ODTRA-Based Task Offload Optimization for Industrial Internet of Things: Improving Efficiency and Performance With Digital Twins and Metaheuristic Optimization

Abstract

The Industrial Internet of Things (IIoT) has revolutionized several industries by improving the communication of sensor data among interconnected machines and systems. IIoT frameworks, on the other hand, can be challenging to set up and keep functioning several times. This paper introducing Task Offloading (TO) into the Internet of Things (IoT) brings about plenty of challenges, which will be addressed throughout this study. Tasks that use an enormous number of resources are allocated to remote server locations in the cloud in order to be performed. As a result of the objective of optimizing decisions involving Optimize Task Offloading (OTO), it should be considered that Digital Twins (DT) be developed. DTs are automated backups of physical objects or systems that can be used to perform data collection in real-time, management's decisions, and optimization. DTs are also identified as digital copies of things. Using DT, constantly tracked in real-time, and Metaheuristic Optimization (MO) computational approaches, this research recommends a Task Offloading Model (TOM) for the IIoT. In the model, the Task Execution Time (TET) is reduced to a minimum by limited factors such as the server's computing power, the constraints on bandwidth, and the Energy Consumption (EC) of the device. In order to efficiently increase OTO results, the Offloading with Digital Twins and Raindrop Algorithm (ODTRA) method that has been industrialized makes use of the Water Cycle Metaphor (WCM) and the Probabilistic Recursive Local Search (PRLS) technique. It is feasible to implement the algorithm in real-time IIoT environments through the presence of DT in order to progress decision-making and provide real-time monitoring. This work reviews the assumptions of a research study that analyzes the performance of OTO in IIoT environments, with a detailed emphasis on the potential for real-time deployment.