Information Theory: Mohamed Hamada

Download as pdf or txt
Download as pdf or txt
You are on page 1of 19

Information Theory

Mohamed Hamada
Software Engineering Lab
The University of Aizu

Email: hamada@u-aizu.ac.jp
URL: http://www.u-aizu.ac.jp/~hamada
Today’s Topics

• Watching a Coding Video (50 mins.)


• What is Information Theory
• Information Source
• Introduction to Source Coding
• What is Information Theory
Information Theory

- What is Information Theory?

- What is the purpose of information theory?

- Why we need to study information theory?

What is information theory (IT) ?

IT is the science that deals with the concept “ Information” :


Its measurement & its application

What is the purpose of IT?

Transmission of information in an efficient way : minimum time & space

Why we need to study information theory?

Because of the revolution of communication dealing efficiently with information and its
transmission becomes a necessary requirement for a computer engineer.
What is Information?

There are 3 types of Information

Syntactic Information Semantic information Pragmatic information

Related to the structure Related to the Related to the usage and


of the messages meaning of the effect of the messages
(characters) that forms messages
the information

Example:
i. I eat sushi by hashi
ii. By hashi I eat sushi
iii. There is a typhoon in Japan
iv. There is a typhoon in Kanto area in Japan
i. and ii. are syntactically different but semantically & Pragmatically equal
iii. and iv. are syntactically, semantically & pragmatically different
(iv. gives more information than iii.)
INFORMATION TRANSFER ACROSS CHANNELS

Sent Received
messages messages
symbols

source encoderChannel
Source
channel
channel
Channel
decoderSource receiver
coding coding decoding decoding

Compression Error Correction Decompression


Source Entropy Channel Capacity

Capacity vs Efficiency
Digital Communication Systems

Information User of
Source Information

Source Source
Encoder Decoder

Channel Channel
Encoder Decoder

Modulator De-Modulator

Channel
Information Source

Examples:

1. English text
2. A man speaking
3. Photographs
4. Motion of films,
....... etc.

A chief aim of information theory is to study how such sequence of symbols (signals)
can be most effectively encoded for transmission (by electrical means).

For Information source we have:

· Information expressed as a finite set of symbols : Source Alphabet

· A ( discrete) information source : is a source that generates a sequence of symbols

· The symbols are denoted by a1, a2, …., am and the alphabet by A = { a1, a2, …, am }

· A finite sequence of symbols is called word. The set of all words is denoted by A*
Information Source

Memoryless

Memoryless means the generated symbols (of a source message ) are


independent.

Stationary

The idea of stationary of a source demands no change with time

Discrete
The source produces independent symbols in different unit times
Memoryless

Memoryless means the generated symbols (of a source message ) are independent.

X
source

P(X = i | X1, X2, …, Xn ) = P ( X = i )

i.e. The probability of the output X is conditionally independent of previous channel


inputs or outputs X1, .., Xn

Example:

Coin toss 8 times : the probability to get head


Each time is ½ x …x ½ = (½)8 = 1/256 = 0.0039 = 0.4 %
Coin toss number 9 : The probability to get head is still ½

So it is independent of the previous 8 toss


Stationary

A process is called Stochastic if its output is associated with a probability distribution.

A stochastic process is said to be stationary when the process is (temporally


homogeneous ) remain invariant under every translation of the time scale

Example 1 : Assume a source produces an infinite sequence of the form:

Source AE AE AE AE AE etc.

i.e. What comes later is like what has gone before. Stationary is a designation of such
source of characters

Example 2 : Assume the source that produces:

Source AE AAEE AAAEEE etc.

such source is not stationary

The idea of stationary of a source demands no change with time

i.e P(X = i ) = C (constant)


Digital Communication Systems

Source of User of
Information Information

Source Source
Encoder Decoder

Channel Channel
Encoder Decoder

Modulator De-Modulator

Channel
Source Coding

· We consider memoryless, stationary, discrete information source S ,

a1 a2 …. am
S= where pi = p(Xn =ai)
p1 p2 …. pm

so that

P(Xn = ai) = C for a constant C

i.e. probability doesn’t depend on the trial ( n )


Source Coding

-Information source encoding (enciphering) :

is a procedure for associating words constructed from a finite alphabet


of a language with given words of another language in a one-to-one
manner.

i.e. encoding is a procedure for mapping a given set of messages


{m1, m2, ... , mi } onto a new set of encoded messages {c1, c2, ... , ci }
in a one-to-one manner.

-A measure of the encoded information is entropy :


which is the shortest mean encoding.
The unit measure of entropy is bit : binary digit.

Example: The entropy of a fair coin toss is 1 bit

Later on we will study entropy in more details.


Source Coding

-The goal of source coding is to make encoding as short as


possible

- Code : For the alphabet A = { a1,…., am }


a code C is a nonempty set of words, i.e. C is a subset of A*

-Code word : Is an element of the code

- Code word length : Is the number of symbols is the code word.


For a code word x, l(x) denote its length
Source Coding

Example : for the source information { Red, Blue} and the code alphabet is {0,1 }.
Let C (Red) = 00 and C (Blue) = 11
Then C = { 00, 11 } is a subset of { 0, 1 }* and the length is l( 00 ) = 2 and l(11) = 2

- Binary Source Coding : Is the code whose alphabet is { 0, 1 }


Decoding

Is the inverse process of encoding.


i.e. The operation of assigning code words to the corresponding words of the
source information.

ck Source sk Information
Decoder Destination
Decoding

Is the inverse process of encoding.


i.e. The operation of assigning code words to the corresponding words of the
source information.

Various Coding Techniques :

- Comma Code : is a code C with the code words


1, 01, 001, 0001, ….., 0…01, 000…0
where the last two codes have n-1 and n 0's respectively.

-Variable length Code : is the code with codeword of different lengths.

- Instantaneous code : is the code in which each codeword in any string


of codewords can be decoded (reading from left to right ) as soon as it is received.

- Singular Code : A code is singular if some codewords are not different


Examples of Information Source Coding

Symbol Pro C1 C2 C3 C4 C5 C6
A 0.4 00 1 1 1 0 0

B 0.3 01 01 01 00 1 0

C 0.2 10 001 001 010 10 10

D 0.1 11 0001 000 011 01 01

MCL 2 2 1.9 1.9 1.3 1.3


MCL = Mean code length

C1 : fragile ( recovery impossible if letters drop out)


C2 : robust ( recovery possible if letters drop out);
variable-length code (non-equal length code );
C3 : smarter comma code (incorporating maximum length)
C3, C4 : also (variable- length) instantaneous codes;
C5 : decoding impossible;
D A A C
010 01 0 or 0 10
C6 : singular code (so decoding impossible)
Coding/Decoding Example

Bad Code
Example: Code I
One input can be interpreted
to many different outputs.
Source Binary Code
Symbol Representation

a 0
Information aad Source 0011
Source Encoder b 1

Error Free c 00
aad or, Channel
aabb or,
Source d 11
cd or, Decoder
cbb 0011

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy