=Paper= {{Paper |id=None |storemode=property |title=A Data Transfer Model of Computer-Aided Vehicle Traffic Coordination System for the Rail Transport in Ukraine |pdfUrl=https://ceur-ws.org/Vol-1000/ICTERI-2013-p-178-186.pdf |volume=Vol-1000 |dblpUrl=https://dblp.org/rec/conf/icteri/Arkatov13 }} ==A Data Transfer Model of Computer-Aided Vehicle Traffic Coordination System for the Rail Transport in Ukraine== https://ceur-ws.org/Vol-1000/ICTERI-2013-p-178-186.pdf

A Data Transfer Model of Computer-Aided Vehicle
Traffic Coordination System for the Rail Transport in
Ukraine

Denis B. Arkatov1
1
National Technical University "Kharkov Polytechnic Institute", Kharkov, Ukraine

denarkatov@gmail.com

Abstract. This paper gives a general layout of subsystem operation used for the
rolling stock traffic coordination. A principle of selection of information tech-
nologies applied for the realization of the communication and data transfer sys-
tem has been described. GSM and GPRS technologies that are used for the
transmission of navigation information on a vehicle locus and for the informa-
tion exchange in the system have been described. To provide a required level of
transmission capacity through the frequency reuse the guard period between the
base transceiver stations has been calculated. To calculate the GPRS channel
capacity and that of information exchange via the Internet GPRS network an al-
gorithm for the simulation modeling of packet arrival in the prescribed time
space has been constructed. Using this algorithm an experiment was carried out
for different intensity values of packet arrival to the system.

Keywords. Traffic coordination, information technologies, transmission capac-
ity, algorithm, simulation model, communication channel

Key terms. Development, Model, MathematicalModel

1 Introduction

The traffic control automation is a topical and up-to-date problem of the rail transport
in Ukraine. An important part of this problem is the development of algorithms for
the coordination and control of the large amount of rolling stock, situated in a zone of
railway traffic control points. An important requirement set to such algorithms is to
provide safe and regular traffic of the whole collection of trains and to make optimal
decisions from the economic standpoint.
To provide safe traffic for rolling stocks at the stage of departure or arrival a cer-
tain (permissible minimum) time interval should be provided, which is always taken
into consideration while making a train timetable. Despite this fact the train timetable
is sometimes disturbed due to many reasons and situations arise when the rolling
stocks are found to be undivided by the safe time interval and the groups of conflict-
A Data Transfer Model of Computer-Aided Vehicle Traffic Coordination … 179

ing trains are formed. In the case of high traffic intensity the groups of conflicting
trains at stations or at a train overtaking locus can be rather large. On the other part
the malfunction of train timetable can result in subsequent traffic disturbance (train
delay, train overtaking conflicts, etc.) The above problems can be resolved through
the automation of operative traffic control of rail transport. The experience gained by
foreign countries [1,2] shows that the efficient solution of traffic control with regard
to the rail transport is only possible if the latest information technologies are used.
The use of inexhaustible potential of railway information systems for the benefit of
entire transport system of the country allows to reduce control costs required for the
management and realization of domestic and international traffic. This also provides a
considerable improvement in the quality of transport and logistics-related services and
safety of traffic.
The examples of computer-aided systems that can fully or partially solve the above
problems are the European Train Control System (ETCS) [3], American Positive
Train Control (PTC) [4] and ILSD-U system [5] used by Russia.
A specific feature of the developed computer-aided system used for the rolling
stock (RS) traffic coordination is the introduction of contemporary satellite technolo-
gies, communication and data transfer systems into a routine work of rail transport in
Ukraine. The information exchange technology can be described as follows.
A GPS/GPRS –modem receives the navigation information from satellites, after
that using the GPRS technology the coordinates, current speed and other information
are transmitted via the operator’s server of mobile communication to the database
server. The RS locus data are transmitted to the workstation of railway station dis-
patcher and to the workstation of railway dispatcher.
The data should be processed in a real time mode. An important aspect is that a
time interval during which these data gain currency should be taken into considera-
tion. This fact considerably constrains not only the algorithm used for the solution of
coordination problem but also data transfer technologies.
This paper consists of several sections. First of all, the technologies used for the
data collection, processing and transfer will be described. Then, we will estimate the
channel data transfer capacity and do appropriate computations that prove the ob-
tained results.

