物联网可信感知与识别关键技术研究

发布时间：2018-01-22 23:20

本文关键词： 物联网无线传感器网络可信感知恶意节点虚假数据攻击　出处：《长沙理工大学》2014年硕士论文　论文类型：学位论文

【摘要】：近年来,物联网受到各界广泛的关注,作为一种新型的泛在网络,物联网感知层相比于其他网络,对物理世界具有更深层次、更广范围的接触的特点。因此,感知获得的数据具有“异构、实时、分布、海量”的特点。通过无线传感器获取数据技术是物联网感知层所采用的关键技术之一。感知数据的可靠性、有效性、安全性及稳定性是物联网感知层领域的研究热点和难点。本文在总体介绍物联网的发展背景、相关理论及应用的基础上,从物联网可信感知技术入手,介绍了无线传感器网络的特点,并将无线传感器网络与传统网络进行对比,分析无线传感器网络安全问题的特点以及目前常见的安全防御思想；同时,重点介绍了无线传感器网络易遭受的攻击类型,对目前业界提出的经典算法进行了总结与归纳,并对这些经典算法的优点与不足进行了分析。针对Hur算法容纳度低、所需先验知识多,提出对特征节点分析的恶意节点检测算法(DAFNA)。该算法首先对事件源的能量值估计,且在此过程中过滤保留良性特征节点；然后以特征节点为参照建立坐标系,通过分析待检测节点与事件源的距离计算值与距离感知值之间的差异,进行恶意节点的判断；最后通过仿真实验,对该算法性能进行分析,并与Hur算法对比,得出DAFNA算法所需先验知识少,恶意节点容纳度比Hur算法更好。此外,本文针对虚假数据攻击还提出基于三角形内点几何性质的判断算法。首先将感知节点排序成能量值序列,并由判断节点的后续点对构造一系列三角形的集合；然后分别提取出各三角形内点能量感知最小值,重新构成集合,另外提出可信度函数,根据三角形内点性质与可信度函数提出对提供偏小值的恶意节点判断依据,同时,通过分析三角形内点能量感知最小值集合,根据三角形内点性质,也能得到对提供偏大值的恶意节点判断依据；最后,通过仿真实验,对该算法进行性能分析。
[Abstract]:In recent years, the Internet of things has received extensive attention from all walks of life. As a new type of ubiquitous network, the perception layer of the Internet of things has the characteristics of deeper and wider contact to the physical world than other networks. Perceptual data has the characteristics of "heterogeneous, real-time, distributed and massive". Data acquisition through wireless sensors is one of the key technologies used in the perception layer of the Internet of things. The reliability and effectiveness of perceptual data. Security and stability are the hot and difficult points in the field of Internet of things perception layer. This paper introduces the background of the development of the Internet of things, related theories and applications, and starts with the trusted perception technology of the Internet of things. This paper introduces the characteristics of wireless sensor networks, compares wireless sensor networks with traditional networks, analyzes the characteristics of security problems in wireless sensor networks and the current common security defense ideas. At the same time, the emphasis is placed on the types of attacks that wireless sensor networks are vulnerable to, and the classical algorithms proposed by the industry are summarized and summarized. The advantages and disadvantages of these classical algorithms are analyzed. Aiming at the low capacity of Hur algorithm, the prior knowledge is much needed. A malicious node detection algorithm for feature node analysis is proposed. Firstly, the energy value of event source is estimated, and the benign feature nodes are filtered and retained in the process. Then using the characteristic node as reference to establish the coordinate system, by analyzing the difference between the distance calculation value and the distance perception value between the detected node and the event source, the malicious node can be judged. Finally, through the simulation experiment, the performance of the algorithm is analyzed, and compared with the Hur algorithm, it is concluded that the DAFNA algorithm requires less prior knowledge, and the tolerance of malicious nodes is better than that of the Hur algorithm. In this paper, a judgment algorithm based on geometric properties of interior points of a triangle is proposed for false data attacks. Firstly, the perceptual nodes are sorted into a sequence of energy values, and a series of triangles are constructed by the subsequent points of the judgment nodes. Then, the minimum energy perception of each triangle is extracted, and the set is reconstituted, and the credibility function is proposed. According to the properties of interior points and credibility function of triangles, the judgment basis of malicious nodes providing smaller values is put forward. At the same time, by analyzing the set of minimum values of energy perception in triangles, and according to the properties of interior points of triangles. We can also get the judgment basis of the malicious node which provides the larger value; Finally, the performance of the algorithm is analyzed by simulation experiments.
【学位授予单位】：长沙理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP391.44;TN929.5

【参考文献】