This paper illuminates the intrusion detection system monitoring process and puts forward that in intrusion detection methods of analysis by the iterative AdaBoost framework, in each iteration, the algorithm produces a belt of the weight values classifier, the iterative end into multiple classifier. Finally, the classifier are weighted joint to get a higher rate of classifier, and hence overcome classification algorithm USES a single produced to meet the requirements of the recognition system defect, so as to improve the system to attack rate,reduce false alarm rate of purpose, in KDD99 were selected as the experimental data. The simulation experiments show that the method is accurate in early warning detection.