THE landscape for railway telecommunications in Europe has changed drastically since the turn of the century. From 35 separate analogue systems in 2000, a single, interoperable railway communications system now exists across much of the European railway network: GSM-R.

The system, which was drawn up by European railways during the 1990s and ultimately chosen over Tetra, provided a positive step towards the digital age. It offered interoperability between different networks for the first time, and enhanced functionality which improved safety and responsiveness. And it has been a major success. In 2016, 60 countries on five continents were using GSM-R, with more than 100,000km of lines covered in Europe, and both green and brownfield networks set to continue rolling it out over the next few years.

Railway radio 1That they continue to do this in spite of advances in radio technology is due to a lack of a proven and widely-approved alternative. The EU now mandates its members and other associates to adopt GSM-R under the 2012 Control-Command Signalling Subsystems Technical Specification for Interoperability (CCS TSI). And as the telecommunications component of ERTMS, other networks around the world have followed suit as they have embarked on their own ERTMS rollout and analogue radio replacement programmes.

But there is a problem. GSM-R is a second-generation telecommunications system, which means it is a long way behind today’s 4G technology, let alone 5G, which is expected to emerge around 2020. Obsolescence is, then, a major issue. And while providers have committed to maintain GSM-R up to 2030, beyond this it will become increasingly difficult, and expensive, for infrastructure managers to retain the same quality of service.

In response, work is now underway to prepare the industry for the transition to a new radio system and associated technologies.

As the system authority for ERTMS, and following the principles outlined in the 2012 MoU on ERTMS, the European Agency for Railways (the Agency) is coordinating this process in Europe. It is joined in the work by the International Union of Railways (UIC), which has established the Future Railway Mobile Communications System (FRMCS) group consisting of operators, infrastructure managers, and suppliers to oversee the preparation of specifications for a successor to GSM-R. The Technical Committee for Railway Telecommunications (TCRT) at the European Telecommunications Standards Institute (ETSI) has similarly founded a working group on the issue and is liaising with the 3rd Generation Partnership Project (3GPP), which develops the standards for mobile communications in Europe.

The Directorate General for Mobility and Transport (DG Move) at the European Commission is expecting to receive the first recommended high-level changes to CCS TSI, including a strategy for migration and timeline for implementation, from the Agency by the end of 2018. Work will then proceed to begin the initial rollout of a new technology by 2022. Yet as Mr Thomas Chatelet, project officer in the ERTMS Unit at the Agency, admits, there is a great deal of uncertainty at this stage of how the project will play out.

“Some countries including Switzerland and the Netherlands are capable of turning off their GSM-R system soon after the rollout of the successor, and they might be ready for 2022,” Chatelet says “Finland has decided to switch GSM-R off already and is temporarily using its national Tetra network as an alternative while it waits for the future system to emerge. However, there are some countries - Croatia, Slovenia, Slovakia and Hungary - which are currently in the process of rolling out GSM-R as a first-generation network, while France, Germany and Britain now possess mature networks. They are asking the question of whether they should look to renew their system in five to seven years and be the pioneer, or wait for the new system to gain wider acceptance.”

One of the areas causing particular uncertainty for member states, and railways, concerns the radio technology on which the new platform is based. 4G Long Term Evolution (LTE) would appear the natural choice at this point. However, with 5G technologies set to become available by the end of the decade, for some this would be a more logical choice on which to base a new system, particularly as 4G is likely to be reaching obsolescence by 2030.

“There is now a big push towards converging 5G in Europe, and for Europe to take a leading role in 5G development,” says Mr Michael Mikulandra, head of industry appliance next generation RDN at Kapsch CarrierCom. “Transport in particular is one area where 5G can play a role. And from a time perspective, the availability of the first 5G products is in line with the future generation of railway communication at around 2020-21.”

However, Mikulandra adds that first and foremost the new platform must demonstrate flexibility because in some rural areas it might not make economic sense to install 5G equipment where other systems will suffice. Instead the emphasis should be on developing a larger architecture on which certain functionalities are underpinned and not impaired if the radio coverage changes.

