Analysis of the vulnerability of machine learning regression models to adversarial attacks using data from 5G wireless networks

TL;DR

This study investigates how machine learning regression models used in 5G wireless networks are vulnerable to adversarial attacks, demonstrating significant performance degradation and the effectiveness of binary classifiers in detecting malicious data.

Contribution

The paper presents an analytical framework for assessing regression model vulnerability to FGSM adversarial attacks in 5G network data, highlighting detection methods and impact analysis.

Findings

FGSM attacks increase MSE by 33% and decrease R2 by 10%.

Binary classifiers detect adversarial data with 98% accuracy.

Regression models are vulnerable but can be protected through rapid traffic analysis.

Abstract

This article describes the process of creating a script and conducting an analytical study of a dataset using the DeepMIMO emulator. An advertorial attack was carried out using the FGSM method to maximize the gradient. A comparison is made of the effectiveness of binary classifiers in the task of detecting distorted data. The dynamics of changes in the quality indicators of the regression model were analyzed in conditions without adversarial attacks, during an adversarial attack and when the distorted data was isolated. It is shown that an adversarial FGSM attack with gradient maximization leads to an increase in the value of the MSE metric by 33% and a decrease in the R2 indicator by 10% on average. The LightGBM binary classifier effectively identifies data with adversarial anomalies with 98% accuracy. Regression machine learning models are susceptible to adversarial attacks, but rapid analysis of network traffic and data transmitted over the network makes it possible to identify malicious activity