Real-time IoT data analytics for chronic patient monitoring based on the horizontal oneM۲M architecture

The rapid growth of the Internet of Medical Things (IoMT) has enabled continuous monitoring of patients through wearable sensors and connected healthcare devices. However, many existing IoT-based healthcare systems primarily focus on data collection and interoperability, while real-time analysis of physiological data remains limited. This limitation is particularly critical for chronic disease monitoring, where early detection of abnormal health conditions is essential. This paper proposes a real-time IoT data analytics framework for chronic patient monitoring based on the horizontal oneM۲M architecture. The proposed system integrates wearable medical sensors, an edge computing layer, and a standardized oneM۲M service layer to ensure interoperability and scalable device management. Real-time physiological data streams are processed using a streaming analytics module, while an artificial intelligence model based on Long Short-Term Memory (LSTM) networks is employed to detect anomalies in patient vital signs. To evaluate the proposed framework, experiments are conducted using publicly available physiological datasets. The system is assessed in terms of detection accuracy, processing latency, and scalability under increasing numbers of connected devices. The results demonstrate that the proposed approach can effectively detect abnormal physiological patterns while maintaining low latency in real-time monitoring scenarios. The proposed architecture highlights the potential of combining oneM۲M-based IoT platforms with real-time streaming analytics and artificial intelligence to support intelligent healthcare monitoring systems for chronic patients.