Distributed Constraint Satisfaction in Dynamic Environments

Dynamic constraint satisfaction solves problems where the environment changes over time, i.e. new constraints and/or variables are added/deleted. Although a substantial amount of research has been carried out on dynamic constraints in centralised systems, little effort has been devoted to this field in distributed environments. In this project we investigate dynamic constraints in distributed environments and we develop algorithms for solving problems where constraints change over time.

We have developed several constructive algorithms for dynamic constraint satisfaction. These algorithms handle problem changes by modifying the existing solution while raising the knowledge gained from solving the original problem. Empirical results suggest faster problem resolution and increased stability.

Funding

• School of Computing (RGU)

Research Team

• Hatem Ahriz
• Ines Arana (PI)
• Bayo Omomowo (PhD student)

Related project

• Penalty—driven Algorithms for Distributed Constraint Satisfaction
• Hybrid Algorithms for Distributed Constraint Satisfaction

Publications

Omomowo, B., Arana, I & Ahriz, H, DynABT: Dynamic Asynchronous Backtracking for Dynamic DisCSPs. In Proceedings of the 13th International Conference on Artificial Intelligence: Methodology, Systems, Applications (AIMSA), Varna, Bulgaria, 4-6 September 2008. Lecture Notes in Computer Science. Springer, ISBN 978-3-540-85775-4, Volume 5253/2008, DOI 10.1007/978-3-540-85776-1_24, pp. 285-296.