【作者】 刘刚；

【导师】 杨向东；

【作者基本信息】 四川大学 ， 原子与分子物理， 2004， 博士

【摘要】 光与物质的相互作用的研究不仅有重要的理论意义,而且有广泛的应用价值,在通信和信息处理领域也不例外。本文较详细而全面的研究了不同实现方法下全光波长变换。首先研究了光在半导体光放大器中的交叉增益波长变换过程。利用非线性薛定谔方程建立半导体光放大器的啁啾模型,并且通过分析该模型研究了其变换过程中的啁啾特性,以及利用载流子速率方程研究了消光比、信噪比与误码率间的关系。最后利用计算机模拟了入射功率、波长和偏置电流的变化对变换信号特性的影响。研究结果表明: 饱和增益越大SOA啁啾越小,入射光子能量越高啁啾也越低,减小入射功率也可以减小SPM带来的啁啾。增加有缘层长度与有效截面面积,也有利于减小SOA的啁啾。自发辐射对信噪比的影响较大,增大变换信号的消光比可以减小了码间干扰,探测光功率增加有利于提高信噪比,而不利于提高消光比。SOA交叉增益型波长变换即可以向长波方向变换也可以向短波方向变换。减小SOA偏置电流的时,信噪比恶化严重,码间干扰大。本文进一步研究由半导体光放大器组成马赫-增德尔干涉仪光路实现的波长变换。通过把光路损耗集中处理,解决了光功率在马赫-增德尔干涉仪两臂上分布式损耗所带来的研究困难,不仅研究了全光变换的强度噪声,和相对强度噪声,而且仿真了泵浦波长、信号波长与偏置电流对其变换信号特性的影响。研究得到以下结论:在以光纤为克尔介质时,噪声指数和损耗系数与长度积密切相关。当时,对应的就是损耗集中在输入端的情况;相反当时对应的就是损耗集中在输出端的情况。当在NMZI量子阱限制下,相位不敏感的干涉仪的噪声指数在3dB左右。探测光相对噪声提高因子随信号光功率的增加而增加,当信号光功率到一定值时,提高因子也将会趋于某一定值。探测光功率的增加可以减小探测光相对噪声提高因子,并且随着探测光功率的增加,<WP=3>提高因子逐渐减小到某一定值。在小的信号光功率与大的探测光功率作用下,当Mach-Zender干涉仪信号光与探测光反向输入时,变换信号可获得小的相对强度噪声。另外大的探测光功率与大的信号光功率可以更加有效的获得两臂的相位差,但是大的信号光功率可以加大相对噪声强度,对系统不利。同时大的探测光功也率会对输出信号产生大的啁啾。高探测光功率变换后信号要略好于低探测光功率下的变换信号大的SOA偏置电流下变换后信号要明显好于小的SOA的偏置电流下的变换信号。变换信号既可以向长波长方向变换也可以向短波长方向变换,但是向短波长方向变换后的信号更利于在光纤中传输。如何利用半导体光放大器的高阶非线性实现四波混频效应的波长变换本文也进行了详细研究。在现有的四波混频理论分析方法中引入了高频失谐条件,研究了频率失谐时泵浦波长及泵浦功率对输出变换光脉冲的影响。为简化研究,最终建立了四波混频完成全光波长变换的集总模型,并且通过计算机模拟输出变换信号的光谱图、眼图以及变换信号的转换效率。研究结论显示:当探测光角频率低于泵浦光角频率(角频率差)时,获得的耦合光功率明显较大,这符合能量由高频向低频转移的趋势。耦合光的变换效率大约在负20dB到负30dB之间。对四波混频原理的全光波长变换根据泵浦方式得不同分三种情形进行了讨论,并且对输出信噪比做了比较。根据模拟结果,当信号光为泵浦光时在强泵浦弱探测光的条件下,探测光波长对变换信号波长的影响几乎不存在。当泵浦光功率为2mW,增加探测光功率到1mW时,可以得出变换光的变换效率下降到-70dB以下,效率非常低第四基于物质在光场作用下,电荷分布发生形变形成等效偶极子体系的物理思想,用分段处理并建立简化模型的方法研究了利用AlGsAs的和频效应以及差频效应实现全光波长变换的物理特性。在研究和频变换时,得到了和频变换效率的最优条件。通过研究发现影响AlGsAs的和频效应变换效率除了泵浦光和探测光的功率外,主要是波导损耗的影响,结果与实验结果符合的较好。同样在考虑大的传输损耗的条件下,泵浦光功率和和频光功率成线性关系。所以为了获得大的和频光功率应尽可能的使用大的泵浦功率和信号功率。而差频效应实现全光波长低的变换效率主要来自于均匀反应长度限制,损耗和相位失谐条件的影响,并对相位失谐条件对变换光脉冲的影响进行了计算。首先分析了准相位匹配条件改变对差频变换输出光强的影响。相位匹配的正负<WP=4>失谐都会对差频输出光强造成严重的影响。传输损耗对变换光输出光强也有一定程度的影响。当传输损耗为-6dB时,变换光的强度随器件的长度增加而逐渐增加,当长度达到5mm时增加趋势变缓并逐渐趋于某一定值。传输损耗为-12dB时,开始变换光强度随长度的增加而增加,但是当长度为3.5mm左右时,变换光强度随长度增加逐渐减弱。所以在传输损耗的影响下,器件有一个最佳长度。研究结果说明了利用这个模型所做的理论分析和试验结果符合得较好。最后在上述研究基础上提出了了时分复用与波分复用信道间的全光多路波长变换方案,并论证了方案的可行性。该方案使用了全新的设计思想。仿真结果显示该设计方案可以有效的进行时分复用与波分复用网间的高速数据交换。论文在重点研究全光波长变换理论的同时,也进行了大量的计算机数值更多还原

