SECURING LOW-POWER EDGE AI: A VULNERABILITY ANALYSIS AND CYBERSECURITY FRAMEWORK FOR RESOURCE-CONSTRAINED DEVICES

The rapid deployment of Edge AI systems powered by low-power technology has creatednew operational challenges in healthcare facilities, autonomous vehicles, and industrial IoTdevices, leading to increased security vulnerabilities. Security mechanisms require strongcomputational power, which these efficiency and real-time-oriented systems do not possessby design. Predictable threats against Edge AI systems are now more prevalent due to theirlimited computing power. Security frameworks prove inefficient when used in Edge AIenvironments, creating a vital protection weakness. This research paper focuses on a newlightweight cybersecurity system designed for Edge AI systems that lack sufficient resources.The proposed security construct enables the deployment of effective cryptographic systemsand AI-based anomaly identification along with defence mechanisms suitable for real-timeedge devices operating with restricted power consumption. A realistic dataset, known as theCyber Threat Detection Dataset, was utilised to test the framework, which incorporatedmultiple normal and attack behaviours, as well as several different attack varieties. The highspeedintrusion detection system relies on Random Forest (RF), while a LightweightConvolutional Neural Network (CNN) handles anomaly detection activities within theframework, and Federated Learning (FL) decentralises learning across different edge nodeswith privacy protection. According to test results, the developed framework deliversadvanced threat detection accuracy, better energy efficiency, and faster inference speeds.Each security model partakes specific capabilities that enhance a layered defence, resulting ina protected, adaptable and scalable protection for Edge AI systems.