Congestion control method and device, computer equipment and storage medium

By reporting congestion feedback information on network devices and determining the quality characterization value of the data flow, and adjusting the congestion window size, the problem that traditional congestion algorithms cannot effectively consider the difference in data flow experience quality is solved, and more efficient congestion control and user experience improvement are achieved.

CN119030932APending Publication Date: 2024-11-26TENCENT TECHNOLOGY (SHENZHEN) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202310611933.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-05-26
Publication Date
2024-11-26

Smart Images

  • Figure CN119030932A_ABST
    Figure CN119030932A_ABST
Patent Text Reader

Abstract

The invention relates to a congestion control method and device, computer equipment and a storage medium. The congestion control method comprises the following steps: in response to congestion feedback information reported by network equipment, determining respective quality characterization values of a plurality of data streams transmitted by the network equipment; based on the respective quality representation values of the plurality of data streams, determining a target data stream of which the size of a congestion window needs to be updated from the plurality of data streams, and determining the size of a target congestion window of the target data stream; based on the target congestion window size of the target data flow, sending a congestion window setting instruction for the target data flow to a data sending end of the target data flow; the congestion window setting instruction is used for indicating the data sending end to update the current congestion window size of the target data stream based on the target congestion window size, so that the quality representation values of the plurality of data streams tend to be consistent. By adopting the method, the reasonability of congestion control can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of congestion control, and in particular to a congestion control method, apparatus, computer equipment and storage medium. Background Art

[0002] With the rapid development of information technology, users are using the network to transmit more and more data. The larger the amount of data transmitted in the network, the greater the possibility of data transmission congestion. In the case of data transmission congestion, the network transmission delay, packet loss rate, etc. will increase significantly. Therefore, congestion algorithms are needed for congestion control.

[0003] Traditional congestion algorithms seek to ensure that each data flow has the same bandwidth while fully utilizing bandwidth resources and not causing congestion.

[0004] However, if each data stream is allowed to obtain the same bandwidth, since some data streams require a larger bandwidth to ensure the quality of experience, while some data streams require less bandwidth to ensure the quality of experience, the same bandwidth is insufficient for some data streams and redundant for some data streams, resulting in low rationality of congestion control. Summary of the invention

[0005] Based on this, it is necessary to provide a congestion control method, apparatus, computer device, computer-readable storage medium and computer program product that can improve the rationality of congestion control in response to the above technical problems.

[0006] On the one hand, the present application provides a congestion control method. Applied to a control device, the method includes: in response to congestion feedback information reported by a network device, determining the quality characterization value of each of the multiple data streams transmitted by the network device; the quality characterization value of the data stream is used to characterize the experience quality of the data stream; based on the quality characterization values ​​of each of the multiple data streams, determining the target data stream whose congestion window size needs to be updated from the multiple data streams, and determining the target congestion window size of the target data stream; based on the target congestion window size of the target data stream, sending a congestion window setting instruction for the target data stream to the data sending end of the target data stream; the congestion window setting instruction is used to instruct the data sending end to update the current congestion window size of the data stream based on the target congestion window size, so that the quality characterization values ​​of each of the multiple data streams change in a direction tending to be consistent.

[0007] On the other hand, the present application also provides a congestion control device. The device includes: a feedback instruction response module, which is used to determine the quality characterization value of each of the multiple data streams transmitted by the network device in response to the congestion feedback information reported by the network device; the quality characterization value of the data stream is used to characterize the experience quality of the data stream; a window size determination module, which is used to determine the target data stream whose congestion window size needs to be updated from the multiple data streams based on the quality characterization values ​​of each of the multiple data streams, and determine the target congestion window size of the target data stream; a setting instruction sending module, which is used to send a congestion window setting instruction for the target data stream to the data sending end of the target data stream based on the target congestion window size of the target data stream; the congestion window setting instruction is used to instruct the data sending end to update the current congestion window size of the data stream based on the target congestion window size, so that the quality characterization values ​​of each of the multiple data streams change in a direction tending to be consistent.

[0008] In some embodiments, the window size determination module is also used to calculate the mean of the quality characterization values ​​of each of the multiple data streams to obtain an average characterization value; for each of the data streams, when the quality characterization value of the data stream and the average characterization value are inconsistent, the data stream is determined as a target data stream for which the size of the congestion window needs to be updated.

[0009] In some embodiments, the window size determination module is further used to determine a window adjustment coefficient based on the quality characterization value of the target data flow and the average characterization value; and determine the target congestion window size of the target data flow based on the window adjustment coefficient and the current congestion window size of the target data flow.

[0010] In some embodiments, the congestion feedback information includes data flow information of each of the multiple data flows; the setting instruction sending module is also used to generate a congestion window setting instruction for the target data flow based on the data flow information of the target data flow and the target congestion window size of the target data flow; the data flow information and the target congestion window size are used by the data sending end of the target data flow to update the current congestion window size of the target data flow.

[0011] In some embodiments, the feedback instruction response module is further used to determine, for each of the data streams, a preset quality assessment strategy that matches the data stream from a set of preset quality assessment strategies; and determine a quality characterization value of the data stream based on the preset quality assessment strategy that matches the data stream.

[0012] On the other hand, the present application further provides a computer device, wherein the computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps in the above congestion control method when executing the computer program.

[0013] On the other hand, the present application further provides a computer-readable storage medium having a computer program stored thereon, and the computer program implements the steps in the above congestion control method when executed by a processor.

[0014] On the other hand, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps in the above congestion control method are implemented.

[0015] The above-mentioned congestion control method, apparatus, computer equipment, storage medium and computer program product, in response to the congestion feedback information reported by the network device, determine the quality characterization value of each of the multiple data streams transmitted by the network device, determine the target data stream whose congestion window size needs to be updated from the multiple data streams based on the quality characterization values ​​of each of the multiple data streams, and determine the target congestion window size of at least one target data stream, and send a congestion window setting instruction for the data stream to the data sending end of the target data stream based on the target congestion window size of the target data stream. Since the quality characterization value of the data stream is used to characterize the experience quality of the data stream, the congestion window setting instruction is used to instruct the data sending end to update the current congestion window size of the target data stream based on the target congestion window size of the target data stream, so that the quality characterization values ​​of each of the multiple data streams change in a direction tending to be consistent, thereby alleviating congestion while improving the fairness of user experience quality and improving the rationality of congestion control.

[0016] On the other hand, the present application provides a congestion control method. Applied to a data transmitting end, the method includes: receiving a congestion window setting instruction for a target data stream issued by a control device; the congestion window setting instruction includes a target congestion window size of the target data stream, the target congestion window size is determined based on the quality characterization values ​​of each of a plurality of data streams transmitted by a network device, and the target data stream is transmitted from the data transmitting end to the network device; based on the target congestion window size of the target data stream, the current congestion window size of the target data stream is updated so that the quality characterization values ​​of each of the plurality of data streams change in a direction tending to be consistent.

[0017] On the other hand, the present application also provides a congestion control device. The device includes: a setting instruction receiving module, which is used to receive a congestion window setting instruction for a target data stream issued by a control device; the congestion window setting instruction includes a target congestion window size of the target data stream, and the target congestion window size is determined based on the quality characterization values ​​of each of multiple data streams transmitted by a network device, and the target data stream is transmitted to the network device by a data sending end; a window size updating module, which is used to update the current congestion window size of the target data stream based on the target congestion window size of the target data stream, so that the quality characterization values ​​of each of the multiple data streams change in a direction tending to be consistent.

[0018] In some embodiments, the current congestion window size of the target data stream is stored in a data stream window table; the window size update module is further used to update the current congestion window of the target data stream stored in the data stream window table according to the target congestion window size of the target data stream; when a data packet of the target data stream is to be sent, the current congestion window size of the target data stream is queried from the updated data stream window table; the current sending window size of the target data stream is updated according to the current congestion window size of the target data stream, and the data packet is sent based on the updated current sending window size.

[0019] On the other hand, the present application further provides a computer device, wherein the computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps in the above congestion control method when executing the computer program.

[0020] On the other hand, the present application further provides a computer-readable storage medium having a computer program stored thereon, and the computer program implements the steps in the above congestion control method when executed by a processor.

[0021] On the other hand, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps in the above congestion control method are implemented.

[0022] The above-mentioned congestion control method, apparatus, computer equipment, storage medium and computer program product receive a congestion window setting instruction for a target data stream issued by a control device; the congestion window setting instruction includes a target congestion window size of the target data stream, and the target congestion window size is determined based on the quality characterization values ​​of each of the multiple data streams transmitted by the network device, and the target data stream is transmitted to the network device by the data transmitter; based on the target congestion window size of the target data stream, the current congestion window size of the target data stream is updated so that the quality characterization values ​​of each of the multiple data streams change in a direction that tends to be consistent. Thereby, while alleviating congestion, the fairness of user experience quality is improved, and the rationality of congestion control is improved.

