A modified genetic algorithm for multi-criteria optimization based on Eucalyptus

Eucalyptus is an open source cloud platform, scholars can study many problems of cloud computing from its extendibility and open source. In terms of the research about resource scheduling, Eucalyptus has its own three algorithms, Greedy algorithm, Round Robin algorithm and Power Save algorithm. They each only consider part of resource scheduling. The existing algorithms can not meet a comprehensive need, such as load balancing, migration cost, QOS and so on. Therefore, a new algorithm which can get global optimization solution is proposed. Genetic algorithm has the features of global optimal and is easy to be implemented. To satisfy multiple goals about resource scheduling on Eucalyptus, a modified genetic algorithm for multi-criteria optimization based on Eucalyptus is presented. Experimental results show the merits and feasibility of the proposed algorithm.