2 Data Transfer, Processing, and Collection Technology

At the present time the railway adopts the computer-aided system (CAS), which in-
cludes the following basic subsystems: onboard intellectual system, which provides
positioning, control and information support for the rolling stock; surface intellectual
system (SIS), which provides control and coordination in a real time mode; communi-
cation and data transfer system based on the mobile GSM communication; navigation
charts that reflect a real railway infrastructure. The SIS structure and CAS system on
the whole are given in Fig.1.
A special place in SIS is occupied by the information system designed for the roll-
ing stock traffic coordination, which is the subject of this research. The navigation
180 D. B. Arkatov

data collection system is used for the automatic identification of a rolling stock locus
and also for securing the safety of traffic. The information about a rolling stock locus
is required for the optimal use of traffic and carrying capacity of railways and for the
elimination of dangerous situations (dangerous passing approach, passing the traffic
lights with forbidden signals, siding motion, exceeding the allowable speed in the
places of its restriction, etc.) A locus of each RS is transmitted to the traffic control
service via the navigation system for further data processing and traffic control. The
basic principles of monitoring are the safety of traffic, time fulfillment of a transporta-
tion schedule and costs minimization.
System of connection and data

Special navigation maps
transmission

Fig. 1. CAS and SIS structures

In order to create the appropriate computer-aided communication and data transfer
system for the rail transport the adoption of communication standard is required to
meet the entire system operation requirements, in addition to the use of satellite navi-
gation. The required communication system should provide high safety, reliability
and solutions proved in practice and it should also be an innovative and high perform-
ance system. The costs required for the total system deployment should be reduced to
a minimum.
A Data Transfer Model of Computer-Aided Vehicle Traffic Coordination … 181

Taking into consideration the European experience gained in the realization of
similar data transfer systems telecommunication technologies should be developed
meeting the GSM-R standard (Global System for Mobile Communications-
railway)[6,7], which engineering and economic efficiency has been proved not only
by tests but also by real application for different railway systems in developed Euro-
pean countries. However, at this stage of the development of information technologies
in Ukraine we can make to the conclusion that the realization of formulated problem
is possible and it is highly recommended to use the GSM mobile communication
standard. The GPRS technology is recommended for the data transfer.
The GPRS networks split the transmitted information into individual packets that
are delivered from a transmitter to a receiver. If errors have been detected the received
packets can be transmitted once again. The original message is designed by the re-
ceiver party using the obtained packets. The data transfer in packet- switching net-
works differs from the data transfer in channel-switching networks in the way that the
required channel resource is allocated exclusively for the time of transmission of ap-
propriate information packets. The rest of the time it is at disposal of a network. In
the case of GSM/GPRS networks this allows us to use one physical channel for the
transmission of packets to several subscribers and to simultaneously allocate several
physical channels for the transmission of packets to one subscriber. The packets are
transmitted aside from each other in different directions.
The GPRS defines the effective use of a channel resource. The same physical
channel is provided for the group of subscribers to receive messages sent from the
base transceiver station (BTS) to a mobile station (MS) and packets are transmitted as
soon as they arrive depending on the volume of information and the priority of sub-
scribers. Each packet contains an identifier or address, which is used for the delivery.
A subscriber is continuously connected to the packet network, which provides for him
a virtual channel. It becomes a real (physical) radio channel during the packet trans-
mission. The rest of the time this physical channel is used for the transmission of
packets of other users.
Due to the fact that the same channel resource is used by several subscribers and
during the communication session the packets of different users can simultaneously
arrive the waiting list of transmitted packets can be originated, which will result in the
communication delay. The allowable value of packets delay is one of the attributes
defining the quality of a subscriber service.
The paper [8] defines three classes of delay depending on the delay norms and
packet length. The top priority is assigned to the Class 1, the normal priority is given
to the class 2 and the class 3 enjoys the least priority. The delays have not been de-
fined for the Class 4, because the service of packets of this class is performed adher-
ing to the “best effort” principle.
The intensity values of packets arrival can be selected on the basis of statistical re-
search. A statistics shows that a number of packets arriving per time units to the input
of GPRS circuit changer can vary in a wide range from hundreds of packets /s at input
sites to several thousands of packets/s at backbone sites.
To provide the appropriate level of data throughput we use the basic principle of
the construction of cellular communication networks, which is based on the frequency
182 D. B. Arkatov

reuse [8]. The main essence of it consists in that the neighboring (adjacent) cells of a
mobile communication system use different frequency systems and in non-adjacent
cells located at sufficient distance from each other the used frequency bands are re-
peated. In practice the cities and regions with a solid cellular coating use clusters in
which each cell is divided into the three sectors, using the directional radiation an-
tenna with the directional pattern width of 120.
The base stations that allow frequency reuse are located at a distance D from each
other. This D distance is measured between the centers of hexagonal cells and it is
called a guard interval.
Proceeding from geometrical reasons the parameter D can be defined as follows:

