Wednesday 26 October, 2016, 19:12 - Spectrum ManagementIt seems that consultants RealWireless have been at it again. At what again, you ask? Back in May 2014, Wireless Waffle reported that a report, produced for Ofcom had been modified to seemingly correct a factor of 1000 error in the forecasts for future mobile data traffic thus fatally skewing their forecasts for the amount of spectrum needed for mobile services.
Posted by Administrator
Posted by Administrator
This time, in a new report for the European Commission entitled, 'Identification and quantification of key socio-economic data to support strategic planning for the
introduction of 5G in Europe' (admittedly led by Tech4i˛ but where you can bet your bottom dollar that RealWireless were the spectrum experts), the mathematical brainiacs have declared that the amount of radio spectrum required for the next generation of mobile service (5G) by 2025 will be between 19 and 76 GHz depending on the 'sharing scenario', meaning that if operators are prepared to get along nicely and co-operate to use the same spectrum, somewhat less will be needed (as if that is going to happen!)
The report specifically considered a number of scenarios and concluded that it is in the use of 5G on motorways where the demand for spectrum is highest. This is based on the notion that there are :
1000 vehicles along a 1km stretch of motorway, most of which (75%) are using high rate (4K/UHD) and pervasive video applications and devices operating simultaneously in vehicles. The usage in vehicles on a busy motorway within a Smart City with traffic building up due to an accident, is estimated to be 215 Mbps per vehicle as described in the transport ‘day in the life of’ story.
Er, firstly, it doesn't take 215 Mbps to deliver 4K/UHD video: Netflix purportedly manage to deliver 4K video at 18 Mbps, so what is the other 200 Mbps per vehicle for? Obviously that is the driver playing 'Interactive 4D Battle Death Wars XIV' which streams live data from 200 other players to create a 'fully-immersive real-time near-death experience'. Obviously the driver can do this because the car drives itself.
Of course the amount of spectrum needed to deliver this connectivity, even if it did prove necessary, could be reduced by increasing the number of base stations. One every 4 to 5 metres should do it.
These silly preductions, and it's not just RealWireless, because based on the latest ITU estimates, we will all be streaming 4K video 24 hours of every day by 2043, are designed for one purpose only: to allow the mobile industry, most notably the manufacturers of network infrastructure (such as Ericsson and Huawei) to convince the world that more spectrum is needed, so that they can flog a load more of new equipment.
Are the mobile operators themselves clamouring for new spectrum? There is little evidence to suggest that they are. The latest spectrum auction in the US has failed, so far, to raise enough money from the mobile operators to pay the broadcasters to budge over and make way for the 'big boys'. And there are several spectrum bands that have been available for mobile services for 10 years or more (notably the 2 GHz TDD bands, and the 3.4-3.6 GHz band) for which no widespread services have been launched.
It's hard to see how the estimated 76 GHz of spectrum could possibly be needed in 9 years time, and even less so, why operators would invest in spectrum and infrastructure to allow bored drivers to play 'Interactive 4D Battle Death Wars' whilst stuck in a traffic jam. Wireless Waffle predicts that the real amount of spectrum needed by 2025 will be at least 10 times less than this. So, the gauntlet has been thrown down, let's come back in 2026 and see who was right!
The main frequency that SETIologists believe would be the primary real-estate for aliens wishing to make themselves known is around what is termed the 'water-hole' at frequencies close to 1.5 GHz. This frequency range includes the so-called hydrogen line (1420.4 MHz) and hydroxyl line (1666 MHz) which are frequencies that are generated by naturally occuring processes within atoms. The other advantage is that Earth's atmosphere, which comprises quite a large percentage of water, is relatively transparent at these frequencies making Earth based observations easier. The logic is that any intelligent life would be as aware of these lines in the spectrum as we are, and would figure that these are good places to transmit. Whether the transparency of the atmosphere at these frequencies on the planets they inhabit would also point them in the direction of the water-hole is a moot point.
Other frequencies that have been considered are twice the water-hole frequencies (after all, intelligent life would surely be able to multiply by 2 - assuming their number system used the same integers as we do), or 4.462 GHz which is the hydrogen line times pi (because circles are a universal phenomena, right?)
But could we actually receive a transmission on such a frequency? Let's do the link budget calculation...
- The path loss from our nearest star, Alpha Centauri, which is 4.37 light years (41,315,094,156,000 km) away at 1.5 GHz is 368 dB.
- If we use a BIG dish, say 30 metres in diameter, to receive the signal, it would have a gain of 51 dBi.
- Let's also assume that the aliens are transmitting using a similar sized dish (and they are pointing it directly at Earth).
- If the signal is very low bandwidth data (say 100 bits per second) we would need a receiver bandwidth of around 100 Hz, giving a noise floor for a cryogenically cooled receiver of -168 dBm.
