Search

CN-122001817-A - Congestion control method and related device

CN122001817ACN 122001817 ACN122001817 ACN 122001817ACN-122001817-A

Abstract

A congestion control method and a related device are applied to the technical field of communication. The present application obtains congestion information for flow control by processing characteristic information of at least two flows of a first output port, and performs flow control on the flows of the first output port based on the congestion information. The application can realize flow control of the flow level, and can avoid network congestion in advance, so as to realize faster and more accurate network congestion avoiding effect and improve network performance. The application can further support the determination of congestion information through the artificial intelligence AI model, thereby improving the accuracy of predicting the network congestion condition.

Inventors

  • XING XIN
  • LV XIANBIN
  • ZHANG ZHENXING

Assignees

  • 北京华为数字技术有限公司

Dates

Publication Date
20260508
Application Date
20241108

Claims (20)

  1. 1. A method of congestion control, the method comprising: Acquiring characteristic information of at least two flows, wherein the characteristic information is used for indicating time characteristics of messages of the flows and/or data volume characteristics of the messages of the flows, the at least two flows are sent to a first output port, the first output port belongs to first network equipment, and the first network equipment is used for forwarding the messages of the at least two flows; Obtaining congestion information based on the characteristic information of the at least two flows, wherein the congestion information comprises a predicted congestion state of each flow in the at least two flows and/or a predicted congestion state of the first outlet port; The congestion information is used for controlling the flow of the flow passing through the first outlet port.
  2. 2. The method of claim 1, wherein the characteristic information comprises one or more of message length information, message interval information, and periodic flow rate information, The message length information comprises message lengths of at least two messages, the message interval information comprises at least two message intervals, and the period flow rate information comprises period flow rates of at least two periods.
  3. 3. The method of claim 2, wherein the characteristic information of the at least two streams includes message length information and message interval information of the at least two streams, or wherein the characteristic information of the at least two streams includes periodic flow rate information.
  4. 4. A method according to any of claims 1-3, wherein said deriving congestion information based on characteristic information of said at least two flows comprises: Based on the characteristic information of the at least two flows, obtaining a congestion analysis result of each flow in the at least two flows; and obtaining the congestion information based on the congestion analysis result of each flow in the at least two flows.
  5. 5. A method according to any of claims 1-3, wherein said deriving said congestion information based on characteristic information of said at least two flows comprises: And obtaining the congestion information based on the characteristic information of the at least two flows and an artificial intelligence AI model.
  6. 6. The method according to any one of claims 1-5, wherein said deriving congestion information based on characteristic information of said at least two flows comprises: and obtaining the congestion information based on the characteristic information of the at least two flows and the resource use condition of the first outlet port.
  7. 7. The method of claim 6, wherein the deriving the congestion information based on the characteristic information of the at least two flows and the resource usage of the first egress port comprises: and obtaining the congestion information based on the characteristic information of the at least two flows, the resource use condition of the first outlet port and an AI model.
  8. 8. The method according to any one of claims 1-7, wherein the obtaining characteristic information of at least two streams comprises: Acquiring the characteristic information of the flow to which the first message belongs when a new first message appears at the first output port, wherein the flow corresponding to the first message belongs to the at least two flows; Or acquiring the characteristic information of the at least two streams in a period of a first time length, wherein the first time length is predefined.
  9. 9. The method according to any one of claims 1-8, further comprising: Updating a first corresponding relation set based on the congestion information, wherein the first corresponding relation set is used for indicating the corresponding relation between the identification of the flow and the information of the flow, the information of the flow comprises the congestion state of the flow, and the first corresponding relation set comprises the corresponding relation between the identification of the at least two flows and the information of the at least two flows.
  10. 10. The method of any of claims 1-9, wherein the congestion information comprises a congestion status of the first egress port, the method further comprising: A congestion status of a flow through the first egress port is determined based on the congestion status of the first egress port.
  11. 11. The method according to any of claims 1-10, wherein the congestion status of the flow comprises a congestion level of the flow, the congestion level of the flow being one of a plurality of levels, each of the plurality of levels being indicative of a different congestion situation.
  12. 12. The method of any of claims 1-11, wherein the congestion information comprises a predicted congestion status of the first egress port, the method further comprising: and controlling the flow of the flow passing through the first outlet port based on the congestion state of the first outlet port.
  13. 13. The method of any of claims 1-11, wherein the congestion information includes a predicted congestion status of each of the at least two flows, the method further comprising: and controlling the flow of the first flow based on the congestion state of the first flow, wherein the first flow belongs to the at least two flows.
  14. 14. A congestion control method, applied to a first network device, where the first network device is configured to forward a packet of a flow, and the first network device includes a first output port, the method includes: providing feature information of at least two streams to a processing device, wherein the feature information is used for indicating time features of messages of the streams and/or data volume features of the messages of the streams, and the at least two streams go to the first output port; Obtaining congestion information from the processing device, the congestion information including a predicted congestion status of the each flow and/or a predicted congestion status of the first egress port; and controlling the flow of the first outlet port according to the congestion information.
  15. 15. The method of claim 14, wherein the characteristic information comprises one or more of message length information, message interval information, and periodic flow rate information, The message length information comprises message lengths of at least two messages, the message interval information comprises at least two message intervals, and the period flow rate information comprises period flow rates of at least two periods.
  16. 16. The method according to claim 14 or 15, characterized in that the method further comprises: providing the processing device with the resource usage of the first output port, wherein the congestion information is related to the resource usage of the first output port.
  17. 17. The method of any of claims 15-16, wherein an artificial intelligence AI model is deployed within the processing device, the AI model to derive the congestion information based on characteristic information of the at least two flows.
  18. 18. The method according to any one of claims 14-17, further comprising: Acquiring the characteristic information of the flow to which the first message belongs when a new first message appears at the first output port, wherein the flow corresponding to the first message belongs to the at least two flows; Or acquiring the characteristic information of the at least two streams in a period of a first time length, wherein the first time length is predefined.
  19. 19. The method according to any one of claims 14-18, further comprising: Updating a first corresponding relation set based on the congestion information, wherein the first corresponding relation set is used for indicating the corresponding relation between the identification of the flow and the information of the flow, the information of the flow comprises the congestion state of the flow, and the first corresponding relation set comprises the corresponding relation between the identification of the at least two flows and the information of the at least two flows.
  20. 20. The method of any of claims 14-19, wherein the congestion information comprises a congestion status of the first output port, the method further comprising: A congestion status of a flow through the first egress port is determined based on the congestion status of the first egress port.

