PULSE CODE MOULATION (PCM)

COURSE CONTENTS
S/No. Topics 1. Introduction of TDM 2. Mentis of TDM 3. Working principle of P.C.M system 3.1. Sampling 3.2. Quantization 3.3. Types of Quantization 3.4. Encoding/Decoding 3.5. Construction of 2 M bit/s pulse frame & Multi-frame 3.6. AMI and HDB3 Line Codes 3.7. A and Laws 3.8. PCM transmission systems and corresponding Characteristics 4. Function and description of regenerative repeater 17 18 Page No. 4 7 8 8 9 10 11 13 15 17

1. INTRODUCTION OF TDM
When telephone communication began individual connecting paths were used i.e. separate pair of wires was used for every telephone connection. This was known as space-division multiplex (SDM) on account of the fact that multitude of lines were arranged physically next to each other. Since a particularly large proportion of capital is invested in the line plant, efforts were made at an early stage to make multiple use of atleast those lines used for long-range communications. This led to the introduction of frequency-division multiplex (FDM). This involves subdividing a wide frequency band into narrower sub-bands. Fig. 1 shows a 48 KHz band subdivided in to 12 sub bands. The sinusoidal signal of a sub-band (carrier) is modulated by a telephone signal. Since a sinusoidal signal acts as the carrier for a telephone signal. This process is known as carrier transmission. Following demodulation on the receive side the telephone signals are again available at their original frequencies. It is not the only way of making multiple uses of lines however. Another possibility is offered by time-division multiplex (TDM). Here the transmitted telephone signals are separated in time. Fig. 2 shows a period containing 32 time slots each time slot is approximately of 3.9 s time. This sub-division is repeated every 125 s in consecutive periods. One time slot in each of the consecutive periods is allocated to each telephone signal. In TDM, the analog or Digital samples of a number of telephone channels are transmitted over the same line.

Fig.1 Frequency Division Multiplex

subdivided into 32 time slot, each approx. 3.9 s Fig.2 Time Division Multiplex

The principle of time-division multiplex is based on the theory that a complete waveform is not required in order to transmit signals such as those encountered in telephony. It is sufficient to sample the waveform at regular intervals and to only transmit these samples (see Fig. 3). When a waveform is sampled, a train of short pulses is produced. The amplitude of each pulse represents the amplitude of the waveform at the specific sampling instants. This conversion is known as pulse amplitude modulation (PAM). The envelope to the PAM signal reflects the original form of the curve (see Fig. 4). Relatively large intervals occur between each sample. These intervals can be used for transmitting other PAM signals, i.e. the samples of several different telephones signals can be transmitted one after the other in repeated cycles. When the pulses of several PAM signals are combined they form a PAM time-division multiplex signal (see Fig. 5)

Fig. 3 Periodic sampling of the analog telephone signal a

Fig.4. PAM signal consisting of the samples of analog telephone a

Fig.5. PAM time-division multiplex signal consisting of samples taken from the three analog telephone signals, a, b and c in repeated cycles.

Fig.7. PCM time division multiplex signal consisting of the coded samples of analog Telephone signals a, b. If the waveform samples, i.e. the pulses with differing amplitudes, are converted to binary character signals, the term pulse code modulation (PCM) is used. During this process the pulse-like samples are quantized and coded 8 bits are normally used here. The digital signal in Fig.6 is shown in simplified form with 4-bit character signals (PCMwords) instead of 8-bit PCM words. When the PCM signals of several telephone signals are interleaved they produce a PCM time-division multiplex signal (see Fig.7).

2. MERITS OF TDM.
Digital transmission has number of merits over analog. Some of these advantages are:

1.

Transmission quality is very largely unaffected by distance and environmental change.

2.

Process of regeneration at suitable intervals largely minimizes these accumulations.

3. 4. 5. 6.

Expensive filters are not required in the multiplexing equipment. Testing procedures are simplified. High noise immunity PCM time-division multiplex signals permit the multiple use of lines and electronic circuits.

7.

PCM signals are much less sensitive to interference than are analog signals (e.g. PAM signals).

8.

Low space requirements.

3.

Working principle of PCM.

