基于BP神经网络的网络安全评价方法研究.pdf

第29 卷 第8 期 计算机工程与设计 2008 年4 月 Vol . 2 9 N o . 8 Computer Engineering and Design Apr. 2008 BP 基于 神经网络的网络安全评价方法研究 1 2 于 群 ， 冯 玲 (1. 山东农业大学信息学院，271018 ；2. 泰山学院信息科学技术系，山东泰安27102 1) 摘 要：网络安全涉及计算机、通信、物理、数学、生物、管理、社会等众多领域，是一项复杂的系统工程。因此，必须采用系统 工程的思想和方法，对整个网络的安全状况进行综合评价，才能得出科学的评价结果。以层次分析法为基础，构建了BP 神 经网络评价模型，可用于网络安全等级的综合评价，从而得到更科学、合理的评价结果。研究工作为全面评价计算机网络安 全状况提供了新的思路和方法，对网络安全测评认证工作具有较高的理论价值和广阔的应用前景。 关键词：网络安全; 安全评价; 层次分析法; 权值; 人工神经网络 中图法分类号：TP393 文献标识码：A 文章编号：1000-7024 (2008) 08-1963-04 Attribute-weighted clustering algorithm based on rough set 1 2 YU Qun , FENG Ling (1. College of Information Science, Shandong Agricultural University, Taian 271018, China ； 2. Department of Computer Science and Technoloy, Taishan University, Taian 27102 1, China) Abstract ：Network security involves computer, telecommunication, physics, math, biology, management, social and many other fields. It is complex system engineering. Therefore, the idea and method of system engineering is applied to assess the network security comprehensively. Only in this way, scientific assessing results is obtained. BP neural network evaluation model is constructed based on AHP, which can be used on comprehensive assessment of network security. By this way, more scientific and reasonable results can be obtained. The new idea and method are provided for overall assessing security status of computer network system. The study results have important theoretical value and broad application prospects on the evaluation and certification of network security. Key words ：network security; se