Tuesday 5 April, 2011, 13:15 - Spectrum ManagementApparently, mobile phone operators are beginning to run out of capacity on their networks due to all the data traffic that is being generated by smart-phones and people using broadband dongles in their laptops. Of course, whether or not this is true or not today or whether it is just an excuse for poor service quality, there will almost certainly come a time in the relatively near future when it the squeeze on spectrum becomes reality.
Posted by Administrator
Posted by Administrator
A typical mobile operator in an average European country will currently have access to something like 100 MHz of radio spectrum - 50 MHz for the uplink from phones to base stations and another 50 MHz in the opposite direction - more commonly written 2 x 50 MHz. This will be in usually two (or more) of the commonly available mobile bands, such as:
* 900 MHz (actually 880 - 915 and 925 - 960 MHz)
* 1800 MHz (1710 - 1785 and 1805 - 1880 MHz)
* 2100 MHz (1920 - 1980 and 2110 - 2170 MHz)
So when they do run out of spectrum, what can they do? Well help is at hand in the short term through two new bands which are being released. The first, known as the 'digital dividend', has become available due to the more efficient planning of television broadcast networks that has arisen as a result of the switch-over from analogue to digital broadcasting. The second, at 2.6 GHz, was (and in some countries still is) used for a multitude of purposes including wireless video cameras, wireless cable networks and fixed wireless broadband. Together, these two bands make just over another 200 MHz of spectrum available:
* 800 MHz (791 - 821 and 832 - 862 MHz) - the 'digital dividend'
* 2600 MHz (2500 - 2570 and 2620 - 2690 MHz) - note that the gap between 2570 and 2620 MHz is also available
If you assume that an average country has 3 or 4 mobile operators, this equates to something like another 60 MHz (2 x 30 MHz) each, resulting in a 60% increase in their capacity.
So what's the problem? Some observers (eg Cisco) claim that mobile data traffic is doubling roughly every year, so this 60% increase in capacity will amount to about 8 months of traffic growth, then the problem starts all over again. New technology will deal with some growth. Newer mobile technologies from HSPA+ to LTE and LTE-Advanced may offer a doubling in capacity over current 3G (UMTS) networks for each unit of spectrum. Another year dealt with, and only at the cost of changing over all of the network equipment and handsets!
On this front, it is perhaps no surprise that UK mobile operator O2 recently announced plans to offer free WiFi for all. Why is this no surprise? If the traffic from smartphones and laptops can be offloaded from the mobile network to WiFi hotspots, this will ease the burden on the mobile network. But this is a relatively short-term fix too. In the long term, the only way that mobile operators will be able to deal with the growth in data traffic is to get access to more spectrum. But where will this spectrum come from?
It has long been recognised that to offer a sensible (in terms of cost, coverage and capacity) mobile network, frequencies in the range 300 to 3000 MHz are best. Go any higher and things such as Doppler shift and cell handover become real problems. Go any lower and antennas become too large and unwieldy. The problem is that the remaining frequencies in this range are already being used. In general terms:
* 300 to 430 MHz is military territory
* 430 to 440 MHz is radio amateur land
* 440 to 470 MHz is full with PMR systems
* 470 to 790 MHz has UHF television broadcasters in it
* 790 to 862 MHz is already mobile
* 862 to 880 MHz houses all manner of low power devices
* 880 to 960 MHz is already mobile
* 960 to 1350 MHz is where aircraft radars and some radio amateurs live
* 1350 to 1710 MHz is for satellites (including GPS), broadcasters and more tanks, planes and guns
* 1710 to 1980 MHz is already mobile
* 1980 to 2110 MHz is partly mobile and partly full of military folk
* 2110 to 2170 MHz is already mobile
* 2170 to 2400 MHz is mostly military
* 2400 to 2500 MHz is WiFi and bluetooth land
* 2500 to 2690 MHz is already mobile
* 2690 to 2700 MHz is where radio astronomers hang out (mostly in cardigans)
* 2700 to 3100 MHz is aircraft and maritime radars
If any more space is going to be made for mobile services, someone else is therefore going to have to give up their claim to their territory. In some countries (eg Sweden) 2300 to 2400 MHz is being made available for mobile services but in the majority of European countries it is used by the defence services who, having already vacated other spectrum, are beginning to fight back.
Clearly, anyone who moves out for the benefit of mobile services will either have to stop doing what they do (unlikely) or go and do it somewhere else (costly). For any international service (eg boats and planes) this cannot be done unilaterally and getting international agreement is probably too slow. What's more, radars and things such as that need a lot of spectrum due to the way they work and furthermore, removing them might cause planes to fall out of the sky, which would seriously disrupt tourism in many parts of Europe that aren't very close to where you live (though it might have a potential commercial upside for undertakers).
