LTE for rail: Finding a sucessor to GSM-R

GSM-R is widely used by the rail industry, but as with TETRA the sector's gaze is looking to the future and the migration path to LTE. Sam Fenwick investigates

GSM-R is widely used by the rail industry, but as with TETRA the sector’s gaze is looking to the future and the migration path to LTE. Sam Fenwick investigates


GSM-R is a standard for railway communications developed in the second half of the 1990s. It is widely used in Europe and China for security and train tracking applications through the secure transmission of voice and data, and also provides the communications that allow ETCS (European train control systems) to function. Both GSM-R and ETCS are part of the ERTMS (European rail traffic management system) standard, which is defined by the International Union of Railways (UIC) and is designed to allow interoperability between cross-border traffic. GSM-R supports voice group call, voice broadcast call and prioritisation of these services.

Increasingly the railway community has been looking for a possible successor to GSM-R, because of uncertain vendor support beyond 2030 and society’s general move towards IP and broadband technologies.

Chiel Spaans, ERTMS unit project officer at the European Union Agency for Railways, says that while the current transition timetable is not very detailed “we aim at including in EU legislation the possibility to use systems other than GSM-R around 2022”.

“This does not prevent railways from implementingthe successor earlier [for national or non-interoperable applications]. The standards we would like to refer to in the legal framework have to be selected around 2020 to allow time for proof of concept and operational pilots,” he adds.

Spaans says that the rollout of the GSM-R successor networks will “mainly follow” the technical and economic life cycle of the legacy GSM-R equipment, which varies from country to country. He adds that early adoption may startin 2022 or 2023, with an EU-wide migration period lasting approximately 10 years.

To address migration questions the UIC set out the requirements for a GSM-R replacement in its Future Railway Mobile Communication System paper. This is now beingused by 3GPP Working Group SA1 to assess which of the requirements fall under 3GPP’s scope, and are either provided by Release 14 or need to be developed in Release 15. Spaans says that while both Release 15 and 16 could be the Releases that would allow commercial products to be available in 2022, Release 15 “sounds more realistic” once the 2020 deadline for trials is taken into account.

One of the most important benefits of using LTE to succeed GSM-R and introducing the railway-specific features through 3GPP is the potential to align the railway sector with the wider telecommunications market, in the same way as the PPDR market, so that it can take advantage of the economies of scale and variety of devices seen in the mass market.

With this in mind it was only natural for PPDR and transport user organisations to work more closely together.

“Around February [the UIC and the TCCA] got together – we’d been discussing things and attending joint meetings for a while – but we got together in Lille,” says Robin Davis, chair of the TCCA’s transport working group. “There’s a memorandum of understanding. Both parties have agreed that it makes perfect sense not to duplicate each other’s work. The fact that we are working together in this way is absolutely brilliant. Years ago we wouldn’t have been seen in a pub together!”He adds that as part of the mission-critical push-to-talk standard being introduced in 3GPP Release 13 “we havefed in some of the development around railway emergency call, things like that.... Every now and then there’s a burst of activity. We’re not necessarily doing everything every week.”

When Spaans was asked if the timetable for the development of a successor for GSM-R might make 5G of interest he says: “The 5G philosophy is very attractive for railways and the EU is supporting the development of 5G in many ways.

“But 5G doesn’t exist yet. That’s the reason we are following the developments very carefully and will decide at the latest possible moment. One of our main concerns is the size of the ‘catalogue’ of potential technologies and frequency bands to be supported by on-board equipment,” he continues. “We have to take into account that any equipment, once ordered, has to be able to work in different networks for 10 to 15 years... and SDR [software-defined radio] is not mature yet.”

According to a presentation given by Emanuele Di Liberto, director of business development – transportation at Nokia, at the 2016 UIC ERTMS World Conference a test campaign conducted on a 12 km long high-speed rail line in Antequera (south Spain) found that LTE operating at 2.6 GHz “is able to convey critical railway services with an acceptable performance in terms of delay, jitter, packet loss, services capacity and traffic prioritisation”.

Di Liberto sees the migration path to LTE as a gradual progression; starting with LTE providing a pure data overlay, then a multimedia overlay, before replacing GSM-R. He notes that spectrum will be a key issue, with the need to address GSM-R spectrum refarming, possible new bands, and spectrum sharing.

Davis highlights the transportation sector’s interest in Teltronic’s combined use of TETRA and LTE for Bilbao Metro in Spain. LTE is used to provide real-time video transmission between trains, stations and the Metro’s command and control centre, as well as from the portable devices used by security staff. Video footage from cameras in stations can also be streamed in real time using this system. LTE was overlaid onto the existing TETRA network, which provides PMR voice and narrowband data services.

One consequence of the timeline for the development of a successor to GSM-R is that in some cases a stop-gap solution is required.

