An approach towards resource management and transportation of VBR video traffic

This paper investigates the predictability of VBR video traffic which exhibits long-range dependence (LRD) based on the fractional ARIMA (1,d,0) model. We consider a novel predictor based dynamic bandwidth allocation for the short term resource management and study its queueing performance. This method is simple to implement and is demonstrated to provide superior performance over some existing methods. The performance results are presented.