Poisoning Attacks in Federated Edge Learning for Digital Twin 6G-enabled IoTs: An Anticipatory Study

TL;DR

This paper investigates how adversaries can perform poisoning attacks on federated edge learning models in digital twin 6G IoT environments, demonstrating significant accuracy degradation under different data distributions.

Contribution

It introduces an anticipatory study of poisoning attacks in federated learning for digital twin 6G IoT, highlighting attack methods and their impact on model accuracy.

Findings

Poisoning attacks reduce accuracy from 94.93% to 85.98% with IID data.

Attacks decrease accuracy from 94.18% to 30.04% with Non-IID data.

Successful attacks can severely impair model performance.

Abstract

Federated edge learning can be essential in supporting privacy-preserving, artificial intelligence (AI)-enabled activities in digital twin 6G-enabled Internet of Things (IoT) environments. However, we need to also consider the potential of attacks targeting the underlying AI systems (e.g., adversaries seek to corrupt data on the IoT devices during local updates or corrupt the model updates); hence, in this article, we propose an anticipatory study for poisoning attacks in federated edge learning for digital twin 6G-enabled IoT environments. Specifically, we study the influence of adversaries on the training and development of federated learning models in digital twin 6G-enabled IoT environments. We demonstrate that attackers can carry out poisoning attacks in two different learning settings, namely: centralized learning and federated learning, and successful attacks can severely reduce the model's accuracy. We comprehensively evaluate the attacks on a new cyber security dataset designed for IoT applications with three deep neural networks under the non-independent and identically distributed (Non-IID) data and the independent and identically distributed (IID) data. The poisoning attacks, on an attack classification problem, can lead to a decrease in accuracy from 94.93% to 85.98% with IID data and from 94.18% to 30.04% with Non-IID.