Two dimensional signal processing for storage channels

Abstract

Over the past decade, storage channels have undergone a steady increase in capacity. With the prediction of achieving 10 Tb/in2 areal density for magnetic recording channels in sight, the industry is pushing towards di erent technologies for storage channels. Heat-assisted magnetic recording, bit-patterned media, and twodimensional magnetic recording (TDMR) are cited as viable alternative technologies to meet the increasing market demand. Among these technologies, the twodimensional magnetic recording channel has the advantage of using conventional medium while relying on improvement from signal processing. Capacity approaching codes and detection methods tailored to the magnetic recording channels are the main signal processing tools used in magnetic recording. The promise is that two-dimensional signal processing will play a role in bringing about the theoretical predictions. The main challenges in TDMR media are as follows: i) the small area allocated to each bit on the media, and the sophisticated read and write processes in shingled magnetic recording devices result in signi cant amount of noise, ii) the twodimensional inter-symbol interference is intrinsic to the nature of shingled magnetic recording. Thus, a feasible two-dimensional communication system is needed to combat the errors that arise from aggressive read and write processes. In this dissertation, we present some of the work done on signal processing aspect for storage channels. We discuss i) the nano-scale model of the storage channel, ii) noise characteristics and corresponding detection strategies, iii) two-dimensional signal processing targeted at shingled magnetic recording.