Title: Optimising decentralised grid markets through group selection

Authors: Isaac Chao, Ramon Sanguesa, Oscar Ardaiz, Liviu Joita, Omer F. Rana

Addresses: Computer Architecture Department, Polytechnic University of Catalonia, Spain. ' Computer Architecture Department, Polytechnic University of Catalonia, Spain. ' Department of Mathematics and Informatics, Public University of Navarra, Spain. ' CODAS&IT Department, Culham Science Centre, UKAEA, Abingdon, Oxfordshire, OX14 3DB, UK ' School of Computer Science and Welsh eScience Centre, Cardiff University, Queen's Buildings, 5 The Parade, Roath, Cardiff, CF24 3AA, UK

Abstract: Automatic coordination mechanisms for the grid are required due to the increasing complexity that is exhibited in large-scale distributed systems. Decentralised economic models are being considered as scalable coordination mechanisms for the management of service allocations to clients. However, decentralisation incorporates further dynamicity and unpredictability into the system. Introducing higher levels of adaptation and learning in the coordination protocols helps cope with complexity. We provide a solution based on a self-organised, emergent mechanism that evolves grid market participants through a group selection process. Dynamic congregations organise agents into optimised market segments, maximising utility and thereby improving system-wide performance. We provide a system model and evaluation by simulating the group selection mechanism. We further provide a prototype that shows the practical feasibility of the approach.

Keywords: group selection; selection patterns; decentralised grid markets; self-organisation; grid services.

DOI: 10.1504/IJWGS.2008.022541

International Journal of Web and Grid Services, 2008 Vol.4 No.4, pp.357 - 366

Published online: 14 Jan 2009 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article