中文摘要：

物联网环境下的信任机制是一个亟待研究的重要课题,文中提出物联网环境下层次化的信任架构,满足了不同主体的信任需求,隔离了机构信誉和阅读器信任.使用基于证据理论的方法推导动态运动阅读器的信任,因较短的标签通信距离使恶意事件检测效率较差,文中提出可验证缓存前次交互摘要的方法,有效检测出恶意的终端阅读器.在稳定的机构层,使用信誉机制维护机构信任.层间信任交互构成了＂现象可信-行为可信-节点可信-机构可信-授权可信＂的环流,使得信任得到快速收敛和反馈.实验表明,可验证缓存前次交互信息的方法有效解决了证据理论方法中因物体RFID通信距离短无法被邻居节点检测到的缺陷,层次化的信任机制具有较强的汇聚信任的能力,并有较快的收敛速度.

英文摘要：

The role of the trust model in the Internet of Things（IoT） represents an important domain which is not yet well studied.In this paper,we propose a hierarchical trust model to meet heterogeneous subject trust requirement and isolate organization trust and reader trust,in which the following processes can take place： A verifiable caching interaction digest（VCID） schema is introduced for the purposes of monitoring object-reader interaction,an improved evidence theory is employed to deduce unstable reader-reader trust,and a long-term reputation mechanism is used to manage indirect trust of organizations.A cross-layer trust loop flow is established through the following levels of trust： phenomenon,event,node,organization and authorizing.Within this flow,the reader trust from the lower layer converges up towards the organization reputation on the upper layer by way of supervisor readers quickly.With simulations analyzing the impact of abnormal node ratio,node density,and tag communication range on the organization trust,the hierarchical trust model can effectively detect malicious organizations from their node behaviors and has demonstrated good convergence in the distributed IoT environment.