[9p-PB6-1~21] 3.2 情報フォトニクス・画像工学

英語発表

[9p-PB6-9] MLP-Based Batch Reading and Decoding for Holographic Memory

〇(M2)Fang Gao^1,2, Hironori Ito², Changqiu Yu¹, Satoshi Honma² (1.Hangzhou Dianzi Univ, 2.Univ of Yamanashi)

キーワード：

Holographic memory、Complex amplitude modulation、Deep-learning

Spatial Quadrature Amplitude Modulation (SQAM) is a promising approach for increasing the storage density of holographic memory . We previously proposed a batch reading and decoding method that decodes individual SQAM signals from interference intensity distributions generated by multiple simultaneously reconstructed signals. In this study, a Multi-Layer Perceptron (MLP)- based neural network was developed to directly estimate the amplitude ratios and phase differences among multiplexed SQAM signals from interference intensity distributions. The proposed method was validated through numerical simulation.