Computational intelligence-enabled prediction and communication mechanism for IoT-based autonomous systems. (January 2023)

• Record Type: Journal Article
• Authors: Jin, Bo; Khan, Fazlullah; Alturki, Ryan; Ikram, Mohammed Abdulaziz
• Is Part Of: ISA transactions. Volume 132(2023)
• Page Start: 146
• Page End: 154
• Publication Date: 2023-01
• Subjects: Decision support system -- Internet of Things -- Intelligent sensors -- Computational intelligence -- Intelligent prediction -- IoT-based autonomous systems; Engineering instruments -- Periodicals; Engineering instruments; Periodicals; Electronic journals; 629.805
• DOI: 10.1016/j.isatra.2022.06.007
• Languages: English
• ISSNs: 0019-0578

Abstract

Abstract: Autonomous systems and the Internet of Things (IoT) have become more sophisticated research areas and entered successfully into various daily living activities such as smart homes & buildings, autonomous cars, drones, robots, etc. A crucial and essential aspect of these systems is the precision and accuracy of the decision-making process, i.e., the decision support system. Likewise, developing a completely autonomous system is an open research problem. This paper proposes a computational intelligence-based prediction and communication mechanism for the independent system where IoT is used as a data collection tool. Initially, energy gauge (EG) devices collect helpful information about neighboring devices in the IoT networks. Then, information about the potential relaying devices is broadcasted by the concerned EG device, which uses every member device to adjust routing path(s) in the autonomous system. Furthermore, every EG device has an embedded computational intelligent decision support system that is used to precisely predict the criticality of a neighboring device (preferably relay) in the autonomous systems. Therefore, every device must ensure data transmission via the most reliable path(s), i.e., avoiding critical devices if possible. A device is assumed critical if either its residual energy or received signal strength indicator value is less than the defined threshold values for the autonomous systems. Additionally, the proposed mechanism has ensured a uniform traffic distribution of the transmitted packets in the autonomous system. The operational applicability of the proposed computational intelligence-enabled prediction mechanism in the autonomous system is verified by comparing it with the existing approaches. Simulation results show that the proposed scheme has enhanced the accuracy of the concerned autonomous systems more than other schemes.