MATH/MTHE 477/877
Data Compression and Source Coding
Winter 2018
 Course Outline

Efficient transmission and storage of information is of critical
importance in many branches of science
and engineering. The means by which to achieve this is source coding
(a.k.a. data compression), a discipline
that studies the compact representation of information bearing signals
(such as text, speech, still image, and
video) for the purpose of storage or transmission. Source coding is
part of the general theory of communication,
and is closely related to and information theory, signal processing,
as well as probability and random
processes.
In this course the fundamentals of the theory and practice of data
compression will be studied. The following is a list of topics
that will be covered in more or less detail.
 Fundamentals of RateDistortion Theory: The
ratedistortion
function and its properties, Shannon's lossy source coding theorem, calculation of the ratedistortion function, ratedistortion function for sources with memory.
 Lossy Information Transmission:
Lossy joint sourcechannel coding theorem, Shannon's separation
principle, Shannon limit of communication systems.
 Scalar Quantization: uniform and nonuniform quantization, companding quantization,
predictive quantization, speech coding fundamentals, CELP.
 Frequency Domain Coding: Transform coding, bit
allocation, subband coding, wavelet coding, image coding
fundamentals, JPEG, JPEG2000.
 Vector Quantization:
Optimality conditions, design algorithms (LloydMax and related
methods), lattice quantization, high resolution distortion, advantages of vector quantization
over scalar quantization, robust vector quantization
for noisy channels and lossy joint sourcechannel coding.
Back to: [MATH 477/877
homepage]