A couple of months ago, David Neale was engaged in a discussion concerning the nature of wireless technology, particularly the true definition of “3G,” the oft-touted third-generation wireless network that promises data connectivity as quick as you’d expect from a desktop computer.
Neale, vice-president, new product development with Rogers AT&T Wireless in Toronto, explained to his verbal sparring partner that the type of wireless network used by Aliant Mobility, Bell Mobility and Telus Mobility does not offer 3G service. Oh sure, the operators would say it’s 3G, but in practice it isn’t, Neale explained.
For one thing, the network technology those companies use – a new form of code division multiple access infrastructure called CDMA2000 1X – doesn’t move data fast enough to count as 3G. So far as Neale knew, 3G speeds were supposed to start at approximately 150Kbps. Bell Mobility’s CDMA2000 1X network offers a top speed of just 86Kbps.
By contrast, Rogers’ own network, a general packet radio service-enhanced global system for mobile communication (GPRS/GSM) infrastructure, offers a top speed of 115Kbps, Neale said. And it counts only as 2.5G, one step below 3G.
How could CDMA2000 1X count as 3G with such slow delivery? It can’t, he argued.
Or can it?
The CDMA Development Group (CDG) recently put out a press release claiming CDMA2000 1X networks comprise 99 per cent of the world’s 3G infrastructure. According to the CDG, CDMA2000 1X certainly does count as 3G.
“We defined an approach that’s relatively pragmatic for 3G technology,” said Perry LaForge, the CDG’s executive director in Costa Mesa, Calif., explaining CDMA2000 1X’s apparent popularity. “It didn’t require a lot of spectrum…and it leveraged a lot of the learning people got from CDMA.”
The CDG isn’t alone in suggesting that CDMA2000 1X is 3G. The International Telecommunications Union (ITU) backs that assertion, saying CDMA2000 1X is one of a number of wireless technologies included in its list of 3G infrastructure.
CDMA2000 1X networks compete primarily with wideband-CDMA (WCDMA) for 3G dominance, but WCDMA, the eventual 3G endpoint for GPRS/GSM evolution, is a rare bird at this point in time. Hence CDMA2000 1X’s 3G supremacy.
But Neale, who knows plenty about wireless technology, seemed surprised that the ITU defined CMDA2000 1X as 3G.
“I’m a little intrigued,” he said.
And he’s probably not alone. The notion of 3G might be easy enough to grasp from the user experience – it spells quick downloads and fat applications, a miniaturized World Wide Web in the palm of your hand, no wires attached. But from the perspective of actual download speeds and infrastructure specs, 3G is confusing, a clouded vision that deserves to be made clear.
First off, it’s important to note that 3G has less to do with numbers and more to do with functionality, said Gary Fowlie, the ITU’s public information services officer in Switzerland.
“Focusing on data rate alone is not per se very meaningful,” he said, adding that it takes more than quick downloads for a technology to join the 3G club. The technology must also provide both circuit-switched and packet-switched services, such as IP traffic and real-time video. It must offer seamless coverage across various infrastructure iterations. And it must offer a smooth migration between 2G service and 3G service (read: no major changes required from users).
But judging by Neale’s description of GPRS/GSM functionality, the network that Rogers uses should count as 3G if CDMA2000 1X does. GPRS/GSM offers seamless coverage and provides a tidy evolution path from 2G service to 3G, just as the ITU requests.
So why is GPRS/GSM not on the 3G list? Fowlie’s response:
“GSM is a second-generation system. GPRS is an extension of GSM to provide higher bit-rates. It is seen as an interim step to provide better data transmission capability until 3G systems are introduced. Hence, it is called 2.5G.”
Perhaps a glimpse at the speedometer would help clarify things. As Neale said, Rogers’ GPRS/GSM network tops out at 115Kbps. Bell Mobility’s CDMA networks offer an upper limit of 86Kbps, said Brian O’Shaughnessy, Bell Mobility’s vice-president, wireless technology.
But let’s compare apples and oranges, he suggested. Rogers’ network may offer bursts of 115Kbps, but it cruises at approximately 20Kbps, O’Shaughnessy said. Compare 86Kbps to 20Kbps and CDMA2000 1X comes out as the faster network.
In fact, Bell Mobility’s network is designed to top out at 86Kbps for the moment. As applications require more and more bandwidth, Bell will let the network run at its peak 150Kbps, O’Shaughnessy said.
“We wanted to under-promise and over-deliver,” he said.
Neale, on the other hand, said O’Shaughnessy is lowballing the GPRS/GSM cruising speed, which is closer to 40Kbps that 20 Kbps – plenty fast enough for most fat wireless applications, he added.
Not to mention that industry analysts who tested Bell Mobility’s CDMA2000 1X network said they experienced cruising speeds of approximately 60Kbps. Therefore the gap between CDMA2000 1X (60Kbps) and GPRS/GSM (40Kbps) is actually quite small.
Consider as well penetration rates when comparing wireless dominance. The CDG says 99 per cent of the 3G market belongs to CDMA2000 1X operators. But nearly 70 per cent of the world’s wireless networks employ GSM technology, Neale said. If CDMA2000 1X were compared with GPRS/GSM, it would appear to be the also-ran, rather than the top dog the CDG says it is, he pointed out.
Still, Neale conceded that CDMA2000 1X carriers could own 99 per cent of the 3G market, given a big caveat.
“If – if – 1X is 3G, that claim is not unreasonable,” he said.
O’Shaughnessy said Neale’s skepticism comes as no surprise.
“Anyone that hasn’t deployed [CDMA2000 1X] will say it isn’t [3G]… Our competitors say it as often as they can.”
O’Shaughnessy and Neale did agree on one point: that the user’s experience should come first when assessing various wireless technologies. Even though GPRS/GSM operators battle with CDMA2000 1X carriers for market share, “they all seem to do the same thing,” Neale said, adding that the speed difference is indiscernible from the user’s point of view. “If that’s the case, one can argue that this is a lot of hyperbole.”