Monitoring of Network Topology Dynamics
Abstract
We present software for deriving innovative metrics describing dominant parts of the internal structure of large networks. The algorithm is sufficiently fast for the network metrics to support real time monitoring of network dynamics. The network connections (connectivity matrix) are mathematically constructed by capturing the appropriate header parameters of selected internet/network traffic. Our metrics are in part derived from a network cluster decomposition that is based upon a physical model analogy for the network that is very rapidly evolved revealing cluster structures. Certain of the metrics consist in part of Renyi (generalized Shannon) entropy measures on the resulting network clusters and subclusters. This evolution depends upon the connectivity matrix and revels many qualitative features of the network.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 2004
- Accession Number
- ADA447339
Entities
People
- James L. Sidoran
- Joseph E. Johnson
- Rajesh Madamanchi
- Vladimir Gudkov
Organizations
- University of South Carolina