Composite Agent Service Selection Algorithm for Non-functional Attributes Based on Simulated Annealing

With the development of autonomic computing, composite agent service selection problem for non-functional attributes has been taken as one of the most important issue in field of autonomic computing based on agent. Such a constrained composition selection problem is known to be NP-hard. In this paper, we propose a composite agent service selection algorithm for non-functional attributes based on simulated annealing to find composite agent service having a maximal QoS performance under a certain QoS attribute constraint. The simulation results demonstrate that our algorithm performs excellent performance of overall QoS, cost and running time in compared algorithms. Furthermore, it has the property of rapid convergence.