Optimizing Energy Efficiency in Cloud Data center using an Enhanced Dualist Algorithm with Improved Exploration

Investigation for minimizing energy consumption in data centers is increasing due to their heavy usage. In a data center, virtual machine placement is a procedure that maps virtual machines to physical machines. VMP problem is a complex combinatorial optimization problem with various constraints. In literature, the VMP problem is investigated with different objectives. In this paper, the problem is formulated as a single-objective optimization problem with the goal of minimizing energy consumption in cloud data centers. A metaheuristic evolutionary algorithm called the Duelist algorithm is designed to solve the VMP problem. Two variations are proposed with modifications in the winner's learning strategy. The proposed strategy improved the exploration capability of the Duelist algorithm. The performance of proposed variations is tested using 15 datasets with varying problem sizes. Performance is evaluated using the best, mean, standard deviation, success rate, acceleration rate and convergence speed. Variation 1 and variation 2 are better than the basic Duelist algorithm on all measures.