Software-defined Networking-based Distributed Monitoring And Diagnosis Mechanisms And Simulation Implementations

With the rapid deployment of new-generation information technologies such as 5G networks,cloud computing,and the Internet of Things,especially the large-scale application of emerging technologies such as artificial intelligence and big data,network traffic has exploded.Communication networks are developing in the direction of ubiquity,intelligence,high bandwidth and low latency,and network monitoring and diagnosis have become important supporting technologies to ensure the efficient operation of the network.However,existing networks are difficult to effectively monitor and diagnose due to problems such as slow innovation,poor compatibility,and difficult maintenance.The software-defined networking separates the data plane from the control plane,and builds an open new network architecture with open source thinking,which is conducive to better network monitoring and diagnosis.Therefore,the mechanism and simulation of distributed monitoring and diagnosis based on the soft-defined network are deeply studied,and efficient network monitoring and diagnosis are realized.The main research contents and innovative work are as follows:1.Software-defined networking monitoring and diagnosis theory has been studied.By analyzing the theory of software-defined networking,a software-defined networking monitoring method is proposed,and a software-defined monitoring theory of data forwarding layer and network terminal is constructed.A software-defined networking diagnosis method is proposed,and a link fault diagnosis strategy and a switch packet loss diagnosis mechanism are established.Finally,a software-defined networking monitoring and diagnosis theory and method are proposed,which can effectively monitor network equipment and link conditions.It can quickly obtain network status information such as network topology,network connectivity,network control logic,and network performance parameters such as link throughput,packet loss rate,and delay.And on the basis of network monitoring parameters,network diagnosis is performed to effectively determine link failures,port failures and performance failures.2.Research the specific design process of distributed monitoring and diagnosis.Aiming at the problems of poor scalability and fault tolerance of the network single-point control model,a distributed multi-point control method is designed,and the multi-point control strategy design,controller deployment method design and network topology management design are proposed.A distributed monitoring and diagnosis method is constructed,and a monitoring and diagnosis framework design and a monitoring and diagnosis module design are proposed.The global read-write mechanism,atomic snapshot mechanism and data consistency algorithm are studied,and a data consistency design method is proposed.Finally,a distributed monitoring and diagnosis design method was established,and the system reliability was increased by introducing multi-point control,and the consistency problem of host query was solved,and the consistency of data operation was ensured.3.A simulation platform for distributed monitoring and diagnosis was built.Through the deployment of simulation experiment platform,experiment environment,and experiment parameters,the network configuration and topology,port and link monitoring,link port fault diagnosis,link performance diagnosis and other scenarios in the centralized monitoring and diagnosis mode are analyzed.Simulation experiments show that the proposed theory and method can effectively monitor network ports and links,and effectively diagnose link port failures and link performance.At the same time,the path consistency,switch packet loss,host query delay and other scenarios under distributed monitoring and diagnosis mode are studied.Simulation experiments show that the theory and method proposed in the paper are effective and feasible.