Research On WSN Intrusion Detection Method And Its Optimization Strategy

As an important part of the Internet of things,wireless sensor network has been widely used in military and civil fields,such as environmental monitoring,battlefield monitoring,disaster relief,human health care and so on,and the correspondingsecurity issues are also prominent.As with the traditional network,the wireless sensor network is also faced with the invasion threat of hacker or enemy,and because of its limited node energy,dynamic topology,self-organization and other characteristics,so that some of traditional network security technology cannot be completely applied in wire less sensor network.Therefore,it is very important to study the security technology suitable for wireless sensor networks,and intrusion detection is an aspect of wireless sensor network security research.In this paper,we focus on the characteristics of wireless sensor networks and intrusion detection,analyze some problems currently faced by wireless sensor network intrusion detection methods,and solve these problems.First of all,this paper analyzes the research status of intrusion detection technology in wireless sensor networks,analyzes the characteristics of wireless sensor networks and the characteristics of attacks that have been encountered.For the wire less sensor network data set,there are characteristic redundancy and noise characteristics.At the same time,the CFS feature selection algorithm has the ability of reducing the dimensionality and the GA has the global search characteristic.The GA-CFS attribute selection method is introduced into the wireless sensor network intrusion detection feature selection.The heuristic value obtained by CFS evaluation is used as the fitness function of GA to evaluate the individual,so as to reduce the dimension of the eigenvector,reduce the learning classifier required data volume and enhance the generalization ability.Secondly,In order to improve the detection accuracy and convergence rate of the intrusion detection algorithm for wireless sensor networks based on fusion of PSO and SVM,a wireless sensor network intrusion detection system(CS-CPSO-SVM)based on the fusion of CS-CPSO and SVM is proposed.Chaotic particle swarm optimization algorithm based on complete sine-mapping(CS-CPSO)is used to optimize the parameter of SVM,and the sine-mapping chaotic search is applied to not only the generation of initial population and chaotic perturbation of local optimal for particle swarm optimization algorithm but also the optimization of the inertia weight and the generation of the random constant and the learning factor,moreover,multiple initial values are used to generate a number of chaotic orbits.Finally,the experimental platform is set up,and the rationality of intrution detection optimization strategy is verified by simulation experiment,including GA-CFS feature selection algorithm and CS-CPSO-SVM classification algorithm.The experimental data comes from the Kddcup99 data set At the same time summarize the results of the study comprehensively and clear the next step.