Optimization approach with ρ-proximal convexification for Internet traffic control

Authors

  • Adam Kozakiewicz

DOI:

https://doi.org/10.26636/jtit.2005.3.327

Keywords:

nonlinear programming, price method, convexification, network control

Abstract

The optimization flow control algorithm for traffic control in computer networks, introduced by Steven H. Low, works only for concave utility functions. This assumption is rather optimistic and leads to several problems, especially with streaming applications. In an earlier paper we introduced a modification of the algorithm based on the idea of proximal convexification. In this paper we extend this approach, replacing the proximal method with the ρ-proximal method. The new method mixes the quadratic proximal term with higher-order terms, achieving better results. The algorithms are compared in a simple numerical experiment.

Downloads

Download data is not yet available.

Downloads

Published

2005-09-30

Issue

Section

ARTICLES FROM THIS ISSUE

How to Cite

[1]
A. Kozakiewicz, “Optimization approach with ρ-proximal convexification for Internet traffic control”, JTIT, vol. 21, no. 3, pp. 37–42, Sep. 2005, doi: 10.26636/jtit.2005.3.327.