TELKOM
NIKA
, Vol.14, No
.1, March 2
0
1
6
, pp. 129~1
3
5
ISSN: 1693-6
930,
accredited
A
by DIKTI, De
cree No: 58/DIK
T
I/Kep/2013
DOI
:
10.12928/TELKOMNIKA.v14i1.2742
129
Re
cei
v
ed O
c
t
ober 4, 20
14;
Revi
se
d Ja
n
uary 24, 201
6
;
Accepte
d
Febru
a
ry 8, 20
16
Analysis of Handover Trigger Scheme Based on
Distance for LTE High-speed Railway Networks
Rui Zhang*,
Muqing Wu,
Yifan Zhang
Beiji
ng U
n
ivers
i
t
y
of Posts an
d T
e
lecommun
i
catio
n
s, Beiji
n
g
100
87
6, Chin
a
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: zhangr
uib
upt
@12
6
.com
A
b
st
r
a
ct
In hig
h
-spe
ed r
a
ilw
ay e
n
viro
n
m
e
n
t, freque
nt
han
dov
ers an
d
hig
h
ha
nd
over
failur
e
pro
b
a
b
i
lityresult
in seri
ous co
mmu
n
icati
on i
n
terrupti
ons a
nd cal
l
drops,
w
h
ich beco
m
e pressi
ng pr
o
b
le
m to be so
lved
.
Alternativ
e refe
rence
po
int b
a
s
ed
han
dov
er
sche
m
e c
a
n
be
an
ap
plic
ative
soluti
on for
hi
g
h
-spe
ed r
a
il. T
h
i
s
pap
er a
naly
z
e
s
the i
m
pact of
han
dov
er l
o
ca
tion o
n
th
e h
a
ndov
er p
e
rfor
ma
nce. A
nd
n
u
merica
l a
n
a
l
y
s
is is
utili
z
e
d
to g
i
ve
a metho
d
to d
e
termin
e
the
di
stance fr
o
m
th
e servi
ng e
N
o
deB a
nd th
e a
d
jac
ent o
ne to
th
e
mo
bil
e
ter
m
i
n
al w
hen
ha
n
dover is tri
g
g
e
red. H
and
ov
er failur
e
pr
o
bab
ility is
us
ed as crit
eria
for
perfor
m
a
n
ce.
Simulati
on
res
u
lts sh
ow
that
the
pro
pose
d
sche
m
e h
a
s
a b
e
tter p
e
rfor
ma
nce
tha
n
ty
pical
event bas
ed sc
he
me i
n
hig
h
s
pee
d env
iron
ment.
Ke
y
w
ords
:
ha
ndov
er trigg
e
r, hig
h
-spe
ed ra
il
, handov
er fail
ure pro
b
a
b
ility
Copy
right
©
2016 Un
ive
r
sita
s Ah
mad
Dah
l
an
. All rig
h
t
s r
ese
rved
.
1. Introduc
tion
High
-spee
d railway
s play an increa
sin
g
l
y impor
tant part in pe
opl
e’s lives
re
ce
nt years
[1]. The cu
rrent b
r
oa
dba
nd wi
rele
ss comm
uni
cat
i
on sy
stem
s optimize
d
f
o
r lo
w mo
bi
lity
environ
ment
s can
not mai
n
tain ade
qua
te perfo
rman
ce for
pa
sse
ngers traveling on the
hi
gh
spe
ed train
s
(HST
) any l
o
nger. Typi
cal
hard ha
ndov
er
sup
porte
d
by LTE sy
ste
m
s redu
ce
s t
h
e
compl
e
xity of the system
architectu
re
but simu
lta
n
e
ously b
r
ing
s
about hig
her
hand
over fail
ure
ratio and the
decli
ne of the
user exp
e
ri
e
n
ce in hi
gh-speed mo
bility environm
ent [2].
