A new traffic rate estimation and monitoring algorithm for the QoS-enabled Internet

Abstract

The paper presents a new traffic rate estimation method for the QoS-enabled Internet, employing a flip-flop filter that measures actual changes in the characteristics of a traffic flow in a timely and accurate manner (i.e., it is agile), while ignoring short term variations or spikes in the traffic flow (i.e., it is stable). The proposed algorithm is compared with the familiar exponentially weighted moving average (EWMA) and time sliding window (TSW) estimators under the self-similar traffic model. Simulation results illustrate the effectiveness of the proposed scheme.