3rd International ICST Conference on Scalable Information Systems

Research Article

Optimization of Dynamic Data Types in Embedded Systems using DEVS/SOA-based Modeling and Simulation

Download478 downloads
  • @INPROCEEDINGS{10.4108/ICST.INFOSCALE2008.3504,
        author={Jos\^{e} L. Risco-Mart\^{\i}n and Saurabh Mittal and David Atienza and J. Ignacio Hidalgo and Juan Lanchares},
        title={Optimization of Dynamic Data Types in Embedded Systems using DEVS/SOA-based Modeling and Simulation},
        proceedings={3rd International ICST Conference on Scalable Information Systems},
        publisher={ICST},
        proceedings_a={INFOSCALE},
        year={2010},
        month={5},
        keywords={Embedded Systems Design Evolutionary Computation Discrete Event System Specification Service Oriented Architecture DEVS/SOA.},
        doi={10.4108/ICST.INFOSCALE2008.3504}
    }
    
  • José L. Risco-Martín
    Saurabh Mittal
    David Atienza
    J. Ignacio Hidalgo
    Juan Lanchares
    Year: 2010
    Optimization of Dynamic Data Types in Embedded Systems using DEVS/SOA-based Modeling and Simulation
    INFOSCALE
    ICST
    DOI: 10.4108/ICST.INFOSCALE2008.3504
José L. Risco-Martín1,*, Saurabh Mittal2,*, David Atienza1,*, J. Ignacio Hidalgo1,*, Juan Lanchares1,*
  • 1: Department of Computer Architecture and Automation, Complutense University of Madrid, C/Prof. José García Santesmases, s/n 28040 Madrid (Spain)
  • 2: Arizona Center for Integrative Modeling and Simulation, University of Arizona, ECE Building, 1230 E Speedway Tucson AZ 85721
*Contact email: jlrisco@dacya.ucm.es, saurabh@ece.arizona.edu, datienza@dacya.ucm.es, hidalgo@dacya.ucm.es, julandan@dacya.ucm.es

Abstract

New multimedia embedded applications are increasingly dynamic, and rely on Dynamically-allocated Data Types (DDTs) to store their data. The optimization of DDTs for each target embedded system is a time-consuming process due to the large searching space of possible DDTs implementations. This results in the minimization of embedded design variables (memory accesses, power consumption and memory usage). Till date some effective heuristic algorithms have been developed in order to solve this problem, however unreported, as the problem is NP-complete and cannot be fully explored. In these cases the use of parallel processing can be very useful because it allows not only to explore more solutions spending the same time but also to implement new algorithms. This research work provides a methodology to use Discrete Event Systems Specification (DEVS) to implement a parallel evolutionary algorithm within a Service Oriented Architecture (SOA), where parallelism improves the solutions found by the corresponding sequential algorithm. This algorithm provides better results when compared with other previously proposed procedures. In order to implement the parallelism the DEVS/SOA framework in utilized. Experimental results show how a novel parallel multi-objective genetic algorithm, which combines NSGA-II and SPEA2, allows designers to reach a larger number of solutions than previous approximations. This research also establishes DEVS/SOA as a platform for conducting complex distributed simulation experiments.