INTERNATIONAL JOURNAL FOR INNOVATIVE RESEARCH IN MULTIDISCIPLINARY FIELD

Abstract:    The recent and trending methodology in Wireless Body Area Networks (WBANs) of patient medical data handling is serious for ensuring suitable, protected, and consistent healthcare monitoring, particularly under active physiological environment. This work proposes a novel framework that combines Time-Stamped Deep Reinforcement Learning (TS-DRL) with a Recurrent Neural Network (RNN) to astutely manage, prioritize, and broadcast patient data within a WBAN application, while leveraging block chain technology for data veracity and trust supervision. Wearable and implantable sensors constantly produce heterogeneous physical signals, based on time-stamp are checked  by the TS-DRL agent to adaptively allocate network resources, reduce latency, and handle emergency data with higher priority. The RNN model takes sequential dependent data as input analyzing the patient health, enabling exact calculate of serious events and sustaining proactive managerial decisions. To address issue such as data privacy, block chain is employed as a decentralized ledger to document validated medical connections, ensuring translucent access control and unassailable storage without relying on a common authority. The proposed architecture enhances energy efficiency, reduces packet loss, and improves quality of service by integrating data scheduling and prediction with accurate data sharing. Tentative analysis demonstrate that the incorporated TS-DRL–RNN–blockchain framework appreciably outperforms conservative WBAN data handling approaches in terms of response time, making it well suited for smart health systems.