无线传感网络异常动态信息流过滤仿真

摘要

对传感网络信息流的过滤,能够有效提升传感网络运行安全性.对无线传感网络异常动态信息流的过滤,需要给出信息流的关联特征,计算动态信息流的估计权重,完成异常动态信息流的过滤.传统方法给出信息流功率谱密度函数,获取动态信息流模糊聚类中心,但忽略了计算动态信息流的估计权重,导致过滤效果不理想.提出基于改进混沌关联的无线传感器网络异常动态信息流过滤方法.将每个数据平均互信息量作为重构数据相空间的最佳时间延迟,提取各个数据的关联维数信息,并对数据关联维数信息进行聚类,给出异常动态信息流的关联信息特征,拟合异常动态信息流信息特征,计算出异常动态信息流的估计权重,计算出最接近真实值的估计值来实现异常动态信息流过滤.通过仿真证明,所提方法过滤精度较高,充分地保障了无线传感器网络数据的完整性、可靠性、可用性和安全稳定性.%To filter information flow in sensor network can effectively improve the security of sensor network.A method for filtering abnormal dynamic information flow in wireless sensor network based on improved chaotic relevance is proposed.The average mutual information of data is taken as the best time delay for reconstructing phase space and the correlation dimension information of each data is extracted,and then the correlation dimension information of data is clustered to give related information characteristic of abnormal dynamic information flow.After that,characteristics of abnormal dynamic information flow are fitted.The estimation weight is calculated from the abnormal dynamic information flow.Thus,the estimated value which is most close to the true value is calculated to realize the filtration of abnormal dynamic information flow.Through simulation,we can see that the proposed method has high filtering accuracy.It fully ensures the integrity,reliability,availability and security stability of wireless sensor network.