Improving the Reliability of a Train Positioning System through the
Use of Full Coverage Radio Communication Technologies
Performance Study of a TETRA Network to Transmit Position Information
Roberto Carballedo, Pablo Fernández, Unai Hernández Jayo and Asier Perallos
University of Deusto, Deusto Institute of Technology, Av. Universidades, 24, Bilbao, Spain
Keywords: Wireless Communications, Terrestrial Trunked Radio, Train Positioning System, Railway Industry.
Abstract: Today, it is common for trains to incorporate autonomous positioning systems based on geo-location
technologies similar to those used on road transportation. These positioning systems represent a cost
effective solution for railway companies operating in not evolved regions. Furthermore, these positioning
systems can enhance the reliability of positioning systems based on the occupation of the tracks (which are
most used in the most developed regions). Autonomous positioning systems calculate the position of the
train, but the position has to be sent from the train to the control center. GPRS/3G mobile technologies and
WiFi radio technologies are the most common technologies for transmitting the position to the control
centers. These technologies do not guarantee 100% coverage in certain areas such as tunnels or
mountainous regions. This paper presents the results of the tests on an autonomous positioning system to
add a new communications technology in order to increase its coverage. This technology is TETRA, which
is a radio technology that has been traditionally used for voice transmission, but it can be a good
complement to GPRS/3G when there is no coverage.
1 INTRODUCTION
Train positioning systems are a key element to
ensure the safety of the rail system. Most positioning
systems are based on electronic devices that are
installed on the tracks. These positioning systems
are very robust and reliable, but when they stop
working, security levels are significantly reduced. In
these situations the position of trains is done
manually by live-voice communication between
drivers and control centers. For this reason, in recent
years, there have been many initiatives to develop
backup positioning systems, in which the train
becomes the protagonist (Jiang et. al., 2010) (Bai-
Gen et. al., 2011). These positioning systems
calculate the position of the trains by means of
different position sources: GPS, MEMS gyroscope,
maps, ATP, odometer, etc. Once the position is
calculated, it is sent to the control center using a
wireless communications network, usually GPRS or
3G. The main limitations of these systems are the
coverage of mobile communications (tunnels,
overgrown or isolated areas, etc.), and the cost of the
communications (usually communications network
belongs to a private operator). To tackle this
limitation, this paper presents the results of tests
performed with a train positioning system in order to
improve reliability. The tests were focused on
evaluating the improvement achieved by
incorporating the TETRA as a backup channel in
situations where there is no GPRS/3G coverage.
The paper is organized in 5 sections: Section 2
details the potential of TETRA technology for
transmitting position information in the railways.
Section 3 explains the underlying details of the
positioning system and communications architecture
to send the position from the trains to the control
center. Section 4 shows the results of the tests; and
finally, the paper concludes with Section 5 which
details the conclusions and future lines of work.
2 TERRESTRIAL TRUNKED
RADIO FOR DATA
TRANSMISSION
TETRA (TErrestrial Trunked RAdio) is the
communications standard, defined in 1999 by the
323
Carballedo R., Fernández P., Hernández Jayo U. and Perallos A..
Improving the Reliability of a Train Positioning System through the Use of Full Coverage Radio Communication Technologies - Performance Study of a
TETRA Network to Transmit Position Information.
DOI: 10.5220/0004062003230328
In Proceedings of the International Conference on Signal Processing and Multimedia Applications and Wireless Information Networks and Systems
(WINSYS-2012), pages 323-328
ISBN: 978-989-8565-25-9
Copyright
c
 2012 SCITEPRESS (Science and Technology Publications, Lda.)

European Telecommunications Standards Institute,
for a generation of digital radio products. The main
characteristics of TETRA are:
 It is an open standard air interface so it provides
real interoperability with other communication
systems.
 High spectrum efficiency with short bandwidth
that allows creating private workgroups of users and
high quality of services, for example at voice
communications.
 It can be used on all frequencies below 1000
MHz frequency bands for Private Mobile Radio
(PMR) and public safety applications are assigned
on a national basis.
 In the railway context it offers encrypted and
high-quality communications in noisy environments
and whatever the signal conditions, providing up to
28Kbps bit rate.
 It is a mature technology with several suppliers
of terminals.
