Evolutionary Computing in Network Design: Exploring Pleiotropy and Redundancy Trade-offs

Title: Evolutionary Computing in Network Design: Exploring Pleiotropy and Redundancy Trade-offs

Research Question: How can evolutionary computing be used to optimize the design of client-server networks by balancing pleiotropy and redundancy?

Methodology: The study uses evolutionary computation, specifically genetic algorithms, to model and solve the problem of network design. The algorithms are applied to a client-server based network, where servers perform multiple tasks (pleiotropy) and clients are connected to multiple servers (redundancy). The researchers explore how factors such as link failure probability, repair rates, and network size influence the design choices.

Results: The study found that by using evolutionary computation, it is possible to create a network design that balances pleiotropy and redundancy. The algorithms were able to evolve solutions that minimized cost and maximized reliability and flexibility. The results showed that the design choices were influenced by the link failure probability, repair rates, and network size.

Implications: The research has important implications for the field of network design. It demonstrates that evolutionary computing can be used to optimize network design by balancing pleiotropy and redundancy. This can lead to more efficient and reliable networks that are better suited to meet the demands of modern communication systems. The study also provides a new approach to exploring the trade-offs between pleiotropy and redundancy in network design, which can be applied to other areas of engineering and computer science.

Link to Article: https://arxiv.org/abs/0404017v1 Authors: arXiv ID: 0404017v1