93
Chapter 6
Communication
Availability in
Communications-Based
Train Control Systems
Li Zhu and F. Richard Yu
Contents
6.1 Introduction ...............................................................................................94
6.2 Related Works ............................................................................................95
6.3 Proposed Data Communication Systems with Redundancy .......................95
6.3.1 Overview of CBTC and Data Communication System ..................96
6.3.2 Proposed Data Communication Systems with Redundancy ...........97
6.4 Data Communication System Availability Analysis ....................................99
6.5 Modeling Data Communication System Behavior with DSPNs ...............103
6.5.1 Introduction to DSPNs ................................................................. 103
6.5.2 DSPN Formulation.......................................................................104
6.5.3 DSPN Model Solutions ................................................................105
6.6 Numerical Results and Discussions ..........................................................110
6.6.1 System Parameters ........................................................................110
6.6.2 Model Soundness ..........................................................................111
6.6.3 Availability Improvement ..............................................................111
6.7 Conclusion ...............................................................................................113
References .........................................................................................................113
94 ◾ Advances in Communications-Based Train Control Systems
6.1 Introduction
A data communication system, which is the basis for train control, is one of the key
subsystems for communications-based train control (CBTC). A couple of wireless
communication technologies have been adopted in CBTC, such as Global System
for Mobile Communications-Railway (GSM-R) and wireless local area network
(WLAN). For urban mass transit systems, WLAN is commonly used due to the
available commercial o-the-shelf equipment, open standards, and interoper ability
[1–3].
CBTC systems have stringent requirements for communication availability [2].
Whereas in commercial wireless networks, less service availability means less rev-
enues and/or poor quality of services [4], in CBTC systems, it could cause train
derailment, collision, or even catastrophic loss of life or assets [5]. erefore, it
is important to ensure the train–ground communication availability in CBTC
systems.
ere are several WLAN-based CBTC systems deployed around the world, such
as Las Vegas Monorail from Alcatel [6] and Beijing Metro Line 10 fromSiemens [7].
Most system integrators claim that redundancy is used in their systems. However,
they do not reveal the details about redundancy due to condential considerations.
Moreover, availability analysis is largely ignored in the literature of CBTC systems.
In this chapter, we study the availability issue of WLAN-based data communi-
cation systems in CBTC. e contributions of this chapter are as follows:
1. We propose two WLAN-based data communication systems with redun-
dancy to improve the availability in CBTC systems.
2. e availability of WLAN-based data communication systems is analyzed
using continuous-time Markov chain (CTMC) model [8], which has been
successfully used in call admission control (CAC) in mobile cellular networks
[9] and wireless channel modeling [10], among others. e transmission errors
due to dynamic wireless channel fading and handos that take place when
the train crosses the border of two successive access points (APs)’s coverage
areas are considered as the main causes of system failures.
3. We model the WLAN-based data communication system behavior using
deterministic and stochastic Petri net (DSPN) [11], which is a high-level
description language for formally specifying complex systems. e DSPN
solution is used to show the soundness of our proposed CTMC model.
DSPN provides an intuitive and ecient way of describing the com-
plex system behavior and facilitates the modeling of system steady-state
probability.
4. Using numerical examples, we compare the availability of the two proposed
WLAN-based data communication systems with an existing system that has
no redundancy. e results show that the proposed data communication sys-
tems with redundancy have much higher availability than the existing system.
Communication Availability in Communications-Based TCSs ◾ 95
e rest of this chapter is organized as follows: Section 6.2 introduces the related
works. Section 6.3 presents an overview of CBTC and the proposed data commu-
nication systems with redundancy. Section 6.4 discusses the CTMC model, with
its state space for each conguration. Section 6.5 describes the DSPN modeling
approach. Section 6.6 presents numerical results. Finally, Section 6.7 concludes the
chapter.
6.2 Related Works
Documented research has investigated the availability issue of commercial wireless
networks. e authors of [12] analyze the availability and reliability of wireless
multi-hop networks with stochastic link failures. ey provide a method to forecast
how the introduction of redundant nodes increases the reliability and availability of
such networks. A novel wireless communication infrastructure is presented in [13]
for emergency management. Several schemes are proposed in the infrastructure to
improve the system reliability. In [14], techniques for composite performance and
availability analysis are discussed in detail through a queuing system in a wireless
communication network. ree modeling approaches are illustrated for composite
performance and availability analysis in their work. In [15,16], dierent WLAN
topologies are considered in home WLANs, and the simulation results show that
redundancy can greatly improve the system availability of home WLANs.
ere are also some works about availability in CBTC data communication
systems. For trunk railway systems, the authors of [17] study the GSM-R systems
with stochastic Petri net (SPN) model in European train control systems (ETCSs).
