【作者】 唐毅；

【导师】 潘龙法；

【作者基本信息】 清华大学 ， 仪器科学与技术， 2009， 博士

【摘要】 高密度、大容量是光存储发展的趋势。多阶存储是提高存储容量的技术路线之一。本文研究了一种新提出的多阶存储方法——波形调制多阶。它基于普通DVD盘片生产设备和读出光学头,具有独特的物理格式。本文对该多阶方法的一些关键技术进行了研究。本文研究了波形调制多阶的存储原理和具体实现方案。波形调制多阶是在原有二值存储的原始坑和岸中加入子岸和子坑,以改变读出信号的波形,并把波形作为阶次识别的标志。相对于传统的利用信号幅值来识别阶次的多阶方法,它能克服其在存储容量和伺服性能上的不足。本文给出了它在DVD只读光盘平台上的具体实现方案。由于波形调制多阶光盘在写入过程中存在子坑/子岸的加入,它的写策略优化跟其他存储方法有很大的不同。本文提出了一套基于仿真计算的写策略优化方法,建立了从母盘刻录到信号读出的完整计算模型。把阶次易于分辨和游程不被改变作为优化的目标,得到了各个游程阶次的写策略参数。在商用DVD生产线和读出系统对写策略优化结果进行了验证,得到的信号波形能跟仿真结果吻合,实际信号的抖晃值在8%以下,游程错误率能保持在2×10^-4以下。波形调制多阶利用读出信号的波形来识别阶次,它的阶次识别方法也跟其它多阶方法不同。由于读出信号跟符号序列没有对应关系,部分相应最大似然法不能直接用于本多阶存储方法。本文提出了先游程检测再阶次检测的信号检测方案,设计了一套基于最小欧氏距离的自适应阶次识别方法。该阶次检测方法不仅能克服读取过程中的信号波动,还能克服不同批次盘片工艺偏移带来的信号差异。在硬件平台上对该方法用FPGA进行了实现,得到了低于2×10^-4的阶次错误率,实现了高清视频的连续稳定播放。本文还分析了母盘刻录工艺和伺服误差对读出信号的影响,定义了波形偏差参数来定量评估对阶次识别的影响。通过仿真计算,得到了母盘刻录工艺改变和存在聚焦/循迹误差时的读出信号,以及不同改变量下的波形偏差。分析结果对控制母盘刻录工艺和伺服控制的设计有重要的参考价值。更多还原

【Abstract】 High density and large capacity are the trends of optical storage. Multi-level is a candidate technology to increase storage capacity. A new proposed multi-level method, signal waveform modulation （SWM） multi-level is researched. It is based on conventional DVD production equipment and readout optical pick-up, and it has exclusive physical format. Research on some key technologies on this multi-level method is presented.Storage principle and specific implementation are presented. For SWM multi-level, a sub-pit/sub-land is inserted into the original land/pit in order to change the waveform of readout signal. The waveform is used for level differentiation. Compared to conventional multi-level, which uses signal amplitude for level differentiation, the SWM multi-level has advantages on storage capacity and servo performance. The specific implementation on read-only DVD platform is presented.Due to the insertion of sub-pit/sub-land, the write strategy optimization is different from that of other storage method. A write strategy optimization method based on simulation is proposed. A complete calculation model from mastering to readout signal is established. The optimization target is that the levels are easy to differentiate and the run-length is varied by the sub-pit/sub-land. The write strategy vector for every run-length and every level is achieved by simulation calculatiion. The result is validated on the commercial DVD production line and readout system. The readout waveform agrees with the simulation result. The jitter is below 8% and the error radio of run-length can be kept less than 2×10^-4.The SWM employs waveform for the level differentiation, and thus its level detection is difference from other multi-level method. Because the readout signal does not respond to a symbol sequence, the partial response maximum like-hood （PRML） detection cannot be used to the SWM multi-level. A signal detection solution, where level detection is after run-length detection, is proposed. An adaptive level detection method based on minimum Euclid distance is presented. This level detection can overcome the readout signal fluctuation as well as the processing variation of disk production. This detection method is implemented on the hardware platform with FPGA. A level error ratio of less than 2×10^-4 is achieved. Constant and stable playback of high definition video is realized.The influence of mastering processing and servo error on the readout signal is analyzed. A parameter, waveform deviation, is defined to evaluation the influence on level detection. By simulation, the readout signals with mastering processing fluctuation and focusing/tracking error are figured out. As well, responding waveform deviations are achieved. The results provide important reference to the control of mastering processing and servo system design.更多还原