[0023] On the other hand, the present application provides a congestion control method. Applied to a network device, the method includes: performing statistics on the data packets to be forwarded of each of a plurality of data flows to obtain statistical results; generating congestion feedback information when it is determined based on the statistical results that the network device is congested; sending the congestion feedback information to a control device; the congestion feedback information is used to instruct the control device to determine the quality characterization value of each of the plurality of data flows, and update the current congestion window size of the data flow based on each of the quality characterization values, so that the quality characterization values ​​of each of the plurality of data flows change in a direction tending to be consistent.

[0024] On the other hand, the present application also provides a congestion control device. The device includes: a statistical module, which is used to collect statistics on the data packets to be forwarded of each of the multiple data flows to obtain statistical results; a feedback instruction generation module, which is used to generate congestion feedback information when it is determined that the network device is congested based on the statistical results; a feedback instruction sending module, which is used to send the congestion feedback information to the control device; the congestion feedback information is used to instruct the control device to determine the quality characterization value of each of the multiple data flows, and update the current congestion window size of the data flow based on each of the quality characterization values, so that the quality characterization values ​​of each of the multiple data flows change in a direction that tends to be consistent.

[0025] In some embodiments, the statistical module is also used to count the number of data packets to be forwarded of the data flow for each of the data flows to obtain the number of data packets to be forwarded of the data flow; and obtain statistical results based on the number of data packets to be forwarded of each of the multiple data flows; the device is also used to determine that congestion occurs in the network device when at least one of the multiple numbers to be forwarded is greater than a target number threshold.

[0026] In some embodiments, the apparatus is further used to determine a bandwidth of the network device; determine the target quantity threshold based on the bandwidth of the network device; and the target quantity threshold is positively correlated with the bandwidth of the network device.

[0027] On the other hand, the present application further provides a computer device, wherein the computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps in the above congestion control method when executing the computer program.

[0028] On the other hand, the present application further provides a computer-readable storage medium having a computer program stored thereon, and the computer program implements the steps in the above congestion control method when executed by a processor.

[0029] On the other hand, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps in the above congestion control method are implemented.

[0030] The above-mentioned congestion control method, apparatus, computer equipment, storage medium and computer program product collect statistics on the data packets to be forwarded of each of the multiple data streams to obtain statistical results, and when it is determined based on the statistical results that the network device is congested, congestion feedback information is generated and sent to the control device. The congestion feedback information is used to instruct the control device to determine the quality characterization value of each of the multiple data streams, and update the current congestion window size of the data stream based on each quality characterization value, so that the quality characterization values ​​of each of the multiple data streams change in a direction that tends to be consistent. Thereby, while alleviating congestion, the fairness of the user experience quality is improved, and the rationality of congestion control is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A diagram of an application environment of a congestion control method in some embodiments;

[0032] Figure 2 is a flowchart of a congestion control method in some embodiments;

[0033] Figure 3 A flow chart for determining a quality characterization value in some embodiments;

[0034] Figure 4 is a flowchart of a congestion control method in some embodiments;

[0035] Figure 5 A schematic diagram of setting a congestion window in some embodiments;

[0036] Figure 6 is a flowchart of a congestion control method in some embodiments;

[0037] Figure 7 is a flowchart of congestion detection in some embodiments;

[0038] Figure 8 is a flowchart of a congestion control method in some embodiments;

[0039] Fig. 9 is a structural block diagram of a congestion control device in some embodiments;

[0040] Fig.10 is a structural block diagram of a congestion control device in some embodiments;

[0041] Fig.11 is a structural block diagram of a congestion control device in some embodiments;

[0042] Fig.12 is an internal structure diagram of a computer device in some embodiments;

[0043] Fig.13 1 is a diagram of the internal structure of a computer device in some embodiments. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0045] The congestion control method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. The application environment includes a data center. The data center is used to process, store, transmit, exchange, and manage data. Storage systems, routers, switches, firewalls, servers, and application delivery devices are important components of the data center. The data center includes control devices, network devices, and data transmitters. The control device is connected to one or more network devices, and the control device is used to provide congestion control services for the connected network devices. The data center may include multiple network devices, such as Figure 1 The network device 1 and the network device 2 are shown in FIG. The control device can also be called a controller. Each network device can be connected to one or more data sending terminals, for example, Figure 1In the figure, network device 1 is connected to j data transmitting terminals, which are data transmitting terminals 1_1 to 1_j respectively, and network device 2 is connected to k data transmitting terminals, which are data transmitting terminals 2_1 to 2_k respectively. Data transmitting terminals 1_1 to 1_j transmit data streams 1 to 1_j to network device 1 respectively, and data can be transmitted between network devices. For example, a communication link is established between network device 1 and network device 2, and network device 1 can transmit data streams 1 to 1_j to network device 2 through the communication link. In the case of transmission congestion of network device 1, the communication link shared by data streams 1 to 1_j (i.e., the communication link established by network device 1 and network device 2) can be considered as a bottleneck link. Transmission congestion of network device 1 means that at least one port of network device 1 is congested.

[0046] It should be noted that although Figure 1 Only one control device is shown in the figure, but it does not mean that the data center includes only one control device. The data center may include multiple control devices. Figure 1 The network devices in the can be the same type of network devices, for example, all network devices can be interactive machines, or they can be different types of network devices, for example, some network devices are switches, and some network devices are not switches. Network devices include but are not limited to switches, routers, firewalls, bridges, gateways or VPN servers, etc. The data sending end can be the same type of data sending end, or it can be a different type of data sending end. It should be noted that the data sending end here can be any device that can be linked to a network device, for example, the data sending end itself can also be a network device such as a switch. Of course, the data sending end can also be various desktop computers, laptops, smart phones, tablet computers, Internet of Things devices and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart car-mounted devices, etc. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc.

[0047] Figure 1 The data sending end, network device and control device in the embodiment may also be a server. The server may be an independent physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The data sending end, network device and control device may be directly or indirectly connected through wired or wireless communication, which is not limited in this application.

[0048] In some embodiments, the data center may be a cloud data center. A cloud data center refers to a data center in cloud computing. Cloud computing refers to the delivery and use model of IT infrastructure, which refers to obtaining required resources through the network in an on-demand and easily scalable manner; in a broad sense, cloud computing refers to the delivery and use model of services, which refers to obtaining required services through the network in an on-demand and easily scalable manner. Such services may be IT and software, Internet-related, or other services. Cloud computing is the product of the development and integration of traditional computer and network technologies such as grid computing, distributed computing, parallel computing, utility computing, network storage technologies, virtualization, and load balancing. With the development of the Internet, real-time data streams, and the diversification of connected devices, as well as the promotion of search services, social networks, mobile commerce, and open collaboration, cloud computing has developed rapidly. Different from previous parallel distributed computing, the emergence of cloud computing will promote revolutionary changes in the entire Internet model and enterprise management model in terms of concept.

[0049] Among them, the cloud data center can be a private cloud data center or a public cloud data center. Private cloud is to create cloud infrastructure and software and hardware resources within the firewall so that departments within the organization or enterprise can share the resources in the data center. In addition to hardware resources, the creation of a private cloud generally also has cloud equipment (IaaS, Infrastructure as a Service) software. Private cloud computing also includes three levels: cloud hardware, cloud platform, and cloud service. The difference is that cloud hardware is the user's own personal computer or server, not the data center of the cloud computing vendor. The purpose of cloud computing vendors to build data centers is to provide public cloud services to millions of users, so they need to have tens of millions of servers. Private cloud computing only serves friends and relatives for individuals, and only serves employees, customers, and suppliers of the enterprise for enterprises. Therefore, the personal computers or servers of individuals or enterprises are sufficient to provide cloud services.

[0050] Public Cloud usually refers to a cloud provided by a third-party provider for users to use. Public Cloud is generally available through the Internet and may be free or low-cost. The core attribute of Public Cloud is shared resource services. There are many instances of this cloud, which can provide services throughout the open public network today.

[0051] Cloud technology refers to a hosting technology that unifies hardware, software, network and other resources in a wide area network or local area network to achieve data computing, storage, processing and sharing. Cloud technology is a general term for network technology, information technology, integration technology, management platform technology, application technology, etc. based on the cloud computing business model. It can form a resource pool, which is used on demand and flexible and convenient. Cloud computing technology will become an important support. The background services of the technical network system require a large amount of computing and storage resources, such as video websites, picture websites and more portal websites. With the high development and application of the Internet industry, each item may have its own identification mark in the future, and all need to be transmitted to the background system for logical processing. Data of different levels will be processed separately. All kinds of industry data require strong system backing support, which can only be achieved through cloud computing.

