Abstract
This paper introduces an edge-computing based Complex Event Processing (CEP) architecture for Remote Patient Monitoring (RPM) which is an important issue in the context of remote healthcare. In this architecture, the detection of complex events, that may indicate impending health problems, is performed on a mobile device that receives data from sensors attached to the body of a patient. The detected complex events are sent to a back-end hospital server running on a cloud for further processing. Current state-of-the-art RPM techniques use the mobile device as an IoT gateway agent to forward data streams from health sensors to a remote hospital server where complex events are detected. A drawback of this existing methodology is that the mobile phone always needs to remain connected to the hospital server. Also, the mobile network consumption is increased while transferring large volumes of sensor data streams thus leading to an increase in the user cost. Additionally, it leads to an increase in the workload at the hospital server that serves multiple patients. This research investigates a mobile device based CEP system for addressing these issues and demonstrates its viability through a proof-of-concept prototype. A thorough performance analysis is performed using a synthetic workload that provides insights into system scalability and the relationship between system/workload parameters and performance.