At the beginning, TETRA was deployed to satisfy
the needs of the Private Mobile Radio (PMR), Public
Access Mobile Radio (PAMR), Land Mobile Radio
(LMR) and the public applications for security and
protection corps, guard coast, fire-fighter’s services
and ambulances. But later developments at the
standard allowed Package Data Optimized (PDO)
transmission mode and Short Data Services (SDS)
that is comparable with the Short Data Message
(SMS) of GSM.
Package Data Optimized is used by modern
systems to transmit voice message, electronic mail,
data interchange, and so on. In this regard,
Enhanced-PDO known as DAWS (Digital Advanced
Wireless Services) provides a 2Mbps bit rate to be
used in combination with UMTS in high mobility
scenarios as Railroad Digital Networks (Wenlong,
Haige and Hongjie, 2002).
Short Data Services are transmitted over the
control channel in a point-to-point or point-to-
multipoint format without any reception
acknowledgement. Two message formats are
available:
 Status messages that only with 16 bits can
represent 65,535 messages.
 User messages that use a variable package size
from 16 to 2,047 bits.
In Railroad Networks basic TETRA Short Data
Service for polling and transmitting GPS position in
Location Information Protocol (LIP) format can be
easily used with a high reliability in different
scenarios because SDS messages can be send
between subscribers, between subscriber and
dispatcher and subscriber and fixed host in the
network. Moreover, SDS messages can be sending
to individual subscribers or broadcasted on a number
of base stations to all subscribers using that base
station.
In the railroad applications TETRA works in
PDO or SDS modes. These systems are connection
oriented so the mobile node installed in the train
requires first a connection to a specific Base Station
to start the data transmission. As the train is a
moving node, fast handover techniques are needed
to maintain the node connected to the network. This
process produces a delay in the communication that
can break down the link and cause the failure of the
data transmission (Palit, Bickerstaff, and Langmaid,
2009). Moreover, railway networks may also suffer
from the following problems caused by the handover
applied techniques:
 Handover to a wrong cell, which has the same
main control channel frequency as a neighbour cell.
The Mobile Station (MS) then loses all sensible
neighbours as it gets information on neighbours to
the "wrong" cell.
 Cell dragging by an MS, where frequency reuse
is reduced by interference of the transmitting MS
with others that would otherwise be out of range.
3 POSITIONING SYSTEM
ARCHITECTURE
Our positioning system (Carballedo et al., 2010) is
organized conceptually into two functional blocks:
(1) the positioning system itself, and (2)
communications system. The positioning system is
responsible for calculating the position, while the
communications system sends the position from the
train to the control center. Then the particularities of
each are described.
3.1 Positioning System
Train positioning system is primary based on GPS
data. This system is able to generate train
positioning information applying a logical
approximation algorithm for matching railway lines
and GPS coordinates (Guan-Wei et al., 2009). To
generate the most accurate positioning information,
this system uses railway lines lengths (in kilometres)
and traffic signals positions. Based on this
information, the data extracted from the hardware
(GPS, MEMS gyroscope, maps, ATP and odometer)
WINSYS2012-InternationalConferenceonWirelessInformationNetworksandSystems
324

and posi
t
train p
o
p
oint is
tabulate
N
ow
p
osition
i
railway
2004).
B
ionosph
e
systems
that besi
to kno
w
especial
l
accurac
y
separate
d
to dete
c
informa
t
 GPS
data. Th
e
develop
m
b
ased o
n
b
e pos
s
navigati
o

M
E
M
is also i
n

M
ap
s
al., 200
9
coordin
a
informa
t
identify
the inter
n

A
uto
m
are an a
d
differen
t
b
eacons
p
rovide
a
gyrosco
p
b
etween
memory
 Odo
m
a cover
e
error du
e
But it is
In fact,
distance
exact po
3.2
C
The co
m
the posi
t
architec
t
of infor
m
tioning algor
i
o
sition to ki
l
a metric us
e
the lines whe
r
adays, GPS
i
ng solution
environment
s
B
ut due to
G
e
ric propag
a
have not en
o
des the positi
o
w
the exact t
r
l
y complex
b
y
is around t
h
d
by less tha
n
c
t the exac
t
t
ion sources a
r
Coordinates
.
