【作者】 李利军；

【导师】 范戈；

【作者基本信息】 上海交通大学 ， 通信与信息系统， 2008， 博士

【摘要】 近十年来,随着互联网的普及,各种网络业务的相继开展对现有网络的带宽容量提出了更高的要求,通信容量几乎成指数增长。以高速光时分复用(OTDM)技术和密集波分复用(DWDM)技术为核心的全光通信网络,已成为国际上通信领域研究的热点,特别是超高速OTDM技术是实现超大容量(Tbit/s)通信传输速率的首选方案。与DWDM技术相比,OTDM技术具有很多优势:首先,由于OTDM中采用单波长技术,避免了多个波长在光纤中传输时非线性效应引起的信道串扰,对光源波长的稳定性要求降低,波长的管理和控制相对简单,这就避免了DWDM技术的众多缺点;其次,OTDM技术能提供按需分配的带宽服务(BOD),实现不同粒度的灵活的带宽分配,而DWDM的带宽分配则是波长级别的;第三,采用OTDM技术易于实现分组交换和路由,实现真正意义上的统计复用,并实现灵活的组网;最后,因为OTDM中采用的是归零码(RZ码),TDM本身的数字特性非常适合于全光数字信号处理,实现数字再生。由于OTDM技术的以上这些特点,它在中小型城域网、局域网中具有广阔的应用前景;而DWDM技术,由于其超大容量及超长距离的传输能力使得其在大型城域网以及广域网中具有较大优势。所以,未来超大容量的光通信网络必定是二者的结合,优势互补。本文围绕OTDM传输系统的关键技术,对基于SOA干涉仪的全光开关解复用技术、基于SOA四波混频效应的全光解复用技术以及信号色散检测技术进行了理论和实验研究,研究内容主要包括以下四个方面:一、基于干涉仪结构的OTDM全光解复用子系统的核心器件是半导体光放大器(SOA),对SOA的研究首先需要选择合适的基于半经典理论的SOA数学模型。本文在研究了多种面向脉冲传输的SOA模型基础上,选定了一种面向皮秒级光脉冲码流的SOA系统仿真模型。提出了一组层级递进、逐步优化的仿真方案,该组仿真方案为研究基于SOA材料色散、增益色散和二、三阶非线性效应的光信号处理提供了理论基础,也为本文第三章和第四章的研究工作提供了仿真工具。二、合理地选择和设定全光开关的性能参量是评价全光开关在OTDM解复用子系统中解复用性能的首要任务,本文选择全光开关中SOA平均线宽增强因子和器件有效线宽增强因子分析了基于SOA相位调制效应的干涉仪全光开关的性能。平均线宽增强因子是SOA的本征物理属性,它不随输入信号的改变而改变,而器件有效线宽增强因子则是系统相关的,它决定了脉冲的相位变化和增益变化的关联度,是研究基于SOA相位调制效应的光信号处理问题的基础。本文进一步设定了增强型积分对比率(EICR)以便分析光开关的整体积分特性和局部起伏特性,还设定了关联积分对比率(CICR)以便分析光开关上下路信号的分离隔离度,在此基础上,本文进行了基于超高速非线性干涉仪(UNI)光开关的三种时钟/信号波长配置的仿真研究,它们是:近波长光开关、同频光开关和远波长光开关,其中着重给出了详细的基于双波耦合波模型的近波长增益透明光开关研究结果。这里使用的SOA数学模型是J.M.Tang和K.A.Shore构建的宽带大信号模型。在同频光开关和远波长光开关中使用的SOA数学模型是M.Y.Hong等人构建的SOA宽带大信号模型。利用在本文第一章提出的仿真方案,我们具体分析了时钟脉冲能量、时钟脉冲脉宽、偏置时间、SOA腔长对窗口包络的影响。研究结果表明,近波长光开关在增益透明配置条件下相位变化和增益起伏的关联度较低,其xICR指标最好。在对近波长光开关进行仿真研究时,本文还分析了大相移、长时间偏置条件下全光开关窗口双峰效应的成因,并揭示了由带间载流子密度起伏(CDP)导致的光开关模式效应。三、对于基于频移型光开关的OTDM解复用技术,本文借用A.Mecozzi和J.Mork构建的分离扰动模型进行了面向高速OTDM大带宽、高能量的FWM仿真研究。该模型明显的理论意义在于能够给出抽运光和共轭波的独立表达项,这就使准确地计算SOA波长转换效率成为可能,也为滤波器的带宽设计提供了重要的参考依据。研究表明,不论是对单脉冲还是脉冲码流,FWM转换效率对抽运光脉宽都是很不敏感的,但是在单脉冲条件下,开关比(On-Off ratio)对抽运光脉宽非常敏感,虽然在脉冲码流条件下,其敏感度有所降低,但同时其开关比也会有大幅下降。另外,本文修正了N.K.Das对于模式效应的定义,引入了输出信号码流的能量起伏以及共轭波码流的能量起伏两个参量。研究表明,当把全光开关应用于OTDM系统解复用系统时,设定一个适当的时间余量是必要的,这样就可以避免因开关窗口幅度起伏造成的解复用信号强度起伏。四、高速信号具有很小的色散窗口,面向传统的低速光纤链路的色散图谱配置已经不能满足高速信号的色散容限要求。本文从信号色散的时域特性和频域特性两个角度提出并研究了两种色散检测方案。研究表明,基于相位-波形转换的时域方案具有较高的检测范围和检测灵敏度,但是系统代价较大,需要引入复杂的时延干涉阵列;基于边带功率检测的频域方案系统结构简单,操作方便,但是对光电检测器件的检测灵敏度要求较高。OTDM技术的含义是很广泛的,本文仅仅对物理层最基本的两项实现技术进行了初步的研究。除了在物理层涉及到的超窄脉冲信号的产生、多路复用、色散补偿、全光3R再生等关键技术以外,其它的诸如时隙分配、全光地址识别、复用节点和交换节点的设计等问题也都属于OTDM技术的范畴,是我们在该领域从事进一步的研究工作时需要面对的问题。更多还原

