Edge computing boosts cybersecurity with trust based federated intrusion detection
Cybersecurity in Edge Computing: A Trust-Aware Federated Hybrid Intrusion Detection Framework
Cryptography and SecurityMachine Learning
Summary
Edge computing moves data processing closer to devices but increases risks of cyberattacks on vulnerable devices. The authors have created a combined system that uses smart learning models to detect attacks locally while sharing learning progress without exposing raw data. They also developed a way to trust which devices contribute to learning to prevent attackers from corrupting the system. Tested on multiple cybersecurity datasets, their approach detects attacks accurately and handles attempts to poison the learning process.
What this means in practice
- •For network security teams: Improve edge device threat detection by deploying a federated deep learning framework that respects data privacy and counters model poisoning attacks.
- •For industrial iot operators: Secure distributed IoT networks by implementing trust-aware local intrusion detection that collaborates globally without sharing raw network data.
- •For cloud edge service providers: Offer secure edge computing platforms enhanced by a federated hybrid intrusion detection system resistant to deceptive client updates.$Commercial implications: Enables selling robust edge security services that maintain client data privacy and withstand adversarial attacks on model training.
Authors
Zawad Yalmie Sazid, Robert Abbas
Abstract
Edge computing has emerged as a critical computing paradigm in modern distributed systems by migrating data processing closer to end users and Internet of Things (IoT) devices. While this paradigm decentralizes processes, minimizes latency, and reduces backhaul bandwidth congestion, it exponentially enlarges the cyberattack surface. Heterogeneous, resource-constrained edge devices deployed across unmanaged administrative domains present highly vulnerable targets. To address these vulnerabilities without compromising global data privacy regulations, this paper proposes a novel Trust-Aware Federated Hybrid Intrusion Detection Framework (TA-FHIDF). The proposed framework integrates an Autoencoder, a 1D Convolutional Neural Network (1D-CNN), and a Bidirectional Long Short-Term Memory (BiLSTM) model into a unified, localized deep learning engine capable of autonomous spatial and temporal feature extraction. Model training is performed collaboratively via federated learning, ensuring raw network telemetry remains isolated at local gateways. Furthermore, to defend against adversarial model poisoning attacks, we introduce a robust server-side trust-aware aggregation mechanism that evaluates client reliability using a cosine similarity metric before global model integration. Empirical evaluations across multi-vector benchmark datasets (UNSW-NB15, CICIDS2017, and Edge-IIoTset) demonstrate the framework's superior detection accuracy, rapid convergence, and high Byzantine fault tolerance under adversarial attack scenarios.