Online fault detection and recovery in body sensor networks

This paper presents an online solution for fault detection and isolation of erroneous node in body sensor networks to provide sustained delivery of services despite encountered perturbations. This work adopts a multi-tier telemedicine system where a star topology with the sensor nodes sending their data to Personal Digital Assistant (PDA) for data fusion is used. The system performs real-time analysis of sensor data and can generate warnings regarding occurrence of fault which allows fault isolation and system reconfiguration. Time redundancy is used to address detection issues related to persistence of error (intermittent and permanent).