论文部分内容阅读
针对传统信任模型易遭受恶意推荐攻击及提供少数正常服务即可赚取高信誉度的问题,充分考虑无人值守WSN高度自治特点,提出了一种具有激励机制的信任管理模型.利用节点信誉服从Beta分布的理论,计算节点的信任值及其置信度.采用贝叶斯估计方法,将对节点的直接信任与来自邻居节点的推荐信息融合,计算节点综合信任.引入对评价行为的奖惩,激励节点持续地提供真实可信的服务,抵制对信任管理系统的恶意推荐攻击.仿真实验表明,与RFSN相比,引入置信度可避免不良节点通过少数正常行为获得高信任值,在通信量与存储空间显著减少,运算量相似的情况下,判别节点可信性更准确,适合节点资源受限无人值守的传感器网络.
In view of the fact that the traditional trust model is vulnerable to malicious recommendation attacks and provides a small number of normal services to earn high credibility, a trust management model with incentive mechanism is put forward in consideration of the high degree of autonomy of unmanned WSNs. Beta distribution theory to calculate the trust value of nodes and their confidence.Using the Bayesian estimation method, the direct trust of nodes and the recommendation information from neighbor nodes are fused to calculate the overall trust of nodes, and the rewards and penalties, incentives The nodes continuously provide authentic and credible services to resist the malicious recommendation attack on the trust management system.The simulation results show that the introduction of confidence compared with RFSN can avoid the bad nodes from getting high trust value through a few normal behaviors, The space is reduced significantly, and the computational complexity is similar, it is more accurate to determine the node credibility and is suitable for the sensor network with unrestricted node resources.