CS2100 Computer

Organisation
http://www.comp.nus.edu.sg/~cs2100
/
Number Systems and Codes
(AY2013/4) Semester 2

NUMBER SYSTEMS &

CODES
Information Representations

Number Systems
Read up DLD for details!
Base Conversion
Negative Numbers
Excess Representation
Floating-Point Numbers
Decimal codes: BCD, Excess-3, 2421, 84-2-1
Gray Code
Alphanumeric Code
Error Detection and Correction

CS2100

Number Systems and Codes

INFORMATION REPRESENTATION
(1/3)

Numbers are important to computers

Examples

Represent information precisely

Can be processed
Represent yes or no: use 0 and 1
Represent the 4 seasons: 0, 1, 2 and 3

Sometimes, other characters are used

CS2100

Matriculation number: 8 alphanumeric characters (eg:

U071234X)

Number Systems and Codes

INFORMATION REPRESENTATION
(2/3)

Bit (Binary digit)

0 and 1
Represent false and true in logic
Represent the low and high states in electronic
devices

Other units

CS2100

Byte: 8 bits
Nibble: 4 bits (seldom used)
Word: Multiples of byte (eg: 1 byte, 2 bytes, 4 bytes, 8
bytes, etc.), depending on the architecture of the
computer system

Number Systems and Codes

INFORMATION REPRESENTATION
(3/3)

N bits can represent up to 2N values.

Examples:

2 bits represent up to 4 values (00, 01, 10, 11)

3 bits rep. up to 8 values (000, 001, 010, , 110, 111)

4 bits rep. up to 16 values (0000, 0001, 0010, ., 1111)

To represent M values, log2M bits are required.

Examples:

CS2100

32 values requires 5 bits

64 values requires 6 bits
1024 values requires 10 bits
40 values how many bits?
100 values how many bits?

Number Systems and Codes

DECIMAL (BASE 10) SYSTEM

(1/2)

A weighted-positional number system

CS2100

Base or radix is 10 (the base or radix of a number system is the

total number of symbols/digits allowed in the system)
Symbols/digits = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 }
Position is important, as the value of each symbol/digit is
dependent on its type and its position in the number
Example, the 9 in the two numbers below has different values:

(7594)10 = (7 103) + (5 102) + (9 101) + (4 100)

(912)10 = (9 102) + (1 101) + (2 100)

In general,

(anan-1 a0 . f1f2 fm)10 =

(an x 10n) + (an-1x10n-1) + + (a0 x 100) +
(f1 x 10-1) + (f2 x 10-2) + + (fm x 10-m)
Number Systems and Codes

DECIMAL (BASE 10) SYSTEM

(2/2)

Weighing factors (or weights) are in powers of 10:

103 102 101 100 . 10-1 10-2 10-3

To evaluate the decimal number 593.68, the digit in each

position is multiplied by the corresponding weight:
5102 + 9101 + 3100 + 610-1 + 810-2
= (593.68)10

CS2100

Number Systems and Codes

OTHER NUMBER SYSTEMS

(1/2)
Binary (base 2)

Octal (base 8)

Weights in powers of 8
Octal digits: 0, 1, 2, 3, 4, 5, 6, 7.

Hexadecimal (base 16)

Weights in powers of 2
Binary digits (bits): 0, 1

Weights in powers of 16
Hexadecimal digits: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C, D, E, F.

Base/radix R:

CS2100

Weights in powers of R

Number Systems and Codes

OTHER NUMBER SYSTEMS

(2/2)
In some programming languages/software, special

notations are used to represent numbers in certain

bases

In programming language C

Prefix 0 for octal. Eg: 032 represents the octal number (32)8

Prefix 0x for hexadecimal. Eg: 0x32 represents the hexadecimal

number (32)16

In PCSpim (a MIPS simulator)

In Verilog, the following values are the same

CS2100

Prefix 0x for hexadecimal. Eg: 0x100 represents the hexadecimal

number (100)16
8b11110000: an 8-bit binary value 11110000
8hF0: an 8-bit binary value represented in hexadecimal F0
8d240: an 8-bit binary value represented in decimal 240
Number Systems and Codes

BASE-R TO DECIMAL
CONVERSION

Easy!

