Abstract

Volume optical memory represents a unique type of storage channel. It can be characterized as a 2D parallel, space-variant system with potentially time-varying characteristics. Like any other storage channel however, imperfections and noise degrade the SNR of retrieved data and this degradation may result in data errors. As with traditional memory and communications systems, detection and error handling techniques can be used to recover lost SNR and correct data errors.^[1,2] The success of detection theoretic techniques is based on the use of prior knowledge concerning channel characteristics and noise correlations, while error correction techniques are based on the addition of structured redundancy to the stored data. The 2D nature of the volume optical storage channel requires that detection and error correction techniques utilize prior knowledge of the 2D noise and error correlations to produce efficient solutions, thus 2D techniques for detection and error correction are required. In addition, such 2D parallel algorithms will admit 2D parallel implementations which serve to avoid critical data bottlenecks that can arise upon serialization. In this paper we will describe the use of a parallel 2D detection algorithm and a 2D interleaving technique that can improve data fidelity and increase memory capacity in the case of page access volume memory.

© 1997 Optical Society of America