Weibull mixture model to characterise end-to-end Internet delay at coarse time-scales

Abstract

Abstract:Traces collected at monitored points around the Internet contain representative performance information about the paths their probes traverse. Basic measurement attributes, such as delay and loss, are easy to collect and provide a means to both build and validate empirical performance models. However, the task of analysis and extracting performance conclusions from measurements remains challenging.Ideally, performance modelling aims to find a set of self-contained parameters to describe, summarise, profile and easy display network performance status at a time. This can result in the provision of meaningful information to address applications in fault and performance management, hence providing input to network provisioning, traffic engineering and performance prediction. In this work we present the Weibull Mixture Model, a method to characterise endto-end network delay measurements within a few simple, accurate, representative and handleable parameters using a finite combination of Weibull distributions, with all the aforementioned benefits. The model parameters are related to meaningful delay characteristics, such as average peak and tail behaviour in a daily profile, and can be optimally found using an iterative algorithm known as Expectation Maximisation. Studies on such parameter evolution can reflect current workload status and all possible network events impacting packet dynamics, with further applications in network management.In addition, a self-sufficient procedure to implement the Weibull Mixture Model is presented, along with a set of matching examples to real GPS synchronised measurements taken across the Internet, donated by RIPE NCC.

A framework for dependable QoS adaptation in probabilistic environments

“…As discussed in Section 2, this assumption is supported by the results of many recent works (e.g. [6,10,11]). …”

“…Interestingly, in the last few years several works have addressed the problem of probabilistically characterizing the delays in IP-based networks based on real measured data, allowing to conclude that observed empirical delay distributions may be characterized by well known distributions, such as the Weibull distribution [10,6], the shifted gamma distribution [11,4], the exponential distribution [8] or the truncated normal distribution [5]. Based on this, we realized that it would be interesting and appropriate to consider less conservative approaches, by assuming that specific distributions may be identified and thus allowing to achieve better time bounds for the same required coverage.…”

Modelling of inhomogeneity in radar clutter using Weibull mixture densities

“…3a, applying the EM-Bayesian algorithm [14] in which the number of components is set to M = 3, for the same initial condition the following two distinct sets of parameters are obtained [shape, scale, mixture-ratio] …”

A Bayesian Decision Theory Approach for the Techno-Economic Analysis of an All-Optical Router

“…The queuing delay shall be assumed to be Weibull distributed, since this has been shown to accurately capture the queueing delay behaviour of a router with self-similar input traffic [5][6][7]. In this light, the delay probability density function is given by [5]:…”