CIF: Small: Two-Dimensional Channel Modeling, Detection and Coding for Shingled Magnetic Recording

点击翻译按钮获取中文摘要

Current magnetic recording technology is fast approaching its theoretical limit of 1 terabit per square inch (1 Tb/in^2), and unless new technology is developed that can deliver areal densities of up to 10Tb/in^2, the annual increases in hard-disk drive capacity will taper rapidly. Several candidate technologies are currently being developed that will extend the areal densities up to 4Tb/in^2, but only two that can possibly extend it all the way to 10Tb/in^2, namely energy assisted recording on bit-patterned media and two-dimensional magnetic recording. Bit-patterned media requires extreme nano-lithography and imprinting techniques, which are prohibitively expensive, and very tight specifications on the planarization of the disk surface. Two-dimensional magnetic recording, on the other hand, uses conventional heads and media, but requires the development of new two-dimensional signal processing and coding techniques suitable for the magnetic recording channel. This research program aims at providing an in-depth understanding of two-dimensional shingled magnetic recording through: 1) development of realistic channel models at the magnetic grain level; 2) development of techniques to analyze the statistical properties of such models; 3) development of two-dimensional intersymbol interference channel detectors with limited per-symbol complexity; 4) development of a novel and transformative approach to the design of capacity-approaching codes based on fundamental algebraic properties of cyclic codes; and 5) determination of the maximum areal recording density that can be achieved in practice. This research program will help answer the fundamental question of whether it is possible to approach the ultimate theoretical limit of one user bit per magnetic grain using shingled writing.The broader goal of this research program is to help determine the feasibility of hard-disk drives with areal densities of up to 10Tb/in^2 using shingled writing. Its successful completion would establish two-dimensional magnetic recording as the technology of choice for the next generation of magnetic storage devices and potentially shape future research in this field. The significance of achieving such a high recording density is that one could store up to ten times as much data on a drive of the same size as current ones, thus fostering a multi-billion dollar economic impact on the storage industry. In addition, the methods devised for detection and coding for channels with two-dimensional intersymbol interference have a wider application to new two-dimensional transmission and storage systems as they become available, for instance multi-aperture free-space optical communication systems and holographic data storage systems.