Geometric Constraints for Error Detection & Correction
Error-correcting codes add redundancy using geometric constraints. Valid code words occupy specific positions in a constrained subspace. When errors occur, the received word "violates" these constraints, allowing detection and correction.
The Hamming distance between two code words is the number of bit positions where they differ. A code can detect up to d-1 errors and correct up to ⌊(d-1)/2⌋ errors, where d is the minimum Hamming distance between any two valid code words.
Each parity bit enforces a constraint: the sum of bits in specific positions must be even (or odd). When constraints are violated, the pattern of violations (syndrome) identifies the error location.