Wireless Waffle is therefore going to stick it's neck out and make a proposal as to who should lose the battle for this important part of the spectrum and that is ... the broadcasters!
On a completely different, but not unrelated tack, the amount of energy consumed by a terrestrial broadcasting networks is, well, large. Not 'a whole power station' large, but still pretty big. The amount of energy consumed by a satellite is tiny. In fact, once it's up in the sky, it's zero (they are solar powered). A terrestrial broadcast network also delivers much less capacity than a satellite. So broadcasting by satellite consumes much less power (and therefore has a much lower carbon footprint) and offers much greater capacity (for services such as 'The Cartoon Network' in HD). Let's therefore turn off UHF broadcasting and give the spectrum to mobile networks - the broadcasters can go to cable and satellite and can continue to use the VHF band if they really want to.
What would a world without UHF broadcasting look like. In somewhere such as the Netherlands where 90% or more of homes are on cable, not a lot different. It would mean that people's holiday homes might need a satellite dish but these are so cheap and plentiful it should be no big deal. In the UK where most homes still have a terrestrial UHF receiver you might think this would be a bigger deal, but over 50% of homes have either satellite or cable already and again, having to buy a dish is no biggie, so other than the temporary inconvenience of swapping set top boxes and putting a dish up (or getting connected to cable) nothing much would change.
If they so desired, public service broadcasters could continue terrestrial television broadcasting using the VHF band - by switching off those ancient and largely unlistened-to DAB transmitters. DAB could be replaced by DRM and Bob's your uncle - no loss of anything important, just a bit of shuffling around.
If all this sounds far fetched, watch this space. Or, perhaps more accurately, watch outer space!
Thursday 13 January, 2011, 04:22 - Spectrum ManagementPreviously on Wireless Waffle we have discussed ways of checking and even gaining some knowledge of the state of propagation of the short-wave bands. But for truly advanced users, there is a way to find out the actual state of propagation for a particular location in real time. Scattered around the world are a series of ionosondes. These ionosondes are rather like radars in that they transmit a signal to the ionosphere and measure the time taken to get a response. They do this across a range of short-wave frequencies.
Posted by Administrator
Posted by Administrator
The result is a chart called an ionogram. An ionogram is effectively a radar picture of the height of the ionosphere at the location immediately above the ionosonde as well as providing an indication of its refractivity, over a range of frequencies. An example ionogram taken from the ionosonde in Dourbes, Belgium, is shown below.
The ionogram is the ultimate way of assessing short-wave propagation. It tells us exactly what is going on. To help interpret the ionogram, there are also a useful set of figures provided in the diagram which give us some very useful information. So... how do we interpret the ionogram to help understand HF propagation?
In the ionogram above, the strong red/pink line extending from just below 3 MHz to just above 6 MHz shows that the ionosphere above Belgium was refracting radio signals in that frequency range straight back down again (ie at an angle of 180 degrees) - it was acting like a mirror for radio frequencies in this range. As the frequency goes above 6 MHz, the line bends upwards until eventually it goes off the top of the chart. This is the point at which the ionosphere stops refracting signals back down (at 180 degrees), however it will continue to refract signals at higher frequencies which hit it at lower angles (less than 180 degrees).
From this simple data, together with the height of the ionosphere (the scale up the left hand side of the chart) it is possible to calculate a number of very useful figures, and this is done for us.
Firstly, we have the maximum usable frequency (MUF). This is shown amongst the figures to the top right of the chart (in this case 27.62 MHz) and is also repeated at the bottom of the chart (under the label 3000 km). The MUF is the highest frequency which the ionosphere will reliably reflect radio signals. It is also the one which has the lowest refraction angle. What this means is that signals at this frequency will be refracted by the ionosphere (above Belgium in this case) but only where the path between the ends of the link hits it at a low angle, which equates to a path length of around 3000 km. Two stations, each 1500 km away from Belgium, the centre of whose path is above Belgium, will therefore be able to communicate at a frequency of 27.6 MHz. So a station in Western Ireland and one in Romania are likely to be able to communicate on this frequency. Equally one in Spain and one in Sweden might too.
The second useful frequency shown is the one shown as 'foF2' in the diagram (top right). In this example foF2 is 7.15 MHz. foF2 is the highest frequency at which the ionosphere above Belgium will refract signals at an angle of 180 degrees, ie straight back down. If you therefore want to communicate from somewhere in Belgium, to the same place in Belgium, using the ionosphere, this is the highest frequency I can use. How useful! But the best bit is the interpolations between foF2 and the MUF. These are the figures shown at the bottom of the chart under the various distances (from 100 km to 3000 km). These are the maximum frequencies I can use to communicate over the distance shown.
In this example, if my path length is 100 km, the highest frequency I can use is 7.9 MHz. If my path length is 1000 km, the highest frequency I can use is 11.7 MHz. Now this is really useful. If I want to communicate from London to Stuttgart, a distance of approximately 800 km, of which Belgium is roughly half way (in the centre of the path) the highest frequency I could use, in this instance, is 10.2 MHz.
