Dynamic Allocations of Dew Nodes for IoST-Dew-Fog-Cloud System Using a Single-Server Queueing Model with Dropping Function
Ayashkant Mishra, Veena Goswami, Meenakshi Kandpal, Rojalina Priyadarshini, G. B. Mund, Rabindra Kumar Barik · 2024
The Internet of Spatial Things (IoST) has become an indispensable part of daily life and an essential topic for research regarding location-based services. Especially in geospatial data gathering, IoST devices constantly collect and communicate a vast amount of data to be processed to gather necessary information. Leveraging the power of Dew Computing, these data can be stored and processed on-site before being uploaded to the cloud server; it gives the entire system of devices independence and eliminates the chance of loss of data because of loss of connectivity. We propose a model with a dropping function to accept and reject the tasks that arrive at the dew node to increase efficiency and productivity. We provide a mathematical and analytical model for delivering insights and evaluating the performance of IoST-Dew-Fog-Cloud systems. We evaluate and compare different dropping functions to determine the best option for optimizing the performance of the dew nodes. We also find different performance indices like the average length of an idle period, loss ratio, and mean output rate and response time.