The Application Of Artificial Bee Colony Optimization Of BP Neural Network In Intrusion Detection

The Internet has played a more and more important position in people’s daily production and life.while bringing the convenience to the company and individual business, network security has increasingly become a threat. Therefore, the research on the problem of network security is very necessary. According to the characteristic of intrusion detection technology in dynamic response, intelligent control,real-time detection etc.,can be used as an important supplement to the traditional security products.It is currently the main research direction of dynamic security products.According to the characteristics of adaptive,self-learning,self-organization of neural network,it is able to handle some of the background and environment more complex activities.Now based on the neural network method for ntrusion detection has become a trend.This paper discusses it based on neural network method to research on the intrusion detection technology.The traditional BP neural network is easy to fall into local optimum, slow convergence and other shortcomings. To address those issues, researchers went through a variety of genetic algorithm to optimize the neural network. But in the convergence speed and the rate of correct detection has not achieve the desired requirements. In this article, according to the features of global optimization and swarm intelligence of artificial colony bee algorithm, in the neural network parameter initialization, we use the deviation of the neural network as a fitness of artificial bee colony algorithm, select the best fitness of a set of parameters as a nerve power networks and thresholds. Doing so can avoid falling into local optimum neural network and slow convergence problem. The BP neural network model of artificial bee colony optimization applied to intrusion detection, simulation results show that the network model is optimized to accelerate the convergence rate and improve the detection accuracy.the main work of this paper include the following aspects:(1) Analysis the current trend of network security, network security threats has been from the networklayer to the application layer. Explain the basic concepts of intrusion detection, intrusion detectionmodel and composition; the classification of the intrusion detection technology(misuse based andanomaly based intrusion detection) and common methods of intrusion detection are introduced.(2) Introduces the basic concept of BP neural network, analyzes the structure and algorithm of BPneural network. With the traditional BP neural network model to do a simple simulationexperiment. According to the shortcomings of the BP neural network is introduced according tothe characteristics of artificial bee colony algorithm, artificial bee colony algorithm globaloptimization of the BP neural network(weights and thresholds) were optimized.(3) a detailed analysis of the KDD99 data set, extracted from the data set of the data pretreatment, theprocessed data is simulated intrusion detection using BP neural network of artificial bee colonyoptimization, with the traditional BP neural network, BP neural network optimized by geneticalgorithm in convergence speed and detection accuracy contrast ratio, proves the superiority of theoptimization algorithm.