Network traffic anomaly detection method based on combination of convolutional neural network and LSTM

Abstract

The invention provides a network traffic anomaly detection method based on the combination of a convolutional neural network and an LSTM, and the method comprises the steps: collecting network data through an SCADA system, carrying out the preprocessing of the data, and carrying out the screening of data meeting the experiment requirements; converting the preprocessed data into a corresponding grayscale image; establishing a CNN-LSTM model, and determining model optimization parameters through a cross entropy minimization mode; and training the CNN-LSTM model by taking the accuracy rate, the true positive rate, the false positive rate and the F1-score as indexes, and evaluating the detection classification effect based on the trained model. The two methods are combined together, detection is successfully carried out, and compared with a traditional machine learning method, a better detection effect is achieved.

Description

technical field [0001] The invention relates to the technical field of network security, in particular to a network traffic anomaly detection method based on the combination of convolutional neural network and LSTM. Background technique [0002] In the digitalization process of smart grid transformation, advanced communication technology is introduced to realize collaboration and information sharing between substations and remote dispatching centers, so that smart networks and smart substations face information that traditional networks are currently facing, such as network intrusion. security threat. The power grid is controlled and managed using a supervisory control and data acquisition system (SCADA). The centralized controller collects information through remote terminal units and issues control commands to actuators in the grid. The interconnection of grid components introduces the risk of cyber-attacks. Neural networks are widely used in anomaly detection to identi...

AI Technical Summary

Problems solved by technology

[0005] However, the above-mentioned existing technologies have no obvious beneficial effect on anomaly detection. In order to prevent problems before they happen, in the face of the current complex network environment, it is urgent to provide an effective method for anomaly detection.

Images