[0052] Specifically, the network device 1 periodically detects whether it is congested, that is, whether there is port congestion. When it is determined that it is congested, it sends congestion feedback information to the control device. The control device responds to the congestion feedback information, determines the quality characterization value of each of the data streams 1 to data streams j, determines the target data streams whose congestion window size needs to be updated from the data streams 1 to data streams j based on the quality characterization values ​​of each of the data streams 1 to data streams j, and determines the target congestion window size of the target data stream. Based on the target congestion window size of the target data stream, the congestion window setting instruction for the target data stream is sent to the data sending end of the target data stream; the congestion window setting instruction is used to instruct the data sending end to update the current congestion window size of the target data stream based on the target congestion window size, so that the quality characterization values ​​of each of the multiple data streams change in a direction that tends to be consistent. The quality characterization value of the data stream is used to characterize the quality of experience (QOE) of the data stream. The quality of experience is the user experience quality, which refers to the user's subjective feeling of the quality and performance of the device, network and system, application or service. The congestion control method provided by the present application improves the QOE fairness of each data flow while alleviating congestion.

[0053] In some embodiments, Figure 2 As shown, a congestion control method is provided, which can be executed by a terminal or a server, or can be executed by a terminal and a server together. Figure 1 The control device in the example is used to illustrate, including the following steps:

[0054] Step 202, in response to congestion feedback information reported by the network device, determine the quality characterization value of each of the multiple data streams transmitted by the network device; the quality characterization value of the data stream is used to characterize the experience quality of the data stream.

[0055] Among them, multiple data streams refer to at least two data streams. Each data stream is sent from a data sending end to a network device, for example, different data streams are sent from different data sending ends to a network device. The data stream sent from a data sending end to a network device may be generated by the data sending end itself. The data stream sent from a data sending end to a network device may also be a data stream forwarded by the data sending end, for example, each data packet in a data stream is transmitted to a network device via a data sending end. The multiple data streams are transmitted to other devices via a communication link established by the network device, for example, Figure 1 In the example, network device 1 transmits data stream 1 to data stream j to network device 2 through the communication link established with network device 2. When congestion occurs in the network device, for example, at least one port is congested, the communication link established by the network device for transmitting the multiple data streams can be considered as a bottleneck link. Each data stream is transmitted through each network device in the data center and finally transmitted to the corresponding user device.

[0056] Among them, the data stream is finally sent to the user device through the network device. The type of data stream includes but is not limited to a video data stream, an audio data stream or a text data stream. Each of the multiple data streams can be a data stream in any form. Quality of Experience (QOE) is the user's subjective experience. For example, if the data stream is a video data stream, the experience quality can be the user's experience of the video jam or the clarity of the video, if the data stream is an audio data stream, the experience instruction can be the user's experience of the audio jam or sound quality, if the data stream is a text data stream, the experience quality can be the user's experience of the text download speed. The quality characterization value of the data stream is used to characterize the experience quality of the data stream. The larger the quality characterization value of the data stream, the better the experience quality of the data stream. For example, for a video data stream, the larger the quality characterization value, the better the user's experience of the video jam or the clarity of the video.

[0057] Specifically, the congestion feedback information includes data flow information of each of the multiple data flows. The data flow information includes a data flow identifier. The data flow identifier includes a source IP address, a target IP address, a source port number, a destination port number, and a protocol type. Therefore, the data flow identifier can also be represented by a five-tuple, which is (source IP address, target IP address, source port number, destination port number, protocol type). The data flow information can also include at least one of a transmission protocol, a round-trip time (RTT), the number of bytes in transit, or the number of packet losses. The round-trip time can be, for example, SRTT (smooth RTT). When congestion occurs in a network device, the network device generates congestion feedback information based on the data flow information of each data flow, and sends the congestion feedback information to the control device.

[0058] Specifically, a set of preset quality assessment strategies may be stored in the control device. The preset quality assessment strategy is a strategy for calculating a quality characterization value. The preset quality assessment strategy set includes preset quality assessment strategies corresponding to different types of data streams. For example, the preset quality assessment strategy set includes: at least one of a preset quality assessment strategy corresponding to a text data stream, a preset quality assessment strategy corresponding to an audio data stream, or a preset quality assessment strategy corresponding to a video data stream. For each data stream, the control device may determine the preset quality assessment strategy that matches the data stream from the preset quality assessment strategy set; and determine the quality characterization value of the data stream based on the preset quality assessment strategy that matches the data stream. The control device may determine the data type of the data stream, and query the preset quality assessment strategy corresponding to the data type of the data stream from the preset quality assessment strategy set, and determine the queried preset quality assessment strategy as the preset quality assessment strategy that matches the data stream.

[0059] Step 204: based on the quality characterization values ​​of the multiple data flows, determine a target data flow whose congestion window size needs to be updated from the multiple data flows, and determine a target congestion window size of the target data flow.

[0060] The congestion window size refers to the window size of the congestion window. The target congestion window size is used to update the current congestion window size of the data flow. The current congestion window size of the data flow refers to the congestion window size used by the data sender when sending data packets of the data flow at the current time.

[0061] Specifically, the control device may calculate the average of the quality characterization values ​​of the multiple data flows to obtain an average characterization value. For each data flow, if the quality characterization value of the data flow is inconsistent with the average characterization value, the data flow is determined as a target data flow for which the congestion window size needs to be updated.

[0062] In some embodiments, for each target data stream, the control device determines a window adjustment coefficient based on the quality characterization value and the average characterization value of the target data stream, and the control device can determine the target congestion window size of the target data stream according to the window adjustment coefficient and the current congestion window size of the target data stream. The window adjustment coefficient is positively correlated with the average characterization value. The window adjustment coefficient is negatively correlated with the quality characterization value of the target data stream.

[0063] Among them, positive correlation means: when other conditions remain unchanged, the two variables change in the same direction, and when one variable changes from large to small, the other variable also changes from large to small. It can be understood that the positive correlation here means that the direction of change is consistent, but it does not require that when one variable changes a little, the other variable must also change. For example, when variable a is 10 to 20, variable b can be set to 100, and when variable a is 20 to 30, variable b can be 120. In this way, the direction of change of a and b is when a becomes larger, b also becomes larger. But when a is in the range of 10 to 20, b can be unchanged. Negative correlation means: when other conditions remain unchanged, the two variables change in opposite directions, and when one variable changes from large to small, the other variable changes from small to large. It can be understood that the negative correlation here means that the direction of change is opposite, but it does not require that when one variable changes a little, the other variable must also change.

[0064] In some embodiments, when the quality characterization value of the data flow is consistent with the average characterization value, it is determined that there is no need to update the current congestion window size of the data flow.

[0065] In some embodiments, the control device determines the ratio of the average characterization value to the quality characterization value of the target data stream, and uses the determined ratio as the window adjustment coefficient. The control device may perform a product operation on the window adjustment coefficient and the current sending window size of the target data stream, and use the result of the operation as the target congestion window size of the target data stream. The sending window size refers to the size of the sending window. The sending window is a sliding window of the data sending end. When the data packet is located in the sending window, the data sending end sends the data packet.

[0066] Step 206, based on the target congestion window size of the target data stream, send a congestion window setting instruction for the target data stream to the data sending end of the target data stream; the congestion window setting instruction is used to instruct the data sending end to update the current congestion window size of the target data stream based on the target congestion window size, so that the quality characterization values ​​of the multiple data streams change in a direction tending to be consistent.

[0067] The quality characterization value is positively correlated with the size of the congestion window. The congestion window setting instruction of the target data flow includes the target congestion window size of the target data flow.

[0068] Specifically, the congestion feedback information includes data flow information of each of the multiple data flows. The control device can generate a congestion window setting instruction for the target data flow based on the data flow information of the target data flow and the target congestion window size of the target data flow; the data flow information and the target congestion window size of the target data flow are used by the data sending end of the target data flow to update the current congestion window size of the target data flow.

[0069] In the above congestion control method, in response to the congestion feedback information reported by the network device, the quality characterization value of each of the multiple data streams transmitted by the network device is determined, and based on the quality characterization values ​​of each of the multiple data streams, the target data stream whose congestion window size needs to be updated is determined from the multiple data streams, and the target congestion window size of at least one target data stream is determined, and based on the target congestion window size of the target data stream, a congestion window setting instruction for the data stream is sent to the data sending end of the target data stream. Since the quality characterization value of the data stream is used to characterize the experience quality of the data stream, the congestion window setting instruction is used to instruct the data sending end to update the current congestion window size of the target data stream based on the target congestion window size of the target data stream, so that the quality characterization values ​​of each of the multiple data streams change in a direction that tends to be consistent, thereby alleviating congestion while improving the fairness of the user experience quality, that is, improving QOE fairness and improving the rationality of congestion control. And it can improve the overall user experience quality of the network of the data center and improve the overall performance of the network of the data center.

[0070] The congestion control method provided in this application is applied to the scenario of obtaining video data from a data center for playback, which can improve the video playback quality. Figure 1 For example, data stream 1 is a video type data stream, and data stream 2 to data stream j are text type data streams (for example, data streams corresponding to texts downloaded by users). If each data stream is allocated the same bandwidth (that is, it occupies the same bandwidth in the communication link), it is easy for the quality characterization value of data stream 1 to be smaller than the quality characterization value of data stream 2 to data stream j. Due to the low quality characterization value, the video playback quality is poor. For example, the video may be stuck or have low clarity during playback. Therefore, by adjusting the congestion window size to achieve uniform quality characterization values ​​of data streams 1 to data stream j, the video playback quality can be improved without affecting the experience quality of other data streams too much. Therefore, the congestion control method provided in the present application is applied to the scenario of obtaining video data from a data center for playback, and can improve the video playback quality. In particular, for video streams in online video scenarios, the video quality can be improved and the video freeze rate can be reduced without increasing the hardware bandwidth cost.

