【作者】 杨大鹏；

【导师】 贾克斌；

【作者基本信息】 北京工业大学 ， 电路与系统， 2004， 硕士

【摘要】 摘 要 在过去的十几年中，人们对存储技术的大容量和高速传输速率的追求不断升级，特别是随着网络和多媒体技术的广泛应用，这种需求更是急剧膨胀。在这种需求下，体全息存储以其大容量和高数据读取速率的优点进入人们的视野，并被认为是最有希望成为下一代海量存储的技术。但由于缺乏合适的存储材料及实用部件，再加上系统的复杂性，特别是体全息存储受众多噪声的干扰，使得其实用化进程较为迟缓。如何有效地抑制存储系统中的噪声干扰是体全息存储技术的一个重要研究课题。本论文的研究内容就是以此为目标，应用数字信号处理技术，展开对体全息存储系统中噪声抑制方法的研究。 论文首先根据现实体全息存储图像中存在的噪声特点，并针对由于背景光强不均匀及部分固定的失真造成的存储图像噪声进行了研究，为减小它们的影响，在实际存储中采用参考图像的存储方式，在参考图像中提取这些噪声的特征，然后在数据处理中将该部分噪声去除；其次，针对研究室原有存储方案中纠错码纠错能力较弱的问题，引入了纠错能力较强的纠错码，并通过仿真实验证明了，在原有存储实验系统和方案的情况下，经过增强纠错能力后误码率可以达到 10-6 数量级；通过仿真实验还验证了二维移位循环交织的有效性；同时对原有存储方案中调制—阵列码解码速度过慢进行了相应的研究；本论文还从能量的角度分析了在存储中不同 SLM 和 CCD 参数下单个点的能量扩散情况及SLM 和 CCD 参数对像元匹配实验的影响。更多还原

【Abstract】 Abstract The requirements for storage capacity and fast transfer rate have increaseddramatically in the past decades, and are exploding especially with the popularity ofthe Internet and multimedia. For this reason, volume holographic storage, with itshigh storage capacity and data transfer rate, promises to become a next-generationstorage technology. But the absence of storage medium and practical componentsbesides of the complexity of the system result in lots of noise. The noise previouslyprevents the practical condition of the technology. The high fidelity of stored data asthe requirement for practicability makes the efficient noise control in storage systeman important research direction. This is the target of the research that bases on digitalsignal processing to suppressing the noise in holographic storage system. Firstly, according to the noise characteristic of realistic volume holographicimage, the paper studies the noise caused by uniformity of light intensity andpartially fixed distortion. To reduce the influence, reference images are stored inreality. The features of these noises can be drawn from reference image. They help toexclude the partial noises in data processing. Secondly, because of the weakness ofError-Correction Code (ECC), the paper introduces Error-Correction Code withstrong capability into the storage scheme. It has beenproved in simulation that UserBit Error Rate can be 10-6 quantity level after Error-Correction Code decoded. Theefficiency of 2-D cyclic shifting interleaving also can be shown in simulation.Research has been focused on slowness of modulation-code decoding in this paper.Eventually, the paper analyses the point spread function in energy with differentSLM and CCD parameter, and those parameters influencing on pixel-matchedstorage.更多还原