INTRO: Life-expired equipment and new regulations governing the radio spectrum are forcing North American railways to adopt and adapt new communications technologies. BNSF is leading the way, testing in a US environment GSM-R equipment developed in Europe

BYLINE: Fred Gratke

Assistant Vice-President Burlington Northern & Santa Fe Railway

RAPID CHANGE in the wireless communications environment is posing many challenges to North American railways as we enter the 21st century. Burlington Northern & Santa Fe Railway has been researching the capabilities of modern wireless technology for the past two years, and formulating a strategy to accommodate these changes. Although numerous factors led to the BNSF review, they fall into three broad categories.

The first is the evolving US spectrum allocation policy. The Federal Communications Commission has instituted a number of new technical and regulatory requirements, which require railways and other land-based mobile communications users to make more efficient use of the available radio spectrum. BNSF currently uses analogue, VHF-band simplex systems, primarily for voice communications. The new regulations are imposing narrower bandwidths per channel and require conversion to digital modulation. Meeting these standards while maintaining our present capability will require a substantial capital investment.

The second general factor relates to innovation in the wireless infrastructure marketplace. Since the onset of second-generation mobile telephone services, wireless hardware and software has become much more reliable, with greater capacity, and economies of scale are driving down costs. This is most definitely not the case with the proprietary legacy systems found on railways. The limited market for rail-specific mobile devices means development costs are higher than in the mass-production market. This puts a high cost on achieving the efficiencies necessary to meet FCC requirements.

The final broad change is the increasing demand for wireless technology as railways strive to become safer, more efficient and more competitive. Initiatives such as Positive Train Control, on-board work order reporting, monitoring of trackside devices and asset tracking require ever-increasing real-time connectivity and integrated systems. Existing legacy wireless systems simply cannot provide the capacity and coverage to meet these demands. New technology with higher throughput and more robust capabilities is essential. But it has to be reliable and cost effective to meet the needs of both the railway and its customers.

The next steps

The regulatory changes will require substantial investment by the major railways, even if they continue to use their legacy systems. To maximise the return on investment, it would seem a logical step to adopt a more advanced technology and harness the additional potential benefits.

One practical solution might be to adopt commonly used mobile wireless technology. Initially this could operate alongside legacy systems and, as resources permit, be expanded to meet almost all of the railway’s wireless needs. Inspiration for this path comes from experiences with Global System for Mobile Communications - Railway in Europe. Many of the factors present in the USA were responsible for the creation of the European GSM-R standard. Europe too faces outdated analogue hardware, spectrum reorganisation and a need for additional capacity and functionality. The strong European need for interoperability between countries is present in North America among the various railroads.

However, there are some significant differences. Unlike Europe, North American regulatory agencies have not yet set aside a dedicated part of the spectrum for railway use. Here, public second-generation mobile phone operators use the 1900MHz band, as opposed to the 900MHz band in Europe. Unlike in Europe, there is no common consensus among North American railways as to an overall wireless strategy.

Given the success and widespread acceptance of GSM-R in Europe, we felt compelled to include it in our exploration of wireless systems.

Developing the strategy

We knew that our strategy must include an increase in data capability, both along our routes and at terminals. We wanted to preserve some of the voice features in place today, such as direct connections and group calls, while adding privacy and emergency pre-emption capability. Since interoperability is essential, developing a consensus with other operators was extremely important. Finally, it would be necessary to obtain additional spectrum width.

We adopted a clean-sheet approach to technical planning, reviewing several different operating models. We looked at continuing to use our current VHF platform with a controlled migration to a digital narrowband platform. We began to follow European progress closely. We also looked at several commercial mobile technologies in North America, notably the Nextel platform, which offered cellular data and two-way voice capability.

We carried out a ’total cost of ownership’ approach to financial analysis, assuming a 10-year life span. This included the cost of obtaining a new spectrum, capital costs, and operating life-cycle costs, and the benefits of forming a partnership with established mobile operators.

Our study showed that GSM-R offered the greatest value to BNSF. The features required had already been developed and tested, and are now being deployed on railways across Europe. The General Packet Radio Service (GPRS) data capability appears to meet BNSF’s current and projected requirements. The Advanced Speech Call Items (ASCI) features provide more-than-adequate voice capability and overall capacity.

However, GSM-R has its risks. For one thing, it has the highest capital cost outlay of all the options studied. Because no dedicated spectrum is available, negotiations with mobile operators for spectrum use have the potential to be difficult and costly. Furthermore, there is no overall consensus among North American operators towards a GSM-R strategy.

Pilot trials

Because of the differences between the European and North American wireless environments BNSF decided to carry out a series of pilot trials. This would evaluate the feasibility of using the 1900MHz band, and of coexisting within the spectrum rights and network of a public mobile operator.

Other significant evaluation points were a GPRS data network overlay of the GSM-R software suite, the adequacy of the ASCI voice feature set within North American rail practice, and the suitability of GSM-R for the widely-used Advanced Train Control System Spec 200 Centralised Traffic Control code line protocol.

A three-phase testing plan was developed. Phase I was a simple test of GSM audio quality in a railway environment. Phase II will test GPRS data handling capability while working with a GSM-R software base load. Phase III will provide a full test of GSM-R voice capabilities along with GPRS data transmission.

