A common bearer (transport layer) telecoms network takes advantage of the new digital technologies and ‘big data’ applications in order to provide a safe, efficient, reliable railway. In very simple terms, this unseen telecoms network is the ‘glue’ that binds the digital railway together and is therefore hugely important. It will be the heart and veins of the digital railway.
Railways need to modernise and to provide improved capacity and on-time services, especially as the competition from autonomous vehicles is gaining ground all of the time. Reliable, efficient and high-capacity connectivity is essential in order for railways to make efficiencies, innovate and compete. There’s also a growing desire and need to improve mobile connectivity for both passengers and operational services.
There are potentially a lot of use cases to support along the rail corridor. These include operational voice and data services for train control, SCADA for electrification control, remote condition monitoring, CCTV, CIS, GSM-R, IoT, business voice and data, third-party commercial fibre connectivity and broadband track-to-train connectivity.
The digitalisation of the rail network means finally bringing all of these services together on a cost effective, reliable and resilient fibre network that delivers not only on security, with the potential to create virtual private networks (VPN), but also on an ability to expand over a 30-year period.
In some railways, the data services today run across disparate, ageing networks which can be costly to manage and, invariably, fail from time to time. The networks can be near impossible to correlate together and require significant resources to ensure safe, reliable operation. Aging infrastructure (copper, fibre, transmission equipment) can lead to common problems around manageability, to a high cost to maintain and upgrade, and in some cases to operational failures that can lead to train delays.
Globally, many railway infrastructure managers and railway undertakings currently use the interoperable radio communications network, GSM-R (Global System for Mobile Communications – Rail), for operational voice communications and to provide the data bearer for ETCS (European Train Control System). In the European Union this is legally mandated in the Technical Specifications for Interoperability that are applicable in the European Member States. Voice and data communications are also used for various other applications.
GSM-R has been a huge success all over the world, not just Europe, but it is a MOTS (modified off-the-shelf technology) system based around manufacturers’ commercial GSM offerings, enhanced to deliver specific ‘R’ (railway) functionality. Due to the product modifications required to provide this functionality, and the need to utilise a non-commercial radio spectrum, much of the equipment utilised for GSM-R comprises manufacturers’ special-build equipment and/or software variants. The use of MOTS technology for GSM-R has proven expensive for the railways, both interms of capital and operational expenditure.
The predicted obsolescence of GSM-R by 2030, combined with the long-term life expectancy of ETCS (2050) and the railway’s business needs, have led to identifying a successor for GSM-R. This will have to be future proof, learn from past experiences/lessons and comply with railway requirements.
The successor is the Future Railways Mobile Communication System (FRMCS). This is envisaged to provide the same services, plus a higher data speed capability for operational and business purposes (including real time video), with the option of providing passenger mobile connections. Some metro networks are also interested in FRMCS, not just the main-line railways. All this will require each railway to have a reliable, high-bandwidth common bearer.
Every GSM-R or FRMCS failure for ETCS will shut the railway just as surely as a track circuit failure would; so high availability is essential. CBTC systems for metros are also reliant on some form of radio connection. While the future of train control in both ETCS and CBTC will be radio-connection based, radio will only provide the last few kilometres of connectivity, and the majority of the connection path will still be via fixed telecommunications using fibre, routers and switches with a common bearer.
Passenger bandwidth requirements
Mobile coverage and Wi-Fi are increasingly considered as the essential ‘4th utility’, similar to water, gas and electricity, and rail passengers now expect a reliable and seamless service. The government’s current proposals are to provide for ‘uninterrupted’ Wi-Fi and Mobile (5G) broadband speeds of up to 1Gbps on-board all UK mainline train routes by 2025. This is supported by the communications regulator Ofcom, which has set out its vision for the data connectivity that will be required by 2025 on British trains. From its research, Ofcom says that, in seven years’ time, a crowded commuter train is likely to need 3.6Gbps of mobile data capacity to meet the connectivity needs of its passengers.
A report by Kinetic and Exterion into the spending habits of commuters estimates that, across the whole UK, country commuters apparently spend an average of £89 per week using their mobile devices, with London commuters spending £153 per week. The report says that, in total, commuters spend an astonishing £23 billion per year via their mobile devices while on the move. So, the bandwidth demand is there and growing, but how can it be delivered?
The required track-to-train connectivity will involve many different considerations, such as determining the business model on which such a service would be run, how the deployment would be funded, and potential interoperability across multiple routes or TOCs. The UK rail network is a complex one, with lots of stakeholders – Network Rail, train operators, rolling stock providers and mobile networks – so making the change to deliver the required connectivity requires a high level of co-operation. But, at its heart, a high-bandwidth fixed trackside data service bearer will be required, irrespective of whether the radio system is FRMCS or relies on public mobile network operators, as of today.
FRMCS is likely to be based on a private or shared LTE/5G platform and telecommunications specialist Nokia has already successfully deployed private LTE networks in other transport industries, including networks to control autonomous vehicles and freight shipping ports. It is also one of leading players in the development and deployment of the next generation of 5G radio networks all over the world.
Telecoms network requirements
To support a modern railway with its data requirements and truly drive productivity gains, the telecoms network must be fundamentally more. It must be:
Accessible: Networks must provide deeper reach and extend everywhere there is a business or operational requirement. Regardless of access medium, dedicated network connectivity is a must. Various wireless, fixed, IP, optical and microwave technologies must work together to ensure that no site, signalling controller, sensor, worker or customer is left behind and they are all provided with the right priority of service.
