Improvement Of ACO-BP Algorithm And Its Application In Intrusion Detection

With the rapid development of network information technology,people's lives are inextricably linked with the Internet.The network is indispensable for people's work,study,or any aspect of life.Information technology brings convenience to people in all aspects,but the network security problems that come with it cannot be ignored by people.Therefore,research on network security issues becomes particularly necessary.Intrusion detection plays a vital role in network security.Its real-time nature and initiative are indispensable to the current diversification of security issues.In academia,researchers have attached great importance to the research of intrusion detection,and the research on detection methods has become more and more diverse.In the mainstream intrusion detection methods,neural network's characteristics of self-learning,self-adaptation and parallel computing make its application in intrusion detection technology essential.In this paper,the application principle of BP neural network in intrusion detection technology is taken as the starting point.Based on the principle and application of BP neural network algorithm,the optimization BP algorithm is studied.The main works of this article are as follows:Firstly,the characteristics of intrusion detection technology are studied to analyze the BP algorithm.Because the process of adjusting the weights and thresholds in error is based on the gradient descent method,some algorithms are easily trapped in the disadvantage of local optimization.Therefore,the optimization of the weights and thresholds is optimized using the global optimization ability of the ant colony algorithm.Based on the detailed analysis of the working principle of ant colony algorithm,the basic idea of ant colony algorithm optimization BP(abbreviated as ACO-BP)algorithm is discussed in detail.Secondly,in order to improve the optimization ability of ant colony algorithm,the pheromone updating is implemented by a combination of global and local methods.In the formula for globally updating the pheromone,an exponential function is added according to the distribution of the solution to adjust the information residual coefficient so as to update the global pheromone.The residual pheromone of the local pheromone is adjusted according to the minimum error judgment,and the local pheromone is updated.Thirdly,the improved ant colony algorithm is used to optimize the BP algorithm to achieve the optimal choice of weight and threshold.The optimal BP algorithm is applied to the intrusion detection technology.The BP neural network module of the intrusion detection system constructs different BP neural networks for different intrusion features.Fourthly,simulate the optimized BP algorithm.The simulation experiment of the invading samples was carried out through Matlab tool.The effectiveness of the optimized BP algorithm and the correctness rate,false alarm rate and missing rate of attack detection were analyzed.As a result,the improved ant colony algorithm to optimize BP algorithm in this paper has a significant improvement in the detection results.