Quality of service for mixed service tiers
Improving the quality of service in multiple service tiers, particularly during periods of Internet congestion, may be accomplished by scaling back the bitrate from a maximum rate to a minimum reserved rate in a non-linear fashion. The use of an intermediate value for scaling back enables a quick drop off of high maximum rates. In embodiments, the shape of a non-linear curve may be configured and controlled to make the drop-off steeper for heavy users.
1. A method for a non-linear reduction in bitrates for service-flows using shared Internet bandwidth as congestion increases, the method comprising:
identifying a maximum bitrate and a minimum reserved bitrate for each of a plurality of service-flows; and
determining at least one intermediate bitrate for each of the plurality of service-flows to alter a linear scaling between the maximum bitrate to the minimum reserved bitrate, the at least one intermediate bitrate determined to cause, as congestion of the shared Internet bandwidth increases, both:
a reduction in the bitrates for each service flow between the maximum bitrate and the intermediate bitrate that is greater than that of a linear reduction between the maximum and minimum reserved bitrates, and
a greater reduction in bitrates servicing heavier users than a reduction in bitrates servicing comparatively lighter users,
wherein inserting the at least one intermediate bitrate alters the scaling between the maximum bitrate and the minimum reserved bitrate to cause an increasing reduction in a ratio of operative bitrates to respective maximum rates between service-flows as the congestion increases.
2. The method of claim 1 , wherein the bitrate of a service flow is scaled from a maximum rate to a minimum reserved rate in a non-linear fashion.
3. The method of claim 2 , wherein the minimum reserved rate is zero.
4. The method of claim 1 , wherein the at least one intermediate rate determined for a service-flow in the plurality of service flows having a respectively higher maximum rate prevents the service-flow from having a bitrate that is proportional to at least one of the service-flow's maximum rate or minimum reserved rate during a period of increased congestion.
5. The method of claim 1 , wherein at least one of a maximum rate, a minimum reserved rate, or a priority of the plurality of service flows is maintained during the non-linear reduction in bitrates during periods of congestion.
6. The method of claim 1 , further comprising dynamically calculating and updating an intermediate rate of heavy users as congestion increases.
7. The method of claim 1 , wherein the at least one intermediate rate for a heavy user is calculated based on the heavy user's bandwidth utilization measured over a configured prior interval.
8. The method of claim 1 , wherein a heavier user is a user with relatively more usage in a previous period of time of a defined duration than another user of the Internet bandwidth.
9. The method of claim 1 , wherein usage is determined based on a usage in a prior period of time of a limited duration.
10. The method of claim 9 , wherein a first service-flow, with a configured maximum bitrate that is greater than a configured maximum bitrate of a second service-flow, has a higher intermediate rate than an intermediate rate of the second service flow.
11. The method of claim 1 , wherein the maximum rate and the minimum reserved bitrate is the same for two or more of the plurality of service-flows.
12. The method of claim 1 , wherein the intermediate rate of service flow is inserted such that there is less difference in bitrates across service-flows as congestion increases.
13. The method of claim 1 , wherein the intermediate bitrate is further determined to cause a reduction of the ratio of intermediate rate to maximum rate as a maximum rate of the plurality of service flows is increased.
14. The method of claim 1 , wherein the intermediate bitrate is set to at least one of a discrete value determined based on service flow utilizations measured by a CMTS, or a value that varies from a maximum value towards a minimum value based on a measured utilization over a prior interval of a defined duration.
15. The method of claim 1 , wherein the operative bitrates follow a curve between the maximum bitrate and the minimum reserved rate that is pushed lower than the linear scaling by the intermediate bitrate.
16. The method of claim 1 , wherein the reduction in bitrates is a sharper reduction at the onset of a threshold level of congestion.
17. The method of claim 16 , wherein the threshold level of congestion is determined based on a percentage of the available Internet bandwidth that is in use.
18. The method of claim 1 , wherein the intermediate bitrate is varied to vary a penalization of bandwidth available to heavier users.
19. The method of claim 1 , wherein a shape of the non-linear curve that connects the maximum bitrate to the intermediate bitrate to the minimum reserved bitrate can be controlled to have a steeper initial drop for heavy users.
20. The method of claim 1 , further comprising imposing a requirement that all of the service-flows in the plurality of service-flows get their minimum reserved rate before any one service-flow gets more than their minimum reserved rate.