Rate control of elastic connections competing with long-range dependent network traffic

Long-range dependence is regarded as a fundamental property of network traf fic, Using an original approach, cre incorporate this property in a traffic control mechanism for elastic connections that can adapt to the instantane ous network load in a differentiated services-type framework. In this scena rio, the network makes predictions of bandwidth requirements of the high-pr iority traffic and returns feedback information to the elastic source. We i nclude a prediction compensation algorithm that compensates for the larger prediction errors for connections with longer roundtrip delay, and analyze the performance of this algorithm. The specific topology involved in traffi c control for differentiated services is thus harnessed, together with the long-range: dependence, to improve network performance, thereby counteracti ng the undesirable characteristics of self-similarity. Furthermore, an adap tive version of the rate-based control algorithm is studied, based on the u se of real-time estimates of traffic parameters, including the mean, varian ce, and Hurst parameter.