[Paper] Efficient Decoding Method for Holographic Data Storage Combining Convolutional Neural Network and Spatially Coupled Low-Density Parity-Check Code

In this study, we propose an effective data-decoding method for holographic data storage (HDS) by combining convolutional neural network (CNN) and spatially coupled low-density parity-check (SC-LDPC) code. The trained CNN provides output class probabilities accurately demodulates the reproduced from HDS. We focus on these probabilities, wherein only untrainable noise components such as white Gaussian remain. These are used calculating log likelihood ratio in sum-product decoding SC-LDPC demonstrate improvement of approximately 10 dB required signal-to-noise error-free numerical simulations.