Modeling Available Bandwidth for an Efficient QoS Characterization of a Network Path

• Published in: Lecture notes in computer science pp. 316 - 327
• Main Authors: Chobanyan, Alexander, Mutka, Matt W., Mandrekar, V. S., Xi, Ning
• Format: Book Chapter Conference Proceeding
• Language: English
• Published: Berlin, Heidelberg Springer Berlin Heidelberg 2005; Springer
• Series: Lecture Notes in Computer Science
• Subjects: Applied sciences; Computer science; control theory; systems; Computer systems and distributed systems. User interface; Exact sciences and technology; Network measurements; Software; Statistics; Stochastic processes; Channel capacity; Probabilistic approach; Time series; Multipath propagation; Autoregressive model; Distributed system; Real time; Behavior model; Modeling; Heteroscedasticity; Prospective; Bandwidth; Reliability; Time analysis; Service quality; Execution time
• ISBN: 9783540258094; 3540258094
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/11422778_26

Estimating the reliability of an end–to–end network path is critically important for applications that support remote real–time task execution. Available bandwidth, which is defined as a minimum spare capacity of links constituting a network path, is an important QoS characteristic of the path. In this work we demonstrate a new approach to modelling available bandwidth behavior from a time-series analysis prospective. In particular, we introduce a notion of crossing probability–the probability that available bandwidth drops below the QoS critical threshold for the period of time required for a real-time task execution.We estimate “crossing probability” by an application of the ARCH2 (AutoRegressive Conditional Heteroscedasticity) model to available bandwidth behavior. We estimate model coefficients β0 and β1 to quickly output “crossing probability” for arbitrary values of threshold and length of the real-time task. The model was evaluated on real bandwidth measurements across multiple network paths.