Automated Network Vulnerability Detection in IoT: Lightweight IDS Approaches for Enhanced Security
Amol B. Gadewar, Ritesh Vamanrao Patil, Surendra A. Mahajan, Lalit Vasantrao Patil · 2024
Since smart gadgets are quickly surrounding us, the main focus should be on IoT devices' security. Conventional IDS are generally heavyweight models that cannot be implemented particularly on resource-constrained IoT devices. This paper offers a new way of detecting automated network vulnerability in IoT environments using a lightweight conceptualization of IDS. To this end, we suggest a framework that integrates machine learning and anomaly detection designed with IoT -constrained environments in mind. Decision-making is facilitated and computational load is reduced by adopting data weighting, feature sampling, and model adaptation as lightweight processes. We perform several experiments on real-world IoT datasets to determine the performance of our proposed lightweight IDS. Information gathered also shows that this approach provides a better means of identifying weak areas in the network without compromising the devices involved. Further, it has enabled our system to include automatic notification functions that notify the system administrator in real time when it discovers such activities.