Towards optimizing the dynamic placement of tasks by Multi Agent Systems | IEEE Conference Publication | IEEE Xplore