What is the lowest frequency I could use? That is more difficult. What the diagram does tell us, however, is that for short paths, (ie from Belgium to Belgium) the ionosphere was successfully refracting signals at frequencies as low as 3 MHz. How do we know this? There is a nice red/pink reflection on the chart at this frequency. Below it, the picture becomes rather scattered indicating that the refracted signal was not reliable.
So, what can we ascertain:
- The highest frequency being refracted by the ionosphere above Belgium is around 27.6 MHz. This is the highest frequency at which two stations separated by 3000 km for whom Belgium is in the centre of their path, will be able to communicate - the MUF.
- The highest frequency which can be used to communicate from one location to the same location (in Belgium) using the ionosphere is 7.15 MHz - foF2.
- For a range of distances, we can work out the maximum frequency which can be used.
- For short paths, we can find out the lowest possible frequency being refracted by the ionosphere (around 3 MHz in this case) and thus the lowest frequency which can be used.
We can also take a stab at assessing how strongly the ionosphere is refracting. The phantom reflections shown at around 450 km height are signals which were refracted from the ionosphere, then reflected by the earth and then refracted again by the ionosphere. These phantom reflections would tend to suggest that the strength of refracted signals is particularly good, as it has been strong enough to rebound from the earth and refract again! Sometimes, three or even four phantoms can be seen, indicating very strong refractions which would suggest that short wave signals would be very strong.
The ionosondes in Europe include:
- Dourbes, Belgium
- Juliusruh, Germany
- Chilton, United Kingdom (registration is required but is free)
- Warsaw, Poland
- Rome, Italy
Thus GPS devices in themselves can not be used to 'track' the location of users. What they do provide is location information which could then be sent on via some other (radio) connection to enable someones location to be tracked. Standard in-car navigation systems do not have such a facility built in and thus using one does not alert the authorities (or anyone else for that matter) to your location.
That aside, there are in increasing number of uses to which GPS is being put in which the location information it provides is used for control purposes. For example, there are anti-social behaviour tags which monitor the location of offenders and send a signal to the local police or council if the person wearing it goes outside a pre-determined area (or indeed goes inside a particular area).
Similarly, knowing where a vehicle is (for example by sending the GPS location back to a central point via a GSM phone) can be used for road toll or car insurance calculation. There is therefore a growing 'demand' for devices which can stop the GPS receiver working so that the location information for tracking people or cars is not available. Such devices are known as GPS jammers and work in much the same way as the jammers used by various governments to stop international broadcasters.
The most basic GPS jammers operate by producing a high power signal on the main frequency used by GPS receivers, strong enough to ensure that the GPS receiver can no longer hear the (very weak) signals from the GPS satellites and therefore thinks it has lost them and stops working. However, receiver manufacturers have gotten wise to such wheezes and have managed to find ways to overcome this 'carrier' jamming. More sophisticated jammers closely mimic the GPS signal so that not only is the receiver overwhelmed by the local interference but it becomes far more difficult to overcome the jamming as it looks just like a valid GPS signal.
The problem, though, with such jammers is that they don't just wipe out GPS reception by the receiver they are intending to interrupt, but can knock out GPS reception over quite a wide area. Even basic, low power devices (which typically use transmitter powers of around 10 milliWatts) can produce signals strong enough to stop GPS receivers working over a range of several hundred metres. More powerful devices (and there are some easy to get hold of devices which put out a Watt or more) can cause problems for GPS receivers over ranges of over a mile. So without knowing it, someone trying to defeat the GPS monitoring device put in their company car to monitor their movements can unintentionally end up causing aircraft or ships to not be able to identify where they are either.
Now, of course, GPS jammers are illegal to use as they intentionally cause interference which is not just unlawful in that you generally need a licence for any radio transmitter, but that causing harmful interference is a particularly scandalous deed. The problem of GPS jammers is well recognised by the authorities and the impacts of their use so severe that they are one of the very few devices that organisations such as eBay have been asked to stop the selling of on their web-site (try searching there for 'GPS jammer' and you won't find a single one on offer). However, there are ways and means of getting hold of them and like anything that is useful in a 'getting out of paying for something' or 'getting around the law' way, it's almost certain that their sale will continue with devices popping up from new suppliers as the authorities crack down on sellers one by one.
Whilst all this might sound a bit like scaremongering, with aircraft falling out of the sky, that's not the real problem (aircraft don't crash just because they lose their GPS signal). Earlier it was mentioned that GPS satellites transmit both position and time information. The time information is used for a very wide range of applications, from synchronising digital transmitter networks, to ensuring that trades on the stock market are correctly time-stamped. Imagine if these timing signals were lost - TV transmitters would fail and the stock market would come crashing down, so no Coronation Street and no hefty bonuses for city workers (whilst the last of these might not seem too severe, losing Coronation Street would be a national crisis).
