Efficient data uncertainty management for health industrial internet of things using machine learning

Abstract

In modern technologies, the industrial internet of things (IIoT) has gained rapid growth in the fields of medical, transportation, and engineering. It consists of a self-governing configuration and cooperated with sensors to collect, process, and analyze the processes of a real-time system. In the medical system, healthcare IIoT (HIIoT) provides analytics of a huge amount of data and offers low-cost storage systems with the collaboration of cloud systems for the monitoring of patient information. However, it faces certain connectivity, nodes failure, and rapid data delivery challenges in the development of e-health systems. Therefore, to address such concerns, this paper presents an efficient data uncertainty management model for HIIoT using machine learning (EDM-ML) with declining nodes prone and data irregularity. Its aim is to increase the efficacy for the collection and processing of real-time data along with smart functionality against anonymous nodes. It developed an algorithm for improving the health services against disruption of network status and overheads. Also, the multi-objective function decreases the uncertainty in the management of medical data. Furthermore, it expects the routing decisions using a machine learning-based algorithm and increases the uniformity in health operations by balancing the network resources and trust distribution. Finally, it deals with a security algorithm and established control methods to protect the distributed data in the exposed health industry. Extensive simulations are performed, and their results reveal the significant performance of the proposed model in the context of uncertainty and intelligence than benchmark algorithms.