Wireless bandwidth is like land in Manhattan -- it's extremely valuable because they're not making more of it.
But we sure are using more of it. The wireless-industry association CTIA reported in October 2011 that the number of wireless devices in the U.S. had, for the first time, exceeded the number of people.
And Mobile Future, a coalition of vendors and consumers, estimated in a March 2011 report that by 2014, voice traffic will comprise only 2 percent of the total wireless traffic in the United States -- a worrisome statistic because, as the report noted, smartphones consume 24 times more data than old-school cell phones, and tablets consume 120 times more data than smartphones. (See "Data needs bandwidth, but how much?" for details.)
The result: Wireless networks are edging near capacity, not just in the United States, but all over the world. Credit Suisse conducted a survey last year that revealed mobile networks in North America were running at 80 percent of capacity, with 36 percent of base stations facing capacity constraints. The average globally for base station capacity utilization, the report said, was 65 percent.
The problem is going to get worse before it gets better. With advancements in connected cars, smart grids, machine-to-machine (M2M) communication, and domestic installations such as at-home health monitoring systems, wireless demands will only increase.
As with all things mobile, there are no simple answers, if only because potential solutions rely on agreement among a sizable and incompatible array of players -- from spectrum owners (both telcos and broadcasters) and regulators to government agencies and, of course, consumers demanding the latest in cool devices and applications.
With all this sturm und drang, what happens to businesses that are increasingly relying on the productivity that mobile devices deliver? Their numbers aren't inconsiderable: According to a recent IDC report, 75 percent of the North American workforce was mobile in 2010, and iPass reports that 91 percent of mobile workers use their smartphones for work.
Most carriers have already imposed data caps, and industry watchers say the laws of supply and demand indicate that wireless-plan prices can only go up. When AT&T switched on its 4G access in 11 areas in December of 2011, AT&T Business Solutions chief John Stankey warned that if the spectrum situation remains unchanged, pricing will rise, a prediction borne out by AT&T's announcement last week of new pricing structures for its smartphone and tablet customers.
Some experts fear there could be dire consequences. Richard Bennett, senior research fellow for the Information Technology and Innovation Foundation (ITIF), says, "If we can't get spectrum bandwidth for more mobile devices in the next five years, prices will rise, performance will suffer, and innovation will be impaired."
Adds Scott Bergmann, the CTIA's assistant vice president for regulatory affairs, "Companies will have to make some hard choices, including limiting the amount of data employees can use based on job function."
Playing nice -- or not so nice
Other experts -- including some IT executives -- opine that we've encountered spectrum and wireless issues like this before, and sanity has, eventually, prevailed (remember the days before roaming agreements, or the brouhaha over wireless garage door openers?).
But coming to agreement on a workable solution may not be so easy this time around.
The wireless spectrum is spoken for, which means carriers must either do more with what they have -- which would require a costly build-out (as much as $40 billion to double capacity) of cell towers and base stations -- or find bandwidth elsewhere. That "elsewhere" is primarily in the broadcast bandwidth, but broadcasters, unsurprisingly, aren't interested in giving up spectrum that they've controlled for decades.
"Installing base stations is expensive," explains Phil Solis, research director for mobile networks at ABI Research. "Carriers [instead] fight for spectrum, because that's a cheaper way to add capacity. But by getting spectrum, you reduce the chance of new competitors."
At a CQ Roll Call panel called Finite Spectrum, Infinite Demand held in Washington, D.C., last November, Harold Feld noted, "There's a culture of confrontation, not cooperation, around spectrum. We have to come up with ways to approach the issue in a more collaborative fashion." Feld is director of public affairs for Public Knowledge, a nonprofit focusing on Internet openness.
That isn't happening. Charles Golvin, principal analyst for Forrester Research, says, "It doesn't matter if Mother Teresa makes the request. If you're used to something, and someone tries to take it away, you're going to object."
