G2DGA: An Adaptive Framework for Internet-based Distributed Genetic Algorithms

Johan Berntsson

The Internet is different from traditional parallel computing environments, 
and Distributed Genetic Algorithms (DGAs) for the Internet need to be designed 
to address these differences. This paper presents a framework for Internet 
island-model DGAs that uses adaptation methods to maintain efficiency and 
robustness in a volatile and dynamic run-time environment. The applicability 
of the methods is demonstrated on benchmark tests, and a real-world 
optimization problem in VLSI design.