m (Scipediacontent moved page Draft Content 468042286 to Dukkipati et al 2005a) |
|||
| Line 2: | Line 2: | ||
== Abstract == | == Abstract == | ||
| − | Most congestion control algorithms try to emulate processor sharing (PS) by giving each competing flow an equal share of a bottleneck link. This approach leads to fairness, and prevents long flows from hogging resources. For example, if a set of flows with the same round trip time share a bottleneck link, | + | Most congestion control algorithms try to emulate processor sharing (PS) by giving each competing flow an equal share of a bottleneck link. This approach leads to fairness, and prevents long flows from hogging resources. For example, if a set of flows with the same round trip time share a bottleneck link, TCPâs congestion control mechanism tries to achieve PS; so do most of the proposed alternatives, such as eXplicit Control Protocol (XCP). But although they emulate PS well in a static scenario when all flows are long-lived, they do not come close to PS when new flows arrive randomly and have a finite amount of data to send, as is the case in todayâs Internet. Typically, flows take an order of magnitude longer to complete with TCP or XCP than with PS, suggesting large room for improvement. And so in this paper, we explore how a new congestion control algorithm â Rate Control Protocol (RCP) â comes much closer to emulating PS over a broad range of operating conditions. In RCP, a router assigns a single rate to all flows that pass through it. The router does not keep flow-state, and does no per-packet calculations. Yet we are able to show that under a wide range of traffic characteristics and network conditions, RCPâs performance is very close to ideal processor sharing. |
| − | + | ||
| − | + | ||
| − | + | ||
| − | + | ||
| − | + | ||
| Line 17: | Line 12: | ||
* [http://yuba.stanford.edu/~nanditad/RCP-IWQoS.pdf http://yuba.stanford.edu/~nanditad/RCP-IWQoS.pdf] | * [http://yuba.stanford.edu/~nanditad/RCP-IWQoS.pdf http://yuba.stanford.edu/~nanditad/RCP-IWQoS.pdf] | ||
| + | |||
| + | * [http://link.springer.com/content/pdf/10.1007/11499169_22 http://link.springer.com/content/pdf/10.1007/11499169_22], | ||
| + | : [http://dx.doi.org/10.1007/11499169_22 http://dx.doi.org/10.1007/11499169_22] | ||
| + | |||
| + | * [http://yuba.stanford.edu/rcp/RCP-IWQoS.pdf http://yuba.stanford.edu/rcp/RCP-IWQoS.pdf], | ||
| + | : [https://link.springer.com/chapter/10.1007%2F11499169_22 https://link.springer.com/chapter/10.1007%2F11499169_22], | ||
| + | : [http://nms.lcs.mit.edu/6829-papers/RCP1.pdf http://nms.lcs.mit.edu/6829-papers/RCP1.pdf], | ||
| + | : [https://core.ac.uk/display/24572077 https://core.ac.uk/display/24572077], | ||
| + | : [https://dblp.uni-trier.de/db/conf/iwqos/iwqos2005.html#DukkipatiKRM05 https://dblp.uni-trier.de/db/conf/iwqos/iwqos2005.html#DukkipatiKRM05], | ||
| + | : [https://dl.acm.org/citation.cfm?id=2103204 https://dl.acm.org/citation.cfm?id=2103204], | ||
| + | : [https://rd.springer.com/chapter/10.1007/11499169_22 https://rd.springer.com/chapter/10.1007/11499169_22], | ||
| + | : [https://academic.microsoft.com/#/detail/1579768150 https://academic.microsoft.com/#/detail/1579768150] | ||
Latest revision as of 16:33, 21 January 2021
Abstract
Most congestion control algorithms try to emulate processor sharing (PS) by giving each competing flow an equal share of a bottleneck link. This approach leads to fairness, and prevents long flows from hogging resources. For example, if a set of flows with the same round trip time share a bottleneck link, TCPâs congestion control mechanism tries to achieve PS; so do most of the proposed alternatives, such as eXplicit Control Protocol (XCP). But although they emulate PS well in a static scenario when all flows are long-lived, they do not come close to PS when new flows arrive randomly and have a finite amount of data to send, as is the case in todayâs Internet. Typically, flows take an order of magnitude longer to complete with TCP or XCP than with PS, suggesting large room for improvement. And so in this paper, we explore how a new congestion control algorithm â Rate Control Protocol (RCP) â comes much closer to emulating PS over a broad range of operating conditions. In RCP, a router assigns a single rate to all flows that pass through it. The router does not keep flow-state, and does no per-packet calculations. Yet we are able to show that under a wide range of traffic characteristics and network conditions, RCPâs performance is very close to ideal processor sharing.
Original document
The different versions of the original document can be found in:
- https://link.springer.com/chapter/10.1007%2F11499169_22,
- http://nms.lcs.mit.edu/6829-papers/RCP1.pdf,
- https://core.ac.uk/display/24572077,
- https://dblp.uni-trier.de/db/conf/iwqos/iwqos2005.html#DukkipatiKRM05,
- https://dl.acm.org/citation.cfm?id=2103204,
- https://rd.springer.com/chapter/10.1007/11499169_22,
- https://academic.microsoft.com/#/detail/1579768150
Document information
Published on 01/01/2005
Volume 2005, 2005
DOI: 10.1007/11499169_22
Licence: CC BY-NC-SA license
Share this document
Keywords
claim authorship
Are you one of the authors of this document?