Good luck, elmat!
In the meantime, a few more quack, quacks from EMC. Long story short, Sprint and Verizon set to kick ass in the US:
Close
By Mike Woolfrey, Research Analyst - EMC
The recent GPRS launches of Cingular Wireless () and AT&T Wireless () could create the impression that all is rosy for GSM/TDMA operators in the USA, as these enhanced data services are the first to be offered to US consumers. However the road ahead for the number two and three ranked carriers is likely to be a very rough one.
Source: EMC World Cellular Database
Discussion on the various merits of each technology system has been ongoing for many years, but AT&T Wireless' decision to migrate from TDMA to GSM/GPRS/EDGE/W-CDMA () only served to intensify the battle. AT&T Wireless' decision was quickly followed by the announcement that its Canadian partner, Rogers AT&T Wireless was to follow the same strategy. It was naturally assumed that Cingular Wireless would follow suit and announce plans to move its current TDMA subscribers across to GSM, as it has been operating both TDMA services and GSM since 1996 through the PacBell Wireless and BellSouth Mobility DCS networks. Confirmation, however, that Cingular is to overlay its TDMA systems with GSM is still outstanding.
Whilst AT&T has been building its new GSM network and Cingular has been upgrading its GSM properties to GPRS, the CDMA operators, Verizon and Sprint PCS, have been making their move to 3G systems. Both have signed infrastructure deals totalling around $7.5 billion and intend to launch cdma2000 1x services in Q4 2001. These launches could potentially create a large change in the US cellular market. One possible outcome is a significant increase in tariff competition, similar to that of the introduction of the One Rate National calling plans in 1999. Whilst the exact markets that are to be launched are unknown, it is conceivable that the GPRS systems of AT&T and Cingular could be competing with Verizon and Sprint's 1x systems. The wireless industry has now finally accepted that technology does not sell, so merely offering 1x against GPRS would be fruitless. What is at issue is how the carriers use these technologies to market services.
It is quite possible that the CDMA carriers will not use data-capability as a major part of its marketing push. The cellular industry remains unsure of the exact role data is likely to play in the future and there are great uncertainties as to when or whether mass market applications can be developed. Those who have wanted pure data or email applications have found them through offerings such as Cingular Interactive and Go America via a PDA or a BlackBerry. That is not to say that CDMA carriers will totally ignore data. Sprint PCS currently offers data services through its Wireless Web, with around 1.5 million of its 11.5 million subscribers using the service, but these users only contribute about $1 a month to ARPU. Verizon Wireless' Mobile Web users stand at around the one million mark, but again these are believed to have a minimal effect on ARPU. The TDMA/GSM carriers have smaller numbers of data subscribers contributing small amounts to ARPU.
Source: EMC World Cellular Database
Data revenues, and indeed revenues as a whole for western European operators have been largely propped up by SMS. To date messaging has not been a huge success in the USA. This is down to a number of factors, but mainly the lack of network and device interoperability. 2001 has seen these problems resolved, with the leading operators now offering SMS or messaging. To this extent some operators describe messaging customers as mobile data customers, and whilst this is essentially true, messaging is unlikely on its own to offer a business case for data over 2.5 or 3G networks.
Targeting voice
So in the near future it is widely accepted that voice will dominate the marketplace. Reasons for this include the facts that those who require data have had access to it via other methods (Cingular Interactive, BlackBerry and others). Additionally, whilst there remain no real applications for data, operators are unlikely to be attracting new customers. This is not to say that operators will not offer data services. Verizon has declared its support for Qualcomm's BREW platform () and Sprint PCS is also believed to be exploring the options of putting BREW on its network.
The introduction of GPRS networks for Cingular and AT&T will have no positive effect on voice capacity, in fact it is likely to add strain to the spectrum that will be potentially supporting TDMA-1900, GSM-1900 and GPRS services. With TDMA-1900, users are able to fall back onto TDMA-800 or AMPS coverage, this is a luxury GSM users will not have. GSM-800 equipment is being developed, but is unlikely to be in the market until 2002, handsets may arrive even later. Again it must be stressed that Cingular is yet to confirm its migration path for its TDMA operations: if it were to opt for GSM, it is more likely to deploy in the 800MHz band, as it has significantly less PCS (1900MHz) spectrum.
Verizon and Sprint PCS' network upgrade will not be confronted by the same issue. Both have chosen to deploy cdma2000 1x RTT (followed by 1x EV-DO), with networks expected to be launched in Q4 2001. With the deployment of the 1x RTT network, the CDMA operators will not only be capable of achieving data speeds of around 80Kbps (peak 153Kbps), but they will also not have to trade voice against data. This is because voice and data operate in the same channel, with no separate spectrum restricted to data. Depending on the network load, a user may experience varying data speeds - some operators are limiting peak speeds to 60-80Kbps - therefore preventing data reducing voice capacity.
