Congestion control

• US 6,097,697 A
• Filed: 07/17/1998
• Issued: 08/01/2000
• Est. Priority Date: 07/17/1998
• Status: Expired due to Term

• Current Assignee: Barracuda Networks Incorporated
• Original Assignee: Sitara Networks, Inc. (Converged Access, Inc.)
• Inventors: Yao, Jie, Goetz, Thomas
• Primary Examiner(s): MARCELO, MELVIN C
• Application Number:
  US09/118,400
• Time in Patent Office: 746 Days
• Field of Search: 370/229-236, 370/252, 370/253
• US Class Current: 370/230
• CPC Class Codes: H04L 1/0002 by adapting the transmission rate
• CPC Class Codes: H04L 1/187 Details of sliding window management

Abstract

A new approach to congestion control includes features which overcome many of the limitations of the current congestion control approaches. The new approach uses a rate-based congestion control mechanism which uses a combination of multiple indicators of congestion. The transmission rate is decreased when there is an indication of congestion and the rate is increased when there is an indication of little or no congestion. The approach can also limit the transmission rate of multiple data streams destined to the same network node.

Claims

1. A method for controlling congestion in a data communication network by controlling a transmission rate at which a source of data transmits data onto a data network, said method comprising:

deriving values for a plurality of different types of statistical measures related to delivery of data transmitted from the source, each of said plurality of different types of statistical measures providing an indication of congestion on the data network;

adjusting the transmission rate from the source in response to a combination of the derived values for the plurality of different types of statistical measures.

2. The method of claim 1 further comprising:

forming a group of data streams for transmission from the source;

transmitting data from the group of data streams; and

accepting acknowledgments of receipt of the transmitted data from the group of data streams; and

wherein deriving the values for the plurality of different types of statistical measures related to delivery of the data transmitted from the source includes monitoring the transmissions of the data and monitoring the acknowledgments of receipt of the data.

3. The method of claim 2 wherein deriving the values further includes combining acknowledgements for different data streams in the group of data streams.

4. The method of claim 2 wherein forming a group of data streams includes forming a group of data streams which have a common destination on the data network.

5. The method claim 1 further comprising:

computing a maximum transmission rate as a function of the derived values for the plurality of different types of statistical measures; and

wherein adjusting the transmission rate includes limiting the transmission rate to the computed maximum transmission rate.

6. The method of claim 1 wherein deriving the values for the plurality of different types of statistical measures includes monitoring a rate of data loss and a pattern of data loss.

7. The method of claim 6 wherein monitoring the pattern of data loss includes monitoring lengths of sequences of lost data.

8. The method of claim 1 wherein deriving the values for the plurality of different types of statistical measures and the adjusting of the transmission rate is performed in each of a sequence of time intervals.

9. The method of claim 1 wherein adjusting the transmission rate includes:

computing a first factor related to a rate of data loss;

computing a second factor related to lengths of data loss;

combining the first and second factors;

adjusting the transmission rate according to the combined factor.

10. The method of claim 1 wherein the plurality of different types of statistical measures includes a rate of data loss and lengths of data loss.

11. The method of claim 1 wherein the plurality of different types of statistical measures includes a rate of data loss and a pattern of data loss.

12. Software stored on a computer readable medium for causing a computer to perform the functions of:

deriving values for a plurality of different types of statistical measures related to delivery of data transmitted from a source of data over a data network, each of said plurality of different types of statistical measures providing an indication of congestion on the data network;

adjusting the transmission rate from the source in response to a combination of the derived values for the plurality of different types of statistical measures.

13. The software of claim 12 further causing the computer to perform the functions of:

forming a group of data streams for transmission from the source;

transmitting data from the group of data streams; and

accepting acknowledgments of receipt of the transmitted data; and

wherein deriving the values for the plurality of different types of statistical measures related to delivery of the data transmitted from the source includes monitoring the transmissions of the data and monitoring the acknowledgments of receipt of the data.

14. The software of claim 12 wherein the plurality of different types of statistical measures includes a rate of data loss and lengths of data loss.

15. The software of claim 12 wherein the plurality of different types of statistical measures includes a rate of data loss and a pattern of data loss.

16. A congestion control apparatus comprising:

a rate updater for determining a maximum rate of data transmission to a destination over a data network, the rate updater determining the maximum rate using a combination of values for a plurality of different types of statistical measures derived from communication with the destination, each of said plurality of different types of statistical measures providing an indication of congestion on the data network;

storage associating the destination with the determined maximum rate of data transmission; and

a transmission throttle for limiting a rate at which data is transmitted to the destination based on the stored maximum rate.

17. The apparatus of claim 16 wherein the plurality of different types of statistical measures includes a rate of data loss and lengths of data loss.

18. The apparatus of claim 16 wherein the plurality of different types of statistical measures includes a rate of data loss and a pattern of data loss.