A Study on Multifiber WDM Network Architecture

在我们的研究中，我们考虑了一个设计原则的城域核心环，其中的业务量远远超过了一个单一的WDM光纤的容量，因此需要多个光纤。有些人可能在每个核心节点中使用多个WDM光纤和超高容量路由器，尽管他们知道原始设计由于每个节点需要OEO转换和膨胀的路由器容量而非常昂贵。一些高级人员更喜欢在堆栈中使用多个WDM环网。但是，他们也发现每个节点都需要许多OADM。因此，我们看到，没有传统的设计原则给出了一个高容量的城域WDM核心网的成本效益的解决方案。我们相信，多光纤WDM网络（MWN），这已经提出了作者，提供了一个新的机会。然而，MWN需要比这两种设计更多的光纤，而MWN需要简单和经济的OADM。在我们的研究中，我们首次提出了双向多纤WDM网络的基本设计概念。我们表明，双向传输技术可以减半的光纤数量所需的多光纤WDM环网。与传统的双向技术不同，我们的设计不使用WDM交织器。相反，我们使用两组相同的波长在一个单一的光纤上的相反方向，这给出了最好的性能，但一直被认为是不切实际的。我们的数值分析表明，由于瑞利后向散射的光功率损失可以减少小于1 dB的简单的边缘滤波器。我们还研究了一些先进的设计方案。我们相信，以前没有可行的双向设计使用两组相同的波长。最后，我们想强调的是，这些结果使多光纤WDM网络在大容量WDM网络中更具吸引力

In our study, we consider a design principle of metro core ring, where traffic volume far exceeds the capacity of a single WDM fiber, and thus multiple fibers are required. Some people may use multiple WDM fibers and ultra high-capacity routers in every core node, although they know the primitive design is very expensive due to OEO conversion and inflated router capacity needed at every node. Some advanced people will prefer multiple WDM ring networks in a stack. However, they also find that many OADMs are necessary in every node. As a consequence, we see that no conventional design principle gives a cost-effective solution for high-capacity metro WDM core networks.We believe that the multifiber WDM networks (MWN), which has been proposed by the authors, offer a new opportunity. However, MWNs require a more number of fibers than that of those two designs, while MWNs need simple and economic OADMs. In our study, we developed a basic design concept of the bidirectional multifiber WDM networks for the first time. We showed that bidirectional transmission techniques can halve the number of fibers necessary for multifiber WDM ring networks. Our design uses no WDM interleaver unlike conventional bidirectional techniques. Instead, we used two identical sets of wavelengths for opposite directions on a single fiber, which gives the best possible performance, but has long been believed to be impractical.We theoretically analyzed the effect of Rayleigh back scattering noise. Our numerical analysis showed that optical power penalty due to Rayleigh backscattering can be reduced less than 1 dB with simple edge filters. We also investigated some advanced design alternatives. We believe that there is no feasible bidirectional design using two identical sets of wavelengths ever before. Finally, we'd like to stress that the results make multifiber WDM networks more attractive for high-capacity WDM networks for metro areas