Multi-strategy improved sand cat optimization algorithm-based workflow scheduling mechanism for heterogeneous edge computing environment

• Published in: Sustainable computing informatics and systems Vol. 43; p. 101014
• Main Authors: Jayalakshmi, P., Ramesh, S.S. Subashka
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.09.2024
• Subjects: Cloud-edge computing; Energy consumptions; Execution latency; Multi-strategies improved sand cat optimization algorithm; Pareto optimal solutions; Workflow scheduling; Pareto optimal solutions; Multi-strategies improved sand cat optimization algorithm; Cloud-edge computing; Workflow scheduling; Execution latency; Energy consumptions
• ISSN: 2210-5379
• DOI: 10.1016/j.suscom.2024.101014

Edge computing is one of the predominant technologies which facilitates the option of bringing out the computing resources closer to the location of the end users when they are utilized by them. This facility offered by edge computing technology need to reduce the utilization of network bandwidth and response time with respect to the user’s workflow. In this paper, Multi-Strategy Improved Sand Cat Swarm Optimisation Algorithm (MSISCSOA)-based workflow scheduling mechanism is proposed for handling the challenges of workflow scheduling in cloud-edge computing environment. The core objective of this MSISCSOA-based workflow scheduling algorithm targets on minimizing the execution latency and energy consumption to facilitate timely and on-demand end users’ satisfaction of resources. This MSISCSOA scheme is adopted with the improvement introduced using random variation and elite collaborative strategies, such that well-balanced the trade-off between exploration and exploitation is achieved. This improvement is introduced over Sand Cat Optimization Algorithm (SCOA) using the merits of dynamic random search and joint opposite selection strategies that accelerates the convergence of the algorithm with increased global optimization and searching efficiency. It specifically improved SCOA using random variation for escaping from the local point of optimality. It also used well distributed pareto fronts and population evolution multi-strategy that aids in searching solutions with maximized diversity. The simulation experiments conducted using the datasets of Montage, Cybershake, LIGO and SIPHT an average confirmed minimized execution latency of 21.38 % and energy consumptions of 19.56 %, better than the baseline Ant Colony Optimization Algorithm-Based Workflow Scheduling (IACOAWS), Quadratic Penalty Function-based Particle Swarm Optimization Algorithm (QPF-PSOA), Biogeography Optimization (BBO) Algorithm based Multi-Objective Task Scheduling (BBOAMOTS) and Different Evolution-based Task Clustering and Scheduling (DETCS) approaches used for comparative investigation. •In this paper, Multi-Strategy Improved Sand Cat Swarm Optimisation Algorithm (MSISCSOA)-based workflow scheduling mechanism is proposed for handling the challenges of workflow scheduling in cloud-edge computing environment.•The core objective of this MSISCSOA-based workflow scheduling algorithm targets on minimizing the execution latency and energy consumption to facilitate timely and on-demand end users’ satisfaction of resources.•This MSISCSOA scheme is adopted with the improvement introduced using random variation and elite collaborative strategies, such that well-balanced the trade-off between exploration and exploitation is achieved.•This improvement is introduced over Sand Cat Optimization Algorithm (SCOA) using the merits of dynamic random search and joint opposite selection strategies that accelerates the convergence of the algorithm with increased global optimization and searching efficiency.•The simulation experiments conducted using the datasets of Montage, Cybershake, LIGO and SIPHT an average confirmed minimized execution latency of 21.38 % and energy consumptions of 19.56 % better than the baseline approaches used for comparative investigation.