The word “competition” is used a lot today by regulators and governments when talking about the success of mobile services. We have fierce competition between mobile handset suppliers. Network operators are in a daily competitive struggle to attract and retain their customers. System vendors fret about new competition coming from China. Yet very little of this would exist today without the other important “c” word in mobile radio – “cooperation”.
GSM is a story of successful cooperation over technical standards but there would never have been a GSM technical standard had there not been much earlier cooperation by Europe’s radio spectrum managers over the necessary radio spectrum. It is a hidden success story and deserves to be told.
In 1966 the International Union of Railways asked Europe’s spectrum managers for a set of harmonized frequencies for the trains that ran across Europe, crossing one frontier after another. The problem was given to the frequency sub-working group R21 of the CEPT Radiocommunications Working Group (R-group). This was the body where Europe’s spectrum managers met to discuss questions of mutual interest.
Everyone in R21 knew that there was no use looking for any spare channels at VHF frequencies, as desirable as that may have been from an economic point of view. The only place to look was upwards and specifically around 450 MHz. However, finding something that aligned across Europeproved exceptionally difficult. After considerable efforts and many meetings a solution for 21 frequency pairs was found and approved in 1975. It demonstrated the exceptional difficulty of finding aligned spectrum across Europe.
By early 1970’s the shortage of frequencies for land mobile use was becoming a major problem in Europe. It was first discussed in R21 in 1973. The obvious step would have been to look upwards to the 900 MHz band. Unfortunately there was no international allocation for mobile services at 900 MHz. Therefore, the CEPT administrations started preparations for getting the ITU’s Radio Regulations modified at the next World Administrative Radio Conference (WARC) expected to take place later in the decade. In this same R21 meeting Mr. Per Åkerlind from Sweden suggested that one possibility for alleviating the frequency shortage would be the use of public cellular networks. The background to this suggestion is likely to have been the development of the NMT system by the four Nordic countries. The vision was inexpensive public mobile networks that could offer the most economic solution for the majority of those needing mobile communications.
In 1974 R21 took a cautious step towards the 900 MHz solution by drafting a recommendation requesting the CEPT administrations to agree to reserve 40 MHz for the mobile service in the next WARC but no specific reference was made to public networks.
The lesson learnt from the difficulties with the railway frequencies was that Europe’s spectrum managers had to get ahead of the curve and foresee future demands. The then Swiss Chairman of R21 Henry Kieffer had also been in the charge of the difficult work of finding the railway frequencies. He knew that the next pressure would come for the use of radio for public mobile networks. He strongly pressed this view in the 1975 meeting of R21 after the draft recommendation for the railway frequencies had been adopted.
In its 1976 meeting the R-Group took up the idea, and asked R21 to study possibilities for a future integrated network (land, maritime and aeronautical). Aeronautical was soon dropped from the proposal and this future network became known as the ‘integrated mobile network’ in R21 papers. In the following discussions it was felt that 200 channels would be sufficient for this international network with a preference of 900 MHz as its frequency band. This requirement would have meant a need of only 2 x 5 MHz bandwidth.
In the first instance this network was regarded as a long term project but when in 1978 it became clear that the 1979 WARC would be the last opportunity to amend the International radio regulations for some time R21 both confirmed this requirement and stated that the corresponding allocation should be worldwide. In WARC-79 the CEPT countries successfully got mobile services added at 900 MHz in the International Table of Frequency Allocations with an almost a worldwide allocation.
In September 1980, as a follow-up to WARC-79, R21 decided to give priority to an integrated public mobile system in the 900 MHz band and wrote a preliminary draft recommendation for it. But agreement had to be reached with the fixed service users already using the spectrum to concentrate their services (defense radio systems) in certain parts of the band. This would allow a total capacity of 1000 channels (2 x 25 MHz) for new civil mobile use. In the preliminary draft recommendation the central part of 480 channels (2 x 12 MHz) would be for the international use and the remaining 520 channels (2 x 13 MHz) for national networks. It was not yet the time to forget division of channels between international and national use, but the important step forward was the increase in the number of international channels from 200 to 480. The eventual maritime mobile service should have used the international part of the band.
By this time the world’s first automatic cellular mobile system, the Nordic NMT network working on 450 MHz, was in its construction phase. It was also the very first mobile standard to support international roaming. This development was to become an important influence in Europe. It extended the boundaries of cooperation not just from harmonization of spectrum and technical standards but more significantly to the harmonization and internalization of the mobile services themselves.
Perhaps the French had this Nordic cooperation in mind when, in October 1980, they called an informal two-day meeting in Paris to exchange views on mobile radio developments in CEPT countries and to look for common ground on the exploitation of the 900 MHz spectrum. Invitations went out to all those administrations having public mobile services, and experts from Austria, Denmark, Finland, France, Germany, Netherlands, Sweden, Switzerland and the United Kingdom attended.
It was the first time that officials from across Europe had sat down around a table and set out in any great detail what was going on in each country with their national public mobile services and the trend of thinking.
The meeting revealed divergent views on the prospects for public mobile radio services. The Nordic countries believed that there would be about 160,000 – 200,000 customers by 1990 and that this estimate may prove to be low. The Germans were only planning a capacity of 60,000 – 70,000 customers for their new network and thought this would be sufficient for them until 1990. The UK had a similar perspective and stated that did not see a number as high as 200,000 arising in the UK before 2000. France had an automatic network covering Paris with 4000 customers and their market studies had shown that there might be about 12,000 more customers interested in the service! The Nordic countries were clearly well ahead of the rest of Europe in understanding the potential of the mobile telephone.