GPS jammers have been described as a 'clear and present danger' and much is being done to try and minimise both their availability and their impact, but it seems their use will continue, and most likely continue to grow. If you feel like popping out and buying one for any reason, think again, you might as well drop a teaspoon of polonium-210 in your local neighbourhood resevoir - the consequences of both for society may not be that far apart!
Tuesday 1 December, 2009, 01:21 - Spectrum ManagementSome of the biggest brains in Europe, as well as hundreds of millions of Euros of public money are being poured into a concept which has the catchy name of the 'Internet of Things'. The concept in itself is a fairly straightforward one - that as well as people being connected together via the Internet, machines and sensors and all sorts of other electrical and mechanical devices will be connected together as well. So it would be possible for your fridge to talk to your lawnmower, and your kettle to have a chat with your central heating system.
Posted by Administrator
Posted by Administrator
Actually, this is nowhere near as silly as it sounds. From the perspective of saving energy and hence carbon, one of the main problems facing electricity generators is dealing with the peak load. In the USA this occurs on the hottest day of the year when air conditioning units are working overtime, and in the UK typically occurs mid-winter when heating units and lots of TVs are turned on, especially during commercial breaks when everyone gets up to make a cup of tea and turns their kettle on. So at these moments, if the kettle could chat with your freezer, for example, and tell it to stop freezing for a few minutes whilst it boils, no-one would be any the wiser and the net result would be a reduction in peak electricity consumption.
This is all fine and dandy and there are plans for 'smart cities' where lots of devices communicate with each other to the benefit of energy consumption, safety and for lots of other good reasons. But there is a limit to how effective such communication can be. Imagine the following discussion:
Kettle to Fridge Please stop freezing for a bit as I need to boil the water to make a cup of tea.
Fridge to Kettle Sorry, no can do. I've already put freezing on hold for a bit to help the tumble dryer out. Perhaps you could speak with it.
Kettle to Tumble Dryer Please could you stop drying for a bit as I need to boil the water to make a cup of tea?
Tumble Dryer to Kettle You must be joking! I've already had to stop 4 times to let the floodlights come on outside and if I don't get these clothes dry soon, there'll be trouble.
Kettle to TV Hey, TV. Any chance you could turn yourself off for a few minutes whilst I boil some water to make a cup of tea?
TV to Fridge Can you believe it? The Kettle has asked ME, ME the TV to turn off so that it can boil some water, who does it think it is?
Fridge to TV I know. Always trying to steal all the power. Nearly as bad as the iron which is on and off like a faulty switch.
TV to Fridge You are so right. I just pretend Eastenders is on when the iron asks me to turn of, it knows that I couldn't possibly interrupt that programme.
Microwave oven to Fridge and TV Can you two pipe down a bit, I'm waiting for an important message from the vacuum cleaner about who is sharing the power tomorrow morning.
And so on...
So there you have it. Several billion Euros of investment brought down by a neurotic TV and an overly chatty fridge. Not to mention the fruit bowl and the salt pot who block the airwaves with their inane chatter about whether sweet or savoury is best. And therein lies the problem: all this communication needs bandwidth, and given the nature of the devices, they will need wireless bandwidth. A European Commission white-paper on the subject addresses the issue several times in statements such as:
[the internet of things] requires truly ubiquitous wireless capacity that can handle several magnitudes more data.
Communication infrastructure should provide ubiquitous connectivity in the presence of significantly increased traffic load and should be very efficient so as to reduce the cost per bit... Many of the local connections are naturally wireless.It goes on to state
Spectrum must be valued: Radio spectrum is one of the most valuable resources of the digital age. As more and more devices and objects become wireless enabled ... spectrum is becoming a key bottleneck. We have to find ways to manage the spectrum more efficiently so as to maximise data throughput and minimise interference.
The report suggests that one possible solution would be to develop a real-time local market in radio spectrum. What does this mean? It means that when you go to make a call from your mobile (or you fridge wants to open a discussion with the vacuum cleaner) it first interrogates the 'spectrum stock market' and chooses the piece of spectrum which offers the right level of connectivity at the appropriate price. Of course the question remains as to how it does this without, in the process, also using a wireless connection.
What is certain, however, is that we have only started to see the beginning of the squeeze on the radio spectrum and that if it seems congested now, compared to the future it is still a wide open space of nothingness. And like oil as it begins to become rare, it is likely we shall see an increase in the value of spectrum too. We here at Wireless Waffle wonder whether there will eventually be unit trusts and other investment wagons on the stock market that invest in spectrum for a profitable return in the same way as they do in gold, silver, oil, crops and other limited resources. If there are, then as prices spiral, it might just shut the fridge up for a bit.