As 3G Networks Buckle, Ruckus Wireless Sends Smart Wi-Fi to the Rescue
Wi-Fi: it’s the new utility, as crucial to many consumers and businesspeople as electricity and telephones. Wi-Fi chips are built into 800 million new laptops, mobile phones, tablet PCs, printers, video game controllers, and TV set-top boxes every year, and Wi-Fi networks blanket virtually every office, library, airport, hotel, café, and campus.
If only Wi-Fi were as dependable as most other utilities. Unfortunately, the more people who install access points in dense urban environments, the more interference arises between devices, slowing everyone’s connections. On top of that, Wi-Fi signals share the unlicensed 2.4-Gigahertz frequency band with Bluetooth, Zigbee, cordless phones, baby monitors, car alarms, ham radio, and even microwave ovens. (Yes, Wi-Fi devices use roughly the same wavelength as the radio pulses that, at much higher power, can bake a potato.) That’s why you might have a blazing-fast, 30-megabit-per-second Wi-Fi connection in one corner of your office, but lose it a minute later if you shift six feet to the left. Especially if someone is using the microwave at lunchtime.
Part of the problem stems from the fact that Wi-Fi signals, like cellular or FM signals, spread indiscriminately in all directions. That’s a feature, not necessarily a bug: your home or office Wi-Fi router doesn’t know where you are, so it has to send signals everywhere. But what if your router were smarter, and could use modern beam-forming technology to shoot a signal straight at your device—then make that beam follow you as you moved from your desk to your couch to your kitchen? In principle, signals from a dynamic, directional Wi-Fi antenna would have a greater range, and would be far less vulnerable to interference.
Well, it turns out this is an idea radio engineers have been thinking about for a while. And there’s a company in Sunnyvale, CA, called Ruckus Wireless that’s making it work so well that cellular carriers may soon start placing Ruckus Wi-Fi systems in public places, where they could help rescue strained 3G networks by taking on much of the data traffic now squeezing through the 3G channel. That’s called “offloading,” and it’s one of the ways harried operators—whose networks now support unanticipated numbers of 3G devices like iPhones, iPads, and Android phones—hope to survive the interim years before the arrival of true 4G technologies, which could offer Wi-Fi-like speeds across much greater distances.
PCCW, Hong Kong’s leading telecom company, has been installing Ruckus Wi-Fi routers in phone booths, which tend to be located in places where there are also lots of people using cell phones. “Pedestrian gathering places are good places to put Wi-Fi access points,” Ruckus CEO Selina Lo says. “PCCW told me that in peak areas, at peak times, they can see as much as 20 percent offload.” If AT&T were able to hand off that much of its 3G traffic to Wi-Fi, it might be able to mollify many of the iPhone owners who are eager to switch to Verizon, on the (iffy) assumption that the other carrier’s network will have more capacity.
Incubated six years ago in the offices of Sequoia Capital in Menlo Park, CA, Ruckus has collected $51 million in venture backing, from Sequoia as well as Firelake Capital, Focus Ventures, Investor AB, Motorola Ventures, Sutter Hill Ventures, and WK Technology Fund. It’s a scrappy and persistent company—the barking dog in its logo isn’t wholly whimsical—and it has reinvented its product line at least twice over the years to adapt to market trends. But the 278-employee startup, which flirted earlier this year with the idea of going public but put the idea on hold due to the sluggish economy, is now under pressure to find a major market for its smart Wi-Fi access points. So there’s a lot riding on the 3G offloading concept—an idea Ruckus adopted from one of its own customers in India, Mumbai-based Tikona Digital Networks (more on that below).
Lo argues that even after 4G technologies like Verizon’s LTE standard take hold, there will be a need for other shorter-range technologies to fill in the inevitable gaps in coverage. “Licensed spectrum is an expensive resource, so there is always going to be a place for Wi-Fi,” she says. “Over time, making the network as stable as a wire is really critical, and you’ve seen how carriers that have networks that are not stable get embarrassing PR.”
Ruckus’s founders William Kish and Victor Shtrom originally wanted to build what David Callisch, the company’s vice president of marketing, calls a “God box”—a multimedia Internet hub for the home that could handle audio, video, and other kinds of digital data. They knew the box would have to have a radio, to send data to various appliances around homes without forcing residents to snake Ethernet cables through their walls. But at the time, Wi-Fi wasn’t really up to the task of carrying streaming video. So they started to think about how to make a radio connection behave more like a wire.
