WO2023246616A1 - Network congestion control method and apparatus - Google Patents

Abstract

Provided in the present application are a network congestion control method and apparatus. The network congestion control method is applied to an entry device of a campus network, and the entry device is used for being connected to an operator network. The method comprises: acquiring a real-time bandwidth between an entry device and an operator network; and if a data transmission rate represented by the real-time bandwidth between the entry device and the operator network is greater than or equal to a data transmission rate represented by a preset system threshold value, performing packet loss processing on a TCP message from the operator network, wherein the system threshold value is obtained according to a subscribed bandwidth between the campus network and the operator network. In the present application, packet loss processing is performed on a TCP message, which is from an operator network, at an entry device of a campus network, such that the throughput of the entry device is reduced, thereby avoiding network congestion of the campus network that is caused by a real-time bandwidth between the entry device and the operator network being greater than a subscribed bandwidth therebetween, such that the independence and controllability of the campus network is improved, and the dependence of the campus network on the operator network is alleviated.

Description

Network congestion control method and device

This application claims priority to the Chinese patent filed with the China Patent Office on June 23, 2022, with application number 202210726633.5 and the invention title "Network Congestion Control Method and Device", the entire content of which is incorporated into this application by reference.

Technical field

The present application relates to the field of communication technology, and in particular, to a network congestion control method and device.

Background technique

Campus network refers to the office LAN established for enterprises or organizations to realize digital office, production and operation, such as campus network, community network, enterprise and institution network, etc. With the gradual enrichment of campus services, the campus network carries more traffic. When the transmission traffic of the campus network in a unit time exceeds the transmission traffic limited by the contracted bandwidth between the campus network and the operator, network congestion will occur on the campus network. This in turn affects the normal work of users within the campus network.

Contents of the invention

In view of the above, it is necessary to provide a network congestion control method and device. When the data transmission rate of the campus network exceeds the preset data transmission rate, the transmission control protocol TCP packets from the operator network are discarded through the campus network. , to reduce the data transmission rate of the campus network and avoid network congestion in the campus network.

In a first aspect, embodiments of the present application provide a network congestion control method, which is applied to an entrance device of a campus network. The entrance device is used to connect to an operator network. The method includes: obtaining the information between the entrance device and the operator network. The real-time bandwidth between the operator network; if the data transmission rate represented by the real-time bandwidth between the entrance device and the operator network is greater than or equal to the data transmission rate represented by the preset system threshold, then the data transmission rate from The TCP packets of the operator network are subjected to packet loss processing, wherein the system threshold is obtained based on the contracted bandwidth between the campus network and the operator network. Using the above technical solution, the data transmission rate represented by the real-time bandwidth between the campus network and the operator network is greater than the data transmission rate represented by the preset system threshold, and the campus network side actively processes TCP packets from the operator network. Packet loss processing reduces the throughput between the campus network and the operator network by discarding TCP packets, so that the data transmission rate represented by the real-time bandwidth between the campus network and the operator network is within the preset system Within the data transmission rate range represented by the threshold, it avoids network congestion caused by limited bandwidth on the campus network. At the same time, TCP packets are actively discarded through the campus network, and the independence and controllability of the campus network reduce the impact of the campus network on operations. dependence on business networks.

In a possible implementation of the above first aspect, the packet loss processing of TCP messages from the operator network includes: obtaining the priority information of the TCP message; based on the priority information Perform differentiated packet loss processing on TCP packets from the operator's network. Using the above technical solution, different packet loss processing methods can be adopted for TCP messages with different priority information. For example, packet loss processing is not performed on TCP messages with higher priority, but only TCP messages with lower priority are discarded. deal with. In scenarios where bandwidth is limited, services with different needs are provided for different users in the campus network.

In a possible implementation of the above first aspect, obtaining the priority information of the TCP message includes: obtaining a message identifier of the TCP message, wherein the message identifier is used to uniquely identify The TCP message; determines the priority information of the TCP message according to the message identifier.

The above technical solution is adopted to identify TCP messages through message identifiers, and determine the priority information of each TCP message through message identifiers, so as to differentiate TCP messages from the operator network based on the priority information. Packet loss handling. The message identifier can be an IP address or a MAC address.

In a possible implementation of the above first aspect, performing differentiated packet loss processing on TCP messages from the operator network based on the priority information includes: determining based on the priority information that packets are sent from the operator network. The packets to be processed in the TCP packets of the operator network are processed; the packets to be processed are discarded, wherein the priority of the packets to be processed is lower than all the packets from the operator network. The priority of other TCP packets in the TCP packet.

Using the above technical solution, the entrance device of the campus network only discards TCP packets with low priority, and forwards TCP packets with high priority directly to the corresponding client, and because TCP packets with high priority Corresponding to high-priority users, TCP packets of high-priority users are directly forwarded to ensure the user experience of high-priority users in scenarios with limited bandwidth.

In a possible implementation of the first aspect, the packets to be processed include multiple TCP packets, and the packet loss processing for the packets to be processed includes: based on the priority of the packets to be processed. The level information determines the packet loss parameter of each TCP message in the message to be processed; and performs packet loss processing on the corresponding TCP message in the message to be processed according to the packet loss parameter.

Using the above technical solution, packet loss processing is performed on the corresponding TCP packets according to the packet loss parameters corresponding to the priority information, so as to achieve differentiated processing of various TCP packets in the packets to be processed.

Further, the packet loss parameter may be a packet loss rate or a number of packets lost, and TCP messages with different priority information correspond to different packet loss parameters.

In a possible implementation of the above first aspect, performing differentiated packet loss processing on TCP packets from the operator network based on the priority information includes: determining based on the priority information that packets from the Packet loss parameters of TCP messages from the operator network; packet loss processing is performed on TCP messages from the operator network according to the packet loss parameters, where different priority information corresponds to different packet loss parameters.

Using the above technical solution, all TCP messages from the operator network are discarded. Since different priority information corresponds to different packet loss parameters, TCP messages with different priority information are processed through different packet loss parameters. Packet loss processing to implement differentiated packet loss processing for TCP packets from the operator network.

In a possible implementation of the above first aspect, the entrance device stores a packet loss sliding window, and performs differentiated packet loss processing on TCP messages from the operator network based on the priority information. It includes: performing differentiated packet loss processing on TCP messages from the operator network according to the packet loss sliding window.

Using the above technical solution, differentiated packet loss processing is performed on TCP messages from the operator network through the preset packet loss sliding window in the entrance device.

In a possible implementation of the above first aspect, the packet loss processing of the TCP message according to the packet loss sliding window includes: receiving the size of the packet loss sliding window from the operator network. TCP messages; perform packet loss processing on TCP messages with the packet loss sliding window size according to the preset packet loss parameters; if after the packet loss processing, the real-time bandwidth of the entrance device and the operator network The represented data transmission rate is greater than or equal to the data transmission rate represented by the system threshold, gradually increase the packet loss parameter, and slide the packet loss received from the operator network according to the increased packet loss parameter. TCP packets with a window size are discarded until the data transmission rate represented by the real-time bandwidth of the ingress device and the operator network is less than the data transmission rate represented by the system threshold.

