摘要
针对传感网络中的入侵数据进行查询,为提高网络数据管理性能和可靠性。由于在进行入侵数据查询时,需要设定固定的过滤器阈值,使得对入侵数据的遍历不全面。传统的查询算法主要通过收集各节点入侵数据,进行数据查询,没有设置过滤器阈值,导致查询出现冗余数据,查询精度低。提出分簇算法的传感网络中入侵数据查询的改进方法。在传统算法的基础上将网络划分为不同的簇,筛选出各个簇的簇头节点,由簇头节点将数据汇聚至Sink节点,并设置过滤器阈值对汇聚至Sink节点上的入侵数据进行过滤,最后利用入侵数据间的关联特征遍历和收集全部网络中和查询数据关联性较强的入侵数据,生成最终的查询结果。仿真结果证明,基于分簇的传感网络中入侵数据查询方法抗干扰性较强,同时提高了入侵数据查询的精度和效率。
It can improve management performance and reliability of network data to query intrusion data in sensor network. Fixed threshold of filter is needed to be set during querying. It makes traversal of intrusion data incomplete. Traditional query algorithm queries data through collecting the intrusion data of each node. There is no filter threshold. It leads to redundant data and low precision. This paper proposes a modified query method of intrusion data in sensor network based on clustering algorithm. Firstly, the network is divided into different clusters based on traditional algorithm to screen out node of each cluster head. Then the data are converged to Sink node through cluster head node. The filter threshold is also set to filter intrusion data converged to Sink node. Finally, the relevant feature trav- ersal between intrusion data and collected intrusion data correlated with querying data strongly in entire network is used to generate final querying result. The simulation results show that the modified method has high anti - interference property. It can improve accuracy and efficiency of intrusion data querying.
出处
《计算机仿真》
北大核心
2017年第2期314-317,共4页
Computer Simulation
基金
国家自然基金(61472268)
国家自然基金(61472211)
关键词
传感网络
入侵数据
数据关联
查询
Sensor network
Invasion of data
Data association
Querying
