Study On Intrusion Detection With Qpso Optimized Bp Network

With the rapid development of network technology, while it brought convenience for people sharing resources, it also raised a number of network security problems. In recent years, network intrusion events are frequently occurring and information security is facing great threats and challenges. Firewalls, data encryption and other passive safety means of defense have been unable to meet the development needs of network security. Intrusion detection technologies as a means of active and dynamic security defense has become an important issue in future information security research field.Traditional intrusion detection technology still has problems such as single detection method and the detection accuracy is not high currently, therefore try to make neural networks, genetic algorithms, artificial immune particle swarm optimization algorithms and intelligent technology to be applied in intrusion detection. Improved detection methods achieved good results, and they can effectively improve the detection accuracy of the intrusion to some extent.It was analyzed the basic principles of existing intrusion detection techniques and models, as well as the shortcomings. According to the characteristics of BP neural network, it was found that BP neural network is a viable method to be used in intrusion detection. We can take advantages of adaptive learning, distributed parallel storage and other fine features of the neural network technology, and it will improve the deficiencies of the traditional intrusion detection technology to a certain extent.It was analyzed and compared that PSO and quantum particle swarm optimization algorithm still exist the problem which may fall into a local optimum. In this paper, it was presented an improved adaptive mutation quantum particle swarm optimization algorithm, and it improves the convergence of the algorithm to some extent. Because of the shortcomings like slow convergence and it is easy to fall into local minimum of traditional BP algorithm, many kinds of improvements are proposed, and try to use a smart algorithm to optimize BP network parameters, thereby to increase learning the efficiency of BP neural network. The improved quantum particle swarm optimization was used to optimize BP neural network weights and thresholds, then it was got the optimized BP network model.The improved BP network model is applied to intrusion detection in this paper, and a lot of learning and testing samples taken from standard intrusion detection data set to make the simulation experiments. Experimental results show that the convergence of improved BP Algorithm is significantly improved, and it has been improved the efficiency of intrusion detection effectively. Therefore, there are certain advantages and important research value to apply improved BP neural network to intrusion detection.