Han
dover b
e
c
ome
s
a g
r
e
a
t challe
nge
in the systems de
sig
n
e
d
for high
-sp
eed rail
mainly due to
the following
rea
s
on
s: wh
en the ru
nnin
g
spe
ed of HST achieve
s
350
km/h or e
v
en
highe
r in the future, the time interval o
f
every two handove
r
s
can
be as sho
r
t as 15
s, given
a
size of abou
t 1.5km which is typical in Macr
o cell
[3], and it
results in a high han
dov
er
freque
ncy; cl
assical event handove
r
algorithm
trig
g
e
rs a ha
ndo
ver pro
c
ed
ure basin
g on HO
hystere
s
i
s
an
d Time to Tri
gger,
whi
c
h i
n
crea
se
s the
time latency
of handove
r
and me
anwhi
le it
may lead to Radi
o Link F
a
ilure (RL
F
) b
e
fore the u
s
e
r
terminal
s su
ccessfully
acce
ss to the target
eNo
deB. Several
pape
rs h
a
ve provided
optimize
d
scheme
s
for LT
E hand
over. I
n
refe
ren
c
e
[4],
the autho
rs a
nalyze
d
the impact of p
r
op
agation e
n
vironment a
nd velocity of UE
on the ha
ndo
ver
perfo
rman
ce.
The referen
c
e [5] gave
a thre
shol
d
hand
over tri
g
gerin
g sch
e
m
e for
wirel
e
ss
netwo
rks. Re
feren
c
e [6, 7] also introdu
ced alg
o
rith
ms for LTE system
s to improve ha
nd
ove
r
perfo
rman
ce.
In ord
e
r to
so
lve the proble
m
s me
ntione
d
above,
th
e referen
c
e
[8] prop
osed a
h
andove
r
algorith
m
u
s
i
ng the
pa
rticu
l
arity of the
ra
ilway
environment. Acc
o
rding to
referenc
e [9], handover
pro
c
ed
ure co
uld be
execu
t
ed ba
sed
o
n
GPS (Gl
o
b
a
l Position
System) info
rmation. But the
hand
over ref
e
ren
c
e point whi
c
h
i
s
the most
a
ppr
opriate to achiev
e sati
sfying p
e
rform
a
n
c
e
was
not given. So in this pa
per, the ana
lysis of
han
dover trigg
e
r location will
be prop
ose
d
.
Con
s
id
erin
g the path loss, log-n
o
rm
al
shad
ow
i
ng
and freq
uen
cy selective fast fading, the
hand
over
pe
rforma
nce b
a
se
d on
ha
n
dover fail
ur
e
pro
bability
at ce
rtain di
stan
ce from
the
servin
g statio
n can b
e
figu
red out. In this wa
y, the ha
ndover
refere
nce p
o
int ca
n
be determi
n
e
d
throug
h num
erical analy
s
i
s
.
The pa
pe
r is orga
nized a
s
follo
ws: Se
cti
on 2 i
n
tro
duces th
e sy
stem mo
del
and th
e
locatio
n
-b
ase
d
handove
r
p
r
ocedu
re. Nu
meri
cal analy
s
is a
nd han
d
o
ver refe
ren
c
e point choi
ce
are
sho
w
n in
se
ction 3. Se
ction 4
sho
w
s the simu
la
tion
r
e
s
u
lts
an
d
fin
a
lly c
o
nc
lus
i
o
n
s
ar
e
dr
aw
n
in se
ction 5.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 14, No. 1, March 2
016 : 129 – 1
3
5
130
2. Sy
stem Model and Lo
cation
-
ba
se
d Handov
er Procedur
e Introdu
ction
LTE netwo
rk architectu
re
is co
nstituted
of
three pa
rts as
sh
own in Figure 1: e
v
olved-
Nod
e
B (e
No
deB), Mobile
Manage
men
t
Entity
(MME), and Servi
ng Gate
way (S-G
W) / Pa
cket
Data
Net
w
o
r
k Gate
way
(P-GW). Th
e e
N
odeB
perfo
rm
s all
radio
int
e
rface
relate
d
functio
n
s
su
ch
as packet scheduli
ng
a
nd hand
over
m
e
cha
n
ism.
MM
E manag
es
mobility, use
r
equip
m
ent
(UE)
identity and secu
rity para
m
eters. S-G
W
and P-G
W
are two nod
es that termi
nate the interface
towar
d
s E-
UT
RAN an
d Pac
k
et Data
Net
w
or
k r
e
s
p
e
c
tiv
e
ly
[10].
Figure 1. E-UTRAN a
r
chite
c
ture
Combi
ned
wi
th the groun
d-to-t
rain pa
rt, the network ca
n be sh
own a
s
Figu
re 2: the
radio f
r
eq
uen
cy sig
nal
s ge
nerate
d
at B
a
se Ba
nd
Un
it (BBU) a
r
e
firstly co
nvert
ed into o
p
tical
sign
als an
d tran
smitted th
roug
h the fiber to it
s Radi
o Remot
e
U
n
it
(RR
U
).
