Time Scaling Analysis of Jitter Propagation Property

QoS provisioning for user application requires not only the evaluation of performance on end-station but also the evaluation of network parameters on intermediate routers. These network parameters were observed under the condition of self-similarity, and evaluated focusing on first-order performance measures. Second-order performance measures such as jitter are of significance for multimedia data transport and QoS provisioning. In this paper, we show the scaling analysis of jitter propagation properties of Internet paths based on the wavelet analysis using the IP performance active measurement method. After collecting and analyzing the jitter with the global and local scaling analysis, following results were extracted. Firstly, it is deduced from the global scaling analysis that the most typical property for average energy is path dependency, i.e. propagation property through the path. Secondly, local scaling analysis described that short term fluctuations biased the values of partition function and caused multifractal scaling over the middle timescales, and long distance trip to foreign nodes generated comparatively irregular behavior and caused multifractal scaling