An Update on track to train radio
The role of track to train radio communications is becoming ever more important. The early networks of the 1970-80s have long been replaced by GSM-R, a standard that has been adopted on an international scale. However, this is 2G technology which is outdated and has been overtaken in the public mobile networks by more modern developments now centred around 5G.
Whilst GSM-R is serving the railways well and has enabled European Train Control System (ETCS) to be introduced as the future train control system, the radio link now needs to be replaced. The future will be Future Railway Mobile Communication System (FRMCS) and has been under development for some while. The project will be a technical, financial and logistics challenge and will involve both infrastructure and rolling stock.
A recent paper delivered to the IRSE London & South East section by Dan Mandoc, provided an update on the progress achieved to date with a list of the work still to do. Dan is the leader within the Union Internationale Chemins de Fer (UIC) based in Paris, so the project is being undertaken with the intention of it being adopted throughout Europe and beyond.
Background to the project
GSM-R was initially designed to replace national track-to-train radio systems that provided driver to signaller voice communication plus some data messaging, but also to provide communication to trackside workers and information systems on platforms. Interoperability across borders was an important element for trains in Europe travelling from one country to another where a common radio system would be invaluable.
It was always envisaged that GSM-R would be the bearer for ETCS and this has been achieved. However, the data limitations of 2G have meant using General Packet Radio Service (GPRS) in order to get sufficient message throughput in high rail traffic areas.
GSM-R is now installed on 130,000km of railway but expanding the system to provide other facilities is not possible. The system was designed primarily by collaborative efforts of the railway undertakings, working closely with the European Telecoms Standard Institute (ETSI) to develop the special requirements of the railway within the overall GSM standard. The supply industry has guaranteed support until 2035 but the intention is to switch over to FRMCS by 2030.
The vision for FRMCS
As well as providing all the present facilities for voice communication, emergency calls, as an ETCS bearer, the new standard will incorporate the Internet of Things (IoT), smart maintenance opportunities and total wireless connectivity as part of the new specification. Three main elements are included within the overall project: User Requirements; System Architecture; and Frequency Spectrum Availability.
The specification will broadly be based on the 3rd Generation Partnership Programme (3GPP) principles (the 3GPP was set up in 1988 to create Technical Specification Groups and their scope and terms of reference, one such group being FRMCS), including mission critical applications. The migration strategy will be all important as this must cover both the lineside transmitting infrastructure, the control centre tasks and associated screen presentations to signallers and controllers, and the onboard train equipment. During the changeovers, assurances must be given for the continuity of ETCS, otherwise the entire train service will be put at risk.
Five groups have been established to look at functionality, system architecture, technical design, spectrum negotiations, onboard space and adherence to the 3GPP principles. To address all of these, a centralised ETSI Technical Committee has been set up populated by railway engineers, European Union Agency for Railways (ERA) members, and the various stakeholders including the supply industry. The committee reports to UIC HQ.
Major challenges have been identified some of which are: quality of service; ETCS bearer independence; interoperability; spectrum availability across countries; equipment design; cyber security; transition from old to new; cost effectiveness; and multi applications.
User requirements
These are many and not all are listed here. The most significant ones come in three categories:
Critical Communications, which includes: on train communication between controller and driver; communication to trackside workers; railway emergency communication; public emergency calls to typically police, fire, ambulance; monitoring and control of critical infrastructure; voice recording and access to recordings; and train integrity monitoring.
Business Communications, which includes: information help point for public enquiries; WiFi for on train passengers; WiFi for passengers on platforms.
Critical Support Applications, which includes: the provision of assured data communications for such as ETCS; safety application key management for ETCS set up; facilitation for both Automatic Train Protection (ATP) and Automatic Train Operation (ATO); authorisation of communication; and location information of trains.
Many of these requirements exist within the GSM-R specification but many other items have now been added. These requirements need to be understood by many non-railway people globally and not just by railway engineers.