e
GPS chip i
s
m
ent of the
n
standards. T
h
s
ible to mig
r
o
n satellite sy
M
S Gyroscop
e
n
tegrated in t
h
s
. They are
o
9
), (Saab, 2
0
a
tes infor
m
t
ion with gy
r
the exact tra
c
n
al memory
o
m
atic Train
P
d
di
t
ional part
t
kinds of
A
located in th
e
a unique ID
a
p
e-
b
ased trac
k
t
r
acks and
b
of the positi
o
m
ete
r
. It prov
i
e
d distance r
e
e
to wheels
w
very useful
w
for tracks
and speed
d
sition of the t
r
C
ommunic
a
m
munications
t
ion is based
t
ure which al
l
m
ation betwe
e
i
thms, this sy
s
l
ometric poi
n
e
d by the rai
l
r
e its trains ci
r
is a goo
d
because of
s
(Bertran, an
G
PS inherent
a
tion…), G
P
o
ugh accura
c
o
n of trains, i
t
r
ack each tr
a
b
ecause the
h
ree meters,
a
n
2 meters. T
h
t
track, 5
d
r
ea are used:
.
It provides
s
in the on bo
a
positioning
h
erefore, in t
h
r
ate to or
i
stem like Gal
i
e
. It provides
h
e on board h
a
o
rganized as
0
00a) and (S
a
m
ation. Co
r
oscope data,
c
k. All the
m
o
f the positio
n
P
rotection (
A
of signaling
s
A
TPs, but a
l
e
infrastructu
r
a
nd are used t
k
detection
m
b
eacon ID is
a
o
ning system.
i
des relative
p
e
gister. It has
w
earing away
w
hen there is
n
without GP
S
d
ata are used
r
ains.
a
tions S
y
st
e
system that i
s
on a wireles
s
l
ows a full-d
u
e
n trains and
s
tem translate
s
n
ts. A kilom
e
l
way compan
r
culate.
d
and low
its reliabilit
y
d Delgado-P
e
error (multi
p
P
S only b
a
c
y. The reas
o
t
is also nece
s
a
in takes. Th
i
GPS positio
n
a
nd the track
s
h
erefore, in
o
d
ifferen
t
p
os
i
absolute pos
i
a
rd hardware.
s
ystem has
b
h
e future, it w
o
i
ntegrate an
o
i
leo.
angular spee
a
rdware.
in (Guan-W
e
a
ab, 2000b)
w
m
bining
m
it is
p
ossibl
m
aps are stor
e
ing systems.
A
TP) Data.
T
ystems. Ther
e
l
l of them
n
r
e. These bea
c
o
correct pos
s
m
istakes. Rel
a
a
lso stored i
n
p
osition data.
an accumul
a
and wheels s
n
o GPS cove
r
S
coverage,
o
to determin
e
e
m
s
used to tra
n
s
communica
t
u
plex transmi
s
t
he control c
e
s
the
etric
n
y to
cost
y
in
e
nin,
p
ath,
ased
o
n is
s
sary
i
s is
n
ing
s
are
o
rder
i
tion
i
tion
The
b
een
ould
o
ther
e
d. It
e
i et.
with
m
aps
l
e to
e
d in
T
hey
e
are
n
eed
c
ons
s
ible
a
tion
n
the
It is
a
tive
lide.
r
age.
only
e
the
n
smit
t
ions
s
sion
e
nte
r
(G
u
im
p
b
y
H
rail
co
m
mo
d
(on
to
ma
n
an
d
ab
n
the
sw
i
(m
o
suc
an
d
tec
h
inf
o
im
a
tra
n
p
ri
o
of
t
are
GP
R
op
e
rob
the
int
e
ce
n
the
the
TE
T
arc
h
op
e
li
m
u
tiérrez e
t
al.
,
p
lemented an
d
H
irschmann
w
w
ays. In
m
munication
d
ems and a
W
Each train
r
e at each en
d
a maste
r
-sl
a
n
ages comm
u
d
the slave
n
ormality is
d
master.
