【作者】 于艳莉；

【导师】 李克秋；

【作者基本信息】 大连理工大学 ， 计算机应用技术， 2013， 博士

【摘要】 随着信息技术的飞速发展和网络应用需求的不断扩大,融合网络支持、服务支持和应用支持等多种技术于一体的无线传感器网络,得到工业界和学术界越来越多的关注。在无线传感器网络中,由于传感器节点通常布置在无人监控的开放式环境中,极易受到包括信息窃听、篡改、重放在内的多种方式攻击,其安全问题至关重要。要充分发挥无线传感器网络的优势,必须首先解决其安全性、高可用性等关键问题。在当前网络安全技术的研究中,信任管理机制已成为传统安全技术的有效补充,利用信任管理机制解决无线传感器网络的安全问题是一项极具意义的研究课题。本文在总结信任管理机制代表性研究工作的基础上,总结和阐述了无线传感器网络中信任管理机制在抵御各类攻击方面的优势与不足,分析和比较了典型信任管理模型的研究方法。本文从无线传感器网络的特点出发,针对信任管理机制中尚存在的问题进行深入研究,从信任关系和信任值度量展开,分别提出了基于Rasch理论的信任管理机制、基于主观逻辑信任的被动信任模型和基于信任的安全数据融合模型,主要研究内容和贡献包括：(1)提出一种基于Rasch理论的信任管理机制。针对当前无线传感器网络中的信任模型尚未考虑传感器节点的能力及任务难度的关系,导致信任评估缺乏公平性的问题,引入Rasch基本理论定义节点的适应度。在其基础上计算节点的信任值,解决信任的公平性问题,从而提高信任管理机制在无线传感器网络中的性能,在确保信任机制有效性的同时提高传感器节点的使用效率,达到降低耗能、提高整体网络生存周期的目的。并将该信任机制分别应用于分层结构和平面结构的网络中,提出两种拓扑结构下基于Rasch理论的信任管理算法。通过仿真实验评估模型的性能,并与现有的典型信任管理模型进行性能比较和分析。(2)提出一种基于主观逻辑信任的被动信任模型。针对当前面向服务的信任管理模型缺乏灵活性的问题,将被动信任、直接信任与推荐信任相结合,将三种信任关系融入到面向服务的信任管理机制中,采用被动信任反馈算法有效识别恶意节点和鼓励诚实节点。通过仿真实验分析不同恶意节点比例下的信任值和网络性能,表明该信任模型扩展了服务资源信任关系,有效增强信任管理的灵活性,及时识别不诚实节点和不稳定节点。(3)提出一种新的基于信任的安全数据融合模型。针对当前数据融合方法中的安全可靠问题,对采集数据和融合数据的数据准确性完整性进行验证,由此记录融合节点及子节点的行为信息,由基站对各节点进行信任评估计算,构建候选融合节点表和代表节点表。采用代表节点对单元内相同数据进行传输,能够降低网络耗能。通过实验表明，该模型能够及时有效的识别恶意节点,实现融合数据的安全可靠性,并与现有的安全数据融合模型比明显降低了网络能耗。更多还原

【Abstract】 With the development of information technologies and the expansion of network applica-tion requirements, a wireless sensor network has been focused on by industry and research fields more and more. Many technologies are applied in wireless sensor networks, including network support technology, service support technology and application support technology. A wireless sensor network is a highly distributed network, which is deployed using a large number of sensor nodes in harsh environments. Due to inherent vulnerabilities of easily suffering from external and internal attacks via a variety of methods, including eavesdropping, fabrication, modification of packets, clones of nodes and so on. The security issue is a critical problem in wireless sensor networks. As the effective supplement of traditional security technologies,"trust" has been put forward to apply in deferent networks. A trust scheme in wireless sensor networks has been gradually studied by current researchers and remains an open and challenging field.We categorize the various types of attacks and countermeasures related to trust schemes, give a short summarization of classical trust methodologies, emphasize the challenges of trust scheme in wireless sensor network, and analyze the state-of-the-art trust mechanisms in wireless sensor networks. Considering the characteristics of wireless sensor networks, we provide several trust models in wireless sensor networks. The contributions of the dissertation are listed as follows:(1) In order to deal with the problems of energy consumption and fairness of trust eval-uation in wireless sensor networks, a Rasch-based management trust scheme is proposed. By introducing the Fitness Degree of a sensor node, which is used for judging the relationship be-tween the node latent ability and the task difficulty, our proposed model can rationally enhance the fairness of trust evaluation and save energy of sensor nodes. It is the first trust scheme that measure the fairness of trust in wireless sensor networks. Experimental results show that this scheme effectively distinguishes malicious nodes, increases the utilization rate of individual nodes, and prolongs the lifetime of wireless sensor networks.(2) A novel trust management model based on the subjective logic trust is proposed for service-oriented distributed networks. The proposed algorithm involves passive trust of enti-ty and combines the direct trust and recommendation trust. We also propose a novel scheme called passive trust feedback to avoid the deceit of malicious nodes and unstable nodes, and to encourage honest nodes. Our proposed model expands the trust range of service resource via passive trust of entity, which generates a flexible access path to the service resources. Simula-tions show that the proposed trust management model can significantly improve the feasibility of trust management as well as effectively detect malicious entities.(3) A new trust-based securing data aggregation model is proposed to deal with the security issue of data aggregation in wireless sensor networks. To verify the correctness of integrity of sensing data and aggregation results, we can obtain the behavior records of each aggregator and children nodes. The base station will collect all behavior records of nodes and calculate trust values, and then build the candidate representation node list and aggregator node list via the trust value and residual battery. Simulations show that the proposed approach can identify malicious nodes immediately and decrease the energy consumption of network compared with other approach.更多还原