B
B
U sen
d
s
sa
me
sign
als to its
RRUs a
nd
o
n
ly pa
rt of th
e RRUs rece
ive sig
nal
s from the
trai
n
at a
ce
rtain ti
me
interval. For t
he network o
n
the train, vehicl
e
termin
al provid
es th
e last ho
p co
mmuni
cation
for
the sub
s
crib
e
r
s in th
e
ca
rri
age. In
this p
aper,
we
con
s
ide
r
th
e h
a
n
dover bet
wee
n
sou
r
ce
cell
and
target cell.
Figure 2. eNo
deB develop
ment along
ra
ilway
The followi
ng
s sh
ow the
steps of the
lo
cation-b
a
sed h
andove
r
procedure step
s:
Whe
n
the
UE enters the
overlap
re
gio
n
, it
sen
d
s m
easure
m
ent
report
s
p
e
rio
d
i
cally to
the sou
r
ce eNod
eB, inclu
d
ing the sp
e
ed and lo
cati
on informatio
n get from GPS. When the
spe
ed is
high
er than
a cert
ain thre
sh
old
th
V
, the pro
p
o
s
e
d
schem
e wil
l
be ch
osen,
or othe
rwise
typical A3
ev
ent ba
se
d
scheme
should
be u
s
e
d
.
Wh
en the
lo
catio
n
info
rmation
indicates that
the
distan
ce b
e
tween the train
and the ha
nd
over re
fe
ren
c
e point is cl
oser than a thre
shol
d of
th
d
, the
sou
r
ce e
N
o
d
e
B ma
ke
s th
e HO de
ci
sio
n
an
d send
s
the HO requ
est to th
e target e
N
odeB.
The
target e
N
o
d
e
B
save
s the
context, prep
aring
L1/L2
for
HO a
nd
re
spo
ndin
g
to the sou
r
ce wit
h
a
HO
req
u
e
s
t
ACK. The
so
urce e
N
o
deB
tran
sfers al
l
the ne
ce
ssary information
to the
UE in
the
H
O
co
mmand
.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Analysis of Hando
ve
r Trig
ger Schem
e Based o
n
Di
stance fo
r LTE
High-sp
eed
… (Rui Z
han
g)
131
From
then
o
n
, so
urce
e
N
odeB
begin
s
to fo
rward d
o
w
nlin
k
data t
o
the
targ
et
eNo
deB.
Simultaneo
usly, UE detaches from
sou
r
ce
eNodeB
and syn
c
h
r
o
n
ize
s
to the
target on
e. And
then, UE sen
d
s HO co
nfirmation to targ
et eN
od
eB ab
out the su
cce
ss of radio ha
ndover.
Finally, targ
et eNodeB
beg
ins to
send
its b
u
ffered
da
ta re
ceived
from the
so
urce on
e
and
sen
d
s
HO co
mplete
messag
e to i
n
itiate data
p
a
th switchi
ng.
After MME/S-G
W confirm
s
the
path switchin
g, target e
N
odeB will n
o
tice
sou
r
ce
e
N
od
eB to flush its fo
rward downlin
k d
a
ta
buffer and
rel
ease re
sou
r
ce.
3. Numerical
Analy
s
is an
d Handov
er Referen
ce P
o
int Choic
e
In this
sectio
n, the num
eri
c
al a
nalysi
s
of
hand
over
referen
c
e
poi
nt choi
ce i
s
p
r
opo
se
d,
con
s
id
erin
g t
he ha
ndove
r
failure
pro
bability.
For the nu
meri
cal
evaluatio
n, the an
alysis
diagrammati
c sket
c
h is
sho
w
n in Figu
re
3.
Figure 3. Analysis dia
g
ra
m
m
atic sket
ch
The di
stan
ce
betwe
en the
servin
g e
N
od
eB and th
e a
d
jacent e
N
od
eB is
D
. The di
stance
betwe
en th
e
eNo
deB a
nd
railway is
s
d
.
h
D
repre
s
e
n
ts th
e len
g
th of
o
v
erlap
re
gion
. The trai
n
location
x
is th
e distan
ce
b
e
twee
n the
vehicle
stat
io
n and o
r
igin
in X-axis. T
herefo
r
e, th
e
distan
ce b
e
twee
n the veh
i
cle
station a
nd the
sou
r
ce and ta
rget
eNo
deBs
ca
n
be expresse
d as
22
s
s
Dx
d
,
22
()
ts
DD
x
d
resp
ective
ly. OFDM
(Orth
ogo
n
a
l Fre
quen
cy Division
Multiplexing)
is utili
zed fo
r data tran
sm
issi
on.
