Clustering-based Time efficient Concurrent Data Collection Network structure for Shared IoT Applications

Typical enterprise IoT implementation involves setting up of the entire infrastructure right from sensor devices to the cloud and is designed to operate in silos. When multiple entities of IoT applications with similar data requirements exist, each such entity must set up its own sensors, devices, gateways, etc., resulting in over-provisioning. This is specifically true for environments where sharing infrastructure is practical and beneficial. With hundreds of devices in a typical IoT network and multiple processes relying on their data, concurrent data collection in shared IoT infrastructure is a critical challenge. While recent algorithms have aimed to address the issue, they only consider network scenarios where devices can reach the base station(s) directly. As this is not always the case, and with device power constraints, data collection is done independently for each application. This paper presents a Clustering-based Time efficient design for Concurrent Data Collection Network structure in shared IoT applications. To solve this problem, the network is first divided into multiple clusters and then data collection is performed concurrently for each cluster. Later, the data is concurrently transmitted from each cluster to multiple base stations. The proposed network structure minimizes the delay in concurrent data collection and lowers the transmission count, which is validated through extensive simulation results. Through simulation, it is shown that the proposed network structure is efficient in terms of overall time slots as compared to existing concurrent data collection network structures.