Performance and Availability Trade-Offs in Fog-Cloud IoT Environments

摘要

Internet of Things (IoT) is an emerging paradigm that transforms everyday devices (Things) into Internet-connected devices with sensing, processing, and actuation capabilities. These devices have limited storage and processing capacity, so they have been integrated with Cloud computing to overcome these limitations. Cloud computing offers various benefits such as offload data storage and processing burden at the Cloud side. Nevertheless, because Cloud is not an efficient solution for IoT latency-sensitive applications, Fog computing was introduced to address this limitation. Although Fog-Cloud IoT environments have begun to be adopted in the last few years, such environments have not been properly assessed in terms of their capacity to meet the growing demand of IoT devices. In this work, we present a Deterministic and Stochastic Petri Net (DSPN) approach for evaluating Fog-Cloud IoT environments composed of hundreds physical Things. Our approach allows evaluating the trade-offs of many performability metrics (e.g., utilization, response time, throughput, and availability) and, consequently, may help system designers to choose the most suitable Fog-Cloud IoT environment. We demonstrate the feasibility of our approach through a real-world case study. The results revealed that adopting a Fog device can improve availability. However, the performance is only improved in certain conditions like when the environment is not at full capacity.