Network abnormal traffic detection method and system, and storable medium

Abstract

The invention discloses a network abnormal traffic detection method and system and a storable medium, and relates to the technical field of computer network security. Obtaining to-be-detected network flow data, and dividing the to-be-detected network flow data into a training sample and a test sample; inputting the training sample into a multi-target deep reinforcement learning model to train a plurality of Actor-Critic network models; updating parameters of the Actor network and the Critic network by adopting a strategy gradient and a loss function respectively, and storing an updating model; and testing a network traffic test sample through the model, and outputting an abnormal traffic detection result. The method does not depend on a high-performance GPU (Graphics Processing Unit), can quickly train and predict on the CPU, and can remarkably reduce computer resources. In addition, the multi-target deep reinforcement learning model constructed by the invention not only has better convergence, but also can more effectively learn in a high-dimensional and continuous action space, thereby improving the network abnormal traffic detection efficiency and accuracy.

Description

technical field [0001] The invention relates to the technical field of computer network security, and more specifically relates to a method, system and storage medium for detecting abnormal network traffic. Background technique [0002] With the rapid development of Internet technology, the current P2P, streaming media, online games and various new mobile Internet applications have accounted for more than 60% of the network traffic, while new services are also increasing, making the application layer protocol more complex, Abnormal network traffic poses a serious threat to network security. Therefore, how to effectively implement network management and control, traffic anomaly detection, and network planning and construction in the era of data explosion is an urgent problem to be solved. As the basis of network security prevention, abnormal traffic detection technology has become an important technical means of network management. [0003] However, as the rapid growth of n...

AI Technical Summary

Problems solved by technology

When the input data is an image or sound, it often has a high dimensionality, which is difficult to deal with by traditional reinforcement learning

In addition, value-based deep reinforcement learning mainly faces the following two problems: (1) it is difficult to deal with continuous action spaces; (2) it is difficult to learn stochastic policies

However, AC is extremely unstable during the training process, and there is also a problem that it is difficult to converge.

Images