Analog vs. Digital
Most twentieth-century radio systems were designed to transmit sound, which is an analog signal: it varies continuously and is usually represented as a wave. Because electromagnetic radiation is also a wave, this made transmitters and receivers relatively simple. Whether in broadcasting or cellular telephony, a radio terminal is a device for converting sound waves to radio waves and back again. Television uses the same principle, including light as well as sound.
Wireless networks are increasingly being used for computer data, which is inherently digital. Instead of a continuous waveform, it consists of a series of pulses. Many sound and video applications are also beginning to use digital signals, though new digital services are hampered by the large installed base of analog equipment. Figure 2.4 shows the difference between analog and digital.
Encoding analog information as data requires sophisticated electronics, which were not portable or affordable before the 1990s. This means that digital is more likely to be used for newer services, such as mobile telephony. Analog broadcast radio and television reached saturation coverage much earlier, so their move to digital has been much slower. The first digital radio and TV services were launched in Britain in 1999, but radio in particular was hampered by the high cost of reception equipment. TV proved slightly more popular because receivers were subsidized by companies eager to use digital's extra capacity for pay-per-view movies and sports events.
Figure 2.4. Analog and digital waveforms
Digital has several advantages over analog. Among them are
Noise reduction. All communication channels suffer from static interference, a particular problem for wireless networks. With an analog waveform, there is no way for the receiver to distinguish between this noise and the actual signal. A digital waveform is different; it can have only two levels, so anything in between can automatically be discarded by receivers or relay stations. While every piece of interference has an effect on an analog signal, it would have to be very severe to prevent a digital signal from being received intact.
Reliability. Digital signals can be encoded with extra bits called checksums. These are the result of a mathematical calculation performed on the preceding bits; they allow a receiver to check that it has interpreted the signal accurately. If the checksum is wrong, at least one bit has been lost in transmission, so a portion of the signal has to be sent again.
For greater reliability, additional error-correction code can be sent along with the signal, which allows some lost bits to be reconstructed if others are intact. This is known as FEC (Forward Error Correction) because it anticipates errors in advance of their occurrence. The simplest method is to send each bit and its checksum twice, though most systems use more complex formulae that combine the checksum with extra bits. An example is the Hamming code, which calculates a three-bit checksum for every four bits of a signal. A single error in the resultant seven bits can then be corrected, though two or more require a resend.
The obvious problem with checksums and FEC is that they reduce capacity. The more extra bits being sent, the less room there is for real data. The extra bits are known as redundancy; in general, greater redundancy means a more reliable connection. There is a trade-off between capacity and reliability, and wireless network planners need to make a choice based on the importance of their data and the time taken to resend it.
Spectral Efficiency. Thanks to its greater resilience to errors, a digital system can transfer more information than analog over a given amount of spectrum. The digital version of AMPS (Advanced Mobile Phone System) used in many parts of the U.S. carries three conversations in exactly the same frequency bands that the analog version uses for just one.
Digital signals also allow compression, which reduces the amount of capacity needed to send data by looking for repeated patterns. Compression is commonly used in all forms of data transmission, particularly for images, where it is built in to the file formats used by Internet browsers.
Security. Wireless systems are very open to eavesdroppers. A private conversation can easily be picked up by anyone with a suitable radio, so analog cellular was a boon to the nosy and to muckraking journalists. Simple scanners are readily available in electronics shops and are perfectly legal in most countries. They have been used in a few notorious incidents, most famously when Princess Diana and other royals had their affairs exposed across the world's press.
To prevent such scandals, modern digital cellular encrypts all data, using mathematical algorithms. Encryption is possible using analog equipment, but is often unreliable and easy to crack; hence the surge in illegal descramblers for analog cable and satellite TV. Digital signals can be encrypted to arbitrary degrees, depending only on the processing power of the transmitter and receiver. No matter what strength the encryption, an encrypted digital signal uses no more capacity than an unencrypted one.
Timing. Digital information can easily be stored in computer memory, allowing a communication channel to be shared in sophisticated ways. In a TDMA (time-division multiple access) system, each user of a channel uses it for only part of the time, so the data has to be held in memory until it can be sent. A packet-switched system uses capacity only when required, storing data in the memory of a special computer called a router, while calculating the most efficient path to send it. Packet-switching is the basis of the Internet and of the newest wireless networks.
normal: Bits and BandwidthThe term bandwidth originally referred to radio spectrum and described the frequency range allocated to a particular technology or service. For example, the channel between 824 MHz and 893 MHz, commonly used for cellular services in America, has a bandwidth of 69 MHz (because 893-824=69). The channel from 92.7 MHz to 92.9 MHz, used for a radio station, has a bandwidth of 200 kHz. The amount of digitized information that can be sent over a channel is related to its bandwidth through an equation known as Shannon's Theorem, or simply the Bandwidth Theorem. The two are directly proportional, so that in the example above, the cellular network could carry 345 times as much information as the radio station—hardly surprising, as it has to handle many two-way conversations at the same time, while the station has to cope with only a single broadcast. AT&T mathematician Claude Shannon devised the theorem in 1948, as part of a groundbreaking scientific paper called "A Mathematical Theory of Communication." At the same time, he defined the word bit (binary digit), which represents the smallest possible unit of information. To a computer, this would be a switch with two positions, usually represented as 1 and 0. Shannon's Theorem is used so much by engineers that the word bandwidth has come to mean the data capacity of a channel, measured in bits per second. However, the data bandwidth is not always the same as the spectrum bandwidth. It also depends on the signal-to-noise ratio of a channel; the more noise or interference, the less data can be sent. The information content of a message increases exponentially with the number of bits: one bit can represent only two possibilities, two bits can represent four, three would give eight, and so on. This means that for each bit that can't be read because of interference, the information content of the message is halved. |