Research On Internet Of Things Botnet Detection Technology

The rapid development of the Internet of Things has brought great convenience to production and life.However,Io T devices are limited by their own computing power,storage capacity,network bandwidth,and other factors,and the security technology that matches them has not been able to adapt to their development speed,so that there are a lot of insecure vulnerabilities in the Io T The device has serious security risks.The most intuitive manifestation is the proliferation of Io T botnets.Therefore,this article has conducted in-depth research on the detection of Io T botnet traffic.Based on network traffic analysis and neural network related technologies,this paper has improved three aspects in Io T botnet detection.The main research work and innovations are as follows.(1)This paper introduces the development of Io T botnets and the characteristics of botnets from different families,and then analyzes the most typical Mirai botnet in detail.The next step is to compare Io T botnets with traditional botnets and summarize the characteristics of Io T botnets in order to lay the foundation for subsequent detection.(2)In this paper,a botnet detection algorithm based on convolutional neural network is used.The algorithm removes obvious irrelevant data and extracts the overall characteristics of network traffic as a supplement,what't more,batch normalization technology and regular methods are use to improve the convolutional neural network,.Therefore,the detection accuracy and efficiency of Io T botnets have been significantly improved.(3)An Io T botnet feature selection method based on random forest is used.This method extracts statistical features from the original network traffic and filters out the optimal feature subset.Finally an improved convolutional neural network is used for training.Experiments show that the detection model has a certain training speed and generalization ability Promotion.(4)A botnet detection algorithm combining convolutional neural network and long and short-term memory neural network is used.First,the advantages and disadvantages of the convolutional neural network and LSTM are analyzed.The detection model combining the convolutional neural network and LSTM is used to extract the spatial and temporal features of the network traffic,and the architecture of the fusion model is explored.The method has higher detection accuracy.The method proposed in this paper can effectively realize the detection of Io T botnets.Experiments show that the algorithm has higher detection accuracy and lower false positive rate.Finally,it points out the shortcomings of this research and the future improvement directions.