基于行为轨迹属性的软件动态可信模型*

针对软件动态可信度量方法中准确性和效率存在的问题,提出以行为区间划分软件的结构并以行为轨迹属性刻画软件行为的基于行为轨迹属性的软件动态可信模型SBMDB(Software Behavior Model for Dynamic Trustworthiness Based on Behavior Path Properties)。通过对软件功能进行分析,划分软件的行为轨迹区间,提取区间的行为轨迹属性。同时,针对区间的包含、嵌套问题提出了区间化简算法,建立软件的行为模型。该模型以行为轨迹区间为度量基本单位,降低了度量时的整体消耗,提高了度量的效率。实验分析表明,该模型能够精确获取软件的行为信息,有效地检测攻击行为。与其它模型相比,SBMDB可以在保证度量结果准确性基础上提高度量效率。

Software behavior model for dynamic trustworthiness based on behavior path properties

This paper developed a new software model SBMDB(Software Behavior Model for Dynamic Trustworthiness Based on Behavior Path Properties). The model has many software properties which acquired by set software section. The software section is built by software function analysis. To solve the problem about section contain and nest, a section algorithm of the model has been developed in this paper. The model allows the dynamic measurement to be more quickly, because the algorithm reduces the measurement content of the model. Experiment shows that this model can accurately obtain information from software and correctly detect attacks. To compare with other models, this model has lower consumption and higher efficiency in the dynamic measurement.