Title: VHDL implementation of neuro-fuzzy based adaptive bandwidth controller for ATM networks

Authors: M. Sundarambal, D. Chandrakala, P. Anbalagan

Addresses: Department of Electrical and Electronics Engineering, Coimbatore Institute of Technology, Coimbatore – 641 014, Tamilnadu, India. ' Department of Electrical and Electronics Engineering, Coimbatore Institute of Technology, Coimbatore – 641 014, Tamilnadu, India. ' Department of Electrical and Electronics Engineering, Coimbatore Institute of Technology, Coimbatore – 641 014, Tamilnadu, India

Abstract: The modern intelligent asynchronous transfer mode (ATM) networks need a scheme to maintain quality of service (QoS) towards different traffic constraints in real time environment for effective control and optimal system utilisation. This paper proposes the VHDL implementation of neuro-fuzzy based adaptive bandwidth controller. The intelligent control is used to manage the adaptive bandwidth prediction, congestion control and call admission control (CAC) of ATM networks simultaneously. This integrated hybrid mechanism reduces the decision errors of conventional CAC policies and operating errors occurred on fuzzy based CAC in ATM networks. The individual blocks needed to implement adaptive bandwidth controller for ATM networks are modelled using VHDL. The timing analysis for the validation, functionality and performance of the model is analysed using Altera Quartus II (5.1 version) HDL software and the logic synthesis is performed. The required codes are downloaded into FPGA Cyclone device.

Keywords: ATM networks; quality of service; QoS; VHSIC hardware description language; VHDL; fuzzy logic; neural networks; bandwidth controllers; adaptive control; congestion control; call admission control; CAC; FPGA; intelligent control; bandwidth prediction; field-programmable gate arrays.

DOI: 10.1504/IJCNDS.2009.026824

International Journal of Communication Networks and Distributed Systems, 2009 Vol.3 No.2, pp.159 - 174

Published online: 26 Jun 2009 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article