The telephone system has been called the world’s largest machine. We can talk on the phone with the whole world. All that is needed is the right telephone number to reach the party at the other end with whom we wish to speak. Obtaining the number may require personal contact, a telephone book, advertising or surfing the Internet.
In order for a telephone call to take place, a connection must be established and kept open. Thereafter, the call itself must be transmitted in both directions. Sometimes, particularly with international calls, we may notice that it takes some time, usually just a few moments, before the ring signal is heard at the other end. Before this is heard, the telephone station closest to your telephone (in terms of network electronics) has detected that you have lifted the receiver. The station checks for this event about ten times each second. When the entire number has been dialed, other stations are polled and a path for the call is assembled. The stations along this route must cooperate, and a routing for the call should preferably be chosen that avoids traffic jams.
Your call, which must be amplified along the way so that it is not lost among noise and interference, may have its own exclusive channel, but it shares the physical transmission medium with many other calls being carried over the same cable, optic fiber or radio link. Your call may even take a trip out in space, where it is forwarded by a satellite. Reserving an entire channel for a single call, however, is wasteful. Although you may perceive that the call is continually in progress, it requires only a very small portion of the channel’s potential capacity. The rest is unutilized.
This is a description of an analog connection, in which the signal directly corresponds to the pitch and level of your voice and is modulated on to a carrier wave in a simple and direct manner. Another alternative is to code the signal, which not only saves transmission capacity, but also ensures better sound quality. How can these two opposites be combined? By using a computer, of course, which converts the analog signal to digital codes.
Digitization makes it possible to take samples of the speech signal and to transmit the numerical values corresponding to the sound level. If these sample values are taken often enough, the human ear cannot detect any difference when the speech signal is reconstructed again. Because the signal is now digital, consisting of zeroes and ones, it is much more robust and able to withstand noise and interference. It is also possible to add check sums and to retransmit if a problem occurs. Digital signal processing occurs so quickly that you do not notice the difference when you are talking.
Capacity can now be increased by sending several signals that are digitally combined with each other. The four telephone calls A, B, C and D can be interwoven so that they are sent as ABCDABCDABCD. Because not all calls are going to the same destination, other calls can be added at points along the way between the two parties participating in a call. After a while, call A may be placed in a new mix, such as ACGFACGFACGF, which later becomes AFHPAFHPAFHP. It is even possible to change the route along the way, particularly if the call is divided into packet. This is the way the Internet works.
Nothing comes for free, however. If cables, wires, fibers and radio links can be utilized much more efficiently, it is because computing power has been added along the way. The signals are now digital, and they must first be digitized and then converted back to their original form. Processing power in the telephone network is thus essential, and not before semiconductor technology and microprocessors became cheap enough were telephone companies able to switch from analog to digital.
Author: Bengt-Arne Vedin