Cooters, I think it's time for a little humor about GSM/1X. Here's a bit of a dated article the Q actually has in web page that cites Rod "Clueless" Nelson's take on it, which I've boldened:
telecomasia.net
Unnatural selection
Does Qualcomm's convergent GSM-1x technology stand a chance in the global wireless gene pool?
By Kirk Laughlin
"We gotta do something." With those words, Qualcomm CEO Irwin Jacobs kick-started an initiative in November to do what used to seem like a networking impossibility - wed CDMA2000-1x radio with GSM core infrastructure - to produce the still-in-concept solution called GSM-1x.
Jacobs was driven to address the "marketplace of tying GSM and [CDMA]1x together," says Sanjay Jha, senior vice president of engineering at Qualcomm, who recalls the conversation he had with Jacobs.
Jacobs's motive seemed clear: GSM/GPRS wireless operators, especially the spectrally deprived, would have a new voice and data migration alternative. Operators would make faster strides to true 3G by following along to CDMA2000-EVDO, and, at least in theory, discover economic and integration advantages over the road they could have traveled.
By the time the CTIA Wireless 2002 show rolled around in March, Qualcomm had pulled together a handful of infrastructure suppliers, switch makers and handset players to portray GSM-1x as, according to a press release at the time, a "second natural option" for GSM operators to migrate to 3G.
So what's so natural about GSM-1x?
"If there is a genius in this solution its in enabling CDMA2000 to mesh with the GSM network without changing either the GSM core network or the CDMA physical layer," says Jha, whose contends that GSM-1x is a "valid alternative" to UMTS at 2.1 GHz. The solution is also aimed at producing greater spectral efficiencies for operators using 800 MHz, 900 MHz, 1800 MHz, and 1900 MHz.
Jha contends GSM-1x is a plausible alternative because of its simplicity. Operators follow two basic steps - modify the MSN (mobile switching node) to accommodate the new radio-core interface, and roll-out handsets that support recently developed Removable User Identity Module (R-UIM) cards, similar to SIM cards used on GSM networks, enabling roaming across CDMA and GSM networks.
Risky play
GSM-1x may have Qualcomm's full backing, but the solution does stand apart from other promotional campaigns spun out of San Diego. For one, the chipmaker recognized from the start that its initiative could easily become an also-ran, falling into the wrong time-wrong place category. With most of the world's GSM/GPRS operators solidly focused on migration to EDGE and UMTS, Qualcomm admits this is a risky play that may ultimately only produce business in a few, far off Southeast Asian markets, and possibly in China and Australia where select carriers use both GSM and CDMA.
The Jacobs proposal, at least initially, has been puzzling to many in the mobile industry. The timing has clearly not been ideal. Major US carriers AT&T Wireless and Cingular Wireless may have been strong candidates for GSM-1x, about 18 months ago, before both carriers crafted strategies to overlay their networks with GSM/GPRS and later, EDGE.
Some believe Jacobs may have observed signs of sluggishness in the GSM data ramp-up strategy. Perhaps he might have anticipated that if the EDGE data standard continued to lose legitimacy, as seemed the case late last year, GSM-1x would appropriately save the day. But, the most likely cause was to make GSM-1x a convenient technology option for operators who may have entertained thoughts of deserting the GSM battalion. Although worldwide GSM subscribers outnumber CDMA subscribers by a factor of roughly four-to-one, CDMA is ascending faster. According to research firm Strategy Analytics, the global Compound Annual Growth Rate (CAGR) for CDMA is 28%, compared to just above 20% for GSM.
In response to the timing issue, Jha said only recently has a hybrid solution become economically viable. The new R-UIM cards enable roaming across competing network footprints, allowing subscribers to use the cards in either CDMA or GSM phones. Also, new switching platforms include open interface flexibility supporting both GSM-MAP and ANSI-41 protocols. Next generation switch supplier Winphoria, for example, is pushing its WMS-5000 mobile switching center (MSC), which offers service transparency between GSM-1x, CDMA2000 and GSM networks.