1101.1012 = 123 + 122 + 120 + 12-1 + 12-3

572.68 =

2A.816 =

341.245

CS2100

Number Systems and Codes

10

QUICK REVIEW
QUESTIONS (1)

DLD page 42
Questions 2-1 to 2-4.

CS2100

Number Systems and Codes

11

DECIMAL TO BINARY
CONVERSION

Method 1

Sum-of-Weights Method

Method 2

CS2100

Repeated Division-by-2 Method (for whole numbers)

Repeated Multiplication-by-2 Method (for fractions)

Number Systems and Codes

12

SUM-OF-WEIGHTS METHOD

Determine the set of binary weights whose sum

is equal to the decimal number

(9)10 = 8 + 1 = 23 + 20 = (1001)2

(18)10 = 16 + 2 = 24 + 21 = (10010)2

(58)10 = 32 + 16 + 8 + 2 = 25 + 24 + 23 + 21 =
(111010)2

(0.625)10 = 0.5 + 0.125 = 2-1 + 2-3 = (0.101)2

CS2100

Number Systems and Codes

13

REPEATED DIVISION-BY-2
To convert a whole number to binary, use successive
division by 2 until the quotient is 0. The remainders
form the answer, with the first remainder as the least
significant bit (LSB) and the last as the most significant
bit (MSB).
(43)10 = (101011)2

CS2100

2
2
2
2
2
2

Number Systems and Codes

43
21
10
5
2
1
0

rem 1 LSB
rem 1
rem 0
rem 1
rem 0
rem 1 MSB

14

REPEATED MULTIPLICATION-BY2
To convert decimal fractions to binary, repeated

multiplication by 2 is used, until the fractional product is

0 (or until the desired number of decimal places). The
carried digits, or carries, produce the answer, with the
first carry as the MSB, and the last as the LSB.
(0.3125)10 = (.0101)2
0.31252=0.625
0.6252=1.25
0.252=0.50
0.52=1.00

CS2100

Number Systems and Codes

Carry
0
1
0
1

15

MSB

LSB

CONVERSION BETWEEN
DECIMAL AND OTHER BASES
Base-R to decimal: multiply digits with their
corresponding weights.

Decimal to binary (base 2)

Whole numbers repeated division-by-2
Fractions: repeated multiplication-by-2

Decimal to base-R
Whole numbers: repeated division-by-R
Fractions: repeated multiplication-by-R

CS2100

Number Systems and Codes

16

QUICK REVIEW
QUESTIONS (2)

DLD page 42
Questions 2-5 to 2-8.

CS2100

Number Systems and Codes

17

CONVERSION BETWEEN
BASES

In general, conversion between bases can be

done via decimal:
Base-2
Base-3
Base-4

Base-R

Base-2
Base-3
Decimal
Base-4
.
Base-R

Shortcuts for conversion between bases 2, 4, 8,

16 (see next slide)

CS2100

Number Systems and Codes

18

BINARY TO OCTAL/HEXADECIMAL
CONVERSION

Binary Octal: partition in groups of 3

Octal Binary: reverse

(101 1101 1001 . 1011 1000)2 =

Hexadecimal Binary: reverse

(2731.56)8 =

Binary Hexadecimal: partition in groups of 4

(10 111 011 001 . 101 110)2 =

CS2100

(5D9.B8)16 =
Number Systems and Codes

19

QUICK REVIEW
QUESTIONS (3)

DLD page 42
Questions 2-9 to 2-10.

CS2100

Number Systems and Codes

20

READING ASSIGNMENT

Binary arithmetic operations

CS2100

Read up DLD section 2.6, pg 20 21.

Number Systems and Codes

21

PEEKING AHEAD (1/2)

Function simplification (eg: Quine-McCluskey)

In computer-speak, units are in powers of 2

Memory addressing (see next slide)

CS2100

Number Systems and Codes

22

PEEKING AHEAD (2/2)

Memory addressing

Assume 210 bytes in memory, and each word contains 4 bytes.

