Military communication encrypted traffic identification method based on generative adversarial network and model compression

Abstract

The invention provides a military communication encrypted traffic identification method based on a generative adversarial network and model compression. The method comprises the following steps: S10, constructing and training a large model for military communication encrypted traffic identification; s20, establishing a compression model for encrypted traffic identification; s30, constructing and training a generation network for generating encrypted traffic; s40, training a compression model by using the generated encrypted traffic in combination with a knowledge distillation mode to obtain a trained compression model; and S50, carrying out military communication encrypted traffic identification by adopting the trained compression model. The method can solve the problem that the military communication encrypted traffic is difficult to obtain, and can be deployed on military equipment (airborne, handheld and shoulder equipment) with limited power consumption. The method is high in robustness, and the working scene can be a military ad hoc network or a military sensor network.

Description

technical field [0001] The present invention relates to the technical field of wireless communication, in particular to a method for identifying encrypted traffic in military communication based on generative confrontation network and model compression. Background technique [0002] Military communication networks are oriented to military missions and provide reliable and accurate information transmission. The military communication network is different from the general civilian network in: (1) The communication system has higher requirements. Pay more attention to the management of communication, such as individual calls, group calls, emergency calls, multi-priority, over-the-air encryption, end-to-end encryption, etc. (2) Terminal requirements are more stringent. More emphasis is placed on the reliability and security of terminals and the low latency of information transmission, and the volume and power consumption of communication equipment are smaller. With the develo...

AI Technical Summary

Problems solved by technology

The deep learning models mentioned above have achieved high accuracy in identifying encrypted traffic. However, these methods also have some drawbacks.

(1) The deep learning model relies on huge traffic data sets during training. Due to privacy issues, encrypted traffic for military communications is difficult to obtain

(2) The storage space of military equipment (handheld and airborne) is limited, and the deep learning model is huge, making it difficult to deploy deep neural network models on these devices

Although GANs have achieved promising performance in dataset augmentation, few studies have focused on using GANs to solve the problem of unavailable encrypted traffic datasets in military communications

Images