3.1 Sampling.
Sampling theorem. The sampling theorem is used to determine the minimum rate at which an analog signal can be sampled without information being lost when the original signal is recovered. The sampling frequency (fA) must be more than twice the highest frequency contained in the analog Signal (fA): fA>2fs A sampling freq (fA) of 8000 HZ has been specified internationally for the freq band (300 HZ-3400 HZ) used in telephone system i.e. the telephone signal is sampled 8000 times per second. The interval between two consecutive samples from the same telephone signal (sampling interval = TA) is calculated as follows. TA = 1/fA = 1 8000 HZ = 125us

Fig. 8 Generation of a PAM Signal Fig shows telephone signal is fed via L.P.F. to an electronic switch the L.P.F. limits the freq band to be transmitted. It suppresses frequencies higher than half the sampling freq. The electronic switch driven at the sampling freq of 8000 HZ takes samples from the telephone signal once every 125 us. A pulse amplitude modulated signal is thus obtained at the output of the electronic switch a P.A.M signals

Fig. 9. The Sampling Process

Fig. 10 Representation of Sampling and Encoding

Fig. 11 Representation of Time Division Multiplexing

3.2

QUANTIZATION.

The first stage in the conversion to a digital signal in pulse code modulation is Quantization. The whole range of possible amplitudes values is divided in to quantizing intervals.

Uniform Quantizing Fig. 12 Quantization

Non Uniform Quantizing

In this Fig, 16 equal quartizing intervals are indicated, the quartizing intervals are numbered +1 to +8 in the +ve range of the telephone signal and 1 to 8 in the ve range. The quartizing interval is determined for each sample. Decision values form the boundaries between adjacent quatizing intervals. On the transmit side, therefore several different analog values fall within the same quantizing interval. On the receive side one signal value corresponding to the mid point of the quantizing interval is recovered for each quantizing interval.

Fig. 13 The quantizing process

3.3 Types of Quantization.

There are two types of Quantization. 1. 2. Uniform Quantization. Non-Uniform Quantization.

Quantization distortion decreases as the number of quantizing intervals are increased. If the quantizing intervals are made sufficiently small the distortion will be minimum and the noise imperceptible.

Uniform Quantization.
Quantization in which all the quantizing intervals are of equal size. In this type equal and large quantizing intervals are used over the whole amplitude range, relatively large discrepancies will occur in case of small signal amplitudes.

Non-Uniform Quantization.
Quantization in which the quantizing intervals are not all of equal size, small quantizing intervals are usually allocated to small signal values (Samples) and large quantizing intervals to large signal values to make the quantizing distortion ratio nearly independent of the signal level.

Fig 14. Uniform Quantization

Fig.15. Non-Uniform Quantization

3.4 Encoding/Decoding Encoding

The PCM signal to be transmitted is obtained by encoding the quantized intervals. The encoder allocates an 8-bit PCM word to each individual sample. In PCM transmission systems an 8-digit binary code is used for the 128 +ve and 128 ve quantizing intervals (128+128=256= 28 ). The PCM words therefore have 8 bits.

Fig.16. Encoding of quantized samples with 8 quantizing levels (3 binary digits/code word where the most significant bit is used to devote the sign).

Decoding.
On the receive side a signal amplitude V out is allocated to every 8 bit PCM word, it corresponds to the mid point of the particular quantizing interval. The characteristic for decoding is the same as that for non-uniform encoding on the transmit side. The PCM words are decoded in the order in which they are received and converted to a PCM signal, finally the PAM signal is fed to a low pass filter, which reproduces the original analog telephone signal.

3.5 Construction of 2 M bit/s pulse frame Pulse frame

8000 samples per second are transmitted as 8 bit PCM words in both directions for each of the 30 speech circuits. This means that, within a period of 125 s (reciprocal of 8 KHZ) 30 PCM words each with 8 bits, are transmitted consecutively in each direction. In addition to these 30 PCM words a further 28 bits are also transmitted, 8 bits for signaling and 8 bits which contain alternately a bunched frame alignment signal and a signal word. The 30 PCM words together with the other 28 bits form a pulse frame. Pulse frames are transmitted directly one after the other.

Fig-17 Frame structure of the 30 channels PCM system Multi frame.

In the 30 channel PCM system a multiframe is made up of 16 frames within a multiframe the signal time slot contains four bits each for signaling information of the 30 telephone channels. It is a set of consecutive frames in which the position of each frame can be identified by reference to a multiframe alignment signal.

Fig.18 Frame and Multiframe structure for CCITT30 channels PCM system

AMI and HDB3 Line codes

AMI code A code derived from a binary by inverting the polarity of the alternate pulses representing binary 1 and by representing binary 0 with a zero voltage. The AM1 code is therefore a pseudo ternary code. Pseudoternary signals are better suited for line transmission than are binary signals. The alternation between +ve and ve allows the regenerative repeaters to recover the timing signal required for their own synchronization.

