Does white space need to be Weightless?
A new protocol for a new wavelength
Next week will see the formal launch of the Weightless radio protocol, designed for use in the soon-to-be-available white space frequencies. But can a protocol without weight really do it all?
Next Friday (30 September), all the major companies in the UK white space business will be gathered in Cambridge to talk about which protocol should fill the airwaves where television isn't. Alongside the traditional options there's a new contender – Weightless – being proposed as a machine-to-machine standard that could equally fulfil other white space roles.
Despite the constant comparisons to Wi-Fi – encouraged by Google, Microsoft et al – and despite the fact that the FCC keeps calling it "Super Wi-Fi", white space is only similar to Wi-Fi in the manner that the M1 bears a striking resemblance to the Highway Code: that is to say, not at all.
White space is a geographically limited radio band, in which one will be permitted to transmit any radio signal one chooses (subject to power and antenna limits). Devices have to check with an online database to see what frequencies are available locally, but once cleared they are free to interfere with each other as they wish.
For that reason many devices will incorporate sensing technology to avoid other users, just as the most-modern Wi-Fi routers will find the least cluttered channel in which to operate. But that's a lot easier if the majority of users have settled on a single communications standard, as the Wi-Fi band (2.4GHz) has done – but in white space the options are still very open.
Which is what next week's meeting is all about, with the world's first manufacturer of white space kit, Neul, proposing a Special Interest Group be created to endorse its "Weightless" standard for white space communications.
The standard was developed for machine-to-machine communications, and Neul has made noises about building (or helping to build) a national network connecting up electricity meters, cars, and anything else which requires low-power connectivity, but the company reckons the standard can cope with high-speed internet access too (ramping up to 16Mb/sec in tests).
With so much hanging on the standard, and British companies leading the way, we thought it was worth taking a closer look at the wireless standard which, if adopted, could be cheap to buy and free to use within the next year or two.
The weighty matter of Weightless
Weightless operates in an 8MHz-wide channel, to fit into the slots used for broadcast TV (and will thus have to squeeze into 6MHz if used across the pond where TV is smaller). Weightless is a Time Division Duplex (TDD) protocol, so access point and clients take turns to transmit.
When the hub device checks with the national database, it supplies a location and receives a list of 8MHz slots which aren't being used to transmit TV in that location. Weightless will hop between available slots every second or so, skipping any which turn out to be too cluttered (though periodically checking back in case they've cleared).
Showing its M2M roots, a Weightless access point only pages connected devices every 15 minutes, so those devices only need power up the radio four times an hour. Neul reckons that running the radio for two seconds at such intervals results in power consumption roughly equal to the decay rate of an idle battery, so being connected (and idle) has no perceivable impact on battery life.
That means a single Weightless hub can run connections to hundreds devices, across a network spanning 10km or so. Those devices could easily have a battery life measured in years, and be capable of responding with megabytes of data within 15 minutes.
A device which wants to connect to the network won't want to wait that long, and neither will one with something to report. In such circumstances the client can pick up a transmitted frame, which comes every second or two, and register an interest in sending some data upstream.
The security side of Weightless has yet to be worked out, with mutual authentication being considered more important than encrypting the content. Having someone listening in to a meter reading isn't that important, having someone faking a reading is, and content can always be encrypted at a higher level (Weightless will happily carry IPv4 and IPv6 packets).
Once on the network, a device has to wait for the hub to say when it can talk, though it has the chance to request communication slots. The speed of transmission is dependent on the quality of the signal. Each frame is addressed in a basically encoded header; all other devices can switch off their radios once they know the frame isn't addressed to them, and if the receiving device is nearby (as established by the signal strength) then the rest of the frame can be tightly encoded in the knowledge that little will be lost en route.
That means a Weightless hub can speak to hundreds of devices on the same network, with the speed of connection varying between devices. A receiver near the hub might therefore get 10Mb/sec or better, but one operating on the same network, from the same hub, could be running at a few hundred Kb in the same timeframe.
Weightless is proposing itself as the solution to every white space application, but, once again, it is not hard to see the M2M roots. It could, even in its present form, be co-opted into other roles, but as it stands the focus is clearly on keeping low-power devices connected for periods of years rather than allowing someone to download YouTube videos at lower cost.
Weightless isn't the only contender vying to fill the white spaces. The strict out-of-band restrictions imposed by the FCC make white space unsuitable for LTE (Long Term Evolution, the 4G standard which is selling itself as ideal for every radio application at the moment) but Xg Technology is pushing into the space with a credible protocol, and there are other contenders waiting in the wings.
A global standard is a laudable aim, even if wireless internet providers decide to adopt something else for their higher-speed connections and Weightless ends up as a M2M standard. Bluetooth and Wi-Fi happily coexist in 2.4GHz, along with the occasional baby-listener and not forgetting the microwave ovens which originally rendered the band worthless, so even if Weighless doesn't become the standard for all white space communications, it is still a valuable development.
We would recommend that interested parties attend next week's inaugural meeting, but it seems the event is already oversubscribed, so we can only point readers to the waiting list and promise we'll try to keep you updated on how the protocol develops and when it gets deployed. ®
Sponsored: DevOps and continuous delivery