WO2012167571A1

WO2012167571A1 - Network congestion control method and device

Info

Publication number: WO2012167571A1
Application number: PCT/CN2011/082933
Authority: WO
Inventors: 叶思海; 熊华梁; 徐日东
Original assignee: 华为技术有限公司
Priority date: 2011-11-25
Filing date: 2011-11-25
Publication date: 2012-12-13
Also published as: CN102726092B; CN102726092A

Abstract

Disclosed are a network congestion control method and device. The method includes: if the service success rates in N successive control periods are all less than a pre-set service success proportion threshold value, and a network congestion announcement is received, then initiating monitoring of the network status; counting the network time delay of message roundtripping within each control period after initiating monitoring to obtain the average time delay of each control period after initiating monitoring; and, when the average time delay of any control period after initiating monitoring is greater than the upper limit value of a pre-set target time delay interval, reducing the amount of traffic that is allowed to send a service first message in the next control period after any control period. The device includes a starting control module, a statistics module and a first control module. The present embodiments avoid the problem that it is easy to cause errors in control because the current HTR flow control only considers the service success rate, realizing a more effective control of network congestion.

Description

Network congestion control method and device

TECHNICAL FIELD Embodiments of the present invention relate to communication technologies, and in particular, to a network congestion control method and apparatus. Background technique

The network congestion phenomenon can be controlled at the link layer. However, the congestion control of the link layer itself can only ensure the reliable transmission of the message. If the traffic exceeds the transmission capacity of the link, the message will be randomly discarded. The message random discarding damage to the telecommunication signaling service is greater than the damage caused by the stream-based datagram transmission. The telecommunication signaling interaction is required to have a delay. The protocol usually defines an interaction timeout requirement, and the service layer is not allowed to repeatedly retry. A session in the telecommunication signaling interaction process usually includes multiple interaction processes, and any failure of the interaction may cause the service to be damaged. Therefore, the random message discarding damage to the telecommunication signaling service is large. Therefore, for network congestion, only the service volume of the service layer can be reduced to fully utilize the link capability, otherwise it may cause more serious call loss.

In the prior art, the network congestion control of the service layer is a "hard to reach" (HTR) flow control method based on success rate, and the service layer decides to send to the home location register according to the service success rate ( Home Locat ion Reg is ter; hereinafter referred to as: HLR). When the service success rate is greater than a preset threshold, the amount of traffic sent is increased; when the success rate is less than a preset threshold, the amount of traffic sent is reduced.

However, there are many factors affecting the success rate of the service. When the service success rate is used as the basis for congestion judgment, it is usually prone to mis-control, which makes it impossible to effectively control network congestion. For example, when the overload or congestion is not caused by the device itself, the success rate is low at this time, and the flow control is started on the network at this time, that is, the traffic volume reduced at this time cannot solve the congestion phenomenon. Summary of the invention

The embodiment of the invention provides a network congestion control method and device, which avoids the false control phenomenon of the HTR flow control and realizes the control of network congestion.

In order to achieve the above object, an embodiment of the present invention provides a network congestion control method, including: If the service success rate in the consecutive N control periods is less than the preset service success ratio threshold, and the network congestion notification is received, the monitoring of the network status is started, where N is a pre-set positive integer greater than 1.

The network delay of the round-trip of the message in each control period after the monitoring is started is counted, and the average delay of each control period after the startup monitoring is obtained;

When the average delay of any one of the control periods after the startup monitoring is greater than the upper limit of the preset target delay interval, the allowed transmission service of the service first message in the next control period of the any one of the control periods is reduced. the amount;

The average delay of any one of the control periods after the startup monitoring is an average of the network delays of the service success responses received within any one of the control periods after the startup monitoring.

A network congestion control apparatus provided by an embodiment of the present invention includes:

The control module is configured to start monitoring the network status if the service success rate in the consecutive N control periods is less than a preset service success ratio threshold, and the network congestion notification is received, where N is preset a positive integer greater than one;

a statistics module, configured to perform statistics on network delays of messages in and out of each control period after the monitoring is started, and obtain an average delay of each control period after the startup monitoring;

a first control module, configured to: when the average delay of any one of the control periods after the startup monitoring is greater than an upper limit of the preset target delay interval, reduce the next control period of the any one of the control periods The traffic of the first message of the service is allowed to be sent;

