| Abstract: | In transport networks, a multi‐ring architecture is very useful to facilitate network planning and to design and provide more resilient services for customers. Unlike traditional synchronous optical network multi‐rings, the service resiliency of Ethernet‐based multi‐rings is significantly impacted by the ring hierarchy because a link or node failure in a certain level ring triggers filtering database flush actions in all higher level rings as well as in the ring with the failure, and consequently a large amount of duplicated data frames may be flooded. In this paper, we investigate how the ring hierarchy impacts the service resiliency of multi‐ring networks. Based on extensive experiments on various single‐ and multiple‐link failures, we suggest two effective inter‐ring connection rules to minimize the transient traffic and to ensure more resilient multi‐ring networks. In addition, we consider a flush optimization technique called e‐ADV, and show that the combination of e‐ADV and multi‐ring structures satisfying our inter‐ring connection rules results in a more attractive survivability performance. |