[0071] In some embodiments, based on the quality characterization values ​​of each of the multiple data streams, determining the target data stream for which the size of the congestion window needs to be updated from the multiple data streams includes: calculating the mean of the quality characterization values ​​of each of the multiple data streams to obtain an average characterization value; for each data stream, when the quality characterization value of the data stream is inconsistent with the average characterization value, determining the data stream as the target data stream for which the size of the congestion window needs to be updated.

[0072] Specifically, for each data flow, the control device compares the quality characterization value of the data flow with the average characterization value. If the quality characterization value of the data flow is inconsistent with the average characterization value, the target congestion window size of the data flow is determined based on the quality characterization value and the average characterization value of the data flow. If the quality characterization value of the data flow is consistent with the average characterization value, it is determined that there is no need to update the current congestion window size of the data flow.

[0073] In this embodiment, when the quality characterization value and the average characterization value of the data flow are inconsistent, the data flow is determined as the target data flow whose congestion window size needs to be updated, thereby avoiding adjusting the congestion window of the data flow when the quality characterization value and the average characterization value of the data flow are consistent, thereby improving the efficiency of congestion control.

[0074] In some embodiments, determining the target congestion window size of the target data flow includes: determining a window adjustment coefficient based on a quality characterization value and an average characterization value of the target data flow; and determining the target congestion window size of the target data flow based on the window adjustment coefficient and a current congestion window size of the target data flow.

[0075] The window adjustment coefficient is positively correlated with the average characterization value and negatively correlated with the quality characterization value of the target data stream.

[0076] Specifically, the control device determines the ratio of the average characterization value to the quality characterization value of the target data stream, and uses the determined ratio as the window adjustment coefficient. If the quality characterization value of the target data stream is greater than the average characterization value, the window adjustment coefficient is less than 1; if the quality characterization value of the target data stream is less than the average characterization value, the window adjustment coefficient is greater than 1.

[0077] In some embodiments, the control device may calculate the product of the window adjustment coefficient and the current congestion window size of the target data stream, and determine the calculated product as the target congestion window size of the target data stream. In the case where the quality characterization value of the target data stream is greater than the average characterization value, the target congestion window size is smaller than the current congestion window size. In the case where the quality characterization value of the target data stream is less than the average characterization value, the target congestion window size is larger than the current congestion window size.

[0078] In some embodiments, the control device determines the ratio of the average characterization value to the quality characterization value of the target data stream, and uses the determined ratio as the window adjustment coefficient. The control device may perform a product operation on the window adjustment coefficient and the current congestion window size of the target data stream, and use the result of the operation as the target congestion window size of the target data stream. For example, the target congestion window size may be calculated by formula (1).

[0079]

[0080] Among them, w i Refers to the current congestion window size of the i-th data flow. w′ i It refers to the target congestion window size of the i-th data flow. i It refers to the quality characterization value of the i-th data stream. is the average representation value. Refers to the window adjustment coefficient of the i-th data stream

[0081] In this embodiment, a window adjustment coefficient is determined based on the quality characterization value and the average characterization value of the target data flow, and the target congestion window size of the target data flow is determined according to the window adjustment coefficient and the current congestion window size of the target data flow. Since the target congestion window size is used to update the current congestion window size of the target data flow, when the quality characterization value of the target data flow is less than the average characterization value, the current congestion window size of the target data flow can be updated in an increasing direction; when the quality characterization value of the target data flow is greater than the average characterization value, the current congestion window size of the target data flow can be updated in an increasing direction. Since the congestion window size is positively correlated with the quality characterization value, the quality characterization values ​​of each data flow tend to be consistent, that is, tend to the average characterization value, by adjusting the size of the current congestion window size, so that in the process of congestion control, not only congestion is alleviated but also QOE fairness is achieved.

[0082] In some embodiments, the congestion feedback information includes data flow information of each of the multiple data flows; based on the target congestion window size of the target data flow, sending a congestion window setting instruction for the target data flow to the data sending end of the target data flow includes: generating a congestion window setting instruction for the target data flow based on the data flow information of the target data flow and the target congestion window size of the target data flow; the data flow information and the target congestion window size are used for the data sending end of the target data flow to update the current congestion window size of the target data flow.

[0083] Specifically, the data sending end of the data stream refers to a device that transmits the data stream to the control device. The control device generates a congestion window setting instruction for the target data stream based on the data stream information of the target data stream and the target congestion window size of the data stream. The congestion window setting instruction for the target data stream carries the data stream information of the target data stream and the target congestion window size of the target data stream.

[0084] In some embodiments, after the control device generates a congestion window setting instruction for the target data stream, the control device sends the congestion window setting instruction for the target data stream to the data sending end of the target data stream. After receiving the congestion window setting instruction for the target data stream, the data sending end obtains the data stream information from the congestion window setting instruction and obtains the data stream identifier from the data stream information. The data sending end stores a data stream window table, which is used to store the corresponding relationship between the data stream identifier and the current congestion window size. The data sending end searches for the current congestion window size of the target data stream according to the data stream identifier of the target data stream, and updates the current congestion window size to the target congestion window size, thereby realizing the update of the current congestion window size.

[0085] In this embodiment, based on the data flow information of the target data flow and the target congestion window size of the data flow, a congestion window setting instruction for the target data flow is generated, so that the data sending end of the target data flow can quickly update the current congestion window size of the target data flow.

[0086] In some embodiments, determining the quality characterization value of each of multiple data streams transmitted by a network device includes: for each data stream, determining a preset quality assessment strategy that matches the data stream from a set of preset quality assessment strategies; and determining the quality characterization value of the data stream based on the preset quality assessment strategy that matches the data stream.

[0087] Among them, a preset quality assessment strategy set may be stored in the control device. The preset quality assessment strategy is a strategy for calculating a quality characterization value. The preset quality assessment strategy set includes preset quality assessment strategies corresponding to different types of data streams. For example, the preset quality assessment strategy set includes: at least one of a preset quality assessment strategy corresponding to a text data stream, a preset quality assessment strategy corresponding to an audio data stream, or a preset quality assessment strategy corresponding to a video data stream.

[0088] Specifically, the control device can determine the preset quality assessment strategy that matches the data stream from the preset quality assessment strategy set according to the request information of the data stream. The request information of the data stream includes at least one of the request mode, the request header or the request body. The request header can include the data type of the data stream, and the data type includes at least one of the text type, the picture type, the audio type or the video type. For example, the request header includes a MIME type, and the MIME type is used to indicate the data type of the data stream. The MIME (Multipurpose Internet Mail Extensions) type is a standardized way to indicate the data types of various file formats transmitted on the Internet. The MIME type includes at least one of the text type, the picture type, the audio type or the video type. Each preset quality assessment strategy in the preset quality assessment strategy set can correspond to different data types respectively, and the control device can determine the data type of the data stream according to the request information of the data stream, and query the preset quality assessment strategy corresponding to the data type of the data stream from the preset quality assessment strategy set, and determine the queried preset quality assessment strategy as the preset quality assessment strategy that matches the data stream. In some embodiments, the preset quality assessment strategy is used to calculate the quality characterization value according to the delay information. For the preset quality assessment strategy corresponding to the video type, the preset quality assessment strategy can also be used to calculate the quality characterization value according to at least one of the freeze rate, the resolution of the video frame, the screen size of the device receiving the video, and the screen resolution of the device receiving the video. The specific form of the preset quality assessment strategy is not limited in this application.

[0089] In some embodiments, the control device includes an information analysis module, and the quality characterization value can be determined by the information analysis module. Figure 3 As shown, a flow chart of the information analysis module determining the quality characterization value is presented. Figure 3 In the description, "looping to obtain data streams" means obtaining the data packets of each data stream in turn to analyze the quality characterization value of each data stream. "Obtaining data stream request information" means obtaining the request information of the data stream. "Finding the corresponding QOE calculation formula" means determining the matching preset quality assessment strategy based on the request information. "Inserting the delay information carried by the data stream into the formula" means calculating the quality characterization value through the matching preset quality assessment strategy and delay information, and "obtaining the QOE information of each data stream" means obtaining the quality characterization value of each data stream.

[0090] In some embodiments, the control device further includes a window calculation module. The process of determining the target congestion window size can be implemented by the window calculation module. The window calculation module is implemented together. The theoretical analysis of the working principle of the window calculation module is as follows: In order to achieve QOE fairness of the bottleneck link, the problem can be summarized as a maximum minimization QOE problem (i.e., formula (2)).

