June 2025 – We are thrilled to announce that the paper “eIDPS: A Real-Time eBPF-based and Machine Learning-powered Network Intrusion Detection and Prevention Solution” by Stamatios Kostopoulos, Dimitra Papatsaroucha, Ioannis Kefaloukos, and Evangelos Markakis has been accepted at the 6th International Conference in Electronic Engineering & Information Technology.
About the Research
As cyber threats grow in sophistication, traditional Network Intrusion Detection and Prevention Systems (NIDPS) struggle with performance bottlenecks and high false-positive rates. This work, developed by researchers from Hellenic Mediterranean University (HMU), introduces eIDPS, a novel solution leveraging eBPF (extended Berkeley Packet Filter) and Machine Learning (ML) for real-time, high-performance network security.
Key Innovations
- eBPF for High-Speed Packet Processing: Utilizes low-level kernel hooks for efficient, low-overhead traffic inspection.
- Machine Learning for Threat Detection: Enhances accuracy in identifying zero-day and evolving attacks.
- Real-Time Prevention Capabilities: Blocks malicious traffic with minimal latency, suitable for high-speed networks.
- Scalability & Efficiency: Designed for modern cloud and edge computing environments.
Why This Matters
With increasing cyberattacks targeting critical infrastructure, eIDPS offers a next-generation approach to intrusion detection—combining speed, precision, and adaptability. This research bridges the gap between traditional rule-based NIDPS and modern AI-driven security solutions.
Next Steps
The paper will be presented at the 6th International Conference in Electronic Engineering & Information Technology (June 2025). We congratulate the authors on this achievement and look forward to the impact of their work on cybersecurity practices.
