Abstract
We investigate an early stopped version of the Euclidean algorithm in the decoding of BCH codes. We show that this early stopped version is equivalent to the early stopped version of the Berlekamp-Massey algorithm and can be implemented by an algorithm which requires only (t+e) syndromes, instead of 2t syndromes in the conventional Euclidean algorithm or the conventional Berlekamp-Massey algorithm, and has multiplicative complexity te+e2-1, where t is the designed error-correcting capability of the BCH code and e is the number of errors actually occurring.