Explain about hamming error correcting code, Computer Engineering

Q. Explain about Hamming error correcting code?

Richard Hamming at Bell Laboratories worked out this code. We will only introduce this code with help of an illustration for 4 bit data. Let's presume a 4 bit number b4, b3, b2, b1.  In order to create a simple error detection code which detects error in one bit only we can just add an odd parity bit.  Though if we want to find which bit is in error then we may have to employ parity bits for different combinations of these 4 bits in such a way that a bit error can be recognized uniquely.  For illustration we can create four parity sets as below

Source Parity                            Destination Parity

b1, b2, b3                                                        P1                                            D1

b2, b3, b4                                                        P2                                            D2

b3, b4, b1                                                        P3                                            D3

b1, b2, b3, b4                                                  P4                                            D4

Now very interesting phenomenon can be noticed in above displayed parity pairs.  Assume data bit b1 is in error on transmission then it will cause alteration in destination parity D1, D3, D4.

ERROR IN                                                     Cause change in Destination Parity 

(One bit only)

b1                                                                                             D1, D3, D4

b2                                                                                             D1, D2, D4

b3                                                                                             D1, D2, D3, D4

b4                                                                                             D2, D3, D4

Figure: The error detection parity code mismatch

So by simply comparing parity bits of source and destination we can recognize that which of four bits is in error. This bit then can be complemented to eliminate error. Please note that even source parity bit can be in error on transmission but under assumption that just one bit (irrespective of data or parity) is in error it would be detected as just one destination parity would be different. What should be length of error detection code that detects error in one bit? Before responding this question we have to look in comparison logic of error detection. Error detection is done by comparing two ‘i’ bit error detection and correction codes fed to comparison logic bit by bit (see figure below). Let’s have comparison logic that produces a zero if compared bits are same or else it generates a one. Consequently if similar Position bits are similar then we obtained zero at that bit Position however if they are dissimilar that is this bit position may point to an error then this Particular bit position would be marked as one. This way a matching word is built. This matching word is ‘i’ bit long so can signify 2i values or combinations.

Posted Date: 7/17/2013 6:58:56 AM | Location : United States







Related Discussions:- Explain about hamming error correcting code, Assignment Help, Ask Question on Explain about hamming error correcting code, Get Answer, Expert's Help, Explain about hamming error correcting code Discussions

Write discussion on Explain about hamming error correcting code
Your posts are moderated
Related Questions
In a single stage network: (A)  There is no redundancy (B)  There is redundancy (C)  Alternative cross points are available (D)  Alt

An 8086 interrupt can take placedue to the following reasons: 1.  Hardware interrupts caused by some external hardware device. 2.  Software interrupts that can be invoked wit

(i)  Input a single ASCII character into BL register without echo on screen  CODE SEGMENT  MOV AH, 08H;         Function 08H  INT 21H          ;         the character inpu

Automated Reasoning - first-order logic: The topic known as "Automated Reasoning" in "AI"concentrates mostly on deductive reasoning, here new facts are logically deduced from

The 2's complement of the number 1101110 is ? Ans. 1's complement of 1101110 is = 0010001 ans hence 2's complement of 1101110 is = 0010001 + 1 = 0010010.

draw input and output charectoristics of BJT and justify CE configuration provides large current amplification

a. Define parsing? Give difference among top down parsing and bottom up parsing. b. Determine the self-relocating programs? Why self-relocating programs are less efficient then

Q. Determine the performance of a parallel algorithm? One more method of determining the performance of a parallel algorithm can be performed after calculating a parameter know

Perl is an interpreted language (not compiled, like Java) which is ideally suited for CGI programming. It has its roots in UNIX system administration and offers various features li

Shape : A shape is any area that can be perceived to have height and width. Shapes within an image can be created by the use of line, or by the edges created by variations in c