Addresses

Memory

binary
decimal
00101101
0000000000
0
01010101
0000000001
1
10111100
0000000010
2
01111001
0000000011
3
11001100
0000000100
4
10000101
0000000101
5
11010111
0000000110
6
00011000
0000000111
7
01101101
0000001000
8
10011011
0000001001
9
11010101
0000001010
10
01000001
0000001011
11
.
.

.
.

1111111111
1023

CS2100

Number Systems and Codes

23

NEGATIVE NUMBERS

Unsigned numbers: only non-negative values.

Signed numbers: include all values (positive and
negative)
There are 3 common representations for signed
binary numbers:

CS2100

Sign-and-Magnitude
1s Complement
2s Complement

Number Systems and Codes

24

SIGN-AND-MAGNITUDE
(1/3)
The sign is represented by a sign bit

0 for +
1 for -

Eg: a 1-bit sign and 7-bit magnitude format.

magnitude

sign

00110100 +1101002 = ?

10010011 -100112 = ?

CS2100

Number Systems and Codes

25

SIGN-AND-MAGNITUDE
(2/3)
Largest value:
01111111 = +127
10

Smallest value: 11111111 = -12710

Zeros:

Range: -12710 to +12710

Question:

CS2100

00000000 = +010
10000000 = -010

For an n-bit sign-and-magnitude representation, what

is the range of values that can be represented?

Number Systems and Codes

26

SIGN-AND-MAGNITUDE
(3/3)
To negate a number, just invert the sign bit.

Examples:

How to negate 00100001sm (decimal 33)?

Answer: 10100001sm (decimal -33)

How to negate 10000101sm (decimal -5)?

Answer: 00000101sm (decimal +5)

CS2100

Number Systems and Codes

27

1s COMPLEMENT (1/3)

Given a number x which can be expressed as an n-bit

binary number, its negated value can be obtained in 1scomplement representation using:
-x = 2n x 1
Example: With an 8-bit number 00001100 (or 1210), its
negated value expressed in 1s-complement is:
-000011002 = 28 12 1 (calculation in decimal)
= 243
= 111100111s
(This means that -1210 is written as 11110011 in 1scomplement representation.)

CS2100

Number Systems and Codes

28

1s COMPLEMENT (2/3)

Essential technique to negate a value: invert all

the bits.
Largest value:
01111111 = +12710

Smallest value: 10000000 = -12710

Zeros:

Range: -12710 to +12710

The most significant (left-most) bit still

represents the sign: 0 for positive; 1 for
negative.

CS2100

00000000 = +010
11111111 = -010

Number Systems and Codes

29

1s COMPLEMENT (3/3)

Examples (assuming 8-bit numbers):

(14)10 = (00001110)2 = (00001110)1s
-(14)10 = -(00001110)2 = (11110001)1s
-(80)10 = -( ? )2 = ( ? )1s

CS2100

Number Systems and Codes

30

2s COMPLEMENT (1/3)

Given a number x which can be expressed as an n-bit

binary number, its negated value can be obtained in 2scomplement representation using:
-x = 2n x
Example: With an 8-bit number 00001100 (or 1210), its
negated value expressed in 2s-complement is:
-000011002 = 28 12 (calculation in decimal)
= 244
= 111101002s
(This means that -1210 is written as 11110100 in 2scomplement representation.)

CS2100

Number Systems and Codes

31

2s COMPLEMENT (2/3)

Essential technique to negate a value: invert all

the bits, then add 1.
Largest value:
01111111 = +12710

Smallest value: 10000000 = -12810

Zero:

Range: -12810 to +12710

The most significant (left-most) bit still

represents the sign: 0 for positive; 1 for
negative.

