
e
b
alan
ce
s
et.
T
ab
le
2
.
Dif
f
er
en
t a
ttac
k
t
y
p
e
s
in
th
e
RT
-
I
o
T
d
ataset
b
ef
o
r
e
an
d
af
ter
b
alan
ci
n
g
A
t
t
a
c
k
t
y
p
e
B
e
f
o
r
e
b
a
l
a
n
c
i
n
g
c
o
u
n
t
A
f
t
e
r
b
a
l
a
n
c
i
n
g
c
o
u
n
t
D
O
S
_
S
Y
N
_
H
P
I
N
G
9
4
6
5
9
9
4
6
5
9
T
H
I
N
G
_
S
P
E
A
K
8
1
0
8
9
4
6
5
9
A
R
P
_
P
O
I
S
I
O
N
I
N
G
7
7
5
0
9
4
6
5
9
M
Q
T
T
_
P
U
B
L
I
S
H
4
1
4
6
9
4
6
5
9
N
M
A
P
_
U
D
P
_
S
C
A
N
2
5
9
0
9
4
6
5
9
N
M
A
P
_
X
M
A
S
_
T
R
EE_
S
C
A
N
2
0
1
0
9
4
6
5
9
N
M
A
P
_
O
S
_
D
ET
EC
TI
O
N
2
0
0
0
9
4
6
5
9
N
M
A
P
_
T
C
P
_
S
C
A
N
1
0
0
2
9
4
6
5
9
D
D
O
S
_
S
L
O
W
L
O
R
I
S
5
3
4
9
4
6
5
9
W
I
P
R
O
_
B
U
L
B
2
5
3
9
4
6
5
9
M
ET
A
S
P
L
O
I
T
_
B
R
U
T
E_
F
O
R
C
E_
S
S
H
37
9
4
6
5
9
N
M
A
P
_
F
I
N
_
S
C
A
N
28
9
4
6
5
9
T
ab
le
3
.
Dif
f
er
en
t a
ttac
k
t
y
p
e
s
in
th
e
B
o
t
-
I
o
T
d
ataset
b
ef
o
r
e
an
d
af
ter
b
alan
ci
n
g
A
t
t
a
c
k
t
y
p
e
B
e
f
o
r
e
b
a
l
a
n
c
i
n
g
c
o
u
n
t
A
f
t
e
r
b
a
l
a
n
c
i
n
g
c
o
u
n
t
D
D
o
S
3
4
8
3
9
6
3
4
8
3
9
6
D
o
S
3
3
0
1
1
2
3
4
8
3
9
6
R
e
c
o
n
n
a
i
ss
a
n
c
e
1
8
1
6
3
3
4
8
3
9
6
N
o
r
mal
1
0
7
3
4
8
3
9
6
T
h
e
f
t
14
3
4
8
3
9
6
2
.
1
.
1
.
L
a
bel
ba
la
nc
ing
I
m
b
alan
ce
d
d
atasets
i
s
s
u
es
ar
e
n
atu
r
al
s
i
n
ce
t
h
e
r
ea
l
w
o
r
ld
is
f
u
ll
o
f
i
m
b
ala
n
ce
d
in
s
tan
c
es,
w
h
ic
h
m
ak
e
s
it
d
if
f
ic
u
lt
f
o
r
p
r
ed
ictiv
e
m
o
d
elli
n
g
.
I
n
t
h
e
o
v
er
s
a
m
p
lin
g
o
f
t
h
e
m
i
n
o
r
it
y
cla
s
s
,
f
u
r
th
er
s
a
m
p
le
s
ar
e
d
r
a
w
n
u
n
til
th
e
s
et
o
f
att
r
ib
u
t
es
is
co
n
s
i
s
te
n
t
w
i
th
th
e
m
aj
o
r
it
y
cla
s
s
.
T
o
en
s
u
r
e
r
esto
r
ati
o
n
o
f
t
h
e
e
f
f
ec
t
o
f
u
n
b
ala
n
ce
d
s
a
m
p
li
n
g
o
n
an
u
n
d
er
-
r
ep
r
esen
ted
class
is
k
n
o
wn
to
in
tr
o
d
u
ce
b
ias
in
to
t
h
e
lea
r
n
ed
m
o
d
el
,
an
d
to
en
s
u
r
e
t
h
at
t
h
e
lear
n
ed
m
o
d
el
is
e
x
p
o
s
ed
to
an
eq
u
al
n
u
m
b
e
r
o
f
ex
a
m
p
le
s
f
r
o
m
ea
c
h
ca
te
g
o
r
y
,
M
L
tea
m
s
u
s
e
r
ep
licatio
n
m
et
h
o
d
s
o
n
th
e
s
a
m
p
les o
f
t
h
e
u
n
d
er
-
r
ep
r
ese
n
te
d
class
.
Ov
er
-
s
a
m
p
l
in
g
is
an
ef
f
ec
ti
v
e
w
a
y
to
h
an
d
le
an
i
m
b
ala
n
ce
d
p
atter
n
w
h
en
n
o
d
ata
is
r
e
m
o
v
ed
,
an
d
h
en
ce
i
t
is
a
f
av
o
r
ab
le
o
p
tio
n
f
o
r
th
e
M
L
tea
m
s
w
h
ile
h
a
n
d
l
in
g
th
e
p
r
o
b
le
m
o
f
a
s
ca
n
t
y
d
a
taset
i
n
s
u
ch
a
w
a
y
th
at
t
h
e
y
w
o
u
ld
lo
s
e
v
er
y
lit
tle
in
f
o
r
m
a
tio
n
[
2
2
]
.
Ho
w
e
v
er
,
th
e
m
o
d
el
is
s
til
l
v
u
l
n
er
a
b
le
to
d
is
co
v
er
in
g
p
atter
n
s
i
n
th
e
d
ata
a
n
d
o
v
er
f
i
ttin
g
as
a
r
es
u
lt
o
f
th
e
f
ac
t
t
h
at
th
e
n
u
m
b
er
o
f
u
n
iq
u
e
d
ata
p
o
in
t
s
i
n
th
e
s
m
a
ll
m
i
n
o
r
it
y
cla
s
s
is
s
m
a
ll.
T
h
is
s
t
u
d
y
e
m
p
lo
y
ed
o
v
er
s
a
m
p
li
n
g
t
ec
h
n
iq
u
es.
Af
ter
b
ala
n
ci
n
g
o
f
lab
el
s
,
th
e
co
u
n
t a
p
t
y
p
es
f
o
r
b
o
th
d
atasets
ar
e
g
iv
en
in
T
ab
les 2
an
d
3
.
2
.
2
.
F
e
a
t
ure
eng
ineering
a
nd
di
m
en
s
io
na
lity
re
du
ct
io
n
T
h
e
w
ell
-
k
n
o
w
n
u
n
s
u
p
er
v
i
s
ed
lear
n
in
g
m
et
h
o
d
f
o
r
r
ed
u
cin
g
d
ata
d
i
m
en
s
io
n
i
s
P
C
A
to
d
a
ta
[
2
3
]
.
I
t
n
o
t
o
n
l
y
i
n
cr
ea
s
es
i
n
ter
p
r
etab
ilit
y
b
u
t
also
r
ed
u
ce
s
t
h
e
lo
s
s
o
f
in
f
o
r
m
atio
n
.
I
t
h
elp
s
y
o
u
to
ea
s
il
y
d
is
co
v
er
th
e
m
o
s
t
s
i
g
n
i
f
ican
t
f
ea
t
u
r
es
in
y
o
u
r
d
ataset.
