Lots of shiny new mobile data consuming devices made their debut at the recently concluded Mobile World Congress in Barcelona. You can Ray Wong's excellent round-up of all about new hardware that you should care about from LG, Nokia, Samsung, and others here and here. Each of these shiny new mobile devices means more people merging onto the 4G LTE super highway cellular frequencies. And as anyone who has driven on the 405 in LA or the Long Island Expressway during rush hour knows, more users means massive jams. In the case of mobile data, that means frequency jams and intolerably slow speeds.
According to a Cisco, global mobile data traffic grew 81 percent in 2013, and fancy 4G devices generated 14.5 times more traffic on average than non-4G deices. With more folks upgrading to 4G LTE phones, 2014 mobile traffic could easily double compared to last year. The tractor trailers of this LTE highway are bulky video downloads, which accounted for 53 percent of downloads last year, and will account for more than two-thirds of all mobile data traffic by 2017. This jamming LTE traffic doesn't bode well for increasing cluttered spectrum.
So what's the knee-jerk solution to crowded information superhighways? Why, build more wireless broadband highways, of course, which means adding more spectrum. Over the next 18 months or so, the FCC will be wielding its virtual gavel over a series of spectrum auctions, including an "incentive" auction in mid-2015 in which TV broadcasters will be seduced into selling some of their spectrum to the government for subsequent auction to mobile carriers.
One wireless guru, however, insists that we don't need more spectrum — Martin Cooper, the father of the cellphone. Cooper says that adding spectrum is a panacea, and not a real solution to the pending mobile data traffic jam.
Don't Widen The Road, Shrink The Cars
Back in the winter of 1972-73, Martin Cooper headed Motorola's mobile phone and system development, which you can read all about here. To stretch the highway analogy to its illogical conclusion, instead of building new roads, Cooper insists that we try to fit more traffic on the existing ones. In more literal terms, we waste a LOT of spectrum through massive inefficiencies by "a 99 order of magnitude." Solutions, Cooper says, abound.
One of these spectral efficiency solutions is SDMA — Spatial Division Multiple Access. Developed by ArrayComm in the mid 1990s, SDMA, aka multi-user MIMO (MU-MIMO) spatial multiplexing "and other terms designed to confuse rather than clarify," Cooper says. MIMO, SDMA, or whatever you call it, is used in many new 802.11ac routers, and allows multiple folks to talk on the same frequency simultaneously. "Deployment of SDMA could effectively treble available bandwidth without adding any additional spectrum," Cooper claims.
A proprietary SDMA standard, field proven by deployment in 22 countries in the late 1990s, was never widely adopted. In late 2011, Apple was awarded a patent for a MU-MIMO/SDMA system – for what use, we don't know.
The Radio Knows
Then there's Dynamic Spectrum Access (DSA), which encompasses a myriad of spectrum efficiency techniques. DSA adds more aggressive spectrum sharing based on a concept known as "cognitive radio" (which "knows" what users are doing in any given spectrum band), and more intelligent use of frequency band selection.
Ultimately, Cooper says, an efficient spectrum user will:
- Use the minimum bandwidth required for his or her specific transmission,
- Use that bandwidth for the minimum time and then release it for others to use,
- Occupy the least possible area with the users transmission energy, and
- Use the frequency band most suited to the specific transmission (a transmission in a rural area with a large cell might use the 700 MHz band while a user transmitting from one room to another might use 28 GHz).
"Techniques that achieve these four objectives already exist in some form or are in development," Cooper insists, "and hooks exist in LTE to add these techniques in the future."
Implementing Efficient Solutions
Most of the SDMA intellectual property is in the public domain, eliminating complicated licensing issues. "Other techniques will come later," Cooper reports, "but SDMA offers enough improvement to carry us another five or more years." Cooper also doesn't see the cost of implementing these spectral efficiency technologies as a major issue. "Most of these solutions are needed only for congested urban areas, where the cost per user is far less with efficient spectrum use," Cooper says.
The challenge is to get cell site companies such as Ericsson, Huawei, Nokia, Lucent, et al to implement the new technology. A perfect opportunity exists as carriers begin their rollout of LTE Advanced. In fact, at MWC, AT&T declared its own LTE Advanced implementation was aimed at spectrum efficiency, not necessarily just at faster 150 Mbps speeds.
Crisis Promotes Solutions
Two new proposed technologies also could ease congestion. The first is an LTE-compatible solution called pCell, from a company called Artemis, founded and led by Steve Perlman, who brought us WebTV nearly 20 years ago. pCell adds a series of antennas to cell stations connected to data centers, where sophisticated mathematical calculations are performed to essentially create "private" cells (hence "pCell"). pCell has garnered a lot of publicity, with high-profile profiles of Perlman in The New York Times, CNET, Time, Wired and others. But Cooper thinks pCell is simply SDMA "hyped up with 'magic'" — SDMA processing done in pCell's data centers, instead of at the cell site.
A slightly more mature technology, mmWave, or EHF (Extremely High Frequency), are new higher frequency millimeter wave bands such as 28 GHz, 60 GHz, or higher, designed to offload short distance data traffic. But mmWave still has technical hurdles to overcome, one of which is interference caused by rain and snow.
Some of the current spectrum issues will be solved as we move into nearly empty LTE Advanced bands, just as some of the 3G traffic issues were solved when we moved to 4G LTE. But thus far, LTE Advanced is available only on a test basis from T-Mobile in Dallas. AT&T and Verizon both plan rollouts, if not this year then next, once again kicking the spectrum crowding issues — and solutions — farther down the road.
"The spectrum crisis has not happened yet," Cooper admits, "but when it does, technologists will come through again and the crisis will be solved by spectrally efficient technology, not redistributing the radio spectrum."