Dynamic DFS tree in ADOPT-ing

Marius C. Silaghi, Makoto Yokoo

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

2 Citations (Scopus)

Abstract

Several distributed constraint reasoning algorithms employ Depth First Search (DFS) trees on the constraint graph that spans involved agents. In this article we show that it is possible to dynamically detect a minimal DFS tree, compatible with the current order on agents, during the distributed constraint reasoning process of the ADOPT algorithm. This also allows for shorter DFS trees during the initial steps of the algorithm, while some constraints did not yet prove useful given visited combinations of assignments. Earlier distributed algorithms for finding spanning trees on agents did not look to maintain compatibility with an order already used. We also show that announcing a nogood to a single optional agent is bringing significant improvements in the total number of messages. The dynamic detection of the DFS tree brings improvements in simulated time.

Original languageEnglish
Title of host publicationAAAI-07/IAAI-07 Proceedings: 22nd AAAI Conference on Artificial Intelligence and the 19th Innovative Applications of Artificial Intelligence Conference
Pages763-769
Number of pages7
Volume1
Publication statusPublished - 2007
EventAAAI-07/IAAI-07 Proceedings: 22nd AAAI Conference on Artificial Intelligence and the 19th Innovative Applications of Artificial Intelligence Conference - Vancouver, BC, Canada
Duration: Jul 22 2007Jul 26 2007

Other

OtherAAAI-07/IAAI-07 Proceedings: 22nd AAAI Conference on Artificial Intelligence and the 19th Innovative Applications of Artificial Intelligence Conference
CountryCanada
CityVancouver, BC
Period7/22/077/26/07

Fingerprint

Trees (mathematics)
Parallel algorithms

All Science Journal Classification (ASJC) codes

  • Software

Cite this

Silaghi, M. C., & Yokoo, M. (2007). Dynamic DFS tree in ADOPT-ing. In AAAI-07/IAAI-07 Proceedings: 22nd AAAI Conference on Artificial Intelligence and the 19th Innovative Applications of Artificial Intelligence Conference (Vol. 1, pp. 763-769)

Dynamic DFS tree in ADOPT-ing. / Silaghi, Marius C.; Yokoo, Makoto.

AAAI-07/IAAI-07 Proceedings: 22nd AAAI Conference on Artificial Intelligence and the 19th Innovative Applications of Artificial Intelligence Conference. Vol. 1 2007. p. 763-769.

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

Silaghi, MC & Yokoo, M 2007, Dynamic DFS tree in ADOPT-ing. in AAAI-07/IAAI-07 Proceedings: 22nd AAAI Conference on Artificial Intelligence and the 19th Innovative Applications of Artificial Intelligence Conference. vol. 1, pp. 763-769, AAAI-07/IAAI-07 Proceedings: 22nd AAAI Conference on Artificial Intelligence and the 19th Innovative Applications of Artificial Intelligence Conference, Vancouver, BC, Canada, 7/22/07.
Silaghi MC, Yokoo M. Dynamic DFS tree in ADOPT-ing. In AAAI-07/IAAI-07 Proceedings: 22nd AAAI Conference on Artificial Intelligence and the 19th Innovative Applications of Artificial Intelligence Conference. Vol. 1. 2007. p. 763-769
Silaghi, Marius C. ; Yokoo, Makoto. / Dynamic DFS tree in ADOPT-ing. AAAI-07/IAAI-07 Proceedings: 22nd AAAI Conference on Artificial Intelligence and the 19th Innovative Applications of Artificial Intelligence Conference. Vol. 1 2007. pp. 763-769
@inproceedings{1be56f1137eb40e4a87b1656b9514c74,
title = "Dynamic DFS tree in ADOPT-ing",
abstract = "Several distributed constraint reasoning algorithms employ Depth First Search (DFS) trees on the constraint graph that spans involved agents. In this article we show that it is possible to dynamically detect a minimal DFS tree, compatible with the current order on agents, during the distributed constraint reasoning process of the ADOPT algorithm. This also allows for shorter DFS trees during the initial steps of the algorithm, while some constraints did not yet prove useful given visited combinations of assignments. Earlier distributed algorithms for finding spanning trees on agents did not look to maintain compatibility with an order already used. We also show that announcing a nogood to a single optional agent is bringing significant improvements in the total number of messages. The dynamic detection of the DFS tree brings improvements in simulated time.",
author = "Silaghi, {Marius C.} and Makoto Yokoo",
year = "2007",
language = "English",
isbn = "1577353234",
volume = "1",
pages = "763--769",
booktitle = "AAAI-07/IAAI-07 Proceedings: 22nd AAAI Conference on Artificial Intelligence and the 19th Innovative Applications of Artificial Intelligence Conference",

}

TY - GEN

T1 - Dynamic DFS tree in ADOPT-ing

AU - Silaghi, Marius C.

AU - Yokoo, Makoto

PY - 2007

Y1 - 2007

N2 - Several distributed constraint reasoning algorithms employ Depth First Search (DFS) trees on the constraint graph that spans involved agents. In this article we show that it is possible to dynamically detect a minimal DFS tree, compatible with the current order on agents, during the distributed constraint reasoning process of the ADOPT algorithm. This also allows for shorter DFS trees during the initial steps of the algorithm, while some constraints did not yet prove useful given visited combinations of assignments. Earlier distributed algorithms for finding spanning trees on agents did not look to maintain compatibility with an order already used. We also show that announcing a nogood to a single optional agent is bringing significant improvements in the total number of messages. The dynamic detection of the DFS tree brings improvements in simulated time.

AB - Several distributed constraint reasoning algorithms employ Depth First Search (DFS) trees on the constraint graph that spans involved agents. In this article we show that it is possible to dynamically detect a minimal DFS tree, compatible with the current order on agents, during the distributed constraint reasoning process of the ADOPT algorithm. This also allows for shorter DFS trees during the initial steps of the algorithm, while some constraints did not yet prove useful given visited combinations of assignments. Earlier distributed algorithms for finding spanning trees on agents did not look to maintain compatibility with an order already used. We also show that announcing a nogood to a single optional agent is bringing significant improvements in the total number of messages. The dynamic detection of the DFS tree brings improvements in simulated time.

UR - http://www.scopus.com/inward/record.url?scp=36348968268&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=36348968268&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:36348968268

SN - 1577353234

SN - 9781577353232

VL - 1

SP - 763

EP - 769

BT - AAAI-07/IAAI-07 Proceedings: 22nd AAAI Conference on Artificial Intelligence and the 19th Innovative Applications of Artificial Intelligence Conference

ER -