Using the above technical solution, in the process of discarding TCP messages with the sliding window size of packet loss from the operator network, by gradually increasing the packet loss parameters, the ingress device and the operator can be quickly connected. Network real-time bandwidth The data transmission rate characterized is reduced to less than the data transmission rate characterized by the system threshold.

In a possible implementation of the first aspect above, if the data transmission rate represented by the real-time bandwidth between the entrance device and the operator network changes by a data transmission rate greater than or equal to the system threshold, To ensure that the data transmission rate is less than the data transmission rate represented by the system threshold, the method further includes: gradually reducing the packet loss parameter, and analyzing the received TCP packets with the packet loss sliding window size based on the reduced packet loss parameter. Packet loss is performed until the packet loss parameter is 0, and then packet loss processing is no longer performed on TCP messages from the operator network.

Using the above technical solution, the data transmission rate represented by the real-time bandwidth between the entrance device and the operator network changes from a data transmission rate represented by being greater than or equal to the system threshold to being less than the system threshold. The data transmission rate is gradually reduced by gradually reducing the packet loss parameter, and packet loss is performed based on the reduced packet loss parameter.

In a possible implementation of the first aspect above, the method further includes: data transmission rate represented by the real-time bandwidth between the entrance device and the operator network and data represented by the system threshold. The packet loss parameters are adjusted according to the transmission rate; and the TCP messages from the operator network are packet lost according to the adjusted packet loss parameters.

In a possible implementation of the above first aspect, if the data transmission rate represented by the real-time bandwidth of the entrance device is greater than the data transmission rate represented by the preset system threshold, the data transmission rate from the operator network is Packet loss processing for TCP messages includes: if the system threshold is larger, the data transmission rate represented by the system threshold is smaller, and when the actual bandwidth is greater than or equal to the system threshold, the packet loss processing from the operator network is The TCP message is subjected to packet loss processing; if the system threshold is larger, the data transmission rate represented by the system threshold is smaller. When the absolute value of the difference between the contracted bandwidth and the actual bandwidth is less than or equal to the system threshold, perform packet loss processing on the TCP packets from the operator network.

Specifically, when the system threshold is larger and the data transmission rate represented by the system threshold is smaller, the data transmission rate represented by the real-time bandwidth between the entrance device and the operator network is greater than or equal to the preset value. The data transmission rate represented by the system threshold is: the actual bandwidth is greater than or equal to the system threshold.

When the system threshold is larger, the data transmission rate represented by the system threshold is smaller, and the data transmission rate represented by the real-time bandwidth between the entrance device and the operator network is greater than or equal to the preset system threshold. The characterized data transmission rate is: the absolute value of the difference between the contracted bandwidth and the actual bandwidth is less than or equal to the system threshold. In a possible implementation of the above first aspect, the data transmission rate represented by the packet forwarding performance of the ingress device is greater than the data transmission rate represented by the contracted bandwidth.

In the second aspect, embodiments of the present application provide an ingress device, which includes a receiving ingress module, a traffic detection module, a backpressure control module, a sliding packet loss module, a forwarding module, and an egress forwarding module.

Among them, the receiving portal module is used to receive data sent from the operator's network, and the traffic detection module is used to detect the amount of data transmitted by the portal device within a preset time to obtain the real-time bandwidth between the portal device and the operator's network; if the portal The data transmission rate represented by the real-time bandwidth of the device is less than the data transmission rate represented by the system threshold. The forwarding module is used to forward the data sent from the operator network to the egress forwarding module, and the egress forwarding module is used to forward the data sent by the forwarding module; If the data transmission rate represented by the real-time bandwidth of the ingress device is greater than or equal to the data transmission rate represented by the system threshold, the backpressure control module and the sliding packet loss module cooperate to perform packet loss processing on the TCP packets from the operator network. And forward the TCP message after packet loss processing to the egress forwarding module, so that the egress forwarding module can forward the received data.

In a third aspect, embodiments of the present application provide an entry device, including: one or more processors; a storage device for storing one or more programs; when the one or at least one program is used by the one or more The processor executes, so that the one or more processors implement the network congestion control method according to any one of the first aspects.

In a fourth aspect, embodiments of the present application provide a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the network congestion control method as described in any one of the first aspects is implemented.

In a fifth aspect, embodiments of the present application provide a campus network, including an ingress device, where the ingress device is configured to perform any one of the network congestion control methods of the first aspect.

It should be understood that the technical effects brought by any design in the second to fifth aspects can refer to the beneficial effects in the corresponding methods provided above, and will not be described again here.

Description of the drawings

Figure 1 is a schematic diagram of the system architecture of a system network.

Figure 2 is a schematic flowchart of a network congestion control method provided by an embodiment of the present application.

Figure 3 is a schematic flowchart of a packet loss processing method provided by an embodiment of the present application.

Figure 4 is a schematic diagram of the sliding window packet loss processing provided by the embodiment of the present application.

Figure 5 is a schematic diagram of the packet loss parameters of the ingress device changing over time according to the embodiment of the present application.

Figure 6 is a schematic diagram of a state machine of an entrance device provided by an embodiment of the present application.

Figure 7 is a schematic module diagram of an entrance device provided by an embodiment of the present application.

Figure 8 is a hardware schematic diagram of the entrance device according to the embodiment of the present application.

Detailed ways

Hereinafter, the terms “first” and “second” are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity of indicated technical features. Thus, features defined as “first” and “second” may explicitly or implicitly include one or more of the described features. In the description of the embodiments of this application, words such as “exemplary” or “for example” are used to identify examples, illustrations or illustrations. Any embodiment or design described as "exemplary" or "such as" in the embodiments of the present application is not to be construed as preferred or advantageous over other embodiments or designs. Rather, use of the words "exemplary" or "such as" is intended to present the concept in a concrete manner.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terms used in the description of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. It will be understood that in this application, unless otherwise stated, "plurality" means two or more than two, and "and/or" includes any and all combinations of one or more of the associated listed items.

First, the technical terms involved in the embodiments of this application are introduced:

1. Bandwidth (Max net bitrate)

Bandwidth refers to the amount of data that can be transmitted within a unit of time (usually 1 second). The contracted bandwidth is the theoretical maximum bandwidth from the operator's access equipment to the user equipment stipulated in the broadband service agreement signed between the user (such as the campus network, the campus network is the user who signed a contract with the operator's network) and the operator. Reachable rate.

2. Throughput

Throughput refers to the number of data (measured in bits, bytes, packets, etc.) successfully transmitted per unit time on a network, device, port, virtual circuit, or other facility.

3. TCP congestion control

TCP congestion control is an algorithm used by Transmission Control Protocol (TCP) to avoid network congestion and is a congestion control measure. TCP congestion control uses a diverse network congestion control method based on the line increase product decrease mode (including slow start and congestion window modes) to control congestion.

For ease of understanding, the technical solutions in the embodiments of the present application are described below in conjunction with the accompanying drawings.

