Re: Any Colony Optimization

[Andre Costa]

On Wed, 23 Aug 2000, john harvey wrote:

>   Do you know any examples of ant colony optimization models using Swarm

Hi,

I've written a Swarm simulation in Java to model artificial ant-like
agents for
routing in a telecommunications network. This is a very specific
application of the Ant Colony Optimization concept, but is not a
general-purpose Ant Colony Optimization package. 

I am in the process of updating to Swarm 2.1.1 and updating the
ant-routing code, but will be happy to send it to
you when I'm done. I have only implemented an "ant" layer so far, and no
data layer as yet, but I will be doing this soon as well.

As you probably already know, there are papers available at

        http://iridia.ulb.ac.be/~mdorigo/ACO/ACO.html#NET

which describe the telecommunication application of ACO and simulation
results, although Swarm was not used for those particular simulations.

Although completely feasible, I don't think it would be especially useful
or efficient to use
Swarm, a simulation package, to code a general-purpose optimization
algorithm, unless one infact wants to run a simulation environment per-se.

The ant-based routing system is intended for a dynamic environment where
the "optimum solution" varies with time, and thus one would simulate such
a system to see how it evolves and "tracks" the optimal solution. Swarm is
great for this kind of simulation, but is perhaps not the most efficient
environment for a straight ACO application to something like the (static)
TSP ?

cheers,
Andre.

---------------------------------------------------------------------
Andre Costa                             
Research Student   

Teletraffic Research Centre           Telephone: (08) 8303 3237
Department of Applied Mathematics     Fax:       (08) 8303 4395
University of Adelaide                Email:address@hidden
South Australia  5005
----------------------------------------------------------------------          



                  ==================================
   Swarm-Modelling is for discussion of Simulation and Modelling techniques
   esp. using Swarm.  For list administration needs (esp. [un]subscribing),
   please send a message to <address@hidden> with "help" in the
   body of the message.
                  ==================================