Research On Key Technologies Of Intrusion Detection In Cloud Environment

With the popularization of computer network and the development of information technology,as a new technology,cloud computing is increasingly used in various fields of social life.The powerful computing power and storage resources of cloud computing bring convenience for many businesses and individuals.While at the same time,its huge economic value makes it become the target of many hackers.Today,cloud security issues have become the first challenge in the development process of cloud computing.In the face of new security threats,how to protect the information security of the cloud environment is an urgent problem to be solved in the current situation.Intrusion detection system is an important measure of dynamic network security protection.The intrusion detection system based on cloud computing has become an important part of the clo ud security system,and the intrusion detection system in cloud environment is a new development direction in the field of intrusion detection and cloud computing.This paper analyzes the intrusion detection technology and the characteristics of cloud computing.Then combined with the security requirements of the cloud environment,design a cloud environment intrusion detection system model based on the common intrusion detection standard model CIDF.On the basis of artificial neural network,SOM neural network is introduced into the core algorithm of intrusion detection.Considering that the random initialization of the SOM neural network connection weight vector may reduce the intrusion detection efficiency,this paper proposes optimization algorithm.Firstly,use the particle swarm optimization algorithm with simulated annealing to obtain a group of global optimal solutions,and then set them as the initial connection weight vector of SOM neural network.Secondly,use the SOM neural network to train and classify the model.Build a Hadoop cloud platform,and deploy the designed cloud environment intrusion detection system model.Combine the optimized intrusion detection algorithm and the Map Reduce programming model to make algorithm parallel.The compared experiment validate that the optimized algorithm can effectively improve the efficiency of intrusion detection and ensure the safety of the cloud environment.