- The necessary transmitter power to overcome the noise is therefore 368-51-51-168 or 98 dBm or 6.3 MegaWatts.
But path loss is dependent on frequency, and so if a lower frequency was used, say 100 kHz instead of 1500 MHz, the path loss (from Alpha) drops from 368 dB to a more managable 284 dB (84 dB less). Unfortunately the gain of the receive dish also falls from 51 dB to a measly -33 dB (also 84 dB different). But instead of a dish, we could use a long-wire to receive the signals, at 1.5 km long, it would have a gain of 2 dBi, so overall we would gain 33 dB in our link budget calculation.
"A-ha", you say, "but the transmitting antenna would have a lower gain too, so nobody really wins. If I do the maths right, the required transmitter power is now 284+0+0-168 which is 116 dBm or 400 MegaWatts, which is a bit far fetched isn't it?" Maybe, but it's easier to generate 400 Megawatts at 100 kHz than it is at 1.5 GHz (or it is for humans anyway). In fact, this kind of power is generated every day on Earth by... lightning storms.
It may come as no surprise, therefore, that researchers at the University of St Andrews believe that signals that were received in 2009 from exoplanet HAT-P-11b might well have been caused by lightning storms on the distant planet. But what use is this, it doesn't represent extra-terrestrial intelligence, just extra-terrestrial weather (and we already know that even the other planets in our own solar system exhibit different weather characteristics).
The point, if there is one, is that if we could modulate lightening storms, or perhaps induce them in a way that allowed them to occur in a predictable fashion, we could make signals big enough to be transmitted across inter-stellar space. According to Climate Viewer:
DARPA wants to trigger lightning to protect infrastructure, satellites, and use the artificially generated ELF waves to send messages worldwide. Lightning strikes are “triggered” at the University of Florida and University of Arizona, a network of sensors called the Holographic Array for Ionospheric Lightning (HAIL) collects info on these strikes, and HAARP has a large role in the whole process.
But if lightning can be heard many light years away, perhaps these artificially induced lightning strikes are nothing to do with their stated objective but are actually to 'send messages extra-terrestrially'. Maybe the signals received from HAT-P-11b are not just random lightning storms and despite the Daily Mail claiming that these are not messages from space, perhaps they are an attempt at communication after all. After all, stranger things have been true!
Thursday 24 March, 2016, 12:50 - Broadcasting, Pirate/Clandestine, Spectrum ManagementThis week, Wireless Waffle took a rare trip to the far reaches of the UK. Or Birmingham to be precise. To be even more precise, the motorways surrounding the UK's second city, whilst en route to elsewhere. The journey gave the opportunity to do a bit of tuning around the FM band to see what's happening in the West Midlands these days. Other than the various commercial stations (including local regional stations Touch FM and Free Radio), and a handful of community stations (including some in neighbouring Coventry), the thing that was most different to the airwaves in the South East was the absence of much in the way of pirate radio.
Posted by Administrator
Posted by Administrator
Only two pirate stations were heard during the journey:
It's probably about time that Wireless Waffle stopped banging on about pirate radion stations, after all it's now just a year shy of the 50th anniversary of the Marine etc Offences Act whose purpose was primarily to bring down the original pirate stations of the day. The fact, however, that such stations continue to abound suggests that the mainstream UK radio market is failing certain sections of society, which appear to be certain ethnic minorities (Caribbean, African and Turkish in particular) and those who like gruff-beat, wacka-jam and gutter-beat music (or genres with similiarly bizarre titles). It's also clear that Ofcom's efforts to take the stations off-air is not having the desired effect (meaning 'to force pirates off-air').
Information on enforcement activity released by Ofcom under the freedom of information act reveals, for example, that between 2007 and 2012, Kriss FM had its transmitters taken off-air by Ofcom 17 times and its studios raided 6 times. Hot92 had its transmitter raided 42 times and its studio raided 3 times. Sting had an astounding 60 transmitter raids (and 3 studio raids) over the same period.
We recently discussed the idea that pirate radio could move to small-scale DAB radio services, or conversely that the closure of regular FM services might open up the FM band as a playground for even more pirates. But could there be yet another option... If Ofcom were to cease all enforcement activity on pirate radio (except in casese where it was causing interference to safety-of-life services) and let the illegal broadcasters run riot, the amount of intereference they would cause to legitimate stations would increase and this might be the incentive needed for those listeners to finally go out and buy a digital radio! So instead of continued enforcement, why not leave the pirates alone and see if that has the desired effect (in this case meaning 'forcing a move to digital radio').
