基于ABLSTM的SQL注入攻击检测研究  

Research on SQL Injection Attack Detection Based on ABLSTM

作　　者：沈伍强 崔磊 许明杰 杨春松 SHEN Wuqiang;CUI Lei;XU Mingjie;YANG Chunsong(Information Center of Guangdong Power Grid Co.,Ltd.,Guangzhou 510000,China;Guodian NARI Technology Co.,Ltd.,Nanjing 210000,China)

机构地区：[1]广东电网有限责任公司信息中心,广东广州510000 [2]国电南瑞科技股份有限公司,江苏南京210000

出　　处：《微型电脑应用》2023年第3期43-46,共4页Microcomputer Applications

基　　金：南方电网公司科技项目资助(037800KK52190012)。

摘　　要：SQL注入作为最常见的注入攻击方式之一,具有多样性、突变性和隐蔽性,由于传统的攻击检测方法在对电力信息系统提供安全防护方面稍显不足,因此文章提出了一种基于ABLSTM的SQL注入攻击检测方法。一方面,通过数据正样本生成非平衡数据集来平衡数据的分布,缓解过拟合;另一方面,引入Attention机制到Bi-LSTM模型中,在进行特征选择时有效增强关键特征的权重,提高分类的准确性。通过对比实验验证,所提出的方法在检测效果和准确性方面相比其余方法具备显著的优势。As one of the most common injection attacks,SQL injection has the characteristics of diversity,mutability and concealment.The traditional attack detection methods are slightly inadequate in providing security protection for power information system.Based on this,this paper proposes an SQL injection attack detection method based on ABLSTM.On the one hand,the generation of positive data samples expands the unbalanced data set which will balance the distribution of data and alleviate overfitting.On the other hand,the introduction of Attention mechanism into the BI-LSTM model can effectively enhance the weight of key feature during efficient feature selection,which can improve the accuracy of classification.It is verified by comparative experiments that the method proposed has significant advantages compared with other methods in the detection effect and accuracy.

关 键 词：SQL注入 非平衡数据集 过拟合现象 Bi-LSTM 

分 类 号：TP333[自动化与计算机技术—计算机系统结构]