Cognitive optical networks enabled by coherent technologies and filterless concepts

相干调制解调器性能和数字信号处理（DSP）技术的最新发展刺激了对新型敏捷网络架构的探索。申请人小组提出了一种无源广域网（WAN）解决方案，称为无滤波器光网络，作为有源光交换网络解决方案的一种经济有效且可靠的替代方案。所提出的光网络架构是基于这样一个前提，即可以通过发射器的波长调谐和相干接收器的波长识别来提供对敏捷性的需求，这与无线电网络中实现敏捷性的方式大致相同。无滤波器相干网络本质上提供无源广播介质，其中无源非滤波光分配器和合并器在一些节点上用于光纤链路互连。本课题是在无滤波器光网络理论研究和本课题组先进的光层试验台的基础上提出的。该计划旨在提出基于无滤波器和认知概念的相干光网络设计和控制的新方法，并在理论和实验上证明这些概念在实际应用中的适用性以及对成本、性能和能耗的影响。在第一阶段，该项目将探索无滤波器网络设计和控制平面的认知光网络概念，作为通过传输质量、网络状态和能耗意识更好地路由流量和分配资源的潜在手段。在我们的无滤波器设计和控制平面工具中，将测量和集成分支树无滤波器架构中特定于相干传输的损害。在第二阶段，该项目将专注于候选认知柔性光网络架构的设计。将组装一个小规模的无过滤器网络。将设计具体的实验来验证这些新颖的网络概念，并在资源利用、损伤感知和信号质量评估方面评估使用认知的动态性能和效益。拟议研究项目的独创性在于无滤波器光网络概念本身，它看起来越来越像未来高容量光网络的非常重要的候选架构，也因为新的认知网络架构，由连贯的技术和节能设计实践实现，将从理论和实验的角度进行探索，从而允许以新的方式使用可用带宽。最大化网络灵活性。其他一些创新包括：新的设计方法和基于灵活和认知光网络概念的控制平面；新的设计工具和适应相干网络的模拟器；面向内容交付网络的高性价比多播架构新的应用，特别是考虑到无滤波器网络可能比基于传统有源光子交换和光色散补偿元件的光交换网络表现出更低的线路成本和延迟，更高的可靠性，更小的占地面积和更少的节点设备复杂性。本研究计划预计在5年内完成。共有3名博士生和4名硕士生将参与该研究项目。

Recent developments in coherent modem performance and digital signal processing (DSP) technologies have stimulated the exploration of novel agile network architectures. The applicant's group has proposed a passive wide area network (WAN) solution, called the filterless optical network, as a cost-effective and reliable alternative to active optical switching network solutions. The proposed optical network architecture is based on the premise that the need for agility can be provided by wavelength tuning at the transmitter and wavelength discrimination by the coherent receiver, in much the same way as agility is achieved in radio networks. Filterless coherent networks essentially offer a passive broadcast medium in which passive non-filtered optical splitters and combiners are used at some nodes for interconnecting the fiber links. The proposed research program builds on both the theoretical research on filterless optical networking and the advanced optical layer test bed developed by the applicant’s group. The program is aimed at proposing novel approaches for coherent optical network design and control, based on filterless and cognitive concepts, and at demonstrating, both theoretically and experimentally, the applicability of these concepts in real-world applications and the impact on cost, performance and energy consumption. In the first phase, the project will explore cognitive optical networking concepts for the design and for the control plane of filterless networks as a potential means to better route traffic and assign resources through quality of transmission, network status and energy consumption awareness. The impairments that are specific to coherent transmission in a branch-tree filterless architecture will be measured and integrated in our filterless design and control plane tools. In the second phase, the project will concentrate on the design of candidate cognitive flexible optical network architectures. A small scale filterless network will be assembled. Specific experiments will be designed for validating these novel networking concepts and evaluating the dynamic performance and benefits of using cognition, in terms of resource utilization, impairment awareness and signal quality evaluation. The originality of the proposed research project lies in the filterless optical networking concept itself, which is looking increasingly like a very serious candidate architecture for future high capacity optical networks, and also because novel cognitive network architectures, enabled by coherent technologies and energy efficient design practices, will be explored from both the theoretical and experimental perspectives, thereby allowing the use of the available bandwidth in new ways, maximizing network flexibility. Some additional innovations are: new design methods and a control plane based on flexible and cognitive optical networking concepts; new design tools and a simulator adapted to coherent networks; cost-effective multicast architectures for content-delivery networks; and new applications, particularly given that filterless networks are likely to exhibit lower line cost and latency, higher reliability, smaller footprint and less node equipment complexity than optical switching networks based on conventional active photonic switching and optical dispersion compensation elements. The proposed research program will be realized in an expected time frame of 5 years. In total, 3 Ph.D. students and 4 Master's students will be involved in the research program.