ZigBee – The everywhere PAN

In the current flurry of new wireless networking, we have a new player on the block – ZigBee, the new PAN (Personal Area Network) protocol.

It’s been designed to be very low in power consumption and for most of the applications being targeted, its 2 AA batteries should have a life of 6 months to 2 years, but that does vary by application. One of its secrets is that the devices will only communicate when their application deems necessary.

With it’s maximum data rate is 250kbps, compared to 1 Mbps for Bluetooth but as we know, modems speeds are around a fifth of this, so it’s not painfully slow and many applications for devices don’t need loads of bandwidth.

Range is expected to be ~30 meters in a typical home, compared to ~10 meters for unamplified Bluetooth products.

ZigBee networking capabilities include 254 devices per network, compared to 8 for Bluetooth networks. There can also be up to 100 co-located networks.

With its small stack size (28Kbytes), which is about 10th of the Bluetooth – so the computing spec required to run it will be lower.

The entire bill of materials for a radio module is expected to initially be $6.00, coming quickly down to $2.00 to $3.00. Significantly lower than any other wireless network technology.

Products are expected to be available in middle of 2003 and it’s projected that over 400 million units will be produced per year by 2006.

Low cost wireless control within the home has a chance of becomes a reality.

Innovation … from BT?

I was more than surprised to see some ‘connected thinking’ (excuse the management speak) from BT today.

I’ve always said that one of the barriers to the general public taking up of broadband is that fact that they’ve never seen it, and why would you want something that you’ve never see. The same applied to the early days of Sky’s satellite delivered programming. Additionally the terms that are used to promote broadband, such as always-on are pretty meaningless to people with no experience or technology background.

BT payphones has decided, for a limited period, to give FREE broadband access at their new blue kiosks which are scattered around some sixty nine high streets around the UK.

Sky and BT strengthen ADSL relationship

Sky and BT have had a strategic relationship for a long time. Initially it was because Sky needed a return path for their STB’s and back then the only realistic provider of this was a BT phone line. Things went a little frosty a while back when there were rumours that BT might use their network to become a broadcaster but it’s now all lovey-dovey again.

Back in February this year, BT made a know for a while that Sky has had a DSL connected STB/DVR, so clearly a wider take-up for broadband could benefit them in delivering content.

  • Sky customers who subscribe to cheaper packages but will see an advantage for upgrading to get the deal.
  • An unknown is the amount of money that Sky will be paid for each of their subscribers that sign up for the service hasn’t been disclosed.
  • I would also assume that BT has had to give legal assurances that it will not use the ADSL lines to provide content that will compete again Sky.
  • BT will get to strengthen their hold on UK broadband subscribers.

    Quite a good deal to Sky then at first glance.

    MPEG-4 Advanced Video Coding

    A new video encoding system is getting close to approval. H.264 or MPEG-4 Advanced Video Coding as it’s otherwise known, is thought to be able to reach DVD-quality broadcasts at slightly under 1Mbps. Some are sceptical about it being taken up for STB’s, as it will require about three to four times the computing horsepower needed for MPEG-2 – up to now pretty low power chips have been installed in STB’s. It is thought that the proposal should be ratified as part of the MPEG-4 (Moving Picture Experts Group) multimedia standard by year’s end.

    The MS comment was that their proprietary CODEC was much more efficient – but in true MS style, the leader of the overseeing committee is an employee of Microsoft.

    Aggressive wireless rollout in Singapore

    SingTel currently have 100 hotspots setup. Access will be charged at 20 cents per-minute with no subscription to pay and GPRS-enabled phones will also be able to use the hotpots as well as WiFi.

    Singapore is a pretty small place but as it’s configured the offering doesn’t sound that mainstream, with a maximum of only 3,000 being able to use the service by the end of the year when it will reach 150 hotspots. I assume as demand increases they will increase the number and density of the hotspots.

    Denham leads by Broadband example

    Another UK local community is setting up their own wireless network – and hats off to them.

    96 households in Denham, Buckinghamshire had already registered their interest with BT in ADSL-enabling their local exchange but as BT the minimum required that had been set was 450, they fell well short.

    Not seeing this as the end of the boad; two of them decided to set up Denham Broadband to provide wireless access around the area. Subscribers will receive ADSL-like speeds via a small, rooftop antenna, all for an estimated setup cost of 150 UKP plus 30UKP a month.

    It’s going to be interesting to see if this people-power action will mean that other will go to these lengths to live The Broadband Dream.

    The figures in the UK still don’t look great. In the current BT ADSL Demand Tracker only five of the 2323 listed have been converted, and of these 2323, more than ¾ of these exchanges haven’t even had a trigger level set yet.

    At the end of the table, they’ve been kind enough to not order it the by number of households signed in an exchange. To some of these people where none of their neighbours are interested in ADSL like the Bod in Onich (North west of Glasgow, Scotland), The Broadband Dream must feel a long way off.

    I tell you what would be interesting – installing a large number of Denham Broadband-type community networks across the country, as consume.net have been discussing for a long while.

    Then sell the subscribers SIP phones (normal phone handsets that can use the Internet to carry the calls) and give them the benefit of making close-to-free phone calls to other subscribers. If enough people and especially businesses subscribed, it could start to hit BT’s cashflow.

    WiFi Video at home

    Clearly one way to distribute video around the home, be it from a cable signal or household PVR, will be wirelessly, The first company I recall announcing it as a product was Moxi over eight months ago.

    The problem with this lies in the theoretical 11Mps that 802.11a offers. This could get used up pretty fast depending on the strength of signal, the number and quality of simultaneously- viewed video streams and the amount of bandwidth other users on the network are taking up.

    Recognising this, ViXS Systems have designed a chipset to tackle it. Their customers are being offered samples of its XCode video network processor, which takes in number of sources of video, be it an HD signal or an MPEG-1 or MPEG-2 stream and compresses them down. This can then distributed over any IP based network, WiFi or wireline.

    The smart part comes in the fact that the quality of the stream can be adjusted depending on available bandwidth [Acronyms alert – IDS (Intelligent Distributed Video) and the very catchy ABFM (Adaptive Bandwidth Footprint Management)].

    The people behind the RealTime Streaming Protocol standard, RTSP, had the idea a while back and companies such as Apple and Real Networks implemented this. By exchanging performance information between the sender of the video and the client, different qualities of video are delivered to the viewer. The downside of this is, if the video wasn’t originally encoded with a version for really bad network conditions (say 80k), the next closest version is shown. This may a high bandwidth version (say 220k) that, as all of the frame information cannot be received in time will display badly. The reality of producing streaming video is that you cannot practically encode a version for every network condition.

    The beauty of the Vixs approach is it will do this to any current video source, analogue or digital and, one would assume, finely adjusting quality over many different bandwidths.

    They claim to guarantee deliver 30 frames per second (FPS), high quality video over any type of Internet Protocol (IP) network, wired or wireless which is pretty impressive.

    Isn’t it interesting that, as we see the industry grow up, companies measure performance in units that start to mean something to the industry their aiming at rather than tech terms.