A Distributed Approach to Ant Colony Optimization

Systems, Man, and …, 1996

An analogy with the way ant colonies function has suggested the definition of a new computational paradigm, which we call Ant System. We propose it as a viable new approach to stochastic combinatorial optimization. The main characteristics of this model are positive feedback, distributed computation, and the use of a constructive greedy heuristic. Positive feedback accounts for rapid discovery of good solutions, distributed computation avoids premature convergence, and the greedy heuristic helps find acceptable solutions in the early stages of the search process. We apply the proposed methodology to the classical Traveling Salesman Problem (TSP), and report simulation results. We also discuss parameter selection and the early setups of the model, and compare it with tabu search and simulated annealing using TSP. To demonstrate the robustness of the approach, we show how the Ant System (AS) can be applied to other optimization problems like the asymmetric traveling salesman, the quadratic assignment and the job-shop scheduling. Finally we discuss the salient characteristicsglobal data structure revision, distributed communication and probabilistic transitions of the AS.

Proceedings of the first …, 1991

Ants colonies exhibit very interesting behaviours: even if a single ant only has simple capabilities, the behaviour of a whole ant colony is highly structured. This is the result of coordinated interactions. But, as communication possibilities among ants are very limited, interactions must be based on very simple flows of information. In this paper we explore the implications that the study of ants behaviour can have on problem solving and optimization. We introduce a distributed problem solving environment and propose its use to search for a solution to the travelling salesman problem.

Evolutionary Computation, IEEE …, 2002

This paper introduces the ant colony system (ACS), a distributed algorithm that is applied to the traveling salesman problem (TSP). In the ACS, a set of cooperating agents called ants cooperate to find good solutions to TSP's. Ants cooperate using an indirect form of communication mediated by a pheromone they deposit on the edges of the TSP graph while building solutions. We study the ACS by running experiments to understand its operation. The results show that the ACS outperforms other nature-inspired algorithms such as simulated annealing and evolutionary computation, and we conclude comparing ACS-3-opt, a version of the ACS augmented with a local search procedure, to some of the best performing algorithms for symmetric and asymmetric TSP's.

Swarm intelligence, a branch of artificial intelligence is a part which discusses the collective behaviour of social animals such as ants, fishes, termites, birds, bacteria. The collective behaviour of animals to achieve target can be used in practical applications. One of the applications is ant colony optimization. Ongoing research of ACO, there are diverse applications namely data mining, image processing, power electronic circuit design etc. One of that is network routing. By using ACO, we can find the shortest path in network routing

1995

An analogy with the way ant colonies function has suggested the definition of a new computational paradigm, which we call Ant System. We propose it as a viable new approach to stochastic combinatorial optimization. The main characteristics of this model are positive feedback, distributed computation, and the use of a constructive greedy heuristic. Positive feedback accounts for rapid discovery of

This paper presents a multi-agent distributed framework for Swarm Intelligence (SI) based on our previous work ACODA (Ant Colony Optimization on a Distributed Architecture). Our framework can be used to distribute SI algorithms for solving graph search problems on a computer network. Examples and experimental results are given for SI algorithms of: Ant Colony System (ACS) and Bee Colony Optimization (BCO). In order to use the framework, the SI algorithms must be conceptualized to take advantage of the inherent parallelism determined by their analogy with natural phenomena (biological, chemical, physical, etc.): (i) the physical environment of the swarm entities is represented as a distributed multi-agent system and (ii) entities' movement in the physical environment is represented as messages exchanged asynchronously between the agents of the problem environment. We present initial experimental results that show that our framework is scalable. We then compare the results of the distributed implementations of BCO and ACS algorithms using our framework. The conclusion was that our approach scales better when implementing the ACS algorithm but is faster when implementing BCO.

Rcc, 2001

An Ants System is an artificial system based on the behavior of real ant colonies, which is used to solve combinatorial problems. This is a distributed algorithm composed by a set of cooperating agents called ants which cooperate among them to find good solutions to combinatorial optimization problems. The cooperation follows the behavior of real ants using an indirect form of communication mediated by a pheromone. In this work, we present a new distributed algorithm based on Ant System concepts, called the General Ant System, to solve Combinatorial Optimization Problems. Our approach consist on mapping the solution space of the Combinatorial Optimization Problem on the space where the ants will walk, and on defining the transition probability of the Ant System according to the objective function of the Combinatorial Optimization Problem. We test our approach on the Graph Partitioning and The Traveling Salesman Problems. The results show that our approach has the same performances than previous versions of Ant Systems.

Proceedings of the 5 …, 2003

In this paper the analogy between biological swarms and artificial multiagent systems is pointed out. As an example the steps required to model the artificial optimization technique called Ant Colony Optimization starting from the foraging behaviour of natural ant colonies are explained in detail. During the development of the model the authors use the language of multiagent systems to highlight the suitability of such an approach.

Lecture Notes in Computer Science, 2004

2007 IEEE Swarm Intelligence Symposium, 2007

Network resources management issues in complex and dynamic scenarios require decentralized solutions and adaptive systems to face critical and unattended situations. Bioinspired techniques such as swarm intelligence algorithms, have proved to be robust and suitable for managing tasks like routing, load-balancing or resource discovery. In this paper we describe Solenopsis, a framework for the development, simulation and deployment of ant-algorithms, which is aimed at supporting network management middlewares. The system provides a modular and scalable environment that can be distributed over a network. Ants are coded using a simple programming language, and are able to migrate across nodes. Two basic load-balancing algorithms are presented and evaluated, as an example of how this tool works and can be used in practice.