Elastic: Networks must be dynamic and programmable. As new control and command digital signalling is rolled out, and as new sites are added and demands fluctuate, the network should adjust in an automated fashion to optimise resource utilisation and meet application needs in accordance with the railway’s requirements. The signalling supplier will require access to the telecoms data network in order for him to safely test the overall train signalling system. This may also require the telecom node to share the domestic mains electric supply with the signalling system, with no outage to either the signalling or telecoms equipment. Dynamically optimised connectivity should be established wherever it is needed. Programmatic handling of changes in the connections to (and between) local, edge and hybrid clouds will be essential to the performance of the applications and the viability of key use cases.
High-performance: The network should deliver seamless, deterministic performance across all the applications it supports. While the requirements of each set of applications may vary, performance against stringent guidelines must be independently guaranteed and demonstrated for each application.
Resilient: For applications critical to both business and railway operations, downtime can have catastrophic consequences. Networks must ensure availability at all costs to deliver safety and meet business objectives, with 99.9999 per cent uptime a requirement. That last decimal place is important – 99.999 per cent reliability brings about five minutes downtime per year, with 99.9999 per cent – it’s only 30 seconds. Train delay costs and reputations are at risk and, in some cases, human lives and safety may also be at stake.
Secure: As business perimeters expand and devices proliferate, so does the threat radius. Railways know they are at risk from cyber-attacks and cyber security is an essential part of every safety-case approval. Data networks should be a part of the enterprise security solution, rather than the problem. A smart network fabric can play a role in minimising certain threats and ensure that changes are in strict accordance with enterprise policy.
Scalable: Richer data provides deeper context and higher value. A simple move to video for surveillance or for scene analytics necessitates higher bandwidth at each site. Critical real-time video images are considered to be an effective mitigation measure in relation to hazards that may not be detected otherwise by the train control system. In addition, real time video images can enhance operational performance of the railway system when used to support the end user within the target environment. For example, a video application could be used for automatic train operation (ATO), automated detection of objects on or near tracks in the context of autonomous train operation, supervision of platform and tunnels (either by a remote human user or in an automated way) and monitor the situation in the event of an alarm (supervision of railway track, doors, train, smoke detection). It could also be used to transfer a video image in parallel with voice communication (for example, during Railway Emergency Communication). The FRMCS functional working group has just signed off User Requirements Specification (URS) 4.0.0, which includes real time video as a service for the next generation of train radio, so higher fixed bandwidth for video will be an operational requirement.
Each additional use will require the deployment of additional computational power. Control of automated vehicles, for example, may ultimately require the processing and coordination of data from a wide spectrum of sources, including surveillance cameras, in-vehicle sensors and other devices.
The use of high-fidelity information from a range of sensors will improve automation decisions made across a wide spectrum of industrials. As a result, business-critical infrastructure must operate and grow for periods of a decade or longer.
Networks should be designed in a manner that anticipates and adapts to expansion of bandwidth, processing and other capabilities. Within the duration of the life of the telecoms network there will undoubtedly be many compelling new applications that are unknown today, all of which will require higher bandwidth.
Fundamentally, it is an optical-to-the-edge architecture that would use DWM (dense wavelength division multiplex) technology to deliver very tight services from an operations and maintenance perspective (fibre break detection and location, lambda performance, fibre degradation and prediction). All railway services will be separate (on their own lambda, or optical channel) with full resilience. Each lambda can support (on Nokia silicon) up to 400Gbps and, with over 96 lambdas per fibre pair, one can see how this will scale!
Nokia has addressed the problem by introducing a common bearer (transport layer) in multiple-use cases. This addresses both legacy problems and the safety requirements for fibre-based sub-access connectivity, together with the growth and low latency characteristics required by LTE/5G transport, which will form the basis of the next generation of train radio system. The solution provides the opportunity to bring all of the data networks together whilst both maintaining security and separacy and also providing for the possibility of huge expansion over a 30-year timeline.
Some major rail operators are already embracing FTTE (fibre to the edge) to great effect. One example is Schweizerische Bundesbahnen (SBB – Swiss National Railways), which is moving to a fibre underlay with IP/MPLS overlay, to be delivered, managed and operated by Nokia.
SBB, Switzerland’s largest transportation operator that moves both passengers and freight throughout the country, is upgrading its 8,100km communications network of transmission cables and more than 8,500 components to an advanced, converged communications network by 2020.
SBBs synchronous digital hierarchy (SDH) operational communications network has supported all mission-critical applications, including CCTV, train control, signalling and GSM-R while a separate business IT LAN, similar to the Network Rail Fixed Telecom Network (FTN), has handled non-vital services. SBB seeks to realise efficiencies by upgrading and rationalising the technologies used for both networks and to gain flexibility in the deployment of new services, such as passenger connectivity, as well as advanced applications for growth.
Targeted to be fully operational in 2020, the new nationwide data network will consist of more than 10,000 active elements at over 1,300 sites adjacent to the railway and at approximately 500 offices.
SBB’s existing SDH infrastructure and separate IP platform will be migrated to an integrated IP/MPLS and optical network. An innovative architecture will address all of SBB’s needs and support a future-proof networking solution. This encompasses a fully redundant fibre-optic communications wavelength division multiplex (WDM) transport layer that will carry data from different sources. Two different IP/MPLS networks will run on top: one full redundant network for all mission-critical applications, including train control and signalling, GSM-R, interlocking and other applications; and another for services and applications such as CIS, ultra-broadband passenger connectivity, ticketing and a LAN/WLAN for SBB employees.
Nokia service routers and service aggregation routers, with end-to-end network management provided by Nokia Service Aware Manager, will also be deployed.
The SBB network utilises the same Nokia common-bearer architecture outlined in this article – so if other railways were to adopt similar, they would not be in uncharted territory and therefore would be able to deploy the heart and veins of the digital railway with minimal risk.