D  R 3 ,

where R is a radius of circle circumscribed around the regular hexagon;  is a coeffi-
cient of the frequency reuse. The D ratio is defined as a reduction factor of channel
R
noises.
Thus, for the optimal frequency reuse and an increase in the GPRS channel capac-
ity in Ukraine the guard interval should be:

D  15 3  3  45 (km)

3 Case-Study for Capacity Evaluation

The developed computer-aided system used for the rolling stock traffic coordination
should take into consideration, while transmitting the navigation information to the
mobile operator server, not only the capacity of GPRS channels but also that of cable
or fiber-optic channel of the Internet network, which is used for the transmission of
data to the database server.
By analogy to the mobile communication it is necessary to consider a problem re-
lated to the average duration of delays in the packet switch.
The term “packet switch” implies here a concentrator (statistical multiplexor), a
virtual packet switch (network X*25, Frame Relay, and ATM network) and a router
(IP network).The packet switch can be represented as an element with many input and
output channels (switch/router). Using the Kendall notation such network elements
can be presented by queuing systems of G/G/1 or G/G/n type (arbitrary probabilistic
distributions describing the incoming stream of customers (in our case packets or
protocol blocks) and the time of their serving time. (Let us note that models with one
server, i.e. G/G/1 are often used for the analysis of packet switches).
Let’s assume that in the general case applications arrive to the system input in
compliance with the Poisson distribution law, whose service time is an arbitrary
value. Then the average queue length in the system with the infinite buffer size
(M/G/1) is calculated using the classic Khintchine-Pollaczek formula [9]:
A Data Transfer Model of Computer-Aided Vehicle Traffic Coordination … 183

q  1  C s2 
tq   t s  1    (1)
  2(1   ) 

where q is an average queue length in the considered system (including protocol
blocks PB);
 is the M/G/1 system load intensity, <1;


 ,  are the intensity values of the PB arrival and service in the system, accordingly;
t s is an average time of PB service in the system;
Dt s 
C s2  is a quadratic coefficient of service time variation equal to the ratio of
t 2 2
service time variance and squared expectation value.
The values required for computations using formula (1) were obtained through the
simulation modeling of data transfer in the Internet network, which algorithm is given
in [1].
A one-channel queuing system (QS) of the M/G/1 type has been taken as an exam-
ple. The arrival of the Poisson stream of applications with the constant service time
 j has been simulated. The numerical experiment, which was carried out, allows us
to come to the conclusion that there is no loss of data packets.
The constructed graphs show the time of the arrival of customers in a queue and
the time of channel teardown from the previous application at the application arrival
1
intensities of   1 (see Fig.2.a) and   (see Fig.2.b)
10 20

(a) (b)
Array 1 – The estimated time of the channel clearing
Array 2 – Actual time of the channel clearing
1 1
Fig. 2. QS simulation with an intensity of   and  
10 20

The given Figures show that an increase in intensity of arriving customers does not
result in the lack of channel capacity. This proves the correct choice of information
184 D. B. Arkatov

technologies for the realization of computer-aided system used for the rolling stock
traffic coordination.
To define the data transfer channel capacity in the Internet network the network
duration and delay should be calculated.
Due to the fact that the data transfer channel-switching center services packets it
can be simulated using the system with a constant service time of a M/D/1 type.
  
tq  ts  1  .
 21    

At   1 and   1
10 5

 1
 
 2

Let t in the given example be equal to 0,05. As a result we get:
 1
tq  0,05  1    0,075 ,
 2
which corresponds to the first service class.
At   1 and   1
20 15

 3
 
 4

Let t in the given example be equal to 0,05. As a result we get:

 3
tq  0,05  1    0,125 ,
 2

which corresponds to the first service class.

