<![CDATA[Wireless Health Monitoring Networks (WHMNs) that utilize Biomedical Electronics (BMEs) are rapidly changing how medical information is gathered and used in contemporary Medical Informatics through providing ongoing, real-time, and remote assessment and monitoring of essential physiological parameters; including electrocardiogram (ECG), body temperature, blood oxygen saturation, and heart rate. A number of constraints can limit the practical implementation of WHMNs including: stringent Quality of Service (QoS) requirements; limited energy availability; network scalability limitations; and significant risks associated with the transfer of sensitive medical information via open wireless networks. While conventional centralized routing and learning based approaches provide the ability to route packets and learn network behavior they require high levels of communication, make the network vulnerable to cyber threats, and do little to protect the privacy of patients' data. This paper describes a QoS aware secure federated intelligence framework designed specifically for BME driven WHMNs. The proposed framework will include: (1) Decentralized Federated Learning to enable distributed intelligence among edge nodes; (2) Trust based QoS Aware Routing to enable reliable and energy efficient communication throughout the network; and (3) a Lightweight Blockchain-Assisted Mechanism to provide authentication, integrity, and tamper resistance to the network. In contrast to traditional federated learning where edge node models are aggregated into a central server model for processing, the proposed framework enables edge nodes to train their own local models on their own data and to share only encrypted model updates rather than raw biomedical data; thereby reducing communication costs and increasing patient data privacy protection. Simulation results show that the proposed framework outperform other state of the art WHMN protocols in terms of Packet Delivery Ratio (PDR); End-To-End Delay (E2ED); Residual Energy per Node (REPN); and Overall Network Lifetime (ONL). Additionally, the security analysis provided in this paper demonstrates the proposed framework's robustness against Insider Attacks, Data Manipulation Attacks, and Unauthorized Access Attacks. These findings indicate the strong feasibility of the proposed framework for next generation Intelligent Healthcare and Medical Informatics Applications.]]>
