| Summary: | 碩士 === 國立成功大學 === 電腦與通信工程研究所 === 97 === Recently the requirement of bandwidth increases as technology improvement and the optical technology are maturer than before. An Internet Service Provider (ISP) provides not only the requirement of capacity for clients but also reliability. However, the problem between reliability and cost is trade-off. In order to protect the network from shutdown when failing, the network must be provided spare resource. For the case of WDM network, bandwidth of a fiber grows with the changes of specification from OC-36, OC-48 to OC-192, and the transmission rate of OC-192 rises up to 10Gbps.
There are lots of researches and protection schemes about survival network, such as P-Cycle, Loop-Back recovery, etc. The main contribution of this thesis is that we propose an algorithm which can decompose the network topology into several small sub-topologies, called as Star-Block. This thesis not only introduces implementation of the algorithm but also explains the architecture of star block how to save more spare resource than the one of pure Loop-Back recovery without decomposing topology. In addition considering the decomposed topology, we propose a protocol supporting priority recovery, so that the wavelengths set to be high priority can be recovered faster than the one set to be low priority.
|
|---|
