Distributed evolutionary computation programs often needs increasingly big amounts of computational power when tackling large instances of hard optimization problems, and Peer-to-Peer (P2P) systems could be an option for building the large virtual supercomputer in which they could be run. Even as distributed Evolutionary Algorithms (EA) do take advantage of parallel execution by simultaneously promoting diversity and reducing runtime, there are still many challenges on the parallelization of EAs in P2P systems. In this chapter we present a survey of the state of the art in P2P EAs and our solutions to the main P2P issues such as decentralization, massive scalability and fault tolerance. © 2010 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Jimenez Laredo, J. L., Merelo Guervos, J. J., & Castillo Valdivieso, P. A. (2010). Evolvable agents: A framework for peer-to-peer evolutionary algorithms. Studies in Computational Intelligence, 269, 43–62. https://doi.org/10.1007/978-3-642-10675-0_3
Mendeley helps you to discover research relevant for your work.