Smartphone use makes cellular networks' collapse a real possibility

Crowd in stadium watching car racing, photographing with mobile phones.
Crowd in stadium watching car racing, photographing with mobile phones. (Getty Images)
By Jim Giles
New Scientist
Tuesday, November 30, 2010; 6:00 AM

The first thing to go might be your smartphone's connection to YouTube, with videos becoming increasingly choppy and then one day just failing to download. In your impatience, you decide to scout out the latest posts in the Twittersphere, except that, too, is temporarily down. Your e-mail is stalled, and even a simple text is now too arduous, as the world's phone networks come crashing down. In the following months, it's almost impossible to get a lasting connection, even for a voice call.

Welcome to 2013, and the first mobile meltdown.

Although this is the worst-case scenario, some kind of collapse in cellular networks in the near future is a real possibility. They are already showing signs of strain: Your phone may temporarily cut out in large crowds or at a sporting event or music gig, and if you live in New York, San Francisco or London, you may have found it increasingly difficult to make calls in your home city.

Data-gobbling smartphones are, of course, the source of the problem, as they overload networks with requests for Web pages, e-mail and video streaming 24/7. The rapid growth in the use of these devices is behind the dire predictions of a meltdown by 2013, Since many core services depend on wireless communication, the results could be devastating. The only solution, according to many experts, will be an overhaul of the way mobile communications are delivered.

Think of it as a road traffic problem. Governments in Europe and the United States allocate five-megahertz chunks of the electromagnetic spectrum to each operator's network, These chunks correspond to the lanes of a highway, carrying data either to or from the operator's transmitter. Many operators are given just two chunks - one lane each way - though some have as many as five pairs.

Like any road, these highways can hold only so much traffic. Current 3G technologies can send roughly one bit of data - a one or a zero - per second over each hertz of spectrum. That means a cell tower using two chunks of spectrum can transmit just five megabytes of data per second, a handful of streamed videos at most.

Cellphone congestion seemed like a distant prospect a decade ago, when the 3G network was rolled out. At that time, pretty much the only smartphone users were business executives on their BlackBerrys, leaving the 3G network massively underused.

Not anymore. Wireless modems added traffic when they emerged around five years ago. In 2007, Apple launched the iPhone and has now sold 50 million of the devices. And many other companies have joined the fray. Suddenly, lots of people are on the highway, each taking up huge amounts of road space. (A single streaming video occupies as much bandwidth as about 100 phone calls, for example.) As a result, the 3G highway is now overcrowded, especially in cities where lots of people use smartphones .

If the growth in smartphone sales continues at the current pace, mobile traffic will more than double every year for the next four years, according to predictions by the network technology company Cisco. Which means that the occasional congestion of today will become gridlock tomorrow, especially in big crowds in sporting events such as the Olympics.

In the past, cellphone companies used innovative engineering to increase capacity. By making the jump from 2G to 3G (G stands for generation), for example, engineers were able to squeeze five to 10 times as many bits per second into each hertz of spectrum, says Simon Saunders of Real Wireless, a consulting firm based in the United Kingdom. This meant more data could rush down the highway without holdups.

Could a similar technique stave off the wireless crunch? Internet traffic often comes in bursts as users click on a page, read, then click again, says Saunders. 3G networks struggle with this kind of traffic, but their successors - Long Term Evolution (LTE) and WiMax - should do better.

These technologies have spent years in development, yet they will let operators cram only roughly 50 percent more data into the chunks of spectrum before holdups will start happening again, a mere drop in the ocean when faced with the rise and rise of the iPhone. If LTE were the only solution in the pipeline, demand might well trump supply in a couple of years, according to a recent report commissioned by Research in Motion (RIM), maker of the BlackBerry.

Worse still, any successors to LTE will be unlikely to provide the improvements in data transfer rates that would be necessary to avoid the crunch. "LTE is so advanced and complex that it has required the global output of the entire industry to produce," says Peter Rysavy, who produced RIM's report. "If there was an alternative that worked a lot better, they would have found it."

Many cellular operators are optimistic about Option No. 2: widening the road. "If the number of cars on a highway quadrupled without additional lanes, then everything would slow down," says Christopher Guttman-McCabe, a vice president at CTIA - the Wireless Association, which is based in Washington. "We need more lanes." That would mean dishing out more chunks of spectrum to mobile operators.

Before this can happen, governments will have to go through the messy political business of persuading existing owners of spectrum rights to part with underused chunks. That is because much of the spectrum in the range that wireless operators use is spoken for by the military, TV broadcasters and satellite communication firms. The Federal Communications Commission says that various changes and reallocations will eventually free 500 MHz for cellphones. And the British communications regulator, Ofcom, has plans to reallocate close to 300 MHz of spectrum.. But for a variety of reasons, that may not happen anytime soon.

Even once that extra spectrum becomes available, it will soon be eaten up by smartphone users and their data-hungry apps. "Freeing up spectrum would be helpful," says Stirling Essex of CRFS, a British company that sells spectrum-monitoring and management tools. "The demand is insatiable."

One possible solution is essentially to tax the road hogs who are bogging down the networks,

AT&T made a move in that direction.In June, the company announced new price plans for the iPhone that come with monthly caps: 200 megabytes for $15 and 2 gigabytes for $25. The move hasn't seemed to trouble iPhone owners because they can save money by switching from their original $30 unlimited data plan and, in most cases, will not be bothered by the 2 GB limit, which is equivalent to watching more than 100 two-minute videos in a month. These caps may not be onerous, but unpalatable ones could follow unless other ways of dealing with demand are found. Other companies are considering similar tiered plans..

Fortunately, there may be a fourth way that would leave the door open for cheap and extensive Internet use: Install a cellphone transmitter in every home and office. These transmitters, dubbed femtocells, look like wireless routers and would plug into broadband connections. Several companies have begun to offer them. By shifting smartphone traffic onto the Internet, they would bypass conventional cellphone transmitters, which would serve users when they're outdoors.

Femtocells wouldn't be too much of a burden on the home's broadband connection, since the constraints of cell towers have already forced engineers to create smartphones that use data far more efficiently than traditional desktops and laptops.

This approach would also make mobile communication more energy-efficient. Existing cell towers lose 90 percent of their energy when the signal passes through an external wall.

One question with this possible solution is whether femtocells would interfere with one another when packed into urban neighborhoods. Interference is a problem for all transmitters, and engineers routinely monitor transmissions in areas where signals overlap and tweak the output of towers accordingly. As transmitters have become smaller and too numerous to adjust manually, engineers have developed technology that listens to signals from other sources and makes the necessary changes automatically. So far, these systems have coped. But femtocells would add another layer of complexity, and no one knows whether the automated systems are up to the job.

It will take some time, as the operators roll out these possible solutions and jump the inevitable technical hurdles, before we know if the dire warnings about 2013 and beyond can be avoided. Until then, the question looming for many is: Having tasted the wonders of ubiquitous Internet, could we ever live without it?

Giles is a correspondent for New Scientist magazine who is based in San Francisco. This story was excerpted from New Scientist and can be found at full length at

© 2010 The Washington Post Company