Data Coding

1
Data Coding
2
Introduction

• Computing devices use energy to represent data
  and power the devices.
• Consequently, computing devices are electrical
  (power to operate) and electronic (energy to
  represent data). The data can be represented with
  electricity, light, radio waves, microwaves, or
  other energy in the electromagnetic spectrum.
• The data are represented in a two state form
  called binary. The light or electricity is either
  on or off. Symbolically, these states are
  represented by 0s and 1s called bits, the
  smallest units of information a computer can
  store.

3
Introduction

• By themselves, the bits are not very useful
  because they can represent only two states.
• Grouping the bits, however, allows for numerous
  combinations.
• A group of three bits has eight combinations and
  a group of eight bits has 256 combinations.
• Grouping, therefore, allows associations to be
  created with specific items such as characters,
  numbers, or commands. This association is called
  coding.

4
Early Codes

• A code is a predetermined set of symbols that
  have specific meanings.
• Morse Code Developed by Samuel Morse in 1838
  for use in telegraph communications.
• The code is a sequence of dots and dashes.
• A unique aspect of this system is that the letter
  codes have varying lengths
• the letter E corresponds to a single dot and
  the letter H has four dots.
• The varied code length allows messages to be sent
  more quickly than codes with the same length for
  each letter.

5
Baudot, Morse, and BCD Codes
6
Early Codes

• Baudot code- It uses 5 bits for each character
  and letter and was developed by Jean Baudot for
  the French telegraph.
• How many combinations does 5 bits allow?
• How many letters and digits are there?
• Are there duplicates?
• If yes, how can you determine the difference
  between a letter and a digit?
• A shift down (11111) and shift up (11011) are
  used. Upon receiving a shift down all subsequent
  codes are interpreted as letters until a shift up
  is received.

7
Machine Codes

• The telegraph codes are satisfactory for human
  use, but not for computers. Computer codes need
  the following attributes
• binary format
• all characters have the same length
• all the bits are perfectly formed
• all bits are of the same duration

8
Specific Computer Codes

• Binary-Coded Decimal (BCD)
• American Standard Code for Information
  Interchange (ASCII)
• Extended Binary Coded Decimal Interchange
  (EBCDIC)
• Unicode (16 bit)

9
Specific Computer Codes

• Binary-coded decimal- developed to facilitate the
  entry and computation of numeric data.
• Base 10 digits one (0001) through nine (1001)
  were represented in 4 bits in binary format and
  zero was represented as 1010.
• The digits were stored in BCD and calculations
  were conducted in BCD.
• As character data became more important, the BCD
  code was expanded to include other characters.
  The expanded code was eventually called
  binary-coded decimal interchange code (BCDIC).

10
Specific Computer Codes

• ASCII code- The most widely used code today.
• It has seven and eight bit code versions that are
  used to assign unique combinations of bits to
  each keyboard key stroke and special printable
  and non-printable characters. The non-printable
  characters include the line feed, tab, or
  carriage return. In the 8-bit or extended version
  of ASCII, special characters such as the corner
  of a box are included.

11
Specific Computer Codes

• EBCDIC code- used primarily on IBM mainframe
  computers and peripherals.
• It is an 8-bit code that allows 256 characters.
  Like ASCII, it has printable and non-printable
  characters.

12
Specific Computer Codes

• Unicode- With the internationalization of
  networking applications, the ASCII and EBCDIC
  codes have become too inflexible. Unicode was
  developed to address these issues.
• It is a 16-bit coding scheme that supports many
  scripts, collections of mathematical symbols, and
  special characters that exist in particular
  languages.
• Examples include Arabic, Latin, Greek, Gothic,
  and Cyrillic scripts. Unicode has already defined
  codes for more than 90000 characters.
• A special extension mechanism has been created
  encode more than 1 million characters.

13
Data Compression/Compaction

• Definition Data compression is reducing the
  amount of data or bits moving across a network
  connection. The compression improves bandwidth
  utilization and the speed of transmission.
• The key to data compression is to determine if
  there is redundancy in the original data and then
  eliminate it. Redundancy can exist in almost any
  type of data.
• Examples include reoccurring letters, numbers or
  pixels. Compression codes that attempt to reduce
  redundancy are frequency-dependent codes.

14
Data Compression/CompactionFrequency-Dependent
Codes

• Huffman Code-
• Redundant data are counted and frequencies
  calculated.
• Then a Huffman code is assigned to each piece of
  data such as a character. Characters with high
  frequencies are represented with short bit
  strings and those with lower frequencies are
  represented with longer bit strings.

15
Frequencies for the Letters A Through E

• Letter Frequency()
• A 25
• B 15
• C 10
• D 20
• E 30

16
Rules for Creating a Huffman Tree

• Associate a one node binary tree with each
  character and assign the characters frequency or
  weight.
• Look for the 2 lightest weight trees. If there
  are more than 2, choose among them randomly.
  Merge the 2 selected trees into a new tree with a
  new root node whose right and left subtrees are
  the selected trees. Assign the sum of the weights
  to the new tree.
• Repeat step 2 until only one tree remains.

17
Merging Huffman Trees
18
Huffman Codes for the Letters A Through E

• Letter Code
• A 01
• B 110
• C 111
• D 10
• E 00

19
Receiving and Interpreting a Huffman-Coded Message

• Example
• Bit Stream Transmission
• ?---------------------------
  --
• (01110001110110110111)
• First Sent -? A B E C A D B C ?-
  Last sent
• Does it work?
• (Notice that there is no space or any separator
  in the bit stream.)
• Why?

20
Summary of Huffman Trees

• Huffman coding assumes that receiver/sender
  computers have tables of codewords and
  corresponding ASCII codewords.
• This information must be sent before the
  transmission. This is an overhead for this coding
  scheme.
• Huffman codes reduce the number of bits to send,
  but they require that frequency values be known.
  These codes work best with repeatable patterns of
  bits such as those found in character data.

21
Run-Length Encoding

• Many items that are transmitted over networks,
  such as binary files, fax data, and video
  signals, are not like character data therefore,
  some other form of encoding is needed.
• An alternative is run-length encoding.
• It uses a simple approach that analyses bit
  strings looking for a long run of a 0 or 1.
  Instead of sending all the bits, it sends only
  how many bits are in the run.
• This techniques is especially useful for fax
  transmissions, which have 70 to 80 white space.

22
1st Example of Run-length Encoding

• Consider a screen containing plain black text on
  a solid white background. There will be many long
  runs of white pixels in the blank space, and many
  short runs of black pixels within the text.
• Let us take a single scan line, with B
  representing a black pixel and W representing
  white
• WWWWWWWWWWWWBWWWWWWWWWWWWBBBWWWWWWWWWWWWWWWWWWWWWW
  WWBWWWWWWWWWWWWWW
• If we apply a simple run-length code to the above
  scan line, we get the following
• 12WB12W3B24WB14W

23
2nd Example of Run-length Encoding
24
Character Stripping

• Removes the leading and trailing control
  characters from a message, and inserts them again
  at the receiving device.
• This technique is often used with character
  compression and run-length encoding.

25
MNP5

• MNP5 is a compression algorithm that is a
  combination of Huffman coding and run-length
  encoding.
• It is often implemented in the hardware of
  transmission equipment.