An extended multi-population genetic algorithm for resource allocation in service hosting platforms

Abstract

Nowadays, there is an increasingly rise in the number of needs for processing and the need for larger and more powerful computing systems has been felt and the data centers are rapidly moving towards the required technologies. Since the expansion of the dedicated hosting platforms is not justifiable for economic reasons, there is a strong incentive to use shared resources among services which is called “Shared hosting platforms”. One of the main challenges in the hosting platforms is the resource allocation. The resources should be allocated according to the users' requests with the least possible time and expenses in order to be beneficial for providers. In this paper, we define the resource allocation problem for the shared hosting platforms as an optimization problem and solve it. We have used a special multi-population genetic algorithm to solve the resource allocation problem in shared hosting platforms namely self-adaptive multi-population genetic algorithm (SAMPGA). Extensive simulation experiments were conducted for a comparative performance evaluation with respect to standard genetic algorithm (SGA) using two performance measures. The results demonstrated that the SAMPGA has a much better performance than the SGA.