• DocumentCode
    1333855
  • Title

    Delay-Extraction-Based Waveform Relaxation Algorithm for Fast Transient Analysis of Power Distribution Networks

  • Author

    Roy, Sourajeet ; Dounavis, Anestis

  • Author_Institution
    Dept. of Electr. & Comput. Eng., Western Univ., London, ON, Canada
  • Volume
    2
  • Issue
    12
  • fYear
    2012
  • Firstpage
    2044
  • Lastpage
    2056
  • Abstract
    In this paper, a waveform relaxation algorithm is presented for the efficient transient analysis of large power distribution networks. The network is modeled as an orthogonal grid of transmission lines where a delay-extraction-based macromodel is used to represent each line segment in the time domain. Novel partitioning and iterative techniques are proposed for fast convergence and improved scalability of the proposed relaxation algorithm. The overall algorithm is highly parallelizable and exhibits good scaling with both the size of the circuit matrices involved and the number of CPUs available. Numerical examples are presented to illustrate the validity and efficiency of the proposed algorithm.
  • Keywords
    delays; integrated circuit modelling; iterative methods; transient analysis; transmission lines; CPU; circuit matrices; delay extraction based macromodel; delay extraction based waveform relaxation algorithm; fast transient analysis; iterative techniques; line segment; orthogonal grid; power distribution networks; transmission lines; Algorithm design and analysis; Capacitors; Convergence; Integrated circuit modeling; Mathematical model; Partitioning algorithms; Transient analysis; Convergence analysis; delay; power distribution networks (PDNs); signal integrity; simultaneous switching noise (SSN); transient simulation; transmission line; waveform relaxation (WR);
  • fLanguage
    English
  • Journal_Title
    Components, Packaging and Manufacturing Technology, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    2156-3950
  • Type

    jour

  • DOI
    10.1109/TCPMT.2012.2209118
  • Filename
    6353186