DIGITAL REPRESENTATION OF ANALOG SIGNALS
SAMPLING
TIME DIVISION MULTIPLEXING
THE QUANTIZATION PROCESS
Amplitude Quantization is defined as the process of transforming the sample amplitude m(nTs) of a message
signal m(t) at time t = nTs into a discrete amplitude v(nTs) taken from finite set of possible amplitudes.
Here we will use the symbol ‘m’ in the place of m(nTs). The signal amplitude m is specified by the index k if
it lies inside the interval
Where L is the total number of amplitude levels used in the quantizer.
The amplitudes mk, k = 1, 2, …. L are called decision levels or decision thresholds.
At the quantizer output, the index k is transformed into an amplitude vk, k = 1, 2, 3,…. L. The amplitudes vk
are called representation levels or reconstruction levels and the spacing between the two adjacent
representation levels is called a quantum or step size. Thus the quantizer output v = vk if the input sample m
belongs to the interval Ik.
The characteristic of a quantizer is a staircase function.
A Quantizer can be Uniform or Nonuniform type.
Uniform Quantizer: Here the representation levels are uniformly spaces.
Non Uniform Quantizer: Here the representation levels are non uniformly spaced.
Types of Quantizer:
1. Midtread type:
The figure below shows the input output characteristics of midtread type of quantizer. This is so called
because the origin lies in the middle of the tread of the staircase graph.
2. Midrise type: The figure below shows the input output characteristics of midrise type of quantizer.
This is so called because the origin lies in the middle of the rising part of the staircase graph.
QUANTIZATION NOISE
The use of quantization introduces an error defined as the difference between the input signal m and
the output signal v. This error is called quantization noise.
The figure below illustrates variation of quantization noise as a function of time.
Let the quantization error be denoted by the random variable Q of sample value q. Thus we may write
Consider then an input m of continuous amplitude in the range (-mmax, mmax). Assuming the uniform
quantizer of midrise type, the step size of the quantizer is given by,
Where L is the total number of representation levels.
For uniform quantizer, the quantization error Q will have its sample values bounded by
-∆/2 ≤ q ≤ ∆/2
If the step size is sufficiently small, then quantization error Q is a uniformly distributed random
variable. Its probability density function is given by
This is known as Mean Square Quantization Error or Normalized Noise Power or Quantization Error
in terms of Power.
Let ‘R’ denote the number of bits per sample. Then the quantization level is given by
Let ‘P’ denote the average power of the message signal m(t). We may therefore represent the output
signal to noise ratio of a uniform quantizer as
The above equation shows that the output signal to noise ratio of the quantizer increases exponentially
with increasing number of bits per sample R.
Consider the modulating signal of the form m(t) = Am sin(2πfmt)
The average signal power of m(t) is P = Am2/2 and mmax = Am
Substituting these in equation 7, we get
Expressing output SNR in dB, we get
PULSE CODE MODULATION
In Pulse Code Modulation (PCM) a message signal is represented by a sequence of coded pulses which is
accomplished by representing the signal in discrete form in both time and amplitude.
The basic operations performed in the transmitter of a PCM system are:
1. Sampling
2. Quantization
3. Encoding
This is shown in the figure below.
The LPF prior to sampling is included to prevent aliasing of the message signal. The quantization and encoding
operations are usually performed in the same circuit which is called analog to digital converter. The basic
operations in the receiver are:
1. Regeneration of impaired signals
2. Decoding
3. Reconstruction
This is shown in the figure below.
1. Sampling: The incoming message signal is sampled with a train of narrow rectangular pulses. In order
to ensure perfect reconstruction of the message signal at the receiver, the sampling rate must be greater
than twice the highest frequency component W of the message signal. However a pre-alias filter is
used at the front end of the sampler to exclude frequencies greater than W before sampling.
2. Quantization: The sampled version of the message signal is then quantized thereby providing a signal
that is discrete in both time and amplitude. In certain applications it is preferable to use variable step
size i.e Non Uniform quantizer. In this case the step size increases as the input output charateristics
increases and vice versa. This means weak passages are given more protection than loud passages.
The use of non uniform quantizer is equivalent to passing the baseband signal through the compressor
and then applying the compressed signal to a uniform quantizer. There are two forms of compression
laws that are used in practice namely:
µ Law: This law is defined by
where m and v are normalized input and output voltages. And µ is a positive constant. This is
shown in the figure below.
The case of uniform quantization corresponds to µ = 0. We can see from the graph that µ law
is approximately linear at low input levels and logarithmic at high input levels corresponding
to
A Law: It is defined by
The practical values of A tend to be in the vicinity of 100. The case of uniform quantization
corresponds to A = 1. The figure below shows the above representation
In order to restore the signal samples to their correct relative level, a device must be used in the receiver
with characteristics complementary to the compressor. Such a device is called expander. The
compression and expansion laws are inverse of each other. The combination of a compressor and an
expander is called a compander.
3. Encoding: encoding process translates the discrete set of sample values to a more appropriate form of
signal. Representing each of these discrete set of values as a particular arrangement of discrete events
is called a code. One of the discrete events in a code is called code element or symbol.
In a binary code, each symbol may be either of two distinct values or kinds such as presence or absence
of a pulse. The two symbols of a binary code are denoted as 0 or 1. In a ternary code, each symbol
may be one of the three distinct values. However the effect of noise is minimum in binary code as it
withstands relatively high level of noise and is easy to regenerate.
Suppose a codeword consists of R bits, then it can be used to represent 2R distinct numbers. The table
below shows the binary number system for R = 4.
Need for Non Uniform Quantization
LINE CODES
DIFFERENTIAL ENCODING