As Ruckus Wireless CEO Selina Lo walked out onto the sidewalk when we spoke to her last week she said, ‘Remember, UltraWideBand is dead.’ That mantra signals a war that is bound to develop between 802.11n and real UWB technologies, but not yet for a while we think.
We had already gone over the subject of how in our view the WiMEDIA Alliance version of UWB might be ‘compromised’ but perhaps not quite dead, and explained the Pulse~Link recent shock success in tests against WiMEDIA products that had the best of that bunch offering 50 Mbps, versus Pulse~Link’s 890 Mbps, using genuine carrier free UWB technology. ‘It doesn’t scale she said,’ but of course that only applies to the WiMEDIA version, the Pulse~Link systems scale quite nicely.
Lo didn’t miss stride, and continued to firmly bat in the corner of Wi-Fi over every other wireless technology inside the home and enterprise, and rolled out her pre-planned presentation on a shift in the direction of the so far impressive Ruckus.
‘The company was incubated at Sequoia Capital and I was drawn to it by my own experiences of trying to get a TV signal around my home. The cable systems wouldn’t reach and a satellite system needed guys climbing all over my condo, what I needed was a wireless system and then I remembered that Sequoia had one. I asked about the project, met the founders, co-invested, and took the CEO role. We imagine that she hasn’t had time to watch that TV system at home since, as out of nowhere Lo took the company out of the US into Europe and the Far East and signed up well over 100 trials for the Ruckus Wireless system which promised to carry TV signals around a home using Wi-Fi.
‘About 95% of those early trials either are already or are planning to become customers,’ Lo boasted, and said that to date the company has shipped 500,000 units. The Ruckus website only owns up to 150,000, but perhaps that’s because the company has been doubling or trebling shipments and revenues for the past few years and the web site can’t keep up. One of the reasons for this has been Lo’s idea of taking the product to market via operators.
IPTV operators were the first port of call and Italy’s Fastweb, PCCW in Hong Kong, SingTel in Singapore, Belgacom in Belgium, Portugal Telecom, Pioneer and MTA Communications in the United States are all examples of IPTV systems which are happy to offer Ruckus as a way of sending the TV around the home.
‘This is done in a variety of business models,’ says Lo, ‘some will bundle it with the service like Telef’nica del Sur in Chile, but most will offer it as a paid option or a slightly subsidized one.’
One of the Ruckus customers is the advanced quad player PCCW which has installed 3,000 hotspots all over Hong Kong which it uses to extend the reach of its Now TV channels which run on IPTV, mobile TV and over the web through a hotspot, to either a PC or a handset.
Ruckus has also had some significant wins among hospitality IT companies such as Maginet and Quadriga, which take care of the Wi-Fi for many of the world’s largest hotel chains.
The early BeamFlex products were built around the 802.11g standard and it renamed the range to MediaFLex when it introduced the HD capable 802.11n product at the beginning of 2007. Before the HD product Ruckus couldn’t really get buy-in in its home country of the US, so it has developed its market back to front, overseas first, then taking on the US.
The difference between usual Wi-Fi systems is all on the antenna and the software. Ruckus systems have 6 directional antennas built in and the system is constantly dynamically searching for the best route around a home with up to 63 different discreet patterns that the wireless gateway can select picking through different channels and directional routes, using multipath effects of the OFDM signal to try out and construct new ways around interference. Much of this is already handed for it by the Atheros chips that it uses, and Atheros has become a highly standard way around this problem in other products.
Without some form of MIMO or dynamic reconfiguration Wi-Fi will drop around 15% of its packets from room to room in a home. That will mean that TV will break up into pixel blocks, as the key reference I-frames get damaged by late or non-arriving packets. Atheros adds the MIMO capability, and some of the basic re-configuration tools, and Ruckus has added the software stack on top that also ropes in the idea of holding a buffer of the video it has sent, and resending any lost packets and asking for feedback from the device on how well the current channels are working.
In effect Ruckus has made Wi-Fi behave in a predictable fashion, but those Atheros chips are now turning up everywhere. Apple TV relies on one in the Apple Airport Extreme that also uses a Wi-Fi N chip, and there are now a swathe of imitators entering the market.
They may not be quite so effective as the Ruckus system, and they may only be retail products that have to be installed and managed by the consumer, but if all that Ruckus retains as an advantage is being first to market, and holding the operator’s hand a little, how long will it retain its edge?
And that’s just the point of Lo’s whistle-stop tour to promote the new direction of Ruckus. She has three shiny new products to show, the N based Zoneflex Wi-Fi access point for enterprises and hotspots, the MediaFlex 802.11g router for the home, which can also be used as a hotspot from the home, and its FlexMaster global Wi-Fi management platform.
‘We have already made Wi-Fi predictable, now we want to make it manageable,’ said Lo. The FlexMaster suit relies initially on yet another widespread standard, the TR 069 from the DSL Forum, which is a SOAP/HTTP protocol which provides for two way communication between any broadband based CPE and servers which can automatically configure them. But when you combine this with the operator relationships that Lo has built and provide higher level software for plugging it into existing operational support systems and network management layers, it means that operators can virtually forget about Wi-Fi router management. The device is installed, it wakes up and its factory settings tell it to look for the nearest manager server, it finds it, downloads the right configuration and begins working.
This is pretty smart for major enterprises, but it’s absolutely essential for the less smart enterprises like hotel and hospitality based hotspot chains, especially where some of them need capabilities like mesh or self backhaul configured into these wireless networks, and especially where those same networks find themselves being used for carrying video. Lo says that the management platform adds fault detection, performance management and reporting to the basic TR 069.
The new N based Zoneflex model is about carrier class Wi-Fi, while the MediaFLex is a G class product which bundles VPN protection and multiple SSIDs into a Wi-Fi device that extends that proposition out of the home, offering spare capacity to people passing the home without the worry of mixing up all the traffic.
All of this brings us back neatly to the where the company is going and why Lo dislikes the notion of UWB. Any shift in the perception that Wi-Fi is the technology of choice capable of carrying HD TV signals around the home, undermines her one key strength, that simple hand holding relationship with the operators. But while they are fixated on dual mode cellular devices, femto cells and bringing Wi-Fi and VoIP to the handset, she needn’t worry. By the time more than a handful of operators need to send not one or two HD signals around the home, but four or five and who also want to send uncompressed video point to point, and reach for UWB, Ruckus will already be generating multiple $100s millions in annual revenues. Which is just about the size of company that Lo has run before prior to exit through acquisition.
When we talked to the company before we noted that Motorola was one of its trade investors, along with T-Online and thought that it would be a nice property to add to Motorola’s home network and set top product lines. Now that the company’s reach is further into the operator market, and Motorola on the brink perhaps of distancing itself from handsets, we see even more reasons for Motorola to eventually buy Ruckus.