Abstract
In this paper we address the problem of agents engaged in a distributed costly two-sided search for pairwise partnerships in Multi-Agent Systems (MAS). While traditional two-sided search mechanisms are based on a purely sequential search of all searchers, our mechanism integrates an ability of some of the agents to maintain several search efforts in parallel at each search stage. We show that in many environments the transition to the new mechanism is inevitable since the adoption of the parallel-interactions based search suggests a greater utility for the searching agents. By exploring the appropriate model equations, we present the new dynamics that drive the equilibrium when using such a mechanism in MAS environments. Complementary algorithms are offered, based on the unique equilibria characteristics found, for facilitating the extraction of the agents' strategies. The analysis methodology used supplies a comprehensive solution to a self contained model, and also offers a great value for future work concerning distributed two-sided mechanisms for MAS. Towards the end of the paper we review two of these important models that can benefit from the proposed analysis.
Original language | English |
---|---|
Pages (from-to) | 541-569 |
Number of pages | 29 |
Journal | Artificial Intelligence |
Volume | 172 |
Issue number | 4-5 |
DOIs | |
State | Published - Mar 2008 |
Bibliographical note
Funding Information:* Corresponding author. E-mail addresses: [email protected] (D. Sarne), [email protected] (S. Kraus). 1 The research was partially supported by NSF grant IIS0705587 and ISF. Sarit Kraus is also affiliated with the Institute for Advanced Computer Studies, University of Maryland, College Park, MD.
Funding
* Corresponding author. E-mail addresses: [email protected] (D. Sarne), [email protected] (S. Kraus). 1 The research was partially supported by NSF grant IIS0705587 and ISF. Sarit Kraus is also affiliated with the Institute for Advanced Computer Studies, University of Maryland, College Park, MD.
Funders | Funder number |
---|---|
National Science Foundation | IIS0705587 |
Directorate for Computer and Information Science and Engineering | 0705587 |
Iowa Science Foundation |
Keywords
- Autonomous agents
- Equilibrium analysis
- Matching
- Multi-agent systems