无线传感器网络容错性时间同步算法的研究

发布时间：2019-06-21 12:36

【摘要】：时间同步技术在无线传感器网络的数据融合、节点定位跟踪、路由选择、传输调度等应用领域发挥着重要的作用,高精度时间同步技术的研究已成为近几年的研究热点。无线传感器网络节点通常被部署在恶劣的环境条件下,节点间的数据传输易受外界环境因素的干扰,因此,当节点间通信数据受到干扰时,如何使系统具有良好的容错性,确保数据和信息安全可靠的传输是本文研究的重点。在经典同步协议中,泛洪时间同步协议(Flooding Time Synchronization Protocol,FTSP)兼顾了同步精度、能耗、鲁棒性、可扩展性等问题。因此,本文是在FTSP协议的基础上进行分析与改进的,鉴于协议在节点时间同步容错性方面考虑甚少,文章做了以下几方面工作;(1)FTSP用线性回归来估计时钟漂移率,然而线性回归估计法易受异常数据影响,估计值波动较大。针对此问题,本文提出了一种动态加权平均算法,对历史所测有限个时钟漂移率进行加权求解得出新的漂移值,新算法的加权系数是根据节点间收发时间标动态确定,从而减小异常数据对线性回归的影响。(2)针对无线传感器网络节点的时间同步误差随着节点跳数增加而增加的问题,本文采用多源系统代替传统的单源系统来降低节点的跳数,以此来降低节点的同步累计误差。在网络拓扑结构相同的情况下,通过实验仿真分析,多源系统的节点平均跳步数和最大跳步数都比单源系统低,系统节点的累计误差也有明显的减小。(3)针对多源系统提出了一种多源抗攻击算法,算法在FTSP原始同步信息的数据包中加入了发送节点ID和黑名单两个字段,根据同一节点在连续两个周期里时钟漂移率的差值来判断发送节点性质,当差值大于预先设定的阈值时,则认为发送节点为恶意节点,将节点的ID号加入黑名单发送出去,通知等待同步的节点拒绝接受恶意节点的同步信息,以此来提高系统的鲁棒性和安全性。最后,通过仿真结果表明,改进的时间同步算法在时间同步误差、容错性以及安全方面都有较大的改善。
[Abstract]:Time synchronization technology plays an important role in the application fields of wireless sensor networks, such as data fusion, node location and tracking, routing, transmission scheduling and so on. The research of high precision time synchronization technology has become a hot research topic in recent years. Wireless sensor network nodes are usually deployed under harsh environmental conditions, and the data transmission between nodes is easily interfered by external environmental factors. Therefore, when the communication data between nodes is interfered, how to make the system have good fault tolerance and ensure the secure and reliable transmission of data and information is the focus of this paper. In the classical synchronization protocol, flooding time synchronization protocol (Flooding Time Synchronization Protocol,FTSP) takes into account the synchronization accuracy, energy consumption, robustness, scalability and so on. Therefore, this paper analyzes and improves on the basis of FTSP protocol, in view of the fact that the protocol has little consideration in node time synchronization fault tolerance, the following work has been done in this paper. (1) FTSP uses linear regression to estimate clock drift rate, however, linear regression estimation method is easily affected by abnormal data and the estimated value fluctuates greatly. In order to solve this problem, a dynamic weighted average algorithm is proposed to obtain a new drift value from the weighted solution of the limited clock drift rate measured in history. The weighting coefficient of the new algorithm is determined dynamically according to the inter-node transceiver time scale, so as to reduce the influence of abnormal data on linear regression. (2) aiming at the problem that the time synchronization error of wireless sensor network nodes increases with the increase of node hops, In this paper, the multi-source system is used instead of the traditional single-source system to reduce the hop number of the node, so as to reduce the synchronous cumulative error of the node. Under the condition that the network topology is the same, through the experimental simulation analysis, it is found that the average hop number and the maximum jump step number of the multi-source system are lower than those of the single-source system, and the cumulative error of the system node is also obviously reduced. (3) A multi-source anti-attack algorithm is proposed for the multi-source system. The algorithm adds two fields of sending node ID and blacklist to the packet of FTSP original synchronization information. The nature of the sending node is judged according to the difference of the clock drift rate of the same node in two consecutive cycles. When the difference is greater than the predetermined threshold, the sending node is considered to be a malicious node. The ID number of the node is added to the blacklist to notify the node waiting for synchronization to refuse to accept the synchronization information of the malicious node, so as to improve the robustness and security of the system. Finally, the simulation results show that the improved time synchronization algorithm has great improvement in time synchronization error, fault tolerance and security.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP212.9;TN929.5

【参考文献】