CS2100

00000000 = +010

Number Systems and Codes

32

2s COMPLEMENT (3/3)

Examples (assuming 8-bit numbers):

(14)10 = (00001110)2 = (00001110)2s
-(14)10 = -(00001110)2 = (11110010)2s
-(80)10 = -( ? )2 = ( ? )2s

Compare with slide 30.

CS2100

1s complement:
(14)10 = (00001110)2 = (00001110)1s
-(14)10 = -(00001110)2 = (11110001)1s
Number Systems and Codes

33

READING ASSIGNMENT

Download from the course website and read the

Supplement Notes on Lecture 2: Number
Systems.
Work out the exercises in there and discuss
them in the IVLE forum if you have doubts.

CS2100

Number Systems and Codes

34

COMPARISONS

Important!

4-bit system
Positive values

Negative values

Value

Sign-andMagnitude

1s
Comp.

2s
Comp.

Value

Sign-andMagnitude

1s
Comp.

2s
Comp.

+7
+6
+5
+4
+3
+2
+1
+0

0111
0110
0101
0100
0011
0010
0001
0000

0111
0110
0101
0100
0011
0010
0001
0000

0111
0110
0101
0100
0011
0010
0001
0000

-0
-1
-2
-3
-4
-5
-6
-7
-8

1000
1001
1010
1011
1100
1101
1110
1111
-

1111
1110
1101
1100
1011
1010
1001
1000
-

1111
1110
1101
1100
1011
1010
1001
1000

CS2100

Number Systems and Codes

35

COMPLEMENT ON
FRACTIONS
We can extend the idea of complement on

fractions.
Examples:

Negate 0101.01 in 1s-complement

Answer: 1010.10

Negate 111000.101 in 1s-complement

Answer: 000111.010

Negate 0101.01 in 2s-complement

Answer: 1010.11

CS2100

Number Systems and Codes

36

2s COMPLEMENT
ADDITION/SUBTRACTION (1/3)
Algorithm for addition, A + B:

1.
2.
3.

Perform binary addition on the two numbers.

Ignore the carry out of the MSB.
Check for overflow. Overflow occurs if the carry in and carry
out of the MSB are different, or if result is opposite sign of A
and B.

Algorithm for subtraction, A B:

A B = A + (-B)

1.
2.

CS2100

Take 2s-complement of B.
Add the 2s-complement of B to A.

Number Systems and Codes

37

OVERFLOW

Signed numbers are of a fixed range.

If the result of addition/subtraction goes beyond this
range, an overflow occurs.
Overflow can be easily detected:

positive add positive negative

negative add negative positive

Example: 4-bit 2s-complement system

Range of value: -810 to 710

01012s + 01102s = 10112s

510 + 610 = -510 ?! (overflow!)

10012s + 11012s = 101102s (discard end-carry) = 01102s

-710 + -310 = 610 ?! (overflow!)

CS2100

Number Systems and Codes

38

2s COMPLEMENT
ADDITION/SUBTRACTION (2/3)

Examples: 4-bit system

+3
+ +4
---+7
----

0011
+ 0100
------0111
-------

-2
+ -6
----8
----

1110
+ 1010
------11000
-------

+6
+ -3
---+3
----

0110
+ 1101
------10011
-------

+4
+ -7
----3
----

0100
+ 1001
------1101
-------

Which of the above is/are overflow(s)?

CS2100

Number Systems and Codes

39

2s COMPLEMENT
ADDITION/SUBTRACTION (3/3)

Examples: 4-bit system

-3
+ -6
----9
----

1101
+ 1010
------10111
-------

+5
+ +6
---+11
----

0101
+ 0110
------1011
-------

Which of the above is/are overflow(s)?

CS2100

Number Systems and Codes

40

1s COMPLEMENT
ADDITION/SUBTRACTION (1/2)
Algorithm for addition, A + B:

1.
2.
3.

Perform binary addition on the two numbers.

If there is a carry out of the MSB, add 1 to the result.
Check for overflow. Overflow occurs if result is opposite sign of
A and B.

Algorithm for subtraction, A B:

A B = A + (-B)

1.
2.

