Hakimi, Rifqy and Reed, Martin J (2025) Feature Analysis and Selection for BGP Anomaly Detection. In: 2025 28th Conference on Innovation in Clouds, Internet and Networks (ICIN), 2025-03-11 - 2025-03-14, Paris, France.
Hakimi, Rifqy and Reed, Martin J (2025) Feature Analysis and Selection for BGP Anomaly Detection. In: 2025 28th Conference on Innovation in Clouds, Internet and Networks (ICIN), 2025-03-11 - 2025-03-14, Paris, France.
Hakimi, Rifqy and Reed, Martin J (2025) Feature Analysis and Selection for BGP Anomaly Detection. In: 2025 28th Conference on Innovation in Clouds, Internet and Networks (ICIN), 2025-03-11 - 2025-03-14, Paris, France.
Abstract
Automated incidence reporting requires accurate and timely anomaly detection. This paper considers this in the context of the Border Gateway Protocol (BGP). BGP is crucial for Internet routing but is vulnerable to attacks due to a lack of widespread authentication. While rare, BGP disruptions can be highly detrimental to Internet performance and security. Detecting BGP anomalies helps network operators protect their networks and improve Internet reliability. In this work, we investigate and develop anomaly detection for BGP using Machine-Learning. We extract relevant features of BGP control plane messages from RIPE RIS and RouteViews public datasets. After extracting features, we employ various feature selection algorithms to extract the most relevant features and explore balancing the datasets. Finally, we explore detection latency, an important operating parameter for automated anomaly detection. The results show that BGP anomalies can be detected in the order of seven minutes when using real attack data and that the observation point for the BGP data has a significant effect on anomaly detection.
| Item Type: | Conference or Workshop Item (Paper) |
|---|---|
| Uncontrolled Keywords: | Technological innovation, Soft sensors, Machine learning, Feature extraction, Routing, Border Gateway Protocol, Security, Reliability, Anomaly detection, Faces |
| Subjects: | Z Bibliography. Library Science. Information Resources > ZR Rights Retention |
| Divisions: | Faculty of Science and Health > Computer Science and Electronic Engineering, School of |
| SWORD Depositor: | Unnamed user with email elements@essex.ac.uk |
| Depositing User: | Unnamed user with email elements@essex.ac.uk |
| Date Deposited: | 05 Jun 2026 13:16 |
| Last Modified: | 05 Jun 2026 13:16 |
| URI: | http://repository.essex.ac.uk/id/eprint/40794 |
Available files
Filename: Feature analysis and selection for BGP anomaly detection.pdf
Licence: Creative Commons: Attribution 4.0