Today’s wise technology decisions drive tomorrow’s superior profits. Knowing that, TDMA op-erators find themselves at an especially critical juncture. Faced with swapping out their networks for a stronger evolutionary path, these operators must assess how impending capital outlays will impact their long-term returns on investment and ability to generate more revenue from current subscribers while simultaneously attracting new users.
Over the past decade, TDMA thrived in the Americas, parts of Russia and a handful of other markets. Yet lacking a technological migration path to enable higher voice capacity and bandwidth-intensive mobile data services, TDMA has reached its limit. As such, TDMA operators recognize that their future success requires adoption of a different air interface technology which:
-Is suitable to the spectrum available
-Provides for significant revenue and growth opportunities
-Affords technical superiority though enhanced voice capacity and high data rates
-Future proofs their investment with backward and forward compatibility and flexible upgrade options
-Enables rapid deployment and minimal disruptions to existing users
TDMA operators are studying two different technology tracks: GSM/WCDMA or CDMA2000. A critical evaluation shows that the CDMA2000 path provides a technically and economically sound solution.
Third-generation (3G) CDMA2000 1X technology is being deployed at 800MHz, 1900MHz and 2.1GHz, and trials are ongoing at 450MHz in Europe. Just as incumbent cdmaOne™ operators are upgrading directly to 3G through the simple addition of CDMA2000 1X technology, TDMA operators switching to CDMA2000 would jump immediately to 3G via a single evolutionary step using technology that is available right now.
The advantage of this solitary investment is dramatic when compared to the four major technology steps and investments facing TDMA operators who follow the GSM evolutionary path. First comes the buildout of a new network based on GSM. Moving to GPRS entails a new network overlay and new air interface while subsequent migration to EDGE demands yet another new air interface. Though all three of those steps rely on TDMA-based technologies, the final step, deploying an entirely new WCDMA network, would use a completely different CDMA-based air interface.
Since demand for advanced services is uncertain and most demand over the next four to five years will be driven by voice, particularly in developing markets like Latin America, CDMA2000’s spectral efficiency is very well suited for real commercial needs. The technology requires only a small amount of spectrum—1.25MHz—for deployment of advanced services while preserving most capacity for voice.
Notwithstanding those benefits, CDMA2000 is a logical choice for TDMA operators because not all of them may attain enough spectrum to initiate a conversion to WCDMA, which requires a 5 MHz channel. Because CDMA2000 operates in 1.25 MHz bandwidth, operators can conserve valuable spectrum as they migrate to 3G, clearing a minimal amount of spectrum before they deploy their first CDMA RF channel.
Major operators have deployed cdmaOne across Asia and the Americas at 800MHz and 1900MHz. Because CDMA2000 is backward compatible and uses the same infrastructure as cdmaOne, ample quantities of network equipment are already widely available for any TDMA operators migrating to CDMA2000. While GSM infrastructure is reportedly being made available for deployment within the 800MHz band, significant quantities and timely delivery of GSM equipment for widespread deployments in this spectrum have yet to be seen.
There have not yet been any GSM800 deployments, and deployments of GSM in the 1900MHz band have created islands of coverage. If, however, TDMA operators in the Americas were to migrate instead to CDMA2000 at 800MHz and/or 1900MHz, their subscribers could roam onto the extensive array of cdmaOne networks already operating in that spectrum across the Americas and Asia.
CDMA2000’s flexible migration path offers operators a number of options for customizing their networks based on services they wish to provide. CDMA2000 1X deployed today enables delivery of data at speeds up to 144kbps for the forward link. Early next year, some CDMA2000 1X operators will start adding CDMA2000 1xEV-DO (Evolution-Data Only) for data delivery at 2.4Mbps using a dedicated data channel. The next evolution of CDMA2000 technology, CDMA2000 1xEV-DV (Evolution-Data and Voice), will offer data speeds exceeding 4.8Mbps with voice and data on a single channel.
Customer satisfaction with CDMA2000 1X is already evident. A recent Lehman Brothers report on KT Freetel’s experience with CDMA2000 1X in Korea shows that CDMA2000 1X subscribers are generating close to $5.00 in data average revenue per user (ARPU) and close to $7.70 data ARPU for those using devices with color screens. Further, the operator anticipates data ARPU may reach as high as $9.00 in 2002. In addition, Lehman notes that during July 2001, KTF CDMA2000 1X subscribers were consuming an average of 63 packets (512 bytes per packet) per day. This was more data consumption than that of NTT DoCoMo’s i-Mode (200 packets per day, at 128 bytes/packet) in Japan. Lehman says KTF feels CDMA2000 1X packet consumption could more than triple next year.