4 Conclusions

Today the railways are the major branch of the economy of Ukraine and they serve as
a basis of the Ukrainian transportation system. Due to the rapidly changing demands
for freight services and carriage of passengers a permanent control over the required
amount of rolling stocks should be investigated. The proper amount of such vehicles
can be defined through the traffic analysis.
The information on the timetable of railway traffic and actual amount of rolling
stocks involved in traffic serves as a source information for further computations of
the capacity of a communication and data transfer system of the computer-aided sys-
tem. This will provide the fulfillment of actuality condition of obtained data and the
solution of the traffic coordination problem in a real time mode.
A Data Transfer Model of Computer-Aided Vehicle Traffic Coordination … 185

To provide the functioning of computer-aided system on the whole the following
algorithms should be realized. The algorithm used for the determination of the amount
of conflicting trains is the first step towards the problem solution. It allows for the
detection of those rolling stocks for which conflict-free conditions are not observed
due to different reasons (timetable violation, technical malfunction, etc.). A decompo-
sition algorithm is required for the reduction of dimension of a solved problem that
would allow the reduction of time interval of data acquisition. The obtained data are
used for the generation of control action and for the reduction of loading for the used
transmission channel, providing thus the observance of principles of control in a real
time mode. An algorithm of the problem related to the coordination of rolling stock
traffic allows for the elaboration of such control actions that provide not only the
elimination of a definite conflict but also the fulfillment of conflict-free conditions for
the definite rolling stock in the future. An algorithm of data transfer in a real time
mode is the basis of information exchange between the rolling stock and the database
server. Our further research is targeted at the development of algorithmic support of
communication and data transfer system to solve the problem on the coordination of
traffic of rolling stocks.
This paper gives the description of GSM and GPRS technologies that are used for
the transmission of navigation information on the locus of rolling stocks and also for
the information exchange in the system. It has been noted that in order to provide the
required capacity level we applied the main principle used for the construction of
cellular communication networks, i.e. the frequency reuse. The cluster structure with
a template of 3/9 has been constructed and the guard interval between the BSs used
for the mobile communication in Ukraine has been calculated.
To calculate the capacity of GPRS channels and that of information exchange
channel in the GPRS Internet network the algorithm of simulation modeling of arriv-
ing packets during the prescribed time interval has been constructed. On the basis of
this algorithm we carried out an experiment for different intensity values of packets
arrival to the system. Using the analysis data we can come to the conclusion that an
increase in the intensity of arriving customers causes no lack of channel capacity.
Our further research is targeted at the development of an algorithm for the data
processing in a real time mode, and also at the development and testing of the infor-
mation system designed for the rolling stock traffic coordination. We will also delve
into the computation of efficiency estimates to evaluate the introduction of informa-
tion technologies into the rail transport operation.

References
1. Arkatov, D. B.: Models and methods for automation of dispatching management for rail-
ways of Ukraine. Modeli i metodi avtomatizacii dispetcherskogo upravleniya dlya
zheleznodorozhnogo transporta Ukraini, Vostochno-Evropeyskiy zhurnal peredovikh
tekhnologiy, Eastern European journal of Enterprise Technologies, N 1/10 (61), pp. 61–63
(2013)
2. Arkatov, D. B.: Synthesis models of coordination of movement of mobile railway transport
of Ukraine. Sintez modeley koordinacii dvizheniya podvizhnikh sredstv
186 D. B. Arkatov

zheleznodorozhnogo transporta Ukraini, Vostochno-Evropeyskiy zhurnal peredovikh
tekhnologiy, Eastern European journal of Enterprise Technologies, N 4/3 (58), pp. 58–60
(2012)
3. ETCS requirements specification and validation: the methodology,
http://www.era.europa.eu/Document-Register/Documents/ETCS_methodology_v_1_2.pdf
(accessed 20 February 2013)
4. Positive train control (2013), http://en.wikipedia.org/wiki/Positive_train_control (accessed
20 February 2013)
5. Integrated locomotive safety device – unified, http://www.irz.ru/products/20/70.htm
(2012) (accessed 20 February 2013)
6. GSM-R (2013), http://en.wikipedia.org/wiki/GSM-R (accessed 20 February 2013)
7. GSM technical specification, http://www.ttfn.net/techno/smartcards/gsm11-11.pdf (1995)
(accessed 20 February 2013)
8. QoS in GPRS, http://doc.utwente.nl/18117/1/00000039.pdf (1999) (accessed 20 February
2013)
9. Tijms, H. C. A first course in stochastic models / Henk C., Tijms.p. cm. Includes biblio-
graphical references and index.