The network congestion control method and apparatus provided by the embodiment of the present invention initiates monitoring of the network status if the service success rate in the continuous N control periods is less than the preset service success ratio threshold and the network congestion notification is received. Start to count the network delay of the message round-trip in each control period after starting the monitoring, and reduce the permission of the service first message of the next control period of any one of the later control periods according to the average delay of any one of the control periods obtained. The traffic is sent. The average delay of the control period can reflect the current network congestion. This embodiment avoids the problem that the existing HTR flow control is easy to cause error control because only the service success rate is considered. More effective control, improving the business success rate after network congestion. BRIEF DESCRIPTION OF THE DRAWINGS In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, a brief description of the drawings used in the embodiments or the prior art description will be briefly described below. The drawings are some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without any inventive labor. 1 is a flowchart of Embodiment 1 of a network congestion control method according to the present invention;

2 is a flowchart of Embodiment 2 of a network congestion control method according to the present invention;

3 is a structural diagram of Embodiment 1 of a network congestion control apparatus according to the present invention;

4 is a structural diagram of Embodiment 2 of a network congestion control apparatus according to the present invention. The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. The embodiments are a part of the embodiments of the invention, and not all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.

FIG. 1 is a flowchart of Embodiment 1 of a network congestion control method according to the present invention. As shown in FIG. 1 , this embodiment provides a network congestion control method, which may specifically include the following steps:

Step 101: If the service success rate in the continuous N control periods is less than the preset service success ratio threshold, and the network congestion notification is received, the monitoring of the network status is started.

This embodiment can be applied to a network congestion scenario in which various signaling interactions in a communication system, where the network refers to a network corresponding to a communication interface between network elements in a telecommunication system, and the applicable communication interfaces may include but are not limited to the following Interface: General Packet Radio Service (hereinafter referred to as GPRS) service support node (Serving GPRS Support Node; hereinafter referred to as: SGSN) and HLR interface (ie Gr interface), Mobile Switching Center (Mobile Switching Center) (hereinafter referred to as: MSC) / Visit location register (Visitor Location Register; hereinafter referred to as: VLR) and HLR interface (ie C / D interface), SGSN / MSC and Radio Network Controller (Radio Network Controller; RNC) interface (ie Iu interface), MSC and base station controller (Base Station Controller; Said: BSC) interface (ie A interface).

In this embodiment, the HTR automatic flow control is triggered by a preset control condition. If the service success rate of a plurality of consecutive periods is lower than a preset service success ratio threshold, and the network congestion notification is received, the network is congested. The advertisement can be reported to the service layer when the link layer detects the network congestion, and then the network condition is determined to reach a preset control condition, and the monitoring of the network status is started, and the service is sent in the subsequent step 1 02-1 03. The amount is controlled. If the above control conditions are not met, no further steps are required to continue monitoring the network status. By setting the starting condition, the problem that the existing HTR flow control is easy to cause error control due to the flow control only considering the service success rate is avoided, and the control of the network congestion is more accurate and effective.

The N is a pre-set positive integer greater than 1, and the N may be pre-configured according to a network statistical experience value.

Step 1 02: Perform statistics on the network delay of the round-trip messages in each control period after the monitoring is started, and obtain an average delay of each control period after the startup monitoring.

When it is determined that the current network condition reaches the preset control condition, and the monitoring of the network status is started, the flow control operation of the embodiment is started. This step is specifically for each control cycle after the monitoring is started. The network delay within the message round-trip is counted. The network delay of the round-trip of the message in the control period is specifically the time elapsed after the device sends a message to the peer network element in the control period until the response of the message returned by the peer network element is received. . When the network condition meets the preset control condition, the network delay of the round-trip message can be used to determine the current network congestion degree. The network congestion degree here may include the congestion degree of the link and the network element, and the network delay. The larger the congestion, the greater the congestion, and the smaller the network delay, the smaller the congestion. In this step, the average delay of each control period after the monitoring is started is obtained, that is, after counting the network delays in one control period, the average delay of the control period is determined according to the network delay of the statistics. The average delay is the average of the network delays of the round trips of multiple messages. Assume that the value of N in the above step is 4, the statistics of the network delay of the round-trip of each control period after the start of monitoring, that is, the round-trip of the messages of each control period after the fourth control period is counted. Network delay. For example, the network delay of the message round-trip of the 5th control cycle, the 6th control cycle... is obtained, and then the average delay of the 5th, 6th, ... control cycle is obtained, wherein, because the control cycle is Flowing one by one. Therefore, the network delay of the fifth control period, the sixth control period, and the like are counted, and the average delay of obtaining each control period is not Parallel, but the network delay of the 5th control period is obtained, and the average delay of the 5th control period is obtained. When the 6th control period is started, the statistics are executed serially to obtain the 6th control period. Network delay, get the average delay of the sixth control period, and so on.

