Immune network algorithm applied to the optimization of composite SaaS in cloud computing

In order to serve the different application needs of the different Cloud users efficiently and effectively, a possible solution is the decomposition of the software or so-called composite SaaS (Software as a Service). A composite SaaS constitutes a group of loosely-coupled applications that communicate with each other to form higher-level functionality. The benefits to the SaaS providers are reduced delivery cost and flexible SaaS functions, and the benefit for the users is the decreased cost of subscription. For this to be achieved effectively, the optimization of the process is required in order to manage the SaaS resources in the data center efficiently. In this paper, the optimization task of composite SaaS is investigated using an Immune network optimization approach. The approach makes use of activation and suppression that are mimicked by the natural immune system triggering an immune response not only when antibodies interact with antigens but also when they interact with other antibodies. Experiments are conducted with a series of SaaS configurations and the proposed immune network algorithm is compared with a formerly proposed grouping genetic algorithm. The results show that the immune network algorithm outperforms the grouping genetic algorithm.