First, a schematic diagram of the system network architecture is introduced with reference to Figure 1. As shown in Figure 1, the system network includes the campus network and the operator network. The operator's network includes an egress device 210, and the operator's network communicates with the campus network through the egress device 210. For example, the operator network sends downlink data to the campus network through the egress device 210. Among them, the egress device 210 may be a router, a switch, and other devices.

The campus network includes an entrance device 110, a switch 130 and a client 150.

The campus network is connected to the operator network through the entrance device 110. The ingress device 110 may be an edge switch, an access switch, or other devices. The entrance device 110 is used to implement communication between the campus network and the operator's network. For example, the campus network receives downlink data sent by the operator network through the entrance device 110 or sends uplink data to the operator network. Among them, the ingress device 110 of the campus network in Figure 1 is directly connected to and communicates with the egress device 210 of the operator's network. It can be understood that in other embodiments, the ingress device 110 of the campus network can be connected to the egress device 210 of the operator network through other forwarding devices or forwarding networks. In this way, the campus network communicates with the operator network through other forwarding devices or forwarding networks. .

As shown in Figure 1, client 150 includes client 1, client 2, client 3 and client 4. It can be understood that this application only takes Figure 1 as an example, but is not limited thereto. The client 150 can be a mobile terminal, a notebook computer, a desktop computer, etc.

It can be understood that when the operator network communicates with the campus network, the operator network sends downlink data from the operator network to the ingress device 111 of the campus network through the egress device 210. The ingress device 110 forwards the downlink data to the corresponding switch 130, and the switch 130 forwards it to the client 150 of the campus network. The ingress device 110 also receives the uplink data forwarded from the client 150 by the switch 130, and forwards the uplink data to the egress device 210 of the operator network. The egress device 210 of the operator network receives the uplink data from the campus network.

Obviously, when the business volume demand of the client 150 in the campus network increases, the amount of data transmitted between the entrance device 110 and the exit device 210 of the campus network in a unit time exceeds the bandwidth limit of the contract between the campus network and the operator. the maximum amount of data, causing network congestion on the campus network. Network congestion will cause the network transmission performance of the campus network to degrade, affecting the normal operation of all services on the campus network.

In response to the above problems, this application provides a network congestion control method by detecting the real-time bandwidth between the entrance device 110 of the campus network and the operator network. When the data transmission speed represented by the real-time bandwidth between the ingress device 110 and the operator network reaches the preset data transmission rate, the ingress device 110 performs packet loss processing on the TCP packets from the operator network. By discarding TCP packets, the overall traffic of the campus network entrance device 110 is reversely controlled, thereby reducing the throughput of the campus network entrance device 110 and ensuring the connection between the campus network entrance device 110 and the operator's network. The real-time bandwidth is always within the contracted bandwidth range between the campus network and the operator to avoid network congestion on the campus network.

Specifically, the sliding window mechanism is usually used to transmit packets between the operator network and the campus network. When the ingress device 110 discards TCP packets from the operator network, the operator network will receive fewer acknowledgment messages from TCP packets sent by the campus network. In this way, the operator network will reduce the sending window of TCP messages according to TCP congestion control, so that less data is transmitted between the operator network and the campus network, and the throughput of the entrance device 110 of the campus network becomes smaller, which in turn reduces the campus network. The real-time bandwidth between the campus network and the operator's network becomes smaller, so that the real-time bandwidth between the entrance device 110 of the campus network and the operator's network is always within the contracted bandwidth between the campus network and the operator.

Please refer to Figure 2, which is a schematic flow chart of a network congestion control method provided by an embodiment of the present application. This network congestion control method can be applied to the entrance equipment of the campus network. The following takes the network congestion control method applied to the entrance device 110 of the campus network in Figure 1 as an example for detailed description.

S201. Obtain the real-time bandwidth between the portal device 110 and the operator's network.

Specifically, the amount of data transmitted by the entrance device 110 within a preset time can be detected, and the real-time bandwidth can be calculated based on the amount of data transmitted and the preset time. For example, the amount of data transmitted by the entrance device 110 of the campus network in 3 seconds is 300MB, and the actual bandwidth between the entrance device 110 and the operator's network is 300MB/3s=100Mb/s.

It can be understood that in the embodiment of the present application, the real-time bandwidth may also be referred to as the throughput of the portal device 110 . Real-time bandwidth may characterize the data transfer rate between the portal device 110 and the operator's network. The amount of transmitted data includes downlink data received by the ingress device 110 from the operator network and uplink data forwarded by the ingress device 110 to the operator network.

S202. If the data transmission rate represented by the real-time bandwidth between the entrance device 110 and the operator network is greater than or equal to the data transmission rate represented by the system threshold, perform packet loss processing on the TCP packets from the operator network.

Among them, the system threshold is obtained based on the contracted bandwidth between the campus network and the operator network. For example, if the contracted bandwidth is 100Mb/s, the system threshold may be 90Mb/s.

It can be understood that the relationship between the real-time bandwidth and the contracted bandwidth is determined by limiting the relationship between the data transmission rate represented by the real-time bandwidth and the data transmission rate represented by the system threshold between the portal device 110 and the operator network. When the data transmission rate represented by the real-time bandwidth between the portal device 110 and the operator network is greater than or equal to the data transmission rate represented by the system threshold, it indicates that the real-time bandwidth between the portal device 110 and the operator network is close to that between the portal device 110 and the operator network. Contracted bandwidth between operator networks, campus networks have the risk of network congestion. When the data transmission rate represented by the real-time bandwidth between the portal device 110 and the operator's network is less than the data transmission rate represented by the system threshold, it indicates that the real-time bandwidth between the portal device 110 and the operator's network is different from the data transmission rate between the portal device 110 and the operator. There is a large unused bandwidth between the contracted bandwidths between the networks, and there is no need to process the data transmitted between the entrance device 110 and the operator network.

For example, if the contracted bandwidth is 100Mb/s, the system threshold can be 90Mb/s. When the real-time bandwidth between the portal device 110 and the operator's network is 92Mb/s, it is greater than the system threshold, indicating that the portal device 110 and the operator's network are The real-time bandwidth between them is close to the contracted bandwidth between the portal device 110 and the operator network. In order to prevent the real-time bandwidth between the campus network and the operator network from exceeding the contracted bandwidth, packet loss processing is performed on the TCP packets in the downlink data sent from the operator network on the entrance device 110 side to reduce the time between the entrance device 110 and the operator network. The real-time bandwidth between the entrance device 110 and the operator network is ensured to be within the scope of the contracted bandwidth between the entrance device 110 and the operator network, thereby avoiding network congestion in the campus network.

It can be understood that the data transmission rate represented by the system threshold may be 90% to 95% of the data transmission rate represented by the contracted bandwidth. This system threshold can not only ensure the bandwidth utilization of the campus network, but also prevent the real-time bandwidth of the campus network from exceeding the contracted bandwidth, causing network congestion on the campus network. Of course, in other embodiments, the system threshold may be determined based on the network status of the campus network and the operator network.

