-1

Consider the example of hamming ECC https://en.wikipedia.org/wiki/Hamming_code

Suppose after receiving you find that parity bits 16 and 8 are incorrect , which bit do you correct

MysticForce
  • 1,253
  • 1
  • 16
  • 27

1 Answers1

0

The question is a bit vague, but here are some possible answers:

1) 20-bit (20, 15) SEC code

If you have the code shown in the table without the "..." section filled in, then it depends on the implementation of the decoder but in theory it should be a detectable error. The decoder would probably raise the "detectable but uncorrectable error" (DUE) signal.

2) 31-bit (31, 26) SEC code

If you're talking about the code in the table with the "..." section filled in, this is a (31, 26) code. The error will mistakenly miscorrect the 8+16=24th bit, causing silent data corruption (SDC).

3) a 21-bit (21, 15) SEC-DED or 32-bit (32, 26) SEC-DED code

If you add an overall parity bit to the code (see section "Hamming codes with additional parity (SECDED)"), then the code can properly detect any two-bit errors. Therefore this error will be properly detected and the decoder will raise the DUE signal.

mbsullivan
  • 29
  • 4