Your browser is not supported. Please update it.

20 December 2006

Ruckus hopes that video stable 802.11n will open up the US

Ruckus Wireless is planning to show of its next generation of smart Wi-Fi system at the Consumer Electronics show in early January based around 802.11n technology and designed to send video around a home.

Speaking to Ruckus this week the company says this should raise the bar for reliable video delivery around the home from its current 20 Mbps, up to around 60 Mbps.

We asked European technology director Scott Reeves about the product this week and he told us a little about how it works, ‘Conventional wi-fi is very susceptible to interference from other device and from obstacles in the home. Wi-Fi loses 10% to 15% of all the IP packets sent over it as video’

‘Ruckus systems have 6 directional antenna built in and the system is constantly dynamically searching for the best route around a home. There are 63 different discreet patterns that our wireless gateway can select,’ Reeves said.

But that’s not the end of the cleverness. In order for the system to know if it is getting interference in the ‘beams’ as the discreet patterns are called, Ruckus systems have a feedback loop built into them and an ability to re-send lost packets. Even with its patented beamforming approach, Ruckus wi-fi still loses the odd packet, so the packet retransmission is essential as video is highly susceptible to packet error.

A single lost internet packet in the reference frame of a codec can mean seconds of distorted video, and given that the wide installed base of set tops has about a 200 millisecond buffer for video data, any packet re-transmission has to be called for across the home network, in substantially less than that time frame. The Ruckus equipment doesn’t check the video for packets that have been lost between the head end and the home, which are far more rare than when traveling over wi-fi, but it can ensure that there are no further lost packets in the journey around the home.

This is not as easy as it sounds and digging beneath the technology issues gives us an idea of why Ruckus has signed major European IPTV customers at Belgacom in Belgium and at Cesky Telecom, now called Telefonica O2 Czech Republic, and why it has trials at 30 other operators around Europe.

We have witnessed demonstrations of TV over Wi-Fi for the past four years, but most of them cannot make it out of the demo environment to the field.

First off there is the small matter of multicast video streams arriving at the home gateway. Multicast uses a type D internet address, and many set tops will ignore traffic sent to them on their own IP address unless it is a multi-task, this is primarily for security reasons. What the Ruckus gear has to do is ‘snoop’ each packet, which means take a sample of it, and find out if the Internet Group Management Protocol, which defines a multicast, is being used. If it is it can attach ‘TV class’ importance to that packet, which means it stores if in case there are any packet losses and holds it long enough to be sure no TV set tops are going to request it.

Any packets that are lost are then unicast to the set top box at the wi-fi layer and an 802.11 acknowledgment is sent back to confirm it has arrived.

The Ruckus system also has what it calls SmartCast, a proprietary way of adding a Quality of Service protocol to an existing Wi-Fi chip. Early Wi-Fi standards didn’t have any form of QoS which can prioritize traffic, but this not only prioritizes traffic based on operator determined criteria, but it can snoop the ‘type’ of traffic it is, voice or video or html and take a best guess at what kind of priority it ‘should’ have been given even if an operator hasn’t labeled it.

Ruckus has patents in what it calls beamforming, the process by which the antennas shape the radio signals, and also the way it handles multicast re-transmission.

Ruckus successes so far have been in Europe, where High Definition TV services are less prevalent than the US, but if the new system passes muster and can deliver 60 Mbps of stable video around a home, then the US can be back on its agenda.

Already Ruckus has had trials with Microsoft IPTV High Definition systems in Europe and says that it this requires around 11 Mbps for a single feed. With that requirement its existing system could only deliver a single HD feed plus maybe one Standard Definition channel. Its software and antenna simply need to be tuned to standard parts that use the 802.11n standard and it should be closer to managing three concurrent HD streams.

If it could retrofit these to the major tier 1 IPTV efforts, it could be on the verge of a successful move back to its origins in the US.

The only working Microsoft HD systems in Europe are all owned by Deutsche Telekom, in Germany, at tiny Slovak Telekom which is a 51% owned subsidiary of Deutsche Telekom and at T-Online Hungary, also a subsidiary that used to be part of Magyar Telekom.

If Ruckus can integrate to Microsoft IPTV and convince Alcatel of its HD pedigree, then that opens up AT&T, BT, Deutsche Telekom, Swisscom, TDC in Denmark, T-Online France and potentially Verizon FiOS TV, as potential targets, along with BellSouth, Bell Canada, Telkom South Africa and Reliance Infocomm in India.

This is perhaps why the T-Online Venture Fund was one of the investors in the last round of Ruckus, and why Motorola, one of the core set top makers for the AT&T contract, is also an investor.

If Motorola decides that Ruckus is man enough to compete with the wired MoCA coaxial networking that Motorola currently uses, then it could be a breakthrough for Ruckus. Motorola is also shipping MoCA set tops to Verizon.

But the key to Ruckus, according to Reeves is that by enabling QoS, at the same time as being first to market with reliable wireless video, it can position itself as an operator supplied triple play component, complete with the TR069 terminal activation and test tools that are bundled with Ruckus and which are now standard in the DSL market.