It can be understood that in some embodiments, if the system threshold is larger, the data transmission rate represented by the system threshold is larger. For example, the system threshold is a bandwidth threshold, which is used to indicate the bandwidth used by the campus network. The specific implementation of step S202 may be: when the actual bandwidth is greater than or equal to the system threshold, the TCP packets from the operator network are processed. Packet loss handling. For example, the contracted bandwidth is 100Mb/s and the system threshold is 90Mb/s. When the real-time bandwidth is greater than or equal to 90Mb/s, TCP packets from the operator network are discarded.

In some embodiments, if the system threshold is larger, the data transmission rate represented by the system threshold is smaller. For example, the system threshold is the unused bandwidth threshold, which indicates the unused bandwidth of the campus network. The specific implementation of step S202 may be: when the absolute value of the difference between the contracted bandwidth and the actual bandwidth is less than or equal to the system threshold, packet loss processing is performed on the TCP packets from the operator network. For example, the contracted bandwidth is 100Mb/s and the system threshold is 10Mb/s. When the real-time bandwidth is 91Mb/s and the absolute value of the difference between the contracted bandwidth and the actual bandwidth is 9Mb/s and less than the system threshold, then the operator will TCP packets on the commercial network are discarded.

To sum up, when the system threshold is larger and the data transmission rate represented by the system threshold is smaller, the data transmission rate represented by the real-time bandwidth between the entrance device 110 and the operator network is greater than or equal to the data represented by the preset system threshold. The transmission rate is: the actual bandwidth is greater than or equal to the system threshold. When the system threshold is larger and the data transmission rate represented by the system threshold is smaller, the data transmission rate represented by the real-time bandwidth between the entrance device and the operator network is greater than or equal to the data transmission rate represented by the preset system threshold. , means: the absolute value of the difference between the contracted bandwidth and the actual bandwidth is less than or equal to the system threshold.

It can be understood that in some embodiments, when the data transmission rate represented by the real-time bandwidth of the entrance device 110 of the campus network is greater than or equal to the data transmission rate represented by the system threshold, in order to ensure that services in the campus network can be performed normally, the entrance device The data transmission rate represented by the packet forwarding performance of 110 is greater than the data transmission rate represented by the contracted bandwidth. The packet forwarding performance is used to indicate the data forwarding capability of the ingress device 110 .

In one embodiment, please refer to Figure 3. Packet loss processing for TCP packets from the operator network specifically includes:

S301. Obtain priority information of TCP packets from the operator network.

Among them, the priority information is used to indicate the priority or importance of multiple different TCP messages.

It can be understood that the data transmitted between the campus network and the operator network includes TCP packets and non-TCP packets (such as UDP packets). It should be noted that the entrance device 110 in the embodiment of this application only processes the received TCP packets, and other non-TCP packets will be directly forwarded to the corresponding client without packet loss processing.

It can be understood that since the multiple TCP messages received by the portal device 110 from the operator network correspond to different clients, that is, the portal device 110 forwards the TCP messages to the corresponding clients. In this way, the priority information of the TCP packet corresponding to each client can be determined through the priority information of the client. Illustratively, the priority information includes low priority and high priority. Take Figure 1 as an example. Assume that Client 2, Client 3 and Client 4 in the campus network have low priority and Client 1 has high priority. Then the priority of the TCP packet corresponding to Client 1 is greater than that of the client. 2. The priorities of the TCP packets corresponding to client 3 and client 4.

In some embodiments, obtaining priority information of TCP packets from the operator network in S301 includes:

Obtain the message ID of TCP messages from the operator network;

Determine the priority information of the corresponding TCP message based on the message identifier.

Specifically, the portal device 110 receives multiple TCP messages sent by the operator network, then obtains the message identifier corresponding to each TCP message, and determines the priority corresponding to each TCP message based on the message identifier.

It can be understood that in the embodiment of the present application, the message identifier is used to identify the TCP message or the client used by the user, so that the association between the TCP message and the client can be established through the message identifier.

It can be understood that the message identifier can be an IP address or a MAC address. Wherein, if the packet identifier is an IP address, the IP address may be the IP address of the client. Of course, the message identifier can also be other parameters, such as user-defined parameters or the destination port number in the TCP message. There is no specific limit here. It only needs to ensure that the parameter can uniquely identify the TCP message and the client used by the user. That’s it.

In some embodiments, the first correspondence relationship is stored in the portal device 110 . The first correspondence relationship includes message identifiers and priority information corresponding to each message identifier. Correspondingly, the priority information of the corresponding TCP message determined based on the message identifier includes:

Determine the priority information of the TCP message based on the message identifier and the first corresponding relationship.

Specifically, the portal device 110 matches the message identifier with the message identifier in the first correspondence relationship to obtain the priority information corresponding to the TCP message.

In other embodiments, the portal device 110 can also obtain the priority information of the corresponding TCP packet through other methods, such as For example, a recognition model is trained through a machine learning algorithm, TCP packets are input to the recognition model, and the recognition model outputs the priority information of the TCP packets.

S302. Perform differentiated packet loss processing on TCP packets based on priority information.

It can be understood that when the campus network is used by multiple users, different users perform different types of services through the campus network, or different users have different user identities, resulting in different users having different priorities, thus making the customers used by multiple users terminals have different priorities. In this way, differential packet loss processing can be performed on multiple TCP packets based on priority information, and TCP packets of relevant users can be processed in different ways based on the user's priority or the importance of the user's service execution. For example, when bandwidth is limited, packets are dropped for TCP packets of non-important users, but TCP packets of important users are not dropped, so that the real-time bandwidth between the campus network and the operator network is limited by the ingress device. 110 and the operator's network to avoid network congestion in the campus network. In addition, by not performing packet loss on TCP packets corresponding to important users or performing packet loss on important users, the packet loss rate is lower than the packet loss rate on ordinary users, thereby ensuring that the campus network bandwidth is limited. The experience of important users (or users with higher priority).

For example, please refer to Figure 1. Assume that the priorities of Client 2, Client 3 and Client 4 are equal, and the priority of the TCP packet corresponding to Client 1 is greater than that of Client 2, Client 3 and Client 4. According to the priority of the TCP packet, the entrance device 110 only performs packet loss processing on the TCP packets corresponding to client 2, client 3 and client 4, but does not perform packet loss processing on the TCP packet of client 1.

S302 will be described in detail below with reference to Embodiment 1 and Embodiment 2.

Example 1:

In Embodiment 1, in S302, differentiated packet loss processing is performed on TCP packets based on priority information, including:

The packets to be processed in the TCP packet are determined based on the priority information, and the packets to be processed are discarded. Among them, the priority of the packet to be processed is lower than the priority of other TCP packets in the TCP packets from the operator network.

Specifically, the portal device 110 divides TCP packets from the operator network into pending packets and normal processing packets based on priority information. The priority of packets to be processed is lower than the priority of normally processed packets. The entrance device 110 performs packet loss processing on the packets to be processed, and forwards the normally processed packets normally. In this way, the entrance device 110 implements differentiated packet loss processing on TCP messages by performing different packet loss processing methods on TCP messages corresponding to different priority information.

