Towards a Genetic Algorithm Based Approach for Task Migrations

Pervasive cloud computing heavily depends on task migrations in order to mitigate resource scarceness in some cloud nodes, especially the light weight nodes. In order to make decisions on task migrations, a number of possibly conflicting objectives should be considered, such as less energy consumption, quick response, in order to find an optimal migration path and optimal configurations. In this paper, we conduct initial exploration on using a Genetic algorithm (GA) based approach which is effective in solving multi-objective optimization problems. The preliminary evaluations we have done shows that the proposed approach is promising.