Introduction
This web report begins with the basic concepts of electronic communication: signals and the representation of information using analogue and digital techniques. Then the major forms of telecommunications and broadcast technology are described. Finally, the increasingly important 'packet switching' approach to telecommunications - as used in the Internet - is contrasted with the 'circuit switching' of traditional telephony.
Communication, Signals, and Data
Most human communication is mediated by light and sound within a limited distance. Electronic communication and information technologies use electrical processes to receive and generate light and sound signals, and to communicate, transform and store the information embodied in these signals.
Human hearing is sensitive to vibrations with frequencies between about 30 and 18,000 Hertz ('Hertz' means 'cycles-per-second' and usually abbreviated to Hz.) However, perfectly intelligible speech communication can occur using frequencies in the narrower range of 300 Hz to 3,000 Hz.
It is a relatively simple to use a microphone to transform the air pressure variations which constitute sound into similarly varying pressure of electrons (ie. a varying voltage) in a wire - and then to transmit that signal over a distance and to amplify it so it can drive a loudspeaker or earpiece. This is an example of analogue communications - the smooth, infinitely fine, changes in air-pressure are converted to similarly changing pressure of electrons in a wire. This signal is amplified resulting in a similarly changing pressure of electrons, but with a greater pressure range and/or flow of electrons, so the signal has a greater power. This signal then drives the coil of wire in a loudspeaker to move the cone, and therefore the surrounding air, in exactly the same way as the original sound moved the diaphragm of the microphone.
Analog electronics is simple to build and understand, but it is difficult to transmit a signal over very long cables - say more than a few kilometers - using analogue signals without encountering fundamental problems which degrade the quality of the signal. Cables have resistance and capacitance - which reduces the signal strength. They can act as antennae, picking up interference from other sources. Amplifiers have noise and distortion problems. Analogue methods of storing sound - such as magnetic tape - all have problems such as these, in addition to signal degradation with time and use. Furthermore, when these sound signals are represented in an analogue form, they are harder to switch, analyze and process then if they were in an digital form.
A digital representation of a human voice - to the quality required for telephony - can be achieved by measuring the analogue electrical voice signal 8,000 times per second, and converting each voltage into a number. Intelligible speech is possible in the presence of perceptible noise and distortion, so a measurement accuracy of one part in only 256 is adequate. This analogue to digital conversion is rather inaccurate. There is significant noise and distortion from the rough voltage measurement, and only frequencies up to 4,000 Hz at maximum can be represented - however this is fine if the requirement is that speech remain clearly understandable.
For music, a much more demanding sampling must be done. Audio to be stored on Compact Discs, for instance, is sampled at 44,100 times a second, with a voltage accuracy of one part in 65,536 - for both the left and right channels. Properly done, this can represent all frequencies that humans can hear, with a signal-to-noise ratio well in excess of that encountered in any realistic music performance or listening situation. (However, manufacturers of esoteric equipment claim that still higher sampling rates and greater voltage resolution are required to reproduce sound without audible degradation).
With 1960s technology, the processes of digitizing an analogue signal, storing, transmitting and transforming it, and converting it back to an analogue signal was too complex and expensive except for military and research activities.
With the advent of large-scale integrated circuits in the mid 1970s, the stage was set for the cost-effective conversion, storage, transmission, and processing of signals in a digital form.
While it is possible to build electronic circuits which process information using decimal numbers - for instance a single wire carrying a decimal digit 0 to 9 by using ten different voltages - such a system would be inordinately complex, unstable and slow. In contrast, it is relatively easy to make electronic circuitry in which each wire has only two states: '0' (eg. 0 volts or thereabouts) and '1' (eg. 5 volts or thereabouts). Such circuitry is only concerned with switching and sensing the 'low' (0) or 'high' (1) states and so can be simple and very fast indeed.
Consequently, in practice, all electronic representation of digital information is done with binary numbers - numbers to the base 2, rather than the decimal system. For instance, binary 0000 equals decimal 0, binary 0001 equals decimal 1, binary 0010 equals decimal 2 and binary 0101 equals decimal 5.
Each one or zero is called a 'bit'. Very often, bits are organized as a group of 8 bits, such as 00100011. This is known as a 'byte' - and 4 bits is known as a 'nibble'. The binary number 00100011 is 67 in decimal terms. Working from the right-most bit, it has one 1, one 2, no 4, 8 or 16, one 32, and no 64 or 128.
There are 256 possible combinations of the 8 bits in a byte - from 0 (00000000) to 255 (11111111) - so the standard telephone analogue to digital conversion described above produces 8,000 bytes per second - which is 64,000 bits per second. CD quality digital audio, in contrast, produces two bytes for each of two channels, 44,100 times a second - 176,400 bytes per second or 1,411,200 bits per second.
For westerners, another important form of information is text. There is an almost universally used standard called ASCII which relates each character, punctuation mark and non-printing character (such as tab and carriage return) in the English alphabet to a specific 8 bit binary number. For instance, the letter 'C' is represented by the binary number 01000011. This makes it convenient to represent each character of text as a byte. So the text on this page can be represented with a few thousand bytes, and the complete Bible, or the 'Lord of the Rings', requires about 4 million bytes - 4 Megabytes or 32 Megabits.
Back to Main Menu Next Page
|