Network Intrusion Detection Algorithm and Simulation of Complex System in Internet Environment

This paper proposes a CNN-BiLSTM intrusion detection model for complex system networks. The model performs data over-sampling on the unbalanced data set, which reduces the gap in the amount of category data. It is based on the integration, cooperation, and selectivity of methods and mechanisms in the intrusion detection system, so as to achieve the idea of optimization. In the intrusion detection system, an intrusion detection system based on a variety of detection methods and technologies is proposed, and an integrated, cooperative, and selective overall structure is established. It will be based on distributed intrusion detection and feature engine analysis of intrusion detection, efficiency an increase of 6.7%.