Saturday 23 January, 2016, 11:49 - Broadcasting, Licensed, Pirate/Clandestine, Radio Randomness, Spectrum ManagementThe number of pirate radio stations on-air in London does not appear to have diminished over the past 10 or more years, despite there now being many legal ways in which stations could reach their audiences, most recently though streaming audio on-line.
Posted by Administrator
Posted by Administrator
Ofcom has recently proposed that many local radio stations could be accommodated on 'small scale DAB' transmitters which provide a localised service, and it is conducting a number of trials around the UK of such a service. The idea is to use low cost hardware and software to develop the DAB signals, for example using a USRP software radio, and the various software tools provided by the Open Digital Radio project. Using these, it is possible to put a (very) low power DAB station on-air for less than GBP1000 and with a suitable power amplifier (for example a 30 Watt Mitsubishi amplifier module, or how's about a 1.2 kW amplifier module), a DAB transmitter with reasonable coverage can be built for not that much more. Indeed research conducted for Ofcom suggests that a 100 Watt e.r.p. service using such technology could be provided to broadcasters for around GBP1400 per year (at a bit rate of 160 kbps which is far higher quality than most of the existing UK DAB services!)
An article in spectrum newsletter PolicyTracker entitled, 'Can DAB save us from the pirates?' (note - a subscription is required to read the full article), makes the point that the criteria which the UK has set to begin the digital switch-over of radio services (e.g. the turning off of analogue services in favour of digital) is almost upon us. The criteria is that 50% of UK listening should be on a digital platform (whether DAB, on-line, cable, satellite or other) and the latest figures show this is now up to 43%, though as Wireless Waffle has pointed out before, the proportion of digital listening appears to have stagnated. If the UK does set a date for the winding down of FM radio, one of two things could happen:
- pirates may move to digital platforms, in order to be found on the same dial as other stations; or
- pirates may take the opportunity of an emptier FM band to choose clearer frequencies, increase their power, or just increase the number of services.
The problem is that in either case there is no guarantee that the pirates would do this in a legal fashion. Pirate radio stations are not renowned for co-operating with each other so why would they pay to be on a DAB platform, rather than buy the equipment themselves and set up digital pirate stations? The answer might come in the form of the reduced number of frequencies available. In the FM band, assuming a station every 300 kHz (which is just about OK from an interference perspective), there is room for 68 stations on the dial. Taking into account the 20 or so legal stations already on-air in London (depending on what you count as London), this leaves a potential for 48 pirates. For DAB radio, which requires 1.4 MHz of spectrum to operate, there is room for just 32 transmissions in any one location. In London 4 of these frequency blocks are in use, leaving 28 'available'.
"Ah", you say, "but each DAB multiplex can carry 10 or more stations, so really the number is 280". That is true, but this requires the pirates to club together to buy and operate the equipment and, as already stated, they aren't that good at this. Perhaps the business model they employ could change, and a smaller number of illegal transmitter operators could provide services to multiple pirate stations. Or perhaps a small number of legal transmitter operators could provide the same service. The problem here seems to be that in order to legitimise the pirates, they would have to be invited to 'come in from the cold', and Ofcom would have to have a set of licensing policies that were sufficiently lax to permit 24 stations all playing the same kind of electro-shed, play-house or garage-door music onto the dial. It is almost certain that the existing legitimate stations would object to this on the grounds that it would provide unfair competition. But the fact is that such competition already exists on the FM band, and at least if it were done under a licensed framework, there would be some control over what went on, and thus greater protection of the existing 'big boys'.
Maybe Ofcom could take a leaf out of the book of the Lebanese regulator (the TRA) who offered all unlicensed FM operators (which was a large proportion of the country's stations at the time), a licence, if they came forward and provided the necessary details of their transmitting facilities (power, frequency, antenna height and so on). Most stations did this, wishing to gain legitimacy for their service. As soon as they did, however, the TRA could begin to change frequencies, powers and so on to bring the stations in-line, through a proper legal framework that allowed them to inflict penalties if people refused.
Another model might be for Ofcom to licence small-scale DAB operators, but not to set any criteria over which stations are carried on the multiplex, other than their normal broadcasting code which is designed, for example, to stop politically motivated stations from using the airwaves in an partial way. Other than a bit of swearing here and there, most pirate stations would probably already meet most of the rules. This is not dissimilar to the way in which Ofcom licenses digital television, with multiplexes being awarded to companies who are then largely at liberty to select which programmes they include.
In either of the above cases, whether the idea that Ofcom could have an 'amnesty' for pirates, if they agreed to go onto small DAB multiplexes, or if the operation of the DAB multiplexes were made flexible enough to accommodate pirate stations being on them, there might be a sufficient increase the listening to DAB services such that the necessary 50% switch-over threshold is reached sooner rather than later. A 'win-win' situation?