5   Summary and Conclusions

This white paper provides an overview of options open to TDMA/IS-136 operators as they choose a migration path to third generation (3G) and addresses issues in choosing a path. It specifically compares GSM and cdmaOne/CDMA2000 1X as migration alternatives.

Perhaps its most important value lies in identifying the implementation challenges that may hinder that migration. By identifying such challenges, operators can be more watchful of transition pitfalls and question their vendors more closely concerning viable means of avoiding those pitfalls and implementing cost-effective alternatives.

We warn operators (and vendors) against focusing on high data rates. As SK Telecom emphasizes, the issue is not high data rates, but profitable data rates. Over the short- to mid-term, these may well be in the range of 30 to 50 kbps. Looked at in this context, for operators to choose—or for vendors to promote—technologies because of their ultimate speeds becomes self-defeating.

We note the daunting engineering challenges inherent in deploying EDGE and, because of them, the possibility that EDGE may never reach commercial fruition. Or should it reach fruition, we point to the likelihood that some GSM operators will bypass it to migrate directly to UMTS.

We describe and document the historical “reality gap” between when vendors promise handsets and when they actually deliver. Tied to the complexity and expense of dual-mode handsets, this virtually ensures their delay to market and a high price tag when they eventually arrive. The dual-mode nature of these phones precludes the economies of scale, which a single-mode device would provide. This will apply to TDMA-GSM 800 handsets and, if produced, TDMA-CDMA handsets, as well. In contrast, single-mode CDMA2000 1X handsets are already available and being offered at low and declining market prices. Single-mode GSM handsets, when and if produced, will be available at low prices, as well.

Of special importance, we reiterate that UMTS spectrum has not been licensed in the Americas. This means that TDMA/IS-136 operators who choose GSM have no foreseeable migration path to full 3G. In contrast, CDMA2000 1X is becoming available, is enabling 3G services on present spectrum, and is providing capacity increases of 50 percent or more.

We note that AT&T, which precipitated the move of TDMA/IS-136 operators toward GSM, may be a special case. Unlike Cingular, AT&T’s major TDMA rival, AT&T holds 1900 MHz spectrum in 9 of the 12 largest markets. Cingular holds 1900 MHz spectrum in only four. This enables AT&T to migrate using GSM 1900. With little 1900 MHz spectrum, Cingular is much more constrained.

We pay close attention to what may be required to deploy GSM onto a TDMA/IS-136 network. In addition to the challenge of developing dual-mode TDMA-GSM handsets, two seldom-discussed factors emerge.

1. The crowding of the 800 MHz spectrum. No one in the world has experience in deploying GSM at 800 MHz and in dealing with the potential interference from different RF technologies. Two of these, AMPS and iDEN, are unique at 800 MHz. Good engineers will overcome such interference. However, doing so will take time and money.
2. The challenge of integrating the incompatible MAP signaling used by GSM with the ANSI-41 signaling used by TDMA/IS-136. Again, no one in the world has ever done this. It would be disingenuous to think that it will prove an easy process.

Not to be overlooked is the question of whether to use dual-mode handsets at all. A powerful argument can be made to forego dual-mode handsets and use only “pure” CDMA2000 1X or GSM 800. This bypasses many, but not all, of the issues of network integration. By building a pure network, operators would spend money on infrastructure rather than handset subsidies.

Throughout our analysis, we return to the themes of costs and profits—or the phrase we sometimes use ,“commercial” relevance. Related to costs and profits is the theme of maintaining a satisfactory end-user experience throughout the transition. A poor end-user experience creates churn. In addition to being costly humiliations, the failed promises of WAP and GPRS alienated end-users to the operators’ detriment.

Overall, we indicate the possible challenges that TDMA/IS-136 operators may face in deploying GSM and the possible advantages that cdmaOne may offer. Based on these possible advantages, we suggest that TDMA operators may find it worthwhile to consider cdmaOne/CDMA2000 1X as a migration path.

Depending upon their specific circumstances, some TDMA/IS-136 operators may choose CDMA2000 1X as the better path for them. This may happen more for those who are licensed primarily to 800 MHz spectrum. Others may choose GSM. This may happen more for those who are licensed primarily to 1900 MHz spectrum.

In sum, we are not endorsing CDMA2000 1X as the transition technology from TDMA/IS-136 to 3G. Rather, we are saying that it appears to offer a promising story that some TDMA operators will find worth considering.