“The fundamental challenge with radio communications is trying to get energy from Point A to Point B,” says Kish, an ArrayComm veteran who is still with Ruckus as chief technology officer. (Shtrom is chief wireless architect.) “Traditional Wi-Fi networks use antennas that are like light bulbs, radiating energy 360 degrees, even though at any given instant they are only communicating with a client in one direction. It’s a very inefficient use of the unlicensed spectrum. What we do is more like a flashlight, focusing only in the direction we want to communicate, which puts more of the energy on the target.”
The steerable Wi-Fi idea became so absorbing that the company abandoned its God box idea and decided to become a builder of better Wi-Fi access points. The first models contained arrays of fixed directional antennas that could be switched on and off independently. Over time, Kish and Shtrom figured out how to do the same job using fewer antennas, by placing what Kish calls “parasitic elements” around a primary antenna. These blade- or fin-like structures turn on and off and act as reflectors and absorbers, steering a signal wherever it needs to go. The whole operation is under the control of software that makes hundreds of adjustments every second, constantly searching for the transmission path with the lowest packet loss (a measure of how much data is getting through to the target device).
Selina Lo met the Ruckus team about nine months in, after they’d created a demo system that could send TV signals across a room. A Hong Kong native, Lo had built up a deep background in networking at Centillion Networks, a maker of Ethernet switches, and Alteon WebSystems, which she had helped sell to Nortelin 2000 for $7.8 billion. She’d also built up a reputation as one of Silicon Valley’s most demanding and difficult executives. According to a 2009 Inc. magazine profile, she was accustomed to imposing her will through “yelling, fist pounding, and stomach-curdling sarcasm.” Callisch, a subordinate at Alteon, had even asked for a different boss, telling Lo, “You cause me huge amounts of anxiety.”
After becoming Ruckus’s CEO (and an early investor), Lo underwent a “personality makeover,” according to the Inc. profile. After realizing that stressed employees aren’t always productive employees, she set up group decision-making processes, decided she didn’t have to win every argument, and cultivated patience through new hobbies like gardening. She even got Callisch to come along with her to Ruckus.
Lo’s new patience must have come in handy when the first market for Ruckus’s Wi-Fi access points—large telecom companies in Europe and Asia, who would in turn rent the equipment to home broadband subscribers—turned out to be a harder nut to crack than anyone expected. Ruckus had built a carrier-class device, but the carriers wanted to pay consumer prices for them.
“It was a zero sum game,” says Seamus Hennessy, Ruckus’s chief financial officer. “We realized about four years ago that we had a nice core technology and a profitable business, but it was never going to be a big business.”
So Ruckus went after an additional market: mid-sized organizations like hospital, hotels, and schools that needed dependable Wi-Fi networks, but weren’t being served by existing manufacturers. “At the high end of the market you had Cisco, Aruba, Airspace, Meru, Trapeze, and everybody going after the Fortune 500,” says Callisch. “And at the low end you had Belkin, Linksys, and Netgear, who didn’t build anything worth deploying. There was a chasm in the middle—people who didn’t have a crew of 15 IT guys, who just needed something that was easy to deploy, low cost, and reliable.”
Around 2007, Ruckus started selling access points to hotels. That was good timing, since that’s exactly when more guests started showing up with Wi-Fi-capable mobile devices like the iPhone, Callisch says. Soon Ruckus had built up a network of 1,500 resellers purchasing equipment and reselling it to hotels and other institutions. The steady demand from these vendors helped to balance out the company’s revenues from telecom carriers, which had been highly bumpy, with “a multimillion dollar order every other quarter,” says Callisch. But neither business was spectacular: Callisch calls them merely “going concerns.”
That’s when Tikona came along. “When they first came to us, we had no idea who they were,” says Callisch. “They said, ‘We want to offer broadband to residences and small businesses in India, and we are going to take your products and put them on buildings and mesh them into zones and start offering service at 2 to 5 megabits per second.’ And we said, ‘Why would you do that?’ And they said, ‘Because fixed lines are sparse, and 3G is too expensive, and we don’t have the money to license new spectrum.’”
