A quality of service approach based on neural networks for mobile ad hoc networks

Author

Khoukhi, L. ; Cherkaoui, S.

Author_Institution

Dept. of Electr. & Comput. Eng., Sherbrooke Univ., Que., Canada

fYear

2005

fDate

6-8 March 2005

Firstpage

295

Lastpage

301

Abstract

In this paper, we propose an intelligent quality of service (QoS) model named GQOS, with service differentiation based on neural networks in mobile ad hoc networks. The model is composed of two plans: the GQOS kernel plan and the intelligent learning plan. New mechanisms have been developed and integrated in the kernel plan in order to ensure the detection and recovery of QoS violations. The intelligent learning plan performs the training of GQOS kernel operations by using a multilayered feedforward neural network. Simulation results show that our model outperforms the SWAN model by about 10% in terms of average delay and throughput at lower and medium mobility.

Keywords

ad hoc networks; feedforward neural nets; learning (artificial intelligence); mobile radio; quality of service; telecommunication computing; QoS; intelligent learning plan; kernel plan; mobile ad hoc networks; multilayered feedforward neural network; quality of service approach; service differentiation; Ad hoc networks; Bandwidth; Delay; Intelligent networks; Kernel; Mobile ad hoc networks; Network topology; Neural networks; Quality of service; Telecommunication traffic;

fLanguage

English

Publisher

ieee

Conference_Titel

Wireless and Optical Communications Networks, 2005. WOCN 2005. Second IFIP International Conference on

Print_ISBN

0-7803-9019-9

Type

conf

DOI

10.1109/WOCN.2005.1436037

Filename

1436037