A QoS-Based Service Discovery Using Meta-Heuristic Optimization in Federated Cloud

In a cloud based service environment, finding the user-centric optimal service composition has become increasingly necessity. A consumer needs the service composition which meets the quality requirements in a cost effective manner. The service level agreement (SLA) signed by both, the consumer and the service provider, includes Quality of Service (QoS) requirements and penalties in case of violation. The violation of a quality requirement for a service composition needs a suitable service substitution which incurs a considerable adaptation cost. Therefore, an efficient tool is needed to address the user-specific service requirement in a cost effective manner. The work proposed a heuristic based approaches using hybrid Genetic and Tabu algorithm to determine the optimal QoS based service composition with maximum utility value and minimum adaptation cost. The blending of a basic Genetic algorithm and Tabu search has been formulated as Memetic approach for service utility prediction and ranking the service compositions. The work also simulates the basic GA and Memetic approach to compare the performance of the two methods in terms of utility, violation cost and adaptation cost of a service composition.