In part 1 of this post, we described the path that a mobile phone call takes across the Atlantic. Here we will describe what is travelling along the ‘through the air’ part of that path. Well, it’s a phone call, right? But in what form? In what follows, some details may be spotted by telecommunications engineers to be overly simplified. But we are doing that to paint a big picture, and unless you are a telecoms engineer, you hopefully won’t mind that we are describing essentially 1G and 2G technology.
Information in waves
So the first part of the big picture is that your voice does not travel across the Atlantic. That should be obvious – after all, you can’t shout more than a few hundred metres. But it is worth thinking about in these terms. When we talk to each other (without phones), the sound travels in the form of pressure waves through the air. These pressure waves cause a response in the listener’s ear, which is perceived as sound. The fluctuations in the pressure waves encode the information in the sound – the rises and falls in pitch and volume, and the changes in timbre. A graph of the pressure in the air over time for about 1 second’s worth of speech might look like this.
If you have ever used sound recording equipment on a PC, this kind of wave shape might be familiar to you.
Electricity, light and radio
The second bit of the big picture is that if we can’t shout that pressure wave across the Atlantic, we need to find a way to send the information in it across the Atlantic, and then convert it back into a pressure wave (for the listener to hear) at the other end.
Early landline phones used electrical signals to carry that information: typically a piezoelectric crystal in the handset would convert the pressure fluctuations into voltage fluctuations. In other words, it would turn the kinetic energy of air molecules into electrical energy. An electrical signal then travelled along telephone cables. The voltage-time graph of the electrical signal was the same shape (hopefully) as the pressure wave graph. It was an analogue signal, which means that its strength varies continually (in the sense of ‘smoothly through the values without step changes’). Then at the other end, we just need a component that can transform the energy in the opposite sense and recreate a pressure wave in the air. That component is called a loudspeaker.
There are other ways to encode the information in the electrical signal and transmit it, rather than sending the electrical signal itself. One is to superimpose the information onto a (much higher frequency) electromagnetic wave, known as a carrier wave (because it carries the information). Many modern landline signals encode the electrical signal onto a beam of light in fibre optic cables to carry the information. However, we aren’t going to dwell on this here because light in fibre optics travels as a digital signal, and we haven’t reached that part of our narrative yet (information here for the interested). Radio signals also encode information onto electromagnetic waves; two important differences between radio waves and landline phone calls are:
- Radio uses the radio wave (!) part of the electromagnetic spectrum to act as the carrier wave, whereas fibre optic phone call signals have higher frequency and operate in the infra-red region of the spectrum
- Fibre optic cables ‘channel’ the call to the right person and no-one else can hear it; radio ‘broadcasts’ to everyone, and people then choose whether or not to tune in and listen
AM and FM radio
Although this post is pretending to be about mobile phone calls, rather than radio, it is worth discussing the difference between AM and FM radio as a step to understanding mobile technology. The process of altering a carrier wave, and thereby encoding information upon it is called ‘modulation’. Three classes of modulation are:
- Amplitude modulation (AM) which we’ll describe here
- Frequency modulation (FM) which we’ll also describe here
- Digital modulation, which will happen in a few paragraphs time (we already said – we haven’t got to that yet…)
So on your radio, when you select AM, FM, or DAB, you are choosing between stations using these three modes of ‘modulation’, which is what the ‘M’ stands for.
In AM, the information is encoded by modifying the amplitude of a carrier wave (shown as the height on these graphs). The information waveform can be thought of as an ‘envelope’, which constrains the carrier wave as it oscillates. When the modulated waveform reaches a receiver, a demodulator ‘subtracts’ the carrier wave in order to reconstitute the information waveform.
If there are any Arctic Monkeys fans who didn’t already know all this stuff, you might just have learned why their album AM has this album cover artwork.
In FM, information is used to modify the frequency of a carrier wave. Each frequency of the modulated wave represents a particular signal level (amplitude) in the information waveform. In a similar way to AM, a demodulator at the receiver can reconstitute the information waveform.
So, back to mobiles phones from radio… 1G mobile phones used frequency modulation in this way, and a microwave carrying the phone call would ‘look’ like the orange wave in the picture above. That’s a slight simplification, because we have drawn a very simple information waveform. If you imagine encoding the voice signal from earlier on this page, then the changes in frequency would look more complicated than the nice regular orange curve above.
Going digital
So what does it even mean to say that something is digital? Above, we described an analogue signal as one which varies continuously with time. In contrast, a digital signal is one that can only take a certain number of discrete values. An analogue signal can be ‘digitized’ by measuring it at regular intervals in time, a process known as ‘sampling’. The diagram below shows an analogue waveform in blue, samples of it, and the resulting digital version of it in green.
Notice that each sample is at one of the intersections of the grid lines, not necessarily exactly ‘on’ the analogue signal. That is because the vertical gridlines denote the times the samples are taken, and the horizontal grid lines show the digital levels available – values in between are not available. You can see that the digital version is not a very faithful representation of the original analogue form. Two ways to make it more faithful are to:
- Have more digital levels available
- Sample more often (up to a limit)
The next diagram shows the same analogue signal sampled at twice the sampling rate (vertical lines with half the original spacing) with twice the available digital levels (horizontal lines with half the original spacing). Notice how the fidelity of the digital signal has increased (which just means that it looks more like the original analogue signal than the previous version does).
Some important advantages of digital over analogue information transfer are that:
- Digital signals are less affected by ‘noise’ (any unwanted electrical signal which can interfere with the desired signal), provided the allowed signal levels are sufficiently well-separated compared to the noise level
- Digital signals can be compressed to permit higher rates of information transfer
- Digital signals are easier to encrypt
Pulse code modulation (PCM)
We said that digital signal are signals with only a certain number of allowed levels. Commonly, the number of digital levels allowed is 2, and those values are 0 (off) and 1 (on). Such a signal is called a binary signal. So the next part of the big picture is that your mobile phone has to encode the electrical signal representing the pressure wave of your voice as a series of digital samples, and then turn the string of those samples into a code consisting of 1s and 0s. How to do that?
Well, it’s conceptually straightforward (if technically difficult) – we just identify each of the digital levels by a code consisting of 0s and 1s, and transmit the resulting code! The diagram below shows the first five samples from our original analogue signal encoded as 11000011010001101000. In fact, we have encoded the levels in binary so that they are equivalent to 0, 1, 2 etc up the signal strength axis – we are counting in binary. If you don’t know how to count in binary, it doesn’t matter. All that counts (for the concept at least) is that there are enough different combinations of 1s and 0s for all the digital levels needed.
How do you transmit 1s and 0s from your phone?
And the next part of the big picture comes from asking “how do you send a series of 1s and 0s through the network?” A simplified picture comes from what is called “frequency shift keying” (FSK). What this means is that two different frequencies of microwave signal are transmitted, one each to represent 1s and 0s. So what actually gets transmitted from your mobile phone in this case would be an electromagnetic wave represented by the blue picture below.
So to answer the question posed at the beginning, what travels through the air from your phone are modulated microwaves like the picture above.
Your 4G phone actually uses a more complicated type of modulation, rather than FSK but the simple example of FSK describes the big picture of rapidly changing properties of a microwave, to represent the 1s and 0s coming from a string of samples of electrical signal, which itself was created from the pressure wave of your voice.
And we’ll finish off by asking this question, which we will answer in a future post: why don’t you get CD quality sound on a phone call? After all, surely that’s just digital sound too?