Synchronization and Detection for Two-Dimensional Magnetic Recording

—— The subject of this project is two-dimensional signal processing —— the joint processing of multiple readback waveforms from neighboring tracks so as to recover the recorded user bits. Two-dimensional signal processing promises significant increases in areal density, in part because it facilitates the elimination of guard band inefficiencies, in part because it enables shingle writing strategies, and in part because it enables the use of larger and more sensitive read heads. The combination of shingle write recording and two-dimensional readback detection is known as two-dimensional magnetic recording (TDMR). Before the promises of TDMR can become a practical reality, however, further advances in signal processing algorithms are required. This research proposal describes a research project that explores two fundamental questions: •• How best to performing timing recovery (downtrack) and position recovery (crosstrack)? This is the 2-D extension of the timing recovery problem. The challenge is to achieve reliable performance with reasonable complexity despite the low SNR expected for a TDMR system. •• How best to implement a soft-output channel detector that efficiently accounts for both the intersymbol interference and the intertrack interference? The challenge is to find the right balance between performance and complexity, since the optimal 2-D soft-output detector is prohibitively complex. We choose to focus on these two questions for two reasons: •• Synchronization and detection algorithms are often the bottleneck to either the performance achieved by the read channel, or to its overall complexity. •• We are interested in the interaction between synchronization and detection, with the hope that a practical synchronization strategy will be able to exploit the error-control coding in an iterative setting to improve performance. Our aim is to develop synchronization and detection strategies that have reduced complexity and yet are capable of approaching the performance of optimal strategies (which are prohibitively complex).