In this chapter, six hierarchical analytic models are developed to evaluate the performance and availability of mobile networks, especially for the next generation of networks based on the Third Generation Partnership Project IP Multimedia Subsystem. Three hierarchical analytic models are developed for the independent software recovery scheme, whereas the other three are proposed for the dependent software recovery scheme. The top level model (described by a reliability block diagram) captures the signaling system consisting of a series of servers in the core network. Each server in the network is modeled in the lower level with alternatives based on (binary) fault tree, multi-state fault tree, stochastic reward net and parallel multi-state systems, the last one solved by the universal generating function methodology. The steady-state system availability is computed to evaluate whether the required performance level is satisfied or not. The accuracy of these models is compared to cross-validate each other. Moreover, the models are compared with respect to the execution time and different input parameters. The influences of the performance demand and redundancy configuration on the system steady-state availability are also evaluated to provide the optimal redundancy configurations for signaling servers, while maintaining high system availability and minimizing the cost
Performance and Availability Evaluation of IMS-Based Core Networks
POSTIGLIONE, FabioMethodology
;
2014
Abstract
In this chapter, six hierarchical analytic models are developed to evaluate the performance and availability of mobile networks, especially for the next generation of networks based on the Third Generation Partnership Project IP Multimedia Subsystem. Three hierarchical analytic models are developed for the independent software recovery scheme, whereas the other three are proposed for the dependent software recovery scheme. The top level model (described by a reliability block diagram) captures the signaling system consisting of a series of servers in the core network. Each server in the network is modeled in the lower level with alternatives based on (binary) fault tree, multi-state fault tree, stochastic reward net and parallel multi-state systems, the last one solved by the universal generating function methodology. The steady-state system availability is computed to evaluate whether the required performance level is satisfied or not. The accuracy of these models is compared to cross-validate each other. Moreover, the models are compared with respect to the execution time and different input parameters. The influences of the performance demand and redundancy configuration on the system steady-state availability are also evaluated to provide the optimal redundancy configurations for signaling servers, while maintaining high system availability and minimizing the costI documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.