Scribd
Upload
109 views

Lecture 10 - Pulse Code Modulation II

This document summarizes key concepts about pulse code modulation (PCM) including quantization, uniform quantization, quantization noise, and non-uniform quantization. Quantization maps a continuous range of input values to a smaller set of discrete levels, which introduces quantization noise. Uniform quantization distributes noise evenly but performs poorly for low amplitude signals. Non-uniform quantization uses finer quantization levels for low amplitudes to improve the signal-to-noise ratio. Methods like companding use logarithmic compression and expansion to enable non-uniform quantization.

Uploaded by

nabeel hasan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
109 views

Lecture 10 - Pulse Code Modulation II

This document summarizes key concepts about pulse code modulation (PCM) including quantization, uniform quantization, quantization noise, and non-uniform quantization. Quantization maps a continuous range of input values to a smaller set of discrete levels, which introduces quantization noise. Uniform quantization distributes noise evenly but performs poorly for low amplitude signals. Non-uniform quantization uses finer quantization levels for low amplitudes to improve the signal-to-noise ratio. Methods like companding use logarithmic compression and expansion to enable non-uniform quantization.

Uploaded by

nabeel hasan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 11

COMSATS Fall 2014 (Rev. 1.

0)

Principles of Communication Systems (PCS)


EEE 351

Lecture # 10
Pulse Code Modulation (PCM) II
COMSATS PCS

Quantization
• Quantization is a process of • A uniform/linear quantizer divides
mapping a large (usually infinite) the entire input signal range into
set of input values to a smaller set equal number of levels known as
of countable values. quantization levels denoted by ‘L’.

• The basic principle used in • Most of the signals to be quantized


quantization is rounding the input have both positive and negative
values to predefined levels. values so quantizer also assumes
the positive and negative values.
• It is inherently an irreversible
process fundamentally because of • The step size between the
the fact that a same output value quantization levels is known as
is shared by multiple input values. quantile interval (denoted as ‘q’).
COMSATS PCS

Quantization

Bernard Sklar, Digital Communications- Fundamentals and Applications, 2nd Edition.


COMSATS PCS

Quantization

Bernard Sklar, Digital Communications- Fundamentals and Applications, 2nd Edition.


COMSATS PCS

Quantization Noise
• Quantization can be referred to as • For uniform quantization,
‘rounding off’ which quantization error (e) is uniformly
approximates/rounds off an input distributed and can be
value to a predefined level. represented as a uniform
probability density function.
• Due to this rounding off, some
useful information is lost which in • The average quantization noise
turn causes noise/distortion in the power equal to its variance, σ2 .
output signal. This noise is know
as ‘Quantization Noise’. • Calculate signal to quantization
noise power and observe the
• Quantization noise is inversely relationship between the
proportional to the number of quantization levels and noise.
quantization levels.
COMSATS PCS

Uniform Quantization
• The quantizer in which all the • If the SNR is calculated for
quantile intervals are equal is known uniform quantizer, the SNR for
as the ‘Uniform Quantization’. low amplitudes signals is poor as
compared to the high amplitude
• For a uniform quantizer, the signals. This is because of the
quantization noise is equal for all the uniform distribution of the noise
levels. across all quantization levels.

• For speech signals, low volume


(Amplitude) signals occur for 50 % of
the time. Usually the signal
amplitudes are less than ¼ of the
RMS value. Only 15% of the time,
the signal value exceed the RMS
value.

Bernard Sklar, Digital Communications- Fundamentals and Applications, 2nd Edition.


COMSATS PCS

Non-Uniform Quantization
• In order to prevent the poor signal to quantization noise ratio, non-uniform
quantization is used.

• In non-uniform quantization, the quantization levels for low amplitude speech signals
are finely distributed which improves the SNR by reducing the quantization noise.

Bernard Sklar, Digital Communications- Fundamentals and Applications, 2nd Edition.


COMSATS PCS

Non-Uniform Quantization
• One way to perform the non- • This compression and expansion
uniform quantization is by first characteristics pair is known as
distorting the input signal with ‘Companding’.
logarithmic compression and
then using the uniform • There are two types of
quantization. companding laws known as ‘µ-
Law’ and ‘A-Law’.
• At the receiver, the inverse • ‘µ- Law’ and ‘A-Law’?
compression characteristics
known as expansion is applied to
recover the signal.
COMSATS PCS

Non-Uniform Quantization

Bernard Sklar, Digital Communications- Fundamentals and Applications, 2nd Edition.


COMSATS PCS

Encoding

Bernard Sklar, Digital Communications- Fundamentals and Applications, 2nd Edition.


COMSATS PCS

References
[1] Bernard Sklar, Digital Communications- Fundamentals and Applications, 2nd
Edition.
[2] Wikipedia [Online] Wikimedia Foundation. [Cited 06 ,03,2014]
http://en.wikipedia.org/wiki/Quantization_(signal_processing)

You might also like