Planning for the trials began early last year, with the selection of a suitable location. We decided to use the Transportation Technology Center Inc test track near Pueblo, Colorado. This closed-loop line operates almost around-the-clock, providing a good mix of traffic with high frequency operations. In addition, locating the test at TTCI would bring it into the public arena and hopefully encourage observation and participation by other railways and vendors.

The second key consideration was to define the test parameters, and we turned to Nortel Networks for help in this area. Nortel has extensive experience of GSM-R, and owns a major test and development laboratory at Richardson, Texas, near BNSF’s Fort Worth headquarters. Nortel agreed to provide two base stations, test handsets, support staff and use of the test call routing and switching centre in the Richardson laboratory. BNSF provided a T-1 data link from Pueblo to Richardson via our internal optic fibre network. The less trivial task of securing 1900MHz band frequencies was solved by an agreement with VoiceStream, North America’s largest GSM operator, giving temporary use of several channels in the TTCI area.

Testing progress

Phase I was successfully completed towards the end of 2000, using a 1900MHz band GSM mobile mounted in a Hi-Rail road-rail vehicle.

As well as proving joint 1900MHz band operation with an incumbent GSM mobile operator, the aim was also to ensure an acceptable Delivered Audio Quality. DAQ is a subjective measurement of received audio quality, rated on a scale of 1 to 5 with 5 being the highest. For rail operations it has previously been determined that DAQ from 3·4 to 3·6 is the minimum acceptable rating. This equates to a signal-to-noise ratio of 20dB in an analogue radio.

Six different situations were tested (Table I). In the Trackside situation, two measurements were made with a hand-held mobile telephone while standing alongside the tracks. The first test call was made to another hand-held mobile telephone, the other to a conventional landline. Four different calling configurations were tested with the Hi-Rail.

Phase II was scheduled to run from February to May, but testing was delayed by the inability to obtain suitable GPRS-capable handsets in the 1900MHz band. The handset issue was resolved in August, and testing finally began last month.

Phase II centres around the data traffic capabilities of GPRS, and the objectives include latency and delay, session handover, and data throughput relative to signal strength. We will also be testing the ATCS Spec 200 protocol, which is the primary means of using radio-frequency delivery of signalling instructions to trackside CTC devices.

The Phase III tests were originally scheduled to begin in June, but we now hope to be able to obtain a small number of prototype locomotive cab handsets in the first half of next year. Phase III will test the voice capabilities of GSM-R with simultaneous GPRS data transmission. The ASCI features of one-to-one and group conference modes have great appeal, as they are similar to our current voice capabilities but span a greater geographic area. We want to ensure that there is adequate bandwidth for various types of packet data traffic at the same time as voice calls are being made.

Encouraging prospects

Although the test programme has not progressed to the point where BNSF can decide whether to commit to GSM-R, we are encouraged by what we have seen so far. Whilst waiting for the equipment needed for testing, we are continuing with non-technical aspects of our wireless strategy.

Work is still required on a cost-benefit study, and we have to determine an appropriate suite of wireless applications. Finally, the ongoing effort to build an industry-wide consensus will be critical to the successful implementation of any new wireless operating strategy, regardless of its origin. n

TABLE: Table I. Summary of Phase I test results

Test DAQ1


Mobile (trackside) to mobile 4·36

Mobile (trackside) to landline 4·93

Hi-Rail vehicle

Mobile (Hi-Rail) to landline 4·25

Mobile (Hi-Rail) to mobile 3·60

Mobile to mobile (Hi-Rail) 4·82

Landline to mobile (Hi-Rail) 5·00

1. Mean of 10 Delivered Audio Quality measurements

CAPTION: GSM-R base station for the tests at Pueblo

CAPTION: Phase I tests at Pueblo were conducted using a Hi-Rail vehicle

BNSF studies digital radio

Life-expired equipment and new regulations governing the radio spectrum are prompting North American railways to adopt and adapt new communications technologies. Burlington Northern Santa Fe is testing a version of the GSM-R equipment developed in Europe, with encouraging results in delivered audio quality. Future progress will depend upon obtaining cost-effective access to 1 900MHz band frequencies and reaching a consensus with other rail operators about future strategies for inter-operability

BNSF expérimente la radio numérique

Des équipements en fin de vie et des nouvelles réglementations appliquées aux fréquences radio poussent les réseaux ferroviaires nord américains vers l’adoption et l’adaptation de nouvelles technologies de communication. Burlington Northern Santa Fe teste une version du système GSM-R mise au point en Europe, avec des résultats encourageants quant à la qualité du rendu audio. Les progrès à venir dépendront de l’obtention de l’accès au meilleur coût de la bande de fréquences des 1900MHz et de l’obtention d’un consensus avec les autres opérateurs ferroviaires quant à la stratégie d’avenir pour l’interopérabilité

BNSF studiert Digitalfunk

Veraltete Ausrüstungen und neue Vorschriften bezüglich Funkspektrum zwingen die nordamerikanischen Bahnen, sich anzupassen und neue Kommunikationstechnologien einzuführen. Burlington Northern Santa Fe testet eine Version der in Europa entwickelten GSM-R-Ausrüstungen, welche in Sachen Sprachqualität vielversprechende Resultate zeigen. Weitere Fortschritte hängen stark davon ab, ob kostengünstige Zugänge zum 1900MHz-Frequenzband geschaffen werden k