Congestion Control Schemes for TCP/IP Networks
By
Darius Buntinas
Abstract
This paper describes six methods for controlling congestion for TCP connections. The first four, Slow Start algorithm [5], Tri-S [8], DUAL [7], and TCP Vegas [1] treat the network as a black box, in that the only way to detect congestion is through packet loss and changes in round trip time, or throughput. The last two, Random Early Detection [4] and Explicit Congestion Notification [3] depend on the gateways to provide indications of congestion.
Table of Contents
- Introduction
- Source based schemes
- Slow Start
- Slow Start and Search: Tri-S
- DUAL
- TCP Vegas
- Gateway based schemes
- Random Early Detection
- Explicit Congestion Notification
- Conclusion
- Works Cited
View the complete report online
Shift-click to download the report in
Adobe Acrobat format (48,317 Bytes) |
Other reports in this series
Complete list of student reports on Recent Advances in Networking
Back to Raj Jain's Home Page