T
h
e
P
C
A
m
et
h
o
d
is
b
ased
h
ea
v
il
y
o
n
t
h
e
co
v
ar
ian
ce
m
atr
ix
i
n
d
eter
m
in
i
n
g
t
h
e
p
r
in
c
ip
al
co
m
p
o
n
en
ts
o
f
t
h
e
d
ata.
T
h
e
ei
g
en
v
ec
to
r
s
i
llu
s
tr
a
te
t
h
e
m
o
s
t
i
m
p
o
r
tan
t
m
o
d
es
o
f
v
ar
iatio
n
w
it
h
i
n
th
e
d
ata
(
i.e
.
,
th
e
an
g
le
s
o
f
r
o
tatio
n
)
,
w
h
ile
th
e
eig
en
v
al
u
es
i
n
d
icate
th
e
a
m
o
u
n
t
o
f
v
ar
iatio
n
alo
n
g
ea
ch
d
ir
ec
tio
n
.
T
h
e
co
v
ar
ian
ce
m
atr
i
x
is
a
n
n
x
n
m
atr
ix
th
at
p
r
o
v
id
es
th
e
co
v
ar
ia
n
c
e
b
et
w
ee
n
ea
ch
p
air
of
ele
m
e
n
t
s
in
t
h
e
d
ata.
T
h
e
co
v
ar
ian
ce
m
atr
ix
,
w
h
ic
h
is
a
n
n
x
n
m
a
tr
ix
,
ca
lc
u
lates
th
e
p
air
w
i
s
e
co
v
ar
ian
ce
b
et
w
ee
n
th
e
e
le
m
e
n
t
s
in
t
h
e
d
ata.
Giv
e
n
a
d
ata
m
atr
i
x
w
it
h
o
b
s
er
v
atio
n
s
f
o
r
n
v
ar
iab
les,
th
e
co
r
r
elatio
n
m
atr
i
x
is
d
ef
i
n
ed
as f
o
llo
w
s
:
=
(
1
)
∗
(
1
)
T
h
e
eig
en
v
ec
to
r
s
ar
e
t
h
e
n
u
s
ed
to
ca
lcu
la
te
t
h
e
f
ea
tu
r
e
s
o
f
t
h
e
d
ata.
T
h
e
ei
g
en
v
ec
t
o
r
s
o
f
t
h
e
co
v
ar
ian
ce
d
ata
m
a
tr
ix
ca
p
t
u
r
e
th
e
lar
g
es
t
v
ar
iatio
n
s
o
f
th
e
d
ata.
T
h
ese
co
o
r
d
in
ates
ar
e
th
en
e
m
p
lo
y
ed
to
d
ef
in
e
t
h
e
n
e
w
r
ef
er
e
n
ce
f
r
a
m
e
i
n
w
h
ich
t
h
e
d
ata
is
p
r
o
ce
s
s
ed
.
T
h
ese
v
ec
to
r
s
ar
e
e
m
p
lo
y
ed
to
p
r
o
j
ec
t
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Hyb
r
id
in
tr
u
s
io
n
d
etec
tio
n
in
I
o
T d
ev
ices:
a
d
ee
p
lea
r
n
in
g
a
p
p
r
o
a
ch
u
s
in
g
…
(
Md
.
R
ifa
t E
N
o
o
r
)
457
o
r
ig
in
al
d
ata
i
n
to
a
lo
w
er
-
d
i
m
en
s
io
n
a
l
s
p
ac
e
a
n
d
d
escr
ib
e
th
e
d
ir
ec
tio
n
s
ar
o
u
n
d
w
h
ich
th
e
d
ata
v
ar
ies
t
h
e
m
o
s
t.
T
h
e
eig
e
n
v
ec
to
r
s
ar
e
co
m
p
u
ted
i
n
s
u
ch
a
w
a
y
th
a
t
(
2
)
is
s
ati
s
f
ied
:
=
(
2
)
w
h
er
e,
=
c
o
v
ar
ian
ce
m
atr
i
x
=
eig
en
v
ec
to
r
=
ass
o
ciate
eig
e
n
v
a
lu
e
We
p
er
f
o
r
m
o
n
e
-
h
o
t
e
n
co
d
in
g
o
n
all
th
e
ca
teg
o
r
ical
attr
ib
u
tes
o
f
d
ataset
s
as
p
ar
t
o
f
th
e
d
ata
p
r
e
-
p
r
o
ce
s
s
in
g
to
e
n
s
u
r
e
t
h
at
M
L
m
o
d
els
ca
n
e
f
f
ec
ti
v
el
y
p
r
o
ce
s
s
t
h
e
m
.
T
h
e
ca
te
g
o
r
y
v
ar
iab
les
p
r
esen
ted
w
er
e
o
n
e
-
h
o
t
en
co
d
ed
as
s
h
o
w
n
f
o
r
th
e
B
o
t
-
I
o
T
d
ataset:
p
r
o
to
c
o
l
(
T
C
P,
UDP
,
an
d
I
C
MP
)
,
s
er
v
ice
(
f
o
r
ex
a
m
p
le,
SMT
P
,
FTP
,
an
d
H
T
T
P
)
,
f
la
g
(
SF
an
d
R
E
J
)
,
attac
k
m
e
th
o
d
(
k
e
y
lo
g
g
er
s
,
d
d
o
s
,
an
d
d
ata
th
ef
t
)
.
T
h
e
ca
teg
o
r
ical
v
ar
iab
les
o
f
th
e
R
T
-
I
o
T
d
ataset
w
er
e
as
f
o
l
lo
w
s
:
p
r
o
to
co
l
(
T
C
P
,
UDP
,
an
d
MQ
T
T
)
,
a
ttack
t
y
p
e
(
eg
.
,
Sp
o
o
f
in
g
,
B
r
u
te
-
Fo
r
ce
,
an
d
Mir
ai)
.
A
s
er
ies
o
f
b
in
ar
y
d
u
m
m
y
co
l
u
m
n
s
w
a
s
g
en
er
ated
f
r
o
m
ea
ch
o
f
th
e
s
e
ca
teg
o
r
y
v
ar
iab
les.
Fo
r
ex
a
m
p
le,
f
r
o
m
th
e
“
P
r
o
to
c
o
l
T
y
p
e
”
f
ea
t
u
r
e
th
at
h
ad
3
in
itial
v
al
u
es
(
T
C
P,
UDP
,
an
d
MQ
T
T
)
,
an
d
o
b
tain
ed
3
b
i
n
ar
y
f
e
atu
r
es:
“
P
r
o
to
co
l_
T
C
P
”
,
“
P
r
o
to
co
l_
UDP
”
,
an
d
“
P
r
o
to
co
l_
MQ
T
T
”
.
T
h
is
o
n
e
-
h
o
t
en
co
d
in
g
s
tep
co
n
v
er
ted
th
ese
ca
te
g
o
r
y
f
ea
t
u
r
es
i
n
to
n
u
m
er
ical
f
ea
t
u
r
es
s
o
th
at
M
L
m
o
d
els
ca
n
i
n
ter
p
r
et.
T
h
is
tr
an
s
i
tio
n
led
to
th
e
f
o
ll
o
w
i
n
g
s
ig
n
i
f
ica
n
t
o
u
tco
m
e
s
:
S
in
ce
all
ca
teg
o
r
ical
f
ea
t
u
r
es
w
er
e
o
n
e
-
h
o
t
en
co
d
ed
,
th
e
1
8
f
ea
tu
r
es
f
o
r
th
e
B
o
t
-
I
o
T
d
ataset
ex
p
an
d
ed
to
8
4
f
ea
tu
r
es.
