Chapter 1
Introduction
1.1 Introduction
Communication networks in the railway sector are critical to the system operation and have restricted requirements for reliability and safety. These types of networks are commonly characterized as “mission critical.” Also, rail communication networks have requirements for interoperability with legacy technology and long life cycle support.
Until recent years, data transmission between trains and ground did not involve the train by itself as a communication component. New requirements from operators for flexibility and customer services are now transforming the train into communicating equipment that must couple to fixed installations for persistent data exchanges. Train integration in a global communication framework is currently required for multiple kinds of services, especially traffic control and supervision services, those addressing interoperability issues, and dynamic passenger information services.
Dynamic information is required by travelers who consider that simple information such as train destination and scheduled arrival time displayed on internal screens is far from sufficient. For these travelers, the train should also display dynamic information such as the next station name, delay estimations, and causes of delays. By integrating the train into a global communication framework, the train itself can receive information directly from the ground.
1.2 Communication Classes and Related Requirements
1.2.1 Real-time data
Real-time data are used for information that is valid only for a limited period of time. This class of data is used for unidirectional exchanges between one source and one or many receivers. Real-time data can be exchanged using protocols that guarantee delivery reliability. However, the maximum delay of the network must be guaranteed. Excessive delay leads processing system to consider the data to be outdated.
1.2.2 Non-real-time data
Specific protocols must be used when message delivery across the network must be guaranteed. Command messages typically use this class of data communication, since the sender must be informed of the transaction result. Such data exchanges imply a form of acknowledgment; thus a single command results in multiple messages on the network. In most cases the delays induced by the network lead to unpredictable latency times; therefore, the term “non-real-time” is used.
Non-real-time messages are also to be used when evolution of data as a function of time is not continuous. In this case non-real-time data exchange is a way to limit the bandwidth usage, since no communication takes place when data are not evolving.
1.3 Expected Services from a Railway Communication System
Railway services may be classified into core services and additional communication services. The core services are usually include critical railway communications, train operational services and data applications. The additional communication services include passenger experience services and all the business process support services such as voice and data train crew communications and train support applications.
There are many expected services from a railway communication system each of them has its specific requirements. Some of these services are: Automatic Train Control (ATC), Passenger Information System and On-board Internet access [1]. The most important service, which will be focused on, is ATC.
ATC is responsible for exchanging train control signals between ground and train. Also, it covers various applications, from train protection to full automatic train driving. ATC systems are based on the real-time data class. Its applications require the communication to guarantee absolute data integrity. The amount of data exchange required for ATC is very low (most often a few kilobits per second). Latency for ATC applications is required to be low to medium (one second is generally considered as a maximum).
1.4 Requirement of Railway Wireless Communication
The wireless communication network for railway has its own requirements for mission critical operations such as accident prevention of trains, immediate reaction to emergency and on-time operation. It should achieve not only safe operation of trains but also advanced railway services provided in the future. Therefore, it should meet the general communication requirements as follows [2]:
High-speed movement:
In general, the maximum speed of the high-speed train is 300 km/h or more. In addition, stable wireless connection should be guaranteed at the moving speed of 500 km/h or more in the future.
Broadband wireless transmission technology:
The railway communication may include video transmission function (for real-time monitoring of passenger and/or car states in case of unmanned operation) as well as various railway customer services. As a result, the broadband wireless transmission technology that can transfer a large amount of data in real time is essential.
To provide high data rate connections between the User Equipment (UE) and the base station, some relay station may be used to relay the data with potentially higher data rate. Thus, the deployment of relay station becomes an important strategy in increasing data rate while preserving the cellular systems coverage in the future wireless systems.
Relay stations can be classified into fixed relay stations and mobile relay station (MRS) according to the moving state of a relay station. MRSs is installed on the high-capacity vehicle such as the HST. If such MRSs are integrated into cellular systems, UEs will have a better way to connect to the cellular systems.
Low latency:
Shorter communication delay time is required if we consider the increasing speed of trains in the future. In addition, the voice call setup and connection should be quickly performed with a delay less than one second for an emergency event.
Network reliability and availability:
The communication wireless network for railway should provide the network reliability and availability. Network reliability refers to the stable transmission of data information for the safety of railway operation; while the network availability stands for continuous use in spite of various situations of failure.
Quality of Service (QoS):
Existing wireless networks support various QoS depending on different types of traffic and services. In particular, the train control signal which can directly affect the safe operation of trains is mixed with general information signals in the integrated wireless network for railway. Therefore, for more efficient and safe management of resources in a wireless network, an adequate QoS control is required.
1.5 Wireless Systems Applicable to Railway Communication Systems
1.5.1 GSM-R:
The GSM system is the most common cellular system across the world. It offers the interesting capability of supporting both voice and data transmission. Depending on the propagation conditions, the bandwidth available for the data channel can change from 300 bps to 9.6 Kbit/s. This restricts the use of GSM to low bandwidth application profiles.
Since a GSM network is shared between multiple users, there is a non-negligible risk that a Base Station (BS) reached by the train has already allocated all available channels. The risk is lowered by the fact that a user always tries to gain contact with three BSs. It normally selects the nearest one in order to lower power consumption, but it can also connect with a more distant one if no more channels are available on the nearest one, which should lower the risk of communication loss.
