Multihead Multitrack Detection with Reduced-State Sequence Estimation

TL;DR

This paper introduces a reduced-state sequence estimation (RSSE) algorithm for multihead multitrack (MHMT) detection in magnetic storage, significantly lowering complexity while maintaining near-optimal performance, thus making high-capacity data storage more practical.

Contribution

The paper proposes an RSSE-based approach for MHMT detection, reducing complexity and enabling practical implementation with near-ML performance, including analysis for symmetric and asymmetric channels.

Findings

RSSE achieves near-ML performance with fewer trellis states.

Effective distance redefinition improves detection accuracy.

RSSE is applicable to both symmetric and asymmetric 2H2T channels.

Abstract

To achieve ultra-high storage capacity, the data tracks are squeezed more and more on the magnetic recording disks, causing severe intertrack interference (ITI). The multihead multitrack (MHMT) detector is proposed to better combat ITI. Such a detector, however, has prohibitive implementation complexity. In this paper we propose to use the reduced-state sequence estimation (RSSE) algorithm to significantly reduce the complexity, and render MHMT practical. We first consider a commonly used symmetric two-head two-track (2H2T) channel model. The effective distance between two input symbols is redefined. It provides a better distance measure and naturally leads to an unbalanced set partition tree. Different trellis configurations are obtained based on the desired performance/complexity tradeoff. Simulation results show that the reduced MHMT detector can achieve near maximum-likelihood (ML) performance with a small fraction of the original number of trellis states. Error event analysis is given to explain the behavior of RSSE algorithm on 2H2T channel. Search results of dominant RSSE error events for different channel targets are presented. We also study an asymmetric 2H2T system. The simulation results and error event analysis show that RSSE is applicable to the asymmetric channel.