A spatial and temporal analysis of internet aggregate traffic at the flow level

Unmasking the Growing UDP Traffic in a Campus Network

“…Port number usage is one of the key measures to understand the type of traffic and often used for identifying applications such as web surfing, online gaming, and peer-topeer transfer with fixed port numbers [13,15,17,22,26]. Figure 4 plots the cumulative UDP traffic volume against the source and destination ports in k-2011.…”

“…The reported trend has not been studied thoroughly yet, and the cause and the impact of growing UDP traffic to the Internet are to be discovered. Although most traffic measurement studies have been about TCP, some previous research papers have looked at the characteristics of UDP traffic in terms of size, arrival, port usage of flows [17,20,22,25]. However, the traffic traces used in those papers do not reflect the most recent trend as they are all collected before 2009 when only a small portion of UDP traffic around 5% or even less is reported.…”

A study on the network traffic of Connexion by Boeing: Modeling with artificial neural networks

“…These equations express network traffic as a function of fractional Browning motion based on theories developed by Norros (1995). The accuracy of these models is dependent upon measured data characteristics, which are often not very accurate themselves (Cleveland and Sun, 2000;Li et al, 2004;Qian et al, 2004), are difficult to collect and correctly characterize (Bregni, 2004;Cleveland and Sun, 2000;Floyd and Kohler, 2003;Fomenkov et al, 2004;Li et al, 2004;Park et al, 2005;Qian et al, 2004;Yousefi'Zadeh, 2002), and are subject to constant change and vary with time, circumstance, and location (Bianchi et al, 2004;Brownlee and Claffy, 2002;Li et al, 2004;Qian et al, 2004;Rodriques and Guardieiro, 2004).…”

A Traffic Characterization Procedure for Multimedia Applications in Converged Networks

“…The k parameter determines the probability that A(t) will overpass Â(t), in the time slot t. If A(t) is a fBm process, the expression for k may be approximated by a residual Gauss distribution and may be expressed as in (6). An extension of the definition in (5) is made in [14], where the envelope A H (t), for H>0.5 is defined as in (7).…”

“…In the recent years, the self-similar nature of Internet traffic has been demonstrated with high quality measures [5,6]. The impact of this kind of traffic in QoS measures has been already analyzed.…”