Skip to Content

Entrepreneur Hopes to Use Interference to Improve the Mobile Internet

WebTV’s creator thinks his forthcoming wireless technology will give us faster, better mobile Internet access.
March 10, 2014

Ten years ago, when most of us still had no idea what a smartphone was, Steve Perlman was contemplating a future in which we’d be watching so many YouTube videos over cellular networks that the radio frequency bands available to wireless carriers would get clogged up.

Phone home: Each of these iPhones is running a different HD video that was sent with Artemis Networks’ pCell wireless technology.

The situation isn’t quite that bad yet, but mobile data use has risen dramatically, with videos now making up more than half the traffic flowing over mobile networks, according to Cisco’s Visual Networking Index. Even if we are still short of a so-called “spectrum crunch,” academics and industry leaders are concerned that we’ll be unable to cope with the rising demand for mobile data.

Now Perlman, a longtime inventor who cofounded WebTV and founded the cloud gaming service OnLive, says he has a solution that can make better use of the wireless spectrum and could eventually lead to cheaper, lighter phones. For the last decade, he’s been working on a technology called pCell—for personal cell—which his San Francisco-based wireless company, Artemis Networks, hopes to roll out late this year. However, while Perlman positions pCell as revolutionary, it’s encountering some skepticism within the very industry it hopes to revolutionize.

On a standard cellular network, cell towers are placed to minimize interference with others, and each transmits a signal that can be picked up by phones within its signal area, which might range from 50 square meters to a couple of square miles. Each phone gets a portion of the tower’s capacity, which can mean poor service if too many people try to use the same node simultaneously.

Perlman says pCell takes a different approach: it embraces signal interference. In his vision, base stations smaller than your typical satellite TV antenna are placed wherever it’s convenient (such as on the roof or the side of a building), and their signals purposely overlap. Those overlapping signals, Perlman says, combine constructively to create a sort of personal cell, a centimeter in diameter, that moves with you as you move around the network. The signal doesn’t diminish as each additional user joins the network. Overall capacity can grow by adding more access points.

The pCell network determines the wireless conditions for each phone on a node by paying attention to the phone’s normal uplink signal, which is passed on to a pCell data center. To get, say, YouTube videos to stream to several different smartphones, the video data would first be sent from YouTube to a pCell data center, where the optimal waveform for any given access point on the network could be determined. The radio waves that would then be sent from the access point to the smartphones would combine as they approached the handsets and generate the necessary signals.

Perlman says that pCell works with existing LTE devices, and that users won’t notice as they move from a regular cellular access point to a pCell node. Artemis is working with wireless carriers and spectrum owners, he says, as well as entrepreneurs interested in using pCell over unlicensed spectrum.

Perlman says the startup has a partner in San Francisco that will allow it to install base stations for free on 350 rooftops, and Artemis plans to start deploying base stations as early as the fourth quarter of this year. He’ll need a lot of base stations to cover a large area if they’re like the one Perlman showed me in his San Francisco office—the range depends on the base station’s power, but that one, a five-watt unit, would cover about two to eight blocks, depending on the frequencies used.

Silver box: Artemis Networks’ pCell technology uses access points called pWaves to communicate between a data center and gadgets running on the network.

Over time, he hopes, smartphones can be made specifically for pCell networks. Such handsets, he says, would be thinner, lighter, and cheaper than existing phones, since much of the current hardware could be eliminated as more data processing happens on remote servers.

PCell is the latest in a long line of Perlman’s technological endeavors. While working for Apple in the 1980s, he helped come up with the QuickTime video platform. Then he created the Internet-connected TV service WebTV in 1995; he sold it to Microsoft in a $425 million cash-and-stock deal barely two years later. His next big project, the cloud-streaming video game site OnLive, was a flop: it fell apart in 2012 after several years of losing money and reëmerged this month with new owners, leadership, and game services.

For years, Perlman has been working on pCell in the background, and he seems giddy that it’s nearly ready to be deployed. During an interview in the dark, airy office for Rearden Companies—an incubator Perlman started in 2000—he speaks rapidly and excitedly about the technology. He’s accompanied by a slide presentation that illustrates things like the way radio waves would flow throughout a pCell network. It’s the longest project he’s ever worked on, he says.

“We got it working a long time ago. We just didn’t get it working in a practical way until recently,” he adds, giggling.

Yet it’s not clear that pCell is truly a breakthrough. Ken Rehbehn, a principal analyst at mobile market researcher Yankee Group who tracks mobile wireless infrastructure, hasn’t seen the technical details needed to evaluate the significance of pCell’s technology. “Call me skeptical,” he says.

Bhaskar Krishnamachari, an associate professor at the University of Southern California who studies wireless networks (and a 2011 MIT Technology Review Innovator Under 35), is more encouraged. He says pCell sounds like a version of distributed multiuser MIMO technology. MIMO, which stands for “multiple input and multiple output,” is a method of using multiple antennas in a base station and in a receiver (such as a cell phone) to move more data over a network at once. With distributed multiuser MIMO, each transmitting antenna and receiver is a distinct device, and the base stations are spread out and connected to a common server in order to operate over a larger area and improve the network’s capacity.

Perlman’s technology was originally called DIDO, for “distributed input, distributed output,” which sounds similar. But he claims that his technology will continuously increase a network’s capacity as you add base stations and servers while MIMO levels off.

After reading several of the patent filings behind the technology, Krishnamachari says, it does sound at least theoretically possible.

Perlman showed me several demonstrations of pCell, including one using five megahertz of spectrum and pCell antennas to stream high-definition video to eight iPhones simultaneously. Normally, he said, that amount of spectrum would only support a fraction of that traffic. Yet here all the displays glowed against a black table with great picture quality.

Of course, this was a completely controlled environment, and in the real world, there are challenges like ensuring the base stations are synchronized, equipping the network to adequately track users as they move around, and maintaining the network—not to mention setting up all the base stations in the first place.

For these reasons, Krishnamachari expects to see pCell technology used not in big citywide deployments but in more contained, heavy-traffic areas like airports, malls, and stadiums, where a smaller number of them will suffice.

“The more boxes you put out there, regardless of what they are doing, the more points you have to fix when something goes wrong,” he says.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.