Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Civil-Comp Proceedings
ISSN 1759-3433
CCP: 80
PROCEEDINGS OF THE FOURTH INTERNATIONAL CONFERENCE ON ENGINEERING COMPUTATIONAL TECHNOLOGY
Edited by: B.H.V. Topping and C.A. Mota Soares
Paper 145

A Flexible and Extendible Compiler for Complex Systems Simulation by means of a Cellular Automata Network

C.R. Calidonna and M. Mango Furnari

Department of Computer Science, Institute of Cybernetics "E. Caianiello", National Research Council (C.N.R.), Pozzuoli, Italy

Full Bibliographic Reference for this paper
C.R. Calidonna, M. Mango Furnari, "A Flexible and Extendible Compiler for Complex Systems Simulation by means of a Cellular Automata Network", in B.H.V. Topping, C.A. Mota Soares, (Editors), "Proceedings of the Fourth International Conference on Engineering Computational Technology", Civil-Comp Press, Stirlingshire, UK, Paper 145, 2004. doi:10.4203/ccp.80.145
Keywords: cellular automata, parallel language, simulation tool.

Summary
The Cellular Automata paradigm is an assessed simulation and programming model for complex phenomena. For this kind of approach, that shows an intrinsic data parallelism, performances assume a crucial role. The efficient exploitation of intrinsic parallelism is a must for CA based applications.

In the past several different modified CA based model where proposed as the classic model presents limitations in simulating real phenomena see Worsh [4].

Our Cellular Automata Network (CAN) model, see Calidonna [1], as an extension of the CA classic model, allows to simulate micro and macroscopic complex phenomena, that can be composed by several component of different nature and/or characterisation.

The whole complex system, decomposed in components together with their couplings and interactions, can be represented in terms of graph where nodes are the components, expressed in terms of CA nodes, and arcs are the coupling relations.

From an execution point of view several CA could evolve according to coupled relations; and when there is a partial coupling relations between nodes they can evolve concurrently.

All this implies two kind of parallelism sources management: data parallelism for each node together, when available, and control parallelism when nodes can evolve at the same time.

The CAN language, with opportune primitive and operators is available for writing CAN application; furthermore in order to exploit efficiently this opportunities see Calidonna [2] tools and/or systems must be designed.

In a first stage the PECANS [3] environment was proposed stressing particularly modelling activity. After an experimentation in which some techniques were studied see Calidonna [2] we decided to design a compiler in order to store and capture some specific language features and mechanism useful to code optimisation and parallelism opportunities exploitation.

In this paper we deal with the designed CAN compiler system description, according a modular and extendible vision.

The choice in individuating modules, especially in the run-time system allow to apply and implementing novel parallelism exploitation techniques without affecting the other modules.

References
1
C.R. Calidonna, C. Di Napoli, M. Mango Furnari, A. Imparato, "Experimenting with Cellular Automata Programming", Developments in Computational Mechanics with High Performance Computing, Ed. B.H.V. Topping, Civil-Comp Press, 175-181, 1999. doi:10.4203/ccp.57.9.1
2
C.R. Calidonna, C. Di Napoli, M. Mango Furnari, "Parallelism Management in Cellular Automata Networks", Computational Engineering using Metaphors from Nature, Ed. B.H.V. Topping, Civil-Comp Press, 199 - 205, 2000.
3
L. Carotenuto, F. Mele, M. Mango Furnari, R. Napolitano, "PECANS: a parallel environment for CA modeling", Complex Systems, 10, 23-41, 1996.
4
T. Worsch, "Programming environment for Cellular Automata", Technical Report 37/96, Universitat Karlshruhe, Fakultat fur Informatik, November 1996.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £95 +P&P)