149
Chapter 8
Novel Handoff Scheme
with Multiple-Input and
Multiple-Output for
Communications-Based
Train Control Systems
Hailin Jiang, Victor C. M. Leung,
Chunhai Gao, and Tao Tang
Contents
8.1 Introduction .............................................................................................150
8.2 Overview of the CBTC Communication Hando Procedure ..................152
8.2.1 Hando Latency of 802.11 and the Communication Latency
in CBTC Systems .........................................................................152
8.2.2 Features of Hando in CBTC Communication System ...............154
8.3 Proposed MAHO Scheme ........................................................................154
8.3.1 MIMO Transmission in the HandoProcedure
intheMAHOScheme .................................................................155
8.3.1.1 Physical Layer Processing ................................................155
8.3.1.2 Synchronization in the Downlink...................................159
8.3.2 Communication Latency in WLANs ...........................................161
8.4 Analysis of Hando Performance ............................................................. 162
8.4.1 Wireless Channel Model ...............................................................162
150 ◾ Advances in Communications-Based Train Control Systems
8.1 Introduction
e IEEE 802.11 standard for wireless local area networks (WLANs) has been
developed to provide wireless access in the oce and campus environments.
However, the procedures for hando between access points (APs) are not well sup-
ported. When a train moves along the rail, its mobile station (MS) would need to
switch from one AP to the next frequently to guarantee continuous data transmis-
sions between the train and the wayside devices, because the coverage of each AP is
quite limited. In general, during the hando procedure data packets will be lost if
no AP forwarding schemes are implemented. is will have serious impacts on the
safety and eciency of train control.
A lot of research has been done on the WLAN hando algorithms. e hando
procedure is divided into three stages [1]: the probe stage, the searching stage, and
the executing stage. In [1], some 802.11b parameters are further adjusted to reduce
the hando interruption time. In the current version of the IEEE 802.11 stan-
dards[2], the formats of probe request and probe response frames are dened and
the fast basic service set transition schemes including authentication and reassocia-
tion schemes are given. In [3], a channel scanning scheme in WLAN hando is
proposed, where the neighbor cells were cached in the buer to reduce the hando
latency in the probe stage. It is proposed in [4] that each MS continuously tracks
nearby APs by synchronizing short listening periods with periodic transmissions
from each base station. In this way, the station can pre-associate with new APs to
reduce the hando latency. A location-based hando scheme is proposed in [5],
and some congurations of the parameters in 802.11 networks are discussed to
reduce the hando latency in the scanning stage. An integrated design approach
is proposed to jointly optimize hando decisions and physical layer parameters to
improve the train control performance in CBTC WLAN systems in [6–8].
ese schemes can reduce the hando latency eciently, but all of them are
so-called break-before-make schemes, where the station needs to dissociate with
the old AP rst, and then nd a new AP and associate with it. In this procedure,
the data transmissions will be interrupted and the transmitted data will be lost.
In a CBTC system employing WLAN technology, the data transmitted include
8.4.2 Optimal Hando Location ........................................................... 163
8.4.3 Error-Free Period ..........................................................................164
8.4.4 FER of the Hando Signaling ......................................................166
8.4.5 Impacts on Ongoing Data Sessions ...............................................167
8.5 Simulation Results and Discussions .........................................................167
8.5.1 Analysis of the Hando Latency ................................................... 167
8.5.2 Error-Free Periods of Traditional Hando Schemes ......................169
8.5.3 FER of Hando Signaling with Dierent Data Rates ................... 170
8.6 Conclusion ...............................................................................................173
References .........................................................................................................174
Novel Handoff Scheme with MIMO ◾ 151
safety-related train control and train position information, and the loss of the data
will lower the performance of CBTC control severely and aect the system safety.
In [9–11], multiple radio transceivers are used to eliminate the hando latency
completely. However, these schemes are required to use two or more radio units to
scan and pre-associate with new APs, which will increase the cost and complexity
of the communication system.
Multiple-input and multiple-output (MIMO) has been the key technology in
mainstream wireless communication technical standards such as 3rd Generation
Partnership Project (3GPP) Long-Term Evolution (LTE) and IEEE 802.11n [12].
Some hando schemes applying MIMO technologies are proposed in [13,14].
In[13], the handover decision method based on detecting the number of antennas
is proposed. In [14], the beamforming and positioning-assisted handover scheme is
proposed where both the source eNodeB and the target eNodeB switch their work-
ing mode from omnidirectional to beamforming to improve the handover success
probability when the train moves into the overlapping region.
In this chapter, we propose a MIMO-assisted hando (MAHO) scheme for
CBTC WLAN system with two antennas or more antennas con–d on train and
each AP. e distinct features of the work are given as follows:
1. Considering the features of the urban railway communication system, the
triggering of the hando is based on the location of the train. In the scheme,
the MS does not hand o to the new AP by comparing the signal strength,
signal-to-noise ratio (SNR), or any other signal transmission-related param-
eters of the APs. Instead, it will start the hando procedure when the train
has reached the hando location. e location-based hando is proposed to
take advantage of the fact that the train in the CBTC system can acquire its
locations accurately in real time.
