Citation: K. Mills, J. Filliben, D. Cho, E. Schwartz and D. Genin, Study of Proposed Internet Congestion Control Algorithms, NIST Special Publication 500-282, May 2010, 534 pages.
This study describes a coherent set of modeling and analysis methods to investigate the behavior of large distributed systems. The methods are applied to compare several proposed Internet congestion control mechanisms operating under a wide range of conditions. The study provides insights and recommendations regarding various congestion control algorithms. The study also evaluates the modeling and analysis methods adopted. The contents of the study might be of interest to those who design and study Internet congestion control algorithms. The study should also be of interest to designers and researchers seeking methods to model and analyze the behavior of large distributed systems.
The draft report may be downloaded chapter-by-chapter by selecting each chapter name below. The entire (23 MB) document may be downloaded by clicking here.
Cover and Title Pages
Acknowledgements & Executive SummaryTable of Contents
List of Figures, Tables and Acronyms
Chapter 1 - Introduction
Chapter 2 - Method and Related WorkChapter 3 - Description of MesoNet
Chapter 4 - Sensitivity Analysis of MesoNet
Chapter 5 - Modeling Congestion Control Algorithms
Chapter 6 - Comparing Congestion Control Regimes in a Large, Fast NetworkChapter 7 - Comparing Congestion Control Regimes in a Scaled-Down Network
Chapter 8 - Comparing Congestion Control Regimes in a Heterogeneous NetworkChapter 9 - Comparing Congestion Control Regimes in a Large, Fast, Heterogeneous NetworkChapter 10 - ConclusionsChapter 11 - Bibliography
Appendix A - Understanding, Improving and Applying Fluid-Flow Models
Appendix B - Computational Requirements: MesoNet vs. Hybrid Model
Appendix C - Supplementary Sensitivity Analysis ResultsAppendix D - 10-Step Graphical Analysis Technique