## Converting Analog Voice to Digital Voice

The last step in understanding how the PSTN supports voice across a digital PSTN relates to how the PSTN converts the analog electrical signals to digital signals, and vice versa. To see the need for the conversion, examine Figure 15-4.

Figure 15-4 Analog Voice Calls Through a Digital PSTN

PSTN

PCM Codec Converts

Figure 15-4 Analog Voice Calls Through a Digital PSTN

PCM Codec Converts

When Andy calls Barney in Raleigh, the circuit is set up by the telco. (Yes, Barney moved to Raleigh since the last example.) And it works! It works because the phone company switch in the Central Office (CO) in Mayberry performs analog-to-digital (A/D) conversion of Andy's incoming voice. When the switch in Raleigh gets the digital signal, before sending it out the analog line to Barney's house, it reverses the process, converting the digital signal back to analog. The analog signal going over the local line to Barney's house is roughly the same analog signal that Andy's phone sent over his local line.

The original standard for converting analog voice to a digital signal is called pulse-code modulation (PCM). PCM defines that an incoming analog voice signal should be sampled 8000 times per second by the analog-to-digital (A/D) converter. A/D converters that are used specifically for processing voice are called codecs (meaning encoder/decoder). For each sample, the codec measures the frequency, amplitude, and phase of the analog signal. PCM defines a table of possible values for frequency/amplitude/phase. The codec finds the table entry that most closely matches the measured values. Along with each entry is an 8-bit binary code, which tells the codec what bits to use to represent that single sample. So PCM, sampling at 8000 times per second finds the best match of frequency/amplitude/phase in the table, finds the matching 8-bit code, and sends those 8 bits as a digital signal.

The PCM codec converts from digital to analog by reversing the process. The decoding process re-creates the analog signal, but not quite exactly. For instance, if the original frequency was 2139.3, the decoded frequency might be 2140. For normal speech, the quality is great. If you were trying to listen to DVD-quality sounds over the telephone, it probably wouldn't sound as good as it would if you were actually there, but it's pretty close.

If you do the math, you will notice that a single voice call requires 64 kbps of bandwidth in the digital part of the PSTN. PCM says that you need to sample the analog signal 8000 times per second, and each sample needs 8 bits to represent it. A bright fellow at Bell Labs, Nyquist, did some research that showed this sampling rate was needed for digitized voice. He noticed that the human voice could create sounds between 300 Hz and 3300 Hz, and that the sampling rate needed to be twice that of the highest frequency. So, to overcome some other physics problems, Nyquist and the team at Bell Labs decided to round that range of frequencies for the human voice to 0 Hz to 4000 Hz. So, because Nyquist's theorem states that you need twice the number of samples as the highest frequency, you need 8000 samples. To make sure the voice sounded good after being decoded, they decided to use 256 different binary values, each representing a different combination of amplitude, frequency, and phase. To represent the 256 values, they needed 8 bits; for 8000 samples per second, 64 kbps is needed for a PCM-encoded voice call.

Because a single call needs 64 kbps, the digital PSTN first was built on a basic transmission speed of 64 kbps. A single 64-kbps channel was dubbed a Digital Signal Level 0—or DS0. In the United States, the phone company (American Telephone and Telegraph [AT&T] by that point in its history) decided to create hardware that could multiplex 24 DS0s onto a single line, so it called that type of line a Digital Signal Level 1—or DS1. The more popular name for a DS1 today, of course, is T1. Some parts of the world followed AT&T's lead for DS1 with 24 DS0 channels, and other parts of the world, mainly Europe and Australia, chose instead to combine 32 different 64-kbps DS0 channels onto a single line, which is the basis for today's E1s. As you might imagine, even faster digital facilities are defined as well, such as a T3-line, which has 28 T1s in it.

Finally, this small history lesson comes to an end. Most of the work on modems and ISDN, and some of the work for DSL, occurred with the expectation that these technologies needed to work over the PSTN.

In summary:

■ The telco switch in the CO expects to send and receive analog voice over the physical line to a typical home (the local loop).

■ The telco converts the received analog voice to the digital equivalent using a codec.

■ The telco converts the digital voice back to the analog equivalent for transmission over the local loop at the destination.

■ The voice call, with PCM in use, uses 64 kbps through the digital part of the PSTN.