A Graph Decomposition Technique Based on a High-Density Clustering Model on Graphs.

Abstract

Complex design problems are characterized by a multitude of competing requirements. System designers frequently find the scope of the problem beyond their conceptual abilities, and attempt to cope with this difficulty by decomposing the original design problem into smaller, more manageable sub-problems. Functional requirements form a key interface between the users of a system and its designers. In this research effort, a systematic approach has been proposed for the decomposition of the overall set of functional requirements into sub-problems to form a design structure that will exhibit the key characteristics of good design: strong coupling within sub-problems, and weak coupling between them. A scan of the graph decomposition algorithms review that none of the existing techniques is particularly well suited for use in Systematic Design Methodology. In this report, a clustering model on a graph is defined, using the concept of natural or high density clusters which are densely-connected subgraphs separated by relatively few links. A graph decomposition techniques based on this high-density clustering model is developed for identifying the best natural partition for a given graph.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 01, 1980
Accession Number
ADA090348

Entities

People

  • M. Anthony Wong

Tags

Communities of Interest

  • Energy and Power Technologies
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Algorithms
  • Clustering
  • Computers
  • Couplings
  • Decomposition
  • Dynamic Programming
  • Electrical Circuits
  • High Density
  • Information Systems
  • Life Cycles
  • Literature
  • Low Density
  • New York
  • Operations Research
  • Schools
  • Software Development
  • Systems Engineering

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Graph Algorithms and Convex Optimization.
  • Systems Analysis and Design