A traffic congestion detection mechanism for QoS using RTFM architecture and IDC

Congestion detection and control mechanism is an integrated function of any quality of service scheme to provide differentiated services during network congestion. This paper presents a new approach to congestion detection and avoidance based on real-time traffic flow measurement and the statistics of index of dispersion for counts. This approach passively monitors the underlying network traffic flow, and detects or predicts the network congestion without relying on transport-layer protocols or feedback signaling. An experimentation of applying this scheme in a validation testbed is presented with results illustrating the effectiveness of the approach in practical scenarios.