The commu
n
i
tch between
t
o
bile or WiFi
)
h
as: priorit
y
d
communic
a
h
nology is us
e
o
rmation, su
c
a
ges. Further
m
n
smitting inf
o
o
rity, such as
t
Figure 1 sho
w
t
he positionin
g
connected
R
S/3G link
s
e
rator, to
e
u
stness) and
a
stations). Al
l
e
grated into t
h
n
ter by means
information
t
figure also il
l
T
RA netw
h
itecture.
Figure 1:
Currently
t
e
rational, but
m
itations:
,
2010). This
d
deployed on
w
hich meets t
h
addition t
o
syste
m
inte
g
W
iFi radio inte
r
r
ides two c
o
of the train)
a
ve scheme.
u
nications wi
monitors i
t
d
etected, the
s
n
ications syst
e
h
e two com
m
)
based on a
n
y
, volume of
a
tion cost.
O
e
d for transm
i
c
h as logs o
f
m
ore mobile
t
o
rmation of
t
he position o
w
s the comm
u
g
system. As
to the con
(each wit
h
e
nsure maxi
m
a
WiFi link (
w
l
subnets (G
P
h
e Ethernet
n
of firewalls,
e
t
ransmitted f
r
l
ustrates
t
he f
u
o
r
k
in
t
Communicati
o
t
he commu
n
m
obile techn
o
architecture
n
a router ma
n
h
e specificati
o
o
the ro
u
g
rates two
G
e
rface.
o
mmunicatio
n
that operate
a
The maste
r
i
th the contr
o
ts behaviou
r
s
lave takes t
h
t
em is progr
a
m
unication tec
h
n
umber of p
a
information,
O
n one ha
n
i
tting large v
o
f
operation
o
t
echnology i
s
small size
a
o
f trains.
m
unications ar
c
can be seen,
n
trol center
h
a differe
n
mum cover
a
w
hich is onl
y
P
RS/3G and
W
n
etwork of t
h
e
nsuring the s
e
r
om the train
s
fu
ture integrat
i
t
he comm
u
o
ns architectur
e
n
ication sy
o
logy has the
f
has been
n
ufactured
o
ns of the
ter, the
G
PRS/3G
systems
a
ccording
r
system
o
l center,
r
. If an
h
e role of
a
mmed to
h
nologies
a
rameters,
coverage
n
d, WiFi
o
lumes of
o
r CCTV
used for
a
nd high
c
hitecture
the trains
via two
n
t phone
a
ge and
active in
W
iFi) are
h
e control
e
curity of
s
. Finally,
i
on of the
u
nications
e
.
stem is
f
ollowing
ImprovingtheReliabilityofaTrainPositioningSystemthroughtheUseofFullCoverageRadioCommunication
Technologies-PerformanceStudyofaTETRANetworktoTransmitPositionInformation
325

 Ther
e
the rail
w
system
g
vegetati
o
 The
network
,
informa
t
These t
w
not ver
y
was do
n
coverag
e
trains t
o
carrying
The
b
eing
u
commu
n
100% o
f
the volu
m
motivati
o
p
resente
d
4 T
E
In this f
o
in a re
obtaine
d
determi
n
alternati
v
the train
s
The
t
goes fro
Spain).
T
a moun
t
numero
u
GPRS/3
G
Further
m
for th
e
commu
n
architec
t
p
rimaril
y
tests h
a
p
erform
a
a new s
o
p
osition
compari
s
defined,
In t
h
current
p
osition
establis
h
has defi
n
commu
n
e
is only 80
%
w
ay line whic
h
g
oes through
m
o
n, and also t
h
transmission
,
has a cost,
w
t
ion transmitt
e
w
o constraint
s
y
reliable. F
o
n
e to analyze
e
in the trans
m
o
the control
out the work
line on whi
u
sed, has a
n
ication. Thi
s
f
coverage an
d
m
e of inform
a
o
n for the s
t
d
below.
E
ST AND
o
urth section
al life scen
a
d
. The main
n
e whether
a
v
e for the tra
n
s
to the contr
o
t
ests have be
e
m Bilbao to
T
he route is a
b
t
ainous area
u
s tunnels.
