Wednesday 30 November, 2005, 06:55 - Spectrum ManagementNapoleon and the Mathematician
Before going any further we need to take a brief side-step and remember Jean Baptiste Joseph Fourier (1768-1830), one time governor of Lower Egypt under Napoleon and master mathematician. Fourier developed an equation (the Fourier Transform), which allows us to transform waveforms in the time-domain into waveforms in the frequency domain. Fourier shows us that a pure sine-wave in the time-domain transforms into a single line in the frequency domain, representing the frequency at which the sine-wave oscillates (simple eh?). The converse is also true, in that a single line in the time-domain (an impulse such as is caused when lightning strikes), converts to a sine-wave in the frequency-domain – having content at virtually ALL frequencies – which is why a nearby lightning strike, which looks like an impulse in the time-domain, can affect reception on every radio receiver in a house from medium wave (~1 MHz) to television (~500 MHz) and even satellite (~10,000 MHz).
Spark Gap Transmitters
Let's now move forward a hundred years to the work of Guglielmo Marconi (1874-1937), father of modern-day radio. During Marconi's experiments into radio transmission, oscillators producing a nice sine-wave on a single frequency weren't available. Instead, Marconi used a spark gap transmitter. This device produced an impulse (much like a lightning strike) by making a spark between two contacts and thus produced a waveform with content on all frequencies. Then, using a crude tuned-circuit, the frequency that was wanted was selected. It was an inefficient way of transmitting but using high-powered sparks, high-powered signals could easily be developed – which were needed because of the poor sensitivity of the receivers. Also, due to the way in which spark-gap signals were generated, they produced a great deal of interference on many radio frequencies other than the one to which they were tuned.
As the use of radio grew, there was an increased need to separate one transmission from another to reduce interference and to that end, in 1927, the International Radio Consultative Committee (CCIR) was formed in Washington DC (these days part of the ITU). The 1927 conference also decided that the use of spark-gap transmitters would be outlawed (and they would have been completely phased-out by 1939 had World War II not started which slightly delayed their demise).
Ultra Wide Band
Ultra Wide Band (UWB) is a relatively new term to describe a technology which had been descrined in the early 1960's as 'carrier-free', 'baseband' or 'impulse' technology. The basic concept is to develop, transmit and receive a burst of radio frequency (RF) which represent from one to only a few cycles of an RF carrier wave – effectively an impulse of RF. The resultant waveforms are extremely wide, so much so that it is often difficult to determine the actual RF centre frequency. Recent implementations of UWB do away with the idea of using an impulse to generate the signal due to difficulties in selecting the wanted frequencies, but the concept of using a transmission spread out over a very wide frequency range remains. In this sense, UWB is the modern-day equivalent of spark-gap transmitters!
The difference, however, is the power levels concerned. Whereas the spark-gap transmitters of old often used powers of 10s of kilowatts, UWB uses powers of only 10s of milliwatts (around a million times less). Further, the fact that UWB signals are purposefully spread over a very wide range of frequencies means that the actual power transmitted on any given frequency is miniscule and in many cases is below the background noise (caused by man-made, natural and even galactic sources). Conversely, spark gap transmitters were few and far between but the intention is that UWB devices will be ubiquitous. Over a country the size of the UK, where there may have been only a handful of spark gap transmitters, there may be millions of UWB devices so the potential for interference is, arguably, almost as great.
In order to try and prevent UWB from causing interference to other radio users, 'masks' have been set by a number of organisations from the Federal Communications Commission (FCC) in the United States to the European Telecommunications Standards Institute (ETSI) in Europe. A radio mask can be considered as similar to a mask used when printing, it prevents radio signals from being transmitted over certain areas (or reduces them) and allows them to be transmitted over other areas, in this way sensitive services (such as aeronautical radar) can have increased protection from interference. The fact that masks are necessary indicates that spectrum administrations are concerned about the interference potential from UWB, despite the very low signal levels involved.
Counterbalancing the interference problems, however, are the benefits that UWB technology claims to offer over other, competing technologies (such as, for example, WiFi and Bluetooth). UWB claims to provide:
· Extremely high bandwidth data links (over very short ranges)
· Precision measurement and location identification (in radar applications)
· Immunity to multi-path reflections of the signal
· Low complexity and therefore (in theory) low cost
UWB is therefore well suited to applications such as high bandwidth LANs (100Mb/s+), collision avoidance sensors, precision location systems, radio tags (cargo, personnel etc), radio imaging and even low powered handheld radios. The benefits are potentially large ranging from life-saving applications to quality of life enhancing applications, but this must be balanced against the potential for interference to other radio users.
Apples and Oranges?
