Optimum Soft-Output of Autoregressive Detector for Offtrack Interference in LDPC-Coded Perpendicular Magnetic Recording

In this paper we consider the behavior of an autoregressive (AR) detector for partial response (PR) signaling against offtrack interference (OTI) environment in perpendicular magnetic recording Based on the behavior we derive the optimum branch metric to construct the detector by the Viterbi algorithm We propose an optimum AR detector for OTI that considers an optimum branch metric calculation and an estimation of noise power due to OTI in order to calculate an accurate branch metric To evaluate the reliability of soft output likelihood values calculated by our proposed AR detector we demonstrate a bit error rate performance (BER) of low density parity check (LDPC) codes under OTI existing channel by computer simulation Our simulation results show the proposed AR detector can achieve a better LDPC coded BER performance than the conventional AR detector We also show the BER performance of our proposal can keep within 05 dB of the case that perfect channel state information regarding Oil is used in the detector In addition we show that the partial response maximum likelihood (PRML) detector is robust against OTI even if Oil is not handled by the detector