Th
us,
the b
a
seba
nd
sign
al received fro
m
t
he
sou
r
ce eNod
eB can be o
b
t
ained a
s
follow:
1
2
0
(
)
()
()
N
jm
f
t
ss
s
s
n
s
n
rt
p
l
s
h
h
t
d
e
t
(1)
Whe
r
e
2
s
s
pl
A
D
re
pre
s
ent
s the
pat
h lo
ss,
A
is a
consta
nt, and
is the
path l
o
ss expo
nent.
2
s
s
h
is the log
-
no
rmally distrib
u
ted sh
ado
win
g
fading,
()
s
ht
is the sm
all scal
e Rayleig
h
fading,
N
is the total n
u
m
ber
of su
bcarri
ers,
n
d
is tra
n
smitted
data
and tran
smit power
2
[|
|
]
dn
PE
d
,
f
is
sub
c
a
rri
er sp
acin
g and
()
s
t
ind
i
cate
s the ze
ro mean Ga
ussian n
o
ise.
Acco
rdi
ng to
[11], the sign
al stre
ngth
re
ceived from
source e
N
o
d
e
B
defined a
s
s
R
in dB
can b
e
cal
c
ul
ated as follo
w:
ˆ
22
ˆ
2
10
l
g
(
|
|
[
|
(
)
|
])
1
0
l
g
(
)
10
l
g
10
l
g
n
sd
s
s
n
ds
s
RP
p
l
s
h
E
k
PA
D
s
h
(2)
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 14, No. 1, March 2
016 : 129 – 1
3
5
132
Whe
r
e
1
1
2
0
0
1
(2
)
(
1
|
|
)
L
lD
s
l
J
fT
x
x
d
x
,
2
l
repre
s
ent
s the sh
ado
w fading
deviation,
D
s
f
T
is
the normaliz
e
d maximum
Doppler f
r
equenc
y
s
h
ift.
ˆ
ˆ
()
n
n
k
is the u
s
eful f
a
ctor
co
ntrib
u
ted to the
th
n
sub
c
a
rri
er, an
d can b
e
expressed a
s
:
1
ˆ
ˆ
2/
ˆ
,
0
0
1
()
(
)
s
L
T
nj
n
l
N
ns
l
l
s
ke
h
t
k
T
d
t
T
(3)
The sig
nal st
rength from ta
rget eNodeB
can b
e
define
d
simila
rly as
t
R
based on Eq
uation (2
).
Whe
n
the train is in the
overlap re
gi
on, if the ve
hicle termina
l
is not trigg
e
red to
hand
over b
e
fore th
e si
gn
al strength f
r
om sour
ce
e
N
od
eB de
ca
ys to a level
lowe
r tha
n
a
threshold
T
whi
c
h is the min
i
mum re
ceive
d
sign
al st
re
ngth to maintain the com
m
unication,
o
r
the si
gnal
st
rength f
r
om t
a
rget
eNode
B is l
o
we
r th
an
T
after the
hand
over i
s
t
r
igge
re
d, the
hand
over fail
ure
occu
rs.
So the ha
nd
over failu
re p
r
oba
bility can
be divide
d i
n
to failure
be
fore
trigge
ring a
n
d
failure after triggeri
ng, d
enoted by
_
fb
P
and
_
fa
P
re
spe
c
tively. Therefore, the point
that has both
minimized
_
fb
P
and
_
fa
P
can b
e
sel
e
cted a
s
the h
andove
r
refe
rence point.
In traditio
nal
GPS-b
a
sed
han
dover a
l
gorithm, th
e
han
dover referen
c
e
poi
nt is a
statistical result based
on
repe
ated m
e
asu
r
em
ent.
