Modeling of the nonlinear perpendicular recording readback signal

Abstract

Perpendicular recording has been proposed as a better recording technology than the traditional longitudinal recording for high-density recording systems. The change in the recording format from longitudinal to perpendicular recording brings about new challenges in head, media, and channel designs. One of these challenges is the increasing amount of read and write nonlinearities. In addition, the nature of these nonlinearities may not be the same as in longitudinal recording. We developed a model for characterizing and modeling nonlinearities, regardless of their nature or severity, based on finite memory Volterra series. It was determined that a Volterra Model does not adequately model the perpendicular recording signal due to the fact that perpendicular recording suffers from very long inter-symbol interference (ISI). The long ISI is due to the presence of droop effect and high pass filter effects of the preamplifier and this translates into a requirement of large memory length L, in order to avoid truncation effects that lead to lack of fit of the model. The new model introduced here is called the composite Volterra model and it requires a moderate memory length and deals with the extended ISI using a linear model. The composite Volterra model is shown to provide a good fit to measured signals with various amounts of nonlinearities and using the composite Volterra model, one is able to synthesize a readback signal with appropriate nonlinearities for any given input data pattern.