Home // ICAS 2019, The Fifteenth International Conference on Autonomic and Autonomous Systems // View article

Adaptive Control of Traffic Congestion with Neuro-Fuzzy based Weighted Random Early Detection

Authors:
Irina Topalova
Pavlinka Radoyska

Keywords: traffic congestion; Quality of Service; Weighted Random Early Detection, fuzzy logic; neuro-fuzzy system

Abstract:
Differentiating class-based traffic and class-based queue management is the most advanced approach for queue management in routers and switches, controlling and preventing the congestion. The combination of a mechanism for prioritizing the IP traffic and the way to dynamically modify the parameters of the packet rejection algorithm is essential for achieving efficient and reliable traffic. In this study a method is proposed, exploring the automatically adapting of new users added to the backbone of the network, to the already defined weighted random early detection parameters. A neuro-fuzzy-logic network is trained to automatically adapt new end users to the quality of service policy, already set in the backbone area. This network is trained with the quality of service parameters of the backbone area and serves to adapt these parameters in the newly-added routers. The results obtained are compared with those, from the study of this problem by the authors, when a multilayer neural network is used.

Pages: 59 to 64

Copyright: Copyright (c) IARIA, 2019

Publication date: June 2, 2019

Published in: conference

ISSN: 2308-3913

ISBN: 978-1-61208-712-2

Location: Athens, Greece

Dates: from June 2, 2019 to June 6, 2019