Federated learning for the detection of malware in IoT devices

The increasing expansion of IoT devices in smart homes has created new security issues, including malware detection. Traditional malware detection approaches often fail on IoT devices due to resource constraints and heterogeneity. Novel malware detection in smart home IoT devices is proposed using deep federated learning. Methods: We employ deep learning models while protecting data privacy by training them jointly across numerous devices. Our solution uses smart homes' dispersed nature to provide a shared malware detection model without compromising device privacy. The study quantifies encrypted communication, differential privacy, and local aggregation success rates across ten IoT devices, averaging 95%. The proposed solution is compared to encrypted communication, privacy, and local aggregations. Novelty: The proposed method may improve smart home security against changing malware threats. We demonstrate the architecture and methods of our deep federated learning-based smart home malware detection system. We test our technique on the dataset and show that it can detect new malware. Our revolutionary malware detection solution for smart home IoT devices improves security and privacy.