Abstract: As the application of streaming media increases exponentially, it is more and more important to improve the performance of streaming server. The paper proposes a load schedule strategy of streaming media server cluster, which comprehensively applies the multi-stage fuzzy system and the technique of artificial neural network to carry on fuzzy prediction for the user request of response delay, as considering prefix caching and patch prefetching technology of the streaming media programs of the node server, proposes the load schedule algorithm based on the threshold value of response delay and the caching situation of requesting streaming program at the node server, and takes the simulated test to it.


Key words: streaming media, cluster server, response times estimation, multi-stage fuzzy system, neural networks, prefix caching, patch prefetching, load schedule