ABSTRACT
The evolution of the Internet of Things (IoT) has witnessed many advancements in the healthcare sector and envisioned a large scale of healthcare applications. IoT collects an enormous amount of data from end users and transmits them to the cloud for its analysis. Though cloud has scalable infrastructure facilities to support compute, storage and network requirements, it fails to satisfy the low latency requirements of emergency healthcare applications. Hence Fog computing is introduced as an emerging architecture as it has capabilities to bring cloud-based services at the edge, closer to or where data is generated. Pre-processing and analysis of data at the network edge before sending to the cloud reduces latency and network bandwidth. Such local processing and analysis in fog environment are generally carried out in the physical devices present in the layer called fog nodes, which are generally resource constrained and heterogeneous in nature. To further leverage the fog nodes and to reduce its computational intensity, this paper proposes an infrastructure with multiple fog nodes grouping together to become a fog cluster. With the fog cluster, tasks can be distributed across nodes in the cluster and processed in a quick manner. The proposed Infrastructure is evaluated with a healthcare application scenario, Chronic Obstructive Pulmonary Disease (COPD) Patient Monitoring System by simulating the whole environment. The results show that the involvement of fog clusters has reduced the latency and energy consumption to a greater extent. Result also shows the true potential of fog computing in IoT applications, especially for applications with low latency requirements.