CDMA2000 1X also provides for longer handset battery life, thus increasing end-user convenience and satisfaction while encouraging more airtime use through longer talk times. These simple benefits set the stage for higher ARPUs, giving operators a financial advantage over operators using other technologies. According to Lehman Brothers, KTF’s strong voice and data ARPUs are enabling the operator to project positive free cash flow for its CDMA2000 1X network in three years and a complete recovery of its CDMA2000 1X capex costs in five to six years.
Though much attention has been focused on CDMA2000’s data capabilities, for many operators the real plus is the increased voice capacity. Such spectral efficiency is a substantial benefit as wireless penetration rates and minutes of use continue to escalate, requiring operators to do more with the same spectrum. CDMA2000 doubles the capacity of cdmaOne networks, a four to six time increase in capacity for TDMA networks.
For TDMA operators eyeing the switch to CDMA2000, it’s important to note that CDMA2000 uses the same ANSI-41 core network that TDMA operators use, whereas GSM relies on a completely different core network called MAP. Remaining within the common service creation environment of ANSI-41 enables service transparency for end users, making for a smoother migration to CDMA2000 for TDMA operators as well as subscribers.
Reusing IS-41 network elements provides for substantial cost savings through having a common MSC and HLR; common servers for SMS, VMS, WAP and other value-added services; plus a common Network Management System (NMS) for both networks.
Nonetheless, the GSM-WCDMA path has won notable converts, such as TDMA operator AT&T Wireless Services, which has so far implemented GSM/GPRS in a handful of U.S. markets at 1900MHz. Cingular Wireless also recently adopted this path for its TDMA networks. Yet AT&T Wireless’ selection has prompted concerns from investment analysts. In a September 2001 research report, Merrill Lynch & Co. cited risks concerning the timing and costs of AT&T Wireless’ technology choice. “We believe that the company’s planned migration to GSM/GPRS/EDGE/WCDMA technology is likely to be a difficult and expensive transition,” wrote the researchers.
Although the Japanese implementation of WCDMA has been deployed in Tokyo by NTT DoCoMo, many analysts predict WCDMA networks in Europe, China and other parts of Asia will not be widely deployed until at least 2004. Independent analyst Andrew Seybold has written that CDMA2000 1X systems will be up and running in Japan, the United States, Latin America and Eastern Europe before a second WCDMA system comes online. In a recent report, Morgan Stanley predicted that worldwide, CDMA2000 will have two to three times more users than WCDMA in the next few years.
Economies of scale from the GSM evolutionary path should be studied closely—particularly at 800MHz, which has a limited number of equipment suppliers. For instance, EDGE has been described as a solution for North and Latin American GSM operators who want to offer advanced services, but don’t have enough spectrum for WCDMA deployment. European operators have shown little interest in EDGE, thus limiting any possible financial advantage for that technology at 800MHz or anywhere else.
The outlook for WCDMA economies of scale is even worse, with substantial commercial deployment in Europe now postponed several years. In Japan, NTT DoCoMo expects to spend nearly $11 billion during the next three years building out its WCDMA network over a relatively small landmass. Such costs will overwhelm European 3G operators, particularly given the ponderous debt loads they assumed for UMTS/WCDMA licenses. Further, it’s obvious that not all of the licensed UMTS networks will be deployed. Broadband Mobile ASA, co-owned by Sonera and Enitel, already returned its Norwegian UMTS license.
One oft-overlooked fact is that the most accurate predictor of potential economies of scale is covered POPs as determined by available spectrum. CDMA2000 enjoys the benefit of 806 million announced POPs vs. 607 million POPs for J-WCDMA and E-WCDMA/UMTS combined. Those numbers reinforce the fact that CDMA2000 will show strong 3G economies of scale.
Regarding subscriber equipment, the few operators currently deploying GPRS are finding limited handset options. The situation looks worse for EDGE, should it be deployed, given the lack of interest in European markets. In the case of WCDMA, at launch there were only two compatible phones and one PC card on the market in Japan.
A similar problem can be expected with GSM800 handsets. The limited volumes of GSM800 terminals will likely mean higher costs. Hence, GSM800 terminals will not necessarily leverage the claimed global GSM economies of scale.
