• DocumentCode
    108527
  • Title

    An Effective Modeling Framework for the Analysis of Interconnects Subject to Line-Edge Roughness

  • Author

    Manfredi, Paolo ; Vande Ginste, Dries ; De Zutter, Daniel

  • Author_Institution
    Dept. of Inf. Technol., iMinds, Electromagn. Group, Ghent Univ., Ghent, Belgium
  • Volume
    25
  • Issue
    8
  • fYear
    2015
  • fDate
    Aug. 2015
  • Firstpage
    502
  • Lastpage
    504
  • Abstract
    This letter proposes a complete and efficient simulation framework to assess the effects of line-edge roughness appearing in on-chip lines. The modeling approach consists of three steps. First, a stochastic macromodel is created for the per-unit-length RLGC parameters of the line. Secondly, random conductor edge profiles are generated using randomized splines. These are combined with the stochastic macromodel to readily provide place-dependent RLGC parameters. Finally, the resulting nonuniform transmission line is analyzed by means of a fast and accurate perturbation technique. To validate the proposed approach, a statistical analysis of the response of a coupled inverted embedded microstrip line is carried out for different roughness parameters.
  • Keywords
    conductors (electric); integrated circuit interconnections; microstrip lines; perturbation techniques; splines (mathematics); statistical analysis; transmission lines; coupled inverted embedded microstrip line; line-edge roughness; nonuniform transmission line; on-chip interconnects; on-chip lines; per-unit-length RLGC parameters; perturbation technique; place-dependent RLGC parameters; random conductor edge profiles; randomized splines; statistical analysis; stochastic macromodel; Capacitance; Conductors; Correlation; Splines (mathematics); Standards; Stochastic processes; System-on-chip; Line-edge roughness (LER); nonuniform transmission lines; on-chip interconnects; perturbation methods;
  • fLanguage
    English
  • Journal_Title
    Microwave and Wireless Components Letters, IEEE
  • Publisher
    ieee
  • ISSN
    1531-1309
  • Type

    jour

  • DOI
    10.1109/LMWC.2015.2440776
  • Filename
    7130684