Real-Time Signal Processing in IoT-Based Embedded Systems Using Hybrid AI-Enhanced Edge Computing

The rapid expansion of the Internet of Things (IoT) has necessitated real-time signal processing capabilities in embedded systems, but traditional cloud-based models often fall short in meeting the low-latency demands of these applications. To address this, we propose a hybrid AI-enhanced edge computing framework that integrates machine learning with conventional signal processing techniques, optimizing real-time processing at the edge. Our framework was tested in a smart factory environment, where edge devices equipped with lightweight AI models were employed for signal classification and processing tasks. The results demonstrate a 40% reduction in latency, with processing times as low as 50 milliseconds per signal, and an accuracy improvement of 15% over traditional methods, achieving a 92% accuracy rate in signal classification. These findings underscore the potential of hybrid AI at the edge to significantly enhance both the speed and accuracy of real-time signal processing in IoT-based embedded systems. This approach not only meets the critical latency requirements of modern IoT applications but also improves overall system reliability, making it a promising solution for applications in smart manufacturing, healthcare, and autonomous systems. Future work should explore further optimization of AI models for edge deployment and scalability across broader IoT ecosystems.