Adaptive techniques for time-critical data processing in IEC

Abstract

Time-critical data processing presents an essential issue in the IEC since real-time decision-making and responsiveness are becoming more and more important in many different IEC applications, specifically those involving vital infrastructure, transportation systems, industrial automation, and healthcare monitoring. These applications require low latency, high bandwidth, the sustainability of devices, and ensuring data integrity to work effectively and reliably. Edge computing has become increasingly popular as an addition to cloud computing, particularly for applications such as industrial control systems that demand guarantees for timely communication. Although edge computing allows for the rapid analysis of data streamed from the Internet of Things (IoT) devices, these devices often do not have the computational power and bandwidth necessary to ensure satisfactory performance for applications that are sensitive to timing. Therefore, developing a dynamic, distributed approach to manage time-critical IoT data streams efficiently across the IEC continuum is necessary. Numerous scenarios, such as flood control and crisis management, employ thousands of energy-aware sensors. These sensors continue monitoring their environment all the time, gathering vital information that helps them make important decisions. However, since they run on batteries, energy efficiency is critical to their long-term viability. Finding adaptive, time-sensitive, cost-effective solutions that maximise their power usage is essential to extending their lifespan. IoT has quickly become a transformative model for linking devices to gather data, share information, and process data efficiently. Faults within IoT systems can arise in multiple scenarios and manifest differently. Understanding and handling these faults is crucial for improving the integrity and reliability of real-time data for making informed decisions and maintaining the effectiveness of IoT applications. Monitoring real-time data streams on a constant basis for abnormalities, faults, or inconsistencies that can be caused by environmental conditions, communication problems, or malfunctioning sensors is necessary to manage data quality and failure. To handle bandwidth optimisation, energy enhancement, data quality monitoring, and healing, this thesis presents multilateral research towards adaptive techniques for the optimisation of time-critical data processing in IEC.