The BBC recently reported that Radio Russia have quietly switched off the majority of their long-wave broadcast transmitters. Whilst the silent passing of Russia's long wave service will not rattle the front pages either in Russia or anywhere else for that matter, it does raise the question of the long-term viability of long-wave broadcasting.

During the early 1990s there was a resurgence of interest in long-wave radio in the UK caused by the success of the pop station Atlantic 252. But by later in the decade, the poorer quality of long-wave broadcasting compared to FM, together with the increased proliferation of local FM services in the UK eventually led to the demise of the station. Similar logic appears to have been used by the Russian authorities who now have a much more extensive network of FM transmitters and clearly feel that the expense of operating long-wave is no longer justified.

One of the great advantages of long-wave broadcasting is the large area that can be covered from a single transmitter. For countries whose population is spread over very wide areas, long-wave offers a means to broadcast to them with very few transmitters. Conversely, the large antennas and high transmitter powers required to deliver the service make it an expensive way to reach audiences. Presumably there is a relatively simple equation that describes the cost-benefit of long-wave broadcasting, i.e.:



Where:

As long as W>0 as A(FM) increases it continues to be worthwhile to broadcast on long-wave as the cost of providing the service is greater than the cost of doing the same thing using FM.

The cost of providing an FM service - C(FM) - is not constant, and will increase with the audience served, and not in a linear fashion either. The final few audience will cost significantly more than the first few. This is because stations which only serve small, sparse communities tend to be more costly (per person) than ones serving densely packed areas.

The cost of providing the LW service - C(LW) - however, is largely constant regardless of how many people listen to it.

It's therefore possible to draw a graph of the cost per person - C/A - of the FM audience and the cost per person of the long-wave audience, as the FM audience increases.



The figures used in the graph above are illustrative only. They assume that:

• The cost per person of providing an FM service increases by a factor of 10 between the first and the last person served;
• The cost per person of providing the long-wave service is initially only a third of that of providing the same service on FM.

Based on these assumptions, it is not until the FM audience reaches almost 90% of the population that the cost per listener of FM is less than that of providing the same service on long-wave. As FM coverage becomes more widespread, it is this factor that is causing many broadcasters to cease long-wave transmissions (the BBC has a plan to end its long-wave service too, though there is no date set for the closure yet).

Of course there are many other factors to take into account, in particular the difference in service quality between FM and long-wave, and the proportion of receivers that have a long-wave function. There are thus other factors that will hasten the end of long-wave as FM coverage increases. The same could largely be said for medium-wave where arguably, the problems of night time interference make it even worse off than long-wave (though more receivers have it).

Wireless Waffle reported back in 2006 on the various organisations planning to launch long-wave services, not surprisingly none of them have (yet) come to fruition.

There is, however, one factor in favour of any country maintaining a long-wave service (or even medium-wave for that matter), and it's this: simplicity. It is possible to build a receiver for long-wave (or medium-wave) AM transmissions using nothing more than wire and coal (and a pair of headphones) as was created by prisoners of war.

Prisoners of war during WWII had to improvise from whatever bits of junk they could scrounge in order to build a radio. One type of detector used a small piece of coke, which was a derivative of coal often used in heating stoves, about the size of a pea.

After much adjusting of the point of contact on the coke and the tension of the wire, some strong stations would have been received.

If the POW was lucky enough to scrounge a variable capacitor, the set could possibly receive more frequencies.

In the event (God forbid) of a national emergency that took electricity (such as a massive solar flare), it would still be possible for governments to communicate with their citizens using simple broadcasting techniques and for citizens to receive them using simple equipment. Not so with digital broadcasting! Ironically, most long-wave transmitters use valves which are much less prone to damage from solar flares than transistors.

So whilst long-wave services are on the way out in Russia and elsewhere, it will be interesting to see whether the transmitting equipment is completely dismantled at all sites, or whether some remain for times of emergency. Of course if every long-wave transmitter is eventually turned off, there is some interesting radio spectrum available that could be re-used for something else... offers on a postcard!

It wasn't that long ago that Wireless Waffle was discussing the need for spectrum for programme making and special events (PMSE). At the time we were considering how the needs of the burgeoning demand for radio spectrum for the Eurovision Song Contest would be met. Most radiomicrophones currently operate in the UHF television band, in the gaps between transmitters. These gaps (which are there to protect neighbouring television transmitters from interfering with each other) are also being eyed by the wireless broadband and machine-to-machine community amongst others and have been given the moniker 'white space' (though one person infinitely more learned in these things believes they should correctly be considered 'grey spaces' as they aren't as white as you might believe).

