Dynamic pricing algorithms for task allocation in multi-agent swarms

Prithviraj Dasgupta, Matthew Hoeing

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Scopus citations

Abstract

Over the past few years, emergent computing based techniques such as swarming have evolved as an attractive technique to design coordination protocols in large-scale distributed systems and massively multi-agent systems. In this paper, we consider a search-and-execute problem domain where agents have to discover tasks online and perform them in a distributed, collaborative manner. We specifically focus on the problem of distributed coordination between agents to dynamically allocate the tasks among themselves. To address this problem, we describe a novel technique that combines a market-based dynamic pricing algorithm to control the task priorities with a swarming-based coordination technique to disseminate task information across the agents. Experimental results within a simulated environment for a distributed aided target recognition application show that the dynamic pricing based task selection strategies compare favorably with other heuristic-based task selection strategies in terms of task completion times while achieving a significant reduction in communication overhead.

Original languageEnglish (US)
Title of host publicationMassively Multi-Agent Technology - AAMAS Workshops - MMAS 2006, LSMAS 2006, and CCMMS 2007, Hakodate, Japan, May 9, 2006, Honolulu, HI, USA, May 15, 2007, Selected and Revised Papers
Pages64-79
Number of pages16
DOIs
StatePublished - 2008
Event1st International Workshop on Coordination and Control in Massively Multi-agent Systems, CCMMS 2007 - Honolulu, HI, United States
Duration: May 15 2007May 15 2007

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5043 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference1st International Workshop on Coordination and Control in Massively Multi-agent Systems, CCMMS 2007
CountryUnited States
CityHonolulu, HI
Period5/15/075/15/07

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Dynamic pricing algorithms for task allocation in multi-agent swarms'. Together they form a unique fingerprint.

  • Cite this

    Dasgupta, P., & Hoeing, M. (2008). Dynamic pricing algorithms for task allocation in multi-agent swarms. In Massively Multi-Agent Technology - AAMAS Workshops - MMAS 2006, LSMAS 2006, and CCMMS 2007, Hakodate, Japan, May 9, 2006, Honolulu, HI, USA, May 15, 2007, Selected and Revised Papers (pp. 64-79). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5043 LNAI). https://doi.org/10.1007/978-3-540-85449-4_5