Long train running

Sam Fenwick looks at the ongoing work in the rail and cellular standardisation communities to replace GSM-R with FRMCS

For some time, the railway community has been working to develop the Future Rail Mobile Communications System (FRMCS) to replace GSM-R, a standard for railway communications that was developed in the second half of the 1990s. GSM-R handles voice and data for security and train tracking applications, and 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.

The drive to move from GSM-R to FRMCS has arisen because of the limited number of vendors that support GSM-R and their uncertain support for the standard beyond 2030, and the potential benefits of aligning the railway communications community with the telecommunications sector.

One of the parallels between the FRMCS work and the work done previously to bring the public safety community into the cellular 3GPP-based world is that the initial emphasis is on duplicating the functionalities in use rather than on new operational functionality. However, this tendency is stronger with FRMCS than with mission-critical LTE, given the latter has been driven more by the desire of the PPDR community to be able to use bandwidth-intensive applications than any concerns about obsolescence.

“There is a specific group in UIC that deals with the collection of user requirements and there is no indication that the main driver for the evolution is an operational need,” says Begona Domingo, ERTMS project officer, European Union Agency for Railways (ERA).

However, she adds that a scenario in which there is full automation and driverless trains would require video-streaming and other bandwidth-intensive applications, which might not be possible to cater for today.

Norman Frisch, chairman of the eLTE Industry Alliance and marketing director of the transport sector for Huawei Enterprise Business Group, says the emphasis on data has increased, driven by the need for real-time video to enable driverless operations under the umbrella of ETCS Level 3.

In addition, Kees Verweij, ICT innovation engineer for the Dutch police, who has been following the FRMCS standardisation process in 3GPP, says there are plans in the railway community to eventually use FRMCS to move away from trackside sensors for signalling and “implement train control based on data and position updates running over mobile protocols”.

Domingo also states that another driver in addition to the need to move away from a GSM-based technology is the desire for a more modular architecture. “We currently have a quite static definition of the system. Today’s technology, GSM-R, is the only one allowed to be used in EU interoperable trains both for voice and ETCS Level 2 communication – we need to make it easy to change the communication system without impacting the rest.”

Huawei’s Frisch adds that the UIC’s aim to have all railway functionality in the application layer “makes total sense” – given that LTE is just a pipe for data. He adds that one reason it may want this flexibility for applications is that “every country still has very specific small items they run slightly different than other countries”. For example, there are different ways of implementing function addressing in GSM-R, but the system has been standardised in such a way that it can cope with these differences. He argues that given some railways are having to make the switch to LTE before the FRMCS standardisation efforts are complete, a similar approach needs to be used.

Thomas Chatelet, ERTMS project officer, ERA (who is also the rapporteur for an ETSI study on FRMCS architecture), says part of the challenge is to try not to develop FRMCS in the same way as GSM-R, alluding to the latter’s more specialised functions. He adds that the differing lifecycles of railway and telecoms products (the latter are replaced at a much faster rate than the former) – a phenomenon he sees as becoming much more important in the future – means that “the focus needs to be set to finding how it will be possible to achieve this modular architecture where the impact of the modification of the radio part [of the system] is minimised, so you could have 4G, 5G, Wi-Fi or even satellite plug-and-play radio modules”.

Chatelet explains that this concept of bearer flexibility would help mitigate the obsolescence risk in the future and would also be useful in situations such as deployments in isolated areas, where a satellite-based system could be a more cost-effective approach than investing in terrestrial infrastructure.

As one might hope, the development of FRMCS is taking place with an eye for what’s gone before. “We’re trying to reuse the 3GPP framework as much as possible to achieve the functionality required by the railway community,” Chatelet says. He adds that once the user requirements have been refined by the UIC and reviewed by the sector and the telecoms industry in the relevant ETSI Technical Committee for Rail Telecommunications, the use-cases are then transferred to 3GPP, at which point it is determined whether or not they could be met using existing 3GPP functionality, such as the MCPTT, MC video and MC data features which have already been developed for the public safety community. He adds this approach helps to both increase the size of the overall ecosystem for FRMCS, while cutting down on the standardisation work required.

