基于二值异或群的多轨迹识别算法  

A multi-track separating algorithm based on binary XOR group

作　　者：朱洪亮[1,2] 李锐[1,2] 辛阳[1,2,3] 杨义先[1,2] 徐国爱[1,2] 

机构地区：[1]北京邮电大学网络与信息攻防技术教育部重点实验室,北京100876 [2]北京邮电大学网络与交换技术国家重点实验室信息安全中心,北京100876 [3]北京安码科技有限公司,北京100876

出　　处：《高技术通讯》2010年第10期1033-1039,共7页Chinese High Technology Letters

基　　金：863计划(2009AA01Z439);国家自然科学基金(60821001;U0835001)资助项目

摘　　要：针对二值异或群循环轨迹集没有有效的识别算法的问题,进行了高效快速的多轨迹识别算法方面的研究。首先建立了多轨迹识别问题的数学模型,定义了一类新的轨迹——二值异或群循环轨迹,然后针对这类轨迹提出了一种新的识别算法,将求解与基轨迹相异或的元素值的问题转化为求解轨迹特征位模式的问题,并在理论上证明了其正确性,通过这样的处理极大地提高了算法可操作性,最后将此算法应用于共享接入检测场景中并进行了验证。实验证明,此算法具有较高的准确度和误报容忍度,能较好地应用于网络监控和信息安全领域。For there are no effective separating algorithms at present for binary XOR group circular track sets, the authors conducted the research on a method for rapid and efficient multi-track separation. First, a mathematical model of the multitrack separating theory was built, and a new type of track called the binary XOR group circular track was defined. Then, a new separating algorithm for the kind of track was proposed, by which the problem of solving the element that performs XOR operation with the base track can be converted into deciding the eigen bit mode of the track. It proved to be correct in theory and improved the operability greatly. Finally, the algorithm was applied to the shared-access host number detection to prove its effectiveness. The results indicate that this approach is with the high accuracy and tolerance of mis-report and can be widely used in network monitoring and the information security field.

关 键 词：多轨迹识别 二值异或群 特征位 共享接入 网络监控 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]