"Clearly the GSM-1x effort is to get some additional CDMA penetration in some of the newer, evolving markets using a more spectrally efficient protocol," says Michael Champa, the CEO and president of Winphoria.
In addition to Winphoria, Qualcomm has leaned on familiar mobile hardware suppliers like Nortel, Samsung and Kyocera to get behind the strategy, at least by participating in a demo at CTIA 2002. Some of them like Nortel emphasize that GSM-1x is still in the conceptual stage.
Despite the appearance of forward momentum and solidarity, skeptics from the GSM community are unmoved. iAT&T Wireless CTO Rod Nelson, for example, wonders if the initiative is a sign of desperation. "To put this heterogeneous air interface technology together doesn't make any sense at all when I can just stick with the GSM evolution and achieve the same results in terms of voice efficiency and data speeds," Nelson says. "I think it's a desperate attempt to try to break into the GSM marketplace."
Although Qualcomm has hopes of setting up a trial with a US operator by year-end, commercial implementation of GSM-1x in North America seems about as likely as Bill Clinton making another White House run.
Nelson sees no chance of GSM-1x getting a trial at AT&T Wireless, particularly because a 1.25MHz channel would need to be cleared to accommodate GSM-1x, he says. "The cost of clearing the spectrum alone is significant and a complex undertaking," he says. "That to me is a tremendous flax in the logic."
Nelson is also quick to defend EDGE, the 3G data-only standard for GSM/GPRS operators with real-life throughput speeds of 75 to 150Kpbs, roughly equivalent to that of CDMA-1x. Qualcomm has been critical of EDGE, saying it offers operators no voice advantages.
"You can make the argument they're making that EDGE doesn't do anything for voice capacity, but you're not really telling the whole story - it's not intended to," says Nelson. "We have AMR [Adaptive Multirate Codec] technology to do that."
Voice capacity increase
AMR is a vocoder technology that adapts data rates to maximize voice transmission on GSM networks, potentially bringing a fourfold increase in voice capacity. AT&T Wireless is currently installing AMR technology at its base stations and in its handsets.
"EDGE plus AMR is exactly what 1x is," says Nelson.
Despite their zeal, Qualcomm is not going to give up on the US carriers. Jha makes a familiar argument that US carriers may have trouble clearing a big chunk of spectrum - 5 MHz - to effectively run UMTS. "They are on this roadmap, and time will tell as to how effective that roadmap is relative to some of their competition," says Jha.
Prospects for GSM-1x in Europe seem equally bleak. Migration strategies aside, operators maintain an almost jingoistic allegiance to GSM and a mutiny appears highly unlikely. "Operators there would be rowing upstream in the tumultuous waters of government mandates that specify the GSM/GPRS/UMTS path quite clearly," says Cliff Raskind, the director of Global Wireless Practice at Strategy Analytics. He stresses that it is implausible for a Western European operator to follow the CDMA2000-EVDO route for a host of "politically charged reasons".
"From a roaming perspective, any GSM operator in Western Europe that implemented GSM-1x would be an island unto itself in a sea of GSM operators, until the right multimode chips from Qualcomm became ubiquitous in Western European handsets," says Raskind.
The outlook in Asia, however, looks comparatively sparkling. It is widely known that CDMA has flourished in South Korea, but Seoul is not the terminus for Qualcomm's Asia foray. Equity research firm Lehman Brothers estimates 28 million new CDMA subscribers in Asia during 2002, built on big deals expected to go down with Reliance Telecom in India, Indonesia Telekom and possibly an operator in Thailand.
The existence of commercially successful CDMA networks in Asia could bolster the case for bridging to GSM-1x. Countries with a patchwork of GSM and CDMA networks may be swayed to seek a solution that co-joins networks for both voice and data.
