Design of LUT-Based LDPC Decoders for Spin-Transfer Torque Magnetic Random Access Memory

Abstract

Recently, low-density parity-check (LDPC) codes have been proposed to improve the reliability of spin-transfer torque magnetic random access memory (STT-MRAM). While several research works focus on the design of LDPC codes to combat the process variation and thermal fluctuation of the STT-MRAM, we aim to tackle the high decoding latency issues by the novel design of a lookup table (LUT)-based LDPC decoder for the forward and backward (FB) processes. In this work, the channel quantizer and LUT operations of the LDPC decoder are jointly designed by minimizing the error probability estimated from the density evolution (DE) algorithm. The simulation results show that the proposed design method for a code rate of 0.9 outperforms the previous methods in terms of bit error rate (BER) performance. Furthermore, the proposed design can support the channel model of the STT-MRAM without an independent and identically distributed (i.i.d.) channel adapter.