Is it really fair to compare modern UWB technology with 100 year old spark-gap transmitters? Yes. And no. Yes, both transmit over a wide range of frequencies. Yes, both are based on the notion of using an impulse to generate radio signals. Yes, both have the potential to cause interference to many other radio users. No, the technology employed is fundamentally different. No, the power levels are several orders of magnitude different. No, the intended usage is different.
So what are the implications for radio users? Large numbers of UWB transmitters working together have the potential to cause interference to users across the spectrum (from around 1 to around 10 GHz for most currently planned devices). This frequency range includes a number of economically and politically important users such as mobile phones (GSM 1800, CDMA 1900 and UMTS), aeronautical and maritime radar, wireless LAN and HiperLAN, governmental and private fixed links, satellite links and in some countries wireless cable services. Despite much research in both corners of the UWB boxing-ring, there is no conclusive evidence to indicate that UWB will not cause interference to these services, then again there is no evidence that it will. Most administrations are being cautious about allowing the introduction of UWB as it's unlikely that once introduced, it would be able to be taken out of service as operation will typically be licence-exempt and thus controlling future usage will be near impossible.
For any country therefore, the decision currently exists as to whether to allow UWB services to be introduced today, delivering the kind of user benefits that the technology can provide, or whether to hold back on making that decision until there is a greater weight of international experience. Is the bright spark worth the spark gap?
Tuesday 15 November, 2005, 07:44 - Spectrum ManagementCDMA - Code Division Multiple Access - is probably one of the most complex methods for accessing the radio spectrum that is in use today. Let's first take a look at the easier to understand methods to see what the differences are.
FDMA - Frequency Division Multiple Access - is a system whereby different users use different frequencies at the same time. Imagine tuning up and down the FM dial, there are lots of stations you can hear but you can separate one from the other by tuning into the different frequencies that each uses. If two occupy the same frequency at the same time, they cause interference to each other and the whole thing falls to pieces.
TDMA - Time Division Multiple Access - is a system whereby different users use the same frequency at different times. So a number of users messages are scheduled one after the other, so that they can all share the same frequency. Imagine a lot of letters passing along the conveyor belt in a postal sorting office. Each message passes after the other along the same conveyor and can be sorted out back into individual messages for delivery. If two letters were to be in the same place at the same time they would crash into each other and the system would fail. From a radio perspective this works only if you take the message concerned and 'squash' or 'compress' it into less time than it would normally take so that at the receiving end it can be de-compressed to fill in the gap that is produced whilst the other messages are being sent.
CDMA is a system whereby all the users use the same frequency at the same time. How can this possibly work without interference being caused? It works by assigning each message a unique code. Think of it as assigning each message a different language and then place yourself in a room where you can hear all the messages together - the one you immediately understand is the one in the language you speak. So if you hear 'Je m'appelle Colin', 'Ich heisse Lyon' and 'My name is Derek', all spoken at the same time, the clearest message is the one in your native language. However this only works if all the messages are roughly the same volume (and in fact works best if all the messages are exactly the same volume), if one language is shouted a lot louder than the others, it inevitably swamps the listener so that he can no longer hear his natural language.
As such, one element of a CDMA radio system is that the receiver (listener) needs to inform all the transmitters (speakers) how strong their signal is (how loud they are being heard) so that the transmitters can adjust their transmitter power (volume) to allow the receiver to hear them all at the same strength. This system of 'power control' is one of the key technologies that allows CDMA systems to operate.
CDMA receivers also suffer from something called 'cell breathing'. Imagine that you can hear 4 different languages, the chances are that you will hear the one you understand quite easily. Now imagine you can hear 64 languages, it becomes much more difficult to hear your native language. So the more transmitters sharing the same frequency, the harder the receiver has to listen. To compensate for this, receivers move closer together and further apart each and every second to make sure that they are sufficiently near to those transmitters they are receiving to hear them well. Clearly this is not possible and what happens is that the coverage of the cell shrinks when it is heavily loaded, complicating coverage planning.
All this being said, CDMA is reckoned to be around twice as spectrally efficient as either TDMA or FDMA techniques, meaning twice as much traffic (data) can be carried in a given amount of spectrum. It is no surprise, therefore, that all of the competing third generation (3G) mobile phone technologies use CDMA: the US-based CDMA-1x series of specifications designed by Qualcomm; the European W-CDMA (Wideband-CDMA) or UMTS (Universal Mobile Telecommunications System) system; and the Chinese TD-SCDMA (Time Division-Synchronised CDMA) system.
What's interesting is that although all 3G mobile technologies use CDMA, other mobile technologies (such as Digital Audio Broadcasting, Wireless LAN and HiperLAN) have forgone the use of CDMA and instead use another spectrum access technique called OFDM - Orthogonal Frequency Division Multiplex. But that's a story for another day.