B
G
coverage r
m
ore, the line
e
moment
n
ications. C
u
t
ure of the
y
on GPRS/3
G
a
ve been
f
a
nce of the c
u
o
lution that in
t
through a T
E
s
on three d
i
as detailed b
e
h
e first scena
r
version of
is transmitte
d
h
communicat
n
ed a virtual
n
ications net
w
%
of coverag
e
h
is using th
e
m
ountainous
a
h
ere are man
y
of informati
o
w
hich depend
s
e
d.
s
, makes the
p
o
r this reaso
n
the simples
t
m
ission of inf
o
center. This
presented in
t
ch the posit
i
TETRA ne
t
s
communica
t
d
it has no co
a
tion transmi
t
t
udy of com
m
RESULT
S
we present t
h
a
rio and an
a
objective of
a
TETRA ne
t
n
smission of
t
o
l center.
e
n done on a
San Sebastia
n
b
out 140 km.
(with lots o
f
B
ecause of
eaches only
8
has a TETR
A
is used
o
u
rrently, the
positioning
G
communic
a
f
ocused on
u
rrent solutio
n
t
egrates the t
r
E
TRA netw
o
i
fferent scen
a
e
low:
r
io (which c
o
the position
i
d
via GPRS/3
ion, the mobi
private net
w
w
ork. This
V
e
. This is bec
a
e
communica
t
a
reas, with lo
t
y
tunnels.
o
n via the m
o
on the volu
m
p
ositioning sy
s
n
, extensive
w
t
way to inc
r
o
rmation fro
m
is the reaso
n
t
his paper.
i
oning syste
m
t
work for
v
t
ion network
st associated
w
t
ted. This wa
s
m
unication d
e
S
h
e field tests
d
a
lyse the re
s
the tests w
a
t
work is a
v
t
he positions
f
r
ailway track
n
(in the nor
t
and runs thr
o
f
vegetation)
the terrain,
8
0% of the t
r
A
network, w
o
nly for
v
communica
t
system is b
a
a
tions. So, al
l
comparing
n
(GPRS/3G)
w
r
ansmission o
f
rk. To make
a
rios have
b
o
rresponds t
o
i
ng system),
G technolog
y
l
e phone pro
v
w
ork (VPN) i
n
V
PN is integ
r
ause
t
ions
ts of
o
bile
m
e of
s
tem
w
ork
r
ease
m
the
n
for
m
is
v
oice
has
with
s
the
e
lays
d
one
s
ults
a
s to
v
alid
f
rom
that
t
h of
o
ugh
and
the
r
ac
k
.
w
hich
v
oice
t
ions
ased
l
the
the
with
f
the
this
b
een
o
the
the
y
. To
v
ider
n
its
r
ated
wit
h
Th
u
the
co
n
Th
u
TE
T
net
w
p
er
f
TE
T
sce
n
fut
u
run
p
o
s
of
n
in
ins
t
p
o
s
p
o
s
tha
t
net
w
rad
i
are
p
o
s
ha
v
sen
co
n
b
yt
e
p
o
s
(so
u
h
the Ether
n
u
s, train posi
t
trains to the
c
In the seco
n
n
trol center a
r
u
s, all com
m
T
RA networ
k
w
orks. This
f
ormance of
T
RA radios.
n
ario is to i
m
u
re. In order
n
ing on the
s
s
itions. Thus,
n
earby trains
(
The third sc
e
the final s
o
t
alled on th
e
s
ition to the
s
ition will rea
c
t
integrates
t
w
ork of the
i
os mentione
d
dedicated ex
c
s
ition.
With the thr
e
v
e been deve
l
ding positio
n
n
trol center.
T
e
s and their
c
s
ition where
t
u
rce) of the
p
n
et network
o
t
ions are tra
n
c
ontrol center
.
Figure 2: Scen
a
d scenario,
b
r
e equipped
w
m
unications a
k
, without th
e
second s
c
the commu
n
The main
o
m
prove the po
s
t
o increase s
a
s
ame section,
each train w
o
(
like radar).
