Possible low-cost DAB platform for small stations

[Guest]

I've just had this brief technical note from Ofcom which might be of interest

An experimental DAB multiplex was set up in central Brighton between 14th September 2012 and 27th January 2013 using new lower-cost technologies. The experiment was privately-funded, and the research conducted in the author's spare time. As no similar study had previously been carried out, the test was a valuable exercise to inform policy makers of the practicalities of low cost DAB solutions when used to serve small areas, particularly from a single transmitter. The transmissions were operated under a Non-Operational Test and Development licence issued by Ofcom.

As well as testing the viability and reliability of new wireless techniques for generating a Eureka 147 DAB compliant signal in a real world scenario and it also allowed various measurements, and comparisons with existing multiplexes to be made. The experiment successfully demonstrated that much of the infrastructure can now easily be implemented in software, and that integration with public IP networks improves accessibility while reducing the capital and operating costs quite dramatically. It was also demonstrated that a trade-off in power amplifier selection could remove the need for large and costly cavity filters to meet ITU spectral masks. The trade-off is a reduction in power efficiency, but the cost of this is negligible at low power levels. Some thought was also applied as to how a small-scale multiplex could fit in with the current and future radio broadcasting spectrum landscapes. It was also shown that a low-cost, low power approach could deliver a reliable, high quality service at minimal opportunity cost by using interleaved spectrum, which is unsuitable for use by larger networks.

The experiment also highlighted the importance of 'site over might'. Low power transmitters sited in urban population centres can often deliver the field strengths required for reliable indoor reception much more effectively than might be achieved with a higher-powered site on the periphery of the population centre.

Adjacent Channel Interference (ACI) to other DAB services was anticipated and thoroughly checked for, but none was found. It seems that ACI 'holes' are not created by low power DAB transmitters sited in an area where the wanted signal levels from other multiplex services transmitted from elsewhere are sufficiently high.

This work demonstrated that it is feasible to deliver DAB transmission infrastructure at much lower cost than currently required for equipment to deliver wide-area coverage. Nevertheless, significant further work is required to identify suitable spectrum for services making use of these technologies. In addition, it will be necessary to consider how they might be licensed to cover particular areas, especially in circumstances where there is a requirement to carry more than one service on the multiplex. When these issues have been resolved, these new techniques could find particular application for Community Radio or smaller scale commercial radio stations. These techniques might also assist in rolling out existing networks to serve more remote population areas where existing approaches might not prove to be cost-effective.

This report is intentionally light on engineering details (although it does use many technical terms) as the goal is to provide an outline of new concepts that can be understood by an audience with varying depths of understanding in technical matters. Although a wide range of different products and tools is given, their inclusion as a reference in this document should not be taken as an endorsement of these products and tools.

[Gordon]

Thanks AIC for that, an interesting report done as a personal special by the look of it!! Certainly on a shoestrng budget.

http://stakeholders.ofcom.org.uk/market-data-research/other/radio-research/Software-DAB-Research/

Software Defined Radio [SDR] has been coming along for some time but it looks as if, with software developers becoming more engaged, an economic cellular style DAB radio network can be envisaged. One application could well be the extension of existing DAB networks at low cost which would mean operating these devices in SFNs. A test of this would be very useful - the equipment clearly has the tools avaialble.

More generally however, as usual, the availability of spectrum is an issue which has yet to be addressed as are the licensing conditions and service areas etc that would have to be addressed. I didn't see anything about this approach in the OfCom report to government published May 2012, annexe B, that covered options for coverage planning for radio networks. This trial seemed to use a different frequency from the local transmissions rather than on SFNs with them - perhaps this was a licence condition for the trial - so that part of the test was inconclusive. Good piece of work though that should be taken further.

[Nevalti]

.....It was also shown that a low-cost, low power approach could deliver a reliable, high quality service at minimal opportunity cost by using interleaved sp.....

That claim, to deliver a 'high quality service', is the bit that worries me. The BBC have, quite rightly, stopped claiming that DAB is 'high quality' but here we have Ofcom making the same old, misleading claim. Unfortunately, the Advertising Standards Authority can do nothing about it as they are not adverts.

[OldTechie]

That claim, to deliver a 'high quality service', is the bit that worries me. The BBC have, quite rightly, stopped claiming that DAB is 'high quality' but here we have Ofcom making the same old, misleading claim. Unfortunately, the Advertising Standards Authority can do nothing about it as they are not adverts.

The report contains some observations about quality:

The quality achieved in DAB transmissions will depend on the sampling rate and bitrate used to code the services and also the technical quality of the encoder. An interesting finding made through listening tests in this trial was that some services currently on DAB seemed to have subjectively poorer sound quality than might be expected for the bitrate used. The services were checked against two different benchmark coders - the test bed and the control sources were checked several times for anomalies, but none were found. There are a few possible explanations for this, including non-linear programme source material or poorly implemented audio pre-processing...
AND
Although there is no substitute for delivering ‘high quality’ by using appropriate bitrates, it is apparent that there is room for some improvement in the subjective sound quality of some services currently on DAB in the UK by addressing all of the factors described above.

However, what is meant by a "high quality service" is a bit context sensitive (and the quotes round the words in the second para above indicate that the author meant it in context and knew exactly what he meant.) In this context the report is referring to the quality of a local service compared to existing DAB services, so I think its not that unreasonable. In fact, because he was using the open source encoder he was hoping for better sound quality and suggesting that the approach might result in quality improvements resulting from open source efforts in this area.

The wording that annoys me is "digital switchover" which implies that turning off the current analogue services will facilitate improvement of the digital services as it did for TV. But they plan to add local services instead of the existing FM services. This report might suggest that a better way to provide those local services is with local DAB. I do wonder whether there is any call for the local services given that all the slightly larger area commercial radio services seem to give up and take national feeds at some times.

[Nevalti]

......I do wonder whether there is any call for the local services.......

Apart from being interrupted by RDS traffic information, I never use local radio. Presumably some people do but I doubt that many would be seriously disappointed if ALL the local BBC stations closed down.

We can't make too much of a fuss as, according to RAJAR, 4.5 times as many people listen to local radio as those listening to R3. I'll have to concede that minority interests are catered for.

Classic FM has almost 3 times the audience of R3! I wonder why that is? Dumbing down? Compression? Less 'arty' talk? Less opera?