Fig. 19. Conversion of AMI to HDB3 Code

HDB3 Code
Long sequences of zeros which would occur when a channel is idle must be avoided since the lack of transitions may allow timing recovery circuit to malfunction. A popular code family which over comes the timing recovery problems is known as high density bipolar (HDB3) In HDB3 code, the maximum number of zero is three. Conversion from AMI to HDB3 involves the following rules. When four consecutive zeros occur, the code is changed to 000 V. Note that is V a violation pulse having the same polarity as the preceding pulse of the AMI sequence. An odd number of pulses occur between successive V-pulses to ensure the successive V pulses are of opposite polarity. An additional pulse is added so that the preceding rule is achieved. This ensures that the mean value of the signal is zero.

Fig. 20 Code forms of AMI and HDB3

3.7 A and Laws 3.7.1 A Law

Law for specifying the 13-segment characteristic for non- uniform quantizing in PCM codecs. Recommended by the CCITT for PCM30 transmission systems.

3.7.2 Law
Law for specifying the 15-segment characteristic for non-uniform quantizing in PCM codes. Recommended by the CCITT for PCM 24 transmission systems.

3.7.3 A/ Law Conversion

Used on international connections when a conversion has to be made from a PCM30 transmission system to a PCM24 transmission system and vice versa. The A/ law conversion is carried out in that country employing the -law.

3.8

PCM Transmission systems and Corresponding characteristics

The transmission systems recommended by the CCITT are the PCM30 system, with 2048 kbit/s (CCITT Recommendation G.732), and the PCM 24 system, with 1544 kbit/s (CCITT recommendation G. 733); these combine 30 and 24 telephone channels per transmission direction respectively to form a time-division multiplex signal. PCM30 transmission systems are used through out Europe and in many non-European countries. PCM 24-transmission systems have been installed mainly in the USA, Canada and Japan. PCM30 and PCM 24 are also known as primary transmission systems or basic systems. Their most important features are given in Table1. Common characteristics Sampling frequency No. of samples per telephone signal Pulse frame period No. of bit in PCM word Bit rate of telephone channel PCM30 and PCM24 8 kHz 8000/s 1/b =1/8000/s = 125 s 8 bit b . d =8000/s . 8bit =64 kbit/s PCM 24 - law 15 24

a. b. c. d. e.

System-specific characteristics f. Encoding/decoding No. of segments in characteristic g. No. of channel time slots per pulse frame h. No. of bits per pulse frame ( = additional bit) i. Period of an 8-bit channel time slot j. Bit rate of time-division multiplex signal

PCM 30 A-Law 13 32 d. g = 8 bit. 32 = 256 bit

d. g + 1 = 8 bit. 24+ 1 = 193 bits c.d/h=125s . 8/193 c.d/h = 125 s. 8/193 = ca 3.9 s = ca 5.2 s b. h = 8000/s. 256 bit = b. h = 8000/s 193 bit 2048 kbit/s =1544 kbit/s

Table 1. Characteristics of the PCM 30 and PCM 24 transmission systems

4.
4.1

Function and description of a regenerative repeater.

Function of Regenerator:

The function of a regenerator is to discriminate between the presence and absence of pulses in the received signal and to transmit a replica of the digital signal transmitted from the terminal or proceeding regenerator. A typical regenerator contains circuits to regenerate the line signal in both directions of transmission, the regenerator circuits for each direction of transmission being separate and independent of each other.

4.2

Description of Block diagram:

It contains an equalizer with amplifier, a timing recovery circuit and a regenerating circuit. The incoming and attenuated, distorted signal is fed to a preamplifier with a fixed equalizer, the amplifier is a two stage differential amplifier (wide band type) with a gain of 52 db. A detector detects a signal amplitude regulated circuit makes the amplitude of the digital signal independent of the line loss at equalizer out put by compensating attenuation of 40 db. Now this signal is fed to a timing recovery circuit via full wave rectifier which produces the retiming signal with the help of pulse synchronized OSC which is locked in phase The delay element use a capacitor to determine sampling instants in the time decision circuit. The +ve and ve pulses amplitude of 2 M bit/s signal are processed separately. It is sampled in time decision circuit with the help of recovered timing. During sampling if a binary digit 1 is detected a time decision circuit produces a new pulse and if a binary 0 is detected no pulse is produced. At the out put stage signal is again amplified up to 3 V P-0 and is transmitted to the next direction.