Find your way round the RF spectrum
2GHz, 10GHz, Wi-Fi, 802.11a,b and g, licences - what's it all mean?
24 November 2003
Barely a week goes by when someone doesn't pop up and explain that the future of fast Internet access is some new wireless equipment with a different acronym, working at a different frequency, using a different standard.
Anyone could be forgiven for being hopelessly confused, particularly when it seems very little in the way of actual services are being offered. This then is an attempt to explain the myriad of possibilities out there and comment on their likely success in the unwired world.
The reason why it is so hard to simply explain what wireless offerings are best is because there are a large range of interacting factors with each. It is best explained in terms of frequency.
Frequency is the number of oscillations that occur every second. If there is one per second, this is one Hertz (Hz). For the purposes of sending large amount of data, the radio frequencies (RF) used vary from tens of thousands to millions of Hertz.
Let's fudge over the physics: data is added to the normal smooth sinusoidal wave, transmitted and decoded by the receiver. (It is completely different with Ultra-Wide Band but that is still in development.
By and large, RF waves sent at a higher frequency will not travel as far but can transmit more data. Higher frequency equipment also tends to be is more expensive (and you need more transmitting aerials to get full coverage of a given area. At higher frequencies, the aerials become more important and may have to be lined up carefully.
An illustration of this point from broadcast radio, is the fact that AM radio (working at a lower frequency) can be easily picked up but an FM radio needs an aerial pointed in the right direction to get a signal. The FM signal is, however, of better quality.
However, since the range of a signal also depends on how much power you put into it, and this in turn depends on regulations, do not be surprised to find some high frequency services travelling faster than some lower frequency ones. That is all down to the politics.
Combined with all the technical factors, there are also political factors. Over time, the balance between cost, distance and quality has meant that particular applications from boat radios to radar guns have chosen particular frequency bands. However, the government has also dictated what frequencies can be used for what application in order to give some sort of order.
Certain important functions such as police radios have been given their own frequencies and others are banned from using them so there is no interference.
There are other bands that companies or organisations, such as mobile phone companies, pay the government to have exclusive access to. If they get a licence for this particular band they can use a lot more power to send their signal and this means that the signal can travel a far greater distance before petering out.
At the same time, the government has designated big chunks of frequency that can be used by anyone. However, in these licence-exempt (not licence free) frequency bands, the amount of power with which people are allowed to send signals is strictly controlled so that signals are localised and will not override other users.
The combination of all these factors has led to a very confusing world of wireless data. And it is set to get more confusing as the government faces calls for more bandwidth to be handed over to individual companies so they can deliver data across the UK. This is the move to the unwired world, but, perhaps more importantly to the government, it is also a chance to make huge sums of money.
The UK government couldn't believe its luck when mobile phone companies paid billions of pounds for the right to use the 1.9GHz and 2.2GHz frequencies for their forthcoming 3G services. It quickly sought to sell off other frequencies to the highest bidder and make more billions. Rather unfortunately, its greed cocked the whole thing up and subsequent auctions for the 3.4GHz and 28GHz frequencies (suitable for fixed wireless use) were complete fiascos.
In fact, the sale and use of different frequencies to send data wirelessly has been a farce following much excited previous conversation. However, wireless data is beginning to take off and within the next decade may take over as the main way we communicate with one another. Which particular methods (and frequencies) ultimately triumph is still wide open but dozens of companies are planning for it to be them.
This then is how the market exists as of the end of 2003.
The world according to wireless
Here is a rundown of current services in terms of frequency, with low frequencies first and the inclusion of all other services along the way.