It is therefore not surprising that the PPDR community is very supporting of the creation of FRMCS. Robin Davis, chair of TCCA Transport Working Group, says: “We have been liaising with ERA and UIC for the last couple of years and are ready to assist the FRMCS project as it develops further. TCCA Transport members include those organisations that do not use GSM-R but do require mission-critical voice services and currently use TETRA with functions that are closely aligned to that of PPDR users. Therefore both GSM-R users and TETRA users in the railway marketplace will ultimately migrate to LTE-A-based solutions using the FRMCS concepts and standards agreed at 3GPP.

“Alignment of these two ecosystems will therefore create a larger market for transport critical communications users, ultimately resulting in cost-effective solutions for end-user organisations. Through the activities of TCCA Transport Working Group and TCCA’s market engagement role at 3GPP, the association is very keen to support the transport community with understanding their future roadmaps and developments and how they might utilise different communications bearers – a tasking that will be commissioned shortly in conjunction with TCCA Critical Communications Broadband Group, chaired by Tero Pesonen.”

Davis adds: “Within TCCA Transport Working Group, we are all excited about future developments and are greatly appreciative of the work the FRMCS project team is doing. We are here to help and therefore are keen to appoint a liaison officer over the next few months that can assist the FRMCS project team and offer advice on work already undertaken. It definitely makes good sense for TCCA to support FRMCS moving forward.”

Lost in translation
Verweij says one of the obstacles to the FRMCS standardisation effort is the difficulty in translating hundreds of pages of railway use-cases, which are rich in operational background and railway terminology but very hard to distil and analyse – “that makes life difficult”. He adds there is still a lot of work to be done in this area and that the railway community’s thinking is still evolving, so some requirements are undergoing iterations and updates – which can result in dramatic changes from a standards point of view.

He notes that one example of this is functional alias handling – which allows a user to (based on their current operational role) show how they will appear to others, and it “was seen in the early phases as something new for mission-critical services and vital to the railway community, so that work was started two years ago to define the requirements [for it]”. At first, this was seen as a relatively simple feature, but “now we see extra requirements coming up which make it quite a complex feature, and that may have quite a big impact on what was already developed within the downstream groups in 3GPP – those responsible for the architecture and detailed protocols – and that’s quite a dangerous thing”.

As I write this, stage 3 of MONASTERY, which will add capabilities for functional aliases, is flagged as red in the 3GPP workplan and is 28 per cent complete.

Synergy between the PPDR and railway worlds could exist if both groups of end-users end up using the same technology and functionality, which is a distinct possibility given the eventual migration of both to 3GPP-based technology. Verweij expects that while some countries might have dedicated networks for railways or PPDR, the majority will end up using mission-critical services over commercial MNOs. He also expects PPDR and rail will use separate private systems connected to an MNO network as there is little overlap in terms of their operations.

In Europe, the EU mandates the adoption of GSM-R under the Control-Command Signalling Subsystems Technical Specification for Interoperability (CCS TSI, which is due to be revised in both 2019 and 2022). A report from ERA will be sent to the EU Commission on System Definition later this year – which will cover functionality, technology, spectrum and migration. Domingo says this will effectively act as the first indication of the changes that will be made to CCS TSI. For the TSI adoption, all the ministries of transport of the various member states provide their input and work with each other, with the European Commission as chair, to agree on a final text.

The changes introduced in 2019 may include rules for the introduction of FRMCS and the decommissioning of GSM-R. However, according to Domingo, they will not include technical specifications, as these are currently being developed in ETSI and other standardisation groups – but it may include some principles that will set how each member state should share details of its migration plans, thereby giving the EU as a whole and the railway companies better visibility of the overall transition.