[0091]

[0092] Among them, QOE i (r i ) represents the i-th data flow in the bottleneck link, at a sending speed of r i The user quality of experience is QOE. Assuming that the QOE formulas of each data stream are continuous, when the QOE of each data stream is equal and the QOE of no stream can continue to increase, the solution of (2) can be obtained. Assuming {r i *} is the optimal speed set for each data stream, then according to formula (2), the QOE of each data stream will be equal. Assume that the sending speed of data stream 1 is r 1 , the QOE achieved is QOE 1 , the average QOE (i.e., the average characterization value) is Then when When , the sending rate of data stream 1 can be modified to achieve QOE balance between the data streams on the bottleneck link. Since the sending rate is affected by the congestion window, the sending rate can be adjusted by adjusting the congestion window. Therefore, the window calculation module is used to calculate the target congestion window size according to formula (1). The network device can regularly detect whether congestion occurs. If congestion occurs, the congestion feedback information reported will be triggered. Each congestion feedback information triggers the window calculation module to calculate the target congestion window size, thereby continuously updating the current congestion window size of the target data stream according to the target congestion window size of the target data stream, and finally making the quality characterization values ​​of multiple data streams tend to be consistent, thereby achieving QOE fairness.

[0093] In this embodiment, for each data stream, a preset quality assessment strategy matching the data stream is determined from a set of preset quality assessment strategies, and the quality characterization value of the data stream is determined based on the preset quality assessment strategy matching the data stream, thereby improving the efficiency of determining the quality characterization value of the data stream.

[0094] In some embodiments, Figure 4 As shown, a congestion control method is provided, which can be executed by a terminal or a server, or can be executed by a terminal and a server together. Figure 1 Any data sending end in Figure 1 Taking the data sending end 1_1) connected to the network device 1 as an example, the method includes the following steps:

[0095] Step 402, receiving a congestion window setting instruction for a target data stream issued by a control device; the congestion window setting instruction includes a target congestion window size of the target data stream, and the target congestion window size is determined based on the quality characterization values ​​of each of a plurality of data streams transmitted by the network device, and the target data stream is transmitted from a data sending end to the network device.

[0096] Specifically, the control device determines the quality characterization value of each of the multiple data streams transmitted by the network device in response to the congestion feedback information reported by the network device. Based on the quality characterization values ​​of each of the multiple data streams, the control device determines the target data stream whose congestion window size needs to be updated from the multiple data streams, determines the target congestion window size of the target data stream, generates a congestion window setting instruction for the target data stream based on the target congestion window size of the target data stream, and sends the congestion window setting instruction for the target data stream to the data sending end of the target data stream.

[0097] Step 404: Update the current congestion window size of the target data flow based on the target congestion window size of the target data flow, so that the quality representation values ​​of the multiple data flows change in a direction tending to be consistent.

[0098] Specifically, the data flow information and the target congestion window size may be directly extracted by the data sending end from the congestion window setting instruction. Alternatively, the data flow information and the target congestion window size may be indirectly obtained by the data sending end from the congestion window setting instruction. For example, the data sending end may parse the congestion window setting instruction to obtain the data flow information and the target congestion window size.

[0099] In some embodiments, the data transmitter can determine the window size of the sending window used when sending the data packet of the target data stream according to the target congestion window size in the congestion window setting instruction. The sending window is a sliding window of the data transmitter. When sending the data packet of the target data stream, if the data packet is within the sending window, the data packet is sent.

[0100] In some embodiments, a data stream window table is stored in the data transmitting end. The data stream window table is used to store the current congestion window size corresponding to the data stream. For example, the corresponding relationship between the data stream and the current congestion window size is stored in the data stream window table. Specifically, the corresponding relationship between the data stream and the current congestion window size can be represented by a key-value pair, in which the data stream identifier is the key and the current congestion window size is the value. The data stream window table can be a table in any form, for example, a hash table.

[0101] In some embodiments, when there is a current congestion window size corresponding to the target data flow in the data flow window table, the data transmitting end can update the current congestion window size of the target data flow in the data flow window table to the target congestion window size of the data flow. When there is no current congestion window size of the target data flow in the data flow window table, the data transmitting end stores the corresponding relationship between the target data flow and the target congestion window size in the data flow window table, so that the target congestion window size becomes the current congestion window size corresponding to the target data flow.

[0102] In some embodiments, the data transmitting end includes a window device module, and steps 402 to 404 may be implemented by a window setting module in the data transmitting end. The window device module in the data transmitting end and the window calculation module in the control device constitute a congestion control system.

[0103] In the above congestion control method, a congestion window setting instruction for a target data stream is received from a control device; the congestion window setting instruction includes a target congestion window size of the target data stream, and the target congestion window size is determined based on the quality characterization values ​​of each of the multiple data streams transmitted by the network device, and the target data stream is transmitted from the data transmitter to the network device; based on the target congestion window size of the target data stream, the current congestion window size of the target data stream is updated so that the quality characterization values ​​of each of the multiple data streams change in a direction that tends to be consistent. This improves the fairness of the user experience quality while alleviating congestion, that is, improves QOE fairness. It can also improve the overall user experience quality of the data center's network and improve the overall performance of the data center's network.

[0104] In some embodiments, the current congestion window size of the target data flow is stored in a data flow window table; updating the current congestion window size of the target data flow based on the target congestion window size of the target data flow includes: updating the current congestion window of the target data flow stored in the data flow window table according to the target congestion window size of the target data flow; when a data packet of the target data flow is to be sent, querying the current congestion window size of the target data flow from the updated data flow window table; updating the current sending window size of the target data flow according to the current congestion window size of the target data flow, and sending the data packet based on the updated current sending window size.

[0105] The sending window size refers to the size of the sending window. The sending window is a sliding window of the data sending end. When the data packet is located in the sending window, the data sending end sends the data packet. The current sending window size of the target data stream refers to the window size of the sending window when sending the data packet of the target data stream at the current time.

[0106] Specifically, the congestion window setting instruction also includes data flow information of the target data flow. The data sending end can determine the data flow identifier based on the data flow information, and store the corresponding relationship between the data flow identifier and the target congestion window size in the data flow window table to update the current congestion window size of the target data flow. When a data packet of the data flow is to be sent, the data sending end obtains the data flow identifier of the target data flow from the data packet to be sent, queries the current congestion window size of the target data flow from the updated data flow window table based on the data flow identifier, updates the current sending window size of the target data flow according to the queried current congestion window size, and sends the data packet to be sent based on the updated current sending window size of the target data flow.

[0107] In some embodiments, the data packet may include a data flow identifier, and the data transmitter may obtain the data flow identifier from the data packet and query the corresponding congestion window size from the data flow window table according to the data flow identifier.

[0108] In some embodiments, the data transmitting end may determine the queried current congestion window size as the updated current sending window size. Alternatively, the data transmitting end obtains the window size of the receiving window of the receiving end of the target data stream, and uses the smaller one of the window size of the receiving window and the queried current congestion window size as the updated current sending window size.

[0109] In some embodiments, the data sending end may run a background process, and the background process is used to receive the congestion window setting instruction sent by the control device. For example, the data sending end may continuously run the background process in the user mode. Figure 5 As shown, the background process receives a congestion window setting instruction.

[0110] In some embodiments, the kernel state of the data transmitting end has a kernel program. The background process of the data transmitting end obtains data flow information and target congestion window size from the congestion window setting instruction, and passes the data flow information and target congestion window size to the kernel program. The kernel program obtains the data flow identifier according to the data flow information, and stores the corresponding relationship between the data flow identifier and the target congestion window size in the data flow window table. Among them, the kernel program can be implemented by e-BPF (extended Berkeley Packet Filter) technology, or it can be implemented by other kernel bypass (Kernel Bypass) technology.

[0111] In some embodiments, Figure 5In the embodiment, when the kernel of the data sending end needs to send a new data packet, the kernel program in the data sending end can analyze the data packet, obtain the data flow identifier of the data flow to which the new data packet belongs from the data packet, and query the corresponding current congestion window size from the data flow window table according to the obtained data flow identifier, and determine to update the current sending window size according to the queried current congestion window size, determine the current sending window according to the updated current sending window size, and send the new data packet when the new data packet is within the current sending window. Among them, the data sending end can adjust the congestion window in the kernel through the bpf_setsockopt function, that is, through the bpf_setsockopt function, determine the current congestion window size of the data flow to which the new data packet belongs according to the queried current congestion window size.

[0112] In some embodiments, the data transmitting end can also determine the update time, which refers to the time for storing the target congestion window size in the data flow window, or the time for updating the current congestion window size. The data transmitting end uses the data flow identifier as a key and generates a key-value pair with (target congestion window size, update time) as a value, and stores the key-value pair in the data flow window table. The data transmitting end can periodically clean up the key-value pairs in the data flow window table according to the update time. Specifically, the data transmitting end can periodically clean up the corresponding relationship between the data flow identifier and (congestion window size, update time) stored in the data flow window table to avoid the old data from affecting the transmission of the data flow. Specifically, the data transmitting end can calculate the time difference between the update time and the current time, and when the time difference is greater than the time difference threshold, delete the corresponding relationship between the data flow identifier and (congestion window size, update time). The period for the data transmitting end to clean up data can be set as needed. For example, the period for the data transmitting end to clean up data can be consistent with the period for the network device to perform congestion detection.

