Efficient Communication and Coordination for Large-Scale Multi-Agent Systems

This thesis proposes three adaptive services at the agent framework level to reduce the agent communication and coordination cost of large-scale multi-agent applications. First, communication locality-aware agent distribution aims at minimizing inter-node communication by collocating heavily communicating agents on the same platform and maintaining agent group-based load sharing. Second, application agent-oriented middle agent services attempt to optimize agent, interaction through middle agents by executing application agent-supported search algorithms on the middle agent address space. Third, message passing for mobile agents aims at reducing the time of message delivery to mobile agents using location caches or by extending the agent address scheme with location information. With these services, we have achieved very impressive experimental results in large-scale UAV simulations including up to 10,000 agents. Also, we have provided a formal definition of our framework and services with operational semantics.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Computer Science

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2015

Author dc:creator

• Jang, Myeong-Wuk

Contributors dc:contributor

• Agha, Gul A.

• Artificial Intelligence

Language dc:language

eng

Identifier

(MiAaPQ)AAI3223620

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/81715