The 1x upgrade only requires a new channel card and a software upload. In moving from cdmaOne to cdma2000 1x RTT manufacturers are promoting an increase of anything to 50-100% in voice capacity. This is feasible, as long as all handsets have cdma2000 chipsets and all channel cards are upgraded (seven 1.25MHz channels in 10MHz of spectrum). Whilst cdma2000 handsets are available in quantity, (LG Electronics, Kyocera, Samsung, Sanyo, Motorola and Ericsson are expected to have handsets for the USA market in Q4 2001) and easily upgradeable from cdmaOne designs, it is unlikely that all handsets will be 1x-enabled as soon as networks are launched. However 2002 is likely to see the majority of handsets being manufactured with cdma2000 chipsets. The onus is then on the operator to switch out old cdmaOne channel cards and replace them with 1x versions, by doing so voice capacity will be increased. Whilst this is a fairly simple procedure, it will not come without expense. Operators will have to evaluate the benefits for voice and data traffic against the cost of upgrading all channel cards in a base station.
In the short term, operators are likely to offer incentives to high-end users to upgrade handsets (as was the case when networks moved from AMPS to digital), and at the same time upgrade channel cards in desired areas. Verizon's announcement that New York will be its first cdma2000 upgrade(), a market that is believed to be at the limit of capacity indicates that it is being driven by voice. Not all markets are going to be easily upgraded to 1x: those with Motorola's radios are believed to be lagging behind and will not be ready until H2 2002. That means markets such as Chicago (Sprint) and Atlanta (Verizon) will be delayed. Verizon also has to continue to migrate its ten million AMPS users (June 2001) to some sort of CDMA service.
Comparison
The problems that the CDMA operators will experience on the route to 1x is minimal to those of the GSM/TDMA operators. AT&T Wireless has seen its business grow through its extensive TDMA footprint, backed up by AMPS. This footprint will not be matched by GSM coverage. AT&T plans to have 40% or around 85 million of its POPs covered by year end 2001. This is an ambitious roll-out timetable, especially compared to the fact that it has taken Sprint PCS five years to reach 185 million with CDMA. It would be possible for AT&T to extend its coverage by roaming onto Cingular and VoiceStream's GSM networks and negotiations regarding such a move are believed to be nearly complete. This would certainly help fill the short term voice coverage issues, but this is not likely to come cheap to AT&T or the user. There would also be GPRS roaming issues to be resolved. At present GPRS roaming is limited, as a dedicated GRX (GPRS Roaming Exchange) and private IP network are needed. So new potential AT&T Wireless customers will have the choice of patchy network coverage for GSM/GPRS or good coverage with limited data services with TDMA. It remains to be seen just how much handset choice a consumer would have for GSM (still a minority system) or TDMA (dwindling technology). The problems are likely to be greater with the introduction of EDGE, GAIT or W-CDMA and any potential device shortage.
Cingular Wireless continues to operate its three different technology systems, and as previously stated it is yet to determine its next move for 3G services.
Source: EMC World Cellular Database, June 2001 Figures exclude Puerto Rico
In operating all three networks, Cingular is more experienced in offering more regional or areas of unconnected coverage. And has no doubt been evaluating its next step for some time. Like AT&T Wireless, it is committed to EDGE, seeing this as the point where its TDMA and GSM networks converge. EDGE itself, however, continues to be an uncertainty. If it does reach the market, the number of devices is likely to be minimal. If this is the case, many of the economies of scale (attraction for TDMA carriers to move to GSM) will be eroded.
80% of Cingular's current subscriber base are on local or regional plans, indicating that a continuous footprint is not the most important factor for a consumer.
Both AT&T and Cingular are advocates of GAIT, and both have indicated that it is part of their network plans. Cingular Wireless looks less dependent on GAIT, for a couple of reasons: the first is its spectrum situation. Unlike AT&T, it is not blessed with 1900MHz spectrum (a situation that may worsen following the NextWave decision), so if it were to opt to move its TDMA operations over to GSM, it would likely use its 800MHz spectrum, although the availability of dual-band 8/19 handsets would restrict this. By using the 800MHz spectrum, Cingular would have to remove TDMA-800 and AMPS users. This would serve to reduce its existing coverage, but at the same time expand its GSM coverage. Cingular is therefore more likely to be in need of dual-band GSM-8/19 equipment, rather than GAIT. Added to this, Cingular already has a comparatively large GSM-1900 footprint.
The routes chosen for AT&T and Cingular (if they were to choose GSM/GPRS/EDGE/W-CDMA) are very much governed by the availability of key products. For example if GAIT handsets were available now, new GSM-1900 coverage could be augmented by the TDMA network, but GAIT products are unlikely to be available until 2002, and even then choice is likely to be limited and high in cost. As mentioned previously, the future of EDGE is unclear with seemingly little progress made since AT&T introduced the possibility of its use in November 2000. Beyond GSM/GPRS deployment, W-CDMA equipment and handsets will be unique to the Americas, with no current UMTS or W-CDMA network builds using 800 or 1900 spectrum.
Conclusion
CDMA carriers appear to have the upperhand in moving to 3G systems, but at the same time they will be improving their core (service), rather than having to re-build it. How they optimise this advantage remains to be seen. With a potential rise in voice capacity they may well opt to increase the amount of free bundled minutes by an amount equivalent to increases gained in voice capacity through cdma2000 1x installation. This way voice competition is sharpened and operators do not necessarily have to look to attract customers through data offerings. But any reduction in voice rates or commensurate increase in bundled minutes needs to be well targeted, as, for the time being at least, voice remains king for US operators.
e-searchwireless.com |
