GSM made its first inroads into the US in New York in December 1995, under the name of PCS1900/GSM. This was one local system among many in the area and failed to spread across the country.
The breakthrough was the result of a long process from 1997–1999 in which Håkan Eriksson, head of technology at Ericsson’s American business unit, persuaded AT&T Wireless to switch technologies.
This can be traced back to AT&T Wireless reaching the conclusion that TDMA, the American standard that ERA had launched, could not provide the data speeds required to compete with GSM/EDGE or CDMA. “The background was that we had opted to make the migration from AMPS to TDMA as simple as possible and so we had retained the AMPS channel bandwidth of 30 kHz. For GSM, the channel bandwidth was 200 kHz,” says Eriksson.
This gave GSM/EDGE an advantage over TDMA by a factor of about seven for the potential data-transfer speeds for an individual user. It also raised the question of which 3G variant AT&T should opt for; this was, however, of only theoretical importance to begin with, while the competition between GSM and CDMA was already unavoidable.
In principle there were four technical alternatives: Multi-Carrier TDMA, OFDM, EDGE and CDMA. The first was complicated, had never been tested in practice and did not offer a good transition to 3G. The second was undeveloped and risked turning into a special solution for one customer. The third would offer advantages of scale and was based on GSM but would require the phones to cope with two technologies at the same time. The fourth was technologically feasible but a commercial no-no.
Then there was the question of who would supply the solution. Ericsson’s American market unit wanted to take the Multi-Carrier TDMA route, because it feared losing AT&T Wireless as a customer. The GSM business unit should have been interested in the EDGE alternative on principle but shied away from the special solutions it would have had to develop to the base stations. Ericsson’s telephone business unit definitely did not want to manufacture a combined TDMA/EDGE phone because the filter requirements and other elements would be far too complex. When Eriksson approached Nokia, he received the same reaction as from Ericsson’s telephone unit.
“In spite of everything the most natural alternative was TDMA/EDGE,” he says. “This was the only technology that had been developed and was an approach that would introduce the customer to the world standard that GSM was in the process of becoming. My problem was that nobody wanted to do it – not the customer, not the business unit for telephones and not Nokia.”
Author: Svenolof Karlsson & Anders Lugn