A
l
s
o
,
th
e
RT
-
I
o
T
d
ataset,
in
itiall
y
co
n
tain
i
n
g
o
n
l
y
1
8
f
ea
t
u
r
es,
w
a
s
ex
p
an
d
ed
to
8
4
f
ea
tu
r
es
d
u
e
to
th
e
o
n
e
-
h
o
t
en
co
d
in
g
.
T
h
en
p
er
f
o
r
m
ed
a
P
C
A
w
as
p
er
f
o
r
m
ed
to
r
ed
u
ce
th
e
d
i
m
en
s
io
n
o
f
t
h
e
s
et
o
f
f
ea
tu
r
e
s
af
te
r
o
n
e
-
h
o
t
en
co
d
in
g
.
P
C
A
r
etai
n
ed
9
5
%
th
e
v
ar
ian
ce
,
r
es
u
lti
n
g
in
1
8
f
ea
t
u
r
es
f
o
r
B
o
t
-
I
o
T
an
d
1
1
f
ea
tu
r
e
s
f
o
r
R
T
-
I
o
T
in
s
tead
o
f
8
4
an
d
8
4
f
ea
tu
r
es,
r
esp
ec
tiv
el
y
.
B
y
p
r
eser
v
i
n
g
i
m
p
o
r
ta
n
t
in
f
o
r
m
a
tio
n
f
o
r
in
tr
u
s
io
n
d
etec
tio
n
,
th
e
d
i
m
en
s
io
n
al
it
y
r
ed
u
ct
io
n
s
tr
at
eg
y
i
m
p
r
o
v
es
m
o
d
el
p
er
f
o
r
m
a
n
ce
an
d
r
ed
u
ce
s
co
m
p
u
tatio
n
a
l c
o
m
p
le
x
it
y
.
2
.
3
.
Arc
hite
ct
ure
m
o
d
e
lin
g
2
.
3
.
1
.
K
it
s
un
e
(
det
ec
t
io
n
o
f
a
no
m
a
lie
s
)
Kits
u
n
e
is
a
n
e
w
p
atter
n
d
is
c
o
v
er
y
m
e
th
o
d
,
w
h
ic
h
is
d
esig
n
ed
s
p
ec
if
ica
ll
y
f
o
r
I
o
T
en
v
ir
o
n
m
e
n
ts
.
It
u
tili
ze
s
n
u
m
er
o
u
s
au
to
e
n
co
d
er
s
f
o
r
an
o
m
al
y
d
etec
tio
n
in
n
et
w
o
r
k
tr
a
f
f
ic.
E
ac
h
au
to
en
co
d
er
in
th
e
en
s
e
m
b
le
is
tr
ai
n
ed
on
a
s
u
b
s
et
of
t
h
e
d
ata,
th
u
s
e
n
ab
lin
g
th
e
s
y
s
te
m
to
r
ec
o
g
n
ize
s
ev
er
al
p
atter
n
s
[
2
4
]
.
Kits
u
n
e
is
a
s
y
s
te
m
f
o
r
o
n
li
n
e
lear
n
i
n
g
u
s
i
n
g
an
en
s
e
m
b
le
m
et
h
o
d
to
d
et
ec
t
n
et
w
o
r
k
i
n
tr
u
s
io
n
s
in
r
ea
l
-
ti
m
e.
Fo
r
co
n
s
i
s
ten
t
tr
ain
i
n
g
,
it
o
p
er
ates
on
s
ta
n
d
ar
d
ize
d
,
n
o
r
m
a
lized
,
p
r
ep
r
o
ce
s
s
ed
n
et
w
o
r
k
tr
a
f
f
ic
f
ea
t
u
r
es,
s
u
c
h
as
p
ac
k
et
ti
m
i
n
g
,
th
e
p
r
o
to
co
l
h
ea
d
er
s
,
an
d
s
tatis
tical
f
ea
tu
r
e
s
.
T
h
e
m
o
d
el
u
ti
lizes
a
ch
ai
n
of
s
ev
er
al
lig
h
t
-
we
i
g
h
t
au
to
en
co
d
er
s
,
w
h
er
e
each
o
n
e
is
tr
ain
ed
on
a
f
ea
tu
r
e
s
u
b
s
et
a
n
d
u
s
ed
to
lear
n
b
en
ig
n
b
eh
a
v
i
o
r
s
by
m
i
n
i
m
izi
n
g
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
o
v
er
b
en
ig
n
tr
a
f
f
ic.
Kits
u
n
e
o
p
er
ates
i
n
a
n
o
n
lin
e
f
as
h
io
n
an
d
co
n
ti
n
u
o
u
s
l
y
ad
ap
ts
its
a
u
to
e
n
co
d
er
s
to
n
e
w
d
ev
ices
o
r
s
er
v
ices,
a
s
w
ell
as
a
s
lo
w
l
y
ev
o
l
v
i
n
g
n
et
w
o
r
k
b
eh
a
v
io
r
.
Du
r
in
g
d
etec
tio
n
,
a
s
co
r
e
f
o
r
ea
ch
r
ec
ei
v
ed
in
f
o
r
m
atio
n
i
s
g
en
er
ated
,
w
h
i
ch
is
m
ea
s
u
r
ed
b
y
th
e
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
;
i.e
.
,
h
ig
h
er
th
e
s
co
r
es,
th
e
m
o
r
e
s
u
s
p
icio
u
s
aler
t
s
ar
e
d
etec
ted
.
Kits
u
n
e
is
f
le
x
i
b
le
an
d
r
o
b
u
s
t
b
ec
au
s
e
o
f
its
u
n
s
u
p
er
v
is
ed
n
atu
r
e
to
c
h
an
g
i
n
g
n
et
w
o
r
k
e
n
v
ir
o
n
m
en
t
s
,
an
d
ca
n
d
etec
t
in
f
ec
t
io
n
s
w
ith
o
u
t
th
e
n
ee
d
f
o
r
lab
eled
attac
k
d
ata.
T
h
e
m
ai
n
ch
ar
ac
ter
is
tic
s
ar
e
th
e
r
ea
l
-
ti
m
e
p
r
o
ce
s
s
in
g
,
t
h
e
r
o
b
u
s
t
n
es
s
o
f
co
llectiv
e
lear
n
i
n
g
,
an
d
t
h
e
d
u
r
ab
ilit
y
o
f
th
o
u
g
h
t
d
r
if
t.
Ho
w
e
v
er
,
it
w
o
r
k
s
o
n
l
y
w
i
th
g
o
o
d
f
ea
tu
r
e
en
g
in
ee
r
in
g
a
n
d
m
a
y
p
r
o
d
u
ce
f
al
s
e
p
o
s
itiv
e
s
o
n
alr
ea
d
y
g
en
u
i
n
e
b
u
t
u
n
e
x
p
lai
n
ed
p
o
ly
p
r
o
b
lem
s
.
Kits
u
n
e
is
esp
ec
ially
e
f
f
ec
tiv
e
f
o
r
d
is
co
v
er
in
g
th
r
ea
ts
s
u
c
h
as
DDo
S
attac
k
s
,
p
o
r
t scan
s
,
an
d
m
al
w
a
r
e
co
m
m
u
n
ica
tio
n
s
o
n
I
o
T
,
en
ter
p
r
is
e,
an
d
in
d
u
s
tr
ia
l c
o
n
tr
o
l s
y
s
te
m
s
.
2
.
3
.
2
.
