神经网络视域下互联网安全态势预测方法分析

Analysis of internet security situation prediction method under neural network vision area

传统层次化网络安全态势评估模型在应用时,主要利用入侵检测系统和警报系统发挥态势评估作用,对警告要素关联性缺乏关注。文章以神经网络为基础背景探讨互联网安全态势预测,意在完善层次化态势评估模型,融入模糊层,为提升网络安全态势评估质量提供保障。通过实践分析可知,模糊层构建后,能够通过警报匹配模式衡量警报成功率数值指标,并且进一步对警报威胁性、警报成功率、警报周期3项指标进行明确,确认其对网络安全态势影响程度。最后,在不同模型结构层级上计算出综合警报态势值,达到优化错报漏报问题、提升评估结果准确度的目标。

When the traditional hierarchical network security situation assessment model is applied,it mainly uses the intrusion detection system and the alarm system to play the role of situation assessment,and lacks attention to the relevance of warning elements.This paper discusses the prediction of Internet security situation based on the background of neural network,aiming to improve the hierarchical situation assessment model,integrate into the fuzzy layer,and provide a guarantee for improving the quality of network security situation assessment.According to the practical analysis of this paper,it is known that after the fuzzy layer construction,the numerical indicators of alarm success rate can be measured through the alarm matching mode,and the three indicators of alarm threat,degree success rate and alarm cycle can be further clarified to confirm its impact on the network security situation.Finally,the comprehensive alarm situation value is calculated at different model structure levels to achieve the goal of optimizing the problem of misreporting and underreporting and improving the accuracy of the evaluation results.