Transmission Aspects Of Voice Telephony

Chapter Objective

 The goal of this chapter is to provide the reader with a firm foundation of the analog voice channel. Obviously, from the term, we are dealing with the transmission of the human voice. Voice is a sound signal. That sound is converted to an electrical signal by the mouthpiece of the subscriber subset. The electrical signal traverses down a subscriber loop to a local serving switch or to a PABX.1 The local serving switch is the point of connectivity with the PSTN.

With the exception of the subscriber plant, the PSTN has evolved to an all-digital network. In most cases the local serving switch is the point of analog-to-digital interface. Digital transmission and switching are discussed in Chapter 6. However, there are still locations in less-developed parts of the world where digital conversion takes place deeper in the PSTN, perhaps at a tandem exchange. Local service inside the local serving area in these cases remains analog, and local trunks (junctions) may consist of wire pairs carrying the analog signals. Analog wire pair trunks are even more prevalent outside of North America.

In this chapter we will define the analog voice channel and describe its more common impairments. The subscriber subset’s functions are reviewed as well as the sound to electrical signal conversion which takes place in that subset. We then discuss subscriber loop and analog trunk design.

Since the publication of the first edition of this text, the traffic intensity of voice versus data have changed places. Whereas voice traffic was predominant in the PSTN, data traffic intensity currently is far greater. Some sources state that only some 5% of the network traffic is voice and the remaining 95% is now data in one form or another. Much of the remainder being data traffic in one form or another. One such form is voice over IP which has the potential for astronomic jumps in traffic intensity. Do we count that as voice or data? It should also be kept in mind that the connectivity to the PSTN—that is, the portion from the user’s subset to the local serving exchange—will remain analog for some time into the future. Here again we may have problems with definitions. The digital subscriber line (DSL) continues increased presence. One can argue whether it is digital or analog. ISDN, of course, is another notable exception.

Definition Of The Voice Channel

The IEEE (Ref. 1) defines a voice-band channel as “a channel that is suitable for transmission of speech or analog data and has the maximum usable frequency range of 300 to 3400 Hz.” CCITT/ITU-T also defines it in the range of 300–3400 Hz. Bell Telephone Laboratories (Ref. 2) defines it in the range of 200–3300 Hz. We remain with the CCITT/IEEE definition.

The Human Voice

Human voice communication depends on the voice-generating mechanism of mouth and throat being the initial transmitter, with the acuity of the ear being the receiver. Frequency components of the human voice extend down to some 20 Hz and as high as 32,000 Hz. The lower frequency components carry the voice energy and the higher frequency components carry emotion. Figure 5.1 shows a distribution of energy and emotion of the typical human voice.

The human ear and many devices and components of the telecommunication network tend to constrain this frequency range. Young people can hear sound out to about 18,000 Hz, and as we get older this range diminishes. People in their sixties may not be able to hear sounds above 7000 Hz.

It is not the intent of the PSTN operator to provide high fidelity communications between telephone users, only intelligible connections. Not only is there the frequency constraint of voice communications brought about by the human ear, there are also constraints brought about by the subset transmitter (mouthpiece) and receiver (earpiece) and the subscriber loop (depending on its length, condition, and make up), but then purposely the electrical voice signal will enter a low-pass filter limiting its high-frequency excursion to 3400 Hz. The filter is in the input circuit of the multiplex equipment. Thus we say that the voice channel or VF (voice-frequency) channel occupies the band from 300 to 3400 Hz.

Figure 5.2 shows the overall frequency response of a simulated telephone network using the standard 500/2500 North American telephone subset.