CS2100

Take 1s-complement of B.
Add the 1s-complement of B to A.

Number Systems and Codes

41

1s COMPLEMENT
ADDITION/SUBTRACTION (2/2)

Any overflow?

Examples: 4-bit system

+3
+ +4
---+7
----2
+ -5
----7
----

CS2100

0011
+ 0100
------0111
------1101
+ 1010
-----10111
+
1
-----1000

+5
+ -5
----0
----

0101
+ 1010
------1111
-------

-3
+ -7
----10
----

1100
+ 1000
------10100
+
1
------0101

Number Systems and Codes

42

QUICK REVIEW
QUESTIONS (4)

DLD pages 42 - 43
Questions 2-13 to 2-18.

CS2100

Number Systems and Codes

43

EXCESS REPRESENTATION
(1/2)
Besides sign-and-magnitude and

complement schemes, the excess

representation is another scheme.
It allows the range of values to be
distributed evenly between the
positive and negative values, by a
simple translation
(addition/subtraction).
Example: Excess-4 representation
on 3-bit numbers. See table on the
right.

Questions: What if we use Excess-2

on 3-bit numbers? Excess-7?

CS2100

Number Systems and Codes

Excess-4
Representation

Value

000

-4

001

-3

010

-2

011

-1

100

101

110

111

44

EXCESS REPRESENTATION
(2/2)
Example: For 4-bit numbers, we may use excess-7 or

excess-8. Excess-8 is shown below. Fill in the values.

Excess-8
Representation

Value

Excess-8
Representation

0000

-8

1000

0001
0010
0011
0100
0101
0110
0111

CS2100

Value

1001
1010
1011
1100
1101
1110
1111
Number Systems and Codes

45

FIXED POINT NUMBERS

(1/2)
In fixed point representation, the binary point is
assumed to be at a fixed location.

For example, if the binary point is at the end of an 8-bit

representation as shown below, it can represent
integers from -128 to +127.

binary point

CS2100

Number Systems and Codes

46

FIXED POINT NUMBERS

(2/2)
In general, the binary point may be assumed to
be at any pre-fixed location.

Example: Two fractional bits are assumed as shown

below.
fraction part

integer part

assumed binary point

If 2s complement is used, we can represent values

like:
011010.112s = 26.7510
111110.112s = -000001.012 = -1.2510

CS2100

Number Systems and Codes

47

FLOATING POINT NUMBERS

(1/4)

Fixed point numbers have limited range.

Floating point numbers allow us to represent
very large or very small numbers.
Examples:
0.23 1023 (very large positive number)
0.5 10-37 (very small positive number)
-0.2397 10-18 (very small negative number)

CS2100

Number Systems and Codes

48

FLOATING POINT NUMBERS

(2/4)
3 parts: sign, mantissa and exponent

The base (radix) is assumed to be 2.

Sign bit: 0 for positive, 1 for negative.
sign

mantissa

exponent

Mantissa is usually in normalised form (the integer part

is zero and the fraction part must not begin with zero)
0.01101 24 normalised
101011.0110 2-4 normalised

Trade-off:

CS2100

More bits in mantissa better precision

More bits in exponent larger range of values
Number Systems and Codes

49

FLOATING POINT NUMBERS

(3/4)
Exponent is usually expressed in complement or excess
format.

Example: Express -6.510 in base-2 normalised form

-6.510 = -110.12 = -0.11012 23

Assuming that the floating-point representation contains

1-bit, 5-bit normalised mantissa, and 4-bit exponent. The
above example will be stored as if the exponent is in 1s
or 2s complement.
1

CS2100

11010

0011

Number Systems and Codes

50

FLOATING POINT NUMBERS

(4/4)
Example: Express 0.1875 in base-2 normalised form
10

0.187510 = 0.00112 = 0.11 2-2

Assume this floating-point representation:1-bit sign, 5-bit

normalised mantissa, and 4-bit exponent.

The above example will be represented as

CS2100

11000

1101

If exponent is in 1s complement.

11000

1110

If exponent is in 2s complement.

