An Evolutionary Multi-agent Preventive Fault-Tolerant QoS Unicast Routing Mechanism

Fault-tolerance needs to be improved and QoS (Quality of Service) routing is necessary in the network, fault-tolerant QoS routing is one of the effective mechanisms. However, due to the dynamic of network parameters, the fuzziness of user requirements, and the imprecise measurements of network parameters, fault-tolerant QoS routing is actually based on the fuzzy information. Therefore, by introducing the knowledge of fuzzy mathematics and probability theory, and using SRLG-disjoint and shared-path protection strategy, in single SRLG (Shared Risk Link Group) failure situation, an evolutionary multi-agent preventive fault-tolerant QoS unicast routing mechanism is proposed. The objective of the mechanism is to find the working path and the protective path under meeting the user QoS constraints. The two paths optimize the end-to-end multiple objectives. Simulation results show that the proposed mechanism is feasible and effective.