The risk of communication loss is generally higher with trains running outside city areas, where the number of BSs is lower. City areas are normally covered by a higher number of BSs. This has a low impact on slow-moving Mobile Stations (MSs), but trains are impacted by the need to change the BS to which they are connected more often. In order to avoid the public GSM limitations for train applications, the GSM-R [3] version has been designed by the "Union Internationale des Chemins de Fer" (UIC) as a universal standard for railway communication.
At the end of the last century, all national railways had their own incompatible analog systems for railway communications. A group of manufactures came together to define a universal standard for railway communication, which resulted in the GSM-R standard. This standard was not only meant to take a leap into the digital age; but also had the goal of interoperability between various railway companies.
Planning a GSM-R network is quite different from a regular GSM network. Normally, a network planner tries to cover a two dimensional area with as little base stations and antennas as possible. But GSM-R doesn't cover a wide area; a train route is rather a long, almost one-dimensional line. That in itself makes GSM-R unique from the perspective of a radio planner.
GSM-R has the same capabilities (voice and data communication) as standard GSM, but it uses a reserved frequency range. The uplink (MS to BS) uses the 876 to 880 MHz band, while the downlink (BS to MS) uses the 921 to 925 MHz band. The entire bandwidth is available for the railway operator, since it is not shared with public users. The UIC recommends that GSM-R BSs are installed along the rail tracks every 3 to 4 kilometers, in order to keep the data rate as high as possible and lower the risk of communication loss when the MS switches from one BS to another.
GSM-R is considered now as an obsolete mobile technology with a number of shortcomings in terms of capacity and capability. These shortcomings become a major issue for railways because they cannot support advanced data services.
GSM-R offers only circuit-switched transmission. This mode of transmission is less efficient than packet-switched. Especially when considering that one of the main applications of GSM-R is delivery of burst control data messages. The lack of packet-switched transmission leads to very low utilization of the GSM-R network.
Another major problem with GSM-R is its insufficient capacity. A small number of channels available for user transmission. This is a consequence of the combined effect of the circuit switched transmission and the reduced band of radio spectrum assigned. In areas with high train concentration such as central train stations there are problems with providing sufficient number of channels to serve all the trains that are to operate there simultaneously. Finally, GSM-R is specifically used for train control instead of passenger communications [4]. GSM-R cannot meet the requirements for future high speed data transmissions.
These shortcomings show how outdated GSM-R technology is from a telecommunication point of view. Now, GSM is being slowly abandoned by commercial operators. This makes GSM-R adaptation by railways becomes invalid. The technology that should be considered as the most likely alternative to GSM-R is LTE-R.
1.5.2 LTE-R:
Railway environments cannot be apart from the current evolution of public communication systems. Advanced broadband wireless communication systems must be developed for high speed rail. GSM-R has an expected end of its lifetime. A new system is thus required to fulfill HST operational needs, offering new services, but still coexisting with GSM-R for a long period of time. The selection of a suitable wireless communication system for HSRs needs to consider performance, service attributes, frequency band, and industrial support.
UIC recommended that GSM-R should be evolved directly into LTE-R [5], which is a broadband railway wireless communication system based on LTE [6, 7] and has the same network structure.
Compared with 3G system, 4G LTE has a simple flat architecture, high data rate, and low latency, making it an acknowledged acceptable bearer for real-time HSR applications. 5G, while currently discussed in 3GPP, will be available only after 2020 and thus is not suitable for the HST timeframe.
In view of the performance and level of maturity of LTE, LTE-R will likely be the next generation of HSR communication systems, and the future vision for HSR wireless technologies will thus rely on Internet Protocol (IP) and broadband.
The UIC announced in 2008 that the presence of LTE communication system influenced the lifecycle of GSM technology and affecting the continuous dissemination and maintenance of GSM-R equipment. The UIC (in Technical Report generated in 2009) presented the results to examine whether LTE communication system would be applicable to the railway wireless communication network.
The main contents were that LTE technology is suitable for the future railway communication network and meet various requirements of railway. It has been considered as the next generation for railway wireless communication network.
There are over 20 years of development separating GSM and LTE technologies, what makes them very different in many aspects. However, here, the focus is put on a few main advantages of LTE that could be highly beneficial from the railway perspective.
The key element differentiating LTE-R from GSM-R is that an LTE-R network is based on packet-switched transmission. It is the first mobile technology adopting the all-IP approach. Packet switched transmission is more flexible in managing available network resources. Hence, it increases network utilization and reduces waste of limited network resources.
LTE includes Quality-of-Service mechanisms that provide packet differentiation. This could be applied to protect railway safety-critical applications such as Train Control System Applications. Another advantage of LTE network is the reduced packet delay. This is achieved by simplifying the network architecture. The LTE network has less logical and physical elements and they are all based on a common technology (IP).