11000

0110

If exponent is in excess-8.

Number Systems and Codes

51

QUICK REVIEW
QUESTIONS (5)

DLD page 43
Questions 2-19 to 2-20.

CS2100

Number Systems and Codes

52

READING ASSIGNMENT

Arithmetic operations on floating point numbers

DLD page 31

IEEE floating point representation

CS2100

DLD pages 32 - 33
IEEE standard 754 floating point numbers:
http://steve.hollasch.net/cgindex/coding/ieeefloat.html

Number Systems and Codes

53

DECIMAL CODES

Decimal numbers are favoured by humans. Binary

numbers are natural to computers. Hence, conversion is
required.
If little calculation is required, we can use some coding
schemes to store decimal numbers, for data
transmission purposes.
Examples: BCD (or 8421), Excess-3, 84-2-1, 2421, etc.
Each decimal digit is represented as a 4-bit code.
The number of digits in a code is also called the length of
the code.

CS2100

Number Systems and Codes

54

BINARY CODE DECIMAL (BCD)

(1/2)
Decimal

Some codes are unused, like

1010BCD, 1011BCD, 1111BCD.
These codes are considered
as errors.

Easy to convert, but

arithmetic operations are
more complicated.
Suitable for interfaces such
as keypad inputs.

CS2100

Number Systems and Codes

BCD

digit
0

0000

0001

0010

0011

0100

0101

0110

0111

1000

1001

55

BINARY CODE DECIMAL (BCD)

(2/2)

Examples of conversion between BCD values

and decimal values:

(234)10 = (0010 0011 0100)BCD

(7093)10 = (0111 0000 1001 0011)BCD

(1000 0110)BCD = (86)10

(1001 0100 0111 0010)BCD = (9472)10

Note that BCD is not equivalent to binary

CS2100

Example: (234)10 = (11101010)2

Number Systems and Codes

56

OTHER DECIMAL CODES

Decimal Digit
0
1
2
3
4
5
6
7
8
9

BCD
8421
0000
0001
0010
0011
0100
0101
0110
0111
1000
1001

Excess-3

84-2-1

2*421

0011
0100
0101
0110
0111
1000
1001
1010
1011
1100

0000
0111
0110
0101
0100
1011
1010
1001
1000
1111

0000
0001
0010
0011
0100
1011
1100
1101
1110
1111

Biquinary
5043210
0100001
0100010
0100100
0101000
0110000
1000001
1000010
1000100
1001000
1010000

Self-complementing code: codes for complementary

digits are also complementary to each other.

Error-detecting code: biquinary code (bi=two,

quinary=five).

CS2100

Number Systems and Codes

57

SELF-COMPLEMENTING
The codes representing the pair of complementary digits
CODES

are also complementary to each other.

Example: Excess-3 code
0: 0011
1:
2:
3:
4:
5:
6:
7:
8:
9:

0100
0101
0110
0111
1000
1001
1010
1011
1100

Question: What are the other self-complementing

codes?

CS2100

Number Systems and Codes

58

GRAY CODE (1/3)

Unweighted (not an arithmetic code)

Only a single bit change from one code value to the next.
Not restricted to decimal digits: n bits 2n values.
Good for error detection.
Example: 4-bit standard Gray code
Decimal
0
1
2
3
4
5
6
7

CS2100

Binary
0000
0001
0010
0011
0100
0101
0110
0111

Gray Code
0000
0001
0011
0010
0110
0111
0101
0100

Decimal
8
9
10
11
12
13
14
15

Number Systems and Codes

Binary
1000
1001
1010
1011
1100
1101
1110
1111

Gray code
1100
1101
1111
1110
1010
1011
1001
1000

59

GRAY CODE (2/3)

Generating a 4-bit standard Gray code sequence.

0
0
0
0

0
0
0
0

0
0
1
0
1
0

0
1
1
0

0
0
0
0

0
1
0
1
0
1
0
1

1
1
0
0

0
1
1
0

1
0
0
1
0
1
0
1
0
1
0
1
0
1
0
1

1
1
1
1
0
0
0
0

0
0
1
1
1
1
0
0

0
1
1
0
0
1
1
0

Questions: How to generate 5-bit standard Gray code

sequence? 6-bit standard Gray code sequence?