There are also moves afoot to squash television broadcasting into even less spectrum to make way for more mobile broadband. At present the spectrum from 470 to 790 MHz is generally available (40 channels - channels 21 to 60). The new plans involve using the spectrum from 694 MHz upwards for more mobile broadband leaving the terrestrial television broadcasters with just 28 channels (channels 21 to 48). And at the moment, there is no guarantee that there won't be further erosion of the UHF television band for other uses.

If TV use is squashed into less spectrum, there will be less 'grey space' available for radiomicrophones, or for anyone else for that matter. To make matters worse, the tuning range of most radiomicrophones (and similar devices) is very limited and each time they are forced to change frequency, new equipment needs to be bought. Of course, this is good news for manufacturers such as Sennheiser and Shure, but is bad news for the end users.

The need for spectrum for radiomicrophones and other PMSE uses is recognised at an European level in the Radio Spectrum Policy Programme (RSPP) article 8.5 of which states:

Member States shall, in cooperation with the Commission, seek to ensure the necessary frequency bands for PMSE, in accordance with the Union's objectives to improve the integration of internal market and access to culture.

So what can be done? Are PMSE users to be left as the nomads of the radio spectrum, packing down their camps, wandering across the desert and re-assembling their tents in a new area every 3-4 years? Or is there a long(er)-term solution that would allow them to lay solid foundations and put down some bricks?

For many years, a band at 1785 - 1800 MHz has been available for wireless microphone use, but only for digital microphones (see CEPT Report 50). Almost no use has been made of the band and the views of Audio-Technica illustrate why this is the case:

The frequency range [1800 MHz] is not really suited for wireless microphones, as the higher frequencies (i.e. shorter wavelengths) create more body absorption and shadow effects due to the directivity, etc. The use of these frequencies will only work adequately when there is a line of sight and a short distance between the transmitter and the receiver.

Using diversity reception (already commonplace in radiomicrophone equipment) and careful antenna placement, there is no reason why the 1.8 GHz band could not prove useful. But one of the other problems with this band is that radiomicrophones are not well suited to using digital technology. To send audio digitally, it must first be converted from analogue to digital. For 'high quality' audio, this would yield a 'raw' data rate of at least 512 kbps, if not more - and more like 1 Mbps by the time error correction is added in. If we were to try to transmit this data in the 200 kHz channels that microphones currently use, we would have to use a high-order modulation scheme (such as 8-PSK or 16 QAM) and this causes problems because:

• transmitters need to be linear meaning they draw more power and would drain batteries much more quickly;
• higher-order modulation schemes require decent signal-to-noise levels and thus higher powered transmitters;
• it takes time to encode and decode complex modulation schemes.

It is this latter point that is perhaps the Achilles Heel of the system. The delay between words being spoken, and the sound coming out of the PA system has to be very small. If it is not, problems of lipsynch soon occur (e.g. the speakers lips are out of synch with the sound you hear). A delay of only a few 10s of milliSeconds is usually regarded as the largest tolerable.

This means that the other way in which digital systems use spectrum efficiently is also a no-go. Compressing the data (e.g. using a compressed audio format such as mp3 instead of the raw digitised audio) also takes time - generally longer than the time taken to transmit the signal digitally. And so we reach an impasse: compressing the audio to use less spectrum takes too long, and transmitting the raw data uses more spectrum than their analogue counterpart and involves a number of other trade-offs. All this means that digital radiomicrophones, whilst slowly being developed, tend to offer no better performance than analogue versions (and at much higher cost).

But the fact is, that if the radiomicrophone industry does not make some strides towards adopting higher frequencies or more spectrum efficient modulation techniques, it might find itself without enough spectrum in which to operate.

