The Research Of Smart Home's Intrusion Detection

Since the concept of Internet of things(IoT)was put forward,Internet of things technology has made great strides and become the core force of the fourth industrial revolution.In the development process of Internet of things,a large number of new products are derived,among which smart home is one of the most representative products of Internet of things,which is closely related to the public and makes the people's life become intelligent and convenient.With the increasing demand for smart home products,the output of smart home products is increasing rapidly,and the related security issues are increasingly prominent.Researchers and smart home manufacturers have proposed a variety of security solutions,but the protocols and standards of smart home products from different manufacturers are not uniform,cause that there is no one solution that can be applied to all smart home security problems.Therefore,it is necessary to study the smart home system with high compatibility and high security.The main works of this paper are as follows:The intelligent home system based on blynk platform is designed.The private key is used to encrypt the communication process of the intelligent home device,so as to ensure the security of the transmission content.Aiming at the threat of internal and external network security,this paper designs a smart home security gateway to resist the potential network attacks.The gateway integrates the intrusion detection system,which can handle and alarm the attacks from the internal and external in time.At present,machine learning has become an important part of intrusion detection research.In the research process,researchers also put forward a variety of methods to solve related problems.In this paper,XGBoost algorithm is selected as the core algorithm of intrusion detection.In view of the complexity and multidimensional nature of network traffic,Approximate Information Entropy algorithm is used to reduce the redundant attributes in the traffic,and then XGBoost algorithm is used to train and test the reduced attributes.An intelligent intrusion detection system based on the combination of approximate information entropy and XGBoost algorithm is designed and implemented.In this paper,UNSW?NB15 network connection data set is used as the main data set of classifier building and testing model.Compared with other classification algorithms in the experiment,XGBoost algorithm is more suitable for intelligent home intrusion detection system in classification task.The intelligent security gateway is designed,which integrates the intelligent home server and the intrusion detection system.The gateway uses a unified way to manage the intelligent home system and the intrusion detection system to reduce the resource load of the gateway.Through the simulation attack experiment,the validity of intrusion detection,access control and log function of smart home security gateway is proved.Finally,the intrusion detection system based on the combination of Approximate Information Entropy and XGBoost algorithm is compared with the traditional intrusion detection system Snort.According to the experimental results,the intrusion detection system based on the combination of Approximate Information Entropy and XGBoost algorithm has a higher detection rate on the basis of reducing detection time.