Research on Dos Attack Simulation and Detection in Low-Orbit Satellite Network

Abstract

Low-Earth Orbit (LEO) satellite networks are increasingly becoming the communication infrastructure of critical sectors because of their extensive coverage and global connectivity. However, due to the unique nature of satellite networks, they face the threat of DoS attacks from malicious actors. In this paper, we propose SDN-based framework and detection method for simulating and detecting DoS attacks in LEO satellite network. In terms of attack simulation, we design dynamic single-path and multi-path forwarding strategies to simulate various DoS attack scenarios encountered in the real network, and construct the marked attack data set. For attack detection, we propose an ensemble learning detection method based on the stacking framework. We incorporate six base classifiers to train and classify the dataset, realize the detection of DoS attacks in LEO satellite networks. Experimental results demonstrate the effectiveness of the attack simulation and detection process. We compare our proposed method with other classifiers and show that the attack simulation framework accurately simulates different types of DoS attack scenarios in LEO satellite networks. Furthermore, our detection method achieves a detection rate of 0.997, validating the effectiveness and feasibility of our proposed approach.

This work was supported in part by the Science and Technology Research Project of the Education Department of Jilin Province under Grant No. JJKH20230850KJ, and the Science and Technology Development Plan Project of Jilin Province under Grant No. 20230508096RC.