Finally, LTE offers much higher throughput over its radio access thanks to a more advanced radio interface. It consists of a number of improvements that increase spectral efficiency of LTE in comparison to GSM. The additional throughput can be consumed in various ways: to serve more users, to provide more applications or to provide bandwidth-demanding applications, which cannot be provided over the low-rate GSM-R radio interface. Some Technical specification of LTE when used in railway networks are listed below:
High Availability and reliability:
LTE network can ensure higher availability and reliability than GSM-R. Bell Labs of Alcatel-Lucent has conducted analysis on the availability and reliability of the LTE network [8]. The network availability represents that a network should be continuously available even when various failures happen. The reliability means that the data should be transmitted safely and reliably. They are very important for railway wireless network since this network includes a train control function for the punctuality of transport and passenger safety. Self-Organizing capabilities of LTE technology must be taken into account for improving reliability and availability.
Real Time Monitoring:
LTE-R provides video monitoring of front rail track, cabinet, and car connector conditions; real time information monitoring of the rail track conditions (e.g., temperature and flaw detection); video monitoring of railway infrastructures (e.g., bridges and tunnels) to avoid natural disasters; and video monitoring of cross-tracks to detect freezing at low temperatures. The monitoring information will be shared with both control center and the high-speed train in real time. Though some of the above surveillance can be conducted by wired communications, the wireless-based LTE-R system is more cost-effective for deployments and maintenances.
Lossless handover and fast re-connection time:
LTE standard support hard handover mechanisms (discussed in details in Ch. 4), which reduces the complexity of the LTE network architecture. However, the hard-handover (HHO) mechanism does not guarantee any data packet losing in handover process. LTE HHO must fulfil the railway service QoS requirements, especially in high speed scenarios.
Safety:
Due to the IP nature of LTE network, the security risks are related with the proper implementation of security protocols for secure IP transportation to avoid data manipulation in control and user plane. LTE implements mechanisms for mutual authentication of the UE and the network, and also for encoding and checking message integrity in the data communication between terminal and eNodeB.
Railway emergency communications:
When natural disasters, accidents or other emergencies occur, it is required to establish communications quickly between accident site and rescue center to provide voice, video, data, and image transmissions. Railway emergency communication systems use the railway private network to ensure rapid deployment and real time response.
Information transmission of control system:
LTE-R provides real time information transmission of control information via wireless communications to enable compatibility with the train control signal which improves the accuracy of train tracking and the efficiency.
Very low dropped call rate.
Low sensitivity to high train speeds of up to 350 km/hour and beyond.
Support for low bandwidth operational applications, such as train control, with performance at least as good as GSM-R.
1.6 GSM-R and LTE-R Coexistence
As GSM-R support by suppliers is committed until at least 2020, it is expected that the co-existence between LTE R and GSM-R will last for a long time. This can be elaborated from several aspects:
Terminal level:
The future HSR terminal should support both GSM-R and LTE-R. The multi-mode mobile terminal with low complexity is a possible solution.
Access network level:
Direct coexistence of access network between GSM-R and LTE-R would be difficult since they both use different access technologies. It is possible for the two networks to share sites at the first step till complete evolution to LTE-R is performed.
Business level:
LTE-R needs to support the traditional applications of GSM-R.
1.7 Thesis Objective
1. Study of LTE Network Structure.
2. Study of LTE-R Handover Procedures, Triggering Events and Problems.
3. Simulate A3 Handover Triggering Event.
4. Power-Distance Handover Algorithm proposal.
5. An optimized Handover measurement procedure proposal.
1.8 Thesis Outline
The rest of the thesis is organized as follows:
Chapter 2 discuss LTE Network Architecture in details. It reviews both core and access networks then gives a detailed explanation of LTE quality of services and its different bearers. Then it demonstrates LTE protocol stack from both control and user plane perspective.
Chapter 3 mainly discusses handover process in LTE. First, it declares the three phases of general handover process, then it clarifies LTE Hard handover Algorithm with its measurement procedure and metrics. Also, it gives a detailed explanation of Handover Triggering Events.
Then, it gives an example for LTE-R handover showing the effect of changing both train speed and eNodeB transmitted power on the Handover Triggering point.
Finally, it discusses the WINNER II Channel models. It concentrates on D2 moving network scenario which is the best model for high speed trains by describing both D2a and D2b models as well as the path loss model used.
Contribution of this thesis is shown in chapter 4 and 5. In chapter 4 a new Algorithm “Power-Distance Algorithm” for LTE-R system is proposed. This algorithm depends only on the eNodeB power and intermediate distances between the two cells and independent on the train speed. This algorithm is used to solve too late handover problem which is discussed also in this chapter.
Two simulation results are given. The first one is for the classical A3 handover triggering algorithm while the other is for our proposed Power-Distance handover Algorithm.
Chapter 5 starts with showing previous work on the mobile relay station for high speed trains. It studies an enhanced scheme for handover measurement with both Quality and Doppler-Based schemes. Then it draws the simulation results for this enhanced procedure.
After that, we add our contribution by optimizing this procedure to be more realistic by taking train speed variation (train acceleration) into account.
Finally, simulation results of our optimized procedure are given for different train speed which aid to declare the importance of this modification.
Chapter 6 presents conclusions followed by some research points recommended for future work on the same scope of this thesis.