Service composition is a technique to recompose various heterogeneous network services for a new business. A modern streaming media system may run multiple businesses. Their resources can easily be improperly distributed and under utilized. This paper studies service composition in the context of streaming media applications. A three-layer model is adopted to describe the system. The optimal service composition is formulated as a constraint optimization problem. By applying linear programming to the proposed cost function, a user demand behavior based adaptive algorithm (UDBA) for service composition is presented. The simulation results show that our UDBA algorithm improves the performance of the streaming media system.
展开▼