Q
AE
cla
s
s
if
ica
t
io
n
T
h
e
QA
E
i
m
p
r
o
v
es
Kits
u
n
e
’
s
n
et
w
o
r
k
i
n
tr
u
s
io
n
d
etec
tio
n
b
y
eq
u
aliz
in
g
co
n
t
in
u
o
u
s
a
n
o
m
al
y
s
co
r
es
to
b
in
ar
y
s
e
v
er
it
y
le
v
el
s
[
2
5
]
.
I
n
o
r
d
er
to
p
r
eser
v
e
th
e
te
m
p
o
r
al
d
im
en
s
io
n
o
f
th
e
r
a
w
r
ec
o
n
s
tr
u
ct
io
n
er
r
o
r
s
co
r
es o
f
K
its
u
n
e
f
ir
s
t
s
tep
co
n
s
i
s
ts
o
f
t
h
e
n
o
r
m
alize
d
i
n
p
u
t
th
at
i
s
s
h
o
w
n
i
n
Fi
g
u
r
e
2
.
T
h
e
q
u
an
tizat
io
n
m
o
d
el
p
r
o
v
id
es
t
w
o
alter
n
at
iv
e
s
:
f
ix
ed
th
r
esh
o
ld
-
b
ased
b
in
ar
y
d
etec
tio
n
u
s
in
g
s
tatis
t
ical
p
er
ce
n
tile
s
o
r
a
lear
n
ed
q
u
an
tizat
io
n
s
c
h
e
m
e
w
it
h
tr
ain
ab
le
th
r
es
h
o
ld
s
an
d
m
o
r
e
g
r
an
u
lar
i
t
y
(
e.
g
.
,
n
o
r
m
al,
s
u
s
p
ici
o
u
s
,
an
d
m
a
l
icio
u
s
)
.
Fo
r
in
s
ta
n
ce
,
a
n
an
o
m
a
l
y
tr
a
f
f
ic
p
atter
n
is
d
escr
ib
ab
le
as
a
l
ar
g
e
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
,
w
h
i
ch
ca
n
b
e
r
eg
ar
d
ed
as
a
m
alic
io
u
s
D
Do
S
attac
k
.
T
h
ef
t
o
r
s
p
o
o
f
in
g
a
ttack
s
m
a
y
b
e
co
n
s
id
er
ed
an
o
m
alo
u
s
,
w
h
er
e
an
er
r
o
r
s
co
r
e
is
in
d
icati
v
e
o
f
d
ev
iatio
n
f
r
o
m
n
o
r
m
al
n
e
t
w
o
r
k
tr
af
f
ic
b
eh
a
v
io
r
s
.
A
n
o
r
m
al
tr
a
f
f
ic
co
n
d
iti
o
n
is
th
e
n
d
etec
ted
w
h
e
n
t
h
e
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
is
m
ai
n
tain
ed
u
n
d
er
s
o
m
e
th
r
e
s
h
o
ld
r
ep
r
esen
ti
n
g
n
o
s
i
g
n
i
f
ic
an
t
d
ep
ar
tu
r
e
f
r
o
m
th
e
le
g
iti
m
ate
tr
a
f
f
ic.
T
h
is
d
e
g
r
ee
-
b
ased
ca
te
g
o
r
y
e
n
ab
l
es
t
h
e
h
y
b
r
id
m
o
d
el
to
b
e
ab
le
to
h
an
d
le
a
v
ar
iet
y
o
f
attac
k
t
y
p
e
s
d
y
n
a
m
icall
y
an
d
lab
el
ea
ch
ab
n
o
r
m
alit
y
ap
p
r
o
p
r
iately
b
ased
o
n
it
s
s
e
v
er
it
y
an
d
f
ea
tu
r
e
s
.
I
n
ad
d
itio
n
to
f
ea
t
u
r
in
g
s
tr
o
n
g
er
,
m
o
r
e
ac
tio
n
ab
le
aler
t
s
,
Kit
s
u
n
e
’
s
o
n
lin
e
f
le
x
ib
ilit
y
ca
p
ab
ili
ties
ar
e
k
ep
t
in
ta
c
t,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
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
.
2
,
A
p
r
il
20
26
:
4
5
2
-
465
458
an
d
th
e
q
u
a
n
tizatio
n
ap
p
r
o
ac
h
m
a
y
ap
p
l
y
to
h
ier
ar
ch
ical
th
r
ea
t
ass
es
s
m
e
n
t
as
w
el
l.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
is
f
le
x
ib
le
i
n
h
an
d
li
n
g
v
ar
io
u
s
d
is
tr
ib
u
tio
n
ca
s
e
s
b
y
co
m
p
r
o
m
i
s
in
g
o
p
tio
n
al
s
u
p
er
v
is
ed
a
s
s
i
s
tan
ce
f
r
o
m
lab
eled
d
ata
w
it
h
u
n
s
u
p
er
v
i
s
ed
d
ep
lo
y
m
en
t.
Fig
u
r
e
2
.
H
y
b
r
id
Kits
u
n
e
–
Q
AE
m
o
d
e
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
I
n
th
is
s
e
g
m
en
t,
a
th
o
r
o
u
g
h
co
m
p
ar
at
iv
e
an
al
y
s
is
o
f
th
r
ee
d
if
f
er
en
t
in
tr
u
s
io
n
d
etec
tio
n
ar
r
an
g
e
m
e
n
t
s
is
d
escr
ib
ed
.
T
h
ese
m
o
d
els
i
n
clu
d
e
t
h
e
i
m
p
r
o
v
ed
Q
A
E
,
t
h
e
s
i
m
p
le
Kit
s
u
n
e
m
o
d
el,
an
d
a
jo
in
t
m
o
d
el
t
h
at
m
i
x
es
th
e
ap
p
r
o
ac
h
es
o
f
th
e
t
w
o
s
tr
ateg
ie
s
.
A
l
g
o
r
ith
m
1
s
h
o
w
s
th
e
o
v
er
all
p
r
o
ce
s
s
in
i
n
tr
u
s
io
n
d
etec
tio
n
.
T
h
e
ev
alu
a
tio
n
u
tili
ze
s
all
f
o
u
r
h
ig
h
-
lev
el
p
er
f
o
r
m
a
n
ce
m
etr
ics
a
s
f
o
llo
w
s
:
F1
s
co
r
e,
p
r
ec
is
io
n
,
m
o
d
el
r
ec
all,
an
d
s
y
s
te
m
ac
cu
r
ac
y
.
T
w
o
o
f
th
e
w
id
el
y
u
s
ed
I
o
T
d
atasets
ar
e:
B
o
t
-
I
o
T
an
d
R
T
-
I
o
T
.
T
h
e
r
esu
lt
w
a
s
r
ep
o
r
ted
as
f
o
llo
w
s
.
A
l
g
o
r
ith
m
1
.
I
n
tr
u
s
io
n
d
etec
ti
o
n
1:
←
∅
2: LOAD_DATASETS (
“
Bot
-
IoT
”
,
“
RT
-
IoT
”
)
3:
repeat
4:
preprocessed
←
preprocess_data(
)
5:
←
initialize_encoders(
preprocessed
)
6:
for
each
i
∈
do
7:
i
←
train_encoder
(
i
,
benign
)
8:
end for
9:
for
each
in
∈
incoming
do
10: e
reconstruction
(
in
) =
∑
=
1
(
in
-
x̂
i
)
2
11: e
quantized
(
in
) = quantize_error(e
reconstruction
(
in
))
12: c
classification
(
in
) =
“
malicious
”
,
if e
quantized
(
in
) >
θ
high
“
suspicious
”
,
if
θ
medium
< e
quantized
(
_
in)
≤
θ
high
“
normal
”
,
if e
quantized
(
in
)
≤
θ
medium
13:
if
c
classification
(
in
) =
“
malicious
”
then
14: ALERT_INTRUSION (
“
Malicious Activity Detected!
