Abstract
Multi-agent systems allow the modelling of complex, heterogeneous, and distributed systems in a realistic way. MARL-Ped is a multi-agent system tool, based on the MPI standard, for the simulation of different scenarios of pedestrians who autonomously learn the best behavior by Reinforcement Learning. MARL-Ped uses one MPI process for each agent by design, with a fixed fine-grain granularity. This requirement limits the performance of the simulations for a restricted number of processors that is lesser than the number of agents. On the other hand, Hitmap is a library to ease the programming of parallel applications based on distributed arrays. It includes abstractions for the automatic partition and mapping of arrays at runtime with arbitrary granularity, as well as functionalities to build flexible communication patterns that transparently adapt to the data partitions.
In this work, we present the methodology and techniques of granularity selection in Hitmap, applied to the simulations of agent systems. As a first approximation, we use the MARL-Ped multi-agent pedestrian simulation software as a case of study for intra-node cases. Hitmap allows to transparently map agents to processes, reducing oversubscription and intra-node communication overheads. The evaluation results show significant advantages when using Hitmap, increasing the flexibility, performance, and agent-number scalability for a fixed number of processing elements, allowing a better exploitation of isolated nodes.
A. Gonzalez-Escribano—This work has been funded by Spanish MINECO and the EU ERDF program under grants HomProg-HetSys TIN2014-58876-P, TIN2015-66972-C5-5-R, CAPAP-H5 network TIN2014-53522-REDT, and COST Program Action IC1305: Network for Sustainable Ultrascale Computing (NESUS).
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Bharambe, A., Pang, J., Seshan, S.: Colyseus: a distributed architecture for online multiplayer games. In: NSDI 2006: Proceedings of the 3rd conference on Networked Systems Design and Implementation, p. 12. USENIX Association, Berkeley, CA, USA (2006)
Blue, V.J., Adler, J.L.: Cellular automata microsimulation for modeling bi-directional pedestrian walkways. Transp. Res. Part B: Methodological 35(3), 293–312 (2001)
Chamberlain, B., Callahan, D., Zima, H.: Parallel programmability and the chapel language. Int. J. High Perform. Comput. Appl. 21(3), 291–312 (2007)
Chen, Y., Cui, X., Mei, H.: Parray: A unifying array representation for heterogeneous parallelism. SIGPLAN Not. 47(8), 171–180 (2012)
Fraguela, B.B., Bikshandi, G., Guo, J., GarzaráN, M.J., Padua, D., Von Praun, C.: Optimization techniques for efficient hta programs. Parallel Comput. 38(9), 465–484 (2012)
Fresno, J., Gonzalez-Escribano, A., Llanos, D.: Blending extensibility and performance in dense and sparse parallel data management. IEEE Trans. Parallel Distrib. Syst. 25(10), 2509–2519 (2014)
Gonzalez-Escribano, A., Torres, Y., Fresno, J., Llanos, D.: An extensible system for multilevel automatic data partition and mapping. IEEE Trans. Parallel Distrib. Syst. 25(5), 1145–1154 (2014)
Gonzalez-Escribano, A., Llanos, D.R.: Trasgo: a nested-parallel programming system. J. Supercomput. 58(2), 226–234 (2011)
Helbing, D., Molnár, P.: Social force model for pedestrian dynamics. Phys. Rev. E 51, 4282–4286 (1995)
Hughes, R.L.: The flow of human crowds. Annu. Rev. Fluid Mech. 35, 169–182 (2003)
Mallón, D.A., Gómez, A., Mouriño, J.C., Taboada, G.L., Teijeiro, C., Touriño, J., Fraguela, B.B., Doallo, R., Wibecan, B.: Upc performance evaluation on a multicore system. In: Proceedings of the Third Conference on Partitioned Global Address Space Programing Models, pp. 9: 1–9: 7. PGAS 2009, NY, USA. ACM, New York (2009)
Martinez-Gil, F., Lozano, M., Fernández, F.: Multi-agent reinforcement learning for simulating pedestrian navigation. In: Vrancx, P., Knudson, M., Grześ, M. (eds.) ALA 2011. LNCS (LNAI), vol. 7113, pp. 54–69. Springer, Heidelberg (2012). doi:10.1007/978-3-642-28499-1_4
Martinez-Gil, F., Lozano, M., Fernández, F.: MARL-ped: A multi-agent reinforcement learning based framework to simulate pedestrian groups. Simul. Model. Pract. Theor. 47, 259–275 (2014)
Reynolds, C.: Steering behaviors for autonomous characters. In: Game Developers Conference, pp. 763–782. Miller Freeman Game Group, San Francisco, California (1999)
Sutton, R.S., Barto, A.G.: Reinforcement Learning: An Introduction. MIT Press, Cambridge (1998)
Vigueras, G., Orduña, J.M., Lozano, M.: A read-copy update based parallel server for distributed crowd simulations. J. Supercomput. 64(1), 156–166 (2013)
Wooldridge, M.: Multi-Agent Systems. Intelligent Agents. MIT Press, Cambridge (2013)
Wooldridge, M., Jennings, N.: Intelligent agents: theory and practice. Knowl. Eng. Rev. 10, 115–152 (1995)
Yilmaz, E., Isler, V., Cetin, Y.Y.: The virtual marathon: Parallel computing supports crowd simulations. IEEE Comput. Graph. Appl. 29(4), 26–33 (2009)
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing AG
About this paper
Cite this paper
Rodriguez-Gutiez, E., Martinez-Gil, F., Orduña, J.M., Gonzalez-Escribano, A. (2016). MARL-Ped+Hitmap: Towards Improving Agent-Based Simulations with Distributed Arrays. In: Carretero, J., et al. Algorithms and Architectures for Parallel Processing. ICA3PP 2016. Lecture Notes in Computer Science(), vol 10049. Springer, Cham. https://doi.org/10.1007/978-3-319-49956-7_17
Download citation
DOI: https://doi.org/10.1007/978-3-319-49956-7_17
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-49955-0
Online ISBN: 978-3-319-49956-7
eBook Packages: Computer ScienceComputer Science (R0)