Bruce Wilson

Analysis of Shingle-Write Readback Using Magnetic-Force Microscopy

Shingle recording experiments were conducted on a precision spin-stand using conventional recording components. Although both head and medium correspond to a 250 Gb/in2 (Gigabits/sq. inch) product, the raw written densities were pushed beyond 1 Tb/in2 (Terabits/sq. inch). High-resolution Magnetic-force microscopy (MFM) was used to examine the resulting written patterns. The MFM images have an apparent resolution down to 16 nm. From these images we are able to characterize the signal and noise in two dimensions up to the highest densities. Two-dimensional signal-processing was applied to the waveforms recovered from these MFM images. At the highest written densities, very poor raw error-rates were encountered and customer data could only be recovered successfully by employing relatively low rate codes. Although the experiments pushed the raw recording density to 1039 Gb/in2, the highest customer information density at which data was successfully recovered was 623 Gb/in2.

Reverse Concatenation with Maximum Transition Run (MTR) Codes for High-Density Perpendicular Recording

We present a reverse concatenation (RC) architecture with maximum transition run (MTR) modulation codes and Reed-Solomon (RS) error-correction codes (ECCs). The scheme employs a high-rate primary (pre-RS) MTR code and a secondary (post-RS) MTR code, which controls error propagation. The two modulation codes are designed in such a way to maximize the overall code rate and maintain simple hardware implementation. Simulation results demonstrate superior performance compared to reverse concatenation with non-MTR codes, especially at a relatively high recording density.

Skew-Tolerant Gray Codes

We consider a particular family of Gray codes having the property that codewords ci and ci+2 differ in a burst of length exactly two. These codes can be applied in the servo track-identification field to identify the correct track during seeks. We call these codes skew-tolerant Gray codes (STGC) and we present explicit constructions

Architecture and Implementation of a First-Generation Iterative Detection Read Channel

This paper explores the architecture of a first-generation iterative detection read channel solution that delivers up to an 8% increase in drive capacity compared to previous generation electronics. Using low density parity check (LDPC) coding, detected data is continually improved by a detection structure that implements multiple iterations through a set of soft-input soft-output (SISO) and message-passing (MP) blocks. A specially constructed LDPC code is used to optimize performance in the presence of inter-symbol interference (ISI). Also, Reed-Solomon error correction coding (ECC) is retained for optimal data integrity. A 65 nm technology is used in the implementation of this design.

Magnetic Recording Near the Grain Size Limit

The writing quality of magnetic recording improves with increasing sharpness of the written transitions. Ultimately, the transition sharpness is limited by the grains themselves, and no longer depends on the head field gradient. In this study, we varied the head field gradient by changing the head-media spacing and characterized the write quality by transition jitter and equalized signal-to-noise (eSNR) measurements. Consistent with a recording in the grain size limit, we discovered that the write quality did not improve with smaller head to medium spacings. We also found that the measured values for the jitter and the eSNR are close to what is estimated for grain size limited recording.

On multiple burst-correcting shortened cyclic codes

In this paper, the multiple burst-correcting shortened cyclic codes are proposed which is optimal in terms of redundancy. We consider linear binary codes and Reed-Solomon codes for correcting shortened cyclic codes.

Pattern-Dependent Noise Predictive Soft Detection in the Post-Processor With Error Filters

This paper investigates a pattern-dependent noise predictive soft detection method for channel architectures that are based on a long target response and post-processing rather than a short target response and base-line wander compensation. We utilize properties of the autoregressive pattern-dependent noise model to compare tentative Viterbi sequence with alternative sequences in the post-processor and efficiently compute soft information. Even though the post-processor cannot consider all possible sequences like trellis-based detectors can, we demonstrate for a short target response that our post-processing solution does not experience any loss in performance compared to the optimal maximum a posteriori trellis-based soft detector. The complexity of the computations in the post-processor grows only linearly with the target length, as opposed to the exponential growth in complexity in trellis-based detectors. In this way we can efficiently perform nearly optimal pattern-dependent soft detection in the post-processor for a very long target response without base-line wander compensation.

Optimal interleaving for burst errors in Generalized Concatenated codes

Generalized concatenation (GC) of Reed-Solomon (RS) codes is a powerful technique to enhance the error-correcting capability of RS codes without resorting to large finite fields. However, interleaving a GC scheme in the standard column-wise way significantly reduces the burst-correcting capability of such scheme. In this paper, we present techniques that optimize the burst-correcting capability of some GC schemes.

GEN03-2: Construction of Distance-Separated Gray Codes

We propose new systematic constructions of Gray codes that improve reliability of the track identification field in the servo portion of a magnetic recording disk. In this application, it is beneficial to obtain good distance properties within a certain range of neighboring Gray codewords, because the servo positioning system typically gives a good estimate of the range of interest. With proposed Gray codes, we provide a high separation range between non-adjacent Gray codewords at Hamming distance one, which guarantees error detection within this range. In addition, we provide a certain level of separation between Gray codewords at Hamming distance two, which can be utilized for error correction.

Thermal decay in perpendicular media

Thermal decay is one of the main limitations in longitudinal recording. In depth understanding the decay processes in perpendicular recording allows understanding the effect of thermal decay on the error rate and improving the design of head, media and channel. In the present paper we investigate thermal decay processes in perpendicular recording media by means of experiments and modeling.