【Abstract】 With the rapid advances of the commutative multi-media services, it becomes evident that the communication capability is being increased at exponential speed, so the higher demands are put forward for the transmission band-width nowaday. It has become a hotspot in the field of telecommunication research throughout the world, which is the all-optical network communication associated the ultra-high-speed optical-time-division-multiplexing (OTDM) technology with the dense-wave-division-multiplexing (DWDM) technology, in particular, the ultra-high-speed OTDM technology is the first choice for the ultra-high-speed and ultra-large-capability communications system (Tbit/s). Compared with DWDM, OTDM has at least four advantages. First of all, in OTDM the interference among different channels can be avoided due to the single wavelength. Consequently, the high level of the stabilization of the wavelength is not necessary. Secondly, DWDM could only provide bandwidth in wavelength level while OTDM could provide more flexible and finer bandwidth. Thirdly, OTDM make it easy to realize statistical multiplexing. Finnally, ODSP (Optical Digital Signal Processing) could be introduced in OTDM since the RZ code is used here.In this work, all optical demultiplexing based on SOA interferometer switches, wavelength conversion based all optical switches in OTDM communication systems as well as all chromatic dispersion monitoring are studied theoretically, which include:(1). The pico-second optical pulse sequence oriented SOA system simulative model is established. A set of simulative schemes, which is developed and optimized gradually, is put forward. The simulative scheme is the theoretical foundation for the OSP (Optical Signal Processing) of SOA accounted with the material dispersion, the gain dispersion and the second nonlinear effect as well as the third nonlinear effect.(2). The average linewidth enhancement factor and the component effective linewidth enhancement factor are introduced for the performance analysis of interferometer all optical switches based on cross phase modulation of SOAs. The average linewidth enhancement is the latent physical attribute of SOA that is the groundwork to investigate the OSP based on cross phase modulation of SOA. The enhancement integrated contrast ratio (EICR) and the conjugate integrated contrast ratio (CICR) are introduced to investigate the general integral attribute and the isolate degree between down and pass signals of all optical switches, respectively. The simulative studies on three pump/probe wavelength configurations of UNI are shown. The gain-transparent all optical switches are investigated in detail where the two-wave coupling model is used that is established by J.M.Tang and K.A.Shore. The broadband and strong signal model built by M.Y.Hong is used in the simulation of same frequency and far wavelength all optical switches. The effect done to the switch window by the clock optical pulse energy, the clock optical pulse width, the time delay and the length of SOA are investigated, respectively. The results show that in the condition of gain-transparent the relationship between the phase-shift and the pulsation of the gain of near wavelength all-optical switch is relatively weak and then the xICR is high. The cause of formation of switch window double-peak effect in the condition of large phase shift and long time delay is given and the all-optical switch data pattern effects due to CDP (the carrier density pulsation) is disclosed.(3). The high speed, strong pulse OTDM signal FWM effect oriented simulative study is performed where the separate disturbing model built by A.Mecozzi and J.Mork is used. Since the pump signal and conjugate signal can be expressed separately, it is realized to calculate the conversion efficiency accurately and to select the appropriate filter bandwidth. The results show that the FWM conversion efficiency is unsensitivity to the pump pulse width no matter in the condition of single pulse or of the pulse train. The on-off ratio is keen to the pump pulse width in the condition of single pulse. In the condition of pulse train the on-off ratio is becoming not as keen as that of single pulse and at the same time the on-off ratio itself is decreasing dramatically. We have modified the definition of data pattern effect of N.K.Das and introduced the energy pulsations of output signal train as well as the conjugate signal pulse. The results show that an appropriate temporal residual is necessary for the application of demultiplexing.(4). High-speed signal has limited chromatic-dispersion window. The traditional low speed dispersion map configuration cannot meet the requirment high-speed signal transmission. We provide and investigate two chromatic dispersion monitoring methods from the perspective of CD temporal attribute and CD spectral attribute. The results shown that the former has the advantages of high monitoring range and monitoring sensitivity at the cost of complicated diff-delay interferometer array. The latter is configuration brief and operation simple, but a high sensitivity of O/E detector is required.The maning of OTDM is comprehensive. Our present work only involves two key techniques and both of them are localized within physical layer. Except for the signal generation, multiplexing, dispersion compensation and all optical 3R regeneration in physical layer, there are many problems required to be overcomed, such as time-slot assignment, all optical address recoganization, the design of multiplexing node and switching node.更多还原