Step 1 03: When the average delay of any one of the control periods after the start of monitoring is greater than the upper limit of the preset target delay interval, reduce the permission of the service first message in the next control period of the any one of the control periods Send traffic.

After the average delay of any one of the control periods after the monitoring is started by statistics, the allowed traffic of the service first message is controlled according to the average delay, that is, the size of the transmitted traffic is determined. Specifically, when the average delay of any one of the control periods after the start of monitoring is greater than the upper limit of the preset target delay interval, the allowable transmission traffic of the service first message in the next control period of the control period is reduced. The target delay interval may be a preset interval configured according to an average delay of a previous control period of any one of the control periods after the start of monitoring. The allowed transmission traffic of the service first message in this embodiment can be used to characterize the capability of the device, indicating the traffic volume of the service first message that the device can transmit in one control period. In a telecommunication service, usually one session includes the interaction of multiple request messages and response messages, wherein the first request message is the service first message. When the degree of congestion of the network is large, this embodiment matches the current network condition by reducing the allowed traffic of the service first message, so as to avoid the phenomenon of randomly dropping the message and realizing the control of the network congestion.

In this embodiment, by controlling the size of the allowed traffic of the service first message, the network congestion degree is finally alleviated, and since the average delay of one control period can directly reflect the network congestion degree, the target controlled by this embodiment is also It can be equivalent to controlling the average delay within the target delay interval. The length of the message buffer queue can be determined by the average delay of the control period. The longer the average delay is, the longer the length of the message buffer queue is. The smaller the average delay is, the shorter the length of the message buffer queue is. If the queue is too long, the overflow may occur, causing the message to be randomly discarded. The protocol timer expires and the service is damaged. If the message buffer queue is too short, the buffer function is not available, and the link capability may not be fully utilized. Therefore, in this embodiment, by controlling the allowed traffic of the service first message, the length of the message buffer queue is indirectly controlled by controlling the average delay, so that the message buffer queue maintains an appropriate value, thereby making the length of the message buffer queue. Within the expected range, the service sent to the network is guaranteed to have a higher success rate. The embodiment provides a network congestion control method, if the service success rate in the continuous N control periods is less than the preset service success ratio threshold, and the network congestion notification is received, the monitoring of the network status is started, and the process starts. The network delay of the round-trip of the message in each control period after the monitoring is started is counted, and the allowed transmission of the service first message of the next control period of any one of the later control periods is reduced according to the average delay of any one of the obtained control periods. Traffic volume; Since the average delay of the control period can reflect the current network congestion degree, this embodiment avoids the problem that the existing HTR flow control is easy to cause error control because the flow control is started only considering the service success rate. More precise and effective control of network congestion improves the service success rate after network congestion.

2 is a flowchart of Embodiment 2 of a network congestion control method according to the present invention. As shown in FIG. 2, the embodiment provides a network congestion control method, which may specifically include the following steps:

Step 201: When the service success rate in the consecutive N control periods is less than the preset service success ratio threshold, and the network congestion notification is received, the monitoring of the network status is started.

In this embodiment, the specific flow control is performed by using a preset control period, and the control period may be specifically configured according to actual conditions, for example, the specific configuration is 10 seconds. The service layer performs the HTR flow control according to the preset control condition. If the service success rate in the continuous N control periods is less than the preset service success ratio threshold, and the link layer is reported, When the network congestion notification is performed, it is determined that the network condition reaches a preset control condition, and the monitoring of the network status is started, that is, the service layer starts to execute the flow control solution in this embodiment. N is a positive integer greater than 1 in advance, and N=4 can be specifically taken here. The service success ratio threshold is preset according to the actual situation, such as the service success measured according to the previous network congestion. The rate is set, which can be 70%. In this embodiment, when the link layer detects that the current network is congested, for example, when the message sending buffer overflows, the network layer is congested in the form of a network congestion notification.

Step 202: Perform statistics on the network delay of the round-trip of the message in each control period after the monitoring is started, and determine the average delay of the each control period according to the network delay of the round-trip of the message in the respective control periods.

After the traffic layer starts to be flow-controlled, the embodiment performs statistics on the round-trip network delay of each message in each control period after the monitoring is started in units of a preset control period. When the network delay of a message is counted, the device may send the message to the peer network element to start timing, until the device receives the response of the message returned by the peer network element, and stops counting. The time is the network delay for the round trip of the message. 4叚 The device is sent to the peer network element within one control period. With 100 messages, but the device only receives the response of 50 messages, this step only counts the network delay of the message round-trip of the 50 messages that received the response. This step also determines the average delay of each control cycle based on the statistical network delay of the message round-trip in each control cycle. The average delay is the average of the network delays of the round trips of multiple messages. The size of the delay can reflect the degree of network congestion. The greater the delay, the more serious the network congestion. The smaller the delay, the lighter the network congestion.

