4th International ICST Conference on Performance Evaluation Methodologies and Tools

Research Article

Parallelising a simulator for the analysis of electromagnetic radiation using MUMPS library

Download641 downloads
  • @INPROCEEDINGS{10.4108/ICST.VALUETOOLS2009.7456,
        author={R. Rico L\^{o}pez and V. Escuder Caba\`{o}as and R. Dur\^{a}n D\^{\i}az and L.E.  Garc\^{\i}a-Castillo and I.  G\^{o}mez-Revuelto and J.A. Acebr\^{o}n},
        title={Parallelising a simulator for the analysis of electromagnetic radiation using MUMPS library},
        proceedings={4th International ICST Conference on Performance Evaluation Methodologies and Tools},
        publisher={ICST},
        proceedings_a={VALUETOOLS},
        year={2010},
        month={5},
        keywords={Parallel Computing MPI Sparse Direct Solvers MUMPS library},
        doi={10.4108/ICST.VALUETOOLS2009.7456}
    }
    
  • R. Rico López
    V. Escuder Cabañas
    R. Durán Díaz
    L.E. García-Castillo
    I. Gómez-Revuelto
    J.A. Acebrón
    Year: 2010
    Parallelising a simulator for the analysis of electromagnetic radiation using MUMPS library
    VALUETOOLS
    ICST
    DOI: 10.4108/ICST.VALUETOOLS2009.7456
R. Rico López1,*, V. Escuder Cabañas1, R. Durán Díaz1, L.E. García-Castillo2, I. Gómez-Revuelto3, J.A. Acebrón4
  • 1: Dep. de Automática Univ. de Alcalá Alcalá de Henares, Spain
  • 2: Dep. de Teoría de la Señal y Comunicaciones Univ. Carlos III Madrid, Spain
  • 3: Dep. de Ingeniería Audiovisual y Comunicaciones Univ. Politécnica de Madrid Madrid, Spain
  • 4: Center for Mathematics and its Applications Lisbon, Portugal
*Contact email: rafael.rico@uah.es

Abstract

The practical experience of parallelising a simulator of general scattering and radiation electromagnetic problems is presented. The simulator stems from an existing sequential simulator in the frequency domain and can be fruitfully used in applications such as the test of coverage of a wireless network, analysis of complex structures, and so on. After the analysis of a test case, two steps were carried out: firstly, a "hand-crafted" code parallelisation was developed within the kernel of the simulator. Secondly, the sequential library, used in the existing simulator, was replaced by the parallel MUMPS library in order to solve the associated linear algebra problem in parallel. For factorising the matrix with MUMPS, two different ordering methods have been considered.