Performance Evaluation in Multi-Rate, Multi-Hop Communication Networks with Adaptive Routing

Abstract

Accurate performance evaluation has always been an important issue in network design and analysis. Discrete event simulation has been known to be accurate, but very time-consuming. Thus, analytical methods/approximation is necessary for real-time estimation, large-scale network optimization, and sensitivity analysis. In a circuit-switched loss network, a particular performance metric of interest is the end-to-end blocking probability. Various analytical approaches and approximation schemes have been suggested for this problem, and among them, the fixed-point method, or reduced load method, has received much attention. However, most of these schemes considered either only single traffic rate situations or multi-rate traffic under fixed routing. The authors have developed an approximation scheme to estimate end-to-end blocking probability in a multi-rate, multi-hop network with an adaptive routing scheme. The approximation results are compared with that of discrete event simulation. An example of application also is provided in which the proposed scheme is linked to the optimization tool CONSOL-OPTCAD to get network design trade-offs. The method described here is readily applicable to the accurate performance evaluation of military networks, which are often large and hybrid.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 1998
Accession Number
ADA440948

Entities

People

  • Archan Misra
  • John Baras
  • Mingyan Liu

Organizations

  • University of Maryland

Tags

Communities of Interest

  • Space

DTIC Thesaurus Topics

  • Algorithms
  • Bandwidth
  • Communication Networks
  • Computational Complexity
  • Department Of Defense
  • Electrical Engineering
  • Engineering
  • Markov Processes
  • Maryland
  • Military Research
  • Networks
  • Optimization
  • Probability
  • Simulations
  • Test And Evaluation
  • Topology
  • Universities

Fields of Study

  • Computer science

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Approximation Theory.
  • Computer Networking