Ensemble Learning Based Multiple Mobile Sink Architecture For Energy Efficient Wireless Body Area Network Towards Disease Centric Patient Group Data Management

Abstract

In critical medical emergency situations, wireless sensor network employed for health care industry is termed as wireless body area network (WBAN). It is equipped as health monitoring systems which transmits the data packets containing critical information of patients' health.  Therefore, to manage the increased traffic load and to provide ubiquitous medical services for highly prioritized disease, a new model for a disease-centric health-care management system using wireless body area networks (WBAN) in the presence of multiple health-cloud service providers (H-CSP) as multiple mobile sink has to be constructed on utilizing the theory of Social Network Analysis (SNA). It is adopted to optimize the computational complexity and the traffic load of the network in an area, considering different disease types and the criticality indices of the WBANs.  Disease-centric Patient Group (DPG) formation among coexisting WBANs ensures optimized traffic load and reduced computational complexity. In this addition, we formulate a pricing model for the efficient mapping of critical WBANs from a DPG to an H-CSP to optimize the expected packet delivery delay and the network throughput for energy efficient data communications. Consequently, to identify the critical WBANs from a DPG, we design an ensemble learning to identify a decision parameter based on an assortment of selection parameters of multiple mobile sinks. The selection of parameters from the multiple mobile sink provides optimized solution for traffic load and energy efficiency of the resources. The performance of the Efficient Healthcare Management (HCM) scheme is analyzed based on distinct measures such as Packet delivery ratio, delay, and throughput. Simulation results exhibit significant improvement in the network performance over the existing schemes.