A Biased Random-Key Genetic Algorithm for the Cloud Resource Management Problem

Abstract

Flexible use options and associated cost savings of cloud computing are increasingly attracting the interest from both researchers and practitioners. Since cloud providers offer various cloud services in different forms, there is a large potential of optimizing the selection of those services from the consumer perspective. In this paper, we address the Cloud Resource Management Problem that is a recent optimization problem aimed at reducing the payment cost and the execution time of consumer applications. In the related literature, there is one approach that successfully addresses this problem based on a Greedy Randomized Adaptive Search Procedure. Due to the fact that consumers require fast and high-quality solutions to economically automate cloud resource management and deployment processes, we propose an efficient Biased Random-Key Genetic Algorithm. The computational experiments over a benchmark suite generated based on real cloud market offerings indicate that the performance of our approach outperforms the approaches proposed in the literature.

A multiswarm for composite SaaS placement optimization based on PSO

“…We have the ACs and the DCs that have to be placed on Cloud servers (computing servers and storage servers) to minimize (or to maximize) an objective function along with different operational (resources) and/or SLA constraints. The approach adopted in the work of Lalla‐Ruiz and Voβ to deal with multiple knapsack assignment problem, namely, a biased random‐key GA, has been adopted to the Cloud resource management problem in the work of Heilig et al…”