Figure 3: Scen
a
e
nario is the
o
lution. A T
E
e
train. Thi
s
T
ETRA net
w
c
h the control
t
he TETRA
n
control cent
d
in the seco
n
c
lusively for
t
e
e scenarios,
l
oped. The t
e
messages b
e
T
he message
s
c
ontent descri
b
t
he train is l
o
p
osition. Dur
i
o
f the contr
o
n
smitted secu
r
.
n
ario 1.
b
oth the trai
n
with a TET
R
a
re made thr
e
interventio
n
c
enario anal
y
n
ications bet
w
o
bjective of
s
itioning syst
e
a
fety levels,
t
,
could exch
a
o
uld know th
e
n
ario 2.
o
ne to be im
p
E
TRA radio
s
radio tran
s
w
ork. And fi
n
l
center throu
g
n
etwork and
t
e
r
. Note the
n
d and third
t
he transmiss
i
several tests
e
sts have co
n
e
tween a trai
n
s have a siz
b
es the train
o
cated and t
h
i
ng testing, t
h
o
l center.
r
ely from
n
and the
R
A radio.
o
ugh the
n
of other
y
ses the
w
een two
this test
e
m in the
t
he trains
a
nge their
e
location
p
lemented
will be
s
mits the
n
ally, the
g
h a VPN
Ethernet
TETRA
scenarios
i
on of the
batteries
n
sisted of
n
and the
e of 572
itself, the
h
e quality
h
e time it
WINSYS2012-InternationalConferenceonWirelessInformationNetworksandSystems
326

takes to
center h
was cor
r
control
c
client.
A
TCP so
c
the sam
p
ossibili
offers T
E
this sen
s
set-up t
i
have be
e
the clos
u
Finally,
message
lost due
into acc
o
Tabl
e
columns
(2) TET
R
message
message
the cont
r
p
ercenta
As
i
scenario
are best
1.28 se
c
problem
lost due
In re
l
2 and
3
transmit
t
observe
d
S
c
Scenario1
Scenario2
Scenario2
Scenario3
Scenario3
get message
s
as been mea
s
r
ect, both the
c
ente
r
have
b
A
ll transmiss
i
c
kets; so that
e order they
ties, the two
m
E
TRA (SDS
a
s
e, since PD
P
i
me and an
e
e
n sent at dif
f
ure and ope
n
note that
s have been
p
to lack of
c
o
unt in the an
a
Figu
r
e
1 shows the
in the table
r
R
A communi
c
sent, (4)
t
is sent from
r
ol cente
r
(th
e
a
ge of messag
e
i
t was expe
c
(where info
r
ones. In thi
s
c
onds. Altho
u
of this scena
r
to lack of co
v
Table
l
ation to the r
3
, which ar
e
t
ing positio
n
d
the delays
a
c
enario
(3G‐‐>VPN)
(Tetra‐‐>Tetra)
(Tetra‐‐>Tetra)
(Tetra‐‐>VPN)
(Tetra‐‐>VPN)
s
from the tr
a
s
ured. To en
s
train positio
n
b
een synchro
n
i
ons have be
e
the messages
were sent.
m
odes of dat
a
a
nd PDO) ha
v
P
mode has
a
e
xpiration ti
m
f
erent interva
l
n
ing of the
P
only corr
e
p
osted, that is,
c
overage, ha
v
a
lysis.
r
e 4: Scenario 3
results of the
r
epresents: (1
)
c
ation mode,
t
he average
the train unt
i
e
smaller the
b
e
s received at
c
ted, the res
r
mation is tra
n
s
scenario the
u
gh this dela
y
r
io is tha
t
20
%
v
erage.
1: Tests result
s
esults obtain
e
e
those that
n
informati
o
a
re between
Mode
Msg.Size
(bytes)
‐‐‐ 572
PDP 572
SDS 572
PDP 572
SDS 572
a
in to the co
n
s
ure that the
t
n
ing system a
s
n
ized using
N
e
n made thr
o
were receiv
e
T
o cover all
a
transmission
v
e been teste
d
a
communica
t
m
e, the mess
a
l
s of time to
f
P
DO connec
t
e
ctly transm
i
the message
e not been t
a
.
tests. Each o
f
)
the test scen
a
(3) the size o
f
delay since
i
l it is receiv
e
b
etter) and (5
)
the destinati
o
ults of the
n
smitted via
3
average del
a
y
is the best,
%
of message
s
s
.
e
d in the scen
a
use TETRA
o
n; as can
3
and 5 sec
o
Received
%
Del
(seco
80,12%
1
99,75%
4
99,80%
4
99,87%
3
99,68%
4
n
trol
time
s
t
he
N
TP
o
ugh
e
d in
the
n
that
d
. In
t
ions
ages
f
orce
t
ion.
m
itted
data
a
ken
f the
ario,
f
the
the
e
d at
)
the
o
n.
first
3G),
a
y is
,
the
s
are
a
rios
for
be
o
nds.
