An efficient model to estimate and optimise the cloud migration costs from on-premises web apps

Abstract

Organisations can now easily scale their operations up or down based on demand, reducing the need for significant upfront investments in infrastructure. This flexibility enables businesses to respond quickly to changing market conditions and remain competitive in today’s fast-paced environment. Despite the undeniable benefits offered by Cloud computing, migrating existing on-premises web applications poses intricate challenges, particularly in terms of cost considerations. This paper presents a novel hybrid algorithm using the genetic algorithm and simulated annealing approaches. The primary objective of this algorithm is to minimise the costs associated with migrating on-premise web applications to cloud environments. We have tested the proposed algorithm’s performance on seventeen different standard test benchmark functions, including seven unimodal and ten multimodal functions. The overall experiments have been conducted in the MATLAB R2023a environment. We compared the overall results with three well-known algorithms: the generalised ant colony optimiser, genetic algorithm, and artificial bee colony. Results have been computed in terms of average and standard deviation metrics. The proposed hybrid genetic algorithm and simulated annealing algorithms perform better in an unknown search space and optimise the migration costs and energy consumption. The algorithm demonstrates superior performance through rigorous testing on several benchmark functions, offering actionable insights and practical solutions to cost estimation and optimisation challenges.