[0113] In this embodiment, the current sending window size of the target data flow is updated according to the current congestion window size of the target data flow, and data packets are sent based on the updated current sending window size. Since the size of the sending window is positively correlated with the sending rate, the sending rate is adjusted through the window size of the congestion window, thereby achieving congestion control.

[0114] In some embodiments, Figure 6 As shown, a congestion control method is provided, which can be executed by a terminal or a server, or can be executed by a terminal and a server together. Figure 1 Taking the network device (such as network device 1 or network device 2) in the example as an example, the following steps are included:

[0115] Step 602: Count the data packets to be forwarded of the multiple data flows to obtain statistical results.

[0116] Each data stream is transmitted from a data transmitter to a network device. For example, different data streams are sent to a network device from different data transmitters. The data stream sent by the data transmitter to the network device may be generated by the data transmitter itself. The data stream sent by the data transmitter to the network device may also be a data stream forwarded by the data transmitter.

[0117] The data packets to be forwarded refer to the data packets to be forwarded by the network device. The multiple data streams are data streams used for transmission by the network device. The multiple data streams refer to at least two data streams, for example, five data streams.

[0118] Specifically, after the network device receives the data packet sent by the data sending end, the data packet is stored in the data buffer area, and the network device reads the data packet from the data buffer area for sending. The data buffer area can be implemented with a queue. The network device can store the data packets of each of the multiple data streams in the same data buffer area, or the network device can store the data packets of different data streams in different data buffer areas respectively. For example, if the multiple data streams are Q1, Q2 and Q4, the data packet of Q1 is stored in data buffer area 1, the data packet of Q2 is stored in data buffer area 2, and the data packet of Q4 is stored in data buffer area 4.

[0119] In some embodiments, the network device may periodically count the number of packets to be forwarded to obtain a statistical result. The statistical result includes the number of packets to be forwarded corresponding to each of the multiple data flows. Specifically, for each data flow, the network device counts the number of packets to be forwarded of the data flow, and uses the counted number as the number of packets to be forwarded of the data flow.

[0120] Step 604: When it is determined based on the statistical result that the network device is congested, congestion feedback information is generated.

[0121] The congestion of the network device may be that at least one port in the network device is congested. The congestion feedback information may include data flow information of each data flow. The data flow information includes a data flow identifier, and the data flow identifier includes a source IP address, a source port number, a destination IP address, and a destination port number. The data flow information may also include at least one of a transmission protocol, a round-trip delay, a number of bytes in transit, or a number of packet losses. The round-trip delay may be, for example, SRTT (smooth RTT).

[0122] Specifically, the statistical result includes the number of data flows to be forwarded respectively. When at least one of the multiple data flows to be forwarded is greater than the number threshold, the network device determines that congestion occurs. For example, when any one of the multiple data flows to be forwarded is greater than the number threshold, the network device determines that congestion occurs.

[0123] In some embodiments, the data flow information of the data flow may be generated based on each data packet to be forwarded of the data flow. The data packet includes a data flow identifier and may also include at least one of a transmission protocol, a round-trip delay, a number of bytes in transit, or a number of packet losses. Specifically, for each data flow, the network device may copy each data packet to be forwarded of the data flow to obtain each copied data packet. The network device may use each copied data packet as the data flow information. Alternatively, in order to reduce the bandwidth occupied by sending the data flow information, the network device may simplify the copied data packet to obtain a simplified data packet. For example, the network device may extract the content required for scheduling from the copied data packet to obtain a simplified data packet. The simplified data packet includes a data flow identifier and may also include at least one of a transmission protocol, a round-trip delay, a number of bytes in transit, or a number of packet losses.

[0124] In some embodiments, the network device may determine the data flow information of each data flow, and generate congestion feedback information based on the data flow information of each data flow. The congestion feedback information may include the data flow information of each data flow. Alternatively, the network device may determine the transmission bandwidth of each data flow and determine the data flow information of each data flow, and generate congestion feedback information based on the transmission bandwidth of each data flow and the data flow information of each data flow. The congestion feedback information may include the data flow information of each data flow and the transmission bandwidth of each data flow. The transmission bandwidth of a data flow refers to the bandwidth actually used by the data transmitter to transmit the data flow to the network device. As shown in Table 1, congestion feedback information is shown.

[0125] Table 1 Congestion feedback information

[0126]

[0127]

[0128] It should be noted that in Table 1, the data flow information set includes multiple data flow information, and each data flow information includes its own source IP, destination IP, source port, destination port, SRTT, number of bytes in transit and number of packet losses.

[0129] Step 606, sending congestion feedback information to the control device; the congestion feedback information is used to instruct the control device to determine the quality characterization value of each of the multiple data streams, and update the current congestion window size of the data stream based on each quality characterization value, so that the quality characterization values ​​of the multiple data streams change in a direction tending to be consistent.

[0130] Specifically, the control device determines the quality characterization value of each of the multiple data streams transmitted by the network device in response to the congestion feedback information reported by the network device. Based on the quality characterization values ​​of each of the multiple data streams, the control device determines the target data stream whose congestion window size needs to be updated from the multiple data streams, and determines the target congestion window size of the target data stream. Based on the target congestion window size of the target data stream, the control device generates a congestion window setting instruction for the target data stream, and sends the congestion window setting instruction for the data stream to the data sending end of the target data stream.

[0131] In some embodiments, the congestion window setting instruction includes a target congestion window size of the target data stream. The data transmitter updates the current congestion window size of the target data stream based on the target congestion window size of the target data stream, so that the quality characterization values ​​of the multiple data streams change in a direction that tends to be consistent.

[0132] In the above congestion control method, the data packets to be forwarded of each of the multiple data streams are counted to obtain statistical results. When it is determined based on the statistical results that the network device is congested, congestion feedback information is generated and sent to the control device. Due to the congestion feedback information, it is used to instruct the control device to determine the quality characterization values ​​of each of the multiple data streams, and update the current congestion window size of the data stream based on each quality characterization value, so that the quality characterization values ​​of each of the multiple data streams change in a direction that tends to be consistent. Thereby, while alleviating congestion, the fairness of the user experience quality is improved, that is, the QOE fairness is improved. And the overall user experience quality of the network of the data center can be improved, and the overall performance of the network of the data center can be improved. In addition, since the congestion feedback information is reported when the network device is congested, that is, when the communication link becomes a bottleneck link, the burden on the network caused by reporting the congestion feedback information is minimized, and the occupation of network resources is reduced.

[0133] In some embodiments, counting the data packets to be forwarded of each of the multiple data streams to obtain statistical results includes: for each data stream, counting the number of data packets to be forwarded of the data stream to obtain the number of data streams to be forwarded; obtaining statistical results based on the number of data streams to be forwarded; the method also includes: determining that congestion occurs in the network device when at least one of the multiple numbers to be forwarded is greater than a target number threshold.

[0134] The target quantity threshold is determined according to the bandwidth of the network device. For example, the target quantity threshold is positively correlated with the bandwidth of the network device.

[0135] Specifically, the network device may compare different to-be-forwarded quantities with quantity thresholds in turn, and when the to-be-forwarded quantity is greater than the target quantity threshold, it is determined that congestion occurs in the network device.

[0136] In this embodiment, when at least one of the multiple numbers to be forwarded is greater than the target number threshold, it is determined that the network device is congested, so that congestion control can be performed on the network device without the network device being too congested, thereby improving the timeliness of congestion control.

[0137] In some embodiments, the method further includes: determining a bandwidth of the network device; determining a target quantity threshold based on the bandwidth of the network device; and the target quantity threshold is positively correlated with the bandwidth of the network device.

[0138] Specifically, the network device can determine the target quantity threshold value according to the bandwidth of the network device, and the target quantity threshold value is positively correlated with the bandwidth of the network device. In the present application, the specific method for determining the target quantity threshold value according to the bandwidth of the network device is not limited. For example, the network device can obtain a preset bandwidth and a preset quantity. The preset bandwidth and the preset quantity can be set based on experience. The network device can calculate the ratio of the bandwidth of the network device to the preset bandwidth, and use the result obtained by multiplying the ratio by the preset quantity as the target quantity threshold value.

[0139] In some embodiments, a bandwidth quantity relationship table may be stored in the network device, and the bandwidth quantity relationship table stores quantity thresholds corresponding to multiple bandwidths. In the bandwidth quantity relationship table, the larger the bandwidth, the larger the quantity threshold corresponding to the bandwidth. The network device can query the quantity threshold corresponding to the bandwidth of the network device from the bandwidth quantity relationship table to obtain the target quantity threshold.