”
)
15:
else if
c
classification
(
in
) =
“
suspicious
”
then
16: ALERT_INTRUSION (
“
Suspicious Activity Detected!
”
)
17:
else
18: ALERT_INTRUSION (
“
Normal Traffic
”
)
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Hyb
r
id
in
tr
u
s
io
n
d
etec
tio
n
in
I
o
T d
ev
ices:
a
d
ee
p
lea
r
n
in
g
a
p
p
r
o
a
ch
u
s
in
g
…
(
Md
.
R
ifa
t E
N
o
o
r
)
459
19:
end if
20:
end for
21:
until
all data is processed
3
.
1
.
K
i
t
s
un
e
ba
s
eline
m
o
del
As o
p
p
o
s
ed
to
th
e
b
asel
in
e,
th
e
Kits
u
n
e
m
o
d
el
r
ec
ei
v
ed
a
n
F
1
s
co
r
e
o
f
8
2
.
2
to
4
3
.
4
,
ac
cu
r
a
c
y
o
f
8
4
.
2
to
8
5
.
4
%
,
p
r
ec
is
io
n
o
f
8
1
.
4
to
8
2
.
1
%
,
an
d
r
ec
all
o
f
8
2
.
5
to
8
3
.
9
%
.
T
o
s
u
m
m
ar
ize,
th
e
s
e
ar
e
th
e
d
ata
w
e
u
s
ed
to
estab
lis
h
an
h
o
n
e
s
t
b
aseli
n
e
v
ia
p
er
f
o
r
m
a
n
ce
,
an
d
to
d
eter
m
i
n
e
w
h
eth
er
Ki
ts
u
n
e
ac
tu
al
l
y
d
etec
ts
i
n
tr
u
s
io
n
s
in
th
e
I
o
T
.
T
h
o
u
g
h
w
e
o
b
s
er
v
ed
th
at
in
co
m
p
ar
is
o
n
w
ith
t
h
e
o
th
er
m
o
d
els,
th
e
p
r
ec
is
io
n
an
d
r
ec
all
co
u
ld
b
e
in
cr
ea
s
ed
.
T
h
ese
ar
e
t
h
e
k
e
y
s
tatis
t
ics
to
g
u
id
e
t
h
e
r
ed
u
ctio
n
o
f
t
h
e
v
ar
iet
y
o
f
f
als
e
p
o
s
itiv
e
s
a
n
d
th
e
m
ax
i
m
izatio
n
o
f
r
ea
l a
ttack
s
t
h
at
ar
e
d
etec
ted
.
3
.
2
.
Q
AE
m
o
del
I
n
all
asp
ec
ts
,
t
h
e
Q
A
E
m
o
d
el
w
a
s
s
u
p
er
io
r
to
th
e
b
aselin
e
Kits
u
n
e
m
o
d
el.
T
h
is
r
es
u
l
ted
in
ac
cu
r
ac
y
s
co
r
es
o
f
8
8
.
3
to
8
8
.
7
%
p
r
e
cisi
o
n
o
f
8
5
.
8
%
r
ec
all
o
f
8
4
.
8
to
8
5
.
6
%
an
d
F1
o
f
8
4
.
6
to
8
5
.
9
%.
T
h
is
is
co
m
p
ar
ed
to
Kit
s
u
n
e
a
n
d
s
h
o
w
ca
s
e
s
h
o
w
w
ell
Q
A
E
d
is
cr
ete
q
u
an
tizat
io
n
p
er
f
o
r
m
s
,
as
w
e
s
ee
a
4
.
5
%
in
cr
ea
s
ed
ac
cu
r
ac
y
,
4
.
4
%
in
c
r
ea
s
ed
p
r
ec
is
i
o
n
,
2
.
7
%
in
cr
ea
s
ed
r
ec
all,
an
d
a
3
.
2
%
in
cr
ea
s
ed
F1
s
co
r
e.
T
h
e
r
esu
lt
s
s
u
g
g
es
t
t
h
at
t
h
e
Q
A
E
m
o
d
el
w
o
r
k
s
w
el
l
w
it
h
b
o
th
d
atasets
,
p
r
o
d
u
cin
g
q
u
a
n
tit
ies
t
h
at
ar
e
j
u
s
t
s
lig
h
tl
y
d
if
f
er
e
n
t (
≤
0
.
4
%)
f
o
r
b
o
th
.
T
h
is
i
m
p
lie
s
th
at
it c
an
g
en
er
al
i
ze
w
ell
ac
r
o
s
s
a
w
id
e
r
an
g
e
o
f
I
o
T
tr
af
f
ic
p
atter
n
s
.
As
s
h
o
w
n
i
n
t
h
e
Fi
g
u
r
es
3
(
a)
an
d
(
b
)
,
Fig
u
r
es
4
(
a)
an
d
(
b
)
,
T
ab
le
4
an
d
T
ab
le
5
s
tr
o
n
g
l
y
s
u
p
p
o
r
t
th
e
f
ac
t
th
a
t
QA
E
e
n
h
a
n
ce
s
t
h
e
co
n
f
id
en
ce
o
f
d
ec
is
io
n
b
o
u
n
d
ar
ies
an
d
r
ed
u
ce
s
th
e
n
u
m
b
er
o
f
f
al
s
e
p
o
s
itiv
e
s
,
w
h
ic
h
lead
s
to
a
b
etter
d
etec
tio
n
r
ate
o
f
attac
k
s
.
(
a)
(
b
)
Fig
u
r
e
3
.
C
lass
if
ier
s
an
d
d
atas
ets
: (
a)
p
r
ec
is
io
n
an
d
(
b
)
r
ec
all
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
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
.
2
,
A
p
r
il
20
26
:
4
5
2
-
465
460
(
a)
(
b
)
Fig
u
r
e
4
.
C
lass
if
ier
s
an
d
d
atas
ets
:
(
a)
F1
s
co
r
e
an
d
(
b
)
ac
cu
r
ac
y
T
ab
le
4
.
A
cc
u
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
an
d
f
1
s
co
r
e
f
o
r
d
if
f
er
e
n
t
m
o
d
els
P
e
r
f
o
r
man
c
e
me
t
r
i
c
K
i
t
su
n
e
w
i
t
h
B
O
T
-
I
o
T
K
i
t
su
n
e
w
i
t
h
RT
-
I
o
T
Q
A
E
w
i
t
h
B
O
T
-
I
o
T
Q
A
E
w
i
t
h
RT
-
I
o
T
H
y
b
r
i
d
w
i
t
h
B
O
T
-
I
o
T
H
y
b
r
i
d
w
i
t
h
RT
-
I
o
T
A
c
c
u
r
a
c
y
8
4
.
2
%
8
5
.
4
%
8
8
.
7
%
8
8
.
3
%
8
7
.
2
%
8
6
.
9
%
P
r
e
c
i
si
o
n
8
1
.
4
%
8
2
.
1
%
8
5
.
8
%
8
5
.
8
%
8
4
.
2
%
8
3
.
8
%
R
e
c
a
l
l
8
2
.
5
%
8
3
.
9
%
8
4
.
8
%
8
5
.
6
%
8
4
.
2
%
8
4
.
5
%
F
1
sco
r
e
8
2
.
2
%
8
3
.
4
%
8
4
.
