Simple ACO and ACO with Neural Network for implementing multicast routing in a network A Comparison based study

Abstract: This paper introduces a comparative study of implementing simple ACO and ACO using neural network concept for multicast routing in a network.ACO optimizes the path by which ants search for their food by laying a chemical substance named pheromone. Ants on the graph move from a node to other node depending on the corresponding probabilities function, and update the pheromone locally and globally on graph when every iteration is finished. The complexity study, iteration wise shows that merging ACS with neural networks provides features of well performance of cost, fast convergence and stable delay to reach optimal solution as compared to simple ACO.

Keywords: Ant Colony System, Ant colony optimization, pheromone, local pheromone update rule, state transition rule, global pheromone updating rule, synaptic weight.