The origins of GSM · 1966–1979
Before GSM: opening Europe’s airwaves
Before Europe could agree on a mobile phone that crossed borders, its railways asked for something more modest. They wanted radio frequencies they could use from country to country. Finding just 21 pairs of radio channels took nine years. Out of that difficult work came a lesson that would help prepare the way for a mobile network nobody had yet designed.
A railway journey meets a national boundary
In 1966, the International Union of Railways brought Europe’s spectrum managers a problem that no single country could solve. Trains crossing borders needed frequencies they could use along an international route without interruption. The request went to R21, the frequency sub-group of CEPT’s Radiocommunications Working Group, where European administrations discussed their shared radio problems.
A frequency is a place on the radio dial. A band is a range of those frequencies. The familiar VHF bands had no spare channels, so the specialists looked higher, towards 450 MHz. Finding frequencies that could be aligned across Europe proved exceptionally difficult.
Henry Kieffer, the group’s Swiss chairman, led the work. It took considerable effort and many meetings before a solution for just 21 frequency pairs was approved in 1975. Nine years had passed since the railways’ request.
The railway agreement closed one negotiation. Kieffer would use its lesson to argue for a different approach to the next.1
Plan for the demand that has not arrived
Even before the railway problem was settled, the shortage of frequencies for mobile use on land was becoming a wider European concern. In 1973, Sweden’s Per Åkerlind suggested a different way of meeting it. Public cellular networks could serve many users economically, rather than treating each new mobile requirement as another search for separate channels.
There was a possible home for these services around 900 MHz, higher up the radio dial. International rules did not yet provide the mobile allocation Europe needed there. The European administrations began preparing to have those rules changed at the next World Administrative Radio Conference. In 1974, the group drafted a proposal to reserve 40 MHz for mobile services, though it did not yet single out public networks.
Kieffer pressed the case for those future public networks in 1975, after the railway recommendation had been adopted. His experience gave the argument practical force. Rather than leave new services facing another prolonged search for frequencies, Europe should plan for their needs in advance.1
The future was still taking shape
The network they began to imagine was broader than a telephone service for people on land. In 1976, R21’s parent Radiocommunications Working Group asked it to study a system for land, maritime and aeronautical users. Aeronautical use soon dropped out, leaving what the papers called an “integrated mobile network”. The ambition to bring different users into one network was taking shape before its detailed technology had been settled.
An initial estimate suggested that 200 channels, preferably at 900 MHz, would be enough, requiring two blocks of 5 MHz. This was a tentative figure for the proposed integrated network, not a revision of the earlier, broader 40 MHz proposal for mobile services. The 21 frequency pairs secured for the railways had addressed a different problem altogether.
At first, the network seemed a long-term project. Then the calendar became urgent. In 1978, the spectrum planners recognised that the following year’s conference would be the last opportunity for some time to amend the international radio rules. They confirmed the network’s requirement and called for a worldwide allocation. Waiting for a finished technical design risked missing the chance to secure its place on the airwaves.1
An opening in the international rulebook
From 24 September to 6 December 1979, delegates gathered in Geneva for the World Administrative Radio Conference, WARC-79. The European countries secured the addition of mobile services around 900 MHz in the international allocation table. The proposed network was still to be designed, but an obstacle that Europe could not remove on its own had been overcome.
This was permission to proceed, not an empty band waiting for GSM. The table listed mobile alongside other services, with different arrangements across regions and countries. Existing users remained. No common technical standard had been chosen. Countries would still have to make frequencies available before the international opening could become a usable service.
The railway request had set off a long search for room in already crowded bands. This time, the planners had secured an international opening for a network that did not yet exist. They had not made GSM inevitable. The countries could now discuss a shared system knowing that the international rules made room for it.12
Space for a service, not yet a system
In September 1980, the frequency group gave priority to an integrated public mobile system at 900 MHz. Existing fixed-service users, including defence systems, still had to agree to concentrate their use in parts of the band. The next stage would depend on national decisions as well as the international agreement.
The unexpected link between railway radio and GSM was a lesson in preparing for the future. Kieffer’s experience helped strengthen the case for anticipating public mobile demand. It did not supply GSM’s technology. Alongside Åkerlind’s proposal and the wider work of Europe’s spectrum managers, it helped create an opportunity for countries to build something together.
Europe had begun making room for a common mobile system. What remained was an agreement about the service itself, from the people it would serve to the technology it would use.1
One band. Nine countries. Different futures
In Paris in October 1980, Nordic delegates envisaged as many as 200,000 mobile customers by 1990. British officials did not expect that many in the UK before 2000. German delegates feared there would not be enough spectrum for portable phones as well as car telephones. Many participants thought practical digital speech was still ten or fifteen years away.
Europe had opened the door to a shared mobile network. In Paris, nine countries were about to discover that agreeing on the airwaves was easier than agreeing on the future.1
Read the next chapter →
Finding frequencies was only the beginning. Follow the agreements, technical work and difficult decisions that brought a common European mobile system together.
Sources and original documents
- Stephen Temple, The Beginnings: Europe’s spectrum managers. The original account draws on CEPT summary records, colleagues’ recollections and Kalevi Teräsvuo’s account of the Paris meeting. It acknowledges the European Communications Office’s help in recovering the records.
- ITU, Final Acts of the World Administrative Radio Conference (Geneva, 1979), published in 1980. PDF page 111 contains the spread labelled RR8-89–RR8-90, including the 890–960 MHz allocation table on RR8-89 and associated provisions on RR8-90. The ITU’s official historical overview, WARC-79 — 30 years and still significant!, records the conference dates and context.
- Home Office: Frequency Requirements (1981). This five-page minute illustrates the national frequency decisions that still had to be made after the international spectrum agreement, including defence requirements, BT’s needs and the possibility of a competing mobile operator.
Where the opening led
Follow the story through the Timeline
Follow the milestones from the early spectrum decisions through the creation of GSM and the later mobile generations.
Inside the Mobile Revolution: UK Seeds of the New Beginning
Read Stephen Temple’s first-hand account of how spectrum became an important part of the negotiations towards a common European mobile system.