Research Of Intrusion Detection System For Smart Device Network

Recently,the appearance and wide utilization of smart device have brought convenience to people's lives and works.At the same time,there are still many security problems in the smart device network,which cause huge losses to society and individuals.How to ensure the reliability and security of information transmission in smart device network and prevent information leakage is the focus of current research.Now,the firewall technology can not satisfy the security requirements of smart device network.In order to locate the malicious behavior before further damage to the system and data,the intrusion detection system is introduced and become a strong barrier of the network security.Machine learning simulates the human cognitive pattern by computers and studies how to handle volume data.Meanwhile,it improves the learning performance by continuous optimization.To use machine learning in the intrusion detection system of smart device network can improve the ability of detection performance with the help of big data mining.Also,this method can reduce the rate of false-positive and false-negative.Machine learning,an effective technique for huge data mining,is becoming a new research area.The main works of this thesis focus on the research of intrusion detection system based on machine learning.First,we investigate the existing machine learning technology and intrusion detection system.Then,the feasibility and necessity of applying the technique of machine learning to intrusion detection system are studied.Based on the decision tree and random forest algorithm,we propose a margin distance optimized random forest pruning algorithm to improve the instruction detection performance with fully considering the characteristics of smart device network.Through theoretical analysis and test data evaluation,it is feasible to apply the margin optimized random forest pruning algorithm to intrusion detection system.The application of machine learning technology in the smart device intrusion detection system provides a great warranty for the usage of smart devices.