6
%
8
5
.
9
%
8
3
.
6
%
8
4
.
1
%
T
ab
le
5
.
A
cc
u
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
s
co
r
e
ad
v
an
tag
e
s
f
o
r
d
if
f
er
en
t
m
o
d
el
s
M
e
t
r
i
c
K
i
t
su
n
e
r
a
n
g
e
Q
A
E
r
a
n
g
e
H
y
b
r
i
d
r
a
n
g
e
Q
A
E
a
d
v
a
n
t
a
g
e
A
c
c
u
r
a
c
y
8
4
.
2
-
8
5
.
4
%
8
8
.
3
-
8
8
.
7
%
8
6
.
9
-
8
7
.
2
%
+
3
.
3
-
4
.
5
%
P
r
e
c
i
si
o
n
8
1
.
4
-
8
2
.
1
%
8
5
.
8
%
8
3
.
8
-
8
4
.
2
%
+
3
.
7
-
4
.
4
%
R
e
c
a
l
l
8
2
.
5
-
8
3
.
9
%
8
4
.
8
8
5
.
6
%
8
4
.
2
-
8
4
.
5
%
+
1
.
9
-
3
.
1
%
F
1
S
c
o
r
e
8
2
.
2
-
8
3
.
4
%
8
4
.
6
-
8
5
.
9
%
8
3
.
6
-
8
4
.
1
%
+
2
.
2
-
3
.
7
%
3.
3
.
H
y
brid
m
o
del
T
h
e
F1
s
co
r
e,
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
an
d
r
ec
all
o
f
h
y
b
r
i
d
m
o
d
el
(
KC
F+
Q
A
E
)
;
(
8
3
.
6
%
-
8
4
.
1
%),
(
8
6
.
9
%
-
8
7
.
2
%),
(
8
4
.
2
%
-
8
4
.
5
%),
r
esp
ec
tiv
el
y
.
Ov
er
al
l,
th
e
h
y
b
r
id
m
o
d
el
d
id
n
o
t
o
u
tp
er
f
o
r
m
Q
A
E
,
th
o
u
g
h
i
t
w
a
s
co
m
p
etiti
v
e
w
it
h
o
th
er
m
o
d
el
s
.
A
lth
o
u
g
h
t
h
er
e
ar
e
ce
r
tain
b
en
e
f
its
o
f
co
m
b
in
i
n
g
m
et
h
o
d
s
,
th
e
d
ata
ex
h
ib
ited
m
u
c
h
b
etter
d
etec
ti
o
n
ab
ilit
y
o
n
Q
A
E
.
T
o
o
p
tim
i
ze
th
e
p
er
f
o
r
m
a
n
ce
o
f
b
o
th
,
b
ased
o
n
th
e
r
es
u
lts
o
b
tain
ed
w
ith
t
h
e
h
y
b
r
id
m
o
d
el,
th
e
r
e
s
u
l
ts
in
d
icate
th
at
f
u
r
th
er
w
o
r
k
w
o
u
ld
b
e
b
e
n
ef
ic
ial
in
th
e
co
m
b
i
n
atio
n
o
f
b
o
th
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
h
y
b
r
id
m
o
d
el
i
s
al
s
o
p
r
esen
ted
in
Fi
g
u
r
e
4
.
A
d
etai
led
d
is
cu
s
s
io
n
o
f
t
h
e
p
r
o
p
o
s
ed
QA
E
m
o
d
el
w
i
th
r
es
p
ec
t
to
th
e
v
ar
io
u
s
I
DS
p
r
esen
ted
in
t
h
e
liter
atu
r
e
d
escr
ib
es
t
h
e
ad
v
a
n
ta
g
es
an
d
d
is
ad
v
an
ta
g
es
o
f
ea
ch
ap
p
r
o
ac
h
.
T
h
e
n
ew
s
t
at
e
-
of
-
th
e
-
a
r
t
m
o
d
e
l
s
in
t
h
i
s
d
o
m
a
in
o
f
in
t
r
u
s
i
o
n
d
e
t
e
c
ti
o
n
a
r
e
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
tw
o
r
k
s
an
d
d
e
e
p
a
u
t
o
en
c
o
d
e
r
s
,
w
h
i
ch
a
ch
i
ev
e
im
p
r
es
s
iv
e
ac
c
u
r
a
cy
[
1
0
]
,
[
1
5
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Hyb
r
id
in
tr
u
s
io
n
d
etec
tio
n
in
I
o
T d
ev
ices:
a
d
ee
p
lea
r
n
in
g
a
p
p
r
o
a
ch
u
s
in
g
…
(
Md
.
R
ifa
t E
N
o
o
r
)
461
B
u
t
th
ese
m
et
h
o
d
s
t
y
p
icall
y
r
el
y
o
n
s
u
p
er
v
i
s
ed
lear
n
in
g
alg
o
r
ith
m
s
.
A
n
n
o
tated
d
ata
is
h
ar
d
to
co
m
e
b
y
an
d
co
llect
in
r
ea
l
-
w
o
r
ld
I
o
T
s
etu
p
s
,
s
o
th
ese
m
o
d
els
n
ee
d
it.
F
o
r
tr
ain
in
g
s
u
p
er
v
is
ed
lear
n
t
m
o
d
el
s
’
d
ata
s
et
w
it
h
lab
els,
it
i
s
to
b
e
p
r
e
-
p
r
o
ce
s
s
ed
w
it
h
in
s
ta
n
ce
s
alr
ea
d
y
as
s
ig
n
ed
lab
els.
T
h
is
ad
d
s
lev
e
l
s
o
f
co
m
p
lex
it
y
i
n
ad
ap
tin
g
to
n
o
v
el
an
d
u
n
a
n
ticip
ated
attac
k
s
u
r
f
ac
es
(
t
h
is
i
s
p
ar
ticu
lar
l
y
p
r
o
b
lem
a
tic
w
i
th
ze
r
o
-
d
a
y
v
u
l
n
er
ab
ilit
ie
s
)
.
T
o
m
iti
g
ate
t
h
e
r
is
k
o
f
p
o
o
r
p
e
r
f
o
r
m
a
n
ce
w
h
e
n
n
o
lab
eled
d
ata
is
av
ail
ab
le,
w
e
p
r
o
p
o
s
e
an
au
to
-
lear
n
m
ec
h
a
n
is
m
as
an
i
n
teg
r
al
p
ar
t
o
f
o
u
r
QA
E
m
o
d
el.
T
h
u
s
,
th
e
Q
A
E
ap
p
r
o
ac
h
is
th
e
m
o
s
t
s
u
itab
le
f
o
r
th
e
I
o
T
ca
s
es
w
h
en
t
h
e
tag
g
ed
d
ata
s
tatis
tics
ar
e
h
ar
d
to
ac
h
iev
e.
An
attr
a
cti
v
e
f
ea
t
u
r
e
o
f
QA
E
w
it
h
r
esp
ec
t
to
s
u
p
er
v
i
s
ed
ap
p
r
o
ac
h
es
is
its
f
le
x
ib
ilit
y
.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
h
as
lo
w
s
tati
s
tical
v
ar
iati
o
n
(
≤
0
.
4
%)
in
I
o
T
tr
af
f
ic
p
atter
n
s
an
d
ex
ce
lle
n
t
p
er
f
o
r
m
a
n
ce
o
v
er
th
e
B
o
t
-
I
o
T
an
d
R
T
-
I
o
T
d
atasets
.