Chatelet agrees, and says whether it’s 4G, 5G, Wi-Fi, or satellite, the emphasis must be on guaranteeing a continuously available and interoperable system. “With GSM-R it is easy because as one system, interoperability is the facet of the system,” he says. “It is one system with one frequency band across the EU.”

Biggest challenge

Nevertheless, it is spectrum allocation rather than radio technology which is described as the biggest challenge facing the deployment of this new technology and the future of railway communications. GSM-R is currently located within the 4MHz of the R-GSM band. However, there is evidence that coexistence of GSM-R and a future system within this band is not possible without a substantial degradation of the level of service.

A 2016 study by LS Telcom, commissioned by the Agency, considered whether LTE/LTE Advanced, as the only practical candidate currently available, could be used effectively on the same band as GSM-R. It found that it is not possible to introduce LTE in the GSM-R band without a number of technical mitigating measures while there is insufficient capacity to allow coexistence without some degradation. The study also concluded that LTE would provide extra data capacity but potentially reduce the capacity of GSM-R. In addition, in areas of high traffic density or border areas, the capacity for both services would be severely reduced.

To counter these problems, railways may be forced to explore the use of a different frequency for the new system, with the frequencies both below and above 1GHz available as possible alternatives. However, this could drastically increase costs due to the requirements to install new infrastructure compatible with an alternative frequency. Rail also faces a challenge from other industries, most notably the road sector and the expected development of Short Range Devices or autonomous vehicles, to secure the spectrum it needs.

In response to this and the continuing uncertainty, at the end of January the Electronics Communications Committee, a division of the European Conference of Postal and Telecommunications Administrations (CEPT), founded a new project team, FM56, to identify solutions for spectrum issues relating to railways.

FM56 consists of national communications regulators and other groups from 48 countries, including Turkey and Russia. The Agency will host the group’s kick-off meeting in Lille at the end of this month. And with members prioritising sector migration, Chatelet says it is hoped that the group will ultimately identify the priorities for spectrum allocation for railways and provide a strong united voice to state rail’s case.

“It is a good opportunity for the railway sector to create an understanding of what it needs and the importance to the future of railway signalling and other applications that are critical to safety,” Chatelet says.

As well as FM56, the industry has also formed the Coordination Radio Working Group. Coordinated by the Agency, and consisting of infrastructure managers, train operators, suppliers, research bodies including Shift2Rail, and authorities including the UIC and ETSI, the group will inform the high-level recommendations that the Agency will present to DG Move at the end of 2018 with an emphasis on maintaining interoperability.

It also hoped these discussions will crystallise the industry’s thinking on the role of public networks in future railway telecommunications.

GSM-R is a closed standard, limited to mission-critical voice communications. The vision for the next-generation system supported by the Agency is for a three-pronged approach, with mission critical services joined by business applications such as those to measure asset performance and market applications including passenger Wi-Fi. Yet the degree to which public networks provide the support for each of these areas remains a matter for debate.

Current onboard passenger services have generally been made available through public networks. However, with the opportunities provided by a new “hybrid” network, railways could assume this responsibility.

The willingness to host public services is unclear at this stage and is likely to vary between country. Indeed, some countries may consider allowing public operators to provide all radio services including mission critical applications, reducing railway’s capex for the next generation technology, although Chatelet says this is considered very unlikely in established markets like Germany, France and Switzerland due to safety concerns.

If the emphasis is on public providers to deliver, they may become obliged by telecoms regulators to guarantee coverage for the railway network. This is currently not the case in many countries, leading to spotty onboard Wi-Fi coverage, although Chatelet reports that telecoms providers are now being forced to do this under regulations recently adopted in France and Spain.

Chatelet adds that the Agency’s recommendations will also strive to offer an open definition of the preferred system in order to enable the market to ultimately come up with the technical solutions. This is intended to avoid a potential monopoly on the next generation solution, and to promote continuing development of the standard as it matures.

