COBGA: MCDM-assisted improved genetic algorithm for scheduling Industrial-Internet-of-Things jobs in cloud computing

Abstract

Industrial-Internet-of-Things (IIoT) revolutionizes industrial operations by integrating data analytics, cloud services, and smart devices to increase productivity and efficiency. Scheduling numerous IIoT tasks in cloud computing is a complex and indispensable problem due to the dynamic nature of resources and the need for efficient matching between IIoT tasks and cloud resources. This paper proposes a Chaotic and Opposition-Based learning Genetic Algorithm (COBGA) to address complex IIoT task scheduling to optimize four objectives. The proposed COBGA extends the Genetic Algorithm (GA) using a piecewise linear chaotic map (PWLCM) and Opposition-Based learning (OBL) to boost the convergence speed and solution quality while avoiding getting trapped in local optima. The COBGA also employs an efficient fitness computation method with linear time complexity rather than quadratic for given tasks and virtual machines. The COBGA also employs the Multi-Criteria Decision-Making technique (MCDM), Multi-Objective Optimization by Ratio Analysis (MOORA), to find the best possible solution based on the trade-off of four contradictory scheduling objectives, i.e., makespan, monetary cost, energy efficiency, and security time requirements. The performance of the proposed COBGA algorithm was compared with improved GA, Particle Swarm Optimization (PSO), Grey Wolf Optimizer (GWO), Whale Optimization Algorithm (WOA), Salp Swarm Algorithm (SSA), Differential Evolution (DE), Greylag Goose Optimization (GGO), and Secretary Bird Optimization Algorithm (SBOA), through extensive experimentation and statistical analysis. The COBGA significantly offers improved results among peer algorithms as it achieves improvements of approximately 87.5%, 86.5%, 50.5%, and 37.5% for makespan, energy consumption, cost, and security time, respectively, for varying task parameters.