So where could microphones go? There are a whole host of frequencies which are currently assigned at a European level by CEPT for radiomicrophone use (as per ERC Recommendation 70-03, Annex 10). These include:

• 29.7 - 47 MHz - manufacturers claim that these frequencies are not ideal as they are too noisy and antennas are too large (fussy lot aren't they)
• 174 - 216 MHz - VHF band III - mostly occupied by TV broadcasting and DAB radio
• 470 - 790 MHz - the aforementioned UHF band that is now being squeezed
• 863 - 865 MHz - licence exempt and shared with other devices
• 1785 - 1805 MHz - 'too high'

The UHF band accounts for over three quarters of the available spectrum, so if it is lost, where next for the radiomicrophones? How's about:

• 1215 - 1350 MHz - mostly an aeronautical radar band but shared with many other uses and therefore presumably sharable with others
• 1350 - 1400 MHz - low capacity fixed links and some mobile services
• 1492 - 1518 MHz - more low capacity fixed links - and already proposed in ERC 70-03 but available in a tiny amount in the UK only
• 1675 - 1710 MHz - a downlink band for meteorological satellites but not heavily used - sterilisation zones around official downlink sites would protect professional users

If the big guns (Sennheiser, Shure, Audio-Technica, AKG) refuse to find a way to use higher frequencies, perhaps the time is right for one of the small guys to. You can bet your bottom Renminbi that if they don't, some enterprising Chinese company will!

With thanks to Keith over at the other Wireless Waffle for bringing this to our attention. Pirate Wogan has to be one of the best pieces of radio broadcasting since Chris Morris's various shows on BBC Radio London and Radio 1. Invented by Peter Serafinowicz the concept is a simple one. Put Terry Wogan (or T-Wog$ as he is now to be known) in front of a microphone on a fictitious pirate radio station and record the ensuing mayhem.



Not only is the result hilarious (up there with '10 things to change the world' - which you can hear on the Cook'd and Bomb'd web-site) but it's available in app format too!

The Pirate Wogan app allows you to loop various drum and bass sounds together with samples of T-Wog$ to produce your very own London pirate sound-alike. In principal you should get bored in no time, but there is something hypnotic about the combination of T-Wog$ dulcet tones and the deep dark bass throb of the music.

Big up the Wireless massive. Easy now.

It is now only a month or so before the annual pan-European musical bun fight that is the Eurovision Song Contest (which is on the 18th of May). The 39 songs that have made it to the competition (which this year is in Malmo, Sweden) are now available to listen to online - with many having professionally produced videos viewable on YouTube. As always there is a mix of complete Euro-nonsense (Greece with their song Alcohol Is Free), songs that could have been written for Eurovision any time from 1960 to the present day (Switzerland's song You And Me falls into this category) copies of last year's winner in the hope that people will want the same thing again this year (Germany's Glorious for example), sickly pop songs with no real substance (Finland's Abbaesque entry Marry Me) and this year, it seems, a whole swathe of rather nice ballads.

For what it's worth, Wireless Waffle is particularly fond of the Icelandic entry (Ég á Líf), the video for which awaits your enjoyment below. If you are of an emotionally sensitive disposition, make sure you have some tissues to hand.



Our selection of songs that ought to be in the top 5, at least if they are able to reproduce the performance in their videos on stage on the night are:

• Eyþór Ingi Gunnlaugsson - Ég á Líf (Iceland)
• Roberto Bellarosa - Love Kills (Belgium)
• Zlata Ognevich - Gravity (Ukraine)
• Despina Olympiou - An Me Thimáse (Cyprus)
• Dina Garipova - What If (Russia)

But dodgy songs aside, the contest has always been a real test of ingenuity and resourcefulness for broadcast engineers. Those who remember the show from the early 1990s will recall the 'fun' that took place trying to establish the video feeds from the countries voting, with cheers going up from the audience on occasions when the link was finally made. These days, making a video connection from one place in Europeland to another is comparatively straightforward, though dealing with the presenters egos is a whole different problem.

What has become more complex, though, is the use of wireless technology for the event. Whilst it's still normal in a concert hall to use wired cameras, the use of radiomicrophones and in-ear monitors (which allow the performer to hear themselves sing without the noise of the surrounding environment) has grown significantly over recent years. Figures provided in response to a European Commission consultation on programme making spectrum show an average annual growth of 12% in the number of such devices employed over the past 15 years (but nearly 40% growth per year over the last 4 years).

Year	Venue	Radiomicrophones used	In-Ear Monitors used
1998	Birmingham, UK	40	2
1999	Jerusalem, Israel	42	6
2000	Stockholm, Sweden	48	16
2001	Copenhagen, Denmark	48	16
2002	Tallinn, Estonia	54	16
2003	Riga, Latvia	54	16
2004	Istanbul, Turkey	54	16
2005	Kiev, Ukraine	54	16
2006	Athens, Greece	54	16
2007	Helsinki, Finland	56	16
2008	Belgrade, Serbia	56	16
2009	Moscow, Russia	56	16
2010	Oslo, Norway	72	32
2011	Düsseldorf, Germany	82	40
2012	Baku, Azerbaijan	104	80

What is driving this growth? One the one hand, the equipment costs are coming down making it more affordable to use wireless microphones. On the other, it's also probably true that in earlier years, microphones would be passed from one user to another as they went on/off stage whereas now each artist and band has their own set of equipment which is tailored to their own needs.