Jha argues it makes little sense to have separate support systems for two networks in one region, suggesting dual standards operators Telstra in Australia, China Unicom and Reliance would all be suitable candidates for GSM-1x.
"Anyone who runs a network has to run two networks, two provisioning systems and two billing systems," Jha says. "It's just a cumbersome thing and being able to combine the two is very attractive."
Likely the biggest catch for GSM-1x would be China Unicom, the fast emerging Central Asia carrier that had 27 million GSM subscribers at the end of 2001, and had hoped for 8 million new CDMA subscribers by the close of this year. Unfortunately for Qualcomm, things haven't worked out that way. Analysts believe Unicom may only attract about 2.4 million CDMA subscribers due to handset shortages and higher prices compared to GSM devices.
Lehman Brothers analyst Tim Luke, who has a "strong buy" rating on Qualcomm, said in a late April recent note that China Unicom's CDMA subscribership lagged because the carrier did not aggressively market the new service. He believes CDMA uptake will increase, once Unicom gets its strategy in order. "We believe that subscriber data from China Unicom may be somewhat volatile as Unicom's network and marketing strategy matures," he wrote.
Southeast Asia targets
Jha would not specify which Southeast Asia carriers are being targeted for GSM-1x. Chances are those with limited spectrum, perhaps lacking 900 MHz licenses, but possession 800 MHz bands, would be most likely to consider the solution. "There are lots of operators, particularly in Southeast Asia, with 800 MHz in which case you have to make very little change, because CDMA vendors already provide 800 MHz support in their equipment," says Jha.
The rocky start for CDMA at Unicom may not bode well for GSM-1x in China, but Qualcomm sees CDMA-1x handset supply eventually working in its favor worldwide.
The cost of handsets is perhaps one of the more combustible controversies in the market share battle between GSM and CDMA. Jha argues that CDMA-1x has a critical headstart over the successors to the GSM-voice handset. "Certainly it is true for GSM-voice only that their volume ramp has happened. They were two years ahead of us," says Jha. "But, for data I believe CDMA is ahead of GSM in terms of delivering volume ramp."
A list of objections to adoption of GSM-1x would not be complete without consideration of commitments by suppliers and the potentially destabilizing realignment of relationships across rival standards camps. The rise of GSM-1x could put traditional CDMA suppliers in an awkward position because, as Jha says, "GSM-1x attempts to replace CDMA vendors with GSM vendors - so it's not always the case that we will have receptive minds to this solution."
Nortel Networks, which sell equipment to both sides in the networking holy war, sounds a note of caution about GSM-1x. "We are really in the evaluation stage," says Mark Morell, Nortel's director of strategic marketing for wireless networks. "We're looking at it as a potential technology for certain operators who have specific needs."
Nortel could not offer any data on the cost per POP for GSM-1x.
Whether Nortel continues to develop infrastructure that would support GSM-1x will partly hinge on Qualcomm's success in securing a trial in Asia, says Morell. "[A trial] obviously influences our decisions."
Another major uncertainty is the willingness of large European mobile equipment manufacturers, like Nokia and Ericsson, to endorse GSM-1x. "By the end of the year, I expect to get a lot more handset vendors, the question is will we get European manufacturers to sign up. I think that will be the issue," says Jha.
Without a solidly aligned group of equipment makers, the viability of GSM-1x would obviously come under question. Jacobs himself may have to decide how much political effort should be placed on technology that leaves many in the GSM community scratching their heads.
But, at least for Jacobs, the initiative was appropriately timed - opportunities to win over GSM operators are quickly diminishing. Even incremental expansion of CDMA's reach will improve on efficiencies of scale. With forecasts of as many as 330 million EDGE subscribers worldwide within five years, campaigning for CDMA should not seem all that surprising. GSM-1x, from that perspective, will likely not be Qualcomm's final attempt to bring CDMA radio to the front yard of the GSM neighborhood. |