I
n
o
r
d
er
to
m
a
n
a
g
e
th
e
s
u
d
d
en
a
n
d
d
iv
er
s
e
p
r
o
p
er
ties
o
f
I
o
T
n
et
w
o
r
k
s
,
it
is
p
r
ef
er
r
ed
f
o
r
th
e
QA
E
to
p
er
f
o
r
m
w
el
l
ac
r
o
s
s
a
d
iv
er
s
it
y
o
f
n
et
w
o
r
k
t
y
p
es a
n
d
attac
k
s
t
r
ateg
ies.
T
h
an
k
s
to
t
h
e
b
etter
ca
p
ab
ilit
y
to
d
etec
t
attac
k
s
,
t
h
e
Q
A
E
m
o
d
el
ca
n
lear
n
b
y
it
s
el
f
.
W
h
en
m
ea
s
u
r
ed
ag
ain
s
t
ea
r
lie
r
tech
n
iq
u
es,
t
h
i
s
m
et
h
o
d
d
em
o
n
s
tr
ated
1
6
%
b
etter
ef
f
ec
ti
v
en
e
s
s
i
n
ze
r
o
-
d
a
y
th
r
ea
t
d
etec
tio
n
.
Z
er
o
-
d
a
y
attac
k
s
ar
e
p
ar
ticu
lar
l
y
h
ar
d
to
d
etec
t
s
in
ce
th
e
y
ex
p
lo
it
f
la
w
s
t
h
at
h
a
v
e
n
o
t
y
et
b
ee
n
d
is
co
v
er
ed
.
T
h
e
Hy
b
r
id
m
eth
o
d
,
w
h
ich
was
th
e
co
m
b
in
at
io
n
o
f
Kits
u
n
e
an
d
Q
A
E
,
p
r
o
v
id
ed
r
esu
lts
th
at
w
er
e
b
et
w
ee
n
t
h
e
t
w
o
.
Si
n
ce
th
e
q
u
an
tizatio
n
o
n
Q
A
E
is
a
p
r
ed
o
m
i
n
a
n
t
o
p
er
ato
r
f
o
r
its
s
p
ee
d
-
u
p
,
th
is
m
o
tiv
ates
t
h
e
id
ea
o
f
u
s
i
n
g
d
if
f
er
en
t
tec
h
n
iq
u
es
i
n
co
m
b
i
n
atio
n
w
ith
t
h
is
.
E
x
is
ti
n
g
s
o
lu
t
io
n
s
o
th
er
t
h
an
b
ein
g
o
r
d
er
s
o
f
m
a
g
n
itu
d
e
m
o
r
e
e
f
f
icie
n
t
t
h
a
n
d
ee
p
au
t
o
en
co
d
er
s
(
w
h
ich
w
as
d
e
m
o
n
s
tr
ated
b
y
Kits
u
n
e)
,
Q
A
E
e
x
ce
ls
in
co
m
p
u
t
in
g
ef
f
icien
c
y
.
W
ith
3
3
%
le
s
s
co
m
p
u
tat
io
n
al
e
f
f
o
r
t,
t
h
e
Q
A
E
m
o
d
el
o
f
f
er
s
r
ea
l
-
ti
m
e
ca
p
ab
il
ities
an
d
p
r
o
v
es
co
s
t
-
ef
f
ec
tiv
e,
esp
ec
iall
y
w
h
e
n
r
eso
u
r
ce
-
co
n
s
tr
ai
n
ed
en
v
ir
o
n
m
e
n
ts
ar
e
f
ac
to
r
s
(
r
ef
er
to
T
ab
l
e
6
)
.
Ow
i
n
g
to
its
ef
f
ec
tiv
e
n
e
s
s
as
w
el
l
as
t
h
e
s
u
p
er
io
r
d
etec
tio
n
ac
cu
r
ac
y
w
i
th
a
lo
w
f
al
s
e
p
o
s
iti
v
e
r
ate,
QA
E
i
s
th
e
o
p
ti
m
al
o
p
tio
n
f
o
r
s
ca
lab
le
an
d
ac
cu
r
at
e
-
d
etec
tio
n
I
o
T
n
et
w
o
r
k
s
.
T
ab
le
6
.
C
o
m
p
ar
ativ
e
s
u
m
m
ar
y
o
f
I
DS
m
et
h
o
d
s
an
d
r
esu
lt
s
f
r
o
m
k
e
y
s
t
u
d
ies
R
e
f
e
r
e
n
c
e
s
M
e
t
h
o
d
s
u
se
d
K
e
y
r
e
su
l
t
s/
c
o
n
t
r
i
b
u
t
i
o
n
s
Al
-
G
a
r
a
d
i
e
t
a
l
.
[
2
]
A
r
e
v
i
e
w
o
f
l
e
a
r
n
i
n
g
me
t
h
o
d
s fo
r
I
o
T
se
c
u
r
i
t
y
H
i
g
h
l
i
g
h
t
e
d
t
h
e
e
f
f
e
c
t
i
v
e
n
e
ss o
f
D
L
i
n
d
e
t
e
c
t
i
n
g
I
o
T
a
t
t
a
c
k
s
L
i
n
d
a
e
t
a
l
.
[
1
1
]
N
e
u
r
a
l
n
e
t
w
o
r
k
s
A
c
h
i
e
v
e
d
9
6
%
a
c
c
u
r
a
c
y
f
o
r
c
r
i
t
i
c
a
l
i
n
f
r
a
st
r
u
c
t
u
r
e
p
r
o
t
e
c
t
i
o
n
H
o
d
o
e
t
a
l
.
[
1
2
]
A
N
N
-
b
a
se
d
I
D
S
D
e
t
e
c
t
e
d
I
o
T
t
h
r
e
a
t
s w
i
t
h
9
4
%
a
c
c
u
r
a
c
y
El
r
a
w
y
e
t
a
l
.
[
1
7
]
S
u
r
v
e
y
o
f
I
o
T
I
D
S
Emp
h
a
s
i
z
e
d
t
h
e
n
e
e
d
f
o
r
l
i
g
h
t
w
e
i
g
h
t
,
h
y
b
r
i
d
a
p
p
r
o
a
c
h
e
s i
n
r
e
so
u
r
c
e
-
c
o
n
st
r
a
i
n
e
d
I
o
T
C
h
a
a
b
o
u
n
i
e
t
a
l
.
[
1
8
]
S
u
p
e
r
v
i
se
d
l
e
a
r
n
i
n
g
(
S
V
M
,
R
F
)
S
V
M
o
u
t
p
e
r
f
o
r
me
d
R
F
w
i
t
h
9
5
%
r
e
c
a
l
l
o
n
I
o
T
-
sp
e
c
i
f
i
c
d
a
t
a
se
t
s
L
i
e
t
a
l
.
[
1
9
]
S
y
st
e
m st
a
t
i
st
i
c
s
l
e
a
r
n
i
n
g
I
mp
r
o
v
e
d
d
e
t
e
c
t
i
o
n
o
f
z
e
r
o
-
d
a
y
I
o
T
a
t
t
a
c
k
s
b
y
2
2
%
o
v
e
r
t
r
a
d
i
t
i
o
n
a
l
me
t
h
o
d
s
M
i
r
s
k
y
e
t
a
l
.
[
2
4
]
En
se
mb
l
e
o
f
a
u
t
o
e
n
c
o
d
e
r
s (K
i
t
su
n
e
)
D
e
t
e
c
t
e
d
a
t
t
a
c
k
s
i
n
r
e
a
l
-
t
i
me
w
i
t
h
0
.
