【作者】 张群；

【导师】 姚勇；

【作者基本信息】 哈尔滨工业大学 ， 电子科学与技术， 2021， 博士

【摘要】 光性能监测技术,是光通信网络中的关键共性技术,在模块及系统层面提供了重要的故障诊断与性能感知能力,进而支持链路的智能选择和网络资源的优化配置,提高了数据传输的效率和稳定性,为光网络的智能化升级提供了有力保障。现有的光通信性能监测方案主要技术缺陷在于监测维度不足与格式依赖性。随着复杂通信设备和先进调制技术的应用,来自于不同维度的系统损伤相互耦合难以分离。已有方案由于缺乏有效损伤解耦方案,存在监测维度不足、监测参数有限的缺点。同时,为了满足智能光网络的动态业务需求,通信系统会配置多种调制格式,这就要求性能监测方案必须具备调制格式透明的能力。然而,现有方案对于调制信号格式的先验信息和训练序列具有依赖性,无法适用于智能光网络。为了满足智能光网络的性能监测要求,论文围绕着调制格式透明与多参数光的性能监测技术开展研究,采取了多参数损伤建模分析和多算法融合方案设计的研究思路,主要完成如下研究内容:(1)研究了多损伤参数下的接收机模型及损伤耦合机理,探究了传统接收机性能监测方案的优缺点及适用场景。在此理论基础上,提出了基于施密特正交法和戈达尔定时误差检测法的级联接收机性能监测方案,该方案将两者算法按损伤引入的相反顺序组成级联算法结构,实现了调制格式透明的多参数接收机性能监测。(2)研究了多损伤参数下的发射机损伤模型及损伤耦合机理,进一步分析了发射机损伤对于本征损伤预处理模块的影响,包括偏振跟踪和相位恢复模块。随后,提出了基于独立成分分析算法的发射机性能监测方案,该方案实现了对发射机损伤不敏感的偏振跟踪和相位噪声补偿,并能够从算法收敛系数中估计出发射机损伤。借助独立成分分析算法在变量分离/格式透明上的天然技术优势,实现了宽范围、格式透明和多参数的发射机性能监测。(3)在背靠背和短距传输场景下探究了系统本征损伤与收发机硬件损伤之间耦合机理,实现了格式透明的发射机与接收机不平衡损伤的分离和监测。在背靠背场景中,着重探究了发射机和接收机硬件损伤的耦合机理,提出了基于频率偏移和偏振旋转的收发机不平衡损伤分离方案。在短距传输系统中,着重探究了色散与收发机损伤的耦合机理,提出了一种色散不敏感的时钟偏移监测算法用于分离色散和收发机损伤。(4)探究了传统误码率估计方案精度不足的主要原因,提出了基于聚类算法辅助的高斯近似误码率估计方案,使用聚类质心代替标准信号用于计算信号均值和方差,从而降低归一化误差,提升误符号率估计精度。同时,该方案依据编码和星座点分布特征实现了通用的误码率转换,扩展了算法的适用性。论文在收发机多损伤建模与耦合机理分析基础上,实现了调制格式透明和多参数性能监测技术。研究结果为光网络提供了有效的性能监测实施方案,可支持灵活的故障检修和系统运维,对光网络的智能化升级具有重要的应用价值。更多还原

【Abstract】 Performance monitoring is a key and standard technology in optical networks,which provides critical fault diagnosis and performance perception functions for the modules and systems.It enables the intelligent configuration of links and optimal allocation of network resources.Consequently,the stability and efficiency of optical transmission networks can be improved and the intelligent upgrading of the optical networks can be implementable.The main technical issues of the reported optical communication performance monitoring schemes are limited monitoring dimension and modulation format dependence.With the application of complex communication equipment and advanced modulation technology,system impairments of different dimensions are coupled,which makes it difficult to separate multiple impairments.Due to the lack of impairment separation method,the existing schemes have the disadvantage of limited monitoring dimensions and parameters.Meanwhile,different modulation format signals are employed in optical networks to meet dynamic service requirements,which requires that the performance monitoring scheme is transparent to modulation formats.However,the previous schemes are dependent on the prior information of modulation format and training sequence,which is not applicable in intelligent optical networks.To meet the specific requirements on performance monitoring in the intelligent optical network,this thesis is aimed to achieve multi-parameters and format-transparent optical performance monitoring.The research approach of the impairments modeling analysis of multi-parameters and the scheme design of multi-algorithm fusion is adopted and the main research work is as follows:(1)The receiver model with multi-impairments and the impairments coupling mechanism are studied.The characteristics and application scenarios of the traditional receiver performance monitoring schemes are investigated.Based on this analysis,a cascaded scheme based on Schmidt orthogonal and Godard timing error detection algorithms is proposed.A cascaded algorithm structure is formed according to the opposite order of impairment introduction.The modulation format transparent and multi-parameters performance monitoring is achieved for the receiver.(2)The transmitter model and impairment coupling mechanism under multi-parameters are investigated,and the influence of transmitter impairment on the preprocessing module of intrinsic system impairment is further analyzed,including polarization tracking and phase recovery modules.Then,a transmitter performance monitoring scheme based on independent component analysis is proposed.The polarization tracking and phase noise compensation are regardless of transmitter impairments,and the transmitter impairments can be estimated from the convergence coefficients.With the natural technical advantages of independent component analysis algorithm in terms of variable separation/format transparency,the wide-range,format transparent,and multi-parameters performance monitoring is achieved for the transmitter.(3)The coupling mechanism between the system intrinsic impairments and the transceiver hardware impairments is investigated in the back-to-back and short-reach transmission scenarios.A transceiver impairment separation/monitoring scheme with format transparency is proposed.In the back-to-back scenario,the coupling mechanism of transmitter and receiver hardware impairments is emphatically discussed,and a transceiver impairment separation scheme based on the frequency offset and polarization rotation is proposed.In the short-reach transmission system,the coupling mechanism of dispersion and transceiver impairments is studied,and a dispersion-insensitive skew monitoring algorithm is proposed to separate dispersion and transceiver impairments.(4)Regarding the traditional bit error rate estimation scheme,the accuracy issue is theoretically analyzed.A Gaussian approximation bit error rate estimation scheme assisted by a clustering algorithm is proposed.In this scheme,the clustering centroid is used to replace the standard signal to calculate the mean and variance of received signals,which mitigates the normalization error and improves the accuracy of bit error rate estimation.Meanwhile,a general bit error rate conversion method is used to improve the algorithm applicability according to coding and constellation distribution characteristics.With the model of multiple impairments and the analysis of impairment coupling mechanism,the thesis achieves transparent modulation format and multi-parameter performance monitoring technology.The research results can provide an efficient solution for performance monitoring for optical networks,which enables flexible fault maintenance and system operation and has important application potential for intelligent upgrading of optical networks.更多还原