Lightweight Fuzzy-Driven Intrusion Detection for Consumer Life-Tech Applications

Abstract

Consumer life-tech applications have significantly benefited from the rapid advancement of cutting-edge technologies, enabling the delivery of intelligent, cost-effective, reliable, and sustainable solutions. As consumer life-tech applications are being extensively embedded in innovative technologies, a key challenge is balancing security and resource efficiency in resource-constrained consumer devices. In this paper, we propose a lightweight fuzzy-driven intrusion detection framework to address these constraints by combining four key techniques: knowledge distillation, fuzzy logic integration, structured pruning, and quantization. We employ knowledge distillation to transfer decision-making capabilities from a large teacher model to a smaller student model. A fuzzy logic layer is further introduced to improve interpretability and robustness to uncertainties, while structured pruning and quantization are used to greatly reduce the model’s computational and memory requirements. Our method achieves over 98% detection accuracy while greatly reducing model size and resource usage. This work offers a practical, interpretable, and high-performing intrusion detection solution for deployment in resource-constrained consumer life-tech ecosystems.