Routing, Wavelength Assignment, Dimensioning and Performance of Optical Networks with Multicast Service

采用波分复用的光通信网络可以使用有限数量的波长信道来支持多个通信会话。在网络的不同部分使用波长复用有助于增加可以支持的会话数量。因此，这些网络可以提供一种手段来服务于那些单独或共同需要大量带宽的应用。具有这种属性的一类重要服务是多点通信，或多点通信，它需要将数据从源传送到一组目的地。虽然一些研究已经解决了在广播和选择光网络中的分组交换多播服务，路由，波长分配，尺寸和性能的波长路由光网络中的电路交换多播服务的主题很少受到关注。 本项目涉及光网络中的多播服务，有和没有波长转换器。这样的应用可以使用波长路由网络来服务。该项目有几个目标。首先，我们要开发的组播树构造算法的光网络，考虑到光网络的约束条件，如功率预算，波长冲突和波长连续性，以及成本和类型的分裂器和波长转换器。在给定流量需求和成本约束的情况下，在多播服务下的最优网络供应和规模确定是另一个目标。我们还计划调查的可能性，制定一个联合问题的最佳供应，最佳wavelengthconverter和分离器的位置。还计划为光网络中的多播业务开发路由和波长选择算法。还计划研究这些算法对网络的尺寸和性能的影响。所有这些都需要建立精确的分析模型，以评估不同路由和波长选择算法下的呼叫阻塞概率。由于多播业务的有效支持需要在网络中的几个点的业务分支，这些模型预计将非常复杂，特别是在这种情况下，不能使用几个标准的独立性假设。除了一般的性能模型，性能上的界限将建立，这将提供一个指导的规定和规模的网络组播服务。这样的界限将建立通过识别极端网络拓扑，这有助于提供这些界限。 这项工作的结果有望促进对未来的，但更多涉及的，交通条件下的光网络行为的理解。同时，这些结果将有助于此类网络的设计和尺寸确定。

Optical communication networks employing wavelength division multiplexing can support severalcommunication sessions using a limited number of wavelength channels. The employment ofwave-length reuse in different parts of the network contributes to the increase in the number of sessionsthat can be supported. As such, these networks can provide a means to serve those applica-tions which individually or collectively require large amounts of bandwidth. An important classof services with such property is the class of multicast, or multipoint traffic, which requires thedelivery of data from the source to a group of destinations. Although a few studies have addressedpacket-switched multicast services in broadcast-and-select optical networks, the topics of routing,wavelength assignment, dimensioning, and performance of circuit-switched multicast services inwavelength routing optical networks have received little attention. This project deals with multicast services in optical networks, with and without wavelengthconverters. Such applications can be served using wavelength routing networks. The project hasseveral objectives. First, we would like to develop multicast tree construction algorithms for opticalnetworks which take into account the optical network constraints into account, such as thepowerbudget, wavelength collisions and wavelength continuity, aswell as the cost and type of splitters andwavelength converters. Optimal network provisioning and dimensioning under multicast service,given the traffic demands and cost constraints, is another objective. We also plan to investigate thepossibility of formulating a joint problem for the optimal provisioning, and the optimal wavelengthconverter and splitter placement. It is also planned that routing, and wavelength selection algo-rithms will be developed for multicast services in optical networks. It is also planned to study theeffect of these algorithms on the dimensioning and performance of the network. All this requiresconstruction of accurate analytical models for the evaluation of call blocking probabilities underdifferent routing and wavelength selection algorithms. As the efficient support of multicast trafficrequires the branching of the traffic at several points in the network, these models are expected tobe very involved, especially that several standard independence assumptions cannot be used in thiscase. In addition to general performance models, bounds on performance will be established, whichwill provide a guide to the provisioning and dimensioning of networks for multicast service. Suchbounds will be established through the identification of extreme network topologies which serve toprovide these bounds. The results of this work are expected to advance the understanding of the behavior of opticalnetworks under futuristic, yet more involved, traffic conditions. At the same time, these resultswill contribute to the design and dimensioning of such networks.