A Method of Network Security Risk Assessment Based on Differential Manifold

Abstract

The invention discloses a differential manifold-based network security risk assessment method. According to the method, the network security risks can be assessed objectively and comprehensively; thesecurity of various devices is measured according to indexes of the devices, the operation is objective, index selection involves the reliability, the availability and the connectivity of a network, the coverage is comprehensive, values of the indexes can be automatically acquired and quantified by an acquisition tool, data acquisition is convenient, and the problem that index quantification is difficult is avoided; and a network topology graph is mapped into a high-dimensional differential manifold by means of the characteristic relation between the topology invariance of a network system andthe indexes, a smooth mapping function of the differential manifold is taken as the distance between every two adjacent nodes, depiction on the risk characteristics of the network system is more detailed, dynamic changes of the network system can be depicted, and therefore the security of the network system is dynamically monitored in real time.

Description

technical field [0001] The invention relates to the technical field of network security, in particular to a method for network security risk assessment based on differential manifolds. Background technique [0002] Among the research methods of network security risk assessment, there are many methods that can be used to construct, measure, and evaluate network system security, among which the most widely used and most practical ones are network security assessment methods based on layers and division dimensions and granularity , hierarchical network security assessment methods, and graph-based security assessment methods. [0003] Hierarchy-based evaluation methods, such as: the patent "Security Situation Aggregation Method for Hierarchical Networks" proposed by Liu Dong et al., publication number: CN201110377711; the patent "Bayesian Network Evaluation of Information Security Based on Analytic Hierarchy Process" proposed by Cai Zhiqiang et al. Method", publication number C...

AI Technical Summary

Problems solved by technology

However, hierarchical analysis also has shortcomings: the division of its internal levels needs to rely on the prior knowledge of experts, and the selection of each index itself and the establishment of weights also require expert opinions, which is too subjective.

However, evaluation methods based on dimension, granularity, and hierarchical division also have disadvantages: because of the dimensions, granularity, and levels considered, the collection of indicators is more subtle, which leads to more complex calculations

However, the graph-based network security assessment method also has defects: for example, in the attack graph, there will be a problem of path explosion, which will lead to the problem of excessive calculation, while the probability graph, Bayesian attribute graph, etc. have strong subjectivity. , the problem of large amount of calculation

[0006] To sum up, there are three problems in the existing network security assessment methods. One is that they are relatively subjective, including the division of levels, the determination of dimensions, the selection of granularity, and the determination of indicators and weights, all of which lack objectivity; The coverage problem is not comprehensive enough to evaluate the entire network system as a whole; thirdly, it is difficult to quantify the indicators, and there are many indicator items, so it is impossible to effectively and clearly evaluate network security

The comprehensiveness and objectivity of the network security assessment are very important. If the network risk assessment is not comprehensive, only part of the network system will be assessed, resulting in incorrect assessment results

The lack of objectivity will lead to poor repeatability of the evaluation, the obtained results are not objective, and cannot be compared with other evaluation methods to verify the correctness of the method

As a result, the above three problems lead to the fact that the existing evaluation methods cannot effectively evaluate network security, the calculation results are inaccurate, and the comparability and repeatability are poor.

Images