Reliability Prediction for Network Systems Using Stochastic Process Algebra
Yu Fu · Xi'an Jiaotong Daxue xuebao · 2011
A reliability prediction method is proposed based on stochastic process algebra(SPA) to deal with the difficulties in the reliability modeling and analysis of network systems with complex structure.SPA modeling of network systems is proposed through analyzing characters of three basic network topologies.Then a labeled transition system(LTS) with time delay marks for the model is generated utilizing the SPA's operational semantics.The prediction value of the system's reliability can be solved out from the LTC's Markovian transition relation and the system's current state.Simulation results show that the proposed method can precisely describe system's stochastic behaviors and interactions among components,and predict the reliability level under different status effectively.