Research On Coding And Decoding Techniques For Multi-dimensional Modulation In Holographic Data Storage

With the rapid development of information technology,mobile communication,and the Internet technique,the global information is exploding exponentially.Data has become the most important resource in the digital world.Information discovery,storage,and protection are becoming increasingly important.However,the rapidly growing data volume has brought unexpected great pressure for the storage technique.Traditional storage techniques have limitations for long-term storage of big data,due to their high energy consumption,short lifetimes,and limited storage capacity.Therefore,how to store the massive amounts of data for a long time at low cost is an urgent requirement in the era of big data.The development of storage technology with larger capacity,faster transmitting rate,longer life time and lower energy consumption is imperative.Holographic data storage(HDS)has the characteristics of two-dimensional encoding and three-dimensional storage.It is expected to be the next generation of optical storage with high capacity and high transmitting rate because of the multi-dimensional modulation of HDS.Due to the pure amplitude modulation of traditional HDS,the capacity is still limited.Recently,studies have been conducted to introduce phase information to improve the coding efficiency of single data page in HDS.However,since the phase cannot be detected directly,complex retrieval algorithms are often required to demodulate the phase information.Indirect phase decoding makes data reading rate lower and the whole experimental system more complicated.The phase data reading is easily affected by environmental noise,which makes it difficult to ensure the transmitting rate and accuracy.How to realize multi-dimensional modulation while ensuring fast decoding of information so as to achieve large-capacity storage and stable data transmission at high rates is the key issue need to be resolved for multi-dimensionally modulated HDS.With this goal in mind,this study conducts research on the coding and decoding technology of multidimensional modulated HDS.First,to improve and optimize the traditional iterative fast Fourier transform phase retrieval algorithm.Designed a phase modulated reference based on the characteristic of the collinear HDS to improve the coding efficiency of the single data page while ensuring the phase decoding rate.A de-noise method based on deep learning is introduced to further improve the decoding accuracy of phase demodulation.Second,a non-interferometric lensless phase retrieval method is proposed.By detecting the intensity features using deep learning and representing the inverse process from intensity image to phase information through convolutional neural network(CNN),the direct phase retrieval from an intensity image is achieved.Third,a non-interferometric complex amplitude retrieval system and method are proposed.Based on the analyzing of the intensity characteristics near field diffraction images,the model descriptions of diffraction intensity to amplitude and phase are established respectively.The direct demodulation of complex amplitude information from a single diffraction intensity image is realized.Further,the complex amplitude modulated HDS is optimized by using a single phase spatial light modulator(SLM)to modulate the complex amplitude information.Finally,the simplification and miniaturization of the HDS system is finally achieved based on the direct decoding of multidimensional coding and decoding.In summary,this study optimizes the traditional iterative phase retrieval method for HDS,proposes a single-shot phase retrieval method based on deep learning and further proposes a complex amplitude modulated HDS and optimizes the system with a double phase hologram.We realized the multi-dimensional modulation and fast demodulation of HDS,which makes the development of high-density recording and fast transmission of holographic data storage possible and helps to realize high-capacity holographic storage system.