Research On Intrusion Detection Based On Improved Whale Optimization Algorithm And ELM

With the rapid development of the Internet and the wide application of information technology,cyberattacks are becoming more and more diversified,complicated and functionalized,which makes malicious intrusion detection more difficult.Therefore,it is very important to establish effective security protection measures.Intrusion Detection System is a software or hardware system which can identify attack behaviors and take relevant preventive measures to avoid significant harm to the network.However,the existing intrusion detection system is difficult to deal with the complex attack behaviors,so designing an efficient intrusion detection model has good research and application value.This thesis proposes an intrusion detection model based on improved whale optimization algorithm and ELM to improve intrusion detection performance after in-deep study the relevant knowledge of intrusion detection,whale optimization algorithm and extreme learning machine.The main research of this work is as follows:(1)An improved whale optimization algorithm(IWOA)is proposed to solve the problems of slow convergence speed,low accuracy and easy to fall into local optimum in the later iteration.First,the algorithm uses elite opposition-based learning strategy to initialize the population,which lays the foundation for the fast convergence of the algorithm.Secondly,the inverse incomplete(38)function is introduced to update the convergence factor,which can coordinate the global exploration ability and local development ability of the algorithm.Finally,the crisscross strategy is used to modify the population and global optimal solution,which can reduce the possibility of the algorithm falling into local optimum while ensuring population diversity.The simulation results of eight test functions show that the IWOA Algorithm has higher precision and faster convergence speed.(2)An intrusion detection model based on IWOA-ELM is proposed to solve the problems of poor intrusion detection accuracy and weak generalization ability caused by random initialization ELM input weights and hidden layer thresholds.The model uses the improved whale optimization algorithm to optimize and adjust the ELM input weights and hidden layer thresholds,and then constructs a ELM intrusion detection model based on the optimal parameters.Finally,the IWOA-ELM model is simulated on the KDD CUP 99 dataset to verify the feasibility and effectiveness of the model.The experimental results show that IWOA-ELM model can effectively improve the detection rate and accuracy rate of intrusion detection,and reduce the false alarm rate and the false acceptance rate.