Illustratively, the ingress device 110 receives 70 TCP packets: Message 1 to Message 70 from the operator network. Next, the portal device 110 obtains the priority information of 70 TCP packets and determines that the priorities of packets 1 to 10 are higher than the priorities of packets 11 to 70. Then the portal device 110 sends packets 1 to 70. Message 10 is forwarded to the corresponding client normally, and packets 11 to 70 are discarded according to the preset packet loss parameters (for example, the packet loss parameter is the packet loss rate).

It can be understood that in other embodiments, if the message to be processed includes multiple TCP messages, the multiple TCP messages have multiple priority information, and the priority information and loss information of the message to be processed are stored in the entrance device 110. Second correspondence between package parameters. Then, packet loss processing for the packets to be processed may include:

Determine the packet loss parameters of the to-be-processed message based on the priority information of the to-be-processed message and the second corresponding relationship;

Perform packet loss processing on multiple TCP packets to be processed according to the packet loss parameters.

Specifically, multiple TCP messages in the messages to be processed have different priorities. The ingress device 110 can set different packet loss rates according to the TCP messages of different priorities, and perform processing on the corresponding TCP messages according to the packet loss rates. Packet loss handling. For example, the messages to be processed include 60 TCP messages: TCP message 11 to TCP message 70, among which TCP message 11 to TCP message 30 have the highest priority, and TCP message 31 to TCP message 50 The priority of TCP packet 51 to TCP packet 70 is the lowest. The entrance device 110 determines based on the priority information and the second corresponding relationship that the packet loss rate of TCP message 11 to TCP message 30 is 10%, the packet loss rate of TCP message 31 to TCP message is 15%, and the packet loss rate of TCP message 51 Report to TCP The packet loss rate of message 70 is 20%, and then the corresponding TCP packet is discarded according to the packet loss rate.

Among them, the packet loss parameter may be the packet loss rate. Of course, in other embodiments, the packet loss parameter may also be the number of packets lost.

It can be understood that the packet loss parameters can be preset in the entrance device 110 for the user. Of course, in other embodiments, the packet loss parameter may be determined based on the network status of the campus network and the operator network.

Example 2:

In Embodiment 2, the TCP packets received by the portal device 110 from the operator network include multiple TCP packets, for example, each TCP packet corresponds to a client. Then, in S302, differentiated packet loss processing is performed on the TCP packets from the operator network based on the priority information, including:

Determine the packet loss parameters of TCP packets based on priority information;

Packet loss processing is performed on TCP packets from the operator network based on packet loss parameters;

Among them, TCP packets with different priority information have different packet loss parameters.

Specifically, the entrance device 110 performs packet loss processing on all TCP messages from the operator network, and performs different packet loss processing on TCP messages with different priority information. That is, the ingress device 110 obtains the packet loss parameters corresponding to each type of TCP message based on the priority information. Since TCP messages of different priorities have different packet loss parameters, the ingress device 110 determines the corresponding types of multiple TCP packets based on the packet loss parameters. The packets are discarded.

For example, the packet loss parameter is the packet loss rate. The ingress device receives 60 TCP messages sent from the operator network: TCP message 11 to TCP message 70, where the priority of TCP message 11 to TCP message 30 is The priority of TCP message 31 to TCP message 50 is the highest, and the priority of TCP message 51 to TCP message 70 is the lowest. The entrance device determines the loss of TCP message 11 to TCP message 30 based on the priority information. The packet rate is 10%, the packet loss rate from TCP packet 31 to TCP packet 70 is 15%, the packet loss rate from TCP packet 51 to TCP packet 70 is 20%, the entrance device 110 calculates the corresponding TCP packet loss rate based on the packet loss rate. The packets are discarded. Of course, in other embodiments, if the ingress device determines that the packet loss rate of TCP packets 11 to TCP packets 30 is 0% and the packet loss rate of TCP packets 31 to TCP packets is 15%, the TCP The packet loss rate of packet 51 to TCP packet 70 is 20%, so the ingress device does not perform packet loss processing on TCP packet 11 to TCP packet 30.

It can be understood that in some embodiments, the packet loss sliding window is preset in the ingress device 110 . The size of the packet loss sliding window can be preset by the user, or can be determined based on the network status of the campus network (such as real-time network speed). Correspondingly, the ingress device 110 can perform packet loss processing on TCP packets from the operator network according to the packet loss sliding window. Specifically, in Embodiment 1 and Embodiment 2, packet loss processing is performed on TCP messages from the operator network based on packet loss parameters, including:

Packet loss processing is performed on TCP packets from the operator network based on the packet loss parameters and packet loss sliding window.

For example, see Figure 4, assuming that the size of the packet loss sliding window is 10 TCP packets, the packet loss parameter is the number of packets lost, and the number of packets lost is 3. Therefore, when the ingress device 110 receives 10 TCP packets from the operator network, 3 of the TCP packets will be randomly discarded according to the packet loss sliding window and packet loss parameters (3) (as shown in the black box in Figure 4 shown).

Of course, in other embodiments, if the ingress device 110 only discards packets to be processed in TCP packets from the operator network, then the ingress device 110 receives the packets to be processed in the size of the packet loss sliding window each time. Packet loss processing is performed based on the size of the packet loss sliding window to be processed. For example, the size of the packet loss sliding window is 10 TCP packets, the packet loss parameter is the number of packets lost, and the number of packets lost is 3. The entrance device 110 receives 50 TCP packets from the operator network, It receives 10 packets to be processed each time and discards the packets to be processed, that is, randomly discards 3 TCP packets in the packets to be processed.

Understandably, please refer to Figure 2 again. The entrance device 110 presets the time for TCP messages from the operator network. After packet loss processing, or after packet loss processing of TCP packets with a preset size (for example, 10MB) from the operator network, the network congestion control method also includes:

Obtain the real-time bandwidth between the portal device 110 and the operator's network;

If the data transmission rate represented by the real-time bandwidth of the ingress device 110 is greater than or equal to the data transmission rate represented by the system threshold, packet loss processing is performed on the subsequently received TCP messages from the operator network, and the packet loss processing is detected. The real-time bandwidth between the portal device 110 and the operator network; if the data transmission rate represented by the real-time bandwidth of the portal device 110 is greater than or equal to the data transmission rate represented by the system threshold, the subsequently received TCP packets from the operator network Packet loss processing is performed, and the cycle is repeated until the data transmission rate represented by the real-time bandwidth between the ingress device 110 and the operator's network is less than the data transmission rate represented by the system threshold. That is, subsequent TCP packets received from the operator network will no longer be discarded.

It can be understood that in some embodiments, packet loss processing of TCP packets from the operator network according to packet loss parameters includes:

Receive TCP packets with the size of the packet loss sliding window from the operator network;

Packet loss processing is performed on TCP packets with the size of the packet loss sliding window according to the packet loss parameters;

If the data transmission rate represented by the real-time bandwidth between the entrance device 110 and the operator's network is greater than or equal to the data transmission rate represented by the system threshold, the packet loss parameter is gradually increased, and the received packet loss parameter is processed based on the increased packet loss parameter. TCP packets with the size of the packet loss sliding window are discarded until the data transmission rate represented by the real-time bandwidth between the ingress device 110 and the operator network is less than the data transmission rate represented by the system threshold.

