Worries mount that upcoming LTE-U deployments will hurt Wi-Fi

 Worries mount that upcoming LTE-U deployments will hurt Wi-Fi

Despite the assurances of its inventors, there’s an ominous lack of proof that LTE-U technology won't muck up current Wi-Fi deployments


LTE-U is a technology developed by Qualcomm that lets a service provider broadcast and receive signals over unlicensed spectrum, which is usable by anybody -- specifically, in this case, the spectrum used by Wi-Fi networks in both businesses and homes. By opening up this new spectrum, major U.S. wireless carriers hope to ease the load on the licensed frequencies they control and help their services keep up with demand.

That demand for connectivity continues to grow on a scale that we're in danger of running out of adjectives to describe -- a recent Cisco estimate suggested that monthly global demand could top 24 exabytes (that's 24 million terabytes) by 2019, a nearly 10-fold increase from today.

+ALSO ON NETWORK WORLD: Qualcomm exec on LTE-U: We're not trying to mess up your Wi-Fi + Sprint layoffs likely as part of cost-reduction plan +

Since an increasing proportion of that demand is being felt by wireless networks, the crunch is on for the carriers, who have spent vast sums on purchasing the rights to chunks of the U.S. airwaves. The industry bought almost $45 billion worth of spectrum in the FCC's most recent auction in February, and has been working hard on ways to squeeze the most out of what they've got. Hence LTE-U.

The catch, of course, is that LTE-U's frequencies are already being used by every Wi-Fi access point in America, which means that it runs the risk of interfering with networks that home and business users rely on. While Qualcomm has sworn up and down that LTE-U will include features that will avoid interference with existing networks, there's little independent proof of those claims.

Unsurprisingly, several outside experiments that pitted standard LTE technology or "simulated LTE-U" technology, in the case of one in-depth Google study, against Wi-Fi transmitters on the same frequencies found that LTE drastically reduced the throughput on the Wi-Fi connection. IEEE member and wireless expert Craig Mathias said that the way LTE deals with interference makes it dominate frequencies that it shares with Wi-Fi.

"LTE-U has a shorter backoff period, so it'll grab the air faster than Wi-Fi can," Mathias said. "So you can expect, in high-density deployments that LTE-U will have a rather deleterious effect on Wi-Fi performance."

Consequently, consumer advocacy groups, as well as the wired broadband industry and other big players that rely on Wi-Fi, have been sounding the alarm about LTE-U. Michael Calabrese, who directs the Open Technology Institute's Wireless Future Project, told Network World that the root of the problem is that Qualcomm and its partners (notably Ericsson) are acting unilaterally, ignoring the well-established standards bodies at the IEEE and 3GPP.

Operating outside of the usual structures, it's difficult for the rest of the industry and the public at large to get any kind of firm assurance that LTE-U technology is going to play well with others, he said.

"Typically, there's a standards process -- and these licensed and unlicensed bands are handled by entirely different bodies, and LTE-U skipped both of them. It's not in 3GPP nor is it in IEEE. It's just a proprietary technology that Qualcomm and Ericsson are just putting out there all of a sudden," Calabrese said.

This isn't to say that the U.S. wireless industry is in lock step on this technology, he added -- Verizon and T-Mobile are both reliant on Ericsson as a primary infrastructure vendor, while Sprint and AT&T are not. (Nevertheless, AT&T has joined the pro-LTE Evolve Coalition and begun its own tests of the technology.)

By contrast, LTE-U's opposition has generally remained united. Philip Berenbroick, a counsel for government affairs at advocacy group Public Knowledge, said that the technology might not cross legal boundaries, but it could easily break the unwritten rules.

"There's a general etiquette in the unlicensed space that you don't disrupt licensed or other unlicensed services," he said. "We would like to see LTE-U comply with the etiquette that is pervasive in that band."

A protocol called "Listen-Before-Talk" could be key to resolving the problem, Berenbroick argued. LBT, which makes transmitters ensure their channels are clear before sending their data, is referred to as a "politeness protocol" in the wireless industry. It provides an even-handed way of ensuring the best possible shared access to a given piece of spectrum, and is legally mandated in the European Union and Japan, though not in the U.S., China or South Korea. An alternative to LTE-U, called LAA, or license-assisted access, incorporates LBT as a safeguard against interference. (LAA is currently under consideration by 3GPP, one of the standards groups that LTE-U's backers bypassed.)

"Because they don't check for other users that are using the bandwidth first, [LTE-U transmitters] can have the effect of slowing or degrading other unlicensed traffic that's out there," Berenbroick told Network World. "We would like to see LTE-U observe Listen-Before-Talk."

LTE-U, instead, uses a coexistence protocol called "duty cycling," which simply creates pre-set on and off periods for an LTE transmitter -- the lengths of which are controlled entirely by the carrier. That's not good enough, according to Jennifer Andreoli-Fang, principal architect at CableLabs, a cable-industry-backed research group.

"There is no requirement to share fairly in time, to avoid interrupting Wi-Fi transmissions mid-stream, or to adapt to different levels of Wi-Fi usage and traffic," she wrote in a September blog post.

The concerns over LTE-U are unlikely to die down soon; reports indicate that Verizon and T-Mobile may start using the technology as early as next year, and although FCC Chairman Tom Wheeler has been reluctant to get involved with the issue, he stated that the wireless industry risks further regulation if the rights of existing unlicensed users aren't sufficiently protected.

None of the involved parties are likely to welcome government oversight, so the preference all around is likely to be for an industry-led solution. Indeed, in the late 1980s, when the current unlicensed band regulations were being developed, nobody thought that the frequencies would ever be used for commercial purposes, according to Mathias.

"Right now, it's a cautionary tale," he said. "The assumption that many of us had made was that the LTE community would rely on handoff to Wi-Fi, rather than operating LTE in the unlicensed bands."

Nevertheless, it's a potentially serious issue -- the increasingly central role of wireless technology to every facet of life in the developed world means that superficially dry and technical discussions about interference can gain real-life consequences in a big hurry.

"When you start looking at the mission-critical nature of Wi-Fi -- all these Internet of Things applications that are coming along, many of which will have medical and security elements to them, plus just regular voice/data/video -- you've got a problem," Mathias warned. "So this is clearly something that multiple entities within industrialized economies are going to have to deal with."

This story, "Worries mount that upcoming LTE-U deployments will hurt Wi-Fi" was originally published by Network World.

From CIO: 8 Free Online Courses to Grow Your Tech Skills
View Comments
Join the discussion
Be the first to comment on this article. Our Commenting Policies