That was in early 2009. Since then, Tikona has purchased 32,000 outdoor access points from Ruckus and scattered them across 30 cities in India.
“When that happened, a light went on in our heads,” says Callisch. “We started looking a lot more closely at the carrier market, and we found that because of all these dual-mode devices—the iPhone and the iPad and BlackBerry and Android—people want fast connectivity, and 3G wasn’t going to suffice. There were going to be different kinds of carriers, some just wanting to take data off other networks.” Then traditional wireless carriers started to reach out to Ruckus, saying they wanted to build concentrated Wi-Fi networks that would plug into their existing cellular networks and take over in the areas with the densest 3G traffic. “That launched us into the whole mobile-operator space, and that is where our focus is going forward,” says Callisch.
It was precisely to offload traffic from its overburdened 3G network that AT&T bought Wi-Fi hotspot operator Wayport in 2008, Lo points out. “It’s a development that is really going on worldwide in all the markets where 3G is common,” she says.
But how would carrier-operated Wi-Fi work at a scale larger than a coffee shop hotspot? And hasn’t this all been tried before in the form of the municipal Wi-Fi networks built by many cities in the mid-2000s—which are now generally seen as a total flop?
“There were quite a number of reasons why municipal Wi-Fi was a failure,” says Kish. “During this whole municipal thing, the laptop was the thing that had Wi-Fi built-in. There weren’t iPhones or Google phones. And people generally don’t walk around with their laptops [meaning there wasn't much demand]. Secondly, these networks were all built with 802.11g, which is limited to 54 megabits per second, and when you start meshing it you get a lot less than 54. With 802.11n [the latest Wi-Fi standard], it’s a whole different story. It’s got five to six times more bandwidth than 11g. And now that you have all these dual-mode phones that people carry with them, higher speed is something that people want.”
The big idea behind 3G offloading would be to place Ruckus devices in crowded public places—think malls, train stations, stadiums, and plazas. Since Wi-Fi access points can’t serve more than one user at any given moment, they rotate through all of the users trying to access them, sending them a few bursts of data at a time. One advantage of Ruckus’s devices is that they’re able remember the last known location of each user. “If we have 100 users on an access point, we have a table in memory that says, ‘For user 45, the last time we talked to him, the best antenna pattern was 378,’ and when it’s his turn again we’ll pull out that pattern,” Kish explains. Basically, the devices redirect their beams hundreds of times each second, like a laser show in a nightclub. And that’s how the Ruckus technology could serve hundreds or thousands of mobile subscribers as they pass through public places.
But none of this going to be put in place overnight. For one thing, for 3G offloading to work, phones would have to be reprogrammed to identify and authenticate themselves to local access points, and to switch over to Wi-Fi automatically when it’s available without requiring the user to enter a password. But carriers are working behind the scenes on protocols that would make all that happen, Kish says.
In addition, carriers would actually have to deploy Wi-Fi gear in thousands of public locations. Kish says that one of the two major U.S. builders of cellular towers—it’s either American Tower or Crown Castle International, he can’t reveal which—has already installed Ruckus access points on a handful of its 20,000 towers. “We joke ’10 down, 19,990 to go,’” says Kish.
The long-term fix for 3G’s problems, of course, is 4G—and networks like Verizon’s LTE, which is faster than Wi-Fi under some conditions, are being deployed right now in dozens of U.S. cities. But Kish and Lo say they’re not worried that 4G might derail Ruckus’s new business before it even builds up steam. If the history of 3G is any guide, after all, the new LTE networks will be overloaded from the moment they’re turned on.
“Wi-Fi has moved from being a band-aid to being a strategic asset,” Kish says. “Based on the perspectives we are getting from a majority of the mobile carriers, Wi-Fi will play a role indefinitely, side by side with LTE.” For example, teams deploying LTE are turning to Ruckus for help building wireless backhaul networks to get data from the Internet backbone to LTE antennas, Kish says.
The question driving Ruckus has always been “how do you make wireless as reliable as a wire,” says Lo. “If an e-mail arrives a few seconds late, you don’t notice, but with video, every dropped packet is very apparent. I don’t want to point fingers at the carriers, but people are aware of the quality problems of the 3G networks—and the air is going to get dirtier, not cleaner. So the need for a technology like ours is going to be bigger and bigger over time.”