Energy Efficient and Resilient Infrastructure for Fog Computing Health Monitoring Applications

In this paper, we propose an energy efficient and resilient fog computing infrastructure for health monitoring applications. We design the infrastructure to be resilient against server failures under two scenarios, without geographical constraint and with geographical constraint. We consider a heart monitoring application where patients send their 30-seconds recording of Electrocardiogram (ECG) signal for processing, analysis and decision making at both primary and backup servers. A Mixed Integer Linear Programming (MILP) is used to optimize the number and locations of the primary and backup processing servers so that the energy consumption of both the processing and networking equipment are minimized. The results show that considering geographical constraint yields a network energy consumption increase by up to 9.36% compared to without geographical constraint. The results also show that, the increasing number of processing servers that can be served at each candidate node can reduce the energy consumption of networking equipment besides reducing the energy increment rate of networking equipment due to increasing level of demand.