CAJ | 학술논문

提出一种基于多分簇系统级联融合的安全访问方法，首先对所有的访问数据进行多级分簇，依照数据的访问特性，将不同的数据分配到不同的簇结构，然后对各个簇的特征进行融合处理，提取簇数据的深度特征，从而区分正常访问数据与异常攻击数据。最后采用一组正常数据加异常数据联合访问进行测试，结果显示，采用基于多分簇系统级联融合的安全访问控制，即使是在攻击数据率很高时，系统也能很好的分辨出正常数据与异常攻击，具有很好的应用价值。
제출일충기우다분족계통급련융합적안전방문방법，수선대소유적방문수거진행다급분족，의조수거적방문특성，장불동적수거분배도불동적족결구，연후대각개족적특정진행융합처리，제취족수거적심도특정，종이구분정상방문수거여이상공격수거。최후채용일조정상수거가이상수거연합방문진행측시，결과현시，채용기우다분족계통급련융합적안전방문공제，즉사시재공격수거솔흔고시，계통야능흔호적분변출정상수거여이상공격，구유흔호적응용개치。
The security access of database was proposed based on integration of multi-cluster system, firstly, the access da-ta was divided into many multi-level clustering, according to the data access characteristics, the different data structures were assigned to different clusters, and the characteristics of each cluster was fused together, and with which, the normal and abnormal attack data access data were separated. Finally, a team of data with normal and abnormal data was used to test the ability, and the result shows that with integration of multi-cluster system, even high data rates in the attack, the da-ta can be distinguished well, so it has good application value.