Application of pattern-output viterbi algorithm to algebraic soft-decision decoding over partial response channels

The algebraic soft-decision decoding algorithm (ASD) requires a reliability matrix as its input. In this paper, a new method to construct the reliability matrix over partial response (PR) channels of interest in magnetic recording is proposed by using recently introduced pattern-output Viterbi algorithm (POVA). A modified bit-level generalized minimum distance (BGMD) algorithm is also proposed with the POVA to achieve performance gains over PR channels that are as large as gains as over AWGN channels.