Boffins teach old radios new channel-hopping tricks
Smart comms for dumb tech
ITU Telecom World Alongside self-congratulatory presentations on how telecommunications is improving life on Earth, ITU Telecom World also hosts detailed technical discussions about making use of new techniques, including the developing field of cognitive radio.
Cognitive radios are supposed to be able to make use of spectrum that is fallow locally, scanning for an empty frequencies while dodging existing users. Sadly that dream just isn't possible: neither the best technology nor the laws of physics allow such a thing, but careful deployment of even limited intelligence is still valuable as two of the presentations to the ITU demonstrate.
White Space devices (unlicensed and operating in unused TV broadcast bands) are expected to be reasonably intelligent once they start being deployed next year. The devices will check with an online database for locally available frequencies, while watching out for other White Space devices, which are free to generate interference and cause momentary blackouts. But for the existing users of White Space – the wireless microphone crowd – momentary blackouts aren't acceptable, and once interference is experienced then it's too late, which is why PMSE (Programme Makers and Special Events) get a few dedicated channels for their exclusive use. But Luis Correia of the Technical University of Lisbon has been working on a scheme to provide 100 per cent reliability in shared White Spaces.
The idea is to have sensors spread around a venue, constantly monitoring the available channels (a list of which comes from the White Space database), wireless microphones being used on stage can then be notified before interference builds up enough to impact with the audio stream, allowing the microphones to switch channels without interruption. An audience member suddenly switching on an impolite White Space device might still cause problems, necessitating an immediate jump, but any brought into the venue or building up from outside could be detected and avoided without affecting the performance.
Wireless mics can only stand a few milliseconds of latency, which still largely excludes the digital signal processing necessary to make a more robust signal, but even if digital signalling does become possible, the consistent availability of clean spectrum would be very valuable.
Correia showed a working prototype of the sensors the team intends to fit around the Berlin conference centre next year, in order to be able to test the system properly, but modelling seems to indicate that 100 per cent availability is possible using unregulated White Space spectrum.
Also carefully modelled are Malte Schellmann's ideas for dynamic spectrum refarming, being developed for Huawei. The idea is to avoid a big-bang approach when migrating customers from 3G to 4G, and instead switch between the two technologies on a second-by-second basis depending on the quantity of devices requesting each generation of telephony.
It wouldn't be possible for a whole block to be dynamically switched, but in his presentation Schellmann demonstrated that if one had a 5MHz channel dedicated to 3G, and another 5MHz dedicated to 4G, then one could then have a third 5MHz channel that can be dynamically switched between technologies depending on the current loading. One can muck about with the numbers, but the premise remains the same: dedicated channels are necessary to ensure customers don't get checked off entirely, but a middle block of spectrum can be passed back and forth on demand.
What's interesting about both these projects is that neither of them achieves the dream of software-defined radios intelligently scanning the radio spectrum for empty space while politely avoiding their peers. Instead they both apply the limited intelligence we do have available to make better use of radio spectrum, and show how even a little bit of intelligence can go a long way if effectively applied. ®
Some good stuff there people are smoking...
...to think that live event techs will accept their shows being run on shared frequencies, especially when the additional unreliability comes in a package with a huge increase in price and complexity, not to mention a mandatory network with Internet access.
This article doesn't really make any sense from a technical perspective. If someone would bother to write the software, the USRP would easily be able to jump between bands - even entire frequency segments. I use a USRP1 for running an OpenBTS tower, and when it's not doing that, it's running as a high-power FM radio transmitter. Simply close one program, switch the aerial outputs, open the other, and it's all good. That's the whole point of an SDR - it's meant to be able to run on any band, and that band is SOFTWARE defined. <_<
Although in the context of this article, good luck using a USRP for something like wireless mics. The mics are what, $200 USD? The USRP is around $1200 USD.
"Although in the context of this article, good luck using a USRP for something like wireless mics. The mics are what, $200 USD? The USRP is around $1200 USD"
That's sort of the problem.
With enough volume in a product hard wired *always* wins in terms of complexity (fewer metaphorical knobs to twiddle), simplicity (works or does not work) and cost (high volume).
It's also likely to be *vastly* more power efficient.
Now if you want *flexibility* that's a different game.