D-PuP: An enhanced prediction algorithm for real-time MPEG-4 VBR video traffic with dynamic adaptation

In previous work [1], we proposed a real-time MPEG-4 VBR video traffic prediction algorithm that calculates the probability density function (PDF) based on the last N frames, rWin, by using Cubic-spline interpolation method, and then utilize it for the prediction of next frame. We fixed the size of rWin as five in the algorithm. Although the algorithm is capable of providing more accurate prediction than those in the research literature, it still has weakness on the adapting the traffic dynamics due to the fixed size of rWin. In this paper, we first study the experimental investigation about the number of rWin, and propose an enhanced traffic prediction algorithm with dynamic rWin.