| System Wide Information Management(SWIM)is a large-scale distributed network system using Service Oriented Architecture(SOA),which is used to transmit and share air traffic operation information.In order to deal with the situation that the SWIM system suffers from natural disasters or network attacks,the emergency response mechanism of the SWIM system based on information subscription / publishing services is researched to ensure the continuity of the SWIM service.The research work of this paper consists of two parts.First,based on the research on the SWIM information service process,a SWIM emergency response mechanism based on subscription / publish service(ERMSP)is proposed.Its core is Subscription publishing and trust management mechanisms.Firstly,based on the situational indicators of SWIM operation,the improved network survivability assessment method based on improved fuzzy analytic hierarchy process(IF-AHP)was studied.Secondly,based on the results of SWIM survivability assessment,a Viterbi based on trust value was proposed.Algorithm to make decisions about SWIM service drift.When the SWIM network survivability index drops beyond the parameter's boundary value,the corresponding information is published to the subscribers,and the subscriber decides whether to perform service drift.Simulation experiments show that the model can guarantee business continuity and realize the emergency response of the SWIM system.Secondly,a SWIM-oriented model based on information propagation.Susceptible Infected Recovered Model(SIR)and a task scheduling algorithm based on subscription combination are proposed.From a micro perspective-task scheduling ensures business continuity in the system.Through the subscription and consolidation of the intermediate nodes,the change rate of the server or intermediate node load curve and its parameters affecting the SIR model are calculated,and the related business is scheduled to the normal node,and the completion time,average response time,average waiting time,and Quality of Service(QoS)analyzes the performance of the algorithm. |