Self-adaptive multi-population genetic algorithms for dynamic resource allocation in shared hosting platforms

Abstract

This paper presents a self-adaptive multi-population approach based on genetic algorithm (GA) for solving dynamic resource allocation in shared hosting platforms. The proposed method, self-adaptive multi-population genetic algorithm (SAMPGA), is a multi-population GA strategy aimed at locating and tracking optima. This approach is based on preventing populations from searching in the same areas. Two adaptations to the basic approach are then proposed to further improve its performance. The first adapted algorithm, memory-based SAMPGA, is based on using explicit memory to store promising solutions and retrieve them upon detecting change in the environment. The second adapted algorithm, immigrants-based SAMPGA, is aimed at improving the technique used by SAMPGA to maintain a sustainable level of diversity needed for quick adaptation to the environmental changes. An extensive set of experiments is conducted on a variety of dynamic resource allocation scenarios, to evaluate the performance of the proposed approach. Results are also compared with those of self-organizing random immigrants GA using three well-known performance metrics. The experimental results indicate the effectiveness of the proposed approach.