Geometry of Dynamic Large Networks: A Scaling and Renormalization Group Approach

Abstract

The research conducted during the period of performance of this grant concerned investigation of large-scale geometry of prototypical and real-life large graphs, and their analytical/numeric estimation of their core congestion. The research included demonstration of delta-hyperbolicity in a large variety of communication and social networks via scaling of the curvature plots through renormalization, detection of graph bottlenecks using a novel techniques based on the Dirichlet eigenvectors of the (normalized) graph Laplacian, proof of the fact that remetrization (by factors away from zero and infinity) does not eliminate O(N sq) core congestion in delta-hyperbolic graphs, derivation of an analytical expression for the average nodal congestion involving the sum of the inverses of the non-zero eigenvalues of the graph Laplacian when random walk routing replaces geodesic routing, and finally determination of the threshold for bootstrap percolation in periodic trees as an upper bounding mechanism for the threshold for more general graphs.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 11, 2013
Accession Number
ADA595914

Entities

People

  • Iraj Saniee
  • Onuttom Narayan

Tags

DTIC Thesaurus Topics

  • Abstracts
  • Air Force
  • Air Force Research Laboratories
  • Boundaries
  • Clustering
  • Communication Networks
  • Computations
  • Congestion
  • Curvature
  • Detection
  • Eigenvalues
  • Equations
  • Geometry
  • Networks
  • Probability
  • Random Walk
  • Social Networks

Readers

  • Computer Networking
  • Graph Algorithms and Convex Optimization.
  • Statistical inference.