Node importance measurement method for network measurement

Abstract

The invention provides a node importance measurement method for network measurement. According to the method, on the basis of a weighted aggregation coefficient and infectious disease dynamics, firstly, the environment where network nodes are located is divided into regions according to the actual situation, then equipment accessing the network is classified according to state types in an infectious disease dynamics model, and then regional information entropy and local aggregation coefficients of the nodes are calculated on the basis of the classified equipment; and finally, the importance of the nodes can be obtained through arrangement. The design of the invention can be used for simplifying the topology when measuring the network performance parameters, can effectively reduce the number of measurement data packets sent during external measurement, and reduces the measurement workload. In a complex topology, the importance of the nodes is measured by weighting the clustering coefficient, and the topology is simplified by omitting the nodes with lower clustering coefficients, so that network fragments which do not need to be excessively concerned can be omitted. When network measurement is carried out, links in some sparse areas can be combined into one link, and the performance of each link does not need to be measured in detail, so that the workload is greatly reduced. In addition, only a part of the disclosed topology can also play a role in protecting personal interests of infrastructure suppliers.

Description

technical field [0001] The invention relates to a node importance measurement method, in particular to a node importance measurement method based on weighted aggregation coefficient and infectious disease dynamics, which will consider environmental perception. Background technique [0002] With the coordinated development of technologies such as mobile communications, Internet of Things, cloud computing, and artificial intelligence, people's lives are increasingly inseparable from Internet technology. In addition, with the development of the Internet, its business volume has also increased sharply, and the corresponding network complexity has also increased day by day. Among them, protocols, topologies, and traffic are important components of the Internet. In-depth research and better understanding of them are of great and far-reaching significance to the operation, maintenance and development of the Internet. For example, the sensitive awareness of network topology and tra...

AI Technical Summary

Problems solved by technology

Passive measurement is based on the existing traffic data in the network, which can reduce the measurement cost, but usually cannot estimate the accurate value of link performance parameters

[0006] In an increasingly complex network environment, the link-level active measurement of the entire network has the following known problems: a large number of probe packets will consume too much bandwidth resources, which will cause interference to devices that need to respond in a timely manner, so that the network The actual behavior may be disturbed, so that the measurement data cannot reflect the true state of the network; in addition, it is obviously impossible to measure the performance parameters of each link in the gradually expanding network scale

[0011] (1) The traditional evaluation method of sorting nodes according to certain attributes is relatively one-sided;

[0012] (2) The measurement method that only considers the structural attributes of nodes cannot meet the needs of network problems;

[0013] (3) Most measurement methods do not consider the perception of the environment

Images