3GPP is adding support for FRMCS in Release 15 (in the form of MONASTERY) and Release 16 (MONASTERY 2), on top of that already present in the mission-critical features created with the input of the PPDR community. Release 15 is due to be complete in September 2018 (functional freeze date with stable protocols), while Release 16 should be completed in December 2019. There was a recent workshop on this work and future rail transport communication needs at ETSI’s headquarters in Sophia Antipolis, France.

It’s not just about Europe
It is worth noting that across Europe there is variation in terms of countries’ readiness to switch from GSM-R to FRMCS – in much the same way that the PPDR community’s transition to LTE is very much dependent on the relative ages of their PMR networks – Switzerland might be ready to transition to FRMCS soon after it rolls out, and France and the UK possess mature GSM-R networks, while the Netherlands and Germany are in the process of replacing their existing GSM-R base stations. Conversely, Croatia, Slovenia, Slovakia and Hungary are still in the process of rolling out their own GSM-R networks. Interestingly, as covered in my previous piece on this topic, Finland has opted to use its PPDR TETRA network as a stop-gap solution, allowing it to retire its old GSM-R network without having to replace it immediately.

However, as Huawei’s Frisch points out, while the work on FRMCS is strongly influenced by the needs of the European railway market, much of the rail community in the rest of the world is in a very different position, with many greenfield scenarios. He explains that this creates a dilemma for such operators – do they implement GSM-R or opt for “something more modern”? There is strong demand for LTE-based systems for railways, particularly in the Asia Pacific region, and Frisch adds that the current FRMCS timeline is at odds with those of many non-European railways. He therefore argues that the UIC should find a path to ensure that FRMCS fits the delivery timelines for all its members.

The most recent example of the trend towards the early adoption of LTE outside Europe was the announcement over in Australia that Huawei will build and maintain a 3GPP-based LTE network that will provide voice and data services across Perth’s rail network in the west of the country. It will operate using the Public Transport Authority’s existing 1800MHz spectrum and is scheduled for completion in 2021.

What about spectrum?
One of the truisms in the world of telecommunications is that nothing can be done without spectrum, and the migration from GSM-R to FRMCS is no exception – especially given the need for a seamless transition with as little downtime or disruption as possible.

Currently in Europe, GSM-R uses 2 x 4MHz (FDD) of spectrum just below and above 900MHz (876-880MHz uplink and 921-925MHz downlink). Chatelet says that one of the scenarios being discussed within FM56, the project team within the Electronics Communications Committee (ECC) – a division of the European Conference of Postal and Telecommunications Administrations (CEPT) – is the addition of a further 1.6MHz of spectrum (including a 200KHz guard band) that would be used for the most critical communications over FRMCS during the transition, while the 4MHz allocated to GSM-R remains untouched. Then once GSM-R is switched off, the full 5.6MHz of spectrum could be used by FRMCS. “This scenario is being heavily pushed by the railway community because having spectrum with similar propagation characteristics aids the reuse of existing infrastructure and allows the co-location of GSM-R and FRMCS base stations,” he adds.

Chatelet explains that another of the scenarios that is being considered is the use of 10MHz (TDD) in the lower part of the 1900-1920MHz band. “There would be potential for EU harmonisation in that band as it was previously assigned to a direct air-to-ground communication system.”

He adds that both scenarios are not mutually exclusive, so it is possible that both would be implemented, allowing the 1.6MHz of spectrum in the 900MHz band for the most critical communications to be complemented with the extra spectrum in 1900MHz for hotspot areas where more data-intensive applications would be required.

As we bring this tale to a close, it’s worth remembering the wider narrative here. With 5G looking to bring mobile operators and verticals together, the work to bring the PPDR and railway communities under the 3GPP umbrella is serving to pave the way. At the same time, it’s clear that the tension between the need to go fast and the need to go together are by no means restricted to the world of public safety.