AI-Powered Intrusion Detection and Mitigation for Critical Infrastructure in Resource-Constrained Environments – Volume 12 Issue 5

Cybersecurity threats targeting critical infrastruc- tures are rising globally, with developing nations facing unique challenges due to limited resources, outdated technologies, and in- adequate expertise. This paper presents the design and evaluation of an Artificial Intelligence (AI)-powered Intrusion Detection and Mitigation System (IDMS) tailored for such environments. The system integrates multiple machine learning algorithms being Random Forest, Convolutional Neural Networks (CNN), and Recurrent Neural Networks (RNN) trained and tested on bench- mark datasets including NSL-KDD, CICIDS2017, and UNSW- NB15. A modular dashboard was developed to support dataset management, model evaluation, real-time alerts, forensic logging, and live packet inspection. Among the tested models, Random Forest achieved the highest detection accuracy of 98.2%, out- performing CNN and RNN while requiring fewer computational resources. The findings demonstrate that AI-driven IDS can pro- vide practical, scalable, and transparent solutions for resource- constrained contexts, thereby strengthening the resilience of critical infrastructure.

This study demonstrated the feasibility of deploying an AI-powered IDS in a resource-constrained environment such as Lesotho. Random Forest outperformed CNN and RNN in terms of accuracy and efficiency, making it more practical for local infrastructures.

[1]A. L. Buczak and E. Guven, “A survey of data mining and machine learning methods for cyber security intrusion detection,” IEEE Commu- nications Surveys & Tutorials, 2016. [2]A. Khan, et al., “Deep learning approaches for intrusion detection in IoT networks,” Computers & Security, 2023. [3]R. Patel and S. Kumar, “Hybrid CNN-RNN models for anomaly detection in network traffic,” Journal of Information Security, 2024. [4]C. Rudin, “Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead,” Nature Machine Intelligence, 2019.

Call for Paper (Hub) Fast Paper Publication IJCT Indexing & Databases Submit Manuscript Call for Paper: Sep–Oct 2025 Call for Paper: Publish Research