Lucius,
<< This exchange really should be on the QCOM or gorilla thread... >>
I think this is very appropriate discussion for this thread as well as the others you reference. Below is an excerpt from a CMP TechWeb article I previously referenced here:
> What no one foresaw, however, was the explosive growth of competing transports. Today's cellular voice and data nets are based on at least eight. To add to the confusion, some transports are known by different names. For example, D-AMPS—one of the chief digital mobile standards in the U.S.—is also called ANSI-136 or IS-41. And with at least a dozen transports slated to be deployed over the next 10 years, things aren't going to get any simpler soon (see Table 2 ).
Why so many? Blame it on the FCC. Instead of going with a single global standard like the rest of the world, the U.S. allows wireless operators to install whatever they want. The result: a broad palette of incompatible services.
Some carriers and vendors insist that unfettered competition is a good thing. "This is what's made America great," says Robert Sellinger, director of wireless architecture for Lucent Technologies Inc. (Murray Hill, N.J.).
It's hard to see how. Since services are based on different technologies, diversity has actually stalled competition and kept prices higher than in other parts of the world. Coverage is patchy. Cell phones haven't taken off nearly as fast in the U.S. as in Europe and Japan. And for once American vendors lag their European counterparts: Nokia and Ericsson have emerged as the leading suppliers of wireless network equipment and phones.
UNCLE SAM STUMBLES
U.S. companies stand to lose even more as mobile IP kicks in overseas, where there's a huge installed base of standardized mobile phone networks just waiting to be upgraded.
That leads to an obvious question. Which new bearer transport is likely to be tomorrow's top choice? Finding an answer means understanding that today's circuit-switched mobile networks can be roughly divided into two categories: those that use TDMA (time-division multiple access) to send data between a mobile device and base station and those that use CDMA (code-division multiple access). The next thing to get a fix on is the size of the customer base for the biggest mobile nets. On the TDMA side of the ledger are GSM (global system for mobile communications), with some 150 million subscribers worldwide; Japan's PDC (personal digital cellular), with 45 million; and the U.S.'s D-AMPS (digital advanced mobile phone system), with 15 million. On the CDMA side there's IS-95a (interim standard 95a, also known as CDMAone), with some 30 million users in the U.S. and Asia.
Do the math and TDMA transports wind up with 210 million subscribers to CDMA's 30 million. Since GPRS is the packet-switched upgrade for TDMA, those numbers put it over the top with room to spare. What's more, the sheer size of its potential customer base translates into lower equipment costs for carriers and blanket coverage for customers. Of course, not all TDMA users will get the GPRS upgrade, particularly in the U.S., which looks likely to deploy alternatives rather than roll out one transport.
CDMA upgrades also are in the works, like CDMA2000. And a handful of GSM operators are pushing ahead with an earlier upgrade—HSCSD (high speed circuit switched data). But all this scheme does is bundle four cellular connections together, wasting a lot of capacity to deliver 28.8 kbit/s. "The market window for HSCSD is too short," says Mikko J. Salminen, technical director of Radiolinja Oy (Helsinki), Finland's second-largest mobile operator.
Ultimately, plans call for all TDMA and CDMA upgrades to converge on a single global standard, IMT-2000 (international mobile telecommunications) from the ITU (International Telecommunications Union). The 2000 refers not to the target deployment date but to the 2 Mbit/s that's the target throughput. IMT-2000 is actually three standards in one: CDMA2000, UMTS (universal mobile telecom system), and UWC-136 (universal wireless communications).
The idea is that TDMA operators will upgrade their nets with GPRS and then upgrade again to UMTS as bandwidth requirements increase. Just to complicate things further, there's an intermediate standard between GPRS and UMTS called Edge (enhanced data rates for global evolution) that's supposed to boost bandwidth to 384 kbit/s.
Much of this muddle will be masked from end-users, since coming wireless devices will likely support multiple bearer standards. What's more, they won't have to upgrade all their gear to get the higher speeds. Many handhelds and laptops already come equipped with a new radio technology called Bluetooth, which enables communications at up to 4 Mbit/s with equipment located very close by. A GPRS connection set up by a cell phone, for instance, could be automatically linked to a PDA or laptop thanks to Bluetooth.
Despite the acronym overload, one thing about mobile IP should be very familiar: the difference between theoretical throughput and the real thing (see Table 3). GPRS, for example, has a theoretical maximum throughput of 115 kbit/s, but customers are more likely to see 22 kbit/s upstream, 44 kbit/s down. This isn't just the usual hype. For one thing, wireless operators want to see what customers want before they overspend on radio capacity, so they're bumping up bandwidth incrementally. For another, vendors are having trouble getting their handhelds to handle the higher speeds. It all comes down to power amplifiers and heat dissipation. "Up to 22 kbit/s upstream is perfectly OK with current technology," says Wavecom's Guillemette. Higher bandwidths need bigger power amplifiers to dissipate the heat, and they can't be crammed into a cell phone or PDA. Nokia, for example, refuses to comment on whether it's experiencing problems with its GPRS terminal and declines to give delivery dates.
Given all these uncertainties, is it possible to predict which wireless services are likely to take off and which will fall by the wayside? To get beyond guesswork, Data Communications surveyed a dozen leading equipment vendors, service providers, and industry organizations. They were asked to assess the current availability of wireless IP services in the U.S. and worldwide and to forecast when and where next-generation technologies will take off over the next two years (see "Mapping the Wireless Future").
The key findings? Europe and some parts of Asia-Pacific are already out in front of the U.S. when it comes to wireless IP services running at 20 kbit/s and above. Only one high-speed offering, Ricochet, is available in three states in the U.S. Mean- while, 19 countries are about to start GPRS trials; 4 (Finland, Hong Kong, Norway, Singapore) also have some HSCSD. The survey also confirms that GPRS is set to become the global standard. Most countries in Europe and Asia-Pacific and nine cities or states in the U.S. will deploy it within the next 12 months. <
- Eric - |