2. e train position information from the MS is sent simultaneously with hand-
o signaling by means of MIMO multiplexing. e signaling and data packets
are recovered at the APs by means of MIMO signal detection algorithms such
as vertical-bell laboratories layered space-time (V-BLAST) [15] algorithms.
3. e train control information from dierent APs is sent in the space–time
block code (STBC) diversity mode, where simple linear interference cancel-
ing algorithms are used to cancel the interference caused by the concurrent
transmissions.
4. e hando performance, including the frame error rate (FER) of the hand-
o signaling, the hando latency, the error free period, and the impacts on
the intersite distance, is analyzed and compared with traditional hando
schemes.
In the proposed scheme, the hando procedure is triggered by the location of the
train. When the hando is triggered, the station keeps connection with old AP by
one antenna on the train and set up a connection with new AP by another antenna
152 ◾ Advances in Communications-Based Train Control Systems
at the same time. e interference at the station between two APs is canceled by
MIMO signal detection techniques and interference canceling algorithms. In
CBTC environments, the APs are planned to assure that even in the border of
coverage area the received power exceeds the minimum receiver sensitivity by more
than 10dB [5], which guarantees the reliability of the data transmission and makes
the signal detection algorithm feasible. With the SNR over certain thresholds, the
MIMO signal detection and interference cancelation algorithms can be reliably
applied and the “make-with-break” can be realized without multiple radio trans-
ceivers at the MS.
e rest of the chapter is organized as follows: An overview of hando proce-
dure in CBTC systems is given in Section 8.2. e MAHO scheme is presented
and the synchronization and interference canceling algorithms are described in
Section 8.3. e hando performance is analyzed and compared with existing
schemes in Section 8.4. Simulation results are presented in Section 8.5. Conclusion
is drawn in Section 8.6.
8.2 Overview of the CBTC Communication
Handoff Procedure
8.2.1 Handoff Latency of 802.11 and the
Communication Latency in CBTC Systems
e timing diagram of the hando procedures in 802.11 is shown in Figure8.1.
ere are three stages in the hando procedure: the probe, authentication, and
reassociation procedures. In passive scanning mode, the MS receives the periodic
Beacon frame from APs in the probe stage, then nds the suitable AP to set up
the communication link. e transmission period of the Beacon frame is about
100ms, so the latency in this stage is several hundreds of milliseconds, which is too
long for the CBTC service. In the active scanning mode, in general, the MS broad-
casts the probe request at all channels one by one, then waits for the probe response
Probe
request
GGGG
Handoff latency
GGG G
Authentication
request
Authentication
response
Reassociation
response
Reassociation
response
Data
transmission
Probe
response
Probe
response
Probe
response
G = DIFS
MS
Figure8.1 WLAN handoff timing diagram.
Novel Handoff Scheme with MIMO ◾ 153
from APs. As there are 11 channels dened in 802.11 [2], the latency in this stage is
quite long. Fortunately, only one channel is used in CBTC communication system
for all APs along the rails. e MS on the train is generally congured to only scan
one channel to reduce the scanning latency.
At the authentication stage, the MS authenticates with the best AP found in
the rst stage. When the 802.1X authentication is implemented in the WLAN,
the latency is more than 150ms [16], which is too expensive for the CBTC com-
munication system. For the authentication procedure of over-the-air fast transition
protocol between APs with four-way handshake dened in 802.11 standard [2], the
latency in this stage is about 40ms [16].
CBTC systems have stringent requirements for communication latency. e
data between the train and the ZC are transmitted periodically. Because a train
needs to get the location information of the train ahead of it, trains and ground
equipment must communicate with each other in every communication period.
When the communication latency is short, a train gets the position of the train
ahead of it in every communication period, and the minimum train headway (min-
imum distance between two successive trains) is calculated considering the braking
distance, the constant safe distance, and the train length. When the communica-
tion latency is long, a train may not be able to get the position of the train ahead
of it at some communication periods. erefore, the train has to suppose that the
preceding train is still at the position which was got from the previous communica-
tion period. e consequence is that the train has to brake in advance, which will
aect the utilization seriously.
When a train moves between successive APs, the channels between the train
and the wayside APs change rapidly due to the fast movement of the train. And
this will result in rapid changes of the received power and signal quality. e rapid
changes of the communication link also have serious eects on the frequent hand-
o procedures, which may result in long transmission latency.
In global system for mobile communication/railway (GSM-R) communication
system, there are two special quality-of-service (QoS) targets to ll for the train’s
operation requirements [18,17]. e rst is the transmission interference period,
which is the period during the data transmission phase of an existing connection
in which no error-free transmission of user data is possible. All the packets will be
lost during the hando procedures if there are no forwarding schemes between
APs. It is clear that the transmission interruption period is just the hando latency
in CBTC communication system because the SNR margin is generally set to be
large enough that packets will not be lost in normal transmission process. And the
interruption due to hando procedures will result in the increase of the end-to-end
communication latency in CBTC communication system.
e other parameter is error-free period, which follows the transmission inter-
ference period to retransmit user data in error and data units waiting to be served.
ere are no standards or drafts for CBTC communications in the urban railway
transition system. However, it is clear that the transmission interruption period,
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.