IP Library › Granted Patent US 12,335,106
Granted Patent B2
US 12,335,106 · App. 17/695,271 · Granted Jun 17, 2025

Network parameter configuration method and apparatus, computer device, and storage medium

Inventors: Siyu Yan (Beijing, CN); Yinben Xia (Shenzhen, CN); Xiaolong Zheng (Beijing, CN); Weishan Deng (Beijing, CN); Zhigang Ji (Beijing, CN); Di Qu (Beijing, CN); Fuxing Chen (Boulogne Billancourt, FR)
Assignee: HUAWEI TECHNOLOGIES CO., LTD.
H04L41/145H04L41/0816H04L41/147H04L43/062
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,335,106
App. No.
17/695,271
Granted
Jun 17, 2025
Kind
B2
Abstract

A network parameter configuration method, where in the method, network running data corresponding to a first period is input into a prediction model, so that the prediction model predicts, based on the input network running data, a value of a parameter of a network device in a second period, and the parameter of the network device in the second period is configured to the value predicted by the prediction model.

Claims (43)

1. A method implemented by a network device, wherein the method comprises:

obtaining, by a parameter calculation circuit of the network device, network running data, wherein the network running data is a characteristic of packet transmission of the network device;

inputting, by the parameter calculation circuit, the network running data into a prediction model;

obtaining, by the parameter calculation circuit, a configuration value, output by the prediction model based on the network running data, of a specified parameter, wherein the specified parameter comprises an explicit congestion notification (ECN) parameter, and wherein the ECN parameter is used to set an ECN flag for a packet forwarded by the network device;

sending, by the parameter calculation circuit, the configuration value;

receiving, by a forwarding chip of the network device, the configuration value from the parameter calculation circuit; and

configuring the specified parameter of the network device to the configuration value in response to receiving the configuration value.

2. The method of claim 1 , further comprising inputting the network running data into the prediction model to obtain an updated prediction model.

3. The method of claim 1 , wherein the network running data comprises a bandwidth utilization, and wherein the method further comprises:

constructing a configuration space comprising at least one association relationship, wherein each of the at least one association relationship is used to indicates a value corresponding to the specified parameter;

establishing a first reward relationship and a second reward relationship, wherein the first reward relationship is used to indicate impact, on network transmission performance, of a quantity of packets buffered by the network device, and the second reward relationship is used to indicate impact of the bandwidth utilization of the network device on network transmission performance; and

constructing an initial prediction model based on the configuration space, the first reward relationship, and the second reward relationship.

4. The method of claim 1 , wherein the network running data comprises at least one of bandwidth utilization, a buffered packet quantity, or a normalized value of the buffered packet quantity.

5. A network parameter configuration apparatus, comprising:

a parameter calculation circuit configured to:

obtain network running data, wherein the network running data is a characteristic of packet transmission of the network parameter configuration apparatus;

input the network running data into a prediction model;

obtain a configuration value, output by the prediction model based on the network running data, of a specified parameter, wherein the specified parameter comprises an explicit congestion notification (ECN) parameter, wherein the ECN parameter is used to set an ECN flag for a packet forwarded by the network parameter configuration apparatus; and

send the configuration value; and

a forwarding chip coupled to the parameter calculation circuit and configured to:

receive the configuration value from the parameter calculation circuit; and

configure the specified parameter of the network parameter configuration apparatus to the configuration value in response to receiving the configuration value.

6. The network parameter configuration apparatus of claim 5 , wherein the parameter calculation circuit is further configured to further input the network running data into the prediction model to obtain an updated prediction model.

7. The network parameter configuration apparatus of claim 5 , wherein the parameter calculation circuit is further configured to:

construct a configuration space comprising at least one association relationship, wherein each of the at least one association relationship indicates a value corresponding to the specified parameter;

establish a first reward relationship and a second reward relationship, wherein the first reward relationship is used to indicate impact, on network transmission performance, of a quantity of packets buffered by the network parameter configuration apparatus, and the second reward relationship is used to indicate impact of bandwidth utilization of the network parameter configuration apparatus on network transmission performance; and

construct an initial prediction model based on the configuration space, the first reward relationship, and the second reward relationship.

8. The network parameter configuration apparatus of claim 5 , wherein the network running data comprises at least one of bandwidth utilization, a buffered packet quantity, or a normalized value of the buffered packet quantity.

9. The method of claim 1 , wherein the ECN parameter comprises at least one of a first threshold indicating a first quantity of packets buffered, a second threshold indicating a second quantity of packets buffered, or a mark probability threshold, and wherein the first threshold is less than the second threshold.

