Smart Hospital Patient Monitoring System Aided by Edge-Fog-Cloud Continuum: A Performability Evaluation Focusing on Distinct Sensor Sources

Abstract The variety of sensors for health monitoring helps professionals in the field make decisions. For certain clinical conditions, it is interesting to monitor the patient's health after discharge. Wearable health monitoring devices (smartwatches) can send data to the hospital. However, to monitor many patients (external and internal), a resilient and high-performance computing infrastructure is required. Such characteristics require high monetary cost equipment. This paper presents an SPN (Stochastic Petri Net) model for evaluating the performance and performability of a multi-tier hospital system architecture (edge-fog-cloud). The model allows you to evaluate the Mean Response Time (MRT), resource utilization level (U), and probability of data loss (DP). A very specific characteristic is to consider two data sources (internal and external). Performance metrics are evaluated in two scenarios based on varying the number of containers for simultaneous processing in the cloud and fog. The model evaluates performability of the MRT metric over the variation of parameters that have the greatest influence on availability (container MTTF) and performance (cloud processing capacity).