A fast non-linear adaptive algorithm for video traffic prediction

Abstract

To guarantee quality of service (QoS), the requirements for video transmission such as delay and cell loss rate (CLR) are very stringent. These constraints are difficult to meet if high network utilization is desired. Dynamic bandwidth allocation is thus needed. Video traffic prediction can play an important role in dynamic bandwidth allocation and traffic management in high speed networks. We propose a fast convergent algorithm to predict the variation of I frames, based on which the bandwidth is assigned. The new algorithm can achieve fast convergence and small prediction error.