Research on NFV-based Fault Diagnosis

Given the emergence of big data and Network technologies, it is difficult for traditional Network equipment and service capabilities to meet people's needs. The Network function Virtualization (Network function virtualization) decoupled the hardware and software from the virtualization of other services, thus reducing the cost of hardware and operation and maintenance and increasing the flexibility of services. In the environment of NFV, the network becomes extremely complex, and the hierarchical decoupling of network elements brings great challenges to fault detection and location. To solve this problem, this paper studies the fault diagnosis of NFV. Based on the research status of fault diagnosis, the network architecture, fault detection and location in the NFV environment were analyzed in detail, and the fault detection and location were studied by combining neural network, adaptive diagnosis, multi-layer network structure, Bayesian probability and other fault diagnosis methods. Finally, the future research direction of fault diagnosis is put forward, and the author thinks that multi-fault detection, rapid fault location and detection, and effective fault prediction algorithm will be the future research direction.