We
su
ppo
se
that the hyst
ere
s
is level i
s
U
(dB). So the hand
over fail
ure proba
bility before
trigg
e
ring
whe
n
the handove
r
trigger p
o
int is
x
can b
e
expre
s
sed a
s
:
2
2
_
2
{|
}
{,
}
{}
1
{1
0
/
|
}
{
}
{}
10
l
g
(
/
)
11
(1
(
)
)
{}
2
s
fb
t
s
s
st
s
s
ts
t
s
s
s
st
st
s
PP
R
R
U
R
T
PR
T
R
R
U
PR
T
PS
U
D
D
S
P
S
d
PR
T
UD
D
Qe
d
PR
T
(4)
Whe
r
e
10
l
g
(
)
10
l
g
ds
TP
A
D
, and
1
0
lg
(
)
1
0
lg
{}
1
(
)
ds
s
TP
A
D
PR
T
Q
. And
the hand
over
failure p
r
ob
ab
ility after triggering
can b
e
expre
s
sed si
milarly as:
2
2
_
2
{|
}
{,
}
{}
1
{1
0
l
g
(
/
)
|
}
{
}
10
l
g
(
/
)
11
(1
(
)
)
2
t
fa
t
t
s
tt
s
ts
ss
t
t
t
ho
st
ho
s
t
PP
R
T
R
R
U
PR
T
R
R
U
PR
R
U
PS
D
D
U
S
P
S
d
P
DD
U
Qe
d
P
(5)
Whe
r
e
10
l
g
(
)
10
l
g
dt
TP
A
D
,
{
}
((
1
0
lg
(
/
))
/
)
ho
t
s
s
t
PP
R
R
U
Q
U
D
D
, and
22
s
t
.
Acco
rdi
ng to
Equation (4
) and (5
), wh
en t
he dista
n
c
e bet
wee
n
the two e
N
od
eBs is
fixed,
_
fb
P
will increase
when
the handover point
x
becom
es l
a
rg
er, whil
e
_
fa
P
ha
s an
oppo
site
tren
d. Therefo
r
e,
con
s
ide
r
ing
both hand
o
v
er failure p
r
oba
bilities b
e
fore a
nd af
ter
trigge
ring, th
e poi
nt ap
propriate
to
execute
ha
ndo
ver
can
be
configured
an
d sele
cted
a
s
th
e
referen
c
e poi
nt, when the failure p
r
o
babi
lity
f
P
is minimum at location
x
as Equatio
n (6):
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Analysis of Hando
ve
r Trig
ger Schem
e Based o
n
Di
stance fo
r LTE
High-sp
eed
… (Rui Z
han
g)
133
__
(1
)
*
(1
)
f
fb
f
a
PP
P
(6)
4. Results a
nd Analy
s
is
The pe
rform
ance of the prop
osed scheme
ba
se
d
on distan
ce
can be vali
dated by
simulatio
n
. T
he
spe
ed
ran
ge of th
e trai
n is supp
o
s
ed
to
be
60
km/h
to
36
0km/h
.
T
h
e
o
v
er
la
p
area of adj
acent cell
s is a
s
sume
d to be 1.5km.
The simul
a
tion pa
ramete
rs are sho
w
n i
n
Table
1. Th
e pro
pag
atio
n model
of Cost23
1-
Hata is u
s
e
d
[12], and it ca
n be expre
ssed as Equ
a
tio
n
(7):
46.
3
33.
9
l
g
13.
82
l
g
(
)
(
44.
9
6
.
5
5
l
g
)
l
g
cb
b
m
Lf
H
H
m
H
d
C
(7)
The
ch
annel
model
of thi
s
scen
e i
s
m
o
untain
enviro
n
ment. Th
e e
s
timated
m
C
can be
o
b
taine
d
from project e
x
perien
c
e a
s
Equation (8):
10.03
m
C
(8)
Table 1. Simulation Para
meters
List
Energ
y
Power
1
Band
w
i
dth of Su
bcarrier
15kHz
2 Carrier
Fre
quenc
y
2.6GHz
3 S
y
stem
Band
w
i
d
t
h
10MHz
4
Transmission Time Interval(T
TI)
1ms
5 Overlap
300m
6
Distance betwee
n
eNodeB an
d R
a
ilw
a
y
Line
100m
7 Transmitting
Po
w
e
r
46dBm
8 Noise
-148.95dB/sub
-channel
9
Path Loss
Refer to Eq.
(7)
10
Log-no
rmal Shad
ow
Fading
Standard D
e
viation=8dB
Mean=0
11
Signal Threshold
T
-58dBm
We
ch
oo
se t
he tri
gge
ring
points withi
n
1300m
~1
600
m re
gion
in t
he ove
r
lap
a
r
ea. Th
e
relation
shi
p
b
e
twee
n the h
andove
r
failure and the
trig
gerin
g point i
s
sim
u
lated.
