无线传感器网络数据融合安全问题的研究

发布时间：2018-08-13 10:45

【摘要】：数据融合作为一种节能技术在无线传感器网络中得到了广泛应用,它能够有效地去除传感器节点在信息采集阶段产生的冗余数据,降低节点间的通信开销和计算负载,在延长网络工作寿命的同时,能够有效缩短系统决策时间。但是,数据融合技术的广泛应用也使无线传感器网络在数据机密性、隐私性和完整性保护等方面暴露出了很多问题。攻击者能够通过窃听、篡改、伪造等恶意行为来破坏网络安全性能。因此,如何解决数据融合过程中出现的安全问题已成为无线传感器网络研究领域内的热点课题。论文通过对无线传感器网络数据融合技术和网络安全技术进行研究和分析,将研究工作的重点集中在数据隐私性保护技术、节点信誉评价机制、同态加密技术以及同态MAC技术上。论文的研究工作得到了国家自然科学基金项目(No.61371071)、北京市自然科学基金项目(No.4132057)、北京市教育委员会学科建设与研究生建设项目的支持。论文的主要研究工作和成果如下：1、提出了基于数据动态分割混合技术的安全数据融合方案ESMART。该方案通过对原始采集数据进行动态差异性分割来为无线传感器网络的数据隐私性提供有效保护。在ESMART方案下,窃听者难以捕获到完整的数据切片,也无法对窃听到的数据切片进行有效复原。该方案在提高隐私数据保护性能的同时,还能够有效地控制节点间的数据传输量,降低了节点的通信开销和计算负载。仿真结果表明,与同类型的数据隐私性保护方案相比,ESMART方案在数据融合精确度、能量消耗水平等方面都有着明显的性能优势,具有很好的实用价值。2、为了抵御网络内部攻击,论文提出了基于信誉的安全数据融合方案SRDA.该方案通过部署Watch Dog和局部离群值检测等机制来对节点行为进行实时监视,并针对节点特征行为提出了相应的评价指标。当节点行为信誉值低于预设门限值时,信誉评价机制将判定该节点为恶意节点并将其隔离。仿真实验数据表明,SRDA方案在节点能耗水平、数据融合精确度以及恶意节点识别效率等方面都有着明显优于其他同类型安全方案的性能水平。3、提出了基于同态加密技术的安全数据融合方案ECC-SDAI,方案中使用了具有加法同态性的椭圆曲线加密机制,保证了节点隐私数据在数据融合过程中始终处于密文状态,融合节点可直接对密文数据进行融合操作,有效地保护了数据的机密性和隐私性,实现了端到端的安全数据融合。在此基础上,ECC-SDAI方案通过构建两个相互独立的树形融合网络,使基站能够对数据融合结果的完整性进行有效验证。仿真实验和安全性能分析结果表明,ECC-SDAI方案在网络安全强度、节点通信开销以及数据融合精确度等方面都有着很好的性能表现。4、提出了基于同态消息认证码技术的安全数据融合方案HMSDA,该方案使用具有加法同态性的对称加密机制来为数据融合过程提供数据机密性和隐私性保护。同时,方案中部署的同态消息认证码机制能够对数据融合结果的完整性进行有效验证。仿真结果表明,与同类型的安全数据融合方案相比,HMSDA方案在数据融合精确度、通信开销以及计算负载等方面都具有一定的优势。
[Abstract]:As an energy-saving technology, data fusion has been widely used in wireless sensor networks. It can effectively remove redundant data generated by sensor nodes in the information acquisition phase, reduce communication overhead and computing load between nodes, prolong the network lifetime and effectively shorten the system decision-making time. The wide application of fusion technology also exposes many problems in data confidentiality, privacy and integrity protection of wireless sensor networks. Attackers can destroy network security by eavesdropping, tampering, forgery and other malicious acts. Therefore, how to solve the security problems in the process of data fusion has become wireless. This paper focuses on data privacy protection technology, node reputation evaluation mechanism, homomorphic encryption technology and homomorphic MAC technology through the research and analysis of data fusion technology and network security technology in wireless sensor networks. The project supported by the National Natural Science Foundation of China (No. 61371071), the Beijing Natural Science Foundation of China (No. 4132057) and the Beijing Education Commission (Beijing Education Commission) is presented. The main research work and achievements of this paper are as follows: 1. A security data fusion scheme ESMART based on dynamic data segmentation and hybrid technology is proposed. It is difficult for eavesdroppers to capture complete data slices and restore eavesdropping data slices effectively under ESMART scheme. This scheme can improve the performance of privacy data protection while at the same time improving the performance of privacy data protection. The simulation results show that ESMART has obvious performance advantages in data fusion accuracy and energy consumption level compared with other data privacy protection schemes. 2. It has good practical value to resist the network. In this paper, we propose a reputation-based secure data fusion scheme SRDA, which monitors node behavior in real time by deploying Watch Dog and local outlier detection mechanisms, and proposes corresponding evaluation criteria for node characteristic behavior. The simulation results show that SRDA scheme is superior to other similar security schemes in energy consumption level, data fusion accuracy and malicious node identification efficiency. 3. A secure data fusion scheme ECC-based on homomorphic encryption technology is proposed. SDAI, which uses the elliptic curve encryption mechanism with additive homomorphism, ensures that the node privacy data is always in the ciphertext state in the process of data fusion. The fusion node can directly fuse the ciphertext data, effectively protecting the confidentiality and privacy of the data, and realizing the end-to-end security data fusion. On the basis of this, the ECC-SDAI scheme constructs two independent tree fusion networks, so that the base station can validate the integrity of the data fusion results effectively. Simulation results and security performance analysis show that the ECC-SDAI scheme has good network security strength, node communication overhead and data fusion accuracy. 4. A secure data fusion scheme HMSDA based on homomorphic message authentication code is proposed, which uses symmetric encryption mechanism with additive homomorphism to provide data confidentiality and privacy protection for data fusion process. The simulation results show that the HMSDA scheme has certain advantages in data fusion accuracy, communication overhead and computing load compared with the same type of secure data fusion scheme.
【学位授予单位】：北京交通大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TN929.5;TP212.9;TP202

【相似文献】