基于改进BP神经网络的无线传感网络DOS攻击检测方法

Wireless Sensor Network Attack Detection Method Based on Improved BP Neural Network

传统无线传感网络拒绝服务(Denial of Service,DoS)攻击检测方法的攻击检测率较低,针对这一问题,文章提出基于改进BP神经网络的无线传感网络Dos攻击检测方法。首先,计算在随机状态下数据中特征信息的空间分布函数,完成无线传感器网络数据的特征提取工作;其次,设计无线传感网络数据与处理模块来对数据进行预处理;再次,基于改进BP神经网络完成DoS攻击检测方法设计,实现无线传感网络DoS攻击检测功能;最后,设计对比实验,将该方法与两种传统方法进行比较。实验结果表明:当设计方法在偏差值为0.85时,攻击检测率达到1.0,性能最优,应用效果最好。

The traditional DoS attack detection method for wireless sensor networks has a low attack detection rate.To address this issue,an improved BP neural network based DoS attack detection method for wireless sensor networks is proposed.Firstly,the spatial distribution function of feature information in the random state data is calculated to complete the feature extraction of wireless sensor network data.Then,a wireless sensor network data and processing module is designed to preprocess the data.Next,based on the improved BP neural network,the DoS attack detection method is designed to achieve the DoS attack detection function of the wireless sensor network.Finally,Design a comparative experiment to compare the proposed method with two traditional methods.The experimental results show that when the deviation value of the design method is 0.85,the attack detection rate reaches 1.0,with the best performance and application effect.