Managing data traffic congestion in network nodes

Patent metadata

Field	Value
Publication number	US-12206591-B2
Application number	US-202318313873-A
Country	US
Kind code	B2
Filing date	May 8, 2023
Priority date	May 9, 2022
Publication date	Jan 21, 2025
Grant date	Jan 21, 2025

Abstract

Example embodiments describe a method for managing data traffic congestion in a network communication node. The method comprising maintaining a marking probability based on a marking ratio indicative of a change in packets responsible for congestion in a network queue in the network communication node over a change in a total number of packets processed by the network queue. The method further comprising classifying packets processed by the network queue as scalable packets or other packets based on an identifier included in the packets; wherein the scalable packets support scalable congestion control. The method further comprising marking the scalable packets responsible for congestion in the network queue with a congestion mark to signal data traffic congestion; and marking the other packets that support unscalable congestion control with the congestion mark based on the marking probability.

First claim

The invention claimed is:

1. A method of managing data traffic congestion in a network communication node;
the method comprising:
maintaining a marking probability based on a marking ratio indicative of a change in a number of packets processed by a network queue and responsible for congestion in the network queue in the network communication node over a change in a total number of the packets processed by the network queue, wherein those of the packets responsible for congestion are causing a congestion parameter to exceed a threshold;
classifying packets processed by the network queue as scalable ones of the packets or other of the packets based on an identifier included in the packets, the scalable ones of the packets supporting scalable congestion control;
marking the scalable ones of the packets responsible for congestion in the network queue with a congestion mark to signal data traffic congestion; and
marking the other of the packets that support unscalable congestion control with the congestion mark based on the marking probability.

2. A method according to claim 1 , wherein the maintaining further comprises identifying those of the packets as responsible for congestion in the network queue before enqueuing the packets in the network queue.

3. A method according to claim 1 , wherein the maintaining further comprises identifying those of the packets as responsible for congestion in the network queue upon dequeuing the packets from the network queue.

4. A method according to claim 1 , wherein those of the packets are identified as responsible for congestion in the network queue in response to a size of the network queue exceeding a size threshold.

5. A method according to claim 1 , wherein those of the packets are identified as responsible for congestion in the network queue in response to a sojourn time of the packets in the network queue exceeding a time threshold.

6. A method according to claim 1 , further comprising:
before enqueuing the packets in the network queue, flagging the scalable ones of the packets for marking and flagging the other of the packets for marking based on the marking probability; and
upon dequeuing, performing the marking with the congestion mark for those of the packets that are flagged for marking and are identified as responsible for congestion in the network queue.

7. A method according to claim 1 , further comprising dropping a number of the packets in response to a size of the network queue exceeding a first drop threshold, or in response to a sojourn time of the packets in the network queue exceeding a second drop threshold.

8. A method according to claim 1 , further comprising dropping the other of the packets that are eligible for marking based on the marking probability and that do not support unscalable congestion control.

9. A method according to claim 1 , wherein the congestion mark and the identifier are included in an explicit congestion notification, ECN, field of an internet protocol, IP, header of the packets.

10. The method of claim 9 , wherein the marking further comprises overwriting the identifier included in the explicit congestion notification, ECN, field with the congestion mark.

11. A method according to claim 1 , wherein the maintaining of the marking probability comprises updating the marking probability and the marking ratio at a predetermined time interval, or at a predetermined change in the total number of packets processed by the network queue.

12. A method according to claim 1 , wherein the marking probability is based on a moving average of the marking ratio.

13. An apparatus, comprising:
a network communication node to manage data traffic congestion; and wherein the network communication node is configured to:
maintain a marking probability based on a marking ratio indicative of a change in a number of packets processed by a network queue and responsible for congestion in the network queue in the network communication node over a change in a total number of the packets processed by the network queue, wherein those of the packets responsible for congestion are causing a congestion parameter to exceed a threshold;
classify the packets processed by the network queue as scalable ones of the packets or other of the packets based on an identifier included in the packets, the scalable ones of the packets supporting scalable congestion control;
mark the scalable ones of the packets responsible for congestion in the network queue with a congestion mark to signal data traffic congestion; and
mark the other of the packets that support unscalable congestion control with the congestion mark based on the marking probability.

14. A non-transitory computer readable storage medium comprising computer-executable instructions for performing the following steps when the instructions are executed on a computer:
maintaining a marking probability based on a marking ratio indicative of a change in a number of packets processed by a network queue and responsible for congestion in the network queue in a network communication node over a change in a total number of the packets processed by the network queue, wherein those of the packets responsible for congestion are causing a congestion parameter to exceed a threshold;
classifying the packets processed by the network queue as scalable ones of the packets or other of the packets based on an identifier included in the packets, the scalable ones of the packets supporting scalable congestion control;
marking the scalable ones of the packets responsible for congestion in the network queue with a congestion mark to signal data traffic congestion; and
marking the other of the packets that support unscalable congestion control with the congestion mark based on the marking probability.

Assignees

• Nokia Solutions & Networks Oy

Inventors

• Tilmans Olivier
• De Schepper Koen

Classifications

• H04L47/56

  implementing delay-aware scheduling · CPC title

• H04L47/12

  Avoiding congestion; Recovering from congestion · CPC title

• H04L47/32

  by discarding or delaying data units, e.g. packets or frames · CPC title

• H04L47/127

  by using congestion prediction · CPC title

• H04L47/629

  Ensuring fair share of resources, e.g. weighted fair queuing [WFQ] · CPC title