Single-Head/Single-Track Detection in Interfering Tracks

Off-track interference (OTI) arises as a result of head misalignment and is likely to become the dominant source of errors in future magnetic storage systems with high track density. In this paper, using a two-track and single-head model for OTI, we compare two different equalization/detection techniques aiming at recovering data from only the main track. One is a conventional approach based on suppressing OTI by a linear equalizer and then performing partial response maximum-likelihood (PRML) detection of the main track data. The other views OTI as the response of the channel to an independent data sequence, simultaneously estimates the two data sequences by joint PRML detection, and then simply discards the adjacent track data at the end. We evaluate the detection performance in terms of effective detection signal-tonoise ratio, and provide bit-error-rate simulation results. We also discuss a technique to adaptively estimate the position of the head as well as compensate for the relative phase difference in the offtrack data.