ant colony optimisation project description

1
Department of Electrical & Electronic Engineering University of Melbourne MB3 - Ant Colony Optimisation in Network Problem Students: Darryl Foo, Johnatan Feuillye and Jun-Huang Chen Supervisor: Dr Marcus Brazil (Senior Lecturer, Telecommunications) Researchers have found that control algorithms based on the socio- biological foraging, pheromone-driven behaviour of ants can help solve difficult mathematical, combinatorial problems, such as finding the shortest path within a given network. This method is called Ant Colony Metaheuristics. Ant Colony Metaheuristics has potentials in a wide range of applications. Varying but promising levels of success have been shown in applications ranging from the traveling salesman problem, the quadratic assignment problem , the vehicle routing problem, scheduling problems, telecommunications networks, and much more. Its success in static combinatorial problems suggests that it can potentially be very useful in dynamic networks such as telecommunication networks. Further research and exploration in this area can help optimize existing telecommunication networks. This project’s objective is to focus on the applications of Ant Colony Metaheuristics in telecommunication networks, and in particular, the internet, by analysing, developing and improving on existing Ant Colony algorithms for telecommunication applications. The group aims to use Ant Colony Metaheuristics to optimize the routing of data in these networks. Using the MATLAB programming language, the group will simulate the internet and implement the “Ant Net” algorithm developed by Marco Dorigo and Gianni Di Caro as a starting point. The “Ant Net” algorithm showed very promising improvement in performance against existing state of the art telecommunication routing algorithms. The group aim to further this research and construct more detailed simulation of the internet and derive optimal algorithmic variants for the internet. In this way a packet-switched data-network such as the internet can be specifically optimized and the question of which algorithmic modifications are best suited for such networks can be explored. Performance of the internet in data transfer can be significantly improved, both through an increase in the transfer rate and an increase in the probability of successful data transfer. Research carried out in this area is of potential commercial interest to organisations using similar data networks.

Upload: darryl-foo

Post on 11-Apr-2015

1.082 views

Category:

Documents


1 download

DESCRIPTION

Ant Colony Optimisation Project Description

TRANSCRIPT

Page 1: Ant Colony Optimisation Project Description

Department of Electrical & Electronic Engineering University of Melbourne

MB3 - Ant Colony Optimisation in Network Problem Students: Darryl Foo, Johnatan Feuillye and Jun-Huang Chen Supervisor: Dr Marcus Brazil (Senior Lecturer, Telecommunications) Researchers have found that control algorithms based on the socio-biological foraging, pheromone-driven behaviour of ants can help solve difficult mathematical, combinatorial problems, such as finding the shortest path within a given network. This method is called Ant Colony Metaheuristics. Ant Colony Metaheuristics has potentials in a wide range of applications. Varying but promising levels of success have been shown in applications ranging from the traveling salesman problem, the quadratic assignment problem , the vehicle routing problem, scheduling problems, telecommunications networks, and much more. Its success in static combinatorial problems suggests that it can potentially be very useful in dynamic networks such as telecommunication networks. Further research and exploration in this area can help optimize existing telecommunication networks. This project’s objective is to focus on the applications of Ant Colony Metaheuristics in telecommunication networks, and in particular, the internet, by analysing, developing and improving on existing Ant Colony algorithms for telecommunication applications. The group aims to use Ant Colony Metaheuristics to optimize the routing of data in these networks. Using the MATLAB programming language, the group will simulate the internet and implement the “Ant Net” algorithm developed by Marco Dorigo and Gianni Di Caro as a starting point. The “Ant Net” algorithm showed very promising improvement in performance against existing state of the art telecommunication routing algorithms. The group aim to further this research and construct more detailed simulation of the internet and derive optimal algorithmic variants for the internet. In this way a packet-switched data-network such as the internet can be specifically optimized and the question of which algorithmic modifications are best suited for such networks can be explored. Performance of the internet in data transfer can be significantly improved, both through an increase in the transfer rate and an increase in the probability of successful data transfer. Research carried out in this area is of potential commercial interest to organisations using similar data networks.