动态传感网络主节点恶意攻击检测仿真

Simulation on the Malicious Attack Detection of Major Node of Dynamic Sensor Network

动态传感网络是一种点对点结构的网络,有着多跳、无中心、自组织网络等特点,主节点会根据需要发生变动,导致网络拓扑结构也随之改变。传统的受恶意攻击主节点检测方法是根据固定拓扑结构设定的属性指标进行检测,针对拓扑结构经常变动的动态传感网络中受恶意攻击主节点检测准确性不高。提出利用加权平均算法的动态传感网络恶意行为检测方法,对节点的变化恶意信息进行加权,将节点特殊标记信息所携带的数据进行加权,使得各个节点的加权因子比较合理,对整个节点网络的加权平均控制在一定范围内,对恶意行为进行检测。通过对改进算法进行仿真验证,结果表明,提出的方法在入侵检测和屏蔽恶意代码攻击方面有着良好的效果。

Dynamic sensor network is a point - to - point structure network, with the characteristics of multi - hop, acentric and self- organizing network. Its primary node changes will occur as needed, resulting in changes of network topology structure. The traditional method of malicious attacks is detected by attribute targets set according to the fixed topology. For the lower detection accuracy of malicious attacks master node in the dynamic sensor network affected by frequent topology changes, this paper presents a method to detect malicious behavior of dynamic sensor networks based on the weighted average algorithm. Firstly, the malicious information of node changes is weighted. Then the carry data of node＇s specific flag information is weighted so that the weighting factor of each node is reasona- ble. Finally, the weighted average of the entire network node is controlled within a certain range, to detect malicious behavior. Through the experiment design and simulation proof, it can be seen that the method of intrusion detection and malicious code attacks blocking has a good effect.