Performance analysis of new multiple access protocols using genetic algorithm
Kandasamy Gunavathi, A. Shanmugam · Indian Journal of Engineering and Materials Sciences · 1998
Genetic Algorithms are robust search and optimisation methods that have emerged to be effective for real-world problems. Most of the technical fields involve some form of optimisation that is required in the process of design. As, design is an open ended problem with many solutions, the quest is to find out the best solution according to some criterion. During the last decade from 1986 to 1996, the extent of change in local area networking has been phenomenal. There is a need for a high-speed network that will extend beyond the area covered by a LAN, but not restricted to normal methods of wide area networking. In this paper, four new and effective multiple access protocols for bus LANs have been proposed. Two of the protocols are proposed for single channel shared systems and two other protocols are proposed for multi channel shared systems. The multichannel protocols are of DAMA kind and are of new proposals of its kind. The protocols are modelled as optimisation problems and are ana lysed for their best and worst performanc9s through Genetic Algorithms. The protocols ensure high throughput and reduced bounded packet delay compared to the known multiple access protocols of the bus network. The fairness, collision-free transmission, distributed control along with the high throughput, bounded and reduced packet delay makes the protocols to be the best alternatives for the existing collision prone protocols of the bus networks.