Description

Congestion control method and related device Technical Field The present application relates to the field of communications technologies, and in particular, to a congestion control method and a related device. Background With the rapid development of 5G and mobile Internet technologies, more and more Internet application programs are developed and popularized, and the technologies greatly improve the life convenience and the quality of digital experience of people from intelligent home, online education to telemedicine, virtual reality and the like. However, with the popularity of mobile devices and the diversification of application scenarios, network traffic has also increased dramatically. The smooth playing of high-definition videos, the real-time interaction of large-scale multiplayer online games and the wide use of various cloud computing and big data application all bring higher requirements to the transmission speed, delay and stability of a network. In this context, network congestion phenomena also become more frequent and severe. Some network congestion avoidance methods are not comprehensive and deep enough in consideration, which results in lack of accuracy, but today's network architecture is increasingly complex, data traffic is explosively increased, causes of network congestion are diversified and complex, and it is difficult to effectively alleviate network congestion conditions by using a traditional congestion avoidance mechanism, and instead, overall performance of a network may be limited, which results in poor user experience. Thus, in complex network environments, these congestion avoidance methods have been difficult to effectively address the increasing network demands, and more efficient solutions are needed. Disclosure of Invention The application provides a congestion control method and a related device, which can obtain congestion information through the characteristic information of at least two flows, and control the flow of a first outlet port through the congestion information, thereby improving the accuracy, the instantaneity and the effectiveness of congestion control, reducing the probability of packet loss and improving the throughput of a network. In a first aspect, the present application provides a congestion control method, including obtaining characteristic information of at least two flows, and obtaining congestion information based on the characteristic information of the at least two flows. The characteristic information indicates time characteristics of messages of the streams and/or data volume characteristics of the messages of the streams, and at least two streams go to a first output port, wherein the first output port belongs to first network equipment, and the first network equipment is used for forwarding the messages of the at least two streams. The congestion information includes a predicted congestion status of each of the at least two flows and/or a predicted congestion status of the first egress port. The congestion information is used for flow control of the flow passing through the first outlet port, i.e. is used in the flow control of the first outlet port, e.g. the operations performed when the flow control of the first outlet port is effected by the congestion information (described below). Alternatively, the method may be implemented by a processing device or congestion control device. In some aspects, the method is applied to a network device, for example, implemented by a hardware module and/or a software module in the network device, for example, implemented by a processing device or a congestion control device in the network device. The network device may include a plurality of ports (or interfaces) from which messages may be received and/or sent. For convenience of description, the congestion control apparatus will be described as an execution subject. The characteristic information of the stream reflects the time characteristics of the messages of the stream and/or the data volume characteristics of the messages of the stream. Since congestion of a network tends to occur at ports of a network device, multiple flows through the same egress port are involved, and the flows are mutually affected. Thus, the present application uses the characteristic information analysis of at least two flows to the same outlet port to obtain congestion information which is further used for flow control of the flow of the first outlet port, wherein the flow control is one of a ramp-up operation and a ramp-down operation. Illustratively, the congestion information may be used for flow control of one or more of the at least two flows, or congestion control of the first outlet port based on the congestion status of the first outlet port, of course, congestion control of the first outlet port, i.e. flow control of flows passing through the first outlet port. Therefore, the application can more comprehensively analyze the congestion state of the flow of the firs