Dynamite Project Autonomous Negotiating Teams (ANTS) Program
Abstract
Distributed resource allocation is a general problem in which a set of agents must intelligently assign their resources to a set of dynamic tasks. Obtaining effective performance in dynamic domains is difficult because agents must deal with significant challenges such as coping with limited resources, ability to respond to changes in tasks, coordinating with other agents and adhering to domain-imposed communication restraints. As part of the DARPA Autonomous Negotiating Teams (ANTs) program, this project addressed shortcomings in current distributed resource allocation research. Significant advancements were made in three areas: 1) Creation of abstract formalizations of the problem that allows the development of general solution strategies, 2) Development of the Adopt algorithm, which addresses solving dynamic distributed constraint problems, and 3) Application of the techniques in a practical domain involving real-world hardware. Target tracking in distributed sensor networks is the selected application area addressed by this report.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2003
- Accession Number
- ADA417450
Entities
People
- Hyuckchul Jung
- Milind Tambe
- Pagnesh J. Modi
- Wei-min Shen
Organizations
- University of Southern California