So what is happening? Here's a look at the players, their perspectives, and recent dealings.
Carriers. Carriers like AT&T and Verizon are in a tough position. On the one hand, they want to entice customers with new services and download speeds. But to do that, they need spectrum. AT&T wanted to acquire T-Mobile not for its customers, but for its spectrum. That's also why it spent $1.9 billion in December 2011 to acquire spectrum from Qualcomm, and why it's rumored to be buying Dish Network.
Ditto for Verizon. Also in December, it paid $3.6 billion for mobile spectrum licenses from SpectrumCo, a joint venture of three cable providers, and paid Cox $315 million for spectrum in another deal. (And don't forget rural carriers: they have a separate set of complaints about too much consolidation of spectrum in the hands of the large carriers.)
Broadcasters. Wireless carriers complain that TV and radio broadcasters got their spectrum for free back in the day. Broadcasters counter that they've invested billions in it since then. ITIF president Rob Atkinson and other experts believe the solution lies in spectrum incentive auctions -- in which broadcasters that wished to sell bandwidth would be compensated by telcos purchasing the spectrum. "Only 10 percent of Americans get their television over the air, yet broadcasters have more frequency than all four major carriers combined," Atkinson points out.
The broadcasting community counters by saying wireless carriers still have lots of unused methods for more wisely exploiting the spectrum they do have. And more to the point, broadcasters don't want to give up spectrum because they want to be involved when mobile video service becomes common. "If data is the central driver," National Association of Broadcasters COO Chris Ornelas told the CQ Roll Call panel, "let's talk about how broadcasters can be part of that."
Regulators. When it comes to authorizing auctions, the FCC complains that Congress is dragging its heels, and vice versa. At this year's Consumer Electronics Show, FCC chairman Julius Genachowski said, "If we don't authorize incentive auctions, we'll get swamped by an ocean of demand and risk losing the competitive advantage to lead the world in innovation."
Congressman Cliff Stearns (R-Fla.), a member of the Subcommittee on Communications and Technology, faults the FCC for moving too slowly in addressing issues within Congressional bills. At the CQ Roll Call panel, he said, "Their response time has to be faster."
At the same time, Stearns wants to wait for the results of a spectrum inventory being conducted by the National Telecommunications and Information Administration (NTIA), the president's principal adviser on telecommunications and information policy, in order to ensure spectrum is distributed fairly.
Public safety advocates. The Department of Defense controls extensive swaths of the wireless spectrum (by some estimates, as much as Verizon and AT&T combined), and insists that national security trumps commerce. Public safety officials tend to agree -- they want spectrum for interoperability of all public safety radio channels to avoid a repeat of the 9/11 communications breakdown, when police, fire, and medical responders were communicating on different frequencies, making coordinated rescue efforts difficult.
To counter that situation, the FCC is in favor of allocating a chunk of spectrum, known as the D Block, to public safety use (FCC white paper.pdf).
How IT copes with the crunch
For the most part, IT execs seem content to let the spectrum drama play out as they cope with more immediate wireless concerns.
At Madison, Wisc.-based CUNA Mutual Group, about 25 percent of the insurance company's 4,000 employees use smartphones to access email remotely. Mark Winger, the company's vice president of IT for product and administration, recommends being "open and collaborative" with carriers; CUNA Mutual has relationships with both AT&T and Verizon. "We share our issues and they share theirs, and they provide us with test devices to gauge how their network works."
At SBLI USA in New York, VP of IT Paul Capizzi has gone the opposite route for the insurance firm's 100 employees, scaling down to one carrier (Verizon) over the last five years. "It's like a friends and family plan, where everyone shares minutes." Capizzi has adjusted the plans so that individual employees get a certain amount of minutes, but with the flexibility that if one person goes over and another goes under, it all averages out. "The last thing you want is have someone come in and say, 'You went over on your data plan.' "
Like CUNA Mutual, SBLI allows employees to use personal devices for work-related activities. That concerns Capizzi, citing the scenario where an employee might be using the device extensively on weekends. To offset increased costs of bandwidth requirements, he says, he would probably recommend subsidizing the cost of the devices based on work-related usage; anything above that would be the employee's responsibility.
