The looming battle over wireless broadband

The war over which standard will become the dominant architecture for broadband wireless infrastructure seems to have ended — but now there’s a new fight.

Of course I am talking about LTE (Long Term Evolution), a 4G (fourth-generation) technology that comes out of the cellular telecommunications world versus WiMax, which has its roots in high tech.

Most carriers are settling on LTE

Following years of GSM versus CDMA and the divergent evolutionary path followed by wireless carriers, LTE is the technology upon which most carriers appear to be converging.

LTE’s first iteration is promised to deliver downlink speeds of anywhere from 3Mbps to a peak of 10Mbps, and uplink speeds of 1.5Mbps to 3Mbps. But of course that depends on the amount of bandwidth a carrier allocates to the service. It can go much higher.

LTE also gives carriers a chance to settle on a single technology worldwide as they move from the multiple 3G technologies to LTE as the replacement 4G technology. With LTE, for example, companies such as Verizon in the U.S. and its part owner Vodafone in Europe could easily offer a single phone with a single technology that will work just about everywhere, unifying the networks at both the radio and core network layers. Today, Verizon uses CDMA and Vodafone uses GSM, so to continue with Verizon as my example, Verizon could only support wireless roaming onto other networks outside the U.S. using a dual-mode CDMA/GSM phone. That limited support would no longer be necessary after it moves to LTE.

It’s true that a few carriers — notably Sprint PCS — have decided to stick with CDMA2000, a 3G cellular technology. But the adoption numbers for CDMA2000 are not in its favor: A recent Gartner report estimates that there are 2.1 billion GSM connections worldwide (or 79 percent of the market) versus 325 million CDMA2000 connections (12 percent of the market).

So LTE is clearly the future for cellular networks.

Or is it?

WiMax may beat LTE to the punch

Although the carriers are largely converging on LTE as their next-gen cellular technology, a new challenge is emerging from an entirely different camp: Wi-Fi, which Intel and computer equipment makers have been pushing. Although WiMax has been promoted for years, its actual deployment finally seems imminent.

The battle between LTE and WiMax reminds me of the current presidential Democratic primary. Only a few short months ago, New York Sen. Hillary Clinton was the expected Democratic presidential candidate. Then the junior senator from Illinois, Barack Obama, took the mantle from her almost overnight. Obama is to WiMax as Clinton is to LTE.

For companies that want high-speed broadband mobile access now, advantage WiMax. This spring trials will begin in Baltimore, Chicago, and Washington, D.C. There are 260 carriers deploying WiMax in 110 countries. Add to this the fact that Intel is incorporating WiMax into its CPU chipsets, which will be available this spring.

In the meantime, LTE is not expected for another two years.

Out of the box, the performance advantage appears to go to WiMax as well. If a carrier uses a 5MHz bandwidth channel, WiMax’s performance is about 10Mbps to 15Mbps, which can be divided in any way the carrier decides for uplink and downlink. With a 10MHz channel you are talking about 20Mbps to 30Mbps.

That brings us to Sprint, which has not signed up for LTE but instead announced more than a year ago that it would deploy a near-national WiMax network. Sprint owns a wide channel of spectrum, between 100MHz and 190MHz of bandwidth, depending on the area. With this bandwidth, Sprint could deliver amazing performance as the updates to the IEEE 802.16e and its follow on 802.16m standard behind WiMax increases how much that bandwidth can be used.

Next year, we may see ratification of the 802.16m standard that would let WiMax use even wider bandwidths and increase performance up to 100Mbps to 200Mbps.

To be fair, there is still some question as to whether or not 802.16m will be backward-compatible with 802.16e. Such compatibility issues could slow down WiMax’s deployments as Sprint waits, or cause Sprint to forgo 802.16m for its next round of capital investments. But the WiMax Forum, an industry association that imposes its own interoperability standards on members beyond the IEEE’s looser requirements, is guaranteeing backward compatibility, notes Mohamed Shakouri, a vice president of the WiMax Forum and vice president of strategy at Alvarion, a mobile equipment manufacturer.

Of course, commercial deployments of 802.16m is about two years away — just like LTE.

End the battle before it begins? Probably not

At the Mobile World Congress in Barcelona, Spain, Vodaphone CEO Arun Sarin called on the standards groups involved with LTE and WiMax to merge. Both technologies do have a lot in common. Both are IP-based and both to some extent use OFDM radio technology, for example.

However, from my talks with folks at the WiMax Forum, I don’t think that will happen any time soon. Using 802.16e, they have about a two-year head start on LTE, and my guess is they are not inclined to relinquish that competitive advantage to merge technologies.

The fight for customers is at the heart of this battle. Manufacturers of all stripes realize that users want the same performance for their applications and services when mobile as when they are tethered to a desk. And they all want to be the one to provide it, which usually means being first. There are not too many second chances in today’s market.

I’ll never forget that when I was in high school, my social studies teacher told the class that whichever country wins the race to the moon will win the Cold War. “The world of undecided nations will be so awed by such an achievement they will want to emulate the winner, whoever it is,” she claimed.

There’s no denying that she called that one pretty well.

And so the fight over standards is a race to give users the same level of performance when mobile as they have when tethered to the desk. Whoever can do that first will win the hearts and minds of customers.