Dynamic adaptation to increased SFU load by disabling video streams

• US 10,075,483 B2
• Filed: 04/20/2017
• Issued: 09/11/2018
• Est. Priority Date: 09/28/2016
• Status: Active Grant

• Current Assignee: Atlassian Pty Ltd (d/b/a Confluence) (Atlassian Corporation)
• Original Assignee: Atlassian Pty Ltd (d/b/a Confluence) (Atlassian Corporation)
• Inventors: Ivov, Emil, Grozev, Boris, Politis, Georgios
• Primary Examiner(s): Scott, Randy
• Application Number:
  US15/493,057
• Publication Number:
  US 20180091564A1
• Time in Patent Office: 509 Days
• Field of Search: 709200, 709204, 709205, 709206, 709217, 370329, 370352, 370396, 725 36, 725 40, 725 60
• CPC Class Codes: H04L 43/0894 Packet rate
• CPC Class Codes: H04L 47/12 Avoiding congestion; Recovering from congestion
• CPC Class Codes: H04L 65/1089 by adding media; by removing media
• CPC Class Codes: H04L 65/403 Arrangements for multi-party communication, e.g. for conferences data switching systems for conference H04L12/18; arrangements for connecting several subscribers to a common circuit, i.e. affording conference facilities H04M3/56; television conferencing systems H04N7/15
• CPC Class Codes: H04L 65/752 adapting media to network capabilities
• CPC Class Codes: H04L 65/765 intermediate
• CPC Class Codes: H04L 65/80 Responding to QoS
• CPC Class Codes: H04N 7/152 Multipoint control units therefor

Abstract

A method for dynamically adapting to increased system load at a selective forwarding unit (SFU) is disclosed. In an embodiment, an SFU sends a plurality of video streams to a plurality of participant computing devices. The SFU monitors a system load value on the SFU according to any of a plurality of metrics. When the SFU determines that the monitored system load value exceeds a stored load threshold value, the SFU selects one or more of the streams being sent to participant computing devices and disables them.

Claims

1. A method comprising:

sending, from a selective forwarding unit (SFU) to each of a plurality of participant computing devices, one or more digital video streams, wherein the SFU stores identifiers of each participant computing device among the plurality of participant computing devices;

monitoring, at the SFU, a system load value which estimates a load on the SFU according to any of a plurality of different load metrics;

determining that the system load value exceeds a particular load threshold value and, in response, selecting a subset of the identifiers of the plurality of participant computing devices;

for each participant computing device that corresponds to one of the identifiers of the subset, disabling one or more of the digital video streams that are sent to the participant computing device from the SFU;

in response to disabling one or more of the digital video streams that are sent to the participant computing device of the subset of the plurality of participant computing devices, waiting a particular period of time;

after waiting the particular period of time, determining that the system load value is below a second load threshold value and, in response, selecting one or more participant computing devices of the subset of the plurality of participant computing devices;

restoring one or more of the disabled digital video streams.

2. The method of claim 1, further comprising, after disabling the one or more of the digital video streams that are sent to the participant computing device, sending a reduced number of the digital video streams to that participant computing device.

3. The method of claim 1, wherein the one or more of the digital video streams comprise digital video streams of a first quality;

the method further comprising replacing the one or more of the digital video streams with one or more digital video streams comprising digital video streams of a second quality, wherein the second quality is lower than the first quality.

4. The method of claim 1, further comprising monitoring the system load value based on network utilization of the SFU, measured as an average egress bitrate on the SFU over a particular period of time.

5. The method of claim 1, further comprising monitoring the system load value based on CPU usage of the SFU.

6. The method of claim 1, further comprising selecting the subset of the plurality of participant computing devices by:

identifying, for each participant computing device of the subset of the plurality of participant computing devices, a particular number of the one or more digital video streams to disable for the participant computing device;

computing a load effect value which estimates an effect on the load of the SFU of disabling the one or more of the digital video streams that are sent to the participant computing device from the SFU, for each participant computing device;

computing a difference between a current system load value and the load effect value;

computing an aggregated participant effect value which estimates a total effect on participant experience of participants of the subset of the plurality of participant computing devices;

identifying each participant computing device of the subset of the plurality of participant computing devices and a particular number of the one or more digital video streams to disable such that the computed difference between the current system load value and the load effect value does not exceed the particular load threshold and such that the aggregated participant effect value is minimized.