Technical decisions
A key decision is that FRMCS will be based on 5G. That seems obvious in today’s world but remember that the work to develop the new standard dates back to the time when 4G was the up-and-coming spec, with 5G being little more than a development envisaged for the future. With that in mind, consideration is being given as to how any 6G or future spec could be built into the specification.
Similarly, the latest 3GPP R17 spec for mission critical services is being used but recognising that an update to R18 will happen before widespread roll out takes place. Multicast gateways will in future be by software solutions not requiring hardware changes. Alliances with all the major radio system suppliers are forming under the controlling eye of ETSI.
Spectrum allocation
Always a sensitive subject and harkening back to the long hours agreeing the frequency allocations for GSM-R. Currently, harmonised pairs of frequencies are allocated for rail use, these being 874.4 – 880.0 MHz in one direction and 919.4 – 925.0MHz in the other. Assurance has been given that these will remain valid when the changeover from GSM-R to FRMCS takes place. An additional 1.6MHz of spectrum is to be added to this allocation in recognition of the additional features now required plus the promise of 10MHz in the 1900MHz band for future developments.
The presumption is thus made that the railways will have sufficient spectrum to cover all foreseen requirements without the need to share frequencies with other users or indeed the public mobile providers. It remains to be seen whether all this materialises in the fullness of time.
5G Rail
This is now the formalised title of the project. The scope and status determination began in November 2020 with a 30-month duration timescale, thus expecting it to be completed in 2023. Ten work packages are within the framework to define all relevant technical detail and framework testing. The intent is to develop first prototypes in the laboratory, then on nominated test tracks which will include a cross border link.
To date the project is seen as being on schedule but a formal meeting between the interested parties is due to happen in April 2022 when a review of progress will be undertaken. Eighteen potential suppliers will be part of this, noting that the first prototypes are already in the laboratory testing stage.
The planning of 5G Rail will cover all the anticipated stages with the first live implementation due to be implemented in 2026. Already the specifications are migrating from v1 in 2019 to a proposed v2 due in 2023 and then to the European trial by 2025. So critical is the project in support of rail operations that it cannot afford for mistakes to be made.
Boundaries and access
5G Rail has to exist alongside other 5G developments so will be approached by consideration of Railway Applications – Service Stratum – Transport Stratum. Clearly, the project must embrace more than pure rail usage and how FRMCS can be accessed by the travelling public, and other third parties engaged in the supply of rail services, must be part of the system design. Engagement with the suppliers will be important in understanding how all of this is going to fit together.
Some infrastructure sharing is anticipated but exactly how remains to be researched. Teaching telecom people the ins and outs of railways and, conversely, teaching railway people the structure of telecoms will be part of the learning and development process.
Implementation, duplication and cost
A major challenge will be how to migrate from GSM-R to FRMCS. Clearly this cannot happen overnight, and lengthy periods of duplicated operation will have to happen. This is most likely to impact on the infrastructure where additional transmitting stations (including radiating cable in tunnels) will be required especially if the 1900MHz band is utilised. All of this will require additional landline connectivity and power supplies.
The position with the onboard train radio looks more hopeful. The new generation of train mobiles as developed in the UK by Siemens Mobility at their Poole premises, have a 4G capability in addition to GSM-R. The company have confirmed that the radio is capable of adaptation to 5G which will mean the train radios will not need replacing when FRMCS begins to roll out. Quite what modification / re-programming will be required is currently unknown.
Already, these new radios have many of the user requirements in place within the design including provision for Driver Advisory Systems (DAS) and real time train monitoring. However, the question as to how trains will pick up either system in a seamless manner is unclear but, in all of this, uninterrupted operation of ETCS must be guaranteed.
The cost of FRMCS is being developed as a business case but it is not going to be cheap in actual terms. An estimate of €25 billion is being put forward for all of Europe but, in context, the building of a new high speed line with new tunnels and bridges would be much more than this. Whatever, this project must happen as the continuance of GSM-R into the future is not possible. While much work has taken place over the past four years, there remain a lot of unknowns still to be resolved. Rail Engineer will keep a close watchful eye on the future.
