Distributed stochastic search and distributed breakout: properties, comparison and applications to constraint optimization problems in sensor networks

This research is motivated by some distributed scheduling problems in distributed sensor networks, in which computational and communication resources are scarce. We first cast these problems as distributed constraint satisfaction problems (DisCSPs) and distributed constraint optimization problems (DisCOPs) and model them as distributed graph coloring. To cope with limited resources and restricted real-time requirement, it is imperative to use distributed algorithms that have low overhead on resource consumption and high-quality anytime performance. To meet these requirements, we study two existing DisCSP algorithms, distributed stochastic search algorithm (DSA) and distributed breakout algorithm (DBA), for solving DisCOPs and the distributed scheduling problems. We experimentally show that DSA has a phase-transition or threshold behavior, in that its solution quality degenerates abruptly and dramatically when the degree of parallel executions of distributed agents increases beyond some critical value. We also consider the completeness and complexity of DBA for distributed graph coloring. We show that DBA is complete on coloring acyclic graphs, coloring an acyclic graph of n nodes in O(n^2) steps. However, on a cyclic graph, DBA may never terminate. To improve DBA's performance on coloring cyclic graphs, we propose two stochastic variations. Finally, we directly compare DSA and DBA for solving distributed graph coloring and distributed scheduling problems in sensor networks. The results show that DSA is superior to DBA when controlled properly, having better or competitive solution quality and significantly lower communication cost than DBA. Therefore, DSA is the algorithm of choice for our distributed scheduling problems and other distributed problems of similar properties.

[1]  Makoto Yokoo,et al.  The Phase Transition in Distributed Constraint Satisfaction Problems: Fist Results , 2000, CP.

[2]  David E. Culler,et al.  Mica: A Wireless Platform for Deeply Embedded Networks , 2002, IEEE Micro.

[3]  Tad Hogg,et al.  Phase Transitions and the Search Problem , 1996, Artif. Intell..

[4]  Deborah Estrin,et al.  Computing aggregates for monitoring wireless sensor networks , 2003, Proceedings of the First IEEE International Workshop on Sensor Network Protocols and Applications, 2003..

[5]  M. Yokoo,et al.  Distributed Breakout Algorithm for Solving Distributed Constraint Satisfaction Problems , 1996 .

[6]  Guoliang Xing.  A Performance-driven Framework for Customizing CSP Middleware Support , 2003 .

[7]  Peter C. Cheeseman,et al.  Where the Really Hard Problems Are , 1991, IJCAI.

[8]  Zhidong Deng,et al.  Distributed problem solving in sensor networks , 2002, AAMAS '02.

[9]  M. Takeda,et al.  Applications of MEMS to industrial inspection , 2001, Technical Digest. MEMS 2001. 14th IEEE International Conference on Micro Electro Mechanical Systems (Cat. No.01CH37090).

[10]  Paul Morris,et al.  The Breakout Method for Escaping from Local Minima , 1993, AAAI.

[11]  Weixiong Zhang,et al.  Distributed breakout revisited , 2002, AAAI/IAAI.

[12]  Makoto Yokoo,et al.  The Distributed Constraint Satisfaction Problem: Formalization and Algorithms , 1998, IEEE Trans. Knowl. Data Eng..

[13]  Gerard Tel,et al.  Introduction to Distributed Algorithms: Contents , 2000 .

[14]  Marko Fabiunke,et al.  Parallel Distributed Constraint Satisfaction , 1999, PDPTA.

[15]  Weixiong Zhang,et al.  An analysis and application of distributed constraint satisfaction and optimization algorithms in sensor networks , 2003, AAMAS '03.

[16]  Makoto Yokoo,et al.  Distributed Constraint Satisfaction: Foundations of Cooperation in Multi-agent Systems , 2000 .

[17]  H. Reichl,et al.  Overview and development trends in the field of MEMS packaging , 2001, Technical Digest. MEMS 2001. 14th IEEE International Conference on Micro Electro Mechanical Systems (Cat. No.01CH37090).

[18]  Stephen Fitzpatrick,et al.  An Experimental Assessment of a Stochastic, Anytime, Decentralized, Soft Colourer for Sparse Graphs , 2001, SAGA.

[19]  Judea Pearl,et al.  Evidential Reasoning Using Stochastic Simulation of Causal Models , 1987, Artif. Intell..

[20]  Weixiong Zhang,et al.  Distributed Stochastic Search for Constraint Satisfaction and Optimization: Parallelism, Phase Transitions and Performance , 2002 .

[21]  Makoto Yokoo,et al.  An asynchronous complete method for distributed constraint optimization , 2003, AAMAS '03.

[22]  相場 亮,et al.  Distributed Constraint Satisfaction : Foundations of Cooperation in Multi-Agent Systems, Mokoto Yokoo (著), Springer-Verlag, ISBN3540675965 , 2001 .

[23]  Athanassios Siapas,et al.  Criticality and Parallelism in Combinatorial Optimization , 1996, Science.

[24]  Stephen Fitzpatrick,et al.  Experiments on Dense Graphs with a Stochastic, Peer-to-Peer Colorer , 2002 .