Prospects for economies of scale and ample handset supplies are even lower when one considers mobiles based on GAIT (GSM/ANSI-136 Interoperability Team), a technological approach designed to allow roaming between GSM and TDMA networks. Only three manufacturers have announced that they will supply the necessary GAIT phones, raising questions about handset availability and the limited range of handset models that will likely greet end users. In addition, these phones are not slated for availability until summer 2002.
On the other hand, terminal providers serving the cdmaOne community are rapidly integrating CDMA2000 1X into their products. Twenty-four CDMA2000 1X handsets are available today, and their numbers will grow substantially in coming months. More than 43 designs of CDMA2000 1X handsets and PC cards are being developed by manufacturers worldwide for availability beginning in late 2001 and throughout 2002. The time-to-market advantage translates into better economies and pricing for CDMA2000 1X phones. In Korea today, CDMA2000 1X phones are selling in the rage of $233 to $310. The Gartner Group predicts this price will come down to below $100 in Asia within two years.
As for the international roaming issue, cdmaOne itself is rapidly spreading across regional strongholds. For instance, in June 2001 China Unicom signed historic roaming Memoranda of Understanding with 13 global CDMA operators. Recently, Japan’s KDDI announced agreements with eight countries across four continents. And as noted earlier, because cdmaOne is widely dispersed across the Americas at 800MHz, operators adopting that technology have immediate access to significant regional roaming opportunities.
Recall that operators deploying GSM800 will still need GSM1900 and GSM1800 handset capability to enable international roaming. For single-mode GSM800 handsets, the roaming will be limited to only AT&T and Cingular in the United States and a handful of other operators elsewhere choosing that path. To enable roaming with WCDMA, a GSM800 operator may need quad-band (800/1900/1800/2100) handsets with multi mode (GSM/GPRS/EDGE/WCDMA) capability. These devices will be expensive and not readily available.
Each time a new frequency band or technology mode is added, direct costs are entailed for com-ponents, memory and processing power, while indirect costs are also prompted due to testing complexity and development time. Multi-band and multi-mode handsets also affect economies of scale because many of markets simply don’t need the extra capabilities except for specialized in-ternational roaming.
Meanwhile, global efforts to develop multi-standard roaming will enable subscribers to travel across networks using disparate air interface protocols. The CDMA Development Group is working with the GSM Global Roaming Forum to define commercial standards and technical requirements for multi-standard roaming. Several companies are developing chipset solutions to make multi-standard roaming a reality, and CDMA/GSM roaming via a removable user identity module (R-UIM) has been demonstrated. Also, in October 2001 Lucent Technologies’ Bell Labs announced a software architecture called Common Operations (COPS) that features a protocol gateway to enable multi-standard roaming. The bottom line is that roaming ability will mainly be dictated by business relationships between operators, not by the technologies they deploy.
Driving home this point is the recent agreement between Virgin Mobile, a UK-based mobile virtual network operator (MVNO), and US network operator Sprint PCS. Rather than partnering with a GSM operator, whose technology would ostensibly complement Virgin’s other MVNO operations in the United Kingdom and Australia, Virgin chose instead to work with a US cdmaOne operator.
Adam Guy, Strategis Group senior analyst for Mobile Wireless Research, observes that such an MVNO agreement is perfect for Sprint PCS, which has extra network capacity. He adds, “The timing of the launch of Virgin Mobile USA in conjunction with the deployment of Sprint PCS' CDMA2000 1X network is not a coincidence,” since the CDMA2000 1X upgrade promises nearly doubled voice capacity. CDMA2000 1X will also enable a host of data services that Virgin Mobile USA might add to its MVNO offerings.
Most importantly, CDMA2000 is delivering on 3G’s promise today. CDMA2000 1X has been commercial for over one year. There are more than 2.5 million CDMA2000 users in Korea downloading games, photos and other data at an average of 60-80kbps. With the software and handset upgrades planned for early next year, data speeds can go up to 307kbps. These customers generate on average 41% more revenue than 2G subscribers do, and the operators report 1.6 to 1.8 capacity improvement in a highly loaded network.
In short, affirmative analyst reports, strong vendor commitments, and ongoing operator triumphs all point to CDMA2000 as a viable option for TDMA operators seeking a clear and profitable migration path. CDMA2000 provides all the ingredients for short-term and long-term success.
Extract from article that will appear in Business Briefing Wireless Technology 2002