Skip to main content
eScholarship
Open Access Publications from the University of California

UCLA

UCLA Electronic Theses and Dissertations bannerUCLA

Absorbing Set Analysis of LDPC Codes and Read-Channel Quantization in Flash Memory

Abstract

High-capacity NAND flash memories achieve high-density by storing more than one bit per cell. Storage systems require extremely low block-error-rates, making powerful error-correcting codes with low-error floors necessary. Low-density parity-check (LDPC) codes are well known to approach the capacity of the additive white Gaussian noise (AWGN) channel, but they often suffer from error floors and require soft information to achieve better performance. This dissertation tackles these two problems.

The first part of this dissertation introduces the cycle consistency matrix (CCM) as a powerful analytical tool for characterizing and avoiding absorbing sets in separable circulant-based (SCB) LDPC codes. Each potential absorbing set in an SCB LDPC code has a CCM, and an absorbing set can be present in an SCB LDPC code only if the associated CCM is not full column-rank. Using this novel observation, a new code construction approach selects rows and columns from the SCB mother matrix to systematically and provably eliminate dominant absorbing sets by forcing the associated CCMs to be full column-rank. Simulation results both in software and in hardware demonstrate new codes that have steeper error-floor slopes and provide at least one order of magnitude of improvement in the low FER region.

This dissertation also shows how identifying absorbing-set-spectrum equivalence classes within the family of SCB codes with a specified circulant matrix significantly reduces the search space of code matrices with distinct absorbing set spectra. For a specified circulant matrix, SCB codes all share a common mother matrix and thereby retain standard properties of quasi-cyclic LDPC codes such as girth, code structure, and compatibility with existing high-throughput hardware implementations. SCB codes include a wide variety of LDPC codes such as array-based LDPC codes as well as many common quasi-cyclic codes. Hence the CCM approach should find wide application.

The second part of this dissertation focuses on coding for flash memory. Traditional flash memories employ simple algebraic codes, such as BCH codes, that can correct a fixed, specified number of errors. This dissertation investigates the application to flash memory of low-density parity-check (LDPC) codes which are well known for their ability to approach capacity in the AWGN channel. We obtain soft information for the LDPC decoder by performing multiple cell reads with distinct word-line voltages. The values of the word-line voltages (also called reference voltages) are optimized by maximizing the mutual information between the input and output of the multiple-read channel. Our results show that using this soft information in the LDPC decoder provides a significant benefit and enables the LDPC code to outperform a BCH code with comparable rate and block length over a range of block error rates. Using the maximum mutual-information (MMI) quantization in the LDPC decoder provides an effective and efficient estimate of the word-line voltages compared to other existing quantization techniques.

Main Content
For improved accessibility of PDF content, download the file to your device.
Current View