A dynamic and minimal overhead framework for securing IoT devices through intrusion detection and encryption

The explosive growth of Internet of Things (IoT) networks has intensified the demand for security mechanisms that are both robust and resource-efficient, particularly for devices with limited computational power and battery life. Traditional cryptographic protocols and intensive intrusion detection systems are often unsuitable for such constrained environments. To address this challenge, we present a novel Lightweight Hybrid Security Framework (LHSF), which combines Elliptic Curve Cryptography (ECC) with a feature-quantized One-Class SVM-based detection module, known as LADS. Experiments conducted using the UNSW-NB15 and IoTID20 datasets—spanning both simulated and real-world IoT traffic—demonstrate that LHSF completes encryption and decryption in 17.8 ms per cycle on a Raspberry Pi 4, reducing energy consumption by 52% compared to RSA-1024. The anomaly detection component achieves 98.64% accuracy, 98.21% precision, and a false positive rate of just 1.07%, all within a 3.1 MB memory footprint and less than 43 ms of added latency. We also introduce a Security-Efficiency Score (SES), which balances performance, latency, and energy efficiency. LHSF achieves an SES of 0.912—an 18% improvement over existing models—demonstrating its viability as a secure, scalable solution for next-generation IoT systems.