Air - Ant Inspired Routing

This paper proposes a new architecture to improve vehicle routing using the Ant Colony Optimization (ACO) algorithm to select paths dynamically. ACO mimics the behavior of ants in the natural world. The path chosen by several ants towards a particular food source from the same initial position always converges to one particular path. This path is characterized by the presence of a large amount of pheromones left by ants on their trails. This paper proposes a system in which the option to choose a different path is always available to the agent. The pheromone values increase in paths selected by the ants which successfully make it to the destination. In addition to this, information from traffic sensors may also be used to simulate a traffic jam or emergency in real life by increasing the rate of pheromone evaporation on selected paths. The performance of such a system can be observed by running the algorithm on an artificially created graph which can be updated with pheromone values signifying different situations. Results show that the solutions converge to optimal paths after a certain number of iterations. Further research can be directed towards finding out optimal values of the parameters of the algorithm.