基于软件行为的检查点风险评估信任模型

Trust model of software behaviors based on check point risk assessment

针对当前软件行为可信研究中往往忽略风险因素影响,为提高软件行为可信性评价的准确性和合理性,通过在软件行为轨迹中织入若干检查点,提出一种基于软件行为的检查点风险评估信任模型(CBRA-TM);通过累积多个有疑似风险的检查点,利用风险评估策略,判定有疑似风险的检查点;采用奖励或处罚机制求出软件行为的可信度,最终判断软件行为是否可信.仿真实验结果表明,该模型能够有效地识别软件行为中潜在的风险,能够较准确地计算软件行为的可信度,验证了模型的有效性和可行性.

In the credible research on current software,risk factors are often ignored.For improving the accuracy and reasonableness of the credible research,by insetting a number of checkpoints into the software behavior tracks,we propose a trust model called（CBRA-TM） which is based on checkpoint behavioral risk assessment.This model can determine the credibility of the software behavior by using risk assessment strategies to determine a suspected risk checkpoint,accumulate multiple checkpoints which are suspected and calculate the credibility of software behavior through the reward or punishment mechanism.Simulation results show that the model can identify potential risks and calculate the credibility accurately,which verifies the validity and feasibility of the model.