Title: Optimal precompensation for nonlinearities in longitudinal magnetic recording using dynamic programming

Authors: Fabian Lim, Aleksandar Kavcic

Addresses: Department of Electrical Engineering, University of Hawaii, Honolulu, HI 96822, USA. ' Department of Electrical Engineering, University of Hawaii, Honolulu, HI 96822, USA

Abstract: We propose a dynamic programming approach to compute optimal precompensation that minimises the Mean Squared Error (MSE) between a nonlinear read-back signal and a linear one. This is done by assuming that the nonlinear read-back signal can be represented by a known finite state machine model. We show how to use finite and infinite horizon dynamic programming techniques to minimise the MSE cost function. Suboptimal techniques are studied, and performance comparisons are performed using computer simulations.

Keywords: dynamic programming; NLTS; magnetic recording; partial erasure; precompensation; finite state machine models; simulation; nonlinearities.

DOI: 10.1504/IJPD.2008.017473

International Journal of Product Development, 2008 Vol.5 No.3/4, pp.410 - 429

Published online: 11 Mar 2008 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article