Paper Details: Downloads: 642
Serial Number: P1120827002
Title: Asynchronous Backtracking with Compilation Formulation for handling complex local problems
Authors: Redouane EZZAHIR, Christian BESSIERE, El Houssine Bouyakhf, Mustapha BELAISSAOUI
Abstract: The Asynchronous Backtracking (ABT) algorithm is a well known algorithm for solving distributed constraint satisfaction problems. However, several works concerned with ABT suppose that each agent owns a single variable. This paper presents the compilation formulation for Asynchronous Backtracking with complex local problems, resulting in the Asynchronous Backtracking with compilation formulation algorithm (ABT-cf). The ABT-cf algorithm and its correctness proof are described in detail. A new interchangeability technique is also presented in detail. The performances of ABT-cf are compared to the standard ABT in which the distributed problem is reformulated by decomposition. Experimental evaluation shows that our interchangeability technique gives a consistent performance improvements in terms of non-concurrent constraint checks, and that ABT-cf increases the performance of the distributed search and outperforms standard ABT by a large scale.
Keywords: DCR, Distributed constraints satisfaction problem, ABT algorithm, complex local problems, interchangeability.
Journal/Conference: International Journal of Artificial Intelligence and Machine Learning
Volume: 8
Issue: 3
Submission Date: 7/4/2008 12:00:00 AM
Review Date: 7/25/2008 12:00:00 AM
Publishing Date: 10/26/2008 12:00:00 AM
Article Downloads: 642
Download:

Facebook