1
%
f
a
l
se
p
o
si
t
i
v
e
s
S
h
a
r
mi
l
a
a
n
d
N
a
g
a
p
a
d
m
a
[
2
5
]
Q
A
E
C
u
t
c
o
m
p
u
t
a
t
i
o
n
a
l
o
v
e
r
h
e
a
d
b
y
4
0
%
w
h
i
l
e
mai
n
t
a
i
n
i
n
g
a
d
e
t
e
c
t
i
o
n
a
c
c
u
r
a
c
y
o
f
9
3
%
K
o
l
i
a
s
e
t
a
l
.
[
2
6
]
R
u
l
e
-
b
a
se
d
a
n
o
m
a
l
y
d
e
t
e
c
t
i
o
n
P
r
o
p
o
se
d
l
i
g
h
t
w
e
i
g
h
t
I
D
S
f
o
r
I
o
T
w
i
t
h
9
2
%
a
c
c
u
r
a
c
y
A
n
a
g
n
o
st
o
p
o
u
l
o
s
e
t
a
l
.
[
2
7
]
S
i
g
n
a
t
u
r
e
-
b
a
se
d
,
b
e
h
a
v
i
o
r
a
l
a
n
a
l
y
si
s
D
e
t
e
c
t
e
d
mo
b
i
l
e
b
o
t
n
e
t
s w
i
t
h
8
9
%
F
1
-
sco
r
e
V
i
n
a
y
a
k
u
mar
e
t
a
l
.
[
2
8
]
D
e
e
p
l
e
a
r
n
i
n
g
(
L
S
T
M
,
C
N
N
)
A
c
h
i
e
v
e
d
9
8
.
5
%
d
e
t
e
c
t
i
o
n
r
a
t
e
o
n
N
S
L
-
K
D
D
d
a
t
a
se
t
R
a
z
a
e
t
a
l
.
[
2
9
]
L
i
g
h
t
w
e
i
g
h
t
I
D
S
(
S
V
E
L
T
E)
R
e
d
u
c
e
d
e
n
e
r
g
y
c
o
n
su
mp
t
i
o
n
b
y
3
0
%
i
n
6
L
o
W
P
A
N
n
e
t
w
o
r
k
s
M
e
i
d
a
n
e
t
a
l
.
[
3
0
]
D
e
e
p
a
u
t
o
e
n
c
o
d
e
r
s
I
d
e
n
t
i
f
i
e
d
I
o
T
b
o
t
n
e
t
s w
i
t
h
9
9
%
p
r
e
c
i
si
o
n
i
n
r
e
a
l
-
t
i
me
T
h
i
s
w
o
r
k
H
y
b
r
i
d
mo
d
e
l
o
f
Q
A
E
a
n
d
K
i
t
s
u
n
e
I
mp
r
o
v
e
d
d
e
t
e
c
t
i
o
n
o
f
z
e
r
o
-
d
a
y
I
o
T
a
t
t
a
c
k
s
b
y
1
6
%,
o
v
e
r
c
o
mi
n
g
c
o
mp
u
t
a
t
i
o
n
a
l
o
v
e
r
h
e
a
d
b
y
3
3
%
w
h
i
l
e
mai
n
t
a
i
n
i
n
g
o
v
e
r
8
5
%
d
e
t
e
c
t
i
o
n
a
c
c
u
r
a
c
y
W
h
ile
th
e
s
tan
d
alo
n
e
Q
A
E
m
o
d
el
o
u
tp
er
f
o
r
m
ed
th
e
h
y
b
r
id
m
o
d
el
in
t
h
r
o
u
g
h
p
u
t
ac
cu
r
ac
y
,
h
er
e
o
p
ted
to
em
p
lo
y
th
e
h
y
b
r
id
m
o
d
el,
as
a
m
o
r
e
co
m
p
lex
s
et
o
f
p
r
o
b
lem
s
in
r
ea
li
s
tic
I
o
T
a
p
p
licatio
n
s
e
m
er
g
ed
th
at
n
ee
d
ed
to
b
e
f
ac
ed
.
T
h
e
h
y
b
r
id
s
y
s
te
m
(
co
m
b
in
in
g
Kit
s
u
n
e
w
it
h
Q
A
E
)
h
as
co
m
p
le
m
e
n
tar
y
s
tr
en
g
t
h
s
th
at
ca
n
b
e
u
s
ed
to
ef
f
icien
tl
y
tac
k
le
d
if
f
er
en
t
a
ttack
class
e
s
an
d
u
n
k
n
o
w
n
t
h
r
ea
ts
.
On
e
s
i
n
g
le
QA
E
m
o
d
el
s
h
o
w
s
th
e
b
est
class
i
f
icatio
n
p
er
f
o
r
m
an
ce
,
b
u
t
h
as
n
o
ca
p
ac
ity
f
o
r
r
ea
l
-
ti
m
e
an
o
m
al
y
d
etec
tio
n
,
wh
ich
is
r
eq
u
ir
ed
f
o
r
an
i
n
tr
u
s
io
n
d
etec
tio
n
s
y
s
te
m
.
T
h
e
p
er
f
o
r
m
an
ce
o
f
t
h
e
Kits
u
n
e
m
o
d
el
is
b
etter
i
n
a
n
o
m
al
y
d
etec
tio
n
,
esp
ec
iall
y
in
id
en
ti
f
y
i
n
g
n
e
w
o
r
u
n
e
x
p
ec
ted
attac
k
s
,
w
h
er
e
QA
E
(
w
h
ic
h
i
s
a
s
i
n
g
le
m
o
d
e
l)
f
ail
s
to
id
en
tify
.
T
h
e
h
y
b
r
id
ap
p
r
o
ac
h
m
ak
es
m
o
r
e
s
e
n
s
e
w
it
h
d
y
n
a
m
ic,
lar
g
e
-
s
ca
le
I
o
T
n
et
w
o
r
k
s
w
it
h
h
i
g
h
l
y
v
ar
y
in
g
tr
af
f
ic
p
atter
n
s
an
d
t
h
e
r
eq
u
ir
e
m
e
n
t
to
d
etec
t/
co
u
n
ter
ac
t
u
n
f
o
r
eseen
attac
k
m
o
d
el
s
b
ef
o
r
e
ca
u
s
i
n
g
an
y
d
a
m
a
g
e.
B
esid
es,
alth
o
u
g
h
t
h
e
Sin
g
u
la
r
QA
E
m
o
d
el
p
er
f
o
r
m
ed
b
etter
in
th
e
co
n
tr
o
lled
ev
alu
atio
n
s
,
th
e
h
y
b
r
id
m
et
h
o
d
en
ab
les a
b
etter
p
er
f
o
r
m
an
ce
o
n
th
e
r
ea
l
-
ti
m
e
a
n
o
m
al
y
d
etec
t
io
n
an
d
th
e
d
y
n
a
m
ic
at
tack
cla
s
s
i
f
icatio
n
.
T
h
is
is
p
ar
ticu
lar
l
y
i
m
p
o
r
tan
t
in
I
o
T
s
ce
n
ar
io
s
,
w
h
er
e
attac
k
s
tr
ate
g
ies
ca
n
ch
a
n
g
e
r
ap
id
ly
.
T
h
e
ab
ilit
y
o
f
th
e
h
y
b
r
id
m
o
d
el
to
d
etec
t
s
u
b
tle
a
n
o
m
al
ies
b
ef
o
r
e
ca
teg
o
r
izatio
n
p
r
o
v
id
es
an
ad
d
itio
n
al
p
r
o
tectio
n
ag
ain
s
t
s
o
p
h
i
s
tic
ated
an
d
co
m
p
lex
attac
k
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.