Dear Manar,

The problem can be implemented using MATLAB. It requires some efforts but it is possible see the following paper please.

http://eprints.lib.hokudai.ac.jp/dspace/bitstream/2115/46831/1/3.%20paper.pdf

Dear Manar

MATLAB is definitely the best to go in this case. It is very easy to learn. The implementation of binary encoded GA could be some challenging for a beginner, however, real coded could be simpler from an implementation perspective, especially in MATLAB..

Check out this publication .. it might help..

http://connection.ebscohost.com/c/articles/88838522/meta-heuristic-based-adaptation-engine-cognitive-radio-systems

I'm aware that ECJ (Evolutionary Computation in Java) has a facility for multi-population cooperative evolution. In principle it's not that hard to use, though the sheer number of classes in ECJ can be bewildering to novice users. See:

http://cs.gmu.edu/~eclab/projects/ecj/

https://cs.gmu.edu/~eclab/projects/ecj/docs/classdocs/ec/coevolve/MultiPopCoevolutionaryEvaluator.html

You might also check out some of the other common evolutionary algorithm frameworks, since some of these may support coevolution, e.g.:

DEAP (python) - https://code.google.com/p/deap/

EO (C++) - http://eodev.sourceforge.net

EvA2 (Java) - http://www.ra.cs.uni-tuebingen.de/software/JavaEvA/

HeuristicLab (C#) - http://dev.heuristiclab.com/

OpenBEAGLE (C++) - https://code.google.com/p/beagle/

If you are familiar with java, there is another implementation, called jMetal (http://jmetal.sourceforge.net/) .  The binary and real coded GA is already implemented. You just need to use them.

Thanks.

http://jmetal.sourceforge.net/

Thanks every one for your replies. I now have a great collection of tools that I can explore for my needs. This is really very helpful.

All the best,

Manar