Improving Intrusion Detection Accuracy in Campus Networks: A Dataset Driven by Real-Time Traffic and Honeypot Simulations
Abstract
This article describes the creation of a domain-specific Intrusion Detection System (IDS) dataset customised for campus networks to overcome the constraints of out-of-date public datasets such as KDD'99 and NSL-KDD. The dataset depicts the various user behaviours, traffic patterns, and device interactions that are unique to educational contexts because it captures network traffic straight from a university. Real-time logs from firewalls, routers, and switches are used as data sources, as is the simulated attack traffic collected by honeypots, which are false open network ports meant to entice malicious behaviour. This technique ensures a balanced mix of normal and attacking actions. Machine learning models trained on this dataset have a 99% detection rate, exceeding models trained on public datasets (95%), while also lowering false positives. The dataset is continually updated to reflect changes in user behaviour, software, and threats, maintaining its long-term usefulness. This work establishes a realistic, adaptive, and effective framework for developing scalable IDS models designed for campus network protection.