CS2100

Number Systems and Codes

60

0
0

GRAY CODE (3/3)

0
11 01
000 0
000 00 1
0 1
0

mis-aligned
sensors

11
111

11
010 1
1
11 1
1
0

mis-aligned
sensors

0
11 00
110 1
000 1 0
11 0
1

0
00 01
11 1
0
000 11 11
1
0

0
11 01
111 0
111 0 1
00 1
0

Binary coded: 111 110 000

CS2100

0
00 00
001 0
111 00 0
0 0
1

1
1

010
1
11

01
010 0
1
11 1
1
0

1
11 11
00 0
0
000 00 11
0
0

0
1

sensors

Gray coded: 111 101

Number Systems and Codes

61

READING ASSIGNMENT

Conversion between standard Gray code and

binary

CS2100

DLD page 36.

Number Systems and Codes

62

QUICK REVIEW
QUESTIONS (6)

DLD pages 43 - 44
Questions 2-22 to 2-25.

CS2100

Number Systems and Codes

63

ALPHANUMERIC CODES
(1/3)
Computers also handle textual data.

Character set frequently used:

alphabets: A Z, a z
digits:
0 9
special symbols: $, ., @, *, etc.
non-printable: NULL, BELL, CR, etc.

Examples

CS2100

ASCII (8 bits), Unicode

Number Systems and Codes

64

ALPHANUMERIC CODES
(2/3)
ASCII

CS2100

American Standard
Code for Information
Interchange
7 bits, plus a parity bit
for error detection
Odd or even parity

Character
0
1
...
9
:
A
B
...
Z
[
\

Number Systems and Codes

ASCII Code
0110000
0110001
...
0111001
0111010
1000001
1000010
...
1011010
1011011
1011100

65

ALPHANUMERIC CODES
(3/3)
A: 1000001
ASCII table
LSBs
0000
0001
0010
0011
0100
0101
0110
0111
1000
1001
1010
1011
1100
1101
1110
1111

CS2100

000
NUL
SOH
STX
ETX
EOT
ENQ
ACK
BEL
BS
HT
LF
VT
FF
CR
O
SI

001
DLE
DC1
DC2
DC3
DC4
NAK
SYN
ETB
CAN
EM
SUB
ESC
FS
GS
RS
US

010
SP
!

#
$
%
&

(
)
*
+
,
.
/

MSBs
011 100
0
@
1
A
2
B
3
C
4
D
5
E
6
F
7
G
8
H
9
I
:
J
;
K
<
L
=
M
>
N
?
O

Number Systems and Codes

101
P
Q
R
S
T
U
V
W
X
Y
Z
[
\
]
^
_

110
`
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o

111
p
q
r
s
t
u
v
w
x
y
z
{
|
}
~
DEL

66

ERROR DETECTION (1/4)

Errors can occur during data transmission. They should

be detected, so that re-transmission can be requested.
With binary numbers, usually single-bit errors occur.

Example: 0010 erroneously transmitted as 0011 or 0000 or 0110

or 1010.

Biquinary code has

length 7; it uses 3
additional bits for
error-detection.

CS2100

Decimal
digit

Biquinary
5043210

0100001

0100010

0100100

0101000

0110000

1000001

1000010

1000100

1001000

1010000

Number Systems and Codes

67

ERROR DETECTION (2/4)

Parity bit

Even parity: additional bit

added to make total
number of 1s even.
Odd parity: additional bit
added to make total
number of 1s odd.

Example of odd parity

on ASCII values.

CS2100

Character
0
1
...
9
:
A
B
...
Z
[
\

Number Systems and Codes

ASCII Code
0110000 1
0110001 0
...
0111001 1
0111010 1
1000001 1
1000010 1
...
1011010 1
1011011 0
1011100 1

68

Parity bits

ERROR DETECTION (3/4)

Parity bit can detect odd

number of errors but not
even number of errors.

