AT&T may be trying to block the merger of Sprint Nextel's and Clearwire's WiMAX activities, but it is also proceeding with its own, more limited, plans for the technology. The telco has been trialling WiMAX-class technologies in rural areas for a few years and now its CTO John Donovan says the system is "at the top of the list" as an alternative to copper.

Donovan said, in an interview with newspaper USA Today that the cost of copper roll-out was making it prohibitively expensive to build DSL for rural communities. This is the usual dilemma that has left the US' rural reaches sadly underserved by broadband and 3G - high cost of deployment, coupled with sparse and often low income populations, and falling broadband tariffs. This ROI-challenging combination could be addressed by broadband wireless, and specifically the standards-based economics of WiMAX, believes Donovan, echoing the view of WiMAX that has driven most of the world's actual deployments, away from the headlines about mobile broadband and 4G - that it is a natural leader for wireless DSL and fixed/nomadic access for underserved areas, and this market will be far easier to dominate than mobile applications.

When BellSouth was added to the AT&T super-telco in 2006, the company fought hard to retain the regional operator's spectrum holdings in 2.5GHz and 2.3GHz. In the end, the FCC forced divestment of the 2.5GHz assets (mainly bought by Clearwire) but allowed AT&T to keep BellSouth's more substantial 2.3GHz holdings. This band has some challenges in the US, despite being a standard WiMAX profile, because it is split into two sections and so does not support the bandwidth of 2.5GHz. However, BellSouth had been making use of the spectrum in some of its rural territories in the southwest, and launched fixed wireless services in mainly small markets in Kentucky, Georgia and eight other states using Navini gear. Under AT&T's wings, it also built out WiMAX in Alaska last year using Alvarion equipment.

Under the FCC ruling on the AT&T-BellSouth merger, the telco had to commit to build out the 2.3GHz spectrum with services to cover 25% of the POPs by July 2010. It also agreed to offer broadband services to 100% of potential users living within the AT&T/BellSouth territory, and that up to 15% of this territory would be provided with broadband connectivity using WiMAX or satellite. These rules, combined with the pressures on the DSL model, are now spurring AT&T into action.

It has been examining ways to optimize the efficiency of WiMAX to enhance its business model and the performance of its 2.3GHz spectrum, which lives in two bands divided by satellite frequencies. Last year it was putting pressure on the WiMAX Forum to add various features to its profiles to improve this performance, in particular by mandating a MIMO smart antenna array with four antennas at each end, rather than the 2x2 formation specified in the 'wave two' profile. These demands were likely geared to making the 2.3GHz spectrum more usable for the operator by supporting more powerful antenna arrays. Now Donovan says AT&T is also considering using femtocells to improve indoor penetration and cost efficiency of build-out.

While AT&T is firmly wedded to LTE for mobile broadband, it appears that WiMAX will play a prominent role in its fixed broadband and quad play strategy for rural areas, and this explains its hostility to the Clearwire venture plan, since despite all its talk of mobile broadband, many of the early target markets for Clearwire have focused on wireless DSL, a more immediate source of uptake and cashflow than advanced 4G-style applications that are currently supported by very few devices.

The other issue for AT&T is whether to expand WiMAX-based wireless DSL in rural territories where it does not hold 2.3GHz spectrum (a total of 28 states). Some of its 700MHz licenses might come into play, since these low frequencies are well adapted to rural coverage, but AT&T may run short of bandwidth, since it also wants to use 700MHz for early LTE deployment and possibly mobile TV.