This paper presents a new architecture for high-speed sustained data recording system, in which real-time and high throughput arc two key points. The FPGA implementation of the system controller is given as well. After the I/O bottleneck of conventional storage system structure is analyzed in this paper, a new recording and storage system framework is put forward by aiming to shorten the data transfer path and consequently to reduce the system resource utilization and data transfer latency, contrarily resulting in the improvement of the sustained data recording speed. To satisfy the real-time requirement, a local bus central logic controller module is designed to take the role of former host-side software and to reduce the utilization of host resource and latency caused by the intervention of the host. A speed-matching logic module between the disk in burst transfer mode and the sustained source acquisition data stream is also designed and implemented on a single FPGA chip. The proposed architecture and implementation approaches can be adapted in many applications where real-time and high I/O throughput are required.

The conventional data recorder is based on the tape for physics experiments. A modular disk recorder is developed that consists of data-splitting modules (DSM), data-recording modules (DRM) and commodity disk drives. This innovative disk recorder has a better performance and higher reliability than the tape recorder. It can be configured flexibly as different redundant array of inexpensive disks (RAID) structures to record acquired data according to different applications.

衍射计算是设计光盘光学系统的基础。由于光盘信息符（坑点）尺寸可以和波长相比较或更小，标量衍射理论的有效性必需进行检验。首先分析了标量衍射理论的能量计算误差。对于典型的坑点尺寸，标量理论的能量计算误差高达19%。然后用耦合波理论分析了入射波偏振态的影响。计算表明TM 波和TE 波的衍射特性与标量理论的结论相差较大，TM波和TE波的衍射行为也有差别。这种差别是由电磁场边界条件引起的，且会随着坑宽的缩小而增大。对于DVD，标量理论的误差已经很大，要得到准确的结果就必须采用矢量衍射理论。

随着光存储技术的发展，如何提高光盘存储密度备受关注。应用信道分析的方法，讨论了影响存储密度的因素和参量，提出光学系统分辨本领及光存储机理是其中最重要的因素。分别从缩短波长、增大数值孔径、超分辨以及缩小记录点、复用记录点及增加存储维度等角度总结了提高系统分辨能力、改变存储机理以提高光盘存储密度的途径。