US : In its newest white paper, In-Band W-CDMA: The Commercial Potential for 850, 900, 1800, and 1900 MHz Deployment, The Shosteck Group examines the potential for “in-band” W-CDMA -- W-CDMA deployed in frequencies already used for GSM, rather than dedicated frequencies to which it is or would be allocated.

THE CASE OF CINGULAR
Cingular is not alone among GSM operators with in-band W-CDMA as its only 3G alternative. However, Cingular is unique in that it is the first operator to deploy it For this reason, it will serve as a global model for the in-band W-CDMA experience.

THE CASE FOR OPERATORS AROUND THE WORLD

So far, the mobile world has viewed in-band W-CDMA as a uniquely American story. However, its potential extends beyond the U.S – into Latin America, Europe, China and India.

GSM operators recognize that at some point, they will need the greater capacity and the functionality of high-speed data applications. To serve those needs, they will have to migrate to 3G. When they do so, in-band W-CDMA, in particular W-CDMA 850/900, may be the preferred choice.

Though operators may recognize in theory the need for more capacity, the reality may prove different. Operators must consider how feasible in-band W-CDMA will be in terms of the technical, commercial, and regulatory issues that may affect its deployment. Three stand out.

First, will regulators authorize in-band W-CDMA? If regulators do not accept in-band W-CDMA, its advantages are moot.

Second, even if regulators allow in-band W-CDMA, will operators be able to deploy it? In core urban areas, GSM networks may be at capacity. In such places, it will be difficult for operators to clear the 5 MHz of spectrum needed for deploying in-band W-CDMA.

Third, to what extent will CDMA 2000 1x EV-DO and wireless broadband solutions such as WiFi, serve as competitive drivers? This will be especially relevant in Latin America where, in common with Cingular, GSM operators use 850 MHz . However, it may also apply to India and China, where GSM 900 operators may feel competitive pressures from those using CDMA 850.

LATIN AMERICA
Regulatory constraints aside, Latin America’s GSM operators are hesitant to move toward W-CDMA for cost reasons. ARPU is lower. Operators have spent heavily for 2G and 2.5G. Regardless of transition path, all are reluctant to spend more, at least for now.

However, by waiting until they need capacity, Latin American operators may face a more daunting task in launching in-band W-CDMA than if they do so early when they still have unencumbered spectrum. Ironically, this radio engineering reality argues for GSM 850/1900 operators to consider their strategic deployment of W-CDMA sooner, rather than later.

By allowing lower-cost W-CDMA 850, regulators can encourage earlier and more extensive deployment of W-CDMA, than otherwise would be the case.

EUROPE

Though European attention to in-band W-CDMA, in particular 900 MHz, would seem superfluous, it is not.

European standards bodies are already evaluating W-CDMA in the GSM 900/1800 frequencies. The standard for in-band 900/1800 W-CDMA will be set by the end of 2005, or near to it.

European operators will likely prefer W-CDMA 900 to W-CDMA 1800 because of the favorable propagation characteristics of the former, in particular for rural and suburban areas.

However, the deployment of W-CDMA 1900/2100 now in progress, and the robustness of GSM/GPRS networks, points to a long delay before the commercial launch of in-band W-CDMA in Europe. Opinion varies on when this will begin, ranging from mid-2009 to late 2012. European operators will not migrate to W-CDMA 900 until the business case warrants – however, this might be closer to 2009.

INDIA AND CHINA
Three factors likely preclude Indian operators from considering in-band W-CDMA into the foreseeable future. First, and unique to India, is the contention between CDMA and GSM operators over the 1900 MHz band. Second, and in common with Europe, is the absence of a standard. Third is the unavailability of equipment.

Overall, developments in Indian and Chinese markets have raised considerable political and economic barriers to the introduction of in-band W-CDMA at 900/1800 MHz. It is very unlikely to be deployed in those countries into the foreseeable future. This means that over the next few years, in-band W-CDMA will be a story of the Americas, primarily North America, and will be available only at 850/1900 MHz.

FOUR KEY ADVANTAGES OF IN-BAND W-CDMA

In-Band W-CDMA: The Commercial Potential for 850, 900, 1800, and 1900 MHz Deployment” analyzes four advantages of in-band beyond the greater capacity, higher data rate, and lower cost of W-CDMA, in general.
First, where regulators have not yet assigned 3G spectrum, in-band W-CDMA enables operators to deploy W-CDMA in response to immediate market needs. Over the near- to mid-term, in-band will be most relevant for operators in Latin America, India, and China, where regulators have not assigned dedicated W-CDMA spectrum.

Second, and related to the above, regulatory permission to deploy in-band W-CDMA would likely rule out spectrum auctions or fees that might otherwise be imposed.

Third, the majority of operators deploy GSM at 850 or 900 MHz. Radio signals at these frequencies propagate for twice the distance as do those at 1700/2100 and 1900/2100 MHz. This means that operators with 850 or 900 MHz assignments could deploy W-CDMA for roughly one-half the infrastructure cost required at the higher frequencies.

Fourth, the 850/900 MHz frequencies provide deeper in-building penetration than do the 1700/1900/2100 MHz frequencies. If regulators grant permission for in-band W-CDMA, operators could extend coverage more rapidly, and serve more subscribers, than otherwise would be commercially feasible.

INTERFERENCE ISSUES
The paper points out that interference will be an early issue, not to be underestimated. Launching W-CDMA on 850 MHz spectrum that is already crowded with TDMA, GSM, CDMA, and in some cases AMPS, will generate interference.

The press will have a feeding frenzy over Cingular’s initial tribulations. In common with the early difficulties of NTT DoCoMo and Hutchison “3,” Cingular will suffer highly publicized ridicule over W-CDMA’s initial failures.

However, these will be the normal problems common to any new radio technology. It may take Cingular 12 to 24 months longer to deploy than it anticipates. However, Cingular’s vendors are not blind to this. There’s too much at stake for them to let Cingular fail.

HANDSET COSTS
However, handset costs will be an issue. By year-end 2007, W-CDMA 850/1900 handsets may still cost operators $200, while GSM/GPRS models will cost as little as $50.

A lower-cost solution could come if operators insist that vendors provide low-cost voice-centric handsets with minimum data functionality. Without it, higher-cost handsets will inhibit adoption of W-CDMA. This will provide openings for Chinese vendors who typically go after markets which might not be perceived as ‘lucrative’ but which will allow them to gain share at the expense of European vendors.

In-Band W-CDMA: The Commercial Potential for 850, 900, 1800, and 1900 MHz Deployment is available free of charge on The Shosteck Group website.