传感器网络对攻击数据检测方法仿真

Simulation of Attack Data Detection Method Based on Wireless Sensor Networks

对传感器网络中攻击数据的检测,能够有效提升传感器安全问题。传感器网络对攻击数据的检测,需要组建攻击数据检测适应度值函数,计算出攻击数据的传输频率,完成攻击数据的检测。传统方法得到数据检测的最优目标函数,获得种群的最优解及种群的最优值,但忽略了攻击数据传输频率的计算,易出现检测结果不准确的现象。提出一种新型攻击数据检测方法:基于改进拟牛顿的传感器网络。利用免疫理论将传感器网络中全部的数据节点坐标映射为一个二维矢量平面,将该矢量平面划分为多个互不相交的子区域,计算矢量平面间的偏差,组建传感器网络攻击数据检测适应度值函数,计算出攻击数据的传输频率,结合拟牛顿理论思想对适应度值进行优化,并更新种群的最优解,以此为依据达到本文研究目的。仿真证明,该方法检测的准确率较高,可确保传感器网络的高效运行。

To test attack data in sensor network can effectively improve sensor security. It is necessary to set up the fitness function of attack data detection and calculate the transmission frequency of attack data. In traditional methods, the calculation of transmission frequency of attack data was often overlooked, which results in low detection accuracy. This article focuses on the method for detecting attack data in sensor network based on improved quasi - Newton method. This method used the theory of immunity to map all data node coordinates in sensor network on a two -dimensional vector plane. Then, the vector plane was divided into several subareas which were mutually disjoint. After that, we calculated deviation between vector planes and established the fitness function for attack data detection in sensor network. In addition, we calculated the transmission frequency of attack data and optimized fitness value combining quasi - Newton theory so as to update the optimal solution of population. On this basis, we detected attack data in sensor network. Simulation shows that the proposed method has high detection accuracy. Moreover, this method can effectively guarantee the security and stable operation of sensor network.