Relative to today's usage patterns, Forrester's Golvin sees few impending changes, saying, "Some businesses might examine their communications budgets more carefully, but by and large, employers see increased productivity as a result of their employees being connected and performing whatever task is in front of them." In other words, the mobile revolution will continue, spectrum crunch or no.
If and when prices do rise significantly or data caps become more onerous, Golvin believes employers will either invest in better management tools, so they can monitor how much employees are working versus playing games, or deploy dual-SIMM devices and require employees to swap out the SIMMs for personal activities -- or both.
ABI Research's Solis agrees that companies won't limit users in the short term, and cites another technological option. IT can make sure that when people are around company locations, they're on the company network, whether over Wi-Fi, or some kind of femtocell or picocell bandwidth extension devices that are part of the private network.
Golvin adds, "Employees already alter their behavior to get the 'best' experience [from their gadgets], whether that's the fastest, the cheapest, or with the lowest latency. End users will continue to develop more awareness of the impact of their behavior, just like they did with voice minutes."
Full-blown crisis or short-term crunch?
Experts vacillate between calling the wireless spectrum situation a crisis or a crunch.
Based on networking experience stretching back more than 20 years, CUNA Mutual's Winger is confident that the situation will correct itself, whether through more bandwidth or better technology. (For more on potential bandwidth stretching technologies, see Busting bandwidth barriers.)
"In the early 90s, we were dealing with narrowband technology, but the technology eventually evolved to spread spectrum, where data could hop. There are advances occurring in multiple solutions, so there are many possible alternatives," Winger says. "We solved those issues, and we'll solve these."
Data needs bandwidth, but how much?
Solving the crunch in wireless bandwidth is a difficult problem made even more challenging because data usage is so inherently different from voice usage, making it difficult for carriers to estimate demand.
"Not only has the demand for capacity on our wireless networks been accelerating significantly, but it's been accelerating in a non-scalable way," says Charles Golvin, an analyst for Forrester Research.
"While we saw tremendous growth in penetration of mobile phones over the last 20 years, each new subscriber's behavior, and the demands they made on the network, were predictable." Carriers could gauge demand based on the number of minutes on each subscriber's plan, and predict capacity needed accordingly.
"Over time, the carriers got better at predicting demand and how much capacity they'd need to handle another n callers on the network."
The current problem: Data usage doesn't map to that model at all. Golvin says, "Any consideration about network load is beyond [users'] consciousness. They want rich media and video when and where they want it. The ability for carriers to model the demand for capacity in total among [their] customer base, but also to model demand on a time-of-day basis, is rudimentary compared to the [voice] situation of 20 years ago."
Busting bandwidth barriers
Here are some possible solutions to the bandwidth crunch.
Unlicensed spectrum. For now, the easiest way to avoid the wireless bandwidth crunch? Stay off that spectrum as much as possible. Groups like the Wireless Broadband Alliance promote offloading as much wireless networking as possible to Wi-Fi networks.
Full duplex. In September 2011, Rice University engineers announced technology that would allow mobile network operators "to double the throughput in their networks without adding any cellular towers." Their plan: Add full-duplex capabilities so that wireless devices can send and receive data at the same time. In February 2011, researchers at Stanford University said they accomplished a similar full-duplex capability via a noise cancellation technique.
White spaces. There are vacant airwaves between TV channels used to prevent interference; the FCC is promoting repurposing white spaces for data.
Silicon Valley-based freelancer Howard Baldwin writes about networking and mobile technology, among other enterprise-related issues.
This story, "Wireless bandwidth: Are we running out of room?" was originally published by Computerworld.