Th
e
in
s
res
u
An
o
2
a
TE
T
del
a
ou
r
sta
t

po
s
tha
t
me
t

the
urb
co
v
use
In
acc
GP
R
tec
h
the
ad
d
p
o
s
TE
T
Ac
c
p
o
s
5
Th
i
to
sys
t
dif
f
p
o
s
AT
is
co
m
Du
e
the
Th
i
reli
sys
t
net
w
TE
T
net
w
the
r
are
ay
nds)
1
,279
4
,851
4
,040
3
,510
4
,090
e
best results
s
cenario 3 (u
u
lts are obta
i
o
ther import
a
a
nd 3, no me
s
T
RA networ
k
Although th
e
a
y of GPRS/
3
positioning
t
ement, are:
The averag
e
s
itioning syst
e
t
in 3 secon
d
t
ers.
The most cri
t
stations. Th
e
an centers. I
n
v
erage and s
o
d.
conclusion,
eptable resul
t
R
S/3G. Th
e
h
nology is th
a
destination.
d
ing to our
s
sibility of t
r
T
RA netwo
r
c
ordingly, thi
s
itioning syst
e
CONCL
U
WORK
i
s paper prese
n
increase the
t
em. The tr
a
f
erent source
s
ition of the t
r
P and odome
t
sent to the
m
munications
e
t
o the natur
e
communicat
i
i
s implies tha
t
able. The rai
t
em is deplo
y
w
ork that is
u
T
RA networ
k
w
ork could b
e
r
e is no GP
R
no studies to
The aim of
t
o
f TETRA te
s
ing PDO).
F
i
ned in stag
e
n
t issue is th
e
s
sages are lo
s
has a covera
g
e
delay of T
E
3
G, this tech
n
system. The
e
speed of
t
e
m is about
8
d
s, the train
t
ical areas wi
e
stations ar
e
n
t
hese areas
o
, GPRS/3G
TETRA
s despite hav
i
e
main ad
v
a
t 100% of
m
Therefore, t
h
communicat
i
r
ansmitting i
n
r
k, will m
a
s
will increa
s
m
propose
d
.
U
SIONS
A
n
ts the result
s
reliability
o
a
in positioni
n
s
of inform
a
r
ain: GPS, M
E
t
er. After cal
c
control cen
architecture
e
of the railw
a
i
ons architect
u
t
the position
i
l
way line in
y
ed has a T
E
u
sed for voic
e
k
has a cov
e
a good alte
r
R
S/3G covera
g
confirm this
h
t
he work don
e
e
chnology are
F
urthermore,
t
e
2 (also wi
t
e
fact that in
s
t. This is be
g
e of 100%.
ETRA is 3
t
n
ology is still
justification
s
the trains
u
8
5 km/h. Th
i
could move
i
th regard to
s
e
usually loc
a
there is mob
transmissio
n
technology
i
ng a delay h
i
v
antage of
m
essages sent
,
h
ese tests co
n
i
ons archite
c
n
formation t
h
a
ke it mor
e
s
e the reliabil
i
A
ND FU
T
s
of the tests
p
o
f a train p
o
n
g system te
s
a
tion to calc
E
MS gyrosco
p
c
ulating the p
n
ter using a
based on
G
a
y line, the c
o
u
re only reac
ing system i
s
which the p
o
E
TRA comm
u
e
communica
t
v
erage of 10
0
r
native for ar
e
g
e, but curre
n
h
ypothesis.
e has been t
o
achieved
t
he worst
t
h PDO).
scenarios
c
ause the
t
imes the
valid for
s
for this
u
sing our
s implies
about 70
s
afety are
a
ted near
i
le phone
n
can be
provides
ghe
r
than
TETRA
arrive at
n
firm that
c
ture, the
h
rough a
e
robust.
i
ty of the
T
URE
p
erformed
o
sitioning
s
ted uses
ulate the
p
e, maps,
o
sition, it
wireless
G
PRS/3G.
verage of
hes 80%.
not very
o
sitioning
u
nications
ions. The
0
%. This
e
as where
n
tly there
o
confirm
ImprovingtheReliabilityofaTrainPositioningSystemthroughtheUseofFullCoverageRadioCommunication
Technologies-PerformanceStudyofaTETRANetworktoTransmitPositionInformation
327

whether the TETRA network is a valid alternative to
improve the reliability of the positioning system. For
this, a series of tests were done. The tests have been
focused on measuring the delay in sending the
position from the trains to central control. It has also
been counted the number of messages sent
successfully.