It can be understood that if the packet loss parameters of TCP packets from the operator network are fixed, then the ingress device 110 performs packet loss processing on the TCP packets from the operator network based on the packet loss parameters, and then the ingress device 110 again Obtain the real-time bandwidth between the portal device 110 and the operator network. If the ingress device 110 determines that the data transmission rate represented by the real-time bandwidth is still greater than the data transmission rate represented by the system threshold, the ingress device 110 can gradually increase the packet loss parameter corresponding to the packet loss process, and based on the increased loss The packet parameter performs packet loss processing on newly received TCP packets to achieve rapid adjustment of the rapidly increasing real-time bandwidth between the ingress device 110 and the operator network.

In addition, by gradually increasing the packet loss parameters corresponding to the packet loss processing, the ingress device 110 can rapidly reduce the data transmission rate represented by the real-time bandwidth between the ingress device 110 and the operator's network to a data transmission rate represented by the system threshold. rate.

For example, assume that the packet loss sliding window is 20 TCP packets, the packet loss parameter is the number of packets lost, and the number of packets lost is 4. If the portal device 110 receives 20 TCP packets from the operator network, the portal device 110 discards 4 TCP packets among the 20 TCP packets, and then obtains the real-time bandwidth between the portal device 110 and the operator network. If the data transmission rate represented by the real-time bandwidth is still greater than or equal to the data transmission rate represented by the system threshold, the number of lost packets is increased to 8, and then 20 TCP messages from the operator network are received, and the ingress device 110 After discarding 8 TCP packets among the 20 TCP packets, the real-time bandwidth between the ingress device 110 and the external network is obtained again. If the data transmission rate represented by the real-time bandwidth is still greater than or equal to the data transmission rate represented by the system threshold, continue to increase the number of lost packets until the data transmission rate represented by the real-time bandwidth is less than the data transmission rate represented by the system threshold. In this way, when the received TCP packets are dropped to reduce the real-time bandwidth between the entrance device 110 and the operator network, the packet loss parameters are gradually increased to quickly reduce the real-time bandwidth and avoid the real-time bandwidth of the campus network. Exceeding the contracted bandwidth.

In this embodiment of the present application, the entrance device 110 can gradually increase the packet loss parameter through a doubling method. For example, assuming that the packet loss parameter is the number of lost packets drop, when the real-time bandwidth of the entrance device 110 and the operator network is obtained for the first time, drop=0 and step=1. Among them, step is the first time the portal device 110 detects the real-time bandwidth between the portal device 110 and the operator's network. When entering When the data transmission rate represented by the real-time bandwidth between the ingress device 110 and the operator's network is greater than the data transmission rate represented by the system threshold, the ingress device 110 can increase the number of lost packets sequentially according to the following multiplication code:
threshold=drop;
drop=drop+step;
step=step*2;

Among them, threshold can be the packet loss parameter preset by the entrance device 100, and the packet loss parameter can be the packet loss rate or the number of packets lost.

Then, the ingress device 110 performs packet loss processing on the TCP packets from the packet loss sliding window of the operator network according to the gradually increasing number of lost packets, until the ingress device 110 detects the real-time bandwidth between the ingress device and the operator network. The data transfer rate characterized is less than the data transfer rate characterized by the system threshold.

Of course, in other embodiments, the entrance device 110 may adopt other methods of increasing packet loss parameters, and this application does not limit this.

It can be understood that in other embodiments, the data transmission rate represented by the real-time bandwidth between the ingress device 110 and the operator network is changed from a data transmission rate represented by being greater than or equal to the system threshold to a data transmission rate represented by being less than the system threshold. When the rate is reduced, network congestion control methods also include:

Gradually reduce the packet loss parameter, and perform packet loss processing on TCP packets in the packet loss sliding window based on the reduced packet loss parameter until the packet loss parameter becomes 0.

For example, the ingress device 110 can gradually reduce the packet loss parameter through the dichotomy method, that is, divide the packet loss parameter by 2 to obtain half of the original packet loss parameter, and slide the packet loss based on the reduced packet loss parameter. TCP packets in the window are discarded until the packet loss parameter is reduced to 0. At this time, all TCP packets from the operator network are not discarded.

It can be understood that the entrance device 110 can adopt other methods of reducing packet loss parameters, and this application does not limit this.

It can be understood that in the above embodiment, by gradually increasing the packet loss parameter until the packet loss parameter increases to the target packet loss parameter, when the TCP packets from the operator network are packet dropped according to the target packet loss parameter and After the data transmission rate represented by the real-time bandwidth between the entrance device 110 and the operator's network is less than the data transmission rate represented by the system threshold, the packet loss parameters are gradually reduced, and packet loss processing is performed based on the reduced packet loss parameters. , and detect the relationship between the real-time bandwidth after packet loss processing and the data transmission rate represented by the system threshold. If the data transmission rate represented by the real-time bandwidth between the entrance device 110 and the operator network is still less than the data transmission rate represented by the system threshold rate, then reduce the packet loss parameter again, and perform packet loss processing based on the reduced packet loss parameter, and cycle in sequence until the packet loss parameter is 0, then stop packet loss processing or if there is a gap between the entrance device 110 and the operator network If the data transmission rate represented by the real-time bandwidth is greater than the data transmission rate represented by the system threshold, the packet loss parameter will be gradually increased, and packet loss processing will be performed based on the increased packet loss parameter.

In some embodiments, if the TCP packets from the packet loss sliding window of the operator network are discarded according to the reduced packet loss parameters, the real-time bandwidth between the ingress device 110 and the operator network is represented by If the data transmission rate changes again to a data transmission rate greater than the system threshold, then the ingress device 110 gradually increases the packet loss parameter, and then performs packet loss processing on the TCP packets from the operator network based on the increased packet loss parameter. Then the packet loss parameter is increased or decreased based on the relationship between the data transmission rate represented by the real-time bandwidth between the ingress device 110 and the operator network and the data transmission rate represented by the system threshold. For example, the packet loss parameter is increased or decreased. If the data transmission rate represented by the real-time bandwidth between the ingress device 110 and the operator network is greater than the data transmission rate represented by the system threshold, the packet loss parameter will be increased. If the data transmission rate represented by the real-time bandwidth between the entrance device 110 and the operator network is less than the system If the data transmission rate represented by the threshold is reached, the parameter is reduced until the packet loss parameter is 0, and the ingress device 110 no longer performs packet loss processing on TCP messages from the operator network.

