PAC Code Construction for Spin-Torque Transfer Magnetic Random Access Memory

Spin-torque transfer magnetic random access memory (STT-MRAM) is one potential candidate to replace the dynamic random access memory (DRAM) due to its superior features of nonvolatility, fast read/write speed, and high scalability. The error correction coding methods are applied to improve the reliability of STT-MRAM which is affected by the process variation and thermal fluctuation. In this paper, we investigate, for the first time, the design and optimization of the polarization-adjusted convolutional (PAC) code for the STT-MRAM channel. A crucial problem for the application of PAC codes to the STT-MRAM channel is the optimization of the index set of the non-frozen bits of the PAC codes. Hence, a rate-profile optimization method based on the genetic-algorithm-assisted bit-swapping is proposed. Simulation results show that the PAC code with the optimized rate-profile outperforms both the polar code and the PAC code with existing generic rate-profiles.