Study On Dynamic Schemes For Optical Network Survivability Based On The Combination Of P-path And P-cycle

Nowadays, the demand of the people for information is increasing sharply, which requires that communication networks can provide greater capacity and higher quality service. So network failure which is a major problem of communications has been got more and more attention.Currently, though survivability technology of the network has been tremendously developed, new ideas and new methods are needed to make up the inadequacy of the existing strategy. So the transformation from static allocation to dynamic allocation and the comprehensive utilization of various survivability technology have become a new research point.In this thesis, firstly, from the view of the significance of the network survivability technology, the concept of network survivability and the performance index, the main contents and the variety of survivability technology, and the analysis of survivability technologies for different networks are discussed. Then two survivability models of optical networks which are developed by the author are introduced.In order to solve the problem that the inaccurate pre-configured P-cycle resources distribution caused by varying traffic in WDM network, the concept of reconfigured P-cycle is developed and the reconfigured P-cycle model based on network performance is established. These problems that the low protection resource utility ratio due to the static P-cycle cann't be reallocated with varying network's status and the congestion caused by inadequate protection resources are resolved by this model. So the protection resources are utilized reasonably best and the network's spending is reduced as possible under the good network status conditions The second model is the combination of dynamic P-path and P-cycle. This model can select the optimum pre-configured structure according to the actual situation of the networks, and determine the reasonable protection resource ratio according to the number of the pre-configured structure. So the redundancy of the network can be reduced largely and a high resource utilization can be achieved by this model. The prominent advantage is that the model can compatible with other models.These two models are simulated on the OPNET simulation platform and based on the pan-European COST239 network topology. The simulation results are in accordance with that of the theoretical analysis, so the advantages of these two models are testified.