For example, please refer to FIG. 5 , which is a schematic diagram of the packet loss parameters of the ingress device 110 changing over time. In Figure 6, the vertical axis is the number of lost packets, and the horizontal and vertical coordinates are time. Before time A, the real-time bandwidth between the entrance device 110 and the operator's network The data transmission rate represented by the width is less than the data transmission rate represented by the system threshold. The portal device 110 directly forwards the TCP packet from the operator network to the client 150. After time A, the real-time bandwidth of the portal device 110 and the operator network The data transmission rate represented is greater than the data transmission rate represented by the system threshold. The entrance device 100 performs packet loss processing on the TCP messages from the operator network, and gradually increases the packet loss parameters. When time B is reached, the entrance device 110 The data transmission rate represented by the real-time bandwidth of the operator network is less than the data transmission rate represented by the system threshold. The ingress device 110 gradually reduces the packet loss parameter, and responds to the TCP packets from the operator network based on the reduced packet loss parameter. When time C is reached, the data transmission rate represented by the real-time bandwidth of the entrance device 110 and the operator's network is less than the data transmission rate represented by the system threshold, the entrance device 110 gradually increases the packet loss parameter, and based on The increased packet loss is performed on the TCP packets from the operator network, and then adjusted according to the relationship between the data transmission rate represented by the real-time bandwidth of the ingress device 110 and the operator network and the data transmission rate represented by the system threshold. The packet loss parameter is adjusted, and the TCP packets from the operator network are discarded according to the adjusted packet loss parameter. Then the entrance device 110 repeats the above steps until the packet loss parameter is 0. The entrance device 110 processes the packet loss from the operator network. TCP packets are no longer discarded.

Please refer to 6 for the state machine of the entrance device 110 provided by the embodiment of the present application. As shown in Figure 6, the entrance device 110 includes four states: idle, congestion, recovery, and probe.

As mentioned above, when the data transmission rate represented by the real-time bandwidth between the portal device 110 and the operator's network is less than the data transmission rate represented by the system threshold, the portal device 110 is in an idle state. For example, the state of the entrance device 110 before time A in FIG. 6 is the idle state.

When the data transmission rate represented by the real-time bandwidth between the ingress device 110 and the operator's network is greater than or equal to the data transmission rate represented by the system threshold, the ingress device 110 is in a congestion state. For example, in Figure 6, the status of the ingress device 110 after time A and before time B, and the status of the ingress device after time C and before time D are all in the congestion state.

When the data transmission rate represented by the real-time bandwidth between the ingress device 110 and the operator's network changes from a data transmission rate greater than or equal to the system threshold to a data transmission rate less than the system threshold, the ingress device 110 is in a recovery state. For example, the state of the entrance device 110 after time B and before time C in FIG. 6 is the recovery state.

When the entrance device 110 and the operator network switch back and forth between the congestion state and the recovery state (for example, in the above embodiment, the data transmission rate represented by the real-time bandwidth of the entrance device 110 and the operator network and the system threshold The packet loss parameters are adjusted according to the data transmission rate, and the TCP packets from the operator network are discarded according to the packet loss parameters), until the state balance point is found (for example, the target packet loss parameter is found), the entrance device 110 is in the detection state. state. For example, in the state of the entrance device 110 after time D in Figure 6, the entrance device 110 continues to detect the real-time bandwidth between the entrance device 110 and the operator's network. When the real-time bandwidth becomes smaller, the packet loss parameter is reduced. When the packet loss parameter When it is 0, the entrance device 110 switches from the recovery state to the idle state.

Please refer to FIG. 7 , which is a schematic diagram of a module of the entrance device 110 . The ingress device 110 includes a receiving ingress module, a traffic detection module, a back pressure control module, a sliding window packet loss module, a forwarding module, and an egress forwarding module.

Among them, the receiving entry module is used to receive data sent from the operator's network.

The traffic detection module is used to detect the amount of data transmitted by the entrance device 110 within a preset time (including the downlink data received by the entrance device 110 from the operator network and the uplink data forwarded by the entrance device 110 to the operator network), to obtain the data volume of the entrance device 110 Real-time bandwidth to and from the operator's network.

The forwarding module is configured to forward the data sent from the operator network to the egress forwarding module when the data transmission rate represented by the real-time bandwidth of the ingress device 110 is less than the data transmission rate represented by the system threshold.

The egress forwarding module is used to forward the data sent by the forwarding module to the corresponding client 150 .

If the data transmission rate represented by the real-time bandwidth of the entrance device 110 is greater than or equal to the data transmission rate represented by the system threshold, The transmission rate, back pressure control module and sliding window packet loss module cooperate to process the packet loss of TCP packets from the operator network, and forward the packet loss processed TCP packets to the egress forwarding module for the egress forwarding module. Forward the received data to the corresponding client 150. Specifically, if the data transmission rate represented by the real-time bandwidth of the ingress device 110 is greater than or equal to the data transmission rate represented by the system threshold, the backpressure control module is used to determine the packet loss parameter (see the above embodiment for the method of determining the packet loss parameter). , The sliding window packet loss module is used to discard TCP messages from the operator network based on the packet loss parameters and the preset packet loss sliding window, and send the discarded TCP messages to the egress forwarding module.

The egress forwarding module is also used to forward the TCP packets processed by the sliding packet loss module to the corresponding client 150 .

It can be understood that FIG. 7 is only used as an example of the entrance device 110 and is not limiting. Portal device 110 may have more or fewer modules than shown in FIG. 7 .

Next, please refer to FIG. 8 , which is a structural diagram of the entrance device 110 provided by the embodiment of the present application. As shown in FIG. 8 , the entrance device 110 includes: a processor 10 , a transmitter 20 , a receiver 30 , a memory 40 and a port 50 . The memory 40, the transmitter 20 and the receiver 30 and the processor 10 may be connected via a bus. Of course, in actual application, the memory 40, the transmitter 20, the receiver 30 and the processor 10 may not have a bus structure, but may have other structures, such as a star structure, which is not specifically limited in this application.

Optionally, the processor 10 may be a central processing unit, an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), one or more integrated circuits for controlling program execution, and a field programmable gate array (Field Programmable Gate). Array, FPGA) developed hardware circuits, baseband processors, etc.

Optionally, the processor 10 may include at least one processing core.

Optionally, the memory 40 may include read-only memory (Read Only Memory, ROM), random access memory (Random Access Memory, RAM), and disk memory. The memory 40 is used to store data required when the processor 10 is running. The number of memories 40 is one or more.

Optionally, the number of ports 50 is one or more, used to connect to the entrance device 110 on the upper or lower layer. If the ingress device 110 is an ingress device 110 connected to a host or server, such as an access switch or an edge switch, port 50 is also used to connect to the host or server.

Optionally, the transmitter 20 and the receiver 30 may be physically independent of each other or integrated together. The transmitter 20 can send the data to the forwarding device of the campus network or the client of the campus network through port 50. Receiver 30 may receive data sent from the operator's network through port 50.

Those skilled in the art will understand that embodiments of the present application may be provided as methods, systems, or computer program products. Accordingly, the present application may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein. These computer program codes may be stored in computer-readable memory that directs a computer or other programmable data processing device to operate in a particular manner.

