Modeling end-to-end packet delay dynamics of the internet using system identification
Teletraffic Engineering in the Internet EraProceedings of the International Teletraffic Congress - ITC-I7, volume 4 of Teletraffic Science and Engineering, Elsevier, (2001)DOI: http://dx.doi.org/10.1016/S1388-3437(01)80189-X
Abstract
Understanding the end-to-end packet delay dynamics of the Internet is of crucial importance since (1) it directly affects the QoS (Quality of Services) of realtime applications, and (2) it enables us to design an efficient congestion control mechanism for both realtime and non-realtime applications In this paper, we propose a novel approach for modeling the end-to-end packet delay dynamics of the Internet. The key idea of our approach is treating the network, seen by specific source and destination hosts, as a black-box, and modeling the end-to-end packet delay dynamics using system identification, having been throughly used in the field of control engineering. The end-to-end packet delay dynamics is modeled as a \SISO\ (Single-Input and Single-Output) system. The input to the system is a packet inter-departure time from the source host, and the output from the system is an end-to-end packet delay variation measured by the destination host. In this paper, the \ARX\ (Auto-Regressive eXogenous) model is used and its coefficients are determined using system identification. Several topics such as model validation and selection of the orders of the \ARX\ model are also discussed. We show that the \ARX\ model accurately captures the end-to-end packet delay dynamics if the orders of the \ARX\ model are appropriately chosen. We also show that the effect of other \UDP\ and \TCP\ traffic can be modeled by white noise by using the end-to-end packet delay variation.