<![CDATA[In challenging environments with intermittent connectivity and the absence of end-to-end paths, delay tolerant networks (DTNs) require robust security measures to safeguard against potential threats. This study addresses these issues by implementing an intrusion detection system (IDS) enhanced with machine learning techniques. Common threats such as distributed denial-of-service (DDoS) and flood attacks are tackled using datasets like network intrusion detection (NID) and flood attack datasets. Multiple machines learning methods, including k-nearest neighbors (K-NN), decision trees (DT), logistic regression (LR), and others, are utilized to improve detection accuracy. A PyCaret-based approach is developed to increase efficiency while preserving attack detection accuracy in DTNs. Comparative research demonstrates that PyCaret outperforms Scikit-learn models, and the artificial bee PyCaret classifier (ABPC) optimizes hyperparameters to improve model performance. NS2 simulation shows the system's ability to thwart attacks, offering useful insights into DTN security and improving communication reliability in various situations.]]>
Copyrights © 2025
