Gossiping Differential Evolution: A Decentralized Heuristic for Function Optimization in P2P Networks

P2P-based optimization has recently gained interest among distributed function optimization scientists. Several well-known optimization heuristics have been recently re-designed to exploit the peculiarity of such a distributed environment. The final goal is to perform high quality function optimization by means of inexpensive, fully decentralized machines, which may either be purposely organized in a P2P network, or voluntarily join a running P2P optimization task. In this paper we present the GoDE algorithm (Gossip-based Differential Evolution), which obtains remarkable results on several test functions. We describe in detail the algorithm design and the epidemic mechanism that greatly improves the performance. Experimental results in a simulated environment show how GoDE adapts to network scale and how the epidemic communication protocol can make the algorithm achieve good results even in presence of a high churn rate.

[1]  Erick Cantú-Paz,et al.  Migration Policies, Selection Pressure, and Parallel Evolutionary Algorithms , 2001, J. Heuristics.

[2]  Juan Julián Merelo Guervós,et al.  Resilience to churn of a peer-to-peer evolutionary algorithm , 2008, Int. J. High Perform. Syst. Archit..

[3]  José Ignacio Hidalgo,et al.  Is the island model fault tolerant? , 2007, GECCO '07.

[4]  Mauro Brunato,et al.  GOSH! Gossiping Optimization Search Heuristics , 2007 .

[5]  A. E. Eiben,et al.  Peer-to-peer evolutionary algorithms with adaptive autonomous selection , 2007, GECCO '07.

[6]  Márk Jelasity,et al.  Peer-to-Peer Optimization in Large Unreliable Networks with Branch-and-Bound and Particle Swarms , 2009, EvoWorkshops.

[7]  Andrew Lewis,et al.  Dynamic search initialisation strategies for multi-objective optimisation in peer-to-peer networks , 2009, 2009 IEEE Congress on Evolutionary Computation.

[8]  El-Ghazali Talbi,et al.  Parallel hybrid multi-objective island model in peer-to-peer environment , 2005, 19th IEEE International Parallel and Distributed Processing Symposium.

[9]  B. Pittel.  On spreading a rumor , 1987 .

[10]  Rainer Storn,et al.  Differential Evolution – A Simple and Efficient Heuristic for global Optimization over Continuous Spaces , 1997, J. Glob. Optim..

[11]  Anne-Marie Kermarrec,et al.  Gossip-based peer sampling , 2007, TOCS.

[12]  Andrew Lewis,et al.  Decentralised distributed multiple objective particle swarm optimisation using peer to peer networks , 2008, 2008 IEEE Congress on Evolutionary Computation (IEEE World Congress on Computational Intelligence).

[13]  Márk Jelasity,et al.  Distributed hyper-heuristics for real parameter optimization , 2009, GECCO.

[14]  Juan Julián Merelo Guervós,et al.  Exploring population structures for locally concurrent and massively parallel Evolutionary Algorithms , 2008, 2008 IEEE Congress on Evolutionary Computation (IEEE World Congress on Computational Intelligence).