Parallel Combinatorial Pyramid Generation with Maximal Independent Directed Edge Sets (bibtex)
by Stefan Fiedler
Abstract:
The combinatorial pyramid is a conceptually simple and efficient representation of plane graphs that is well suited to model irregular graph pyramids. In this report, we show, through theoretical considerations and practical experiments, that maximal independent directed edge sets can be extended to select contraction and removal kernels consisting of independent darts in a combinatorial pyramid. In this way, the adaptation facilitates the parallel processing of combinatorial pyramid generation. Additionally, with the use of the edge preselection mechanism inherent to this method, it is possible to steer the generation process towards any desired set of survivors. While this report only uses a basic preselection to demonstrate its general effectiveness, the approach can be easily adapted by computing different survivor selection criteria.
Reference:
Parallel Combinatorial Pyramid Generation with Maximal Independent Directed Edge Sets (Stefan Fiedler), Technical report, PRIP, TU Wien, 2021.
Bibtex Entry:
@TechReport{TR149,
  author      = "Stefan Fiedler",
  title       = "Parallel Combinatorial Pyramid Generation with Maximal Independent Directed Edge Sets",
  institution = "PRIP, TU Wien",
  number      = "PRIP-TR-149",
  year        = "2021",
  url         = "https://www.prip.tuwien.ac.at/pripfiles/trs/tr149.pdf",
  abstract    = "The combinatorial pyramid is a conceptually simple and efficient representation of plane graphs that is well suited to model irregular graph pyramids. In this report, we show, through theoretical considerations and practical experiments, that maximal independent directed edge sets can be extended to select contraction and removal kernels consisting of independent darts in a combinatorial pyramid. In this way, the adaptation facilitates the parallel processing of combinatorial pyramid generation. Additionally, with the use of the edge preselection mechanism inherent to this method, it is possible to steer the generation process towards any desired set of survivors. While this report only uses a basic preselection to demonstrate its general effectiveness, the approach can be easily adapted by computing different survivor selection criteria.",
}
Powered by bibtexbrowser