An Intelligent Video Transmission Protocol in Grid

The success of Internet allows us to access remote data by using the TCP/IP protocol suite. However, bandwidth and delay have a crucial impact on the quality of a video stream delivered over IP connection. So causal ordered grid architecture is proposed to provide a novel transport protocol to the time-sensitive and ordering-critical video applications. But it indicates that the value of Delta is a system parameter and could not be changed in running time. In fact, the fixed Delta value could not reflect the current network status and different video frame types and sometimes result in poor video quality. We propose an adaptive intelligent Delta-causal ordering protocol based on fuzzy neural network to real-time dynamically choose Delta value according to the current network conditions and delivering video frame types. Experimental results show our algorithms achieve good tradeoff between reliability and real-time performance than fixed Delta value algorithm