This embodiment also provides a computer storage medium. Computer instructions are stored in the computer storage medium. When the computer instructions are run on the entrance device, the entrance device performs the above related method steps to implement the network congestion control method in the above embodiment. .

This embodiment also provides a computer program product. When the computer program product is run on the ingress device, the ingress device performs the above related steps to implement the network congestion control method in the above embodiment.

In addition, embodiments of the present application also provide a device. This device may be a chip, a component or a module. The device may include a connected processor and a memory. The memory is used to store computer execution instructions. When the device is running, The processor can execute computer execution instructions stored in the memory, so that the chip executes the network congestion control method in each of the above method embodiments.

Through the above description of the embodiments, those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above functional modules is used as an example. In actual applications, the above functions can be allocated as needed. It is completed by different functional modules, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above.

In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the modules or the division of modules are only a logical function division. In actual implementation, there may be other division methods, for example, multiple modules or components may be combined. Either it can be integrated into another device, or some features can be ignored, or not implemented. On the other hand, the coupling or direct coupling or communication connection between each other shown or discussed may be through some interfaces, indirect coupling or communication connection of devices or modules, which may be in electrical, mechanical or other forms.

A module described as a separate component may or may not be physically separate. A component shown as a module may be one physical module or multiple physical modules, that is, it may be located in one place, or it may be distributed to multiple different places. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional module in each embodiment of the present application can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module. The above integrated modules can be implemented in the form of hardware or software function modules.

If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present application are essentially or contribute to the existing technology, or all or part of the technical solution can be embodied in the form of a software product, and the software product is stored in a storage medium , including several instructions to cause a device (which can be a microcontroller, a chip, etc.) or a processor to execute all or part of the steps of the methods described in various embodiments of this application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk and other media that can store program code. .

The above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present application shall be covered by the protection scope of the present application. . Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims (16)

1. A network congestion control method, characterized in that it is applied to an entrance device of a campus network, and the entrance device is used to connect to an operator's network. The network congestion control method includes:

   Obtain the real-time bandwidth between the entrance device and the operator network;

   If the data transmission rate represented by the real-time bandwidth between the entrance device and the operator network is greater than or equal to the data transmission rate represented by the preset system threshold, then the transmission control protocol TCP from the operator network Packet loss processing is performed on the packets, where the system threshold is obtained based on the contracted bandwidth between the campus network and the operator network.
2. The network congestion control method according to claim 1, wherein the packet loss processing of TCP messages from the operator network includes:

   Obtain the priority information of the TCP message;

   Perform differentiated packet loss processing on TCP messages from the operator network based on the priority information.
3. The network congestion control method according to claim 2, wherein said obtaining the priority information of the TCP message includes:

   Obtain the message identifier of the TCP message, where the message identifier is used to identify the TCP message;

   Determine the priority information of the TCP message according to the message identifier.
4. The network congestion control method according to claim 2 or 3, wherein the differential packet loss processing of TCP messages from the operator network based on the priority information includes:

   Determine the pending packets in the TCP packets from the operator network according to the priority information;

   Perform packet loss processing on the packet to be processed, wherein the priority of the packet to be processed is lower than the priority of other TCP packets among all TCP packets from the operator network.
5. The network congestion control method according to claim 4, wherein the packets to be processed include a variety of TCP packets, and the packet loss processing for the packets to be processed includes:

   Determine the packet loss parameters of each TCP message in the to-be-processed message based on the priority information of the to-be-processed message;

   Perform packet loss processing on the corresponding TCP message in the message to be processed according to the packet loss parameter.
6. The network congestion control method according to claim 2, wherein performing differentiated packet loss processing on TCP messages from the operator network based on the priority information includes:

   Determine packet loss parameters of TCP messages from the operator network based on the priority information;

   Packet loss processing is performed on TCP messages from the operator network according to the packet loss parameters, where different priority information corresponds to different packet loss parameters.
7. The network congestion control method according to any one of claims 2 to 6, characterized in that a packet loss sliding window is stored in the entrance device, and the packet loss sliding window from the operator network is stored based on the priority information. Differential packet loss processing is performed on TCP packets, including:

   Perform differentiated packet loss processing on TCP messages from the operator network according to the packet loss sliding window.
8. The network congestion control method according to claim 7, wherein the packet loss processing of the TCP message according to the packet loss sliding window includes:

   Receive TCP messages of the packet loss sliding window size from the operator network;

   Perform packet loss processing on TCP packets with the size of the packet loss sliding window according to the preset packet loss parameters;

   If after the packet loss processing, the data transmission rate represented by the real-time bandwidth of the entrance device and the operator network is greater than or equal to the data transmission rate represented by the system threshold, gradually increase the packet loss parameter , and increase according to The subsequent packet loss parameters are used to perform packet loss processing on TCP packets with a packet loss sliding window size received from the operator network until the data transmission rate represented by the real-time bandwidth of the entrance device and the operator network is less than the required The data transfer rate represented by the system threshold.
9. The network congestion control method according to claim 8, characterized in that if the data transmission rate represented by the real-time bandwidth between the entrance device and the operator network is greater than or equal to the data represented by the system threshold The transmission rate is changed to be less than the data transmission rate represented by the system threshold, and the method further includes:

   Gradually reduce the packet loss parameter, and drop the received TCP packets with the size of the packet loss sliding window according to the reduced packet loss parameter until the packet loss parameter reaches 0.
10. The network congestion control method according to claim 9, characterized in that the method further includes:

    Adjust the packet loss parameter according to the relationship between the data transmission rate represented by the real-time bandwidth between the entrance device and the operator network and the data transmission rate represented by the system threshold;

    Packet loss processing is performed on TCP messages from the operator network according to the adjusted packet loss parameters.
11. The network congestion control method according to any one of claims 1 to 10, characterized in that when the system threshold is larger and the data transmission rate represented by the system threshold is larger, the inlet device and the The data transmission rate represented by the real-time bandwidth between the above operator networks is greater than or equal to the data transmission rate represented by the preset system threshold, which is:

    The actual bandwidth is greater than or equal to the system threshold.
12. The network congestion control method according to any one of claims 1 to 10, characterized in that when the system threshold is larger and the data transmission rate represented by the system threshold is smaller, the entrance device and the The data transmission rate represented by the real-time bandwidth between operator networks is greater than or equal to the data transmission rate represented by the preset system threshold, which is:

    The absolute value of the difference between the contracted bandwidth and the actual bandwidth is less than or equal to the system threshold.
13. The network congestion control method according to any one of claims 1 to 12, characterized in that the data transmission rate represented by the packet forwarding performance of the ingress device is greater than the data transmission rate represented by the contracted bandwidth.
14. An entrance device, characterized in that the entrance device includes:

    one or more processors;

    A storage device for storing one or more programs;

    When the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the network congestion control method according to any one of claims 1 to 13.
15. A computer-readable storage medium on which a computer program is stored, characterized in that, when the program is executed by a processor, the network congestion control method according to any one of claims 1 to 13 is implemented.
16. A campus network, characterized in that the campus network includes an ingress device, and the ingress device is used to perform the network congestion control method according to any one of claims 1 to 13.