20kHz - 150 kHz: Short range devices, metal detectors etc
150 kHz to 255 kHz: AM radio
255 kHz - 526 kHz: Radio navigation, marine communication
531 and 1620 kHz: Medium-wave radio
1.7 - 70 MHz: Marine, aeronautical communication; government use
70 - 88 MHz: TV
88 - 108 MHz: FM radio
110 - 890 MHz: TV, Satellite, Radios
890 MHz - 950 MHz: Mobile phones
960 MHz - 1700 MHz: Radio navigation
1.5 GHz: Earmarked for digital broadcasting
1.71 - 1.87 GHz: GSM mobiles
1.9 -2.0 GHz and 2.1- 2.2 GHz: 3G mobiles
2 GHz: BT in particular has been asking the government to licence this frequency as it is extremely useful for sending data. The 3G licences slotted either side of it. Why? Because this is the frequency used by the security services and they want to keep hold of it. The government might be tempted to sell if off if it thought it could make a lot of money but it would be a battle.
2.2 - 2.4 GHz: Government, military
2.4 - 2.5 GHz: This is the most successful of all wireless data frequencies, with large amounts of cheap equipment (due to investment, due to demand) attracting people. Licence-exempt so anyone is free to use it as long as they follow the rules, it is where Bluetooth and two of the Wi-Fi flavours (802.11b and 802.11g). The Wi-Fi wireless LANs -- 802.11g is faster than 802.11b – are used for sending data relatively short distances at hotspots or in offices.
2.5- 2.7 GHz: Being touted as a possible extension area for 3G phones.
2.7 - 2.9 GHz: Radar
3.4 GHz: Fixed wireless links, suitable for distributing broadband access to locations that cannot get DSL or broadband. In June this year, licences for 15 regions across the UK were sold but, due to a poorly thought-out auction process, one company - PCCW - has gained almost exclusive use of the frequency across the UK. In just one area - the South East - does another company exist (PublicHub). In other countries, a variety of companies have been allowed to compete with one another in one area and the roll-out has proven successful. PCCW has so far failed to even outline what its plans are. PublicHub has just announced its first small-scale trial. PCCW could theoretically use its monopoly to launch a good, cheap UK-wide system. But don't hold your breath - it could just as well sell its bandwidth as backhaul for 3G phone companies and avoid the cost and hassle of rolling out its own service.
3.6 GHz: Again, this band is good for fixed wireless access. While the government is talking of another auction for this frequency, there is not much taste for it at the moment following the disastrous auctions at 3.4GHz and 28GHz. Currently, Tele2 pays the government an annual retainer for exclusive use of the band, and this is likely to continue. Tele2 is gradually rolling out services across the UK but mostly only in big towns, where a variety of options for Net access are already available.
5 GHz: This is the other main unlicensed band of frequency and is where the “a” Wi-Fi standard lives. It is more expensive and many predict its demise since the cheaper “g” standard offers the same speed of data exchange. However, corporates may use “a” as it gives them greater security. Its long-term future does not look that rosy though.
10 GHz: Another disaster. At this frequency, a cable-like fixed wireless access service can be offered but there is limited spectrum available, making some fearful that it can't support much of a market. NTL is keen on 10GHz and is pushing it - but again only in big cities. Others though, such as Cable & Wireless, Thus, Eircom have handed back their licences to the government. Others licence holders have still failed to provide any services.
10.8 - 12.8 GHz: - Digital TV
24 GHz: The other auction disaster, the 24GHz band is again suitable for fixed wireless access. It has a lot of potential, especially since it has been harmonised across Europe. Unfortunately, after three auction attempts, a large number of the licences remain unsold. The government's plans to make £1 billion alienated bidders at a time of uncertainty and so it has got off to a bad start. Very few limited services have popped up across the country but bigger and better things should be around the corner.
40 GHz: Another spectrum dropped from intended auction. A market that is being developed across Europe. One great advantage is a large available bandwidth. However the cost of introducing it is such that it would need a large number of subscribers in a small area before it became financially viable. Again, most efforts will tend to focus on cities, where a range of services are already available.
As can be seen, the market is in an uncertain state and investment is limited, meaning that what services are available are currently restricted to tightly defined geographic areas. However, there is a lot of room for movement and as wireless communications becomes more widely understood and accepted, there will be fierce battles for customers, and a variety of services.
Which way the market goes may depend on who is willing to put in the money; just one or two big contracts could make a service viable.