Advertisement

A Network Service for DSP Multicomputers

  • Juan A. Rico-Gallego
  • Jesús M. Álvarez-Llorente
  • Juan C. Díaz-Martín
  • Francisco J. Perogil-Duque
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5022)

Abstract

Programmers of embedded digital signal processors often have to deal with the devices of the platform or with low level hardware abstraction layers in order to reach the better performance from a given algorithm. This complexity increases when the application is distributed on multicomputers such as those by SundanceTM, Hunt EngineeringTM, etc. These machines are loosely coupled networks based on carrier boards hosting modules of digital signal processors and FPGAs. This paper describes the design and implementation of a network layer library for these platforms. We show how, without a significative lost of performance, it improves the portability of target applications by avoiding the hardware communication complexities. More important yet, it broadens the spectrum of network applications and middlewares that these powerful platforms can support.

Keywords

Digital signal processing distributed embedded systems DSP multicomputers 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
  2. 2.
  3. 3.
    Rico-Gallego, J.A., Díaz-Martín,, Álvarez Llorente, J.C., Jesús, M.: An MPI Implementation for Distributed Signal Processing. In: Di Martino, B., Kranzlmüller, D., Dongarra, J. (eds.) EuroPVM/MPI 2005. LNCS, vol. 3666, pp. 475–482. Springer, Heidelberg (2005)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2008

Authors and Affiliations

  • Juan A. Rico-Gallego
    • 1
  • Jesús M. Álvarez-Llorente
    • 1
  • Juan C. Díaz-Martín
    • 2
  • Francisco J. Perogil-Duque
    • 2
  1. 1.Department of Engineering of Computer Systems and Telematics  
  2. 2.Department of Computer and Communication TechnologyUniversity of ExtremaduraCáceresSpain

Personalised recommendations