Harnessing Degrees of Freedom of Spatially Coupled Graph Codes for Agile Data Storage
Homa Esfahanizadeh, Lev Tauz, Lara Dolecek · IEEE BITS the Information Theory Magazine · 2023
With the massive growth rate in the volume of data and the heightened expectation for reliability and speed of modern storage devices, developing high-performance and low-cost error correction techniques that are customized to the emerging storage technologies has become critical. Spatially coupled low-density parity-check (SC-LDPC) codes are a viable candidate for data storage technologies as they offer strong error correction and low-latency decoding. This article presents an overview of recent advances in designing practical SC-LDPC codes that incur minimal overhead in the design process and in deployment, while achieving impressive levels of error correction. In particular, we present methods to reduce the cost of deign by harnessing the symmetries that the structure of SC-LDPC codes offers. We also discuss how the storage model side information and its working conditions can be used in the design to improve the error correction performance. Finally, we present an overview of recent techniques for enabling further flexibility in the design and decoding processes. The article concludes with open problems and suggestions for further investigation.