[0140] In some embodiments, the network device may include a detection module, and the various embodiments implemented by the above network device may be implemented by the network device through the detection module. Figure 7 As shown, a flow chart of congestion detection of the detection module is shown. Figure 7In the above, "regularly checking the current queue status" means that the detection module regularly counts the number of data packets to be forwarded to obtain statistical results. The threshold in "queue exceeds threshold" refers to the target number threshold, and "queue exceeds threshold" means that the number of data packets to be forwarded is greater than the target number threshold. The "data packets to be sent" in "mirror copy the data packets to be sent" refers to "data packets to be forwarded", and "mirror copy" means "copy". "Obtaining bandwidth information" means obtaining the transmission bandwidth of the data flow. "Performing preliminary processing of data packet information" means simplifying data packets. "Sending the processed information together with the bandwidth information to the controller" means the data flow information, and the "controller" means the control device. The detection module in the network device and the information analysis module in the control device constitute a congestion information acquisition system.

[0141] In this embodiment, since the target quantity threshold is positively correlated with the bandwidth of the network device, the rationality of the target quantity threshold is improved.

[0142] In some embodiments, Figure 8 As shown, a congestion control method is provided, which is described by taking the method applied to a control device, a network device and a data sending end as an example, and includes the following steps:

[0143] Step 802: for each data stream among the multiple data streams transmitted by the network device, the network device counts the number of data packets to be forwarded of the data stream to obtain the number of data packets to be forwarded of each of the multiple data streams.

[0144] The network device transmits the multiple data streams through the same communication link.

[0145] Step 804: When at least one of the multiple numbers to be forwarded is greater than the target number threshold, the network device generates congestion feedback information based on the data flow information of each data flow, and reports the congestion feedback information to the control device.

[0146] Among them, the target quantity threshold is positively correlated with the bandwidth of the network device.

[0147] Step 806: The control device determines the quality characterization value of each of the multiple data streams in response to the congestion feedback information, calculates the average of the quality characterization values ​​of the multiple data streams to obtain an average characterization value, and for each data stream, if the quality characterization value of the data stream and the average characterization value are inconsistent, determines the data stream as a target data stream whose congestion window size needs to be updated, determines the ratio of the average characterization value to the quality characterization value of the target data stream as a window adjustment coefficient, multiplies the window adjustment coefficient by the current congestion window size of the target data stream, and determines the result of the operation as the target congestion window size of the target data stream.

[0148] Step 808: The control device generates a congestion window setting instruction for the target data flow based on the target congestion window size of the target data flow, and sends the congestion window setting instruction for the target data flow to the data sending end corresponding to the target data flow.

[0149] For example, if the network device is Figure 1 The control device controls the network device 1 in the control unit, and determines the target congestion window sizes of data stream 1 and data stream 2, then the control device sends a congestion window device instruction to the data sending end 1_1 corresponding to data stream 1, and the congestion window device instruction includes the target congestion window size of data stream 1. The control device sends a congestion window device instruction to the data sending end 1_2 corresponding to data stream 2, and the congestion window device instruction includes the target congestion window size of data stream 2.

[0150] Step 810, the data sending end receives the congestion window setting instruction for the target data flow issued by the control device, updates the current congestion window size of the target data flow based on the target congestion window size of the target data flow, and when a data packet of the target data flow is to be sent, sends the data packet according to the updated current congestion window size of the target data flow.

[0151] In this embodiment, for each data flow, when the quality characterization value and the average characterization value of the data flow are inconsistent, the data flow is determined as a target data flow whose congestion window size needs to be updated, the ratio of the average characterization value to the quality characterization value of the target data flow is determined as a window adjustment coefficient, the window adjustment coefficient and the current congestion window size of the target data flow are multiplied, and the result of the operation is determined as the target congestion window size of the target data flow. Thus, when the quality characterization value of the data flow is less than the average characterization value, the current congestion window size of the data flow can be updated in an increasing direction, and when the quality characterization value of the data flow is greater than the average characterization value, the current congestion window size of the data flow can be updated in an increasing direction. Since the congestion window size is positively correlated with the quality characterization value, the quality characterization values ​​of each data flow tend to be consistent, that is, tend to the average characterization value, by adjusting the size of the current congestion window size, so that in the process of congestion control, not only congestion is alleviated but also QOE fairness is achieved.

[0152] With the rise of cloud services, more and more services are deployed in cloud data centers and provide services to the outside world. There are many types of services in data centers, and different services and even different requests for the same service will have different impacts on user experience. For example, in a data center network, the transmission time of a large file increases from 10s to 11s, which is almost lossless to the user experience. However, for video viewing applications, the video freeze time increases from 0.1s to 1s, which will seriously affect the user experience. For another example, for the same video viewing application, if the picture quality drops from 1080p to 480p, users who watch on mobile phones will get a better user experience than those who watch on 4k TVs. However, in the design of traditional congestion algorithms, the impact of user experience is not taken into account. Traditional congestion algorithms seek to ensure that each flow can obtain the same bandwidth under the premise of fully utilizing bandwidth resources and not generating congestion to ensure the fairness of the algorithm. This goal allows requests that originally require more bandwidth to ensure user experience to obtain the same bandwidth as requests that do not require too much bandwidth to allow users to obtain a good experience, resulting in the failure to maximize the overall user experience quality of the data center. The congestion control method provided in the present application can be applied to a cloud data center in a private cloud or a public cloud to perform congestion control on the network in the data center, thereby alleviating congestion while improving QOE fairness.

[0153] The congestion control method provided in the present application can optimize the QOE fairness for the data flow in the bottleneck link. For the data flow in the non-bottleneck link, the traditional congestion algorithm can be used, which not only ensures the efficiency of the algorithm but also allows the algorithm to have global information to improve the overall QOE of the network.

[0154] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0155] Based on the same inventive concept, the embodiment of the present application also provides a congestion control device for implementing the congestion control method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more congestion control device embodiments provided below can refer to the limitations of the congestion control method above, and will not be repeated here.

[0156] In some embodiments, Fig. 9 As shown, a congestion control device is provided, including: a feedback instruction response module 902, a window size determination module 904 and a setting instruction sending module 906, wherein:

[0157] The feedback instruction response module 902 is used to determine the quality characterization value of each of the multiple data streams transmitted by the network device in response to the congestion feedback information reported by the network device; the quality characterization value of the data stream is used to characterize the experience quality of the data stream.

[0158] The window size determination module 904 is used to determine a target data flow whose congestion window size needs to be updated from the multiple data flows based on the quality characterization values ​​of the multiple data flows, and determine a target congestion window size of the target data flow.

[0159] The setting instruction sending module 906 is used to send a congestion window setting instruction for the target data flow to the data sending end of the target data flow based on the target congestion window size of the target data flow; the congestion window setting instruction is used to instruct the data sending end to update the current congestion window size of the data flow based on the target congestion window size, so that the quality characterization values ​​of each of the multiple data flows change in a direction tending to be consistent.

[0160] In some embodiments, the window size determination module 904 is also used to calculate the mean of the quality characterization values ​​of multiple data streams to obtain an average characterization value; for each data stream, when the quality characterization value of the data stream and the average characterization value are inconsistent, the data stream is determined as a target data stream for which the size of the congestion window needs to be updated.

[0161] In some embodiments, the window size determination module 904 is further used to determine a window adjustment coefficient based on the quality characterization value and the average characterization value of the target data flow; and determine the target congestion window size of the target data flow according to the window adjustment coefficient and the current congestion window size of the target data flow.

[0162] In some embodiments, the congestion feedback information includes data flow information of each of the multiple data flows; the setting instruction sending module 906 is also used to generate a congestion window setting instruction for the target data flow based on the data flow information of the target data flow and the target congestion window size of the target data flow; the data flow information and the target congestion window size are used by the data sending end of the target data flow to update the current congestion window size of the target data flow.

[0163] In some embodiments, the feedback instruction response module 902 is further used to determine, for each data stream, a preset quality assessment strategy that matches the data stream from a preset quality assessment strategy set; and determine a quality characterization value of the data stream according to the preset quality assessment strategy that matches the data stream.

[0164] In some embodiments, Fig.10 As shown, a congestion control device is provided, including: a setting instruction receiving module 1002 and a window size updating module 1004, wherein:

[0165] The setting instruction receiving module 1002 is used to receive the congestion window setting instruction for the target data flow issued by the control device; the congestion window setting instruction includes the target congestion window size of the target data flow, and the target congestion window size is determined based on the quality characterization values ​​of each of the multiple data flows transmitted by the network device. The target data flow is transmitted from the data sending end to the network device.

[0166] The window size updating module 1004 is used to update the current congestion window size of the target data flow based on the target congestion window size of the target data flow, so that the quality characterization values ​​of the multiple data flows change in a direction tending to be consistent.