【Abstract】 Studies of interaction between optical and substance not only have weightilyacademic values but also have widely applied values especially in communica--tion field. In this dissertation physical characteristics of a few all optical wave--length conversion (AOWC)methods are amply researched.In second chapter numerical model of AOWC is established which is base on cross gain modulation in a semiconductor optical amplifier. Nonlinear schrodinger equation is used for studying chirp characteristic in wavelength conversion process. And incidence power, wavelength and bias current affect conversion pulse are simulated. Results as follows: Reducing of SOA chirp in connection with increasing of SOA gain or photon energy; Reducing input light power or increasing length and effective area of active layer, we can reduce the SOA chirp; Code interfering can be reduced when extinction ratio of conversion signal is increased; Enhancing power of prob light, we can improve signal-noise ratio; Wavelength conversion of cross gain modulation of SOA transform along with short wave direction or long wave direction too; When bias current is reduced signal-noise ration will be worsen and code interfering will be augmented.In the next chapter Mach-Zehnder AOWC is researched. Distributed losses of light power in Mach-Zehnder two arms are skillfully dealt by converging method. Intension noise of AOWC is investigated further more relative intension noise. And computer is used for simulating the conversion process. There are five important results to be obtained. The first one is that noise index associate with product of loss coefficient and fiber length. As loss of fiber concentrate in input port, on the contrary, as loss concentrate in output port. The second result is that under NMZI quantum-well qualifications interferometer of nonsensitive phase has about 3dB noise index. The third one is that relative noise factor is increased along <WP=6>with power of signal light, finally it tend to a fix value. Conversion signal can obtain small relative intension noise when signal light and probe light have opposite direction of transmission besides small power of signal and big power of probe. The last one is that conversion signal transform along with short wave direction or long wave direction too but signal of short wave is propitious to transfers more than signal of long wave in fiber. In forth chapter AOWC of four wave mixing (FWM) is detailed investigated with nonlinear high-rank effect of SOA. High frequency unharmonious term is taken in FWM theory. The last collectivity model is set up. Spectrogram, eye-chart and conversion efficiency of conversion pulse are simulated. Big power of coupling light can be obtained when angle frequency of probe light is less than angle frequency of pump light. Base on results of simulation efficiency of conversion is very low about minus 20dB-30dB. Fifthly electric charge distributing is distorted which formed equivalent dipole system when matter is influenced by light field. Found on this physical thought physical characteristics of AOWC about difference frequency and sum frequency are investigated with the method of subsection disposal. Calculation results are in better agreement with experimental datum of literature . Last chapter new all-optical transmultiplexers are researched and feasibilities are demonstrated. And computer is used for designing and simulating the system. Results show transmultiplexers can effective exchange datum between Wavelength Division Multiplexing network and Optical Time Division Multiplexing network . In this thesis the emphasis is theory investigation of AOWC, at the same time, numerical simulation and design are studied too. Holding true areas and Optimization configuration parameters of AOWC are obtained with different methods of wavelength conversion.更多还原