Modeling and Generation of Self-similar Behavior of Mpeg-4 Video Traffic Using Chaotic Maps Method

The rise in popularity of the Internet, and specially the World Web Wide, has resulted in the web browser becoming the most commonly used Internet application. This popularity has driven the increase in multimedia applications with streaming audio and video players becoming almost ubiquitous. These audio and video traffics represent a new class of Internet traffic, which will increase with introduction of various new technologies such as MP3 and MPEG. In order to design networks that can meet the demand of these applications, a solid understanding of video traffic behavior is required. Effective design and performance analysis of such networks thus depends on accurate modeling of video traffic. Modeling of video traffic has always been important research area for design, simulation and performance analysis of packet video networks and Internet streaming. In this paper, we investigate the scaling behavior that is present in MPEG-4 video traffic and to accurately capture the long-range dependent (LRD) properties of the MPEG-4 video traffic, we present a frame-level framework for modeling video traffic, in which the I, B and P frames are modeled using the chaotic maps method. The key findings of our study are that the video encoder output traffic has fractal behavior and this behavior exists regardless of the compression ratio and the simulation results demonstrate that our model not only preserves the temporal burstiness of the empirical traffic traces, but also captures the important statistical features of the original video traces such as the autocorrelation function and the frame-size distribution. Keywords— Video traffic, Chaotic Maps, Self-Similarity, MPEG4