Winning at 3G

A year ago, it seemed like the 3G winner already had been decided. Wideband CDMA (UMTS), building on the success of GSM, was far ahead in the development of standards, and it looked like the migration from GSM to GPRS to EDGE to WCDMA, although awkward, was going to be accomplished in record time. 3GPP was even winding up work on the first commercial release of its 3GPP standard.

By contrast, the portion of the world running on the ANSI-41 network seemed stuck without direction. What was once the envy of the world, with North-America-wide compatibility between AMPS systems, had splintered into AMPS, N-AMPS, D-AMPS, CDMA and GSM. Europe, which had started with many incompatible analog standards, had transitioned successfully in the mid-1990s to one digital standard, and now is claiming about half a billion customers on its networks. America's 100 million seems trivial by comparison.

ANSI-41, the signaling network that supported all North American wireless standards except GSM, struggled to provide compatibility and to move ahead despite direction by companies that were often bitter technological foes. The defection of AT&T Wireless from the ANSI-41 camp was, in one way, a victory for WCDMA, especially when combined with rumors or announcements that other major ANSI-136 carriers such as Cingular and Rogers would join. However, this move also had the effect of solidifying ANSI-41 as the network for the ANSI-95/IS-2000 CDMA family of standards.

Victories With 1XRTT

The 3GPP2 camp (responsible for the development of 3G versions of cdmaOne and cdma2000 standards) can claim some recent victories with the implementation of 1XRTT systems in Korea that meet the ITU definition of 3G data rates (144kb/s in a mobile environment). 3GPP currently only can claim GPRS, a data service running over GSM channels that does not quite meet 3G data rates even at its maximum rate. GPRS also recently was bedeviled by arguments over critical aspects of the standard. This left Motorola as the only company producing quantities of terminals, but also taking a risk that its terminals would all be made obsolete. NTT in Japan also stumbled when the commercial launch of WCDMA, promised in May 2001, was pushed back to October and then turned out to be limited in scope. Even then, its system was not compatible with European versions of WCDMA.

Standards, however, are only part of the 3G story. A bigger problem is that technologists can generate high-speed data services much faster than carriers can figure out how to sell them and consumers can figure out which ones they cannot live without. SMS has turned out to be a desirable data capability using low bandwidths. This was undeniably a victory for GSM because AMPS could not support short messaging, and ANSI-41 could not be used to send messages directly between TDMA and CDMA terminals, let alone to GSM.

Voice Still Rules

The demand for text-based e-mail, using considerably more bandwidth, is growing, with devices such as the RIM BlackBerry becoming popular and newer ones integrating voice and e-mail services. Web services such as WAP have not proved as popular as Japan's i-mode because so many North Americans are spoiled by cheap, high-speed services with much more capable user interfaces at home. Even with these recent victories for data, the killer app on cellular and PCS networks still is voice.

European carriers who paid enormous sums of money for 3G spectrum cannot afford to wait for the market demand to jell. They must load millions of subscribers on their systems almost at once, but that simply is not going to happen. The only service that can be guaranteed to bring customers along in droves is voice, and that hardly makes the best use of spectrum set aside for 3G.

One of the conundrums for carriers is that voice may pay them better than data, at least on a bit-by-bit basis. Digital voice has very modest data demands (8kb/s to 13kb/s), and these demands are as likely to decrease in the future as they are to increase. A focus on bit rates leads to a pricing problem. Should an 80kb/s data user be charged at 10 times the voice rate? What about 800kb/s?

What About Wi-Fi?

A dark horse in the 3G contest is 802.11 (Wi-Fi). It may have a big effect on 3G because companies that have installed these systems using unlicensed spectrum in their offices certainly will not want their employees using pay-by-the-minute 3G systems when the corporate system has unused capacity. The success of Wi-Fi in serving the public marketplace is less sure. Cellular and PCS carriers have much more experience running nationwide systems with mobile devices, providing fraud protection, billing and roaming services.

Wireless carriers need to sell data services (whether 3G or not) based on their service's advantages. Going head to head against cable modems and DSL in urban areas is a losing proposition. The big advantage of wireless is mobility. It is not smart to pretend that a wireless device will have the capabilities of an office computer for the foreseeable future. But it can compete nicely with wired phones and PDAs.

In rural areas, the picture is different. Here, spectrum is cheap; wiring high-speed connections is difficult and expensive; and broadband wireless (point to multipoint) is moribund. 3G services could become a valuable adjunct to wireless carriers with a thin population base. Adding data to their voice packages could increase their revenues while helping to justify a transition from analog or 2G digital to a full 3G system, which also will provide more voice capabilities, as well as integrated voice/data services.

If 3G is, indeed, a contest, the unusual thing is that all contenders that finish will pay out. The odds right now appear pretty even.

---

Crowe (crowed@cnp-wireless.com) is a wireless-standards consultant and editor of Cellular Networking Perspectives, a wireless-standards and -technology bulletin.