Commentary: Why 2.5G Plus WLAN Doesn't Equal 3G

Two plus two equals five.

That's the kind of cockeyed arithmetic that some industry figures are employing when they tell you that today's so-called 2.5G cellular networks, if combined with IEEE 802.11 wireless LAN networks, could provide a cheaper alternative to third-generation (3G) wireless networks in Europe (we'll get to the U.S. and Japan later). Yet it has become a popular mantra among the wireless chattering classes, many of whom now see 3G networks as little more than a huge cash drain and a crashing bore to boot.

Now, don't get us wrong -- we here at Unstrung are very excited about the many, many applications of wireless LAN technology. But talk of a combination of 2.5G technology and wireless LAN public access points supplanting 3G is just plain wrong, especially on the old continent.

Here are the reasons why:

Voice: Ah yes, the other wireless killer app. The real reason for rolling out 3G in Europe is because the operators need to ensure they have enough capacity to support more voice traffic. It has little to do with data or any of that fancy stuff. Their GSM networks are full, yet mobile usage is still growing, and carriers expect more of their fixed-line traffic to go mobile over the next few years.

"If voice is the main market for 3G, then neither 2.5G nor any combination of WLAN or Bluetooth remotely affect it," claims an interesting report issued last year by WestLB Panmure.

Now, when Unstrung was hobnobbing with the wireless cognoscenti in Cannes earlier this year, we heard a couple of operators complaining that they had been conned into this billion-dollar baby called 3G. They sniped that it was the fault of the government and conniving equipment operators, looking for a quick buck.

So why don't the carriers that have bought 3G spectrum just use it to provide more GSM services? Well, that brings us to ...

The regulatory stick: The "3G" licenses are actually UMTS licenses -- to use them, carriers must upgrade to UMTS (Universal Mobile Telecommunications Service). As it did with GSM, Europe has settled on UMTS as the single standard for next-gen services. When the carriers emptied out their piggy banks to pay for the licenses, they were also agreeing to start on the upgrade path to UMTS networks. This may not turn out to be such a bad thing in the end, as the European UMTS license has around 10 times the capacity of GSM, so it provides room to grow.

By contrast, if you look at putting together a voice and data system that combines GPRS networks with WLAN hotspots, you've got a ...

Technology nightmare: Support for voice calls isn't even part of the standard 802.11 specification yet! (See Symbol Talks Up Voice Over 802.11) When that technology is eventually ratified, handset vendors will have to make dual-mode WLAN/WAN handsets in order to enable callers to receive voice over WLAN calls. We're still waiting for the promised flood of inexpensive Bluetooth and GPRS handsets to hit the market (they are coming, right?). Can you imagine how long it's going to take to get dual-mode handsets on the market?

Bluetooth could actually prove to be a partial solution to the VOIP problem, as voice support is part of the short-range wireless connection's specification. However, the basic specification has a very short range, 30 feet compared to 802.211b's 100-foot throw.

Another big question with either WLAN or Bluetooth is how to get those essentially short-range standards to scale up to provide the coverage that 3G networks can. Pretty tricky, unless you're in a very densely populated city (like, say, Tokyo) or have an array of fancy antennas and signal boosters. Although there are companies working on these issues, it is difficult to see how they will be solved before the serious rollout of UMTS systems starts.

Then there is the issue of roaming on wireless LANs and billing for data downloaded. Even now that actual operators, rather than college students or coffee shops, are starting to run WLAN sites in Europe, there are still vast numbers of WLAN "islands." How can operators hope create a system that allows a subscriber to roam across disparate WLAN networks and get the information that will enable them to charge for a range of services?

Diff'rent strokes: In other countries, the story is different. In the U.S., there is a severe spectrum shortage. All the major carriers, aside from Sprint Corp.'s PCS division, say they don't have enough available bandwidth to offer 3G data services. If that situation doesn't change, then WLAN may prove to be a quick and dirty solution for operators working around the spectrum crunch.

In Japan, NTT DoCoMo is already building out WLAN hotspots in Toyko. NTT may be able to take advantage of its dominant position in the domestic market to solve roaming problems and seed the densely populated city with access points.

However, if carriers do start to roll out WLAN services, Unstrung will make one last bold prediction: You can kiss the idea of WLAN services being cheap or even free goodbye. Despite the fact that WLAN bandwidth is cheaper, carriers won't be keen to eat their cellular margins by offering an inexpensive alternative.

— Dan Jones, Senior Editor, Unstrung
http://www.unstrung.com