There were also strongly divergent views expressed on the future of the portable mobile telephone. The UK and France thought there had to be a possibility for using portables (but not from inside buildings). Germanywas the least enthusiastic. They did not believe that sufficient amount of spectrum could be found to accommodate both a car telephone service and a service for portables. The portables would simply generate too much traffic. They were also worried that the low power of portables would complicate the control system particularly if they were used inside buildings. Most of all they were worried about traffic from the fixed network “escaping” onto the mobile network. This reflected the way many countries in Europe at the time did not view a mobile network as distinct from their fixed wire-line network.
Most fascinating of all was the discussion on the next generation of mobile technology. Three countries presented their national views. France was studying a digital system and believed it possible that data rates as high as 30 kb/s could be achieved – allowing a lot of error correction to be included (necessary for good quality telephone calls).Sweden thought that Time Division Duplex technology needed to be examined before frequency allocations were decided. Many in the room did not believe that digital speech coding would be advanced enough for at least another 10 – 15 years.
The UK suggested that the future might be Single Side Band analogue technology. In the UK standard 25 kHz wide radio channel bandwidth used for private mobile services were being cut to 12.5 kHz meet rising demand and research was underway to squeeze telephone calls into 5 kHz wide channels. The other countries present did not see this trend as appropriate for public mobile services at 900 MHz.
The most significant result of this informal Paris meeting was to have identified the scope for a much deeper engagement of European countries over the future cellular systems to be implemented in the new 900 MHz radio spectrum. The French Chairman suggested that a few countries should make a proposal for a new CEPT activity to address the technical and service issues that had come up in the discussion. Other countries present were not ready to move quite so quickly.
R21 finalized the 900 MHz recommendation in its meeting in spring 1981 with defense officials influencing the likely pace of implementation in some countries. Division between international and national use had vanished so that all 1000 channels (2 x 25 MHz) would be reserved for international land mobile and maritime mobile services. The only caveat was that eventual international maritime use should be concentrated to the center part of the band. The word “international” was understood to mean public. However, some countries indicated certain reservations on the availability of all these frequencies for mobile use. By the next R-group meeting in the beginning of 1982 major reservations could be withdrawn. The draft recommendation had also been sent to FCC and Canadian and Japanese administrations in the hope that at least the maritime mobile frequencies could be made worldwide, but as this proved impossible the interest in maritime use faded.
The Paris meeting was crucial in many respects. First, it revealed the potential consensus for using the capacity of 1000 channels for the future public mobile network (no division of channels between international and national use). This paved way for the final form of the 900 MHz recommendation. Second, this meant a single public mobile network standard could be developed for whole of the 900 MHz band. Third, it created the option for a cooperative development of such a system between all interested CEPT countries. Finally, digital modulation emerged on the agenda for the first time for such a future public mobile network. The Paris meeting had set the ball rolling.
The final 900 MHz recommendation was approved by the Telecommunications Commission (T-Com) in its June meeting in 1982. Two other CEPT working groups had also been studying how to proceed with the question of mobile networks and the Netherlands submitted a contribution to the meeting of the Harmonization Coordination Committee (CCH) which took place at the same time. The document was based on the idea that the work should be divided between three permanent working groups (SF, CS and R) and coordinated by CCH. The four Nordic countries strongly supported the Dutch initiative. However, due to the complexity of the task, they proposed that it should be left to a single “special” group of experts reporting regularly to T-Com. The idea of a special group was accepted with the proviso that instead of T-Com it should report to CCH. The mandate of this group, GSM, was finally accepted in a CCH meeting in November 1982 and GSM had its first meeting very soon after.
The work of R21 had made available a large number of channels for public mobile networks. This foresight by the frequency managers had opened the way for the cooperative development of the technical standard which in due time brought a compatible mobile radio service to everybody across Europe and most of the rest of the world. Building on the consensus in the Paris meeting it had been possible to avoid splitting the 900 MHz spectrum public mobile systems in Europe entirely on the basis of different national interests in the critical year of 1980. For its part T-Com decided to entrust the development of this future system to a “special” expert group. The next critical step would depend on the European operators within GSM and manufacturing industry to create a unique way of cooperating over a common mobile radio system standard for Europe. The informal meeting in Paris in October 1980 had laid the foundation for this extended chain of cooperation. It was amongst the most important turning points in European mobile radio history and perhaps every bit as critical as the Bonn meeting of Ministers in 1987 turned out to be for the GSM standard.
We would like to thank the personnel of the European Communications Office (ECO) for finding the relevant summary records. It would not have been possible to write this article without their help. We are also grateful to the old colleagues who have extended their help while gathering the information and delivered their memories of these years. We are also grateful to Krister Björnsjö, Thomas Haug and Klaus Olms who have extended their help. Finally, we owe very much to late Thormod Boe who had delivered his archive material to ECO.
The relevant summary records of the meetings of the Telecommunications Commission of CEPT, its Harmonization Coordination Committee, its Radiocommunications Working Group and its frequency sub-working group R21.
Personal report of the informal meeting inParison 8 – 9 October 1980 by Kalevi Teräsvuo.