7. The method of claim 6, further comprising computing the aggregated participant effect value by:

for each participant computing device, computing an individual participant effect value by dividing the particular number for the participant computing device by a current total number of video streams for the participant computing device;

computing the aggregated participant effect value as a summation of the individual participant effect values for each participant computing device.

8. The method of claim 1, wherein the plurality of participant computing devices comprises conference sets of participant computing devices, and the method further comprising selecting the subset of the plurality of participant computing devices by:

selecting one or more of the conference sets of participant computing devices;

lowering a maximum stream number for the selected one or more of the conference sets;

for each participant computing device in the subset of the plurality of participant computing devices, determining that a number of the one or more digital video streams sent to the participant computing device exceeds the maximum stream number for a particular conference set of the selected one or more of the conference sets which includes the participant computing device.

9. A system comprising:

a selective forwarding unit (SFU) comprising a memory, one or more processors, and one or more instructions stored in the memory which, when executed by the one or more processors, cause;

sending, from the SFU to each of a plurality of participant computing devices, one or more digital video streams, wherein the SFU stores identifiers of each participant computing device among the plurality of participant computing devices;

monitoring, at the SFU, a system load value which estimates a load on the SFU according to any of a plurality of different load metrics;

determining that the system load value exceeds a particular load threshold value and, in response, selecting a subset of the identifiers of the plurality of participant computing devices;

for each participant computing device that corresponds to one of the identifiers of the subset, disabling one or more of the digital video streams that are sent to the participant computing device from the SFU;

in response to disabling one or more of the digital video streams that are sent to the participant computing device of the subset of the plurality of participant computing devices, waiting a particular period of time;

after waiting the particular period of time, determining that the system load value is below a second load threshold value and, in response, selecting one or more participant computing devices of the subset of the plurality of participant computing devices;

restoring one or more of the disabled digital video streams.

10. The SFU of claim 9, wherein the one or more instruction, when executed by the one or more processors, further cause, after disabling the one or more of the digital video streams that are sent to the participant computing device, sending a reduced number of the digital video streams to that participant computing device.

11. The SFU of claim 9, wherein the one or more of the digital video streams comprise digital video streams of a first quality;

wherein the one or more instruction, when executed by the one or more processors, further cause replacing the one or more of the digital video streams with one or more digital video streams comprising digital video streams of a second quality, wherein the second quality is lower than the first quality.

12. The SFU of claim 9, wherein the one or more instruction, when executed by the one or more processors, further cause monitoring the system load value based on network utilization of the SFU, measured as an average egress bitrate on the SFU over a particular period of time.

13. The SFU of claim 9, wherein the one or more instruction, when executed by the one or more processors, further cause monitoring the system load value based on CPU usage of the SFU.

14. The SFU of claim 9, wherein the one or more instruction, when executed by the one or more processors, further cause selecting the subset of the plurality of participant computing devices by:

identifying, for each participant computing device of the subset of the plurality of participant computing devices, a particular number of the one or more digital video streams to disable for the participant computing device;

computing a load effect value which estimates an effect on the load of the SFU of disabling the one or more of the digital video streams that are sent to the participant computing device from the SFU, for each participant computing device;

computing a difference between a current system load value and the load effect value;

computing an aggregated participant effect value which estimates a total effect on participant experience of participants of the subset of the plurality of participant computing devices;

identifying each participant computing device of the subset of the plurality of participant computing devices and a particular number of the one or more digital video streams to disable such that the computed difference between the current system load value and the load effect value does not exceed the particular load threshold and such that the aggregated participant effect value is minimized.

15. The SFU of claim 14, wherein the one or more instruction, when executed by the one or more processors, further cause computing the aggregated participant effect value by:

for each participant computing device, computing an individual participant effect value by dividing the particular number for the participant computing device by a current total number of video streams for the participant computing device;

computing the aggregated participant effect value as a summation of the individual participant effect values for each participant computing device.

16. The SFU of claim 9 wherein the plurality of participant computing devices comprises conference sets of participant computing devices, and wherein the one or more instruction, when executed by the one or more processors, further cause selecting the subset of the plurality of participant computing devices by:

selecting one or more of the conference sets of participant computing devices;

lowering a maximum stream number for the selected one or more of the conference sets;

for each participant computing device in the subset of the plurality of participant computing devices, determining that a number of the one or more digital video streams sent to the participant computing device exceeds the maximum stream number for a particular conference set of the selected one or more of the conference sets which includes the participant computing device.