Abstract
The areal density on a magnetic recording medium can be increased by increasing the linear bit density, or by increasing the track density, or, more commonly, by a combination of both. Equalizing the signal to a higher order partial response polynomial (i.e. EPR4), and employing a trellis code in conjunction with a PR4 channel (i.e. TCPR4) are among the techniques to achieve this goal by providing performance gains over a PR4 system. In this study, the potential increases in linear and track densities that could be afforded by EPR4 and TCPR4 are investigated by simulating a read channel model which incorporates the effects of non-ideal timing recovery, A/D conversion, and finite register-lengths, using spinstand data. In order to quantify the effects of the increased track density, intertrack interference (ITI) is taken into account in evaluating the performance, by defining the interference as a function of the track misregistration. The trade-offs concerning the equalizer loss due to a higher linear bit density and the SNR loss due to a narrower track, as a function of PW50/T, are addressed. The measure of performance used captures the coding gain, the rate loss due to coding, the loss of the equalizer, and the degradation due to ITI. Assuming similar hardware complexity for the detectors, the two alternatives are compared in terms of the areal density increase they provide over a range of user bit densities that are of current practical interest.
