REAL-TIME MALWARE DETECTION FOR IOT SYSTEMS USING DEEP LEARNING TECHNIQUES

Abstract

Security vulnerabilities, especially malware attacks on resource-constrained systems, have arisen due to the development of Internet of Things (IoT) devices. The dynamic characteristics of contemporary threats render conventional malware detection methods sometimes inadequate for ensuring full protection. This research proposes a solution for virus detection in IoT devices using deep learning techniques.The suggested method employs advanced neural network models to independently identify intricate patterns and features from data on network traffic and device activities. The model attains high precision in identifying both established and novel malware variants by the use of architectures like Recurrent Neural Networks (RNN) and Convolutional Neural Networks (CNN). The framework is designed to be versatile and adaptive, geared for efficient operation within the constrained processing capabilities of IoT devices.The proposed approach exhibits markedly improved detection accuracy, precision, and a reduced false positive rate compared to traditional machine learning methods, as shown by the experimental results. Moreover, the technology can identify attacks in real-time, hence enhancing the overall security of IoT ecosystems.