After decades of dealing with a hodge-podge of air interface technologies, the mobile industry is slowly but surely converging on LTE. But even widespread use of a supposedly common technology will not unify the industry because LTE is being deployed on such a multitude of unrelated spectrum bands that if all the LTE frequencies were dumped into a bucket they'd resemble a nest of slithering snakes.
I recently spoke with Peter Carson, senior director of product management at Qualcomm (NASDAQ:QCOM), who cautioned that prospects for significant economies of scale from volume production simply don't exist for LTE and won't for a very long time because too many different spectrum bands are being used for LTE in too many different places. "Right now the biggest issue the industry has to deal with is complexity of the RF front end," Carson said.
Further, he noted that a single LTE product design that could be shipped anywhere in the world is far from fruition. An LTE world phone is just a fantasy at this point.
Frequency bands ranging from 700 MHz to 2.6 GHz have been earmarked for LTE networks around the world. The GSMA's Wireless Intelligence service released a report in December 2011 that offered a good-news, bad-news scenario: There will be 200 live LTE networks in 70 nations by 2015, certainly an auspicious start for a new technology, but they will be deployed using 38 spectrum frequency combinations, a toxic situation if ever there were one.
The GSMA identified the 2500/2600 MHz IMT-extension band as the most globally harmonized for LTE deployments to date, accounting for more than 50 percent of live networks in 2011. However, 700 MHz is a preferred band for LTE in the United States, where operators are exploiting the digital dividend accrued when analog TV broadcasters were forced to give up their 700 MHz spectrum. But 700 MHz is far from being a global LTE spectrum band, and Europe's digital dividend band will be 800 MHz. Further, 1700/2100 AWS spectrum will get extensive use in the Americas but nowhere else. And things get really messy in the bands above 2100 MHz with a mix of FDD and TDD confusing things all the more.
The GSMA issued a call for global spectrum harmonization in order to allow the emerging LTE ecosystem to mature and enable equipment vendors to deliver globally compatible networks, LTE devices and chipsets. That's not a new demand, and it's as unlikely to happen as ever because governments have repeatedly shown that when it comes to finding more spectrum or re-farming frequencies, national interests trump global considerations.
Vendors have proven that supporting multiple frequency bands and technologies in a single device can be done, so there's no reason to think they can't kick in more spectrum bands and air interfaces for next-generation products. But doing so presents challenges to chipset size, power management and manufacturing costs. Add those issues to the standard RF concerns of layout and routing, and it's easy to see why LTE device makers will have their hands full.
With no spectrum harmonization in sight, chipset providers and device manufacturers could ultimately become the puppet-masters controlling LTE's take-up and long-term adoption. Each will assess the addressable market for the spectrum combinations available and will make products accordingly. There may be 38 possible LTE frequency combinations, but no device maker in its right mind will try to address all of them. Instead, vendors will focus product development on the most realistic and lucrative regional band combinations. That means operators with less-than-attractive spectrum allocations may actually find themselves with zero compatible devices.
Moreover, spectrum fragmentation means the LTE industry won't enjoy dramatic economies of scale or full-fledged LTE global roaming for a long time. That's a shame because LTE and LTE-Advanced have so much to offer in terms of performance, but it's not the first time marketplace realities have kept a good idea from being fully executed. --Tammy