Abstract
This paper proposes a parallel model of the self-organizing map (SOM) neural network applied to the Euclidean traveling salesman problem (TSP) and intended for implementation on the graphics processing unit (GPU) platform. The plane is partitioned into an appropriate number of cellular units, that are each responsible of a certain part of the data and network. The advantage of the parallel algorithm is that it is decentralized and based on data decomposition, rather than based on data duplication, or mixed sequential/parallel solving, as often with GPU implementation of optimization metaheuristics. The processing units and the required memory are with linear increasing relationship to the problem size, which makes the model able to deal with very large scale problems in a massively parallel way. The approach is applied to Euclidean TSPLIB problems and National TSPs with up to 33708 cities on both GPU and CPU, and these two types of implementation are compared and discussed.
Access provided by Autonomous University of Puebla. Download to read the full chapter text
Chapter PDF
Similar content being viewed by others
Keywords
References
Papadimitriou, C.H.: The euclidean travelling salesman problem is np-complete. Theoretical Computer Science 4, 237–244 (1977)
Kohonen, T.: Self-organizing maps, vol. 30. Springer (2001)
Angeniol, B., de La Croix Vaubois, G., Le Texier, J.Y.: Self-organizing feature maps and the travelling salesman problem. Neural Networks 1, 289–293 (1988)
Cochrane, E., Beasley, J.: The co-adaptive neural network approach to the euclidean travelling salesman problem. Neural Networks 16, 1499–1525 (2003)
Créput, J.C., Koukam, A.: A memetic neural network for the euclidean traveling salesman problem. Neurocomputing 72, 1250–1264 (2009)
McConnell, S., Sturgeon, R., Henry, G., Mayne, A., Hurley, R.: Scalability of self-organizing maps on a gpu cluster using opencl and cuda. Journal of Physics: Conference Series 341, 012018 (2012)
Yoshimi, M., Kuhara, T., Nishimoto, K., Miki, M., Hiroyasu, T.: Visualization of pareto solutions by spherical self-organizing map and its acceleration on a gpu. Journal of Software Engineering and Applications 5 (2012)
Bentley, J.L., Weide, B.W., Yao, A.C.: Optimal expected-time algorithms for closest point problems. ACM Transactions on Mathematical Software (TOMS) 6, 563–580 (1980)
Reinelt, G.: Tsplib a traveling salesman problem library. ORSA Journal on Computing 3, 376–384 (1991)
NVIDIA: CUDA C Programming Guide 4.2, CURAND Library, Profiler User’s Guide (2012), http://docs.nvidia.com/cuda
Sanders, J., Kandrot, E.: CUDA by example: an introduction to general-purpose GPU programming. Addison-Wesley Professional (2010)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Wang, H., Zhang, N., Créput, JC. (2013). A Massive Parallel Cellular GPU Implementation of Neural Network to Large Scale Euclidean TSP. In: Castro, F., Gelbukh, A., González, M. (eds) Advances in Soft Computing and Its Applications. MICAI 2013. Lecture Notes in Computer Science(), vol 8266. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-45111-9_10
Download citation
DOI: https://doi.org/10.1007/978-3-642-45111-9_10
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-45110-2
Online ISBN: 978-3-642-45111-9
eBook Packages: Computer ScienceComputer Science (R0)