Community-based load balancing for massively multi-agent systems

Naoki Miyata, Toru Ishida

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

6 Citations (Scopus)

Abstract

Recently, large-scale distributed multiagent systems consisting of one million of agents have been developed. When agents are distributed among multiple servers, both the computational and interaction cost of servers must be considered when optimizing the performance of the entire system. Multiagent systems reflect the structure of social communities and artificial networks such as the Internet. Since the networks possess characteristics common to the 'small world' phenomenon, networks of agents on the systems can be considered as small worlds. In that case, communities, which are the sets of agents that frequently interact with each other, exist in the network. Most previous works evaluate agents one by one to select the most appropriate agent to be moved to a different server. If the networks of agents are highly clustered, previous works divide the communities when moving agents. Since agents in the same community often interact with each other, this division of communities increases the interaction cost among servers. We propose community-based load balancing (CLB), which evaluates the communities to select the most appropriate set of agents to be moved. We conducted simulations to evaluate our proposed method according to the network of agents. Our simulations show that when the clustering coefficient is close to 1.0, the interaction cost with CLB can be approximately 30% lower than that with previous works.

Original languageEnglish
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
Pages28-42
Number of pages15
DOIs
Publication statusPublished - 2008 Oct 27
Externally publishedYes
Event1st International Workshop on Coordination and Control in Massively Multi-agent Systems, CCMMS 2007 - Honolulu, HI, United States
Duration: 2007 May 152007 May 15

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
Period07/5/1507/5/15

Fingerprint

Multi agent systems
Load Balancing
Resource allocation
Multi-agent Systems
Servers
Server
Costs
Small World
Evaluate
Community
Interaction
Internet
Clustering Coefficient
Divides
Distributed Systems
Division
Simulation
Entire

Keywords

  • Fault tolerance and dependability
  • Mobile agents
  • Scalability and performance issues: robustness

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Miyata, N., & Ishida, T. (2008). Community-based load balancing for massively multi-agent systems. 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. 28-42). (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_3

Community-based load balancing for massively multi-agent systems. / Miyata, Naoki; Ishida, Toru.

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. 2008. p. 28-42 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5043 LNAI).

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

Miyata, N & Ishida, T 2008, Community-based load balancing for massively multi-agent systems. 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. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 5043 LNAI, pp. 28-42, 1st International Workshop on Coordination and Control in Massively Multi-agent Systems, CCMMS 2007, Honolulu, HI, United States, 07/5/15. https://doi.org/10.1007/978-3-540-85449-4_3
Miyata N, Ishida T. Community-based load balancing for massively multi-agent systems. 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. 2008. p. 28-42. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-540-85449-4_3
Miyata, Naoki ; Ishida, Toru. / Community-based load balancing for massively multi-agent systems. 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. 2008. pp. 28-42 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{b57f1a04be614f7f82f05e3327589857,
title = "Community-based load balancing for massively multi-agent systems",
abstract = "Recently, large-scale distributed multiagent systems consisting of one million of agents have been developed. When agents are distributed among multiple servers, both the computational and interaction cost of servers must be considered when optimizing the performance of the entire system. Multiagent systems reflect the structure of social communities and artificial networks such as the Internet. Since the networks possess characteristics common to the 'small world' phenomenon, networks of agents on the systems can be considered as small worlds. In that case, communities, which are the sets of agents that frequently interact with each other, exist in the network. Most previous works evaluate agents one by one to select the most appropriate agent to be moved to a different server. If the networks of agents are highly clustered, previous works divide the communities when moving agents. Since agents in the same community often interact with each other, this division of communities increases the interaction cost among servers. We propose community-based load balancing (CLB), which evaluates the communities to select the most appropriate set of agents to be moved. We conducted simulations to evaluate our proposed method according to the network of agents. Our simulations show that when the clustering coefficient is close to 1.0, the interaction cost with CLB can be approximately 30{\%} lower than that with previous works.",
keywords = "Fault tolerance and dependability, Mobile agents, Scalability and performance issues: robustness",
author = "Naoki Miyata and Toru Ishida",
year = "2008",
month = "10",
day = "27",
doi = "10.1007/978-3-540-85449-4_3",
language = "English",
isbn = "3540854487",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "28--42",
booktitle = "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",

}

TY - GEN

T1 - Community-based load balancing for massively multi-agent systems

AU - Miyata, Naoki

AU - Ishida, Toru

PY - 2008/10/27

Y1 - 2008/10/27

N2 - Recently, large-scale distributed multiagent systems consisting of one million of agents have been developed. When agents are distributed among multiple servers, both the computational and interaction cost of servers must be considered when optimizing the performance of the entire system. Multiagent systems reflect the structure of social communities and artificial networks such as the Internet. Since the networks possess characteristics common to the 'small world' phenomenon, networks of agents on the systems can be considered as small worlds. In that case, communities, which are the sets of agents that frequently interact with each other, exist in the network. Most previous works evaluate agents one by one to select the most appropriate agent to be moved to a different server. If the networks of agents are highly clustered, previous works divide the communities when moving agents. Since agents in the same community often interact with each other, this division of communities increases the interaction cost among servers. We propose community-based load balancing (CLB), which evaluates the communities to select the most appropriate set of agents to be moved. We conducted simulations to evaluate our proposed method according to the network of agents. Our simulations show that when the clustering coefficient is close to 1.0, the interaction cost with CLB can be approximately 30% lower than that with previous works.

AB - Recently, large-scale distributed multiagent systems consisting of one million of agents have been developed. When agents are distributed among multiple servers, both the computational and interaction cost of servers must be considered when optimizing the performance of the entire system. Multiagent systems reflect the structure of social communities and artificial networks such as the Internet. Since the networks possess characteristics common to the 'small world' phenomenon, networks of agents on the systems can be considered as small worlds. In that case, communities, which are the sets of agents that frequently interact with each other, exist in the network. Most previous works evaluate agents one by one to select the most appropriate agent to be moved to a different server. If the networks of agents are highly clustered, previous works divide the communities when moving agents. Since agents in the same community often interact with each other, this division of communities increases the interaction cost among servers. We propose community-based load balancing (CLB), which evaluates the communities to select the most appropriate set of agents to be moved. We conducted simulations to evaluate our proposed method according to the network of agents. Our simulations show that when the clustering coefficient is close to 1.0, the interaction cost with CLB can be approximately 30% lower than that with previous works.

KW - Fault tolerance and dependability

KW - Mobile agents

KW - Scalability and performance issues: robustness

UR - http://www.scopus.com/inward/record.url?scp=54249153759&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=54249153759&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-85449-4_3

DO - 10.1007/978-3-540-85449-4_3

M3 - Conference contribution

AN - SCOPUS:54249153759

SN - 3540854487

SN - 9783540854487

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 28

EP - 42

BT - 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

ER -