The results of the tests confirmed the initial
hypothesis. Therefore, TETRA technology is a good
alternative to increase the coverage of the
communications architecture.
In the future, efforts will focus on the integration
of the TETRA network in the current
communication system. To do this, we must modify
both the hardware and software of the
communications architecture. On the one hand, we
must connect two new TETRA stations to the router
that manages the communications between the train
and the control center. On the other hand, we must
define a scheme for selecting the right
communications technology (GPRS/3G or TETRA)
according to the existing coverage.
ACKNOWLEDGEMENTS
This work would not have been possible without the
collaboration of EuskoTren (which is a railway
company operating in the north of Spain), and
without funding from the industry department of the
Basque Government.
REFERENCES
Bai-Gen, C., Yi, A., Guan-Wei, S., Jiang, L. and Jian, W.
(July 2011). Application simulation research of
Gaussian particle filtering in train integrated position
system. Prof of IEEE International Conference on
Service Operations, Logistics, and Informatics (SOLI),
ISBN: 978-1-4577-0573-1, pp. 527-531. Beijing,
China.
Bertran, E. and Delgado-Penin, J. A. (September, 2004).
On the use of GPS receivers in railway environments.
IEEE Transactions on vehicular technology, vol. 53,
no.5.
Carballedo, R., Salaberria, I., Perallos, A., Odriozola, I.,
Gutierrez, U. (September, 2010). A Backup System
Based on a Decentralized Positioning System for
Managing the Railway Traffic in Emergency
Situations. Proc. of 13th International IEEE
Conference on Intelligent Transportation Systems - On
the Way to Intelligent Sustainable Mobility (ITSC),
ISBN: 978-1-4244-7658-9, pp. 285-290. Madeira
Island, Portugal.
Guan-Wei, S., Bai-Gen, C., Jian, W. and Jiang, L. (June,
2009). Research of Train Control System Special
Database and Position Matching Algorithm. Proc. of
the IEEE Intelligent Vehicles Symposium. pp. 1039-
1044. Xian, China.
Gutiérrez,U., Salaberria, I., Perallos, A., Carballedo, R.
(April, 2010). Towards a Broadband Communications
Manager to Regulate ‘train-to-earth’ Communications.
Proc. of the 15th IEEE Mediterranean
Electrotechnical Conference (MELECON), ISBN:
978-1-4244-5794-6, pp. 1600-1605. Valletta, Malta.
Jiang, L., Bai-Gen, C., Tao, T. and Jian, W. (August,
2010). A CKF based GNSS/INS train integrated
positioning method. Proc. of International Conference
on Mechatronics and Automation (ICMA), ISBN: 978-
1-4244-5140-1, pp. 1686-1689. Xi’an, China.
Palit, S. K., Bickerstaff, S. and Langmaid, C. (March,
2009). Design of wireless communication sensing
networks for tunnels, trains and building. International
journal on Smart Sensing and Intelligent Systems, Vol.
2, No. 1.
Saab, S. S. (Mar. 2000). A map mathing approach for train
positioning part I: Development and analysis. IEEE
Transactions on Vehicular Technology, vol.49, no 2,
pp. 467-475.
Saab, S. S. (Mar. 2000). A map mathing approach for train
positioning part II: Application and experimentation.
IEEE Transactions on Vehicular Technology, vol.49,
no 2, pp. 467-475.
Wenlong, X., Haige, X. and Hongjie, Y. (2002). TETRA
protocol interfaces features and potential applications
in railway. Proc. of the IEEE Region 10 Conference
on Computers, Communications, Control and Power
Engineering (TENCON). ISBN: 0-7803-7490-8, pp.
1086-1088. Beijing. China.
WINSYS2012-InternationalConferenceonWirelessInformationNetworksandSystems
328