Finland was one of the first countries to adopt GSM-R and as a consequence its GSM-R network (RAILI) needs to be replaced in 2018 as it suffers from interference caused by commercial networks. The Finnish Transport Agency (FTA) needed a bridging technology so that it could retire the GSM-R network without having to replace it.

It will soon be relying on a bearer-independent communication architecture called URCA (Unified Railway Communication and Application), which is being supplied by Frequentis and will allow Finnish railway users to utilise the national TETRA network VIRVE and public mobile networks. Thomas Karl, director of public transport at Frequentis, notes that one advantage of bearer-independent architecture is that it will make the migration to a future mobile communications system significantly easier.

According to Markku Voutilainen, senior inspector and project manager of railway communications at the Finnish Transport Agency, the main challenges for the transition to URCA include a tight schedule for its testing and delivery, the rollout of the TETRA train radios, and improving TETRA coverage in shunting yards and railway tunnels. He adds that URCA is planned to go live in the third quarter of 2017.

However, the country has some unique circumstances that make it difficult for this approach to be adopted elsewhere. “Finland is a railway island,” says Voutilainen. It does not have any border-crossing traffic to any other EU country.

Some of the benefits of using TETRA as a stop-gap include cost savings, as the FTA no longer needs to run its ownradio network, and according to Voutilainen it will eliminate the need for train operators to “swap the existing GSM-R handheld cab radios to interference-improved interoperable GSM-R cab radios”. He explains that this is particularly expensive for old locomotives “where it is difficult to arrange a place for big and expensive GSM-R cab radios”.

There has been talk among some industry players of railways sharing communication systems with public safety agencies. But does this make sense?

“My personal opinion is that railways have to sync their needs with PPDR’s, to prevent a rail-specific niche market and increase a common market for professional users,” says Spaans. “This applies mainly to terminal equipment; the agreed network features will – hopefully – be covered by all network suppliers and may also be available in public networks.

“Recently the UIC and CCBG [Critical Communications Broadband Group] have agreed to investigate the conditions for successful sharing,” he adds. “These conditions are related to operational co-existence (who needs what, where and in what circumstances) and ‘technical’ co-existence (security, subscriber management, capacity allocation). This study is not performed yet, but may provide essential information for network management and for national discussions on sharing.”

However, he notes that even if the study identifies promising outcomes and economic benefits the final decision will rest in the hands of each country’s government. During the transport panel discussion at Critical Communications World he said that the European Commission is seeking two things with the next generation of critical communications for railway users: cost reduction and interoperability. As part of this, member states can make national decisions as long as the solution they chose is interoperable with that used by other member states, when applicable.

“There’s an interesting debate over what is meant by sharing [with PPDR], and some countries have their own transport police so how far does the provision of the service beyond them to railway operators themselves go?” asks Davis. “It’s not there yet, but it is early days and it is being talked about. It makes sense but it hasn’t got to the political level yet.” He adds that in the UK some transport users, such as the British Transport Police and the Highways Agency, already use the public safety Airwave TETRA network.

However, also during the Critical Communications World transport panel, Milan Vignjevic, project manager– communications engineering for Toronto Transit Commission, said that while he thinks network sharing between PPDR and transport would be of benefit, it’s currently not happening in North American transit because transport agencies aren’t seen as as critical as the police, fire and ambulance services.

Speaking in the same session as Vignjevic and Spaans, Frédéric Jans-Cooremans, spectrum and radio project manager at Société des Transports Intercommunauxde Bruxelles (STIB), said that while it was easy for his organisation to become a user of Belgium’s ASTRID TETRA network the usage costs made it “much more interesting to buy and make a new network”. He also raised the issue of the emergency services having priority over other users – “it’s their network, they have priority”.

Railway operators’ needs differ to those of PPDR organisations in that their core requirements for wireless communication (those relating to safety and the performance of their operations) are “not that demanding” in termsof bandwidth, according to Spaans. He adds that future railway-related broadband applications may be deployed on public networks prior to the implementation of a successorto GSM-R, and their subsequent migration to dedicated networks could be an option for those applications considered to be mission-critical. He adds that this is having an impact on the work regarding the additional spectrum that will be required (for at least the migration period).

One issue Davis is keen to stress is the importance of cybersecurity. “I think the future opportunities around M2M interfaces and IoT are clearly interesting for the user groups, but one of the issues of all of this will be security. If you’re looking at sharing networks you have to define [how they will be shared] because if you run telemetry, M2M interfaces, and various train-borne technology over the same network you have to ensure all the end points are secure. There’s quite a bit of work that needs to be done in that area.”

Clearly there is much to be decided and tested before Europe’s railways switch to mobile broadband and/or embrace the use of PPDR networks. However, with this new-found spirit of co-operation between the two sectors anything could be possible.