A similar model for ETCS with a unied modeling language (UML) software
tool, STOCHARTS, is proposed in [18]. Based on the parameters taken from the
ETCS standards, the results show that the communication requirement in ETCS
is very strict for train control systems. For urban mass transit systems, the authors
of [5] study the CBTC data communication system with simple redundancy, and
the reliability and availability are analyzed. However, the redundancy schemes in
their study are exploratory. ey do not consider the impact of AP deployment
space on system availability performance, and the analysis approach may not be
comprehensive.
6.3 Proposed Data Communication
Systems with Redundancy
In this section, we rst present an overview of CBTC and its basic conguration
of data communication system based on WLANs. en, the proposed data com-
munication systems with redundancy are presented.
96 ◾ Advances in Communications-Based Train Control Systems
6.3.1 Overview of CBTC and Data Communication System
A simple view of a CBTC system is illustrated in Figure6.1. In this system,
continuous bidirectional wireless communications between each station adapter
(SA) on the train and the wayside AP are adopted instead of the traditional xed-
block track circuit. e railway line is usually divided into areas. Each area is
under the control of a zone controller (ZC) and has its own wireless transmission
system. e identity, location, direction, and speed of each train are transmitted
to the ZC. e wireless link between each train and the ZC must be continu-
ous to ensure that the ZC knows the locations of all the trains in its area at all
times. e ZC transmits to each train the location of the train in front of it, and
a braking curve is given to enable it to stop before it reaches that train as well.
eoretically, two successive trains can travel together as close as a few meters in
between them, as long as they are traveling at the same speed and have the same
braking capability.
Data communication systems are primarily designed to connect each com-
ponent of CBTC systems: ZCs, APs along a railway, and train aboard equip-
ments. A basic conguration of WLAN-based data communication system is
shown in Figure6.1. Following the philosophy of open standards and interop-
erability [2], the backbone network of the data communication system, which
mainly includes Ethernet switches and ber-optic cabling, is based on the IEEE
802.3 standard. e wireless portions of the data communication system, which
consist of APs along the railway and SAs on the train, are based on the IEEE
802.11.
In contrast to the backbone network, the wireless links are more prone to chan-
nel degradation in railway environments. e low reliability of the wireless portion
of the data communication system is mainly caused by the following:
Coverage of AP1
Coverage of AP2
ZC
Train braking curve
Backbone
AP1
SA SA SA
AP2
Figure6.1 CBTC system.
Communication Availability in Communications-Based TCSs ◾ 97
1. Transmission errors due to dynamic channel fading in railway environments.
2. Handos that take place every time the train crosses the border of two suc-
cessive APs’s coverage areas. e communication link will be lost for a short
period of time during the hando process.
In order to improve the communication availability in CBTC systems, we propose
two data communication systems with redundancy, which will be presented in
Section 6.3.2.
6.3.2 Proposed Data Communication
Systems with Redundancy
In order to describe our system more easily, we rst dene two dierent kinds of
links for the data communication system.
1. Active link: A link that is currently used between an SA and its associated AP
2. Backup link: A link that is currently not used between an SA and its associ-
ated AP, but can be used in case of failure of the active link
If the SA is in the coverage of more than one AP, the active link is the one that the
SA associates with the AP, which provides the better signal-to-noise ratio (SNR).
In case of failure of the active link, the SA can associate with another AP. We call
this link a backup link.
e data communication system with basic conguration is shown in Figure6.1.
In this system, only one AP with directional antenna is deployed in each loca-
tion. e head directional antenna of the train is connected to the SA. ere is
only one active link for the train at any time. No backup link exists in the basic
conguration.
We propose two data communication systems with redundancy congu-
rations. e rst proposed system is shown in Figure6.2. In this system, two
APs each with one directional antenna (facing in opposite directions) are used
in each location, which are, respectively, connected to two backbone networks
(i.e., Backbone network 1 and Backbone network 2). Two directional antennas
(i.e.,head antenna and tail antenna) are connected with two independent SAs
on the train. e two APs in each location have dierent service set identiers
(SSIDs), and the two SAs on the train also have dierent SSIDs. Normally, SAs
can only associate with the AP with the same SSID. No backup link exists in this
redundancy conguration. Instead, there are two active links between the train
and the ground. e communication will not be interrupted if only one of the
active links fails because of deep channel fading or hando. We assume that there
is no chance that handos happen at both SAs at the same time. is assump-
tion is reasonable in practice, because the APs can be appropriately deployed so
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.