Hamming codes
An example: Construct a code that corrects a single error
- We add m check bits to our message
- Can encode at most (2m – 1) error positions
- The unused (all zeros) check means no error
- Errors can occur in the message bits and/or in the check bits
- In n is the length of the original message then 2m – 1 >= (n + m)
- Examples:
- If n = 11, m = 4: 24 – 1= 15 >= (n + m) = 15
- If n = 1013, m = 10: 210 – 1= 1023 >= (n + m) = 1023