For GSM-R, Kapsch, Huawei and Nokia have supplied the industry, while Ericsson is currently addressing the market with its LTE solution. Chatelet says that for the next generation, it is expected that these vendors will continue to lead the way, but with opportunities for others, including current signalling suppliers, to come up with their own solutions. “Everyone should be able to participate and propose enhancements to the successor of GSM-R,” Chatelet says. “It should not be like a club.”

Research

Current research programmes, in particular Shift2Rail, could play a critical role in identifying these prospective solutions. Kapsch is the only GSM-R supplier currently working in Shift2Rail as an associate member, and is a key player in Technology Demonstrator 2.1. This project will collate future railway communications requirements, define standards and specifications, study business models and ultimately provide prototypes for field testing. The X2Rail-1 project was subsequently launched on September 1 with the goal of providing the preparatory stages to support the introduction of an adaptable IP-based communications system. It will conclude in August 2019.

“The biggest benefit for us is working with the different stakeholders involved, especially the signalling companies,” Mikulandra says. “We are able to talk on a technical level and get a better idea of the projects planned such as for automatic train operation, onboard voice and wayside operations for the next five to 10 to 15 years. We are then able to import these findings back into our own thinking.”

However, research is not restricted to Europe or Shift2Rail. Huawei is continuing to push 4G LTE and 4.5G LTE as the preferred solution and is engaged in various research and development activities. This includes through membership of FRMCS and TCRT, and partnerships with the Chinese Railway Science Research Institute and Beijing Jiaotong University on researching the next generation of LTE technologies.

Huawei will take a critical step forward with the rollout of 4G LTE on a China Railway main line in 2017, and has already successfully delivered the solution to the heavy-haul Shuo Huang Railway. The Chinese company is also continuing to develop and implement 4G LTE for metro networks with more than 20 applications in China to date.

However, as Mr Norman Frisch, Huawei’s business development manager for rail transport solutions, points out, the work now underway in Europe means that it remains most likely to take the lead on defining standards for the future use of 4G LTE around the world despite no European 4G LTE main line deployment at this point.

“The pressure is on Europe, and everyone is looking at what Europe is doing,” Frisch says. “If Europe comes up with a standard which allows the successful migration from GSM-R, it will be very important for railways outside Europe because of Europe’s experience of using and standardising GSM-R.”

Frisch adds that 4G LTE is more than capable of delivering what the rail sector will require from its next-generation communications network. And it is advantageous because the technology is already available. He reports that 3GPP Release 13, the current standard for 4G LTE, which was completed at the end of 2016 and provides the basis for hosting mission-critical applications, could offer a firm foundation for GSM-R migration.

Frisch also says that the capabilities of 5G, which will emphasise the Internet of Things (IoT) and have the capacity to connect billions of devices and transfer gigabytes of data per second, may be beyond what railways realistically require from their communications architecture.

“IoT is important for railways for monitoring and preventative maintenance, but this can be built into the 4G network,” Frisch says. “I don’t see how railways will benefit from the difference as 5G will connect billions of devices and transfer many gigabytes of data while a railway probably only needs to connect millions and transfer a gigabyte of data.”

The range of debate and disagreement currently underway and set to take place at conferences and meetings on the subject over the next few years is testament to the nature of the process to identify the preferred solution. Indeed, at this relatively early stage the only certainty is that change is afoot.

Currently GSM-R is not thought to hinder the deployment of advanced technologies such as Automatic Train Operation on main line railways, which are under discussion and in the initial stages of development at railways across Europe. But it cannot be relied on as the telecoms support for these advanced systems in the long-term.

If the transition to GSM-R is anything to go by, however, setting 2030 as the target date for a full rollout of the next generation system is likely to be a major challenge. While medium-sized member states, with a few thousand route-km each, took around seven years from the start of planning to the end of migration, for larger member states, the individual migration has taken up to 19 years.

Experiences from the GSM-R migration may help to speed this process up, particularly if existing infrastructure is adapted and reused, which would reduce costs significantly. Yet the speed at which this needs to take place should not be underestimated or underemphasised. As studies commissioned by the Agency have clearly shown, railways should ignore the need to do anything at their peril. As the debate proceeds, the Agency makes its recommendations, and the technologies are ultimately developed, this point needs to be clearly made. And made again.