INNOVATION

Checking The Claim: A Wireless Network That Streams A Thousand Times Faster?

As wireless data networks face more congestion, will entrepreneur Steve Perlman’s latest idea be the ultimate fix?

March 28, 2014

Steve Perlman hopes his invention, the pCell, will transform how we transmit data from our smartphones. Artemis

Though smartphone plans seem to only be growing more expensive, it may not feel like the quality and speed of service has followed suit. Take, for instance, the last time you tried to use a mobile device at a packed event, like a sold-out football game. The frustrating lag you likely experienced is a microcosm of what more people will encounter as more data-hungry mobile devices bog down existing networks.

The demand for data, which doubled in the past year, is expected to multiply by a factor of 25 by 2020. While that year may seem far off, carriers are already scrambling to ease the growing congestion—lobbying the Federal Communications Commission to wipe limits on the purchase of additional low frequency and investing in small tower systems to prevent larger ones from becoming overloaded.

Entrepreneur Steve Perlman's proposed solution, developed over the past decade, is a radically different solution. Instead of spacing out cellular sites so they can send and receive signals with less interruption within a defined radius, his pCell technology is designed to actually take advantage of interference. And as the signals intersect, he says, they create a network that delivers data at speeds 1,000 times faster than what current networks provide.

Checking The Claim: A Wireless Network That Streams A Thousand Times Faster? — The top drawing shows how conventional cell antennas are positioned; the bottom illustrates how pCells could be placed across a city. Artemis

To visualize how this super-charged wireless network would work, think of conventional cell towers as concession stands at a football game. Using this analogy, ordering and purchasing, say, a hot dog, is similar to how cellular data is transmitted: Uploading and downloading are transactions, carried out on a first-come, first-serve basis. So during peak periods, such as halftime, long lines and gridlock are all but inevitable.

While playing a video on, say, YouTube would typically involve your connected mobile device sending a request to increasingly overtaxed towers, a pCell network would aim to circumvent these kinds of scenarios by routing data between mobile devices and websites through what's called Distributed-Input-Distributed-Output (DIDO), a data transfer technique that uses a cloud-based information center as a middleman. If a smartphone user, for instance, attempts to watch that same YouTube video, the cloud server receives the streaming data immediately and sends it out as customized “radio waveforms” that, regardless of how many other connected devices are in the vicinity, can be transmitted simultaneously, rather than one by one.

With Perlman's method, the requests—like watching a video or clicking on a link—that travel between a device like a smartphone and the cloud server are relayed using a series of small, shoebox-sized "pWave" radio antennas installed nearby.

In this case, having a higher concentration of antennas in a given area is actually a good thing, because interference generates a tiny invisible cell around the device. These "pCells," as Perlman explains it, function as private cell towers that transfer data to each device. In essence, it's kind of like having a hot dog stand all to yourself.

"The pCell is basically this little bubble of full spectrum around each person's phone that no one has to share," he says. "The idea is that instead of you moving around a cell tower, the cell tower follows you."

What makes Perlman's approach particularly disruptive is that it runs counter to how communications technologies work. Efforts to improve data streaming have always operated out of a paradigm that mimalizes signal inteference. In 2011, after experimenting with the concept for several years, Perlman introduced DIDO in a white paper that described how the system would work in practice. His company, Artemis Networks, based out of San Francisco, has since tweaked the pCell technology to ensure it is compatible with existing smartphones with high-speed LTE capabilities. It's also started to demonstrate the technology on video and in person for curious media outlets (albeit within the confines of small, controlled environments like labs and office rooms) in hopes of getting the industry to take notice.

While the technology has its fans, it also has its fair share of skeptics. Steven Crowley, a wireless engineer who consults primarily with foreign carriers, told the New York Times that Perlman's claims “seem difficult to achieve in practice.” CNBC columnist Ina Fried wrote that while Perlman is "no stranger to big ideas," he has "struggled to get mainstream adoption for those technology breakthroughs." (The remark is in reference to the WebTV creator's previous ideas like Moxi, a hi-definition satellite TV recording system, and Onlive, a cloud gaming service; both failed to live up to early hype).

Even if the technology turns out to be feasible, deploying it on a large scale would require a hefty investment from carriers and service providers. They'd not only have to build and run large data centers, but they'd also have the labor-intensive task of outfitting service areas with a sufficient number of pWave antennas.

Still, Perman contends that the budget for setting up a pCell network throughout a city would be about 10 percent of what it costs to erect conventional networks. According to his calculations, eight pWave antennas, which cost $500 each to manufacture, would be sufficient to service customers within the radius of a standard cell tower. And since a large urban city like San Francisco currently has 32 cell sites, a minimum of roughly 256 antennas would be enough to provide wireless Internet access for the entire city's 825,000 residents, provided some entity gets a cloud data center up and running.

Asked if any service providers have expressed interest in licensing the technology, Perlman replied, quite confidently, that several leading firms are "lining up" to discuss the possibility. He added that his company is already collaborating with "major carriers around the world" to run field trials, though he wouldn't divulge specifically who these parties are. Thomas Pica, a Verizon spokesman, told the Times that the carrier was aware of pCell, but didn't say whether it planned to adopt it.

As part of a private trial, the team is in the process of installing a network consisting of 350 "weatherproof" antennas around San Francisco. Perlman hopes, through a series of partnerships, to have the first pCell system running in a U.S. city by the end of the year, with expansion to major markets by the end of 2015. He's also looking into implementing pCell on smaller scales, such as on college campuses, in stadiums or hotels.

"We're showing that a small company with only eight people has figured out something all the universities and labs have overlooked," he says. "It's a game changer, and it's all thanks to the little guys."