Figure 4 sh
o
w
s
the analysi
s
o
f
handover fai
l
ure p
r
ob
ability before trigg
e
ring h
app
eni
ng.
Figure 4. Handover failure
pr
obability before triggering
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 14, No. 1, March 2
016 : 129 – 1
3
5
134
And Figu
re
5 sho
w
s the
tenden
cy variation of
h
andove
r
failu
re proba
bility after trigge
ring
happ
ening
at
given lo
catio
n
s. It can b
e
see
n
t
hat
wh
en trig
geri
ng
happ
en
s cl
oser to th
e
sou
r
ce
eNo
deB, a lowe
r failure prob
ability be
fore trigg
e
ri
n
g
can
be obt
ained, an
d it con
s
tantly g
e
ts
highe
r evid
e
n
tly along
wit
h
the l
o
catio
n
moving
cl
o
s
er to th
e target e
N
od
eB. Mean
whil
e,
the
failure p
r
ob
ab
ility after triggering h
a
s a d
o
wntrend p
r
e
s
entin
g a co
n
t
rary variation
.
The h
and
ove
r
failu
re p
r
o
bability wh
en
trigge
ring
within 130
0m
~1600m
is sh
own
in
Figure 6.
Comprehensiv
ely consi
d
ering the fa
ilure probability before
and after t
r
iggering
happ
ening,
it ca
n b
e
see
n
that it g
o
e
s
u
p
afte
r d
r
oppin
g
. Thi
s
illustrate
s th
a
t
whe
n
spe
c
i
f
ic
scena
rio is gi
ven, a small region
suited t
o
trigge
r han
dover in the
overlap a
r
e
a
can b
e
analy
z
ed
according to
the ch
ann
el environ
ment
, so a
s
to redu
ce the
h
andove
r
failu
re p
r
ob
ability to
accepta
b
le ra
nge.
Figure 5. Handover fa
ilure probability
after
trigge
ring
Figure 6. Handover fa
ilure probability
when
trigge
ring at g
i
ven locatio
n
Figure 7. Performa
nce com
pari
s
on of tra
d
itional and
p
r
opo
se
d sche
me
Comp
ari
s
o
n
betwe
en the prop
osed sch
e
me
and typical event ba
sed schem
e is shown
in Figure 7. It can be seen that they h
a
ve si
milar perform
a
nce i
n
low mobilit
y environment.
Whe
n
the user termin
al ru
ns at a high
spe
ed,
event
base
d
sche
me displ
a
ys
a perfo
rman
ce of
seri
ou
s de
clin
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Analysis of Hando
ve
r Trig
ger Schem
e Based o
n
Di
stance fo
r LTE
High-sp
eed
… (Rui Z
han
g)
135
5. Conclusio
n
This pa
per h
a
s analy
z
ed
a handove
r
schem
e ba
se
d on distan
ce information
for LTE
high-sp
eed
rail networks.
A method
of sele
cting
hand
over
ref
e
ren
c
e
point
is provided
for
scena
rio
s
giv
en
cha
nnel
e
n
vironm
ent.
Simulation
re
sults sho
w
t
hat a
sm
all region
suited
to
trigge
r in
the
overlap
a
r
ea
can
be
obtai
n
ed,
an
d the
h
andove
r
p
e
rf
orma
nce
can
be in
crea
sed
by
redu
cin
g
the
hand
over fail
ure
proba
bility.The pr
opo
sed
scheme
can a
dapt to
the hi
gh-sp
ee
d rai
l
environ
ment
better than typical eve
n
t base
d
sc
he
me
, becau
se it
can eli
m
inate
ping-pong
HO
and a
c
hieve
prefe
r
abl
e ha
ndover failu
re
proba
bility.
Ackn
o
w
l
e
dg
ements
This work wa
s su
ppo
rted b
y
Technol
ogy
Major Proj
ects (No.2011
Z
X
03001
-0
07-03).
Referen
ces
[1]
Che
n
g
w
e
i
L, Mingc
hin C, Mengc
han
g C
.
S
eamless Han
dover for
High-Sp
ee
d T
r
ains Usin
g
F
e
mtocell-B
a
s
ed Mu
ltipl
e
Egress
Ne
t
w
ork
Interfa
c
es.
IEEE Transactions
on Wir
e
less
Co
mmun
icati
o
ns
. 2014; 1
3
(1
2): 6619-
66
27.
