• MOST READ

• LATEST NEWS

As Verizon released videos as the first step in its marketing run-up to LTE services, AT&T did a U-turn on its own next generation strategy, and now plans to implement HSPA+. Previously, the telco had planned to upgrade its nationwide W-CDMA footprint to the 7.2Mbps version of HSPA, and then skip straight to LTE for mobile broadband from 2011. Now it will deploy 14.4Mbps HSPA+ too.

The change of heart is welcome. AT&T's strategy had seemed highly illogical. In the wake of the huge strain on its network from mobile data services, only moving to 7Mbps HSPA - which most European carriers are already leaving behind in favor of 28Mbps or even 42Mbps implementations - seemed inadequate. Having a relatively slow network would also put huge pressure on AT&T to build out LTE very quickly, in any areas where it saw high levels of mobile broadband usage. Although the LTE build it still scheduled to start in 2011, a faster national back-up network would allow AT&T to be more flexible in how intensive it wants its 4G capex hit to be. By stopping at 14.4Mbps iteration (the figures refer to peak download speeds), AT&T does avoid the additional cost and complexity of supporting MIMO antenna arrays, which are necessary for the 21Mbps and higher versions.

John Stankey, CEO of AT&T Operations, outlined the new plan in interviews with the GigaOm and Engadget blogs. He said it would cost AT&T less than $10m to upgrade to HSPA+, with coverage of 250m pops, by the end of this year.

In adopting a middle stage between HSPA and LTE, AT&T is not only speeding up its networks in the interim period before it has widespread 4G, but keeping pace with T-Mobile, which has an aggressive strategy to move to HSPA+. By contrast, Verizon Wireless denied recent reports that it would also have a mid-speed strategy, deploying EV-DO Rev B to upgrade its current network and fill in short term or permanent gaps in LTE coverage. Rev B can achieve peaks around 14Mbps. Sprint was also said to be looking at Rev B to enhance its 3G network outside the areas where it provides 4G services via its Clearwire WiMAX venture, though it has not commented on these reports.

Another reason for AT&T to invest in HSPA+ is that it will have a better device ecosystem than LTE for many years to come, especially on the handset side. Stankey has always said LTE would not be ready for the mainstream until 2014, largely because the early handsets would be unappealing, as the 3G ones were. He told GigaOm: "The vendors are experiencing some challenges on certain features and software, and first implementations in 2011 will be…pretty vanilla." It will take several years, he thinks, fully to address issues such as effective 3G/LTE roaming, battery life and voice. Although HSPA+ requires new devices too, these are less of a step away from current phones, even though early HSPA+ services round the world have mainly been based on data cards and dongles.

HSPA+ can reach 28Mbps using MIMO only but the addition of multicarrier has got it to 42Mbps already. Huawei has demonstrated 56Mbps systems and Ericsson has now shown off 84Mbps, and says the technology can be stretched further over time, to beyond 100Mbps.

Related Stories

COMMENTS

Posted by maxwell on Tuesday 18th May, 2010

Actually, the 21 Mbps download/forward link does not require MIMO but just 64QAM. This is already supported in the hardware that most operators have out in the field today.

So either AT&T still have a pain with their backhaul or some of their suppliers (networks or devices) have problems delievring.

Carrier Broadband Performance: Africa & Middle East
Carriers are using mobile broadband as their weapon of choice in the fight against the commoditization of voice and falling ARPU. This...

Voice over LTE: Market Analysis and Forecasts
VoLTE offer unique advantages in lowering a carrier's voice infrastructure economics and promises to improve voice quality, device...