Effective Distributed Genetic Algorithms for Optimizing Social Utility

Most real-world negotiation involves multiple interdependent issues that makes an agent's utility function nonlinear. Traditional negotiation mechanisms, which were designed for linear utilities, do not fare well in nonlinear contexts. One of the main challenges in developing effective nonlinear negotiation protocols is scalability, which can produce excessively high failure rate when there are many issues due to computational intractability. One reasonable approach to reducing computational cost while maintaining quality outcomes is to decompose the utility space into several largely independent sub spaces. In this paper, we propose a new method for decomposing a utility space based on the interdependency of issues and employing the genetic algorithms in each issue-group. In addition, our experimental results demonstrate that our method can find higher quality solutions than existing works. They also show that our method is highly effective for reducing the execution time.