Example: Assume odd parity,

10011 10001 (detected)

10011 10101 (not detected)

0110 1
0001 0
1011 0
1111 1
1001 1
0101 0

Parity bits can also be

applied to a block of data.

Column-wise parity

Row-wise parity
CS2100

Number Systems and Codes

69

ERROR DETECTION (4/4)

Sometimes, it is not enough to do error

detection. We may want to correct the errors.
Error correction is expensive. In practice, we
may use only single-bit error correction.
Popular technique: Hamming code

CS2100

Number Systems and Codes

70

ERROR CORRECTION (1/7)

Given this 3-bit code C1

{ 000, 110, 011, 101 }

With 4 code words, we actually need only 2 bits.

To add error detection/correction ability, we use

more bits than necessary.

We call this k, the number of original message bits.

In this case, the length of each codeword is 3

We define code efficiency (or rate) by

k / length of codeword

Hence, efficiency of C1 is 2/3.

CS2100

Number Systems and Codes

71

ERROR CORRECTION (2/7)

Given this 3-bit code C1

{ 000, 110, 011, 101 }

Can C1 detect a single bit error?

Can C1 correct a single bit error?

Sometimes, we use 1 error for single bit error,

2 errors for 2 bits error, etc.

CS2100

Number Systems and Codes

72

ERROR CORRECTION (3/7)

The distance d between any two code words in a

code is the sum of the number of differences
between the codewords.

The Hamming distance of a code is the

minimum distance between any two code words
in the code.

Example: d(000, 110) = 2; d(0110,1011) = 3.

Example: The Hamming distance of C1 is 2.

A code with Hamming distance of 2 can detect 1

error.

CS2100

Number Systems and Codes

73

ERROR CORRECTION (4/7)

Given this 6-bit code C2

{ 000000, 111000, 001110, 110011 }

What is its efficiency?

What is its Hamming distance?
Can it correct 1 error?

Can it correct 2 errors?

CS2100

Number Systems and Codes

74

Self-study
ERROR CORRECTION (5/7)

Hamming code: a popular error-correction code

Procedure

Parity bits are at positions that are powers of 2 (i.e. 1,

2, 4, 8, 16, )
All other positions are data bits
Each parity bit checks some of the data bits

CS2100

Position 1: Check 1 bit, skip 1 bit (1, 3, 5, 7, 9, 11, )

Position 2: Check 2 bits, skip 2 bits (2, 3, 6, 7, 10, 11, )
Position 4: Check 4 bits, skip 4 bits (4-7, 12-15, 20-23, )
Position 8: Check 8 bits, skip 8 bits (8-15, 24-31, 40-47, )

Set the parity bit accordingly so that total number of 1s

in the positions it checks is even.
Number Systems and Codes

75

Self-study
ERROR CORRECTION (6/7)

Example: Data 10011010

Insert positions for parity bits:
__1_001_1010

Position 1: ? _ 1 _ 0 0 1 _ 1 0 1 0 so ? must be 0
Position 2: 0 ? 1 _ 0 0 1 _ 1 0 1 0 so ? must be 1
Position 4: 0 1 1 ? 0 0 1 _ 1 0 1 0 so ? must be 1
Position 8: 0 1 1 1 0 0 1 ? 1 0 1 0 so ? must be 0

Answer: 0 1 1 1 0 0 1 0 1 0 1 0

CS2100

Number Systems and Codes

76

Self-study
ERROR CORRECTION (7/7)

Suppose 1 error occurred and the received data

is:
011100101110
How to determine which bit is in error?
Check which parity bits are in error.

Answer: parity bits 2 and 8.

Add the positions of these erroneous parity bits

Answer: 2 + 8 = 10. Hence data bit 10 is in error.

Corrected data: 0 1 1 1 0 0 1 0 1 0 1 0
CS2100

Number Systems and Codes

77

END

CS2100

Number Systems and Codes

78