Completeness and Performance Of The APO Algorithm

Grinshpoun, Tal; Meisels, Amnon

doi:10.1613/jair.2611

Cited by 19 publications

(22 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…APO uses the least number of NCCCs, cycles, and messages with NAPO using slightly more. These results seem to contradict the findings of presented by Grinshpoun and Meisels [5] who reported that APO used more NCCCs on medium density problems across various levels of constraint tightness. The discrepancy between these results can likely be explained by the difference in the experiments that were conducted.…”

Section: Resultscontrasting

confidence: 99%

See 1 more Smart Citation

Solving Distributed CSPs Using Dynamic, Partial Centralization without Explicit Constraint Passing

Mailler

Graves

2012

Principles and Practice of Multi-Agent Systems

View full text Add to dashboard Cite

Dynamic, partial centralization has received a considerable amount of attention in the distributed problem solving community. As the name implies, this technique works by dynamically identifying portions of a shared problem to centralize in order to speed the problem solving process. Currently, a number of algorithms have been created which employ this simple, yet powerful technique to solve problems such as distributed constraint satisfaction (DCSP), distributed constraint optimization (DCOP), and distributed resource allocation.In fact, one such algorithm, Asynchronous Partial Overlay (APO), was shown to outperform the Asynchronous Weak Commitment (AWC) protocol, which is one of the best known methods for solving DCSPs. One of the key differences between these two algorithms is that APO, as part of the centralization process, uses explicit constraint passing. AWC, on the other hand, passed nogoods because it tries to provide security and privacy. Because of these differences in underlying assumptions, a number of researchers have criticized the comparison between these two protocols.This paper attempts to resolve this disparity by introducing a new AWC/APO algorithm called Nogood-APO that like AWC uses nogood passing to provide security and privacy and like APO uses dynamic partial centralization to speed the problem solving process. Like its parent algorithms, this new protocol is sound and complete and performs nearly as well as APO, while still outperforming AWC, on distributed 3-coloring problems. In addition, this paper shows that Nogood-APO provides more privacy to the agents than both APO and AWC on all but the sparsest problems. These findings demonstrate that a dynamic, partial centralization-based protocol can provide privacy and that even when operating with the same assumptions as AWC still solves problems in fewer cycles using less computation and communication..

show abstract

Section: Resultscontrasting

confidence: 99%

“…We refer the reader to [9] and [5] for the complete details of the proofs for APO. Here is a basic outline of the proof for NAPO:…”

Section: Soundness and Completenessmentioning

confidence: 99%

Solving Distributed CSPs Using Dynamic, Partial Centralization without Explicit Constraint Passing

Mailler

Graves

2012

Principles and Practice of Multi-Agent Systems

View full text Add to dashboard Cite

show abstract

“…APO uses the least number of NCCCs, cycles, and messages with NAPO using slightly more. These results seem to contradict the findings of presented by Grinshpoun and Meisels [8] who reported that APO used more NCCCs on medium density problems across various levels of constraint tightness. The discrepancy between these results can likely be explained by the difference in the experiments that were conducted.…”

Section: Resultscontrasting

confidence: 99%

“…The reader is referred to [14] and [8] for the complete details of the proofs for APO. The original proof of APO's soundness and completeness was originally discovered to be incorrect in early 2005.…”

Section: Soundness and Completenessmentioning

confidence: 99%

“…The error was rediscovered in 2007 by Grinshpoun and Miesels [8] who had reimplemented the original version of APO from published descriptions instead of using the reference implementation. During their investigation, however, they were also able to identify another potential condition that may occur due to delayed state updating at the end of a mediation session.…”

Section: Soundness and Completenessmentioning

confidence: 99%

See 1 more Smart Citation

Improving the privacy of the asynchronous partial overlay protocol

Mailler

2012

MGS

View full text Add to dashboard Cite

Dynamic, partial centralization has received a considerable amount of attention in the distributed problem solving community. As the name implies, this technique works by dynamically identifying portions of a shared problem to centralize in order to speed the problem solving process. Currently, a number of algorithms have been created which employ this simple, yet powerful technique to solve problems such as distributed constraint satisfaction (DCSP), distributed constraint optimization (DCOP), and distributed resource allocation.In fact, one such algorithm, Asynchronous Partial Overlay (APO), 1 was shown to outperform the Asynchronous Weak Commitment (AWC) protocol, which is one of the best known methods for solving DCSPs.One of the key differences between these two algorithms is that APO, as part of the centralization process, uses explicit constraint passing.AWC, on the other hand, passed nogoods because it tries to provide security and privacy. Because of these differences in underlying assumptions, a number of researchers have criticized the comparison between these two protocols.This article attempts to resolve this disparity by introducing a new AWC/APO algorithm called Nogood-APO that like AWC uses nogood passing to provide privacy and like APO uses dynamic, partial centralization to speed the problem solving process. Like its parent algorithms, this new protocol is sound and complete and performs nearly as well as APO, while still outperforming AWC, on distributed 3-coloring problems. In addition, this paper shows that Nogood-APO provides more privacy to the agents than both APO and AWC on all but the sparsest problems. These findings demonstrate that a dynamic, partial centralization-based protocol can provide privacy and that even when operating with the same assumptions as AWC still solves problems in fewer cycles using less computation and communication.

show abstract

A colored Petri net model for DisCSP algorithms

Pascal

Pănescu

2017

Concurrency and Computation

View full text Add to dashboard Cite

Summary The aim of this paper is to present a colored Petri net that models a system applying a DisCSP method. Our study considered 3 representative DisCSP algorithms: synchronous backtracking, asynchronous backtracking, and weak‐commitment search. To obtain the model, it was necessary to transpose the operation of a DisCSP‐based system into a discrete event system. The constructed colored Petri net is presented, with details on the information stored in places, procedures associated with transitions, and communication between agents. Through the performed trials (with the n‐queens problem), it was proved that the model is correct, easy to use, and adaptable to different operation conditions. The conducted simulations revealed new results about the way the 3 analyzed algorithms compare one with the other. It was showed that the performance of weak‐commitment search is degrading in cases close to real conditions, namely, when agents use un‐updated information about the state of system. We could also study 3 strategies for improving the performance of DisCSP algorithms, making them more practicable. The proposed model is available for open use, it is independent of the software that carries out a DisCSP method, and it can be enhanced for other DisCSP algorithms and diverse communication schemes.

show abstract

Completeness and Performance Of The APO Algorithm

Cited by 19 publications

References 20 publications

Solving Distributed CSPs Using Dynamic, Partial Centralization without Explicit Constraint Passing

Solving Distributed CSPs Using Dynamic, Partial Centralization without Explicit Constraint Passing

Improving the privacy of the asynchronous partial overlay protocol

A colored Petri net model for DisCSP algorithms

Contact Info

Product

Resources

About