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

Abstract

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..

Holonic coordination obtained by joining the contract net protocol with constraint satisfaction

“…A distributed solution should not be decided based only on advantages of a distributed mechanism, but also considering the way agents are tightly/loosely coupled, they naturally embed certain knowledge in a dispersed manner and the way information can be shared and communicated [6][7][8]. One must take into account issues regarding privacy and availability of different types of information [9], the need of a minimal communication load [10], and issues concerning the moment when a partial solution is to be communicated between agents -sooner or later [11]. Knowing these issues, our approach is exploiting the specificity of a manufacturing environment and the benefits of holonic method.…”

A colored Petri net model for DisCSP algorithms

“…The evaluation of DisCSP algorithms is dependent on the use of appropriate software instruments. In literature, one can find various distributed constraint reasoning simulation tools, which operate according to DisCSP algorithms and allow their analysis and comparison; in these settings, agents have their own execution thread and communicate via message exchange. In the early stage, Yokoo and collaborators evaluate the efficiency of DisCSP algorithms according to a discrete event simulation.…”