[0167] In some embodiments, the current congestion window size of the target data stream is stored in the data stream window table; the window size update module 1004 is also used to update the current congestion window of the target data stream stored in the data stream window table according to the target congestion window size of the target data stream; when a data packet of the target data stream is to be sent, the current congestion window size of the target data stream is queried from the updated data stream window table; the current sending window size of the target data stream is updated according to the current congestion window size of the target data stream, and the data packet is sent based on the updated current sending window size.

[0168] In some embodiments, Fig.11 As shown, a congestion control device is provided, including: a statistical module 1102, a feedback instruction generating module 1104 and a feedback instruction sending module 1106, wherein:

[0169] The statistical module 1102 is used to collect statistics on the data packets to be forwarded of the multiple data flows to obtain statistical results.

[0170] The feedback instruction generating module 1104 is used to generate congestion feedback information when it is determined based on the statistical result that the network device is congested.

[0171] The feedback instruction sending module 1106 is used to send congestion feedback information to the control device; the congestion feedback information is used for the control device to determine the quality characterization value of each of the multiple data streams, and update the current congestion window size of the data stream based on each quality characterization value, so that the quality characterization values ​​of the multiple data streams change in a direction tending to be consistent.

[0172] In some embodiments, the statistical module 1102 is also used to count the number of data packets to be forwarded in the data flow for each data flow, so as to obtain the number of data packets to be forwarded in the data flow; and obtain statistical results based on the number of data packets to be forwarded of each of the multiple data flows; the device is also used to determine that congestion occurs in the network device when at least one of the multiple numbers to be forwarded is greater than the target number threshold.

[0173] In some embodiments, the apparatus is further configured to determine a bandwidth of a network device; determine a target quantity threshold based on the bandwidth of the network device; and the target quantity threshold is positively correlated with the bandwidth of the network device.

[0174] Each module in the above congestion control device can be implemented in whole or in part by software, hardware or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0175] In some embodiments, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Fig.12 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data involved in the congestion control method. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a congestion control method is implemented.

[0176] In some embodiments, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Fig.13 As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface, the display unit and the input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a congestion control method is implemented. The display unit of the computer device is used to form a visually visible image, and can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covered on the display screen, or a button, trackball or touchpad set on the computer device casing, or an external keyboard, touchpad or mouse, etc.

[0177] Those skilled in the art will understand that Fig.12 and Fig.13 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0178] In some embodiments, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above congestion control method when executing the computer program.

[0179] In some embodiments, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned congestion control method are implemented.

[0180] In some embodiments, a computer program product is provided, comprising a computer program, which implements the steps in the above congestion control method when executed by a processor.

[0181] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0182] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0183] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0184] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A congestion control method, characterized in that: Applied to a control device, the method comprises: In response to congestion feedback information reported by the network device, determine a quality characterization value of each of a plurality of data streams transmitted by the network device; the quality characterization value of the data stream is used to characterize the experience quality of the data stream; Based on the quality characterization values ​​of the multiple data flows, determining a target data flow for which a congestion window size needs to be updated from the multiple data flows, and determining a target congestion window size of the target data flow; Based on the target congestion window size of the target data stream, a congestion window setting instruction for the target data stream is sent to a data sending end of the target data stream; the congestion window setting instruction is used to instruct the data sending end to update the current congestion window size of the target data stream based on the target congestion window size, so that the quality characterization values ​​of each of the multiple data streams change in a direction tending to be consistent.

2. The method according to claim 1, characterized in that The determining, based on the quality characterization values ​​of the multiple data flows, of the target data flow for which the size of the congestion window needs to be updated from the multiple data flows comprises: Performing mean calculation on the quality characterization values ​​of the multiple data streams to obtain an average characterization value; For each of the data flows, when the quality characterization value of the data flow is inconsistent with the average characterization value, the data flow is determined as a target data flow for which the size of the congestion window needs to be updated.

3. The method according to claim 2, characterized in that Determining the target congestion window size of the target data flow includes: Determining a window adjustment coefficient based on the quality characterization value of the target data stream and the average characterization value; A target congestion window size of the target data flow is determined according to the window adjustment coefficient and the current congestion window size of the target data flow.

4. The method according to claim 1, characterized in that: The congestion feedback information includes data flow information of each of the multiple data flows; The sending of a congestion window setting instruction for the target data flow to a data transmitting end of the target data flow based on the target congestion window size of the target data flow comprises: Based on the data flow information of the target data flow and the target congestion window size of the target data flow, a congestion window setting instruction for the target data flow is generated; the data flow information and the target congestion window size are used by the data sending end of the target data flow to update the current congestion window size of the target data flow.

5. The method according to claim 1, characterized in that Determining the quality characterization value of each of the plurality of data streams transmitted by the network device comprises: For each of the data streams, determining a preset quality assessment strategy that matches the data stream from a preset quality assessment strategy set; According to a preset quality assessment strategy matched by the data stream, a quality characterization value of the data stream is determined.

6. A congestion control method, characterized in that: Applied to a data sending end, the method comprises: A congestion window setting instruction for a target data stream issued by a receiving control device; the congestion window setting instruction includes a target congestion window size of the target data stream, the target congestion window size is determined based on the quality characterization values ​​of each of a plurality of data streams transmitted by the network device, and the target data stream is transmitted by the data transmitting end to the network device; The current congestion window size of the target data flow is updated based on the target congestion window size of the target data flow, so that the quality representation values ​​of the multiple data flows change in a direction tending to be consistent.

7. The method according to claim 6, characterized in that The current congestion window size of the target data stream is stored in the data stream window table; The updating of the current congestion window size of the target data flow based on the target congestion window size of the target data flow comprises: According to the target congestion window size of the target data flow, updating the current congestion window size of the target data flow stored in the data flow window table; When a data packet of the target data stream is to be sent, querying the current congestion window size of the target data stream from the updated data stream window table; The current sending window size of the target data flow is updated according to the current congestion window size of the target data flow, and the data packet is sent based on the updated current sending window size.

8. A congestion control method, characterized in that: Applied to a network device, the method comprises: Counting the data packets to be forwarded of each of the multiple data flows to obtain statistical results; In the case where it is determined based on the statistical result that congestion occurs in the network device, generating congestion feedback information; The congestion feedback information is sent to a control device; the congestion feedback information is used to instruct the control device to determine a quality characterization value of each of the multiple data streams, and to update a current congestion window size of the data stream based on each of the quality characterization values, so that the quality characterization values ​​of each of the multiple data streams change in a direction tending to be consistent.

9. The method according to claim 8, characterized in that The statistical result obtained by counting the data packets to be forwarded of the plurality of data flows comprises: For each of the data flows, counting the number of data packets to be forwarded of the data flow to obtain the number of data packets to be forwarded of the data flow; Obtaining statistical results based on the number of to-be-forwarded data flows; The method further comprises: When at least one of the multiple numbers to be forwarded is greater than a target number threshold, it is determined that congestion occurs in the network device.

10. The method according to claim 9, characterized in that The method further comprises: determining a bandwidth of the network device; The target quantity threshold is determined based on the bandwidth of the network device; the target quantity threshold is positively correlated with the bandwidth of the network device.

11. A congestion control device, characterized in that: The device comprises: A feedback instruction response module, configured to determine, in response to congestion feedback information reported by a network device, a quality characterization value of each of a plurality of data streams transmitted by the network device; the quality characterization value of the data stream is used to characterize the experience quality of the data stream; A window size determination module, configured to determine a target data stream for which a congestion window size needs to be updated from the multiple data streams based on the quality characterization values ​​of the multiple data streams, and determine a target congestion window size of the target data stream; A setting instruction sending module is used to send a congestion window setting instruction for the target data stream to the data sending end of the target data stream based on the target congestion window size of the target data stream; the congestion window setting instruction is used to instruct the data sending end to update the current congestion window size of the data stream based on the target congestion window size, so that the quality characterization values ​​of each of the multiple data streams change in a direction tending to be consistent.

12. A congestion control device, characterized in that: The device comprises: A setting instruction receiving module, used for receiving a congestion window setting instruction for a target data flow issued by a control device; the congestion window setting instruction includes a target congestion window size of the target data flow, the target congestion window size is determined based on the quality characterization values ​​of each of a plurality of data flows transmitted by the network device, and the target data flow is transmitted from a data transmitting end to the network device; The window size updating module is used to update the current congestion window size of the target data flow based on the target congestion window size of the target data flow, so that the quality characterization values ​​of the multiple data flows change in a direction tending to be consistent.

13. A congestion control device, characterized in that: The device comprises: A statistics module is used to collect statistics of the data packets to be forwarded by each of the multiple data flows to obtain statistical results; A feedback instruction generating module, configured to generate congestion feedback information when it is determined based on the statistical result that the network device is congested; A feedback instruction sending module is used to send the congestion feedback information to a control device; the congestion feedback information is used by the control device to determine the quality characterization value of each of the multiple data streams, and to update the current congestion window size of the data stream based on each of the quality characterization values, so that the quality characterization values ​​of each of the multiple data streams change in a direction tending to be consistent.

14. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 10 are implemented.

15. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 10 are implemented.

16. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 10 are implemented.