[2]
Meng-S
h
iu
an
P,
T
z
u-Ming L,
Men-T
s
uen C
.
An E
nha
nced
Hand
over Sc
heme for Mo
bi
le Re
la
ys i
n
LT
E-A High Spee
d Ra
il Net
w
o
r
ks.
IEEE Transactions on Vehicular Tec
hnology
. 2
0
1
5
; 64(2): 74
3-
756.
[3]
ET
SI
T
R
102.281: Detai
l
e
d
re
quir
e
ments
for GSM operatio
n
on Rai
l
w
a
ys, v
2
.0.0. 200
6.
[4]
Saif-Ur-Re
h
ma
n Q, Syed Ju
nai
d N, Moha
mmad P.
T
he imp
a
ct of pro
pag
atio
n envir
on
me
nt a
n
d
velocity on the
handov
er
perform
ance of LTE
system
s
. Proc
. of Internation
a
l Co
nfere
n
ce
on W
i
rel
e
ss
Commun
i
cati
o
n
s and Si
gn
al
Processi
ng. Su
Z
hou. 201
0: 1-5.
[5]
Limin
g
C, Qin
g
G, Z
heng
yu
N, Kai
y
ua
n J
.
A
T
h
reshol
d
Based H
a
n
d
o
v
er T
r
iggerin
g
Scheme i
n
Heterogeneous Wireless Net
w
orks.
T
e
lko
m
n
i
ka (T
elec
o
m
mu
nicati
on
Co
mp
uting El
ectronics a
n
d
Contro
l)
. 2014;
12(1): 163-
17
2.
[6]
Danis
h
A, Rolf
S.
Improve
m
ent of LT
E Handov
er Perfor
ma
nce thro
ug
h Interference
Coor
din
a
tio
n
.
Proc. of Vehicu
lar T
e
chnolo
g
y
Confere
n
ce (V
T
C
). 2009: 1-5.
[7]
Li J, T
i
an L, Z
hou
Y.
An
ad
aptive
ha
nd
ov
er trigg
e
r sc
he
me
for w
i
rel
e
s
s
co
mmun
icati
ons
on
hig
h
spee
d rail
. IEEE Internatio
nal
Confer
ence
on
Commun
i
cati
o
n
s (ICC). 2012:
5185-
51
89.
[8]
Linl
in L, Mu
qin
g
W
,
Yifan Z
.
A Hand
over Al
gor
ithm for LT
E S
y
stem b
a
s
ed on th
e T
a
rget Cel
l
Pre-
Bearer i
n
Hig
h
-
spee
d Rai
l
w
a
y Envir
onm
ent
.
Internationa
l Journ
a
l of
Adv
ance
m
ents in
Co
mp
uting
T
e
chno
logy
. 2
012; 4(2
2
): 631
-640.
[9]
Linl
in
L, Muqi
n
g
W
,
W
e
i W
.
A GPS-based
han
dov
er al
gor
ithm in LT
E hi
gh-sp
eed r
a
il
wa
y
net
w
o
rks.
Advanc
es in In
formati
on Sci
e
nces an
d Servi
c
e Scienc
es
. 2012; 4(9): 2
05-
213.
[10]
3GPP
T
S
36.423: Evolve
d U
n
iv
ers
a
l T
e
rrestrial Ra
dio Acc
e
ss Net
w
ork (
E
-UT
R
AN) X2
Applic
ation
Protocol (
X
2A
P), V8.3.0. 200
8.
[11]
T
i
an L, Li J, Hu
ang Y. S
eaml
e
ss Dua
l
-Li
n
k H
and
over Sc
he
me inBr
oad
ba
n
d
W
i
rel
e
ss Co
mmunicati
o
n
S
y
stems for
H
i
gh-Sp
ee
d Ra
il.
IEEE Journal
on S
e
lected
Ar
eas inCommunications
. 20
12;
30(4):
708-
717.
[12]
Reka
w
t
SH, T
A
Rahma
n
, AY Abdulr
ahm
a
n
. LT
E Coverage N
e
t
w
ork
Plan
nin
g
an
d
Comp
ariso
n
w
i
t
h
Differ
ent Propa
gati
on Mode
ls.
T
e
lko
m
n
i
ka (T
el
eco
m
mu
nicati
on
Co
mp
uting E
l
ectronics
and
Contro
l)
. 2014;
12(1): 153-
16
2.
Evaluation Warning : The document was created with Spire.PDF for Python.