【作者】 冯震；

【导师】 盛新志；

【作者基本信息】 北京交通大学 ， 光学， 2013， 博士

【摘要】 同步时分交换是大型交换机的主流交换技术,尽管目前基于分组的异步交换技术得到了较快发展,但是在大型的交换局中,同步数字程控交换机以其巨大的交换容量,仍然占有一定的规模。在开发高速全光网的时候,如何将已有的交换设备全光化,省去光电光转换,仍然是一件非常有意义的工作。全光缓存器可以在光域内实现光数据包的时分交换操作,避免了光-电-光的转换过程,有效克服了通信网络中的电子速率瓶颈,是未来全光时分交换设备的关键部件。以半导体光放大器的非线性偏振旋转效应为基础的全光缓存器是一种新型的全光缓存器,它具有结构紧凑、易于读写、可级联性好等优势。本论文在国家863项目“基于全光缓存器的弹性光分组交换环上”以及国家自然基金“并行数字式级联结构全光缓存器的研究”的支持下,针对半导体光放大器的非线性偏振旋转效应开展全光开关及全光缓存器的研究工作,并在此基础上,瞄准同步数字程控交换机全光化,开展了全光时分交换系统的探索。本论文完成的主要创新性研究工作如下：1.通过对SOA控制电流与NPR效应的研究,实现了一种具有高性能的基于SOA的偏振光开关,其分光比与消光比均可达到30dB,并提出了一种基于SOA的NPR效应的2×2空分光开关。2.首次提出使用保偏光纤作为全光缓存器的延迟器件,可使系统获得更好的偏振稳定性,分析了信号光缓存圈数与SOA控制电流的关系,首次实现了基于该缓存器单元的40圈缓存结果,延迟时间为16.4μs,缓存粒度为0.41μs。该缓存器单元只含有一个有源器件SOA,是一种结构紧凑易于集成的可循环式全光缓存器。首次提出使用压电陶瓷对缓存器单元光纤延迟线进行压力调制的方法,改变缓存过程中的传输损耗,得到了10Gbit/s信号光缓存10圈的输出功率均衡的结果。3.首次实现了基于两级串联结构缓存器系统的三数据包时隙压缩与时隙交换实验。结果表明,该系统可实现三数据包之间完备的时隙交换操作,可操作的最长时隙为124ns,时隙压缩可获得的最小安全时间间隔为13ns。并首次搭建了四级串联式的光缓存器结构,对单一数据包的分段缓存进行了实验研究,每个缓存单元的缓存深度达到90%。4.首次实现了一种基于SOA的NPR效应缓存器单元的空分-时分-空分(STS)结构的全光时分交换系统,并实现了数据包{A,B,C,D}到{D,C,B,A}的时隙交换操作。5.首次提出一种4×4结构的全光时分交换节点结构,采用并联反馈式缓存单元的组合,昂贵器件较少,且信号通过后完整性及一致性保持较好,由于SOA的增益作用,这种节点具备一定的自相似整形功能。更多还原

【Abstract】 Synchronous time division is the mainstream of switching technology in large switching machine. Despite the fast developing of the current packed-based asynchronous switching technology, the synchronous digital program-controlled switches still occupied a certain scale in large exchange offices, for their huge exchange capacity. In the development of high-speed all-optical network, how to change the existing equipment all-optical, eliminating the need for the photoelectrical conversion is a very rewarding job. The optical buffer can accomplish the time division exchange operation of optical data avoiding optical-electrical-optical conversion process, overcoming the bottleneck of electron rates in communication network and will be a key component of the all optical synchronous time division switches in future. The optical buffer based on nonlinear polarization rotation in a semiconductor optical amplifier is a novel optical buffer, which has the advantages of easy integration and can be cascaded.In this thesis, with the supporting of the state863Project "optical packet switching ring based on the elasticity of the optical buffer" and the National Nature Science Foundation of Parallel digital cascade structure study of the optical buffer, researches were deployed on all optical switch and buffer base on nonlinear polarization rotation in semiconductor optical amplifiers. And on this basis, targeting all optical synchronous digital program-controlled switches, a systematic exploration of all-optical synchronous time division switching system was carried out.The main contributions of this dissertation are listed as follows:1. By the study of the relationship of injected current in SOA and the NPR effects, we got a high-performance SOA-based polarization switch, and the splitting ratio and extinction ratio can reach30dB, and also we proposed a new2×2optical switch based on NPR in SO As.2. Firstly proposed a delay device composed of all optical buffer based on NPR effect in SOA. With only one SOA in the buffer units, a simple compact structure integrated in all-optical buffer. Experimentally realized a2.5G data optical packet of127bit buffered40laps. The relationship of cache laps with the output power, and gives the power equalization solutions and experimental results by the way of PZT. 3. Realized the time slot interchange of triple data packets in buffer units based on NPR of SOA. The results illuminates this system performs well on the TSI operation between triple data packets, and the longest time gap is124ns. The safe distance of time gap exchanging was also studied and the distance is13ns. Built a four cascade cache structure. Based on this system, segmented cache experimental study of a single data package was performed, each cache unit buffer depth reached90%.4. For the first time, a time slot interchange system with a STS structure of buffer units based on NPR effect of SOA was realized and an experiment four packets complete time slot interchange was completed.5. Proposed a optical time division switch structure of4×4all-optical switching matrix for the first time. The cache part was a combination of parallel feedback caching unit with less expensive devices. Solve the node data conflict node signal integrity and consistency remains good. Because of the gain role of the SOA, this node with a certain degree of self-similarity shaping functions.更多还原