10. The method of claim 1 , wherein the forwarding chip is configured to obtain collected data of the network device.

11. A non-transitory computer-readable storage medium comprising computer executable instructions and that, when executed by a processor, cause a network parameter configuration apparatus to:

obtain, by a parameter calculation circuit of the network parameter configuration apparatus, network running data, wherein the network running data is a characteristic of packet transmission of the network parameter configuration apparatus;

input, by the parameter calculation circuit, the network running data into a prediction model;

obtain, by the parameter calculation circuit, a configuration value, output by the prediction model based on the network running data, of a specified parameter, wherein the specified parameter comprises an explicit congestion notification (ECN) parameter, and wherein the ECN parameter is used to set an ECN flag for a packet forwarded by the network parameter configuration apparatus;

send, by the parameter calculation circuit, the configuration value;

receive, by a forwarding chip of the network parameter configuration apparatus, the configuration value from the parameter calculation circuit; and

configure the specified parameter of the network parameter configuration apparatus to the configuration value in response to receiving the configuration value.

12. The non-transitory computer-readable storage medium of claim 11 , wherein the computer executable instructions further cause the network parameter configuration apparatus to input the network running data into the prediction model to obtain an updated prediction model.

13. The non-transitory computer-readable storage medium of claim 11 , wherein the ECN parameter comprises at least one of a first threshold indicating a first quantity of packets buffered, a second threshold indicating a second quantity of packets buffered, or a mark probability threshold, and wherein the first threshold is less than the second threshold.

14. The non-transitory computer-readable storage medium of claim 11 , wherein the forwarding chip is configured to obtain collected data of the network device.

15. The network parameter configuration apparatus of claim 5 , wherein the ECN parameter comprises at least one of a first threshold indicating a first quantity of packets buffered, a second threshold indicating a second quantity of packets buffered, or a mark probability threshold, and wherein the first threshold is less than the second threshold.

16. The network parameter configuration apparatus of claim 5 , wherein the parameter calculation circuit comprises one or more of a field programmable gate array, a central processing unit, a network processor, or an artificial intelligence chip of the network parameter configuration apparatus.

17. The method of claim 1 , wherein the parameter calculation circuit comprises one of a field programmable gate array, a central processing unit, a network processor, or an artificial intelligence chip of the network device.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Apr 29, 2025
From: YAN, SIYU; XIA, YINBEN; ZHENG, XIAOLONG; DENG, WEISHAN; JI, ZHIGANG; QU, DI; CHEN, FUXING
To: HUAWEI TECHNOLOGIES CO., LTD.
Reel/Frame 070970/0560 →
Priority Claims (1)
CN 201910878276.2 · Sep 17, 2019 · national
Continuity (2)
Continuation PCTCN2020112599 · Aug 31, 2020
Related Publication 20220210026A1 · Jun 30, 2022
References Cited (19)
US 10873533B1 · Ismailsheriff · 2020 [cited by examiner]
US 20070055477A1 · Chickering et al. · 2007 [cited by applicant]
US 20140064079A1 · Kwan · 2014 [cited by examiner]
US 20170076595A1 · Macfarlane et al. · 2017 [cited by applicant]
US 20170111233A1 · Kokkula · 2017 [cited by examiner]
US 20170366398A1 · Mizrachi · 2017 [cited by examiner]
US 20180302296A1 · Yu et al. · 2018 [cited by applicant]
US 20190230009A1 · Yousuf · 2019 [cited by examiner]
US 20190319881A1 · Maskara · 2019 [cited by examiner]
CN 106058851A · 2016 [cited by applicant]
CN 108924127A · 2018 [cited by applicant]
CN 109417514A · 2019 [cited by applicant]
CN 109861925A · 2019 [cited by applicant]
CN 109873712A · 2019 [cited by applicant]
CN 110022264A · 2019 [cited by applicant]
CN 110061927A · 2019 [cited by applicant]
CN 110213784A · 2019 [cited by applicant]
Yifei Lu et al., “Dynamic ECN marking threshold algorithm for TCP congestion control in data center networks”, Computer Communications 129 (2018) 197 208, total 12 pages. [cited by applicant]
Joy, S., et al., “SDN based ECN Adaptation Scheme for Improving Data Center Long Flow Throughput,” IEEE ICC 2015-Workshop on Smart Communication Protocols and Algorithms (SCPA 2015), School of Electrical Engineering and… [cited by applicant]