The paper presents a novel approach to dynamic transmissionbandwidth allocation for transport of real-time variable-bit-rate videoin ATM networks. The authors describe video traffic in the frequencydomain: the low frequency signal captures the slow time-variation ofconsecutive scene changes; the high frequency signal exhibits thefeature of strong frame autocorrelation. The study indicates that thevideo transmission bandwidth in a finite-buffer system is essentiallycharacterized by the low, frequency signal. Since the time scale ofscene changes is usually in the range of a second or longer, the lowfrequency video signal is defined in a well-founded low frequency band.Hence, it is feasible to implement dynamic allocation of videotransmission bandwidth using on-line observation and prediction of scenechanges. Two prediction schemes are examined: the recursive least squaremethod vs. the time delay neural network method. A time delay neuralnetwork with low-complexity high-order architecture, called a“pi-sigma network”, is successfully used to predict scenechanges. The proposed dynamic bandwidth allocation scheme is shown to bepromising and practically feasible in obtaining efficient transmissionof real-time video traffic with guaranteed quality of services
展开▼