Fault detection method and device for network management system

Abstract

The invention discloses a fault detection method for a network management system, and the method comprises the following steps: 1, obtaining an original unbalanced network fault data set NF, carrying out the compression and filtering operation of the NF, and obtaining a processed data set NF', wherein the NF' comprises samples, the samples comprise two types of samples, and the two types of samples are non-fault samples and fault samples; 2, acquiring an imbalance ratio I, wherein the imbalance ratio I = the number of non-fault samples / the number of fault samples; 3, presetting the proportion of the number of two types of samples after sampling, and obtaining the number N of newly synthesized fault data samples; 4, constructing a spatial distribution rule of a newly synthesized fault data sample; 5, constructing a synthesis rule of a newly synthesized fault data sample; and 6, establishing an XGB-RF network fault detection model. According to the invention, the accuracy of network fault detection is effectively improved.

Description

technical field [0001] The invention relates to the technical fields of electronics, communication and information engineering, in particular to a fault detection method and device for a network management system. Background technique [0002] Faced with the current increasingly complex and large-scale network systems, the difficulty of network maintenance has also increased exponentially. Once the network fails, it will cause disastrous consequences no matter in terms of military defense, business, or residents' daily life. Estimated loss and inconvenience. Therefore, how to effectively use the current technology, discover faults, diagnose faults, and restore faults is of great significance to ensure the smooth operation of the network management system and maintain the normal and orderly development of the social environment. [0003] Some traditional data analysis methods are difficult to obtain satisfactory results in the face of increasingly large and complex network b...

AI Technical Summary

Problems solved by technology

However, since SMOTE treats all minority samples equally and does not consider the category information of neighboring samples, sample aliasing often occurs, resulting in poor classification results.

In addition, if the selected minority class samples are surrounded by minority class samples, the newly synthesized samples will not provide much useful information; if the selected minority class samples are surrounded by majority class samples, such samples may be noise, then The newly synthesized samples will have a large overlap with the surrounding majority class samples, making classification difficult

Images