Clearwire\'s shift to LTE is not just a move away from WiMAX, but it cements Clearwire’s shift in strategy from being a retail operator to a wholesale provider — a shift that has been coming for a while. In this week’s announcement Clearwire said that it would sell its LTE network not just to partners such as Sprint, but also to anyone that wants such service.

This isn’t a new field for Clearwire, which has sold wholesale WiMAX access to the cable providers that funded it and to retailers such as Best Buy, but as it moves to LTE and closes down retail locations this wholesale strategy is becoming more clear. And it’s also setting Clearwire up as an odd operator out in the industry as many other cellular companies try to sell services as opposed to pure capacity. However, in Clearwire CTO John Saw’s view "LTE is all about capacity."

Building the LTE wholesale network.

During the call to discuss the transition to LTE, Clearwire emphasized that it would employ a version of LTE known as LTE-Advanced. This refers to releases 10 and 11 of the LTE technology, but Release 10 was just frozen in March and the gear isn’t out yet for it. What Clearwire hopes to deploy if it gets $600 million or so from investors, is LTE Release version 8, and will be upgradable to later releases that are actually LTE-Advanced. The key elements for Clearwire, and the reason it’s using the LTE-Advanced lingo is because it’s using some features built into the version of the standard that are associated with LTE-Advanced. Plus, since LTE-Advanced was the “real 4G” according to the ITU, I’m sure there will be some marketing spin on this later.

Saw says that Clearwire plans to take advantage of features that allow an operator to group different spectrum bands together to create a virtual pipe as well as features known as MIMO that allow multiple antennas on the device and base stations to boost upload speeds. Those two things are the most important reasons Clearwire has switched to LTE.

For the deeply nerdy, it’s using TDD-LTE (GigaOM Pro sub req’d), a different variation from Verizon and AT&T, which are deploying FDD-LTE. The difference is that Verizon and AT&T must deploy their spectrum in equal clumps going upstream and downstream, but Clearwire (and anyone using TDD-LTE) can allocate their spectrum unevenly, with a greater proportion going to downstream and less for upstream use.

Upgrading the existing infrastructure will be easier in cities where Clearwire has recently deployed WiMAX said Saw: "Adding LTE to those markets is as simple as plugging in another line card on the cell site." In some areas Clearwire may have to install new radios and in general it will upgrade the core network and backhaul networks, all for that estimated $600 million mentioned.

What happens to WiMAX?

But amid the Clearwire move to LTE, what happens to the existing WiMAX network? Saw says it will remain intact. The company will use 20 MHz of spectrum for LTE and reserve 10 MHz for WiMAx and operate both networks side-by-side. In most cases the equipment is designed to do just that.

It all depends on devices.

By managing multiple networks, even if it can reuse some of the same equipment, Clearwire avoids the challenge of clearing spectrum and getting people to transition to new devices. But devices will still be a key element in Clearwire’s success with LTE. Because it plans to offer wholesale access, Clearwire will have to rely on device makers to put radios into their products that are TDD-LTE compliant and that work in the 2.3 to 2.7 GHz spectrum band that Clearwire is using. Wireless radios, with their associated IP aren’t cheap, so the key is getting them both inexpensive, but also small enough and power efficient enough that a tablet using Clearwire’s LTE won’t cost a lot more and will still have decent battery life.

Clearwire has teamed up with China Mobile and Vodafone to promote a world band in its spectrum for TDD-LTE as part of the Global TD-LTE Initiative. Saw claims that the members of the GTI represent hundreds of millions of potential subscribers and members are deploying networks this year. He didn’t provide details but said Qualcomm, Broadcom and others are planning chips for the band. When I asked if the chips were sample and if we could expect a 12-18 month time frame before such devices hit the market, he said pre-commercial devices are already available. That’s not a clear answer so figuring out when devices that could use the network will arrive is still an open question. Qualcomm recently made its own spectrum play in India in the 2.3 GHz band suggesting it does have plans to support it with radios.

Clearwire\'s gamble may all depend on cheap chips

Clearing up Clearwire’s business model.

With its plans to deploy LTE only in areas with high demand, Saw explains that Clearwire’s business model will be built around providing capacity offload. This is something other carriers are doing with Wi-Fi today, but having a more mobile option clearly has benefits in areas where one can’t find a hot spot. By offering LTE it would compete against the planned wholesale LTE network from LightSquared, which wants to use a mix of satellite and terrestrial capacity from Sprint to offer service.

Clearwire is up against LightSquared\'s planned satellite network.

Saw bristled when compared with LightSquared pointing out that the company is not only in a fight with the GPS industry over interference, but it also doesn’t have a network. “It’s hard to speculate about LightSquared. It has zero spectrum and no network, and even if they, by some miracle, get their lower 10 MHz approved, that’s a very thin network compared to having 160 MHz in the top 100 markets,” Saw said.

And once again we’re back to capacity, which is what Clearwire has to hope matters. Even with its WiMAX network, Clearwire was hitting the capacity angle hard, because it had those vast spectrum reserves. And while technologists will argue about the poorer spectrum propogation characteristics of the 2.3 and 2.7 bands (they don’t go through buildings easily), that’s the cards Clearwire was dealt (or bought, cheap at auction actually). So the question for Clearwire investors and partners becomes whether or not Clearwire can build out an LTE network designed to offer mobile broadband in areas where carriers and other providers need service and whether that business is big enough to support the costs of building and running such a network.