Ant Colony Intelligence Based Solution for Grid Services Selection

摘要

In Grid environment, application systems are generally implemented by composition and cooperation of several Grid services. In order to get an optimal service composition plan, appropriate services should be selected to satisfy end-to-end QoS requirements. QoS-aware service selection is a complex combinatorial optimization problem. According to the characteristic of Grid services, this paper presents a QoS-aware service selection algorithm based on ant colony intelligence. Firstly, we present a user satisfactory degree model for Grid services. A proposed services composition graph is applied to model the composition problem. Then an extended Ant Colony System using a novel ant clone rule is applied to solve the selection problem. In order to quicken the speed of its convergence, the user synthesis satisfaction function is considered as the heuristic information. Finally, the algorithm is tested for the performance under the stable state and the unstable state.