Equipment is one thing, but what about the radio spectrum issues? Radiomicrophones and in-ear monitors are normally analogue (for reasons that can wait for another day) and use around 200 kHz of spectrum each. If all the devices in Baku (184) were turned on at the same time, they would need 37 MHz of spectrum just to exist. But this is not the full story. Firstly, if devices were 200 kHz apart, there would be not a sliver of a gap between them. Whilst they could all successfully transmit, even the smartest digital receivers would find it difficult to separate them from each other - and we aren't dealing with digital technology. In reality, frequencies are separated by at least 300 kHz and often more to allow the receivers room to breathe.

The situation is even worse than this however. Due to the close proximity of devices to each other, and of devices to receivers, there is a tendency for lots of intermodulation to occur. If you assume that each 200 kHz channel that can be used is numbered, starting at 1 then you can't use the adjacent channel because it's too close. You also, because of intermodulation, can't use any channel which represents the difference or sum of twice the value of one channel being used minus another channel being used. As and example:

• If you are using channels 1 and 3 (channel 2 is adjacent to channel 1, so you can't use that), you can't use channel 4 (as its adjacent to 3) or channel 5 (as its 2*3-1). The next available channel is 6.
• If you are using channels 1, 3 and 6, you now also can't use channel 7 (adj to 6). The next available channel is 8.
• If you are using channels 1, 3, 6 and 8 you now can't use channels 9 (adj to 8), 10 (2*8-6) or 11 (2*6-1). The next available channel is 12.

... and so on.

The diagram below shows the situation. Green channels are those in use. Red channels are adjacent channels that can't be used. Purple channels are intermodulation products. Pink channels are both at the same time!



In the example given above, out of 24 available channels, only 8 are useable. Actually, if you extended the diagram at this point you would find that another 10 channels are already 'wallied out' because of intermodulation. So 8 transmitters has used 34 frequencies! The relationship between the number of transmitters on-air and the number of channels sterilised is not linear but in general something of the order of 1 frequency in 5 can be used for radiomicrophones where they are packed densely in a given location if these problems are to be avoided.

This represents pretty poor frequency efficiency but is fairly representative of what is achieved in real life, which is that only something like one fifth of any spectrum available for radiomicrophones or in-ear monitors can be used in any one venue at the same time. Returning to Baku then, the amount of spectrum required is not 37 MHz, but to support 184 devices simultaneously would require more like 184 MHz of spectrum: give or take 1 MHz per device!

Most radiomicrophones (and in-ear monitors) operate in and amongst television broadcasts in the UHF band which notionally runs from 470 to 862 MHz. In many countries, however, the upper end of this band from 790 MHz upwards has now been set-aside for mobile broadband services, leaving 320 MHz remaining for television broadcasting (and of course radio microphones).

Azerbaijan, last year's host of the Eurovision, is yet to switch over to digital TV and equally has not yet used the upper part of the UHF TV band for mobile services and so finding 180 MHz of spectrum for radiomicrophones is presumably not that difficult. In 2013, however, the contest is in Malmo, Sweden. Not only has Sweden cleared the upper part of the UHF band for mobile services, but it has also gone over to digital broadcasting. Malmo is virtually on the border between Sweden and Denmark meaning that the local TV spectrum will be occupied not only by the 8 Swedish multiplexes but by the Danish ones too. Finding 180 MHz of spectrum for this year's competition is therefore much more challenging.

But move forward 10 years and then what will happen. For starters, at the current rate of growth, and assuming no improvement in the spectrum efficiency of wireless microphone technology, there will be a requirement for 560 MHz of spectrum for the Eurovision. Secondly, the parts of the UHF TV band currently unoccupied and used for these purposes will be full of 'cognitive radio' devices hunting out every last vestige of unused spectrum. What will happen then? The simple fact is that no-one really knows, but the programme making community are worried, and understandably so. If there was a way around the intermodulation problem, then the amount of spectrum required would decrease, so perhaps now is the time for some enterprising RF engineer to find a solution to this problem so that we can continue to enjoy the pageantry of the world's greatest song contest. Either that or sing less? Some would argue that for the Eurovision that would be no bad thing.