xxGA is software library for parallel genetic algorithms. It implements in a unify manner the most well-known parallel models in the literature (like cellular or distriuted parallel models). The main models implemented in this generic frameworks are:

- genGA (Generational Genetic
Algorithm)

This model generates a complete new population in each generation of the GA. This new population replaces the old one. - ssGA (Steady-State Genetic
Algorithm)

In this scheme, only one or a small number of new individual are generated in each iteration of the main loop of GA method. Usually, these new individuals replace the worst solutions of the old population. - cGA (Cellular Genetic Algorithm)

In cGA the population is structured in a grid and a neighborhood structure is defined on it. Then, individuals can only interacts with its neighbors. The most well-known neighborhood (NEWS) includes five solutions: the current one, and its north, east, west, and south neighbor solutions. - dcGA (Distributed Cellular
Genetic Algorithm)

n dcGA the population is partitioned in several subpopulations (islands) and each island execute a cGA scheme. All island cooperate to solve the problem interchanging some information (solutions) among them according to a predefined migration policy.

This website is structured as follows:

- An example of use of xxGA (monoprocessor version): we have
included an example where a cGA is used to solve the ONEMAX problem

(instance of 100 variables). - How a new problem can include to this library.
- Some issues about the compilation of this framework.
- The configuration of the 'island.params' file.
- The self-adaptive methods for the population shape.
- The implementation a distributed version.