Step 203: Determine whether the average delay of any one of the control periods after the monitoring is started is within a preset target delay interval. If yes, execute step 204; otherwise, perform step 207.

In this embodiment, the network congestion condition is determined by the average delay of any one of the control periods after the monitoring is started, and the amount of transmission traffic is controlled according to the network congestion condition. In this embodiment, the window access restriction (Window Acces s L imi t; hereinafter referred to as WAL) control method can be specifically used to control the transmission traffic. In this embodiment, a target delay interval may be set according to actual conditions, and the target delay interval may be an interval range centered on the target delay, and the upper limit of the target delay interval (target delay + delay adjustment) The value Δ ) , the lower limit of the target delay interval is (target delay - Δ ), that is, the target delay interval is [target delay - Δ, target delay + △]. In this step, it is determined whether the average delay of any one of the control periods after the start of monitoring falls within the target delay interval to determine the degree of network congestion of the control period. When the average delay of any one of the control periods after the start of the monitoring falls within the target delay interval, indicating that the current network congestion is low, or there is no network congestion, step 204 is performed to continue with further specific judgment; When the average delay of any one of the control periods after the monitoring is started does not fall within the target delay interval, it indicates that the current network congestion is high, or the network congestion is serious. Then, step 207 is performed.

Step 204: Determine whether the instantaneous delay of any one of the control periods after the start of monitoring is greater than the immediate delay of the previous control period of any one of the control periods after the start of monitoring, and if yes, execute step 205, otherwise step 206 is performed. .

When the average delay of any one of the control periods after the start monitoring is within a preset target delay interval, continue to determine whether the instantaneous delay of any one of the control periods is greater than any one of the control periods after the start of monitoring The instantaneous delay of a control period, where the instantaneous delay of any one of the control periods after the start of the monitoring is the network delay of the last round of the message in any one of the control periods after the start of the monitoring, in this embodiment It can be specifically the immediate delay of the last 1 second of the control cycle. When the instantaneous delay of any one of the control periods after the startup monitoring is greater than the instantaneous delay of the previous control period of the any one of the control periods, indicating the current network time If the delay is in an upward trend, step 205 is performed to reduce the allowed transmission traffic of the service first message in the next control period of the control period. When the instantaneous delay of any one of the control periods after the startup monitoring is less than the instantaneous delay of the previous control period of the any one of the control periods, indicating that the current network delay is in a downward trend, step 206 is performed to increase the control period. The amount of traffic allowed for the first message of the service in the next control cycle.

Further, the optional solution of the embodiment of the present invention may further determine, in the foregoing step, when the instantaneous delay of any one of the control periods is greater than the immediate delay of the previous control period of any one of the control periods, Whether the difference between the instantaneous delay of any one control period and the instantaneous delay of the previous control period of any one of the control periods is greater than or equal to a preset delay interval, and if so, that is, the instantaneous delay of any one of the control periods When the preset delay adjustment value Δ is increased compared with the preset delay adjustment value Δ of the previous control period, the current network delay is determined to be in an upward trend. In this embodiment, the control period is reduced. In the next control cycle, the first message of the service allows the transmission of traffic to implement network congestion control. When the instantaneous delay of any one of the control periods is less than the instantaneous delay of the previous control period, the instantaneous delay of the last control period of any one of the control periods and the instantaneous delay of the any one of the control periods may be further determined. Whether the difference is greater than or equal to a preset delay interval, and if so, that is, when the instantaneous delay of any one of the control periods is reduced compared to the instantaneous delay of the previous control period of the arbitrary one of the control periods, the preset time is reduced When the adjustment value Δ is extended, it is further determined that the current network delay is in a downward trend, and the allowed transmission traffic of the service first message in the next control period of the control period may be increased.

Step 205: Control the allowed transmission traffic of the service first message in the next control period of any one of the control periods after the startup monitoring to be the current allowed transmission traffic of the service first message in the next control period (lm %) times.

The average delay of any one of the control periods after the start monitoring is greater than or equal to a lower limit value of the preset target delay interval (target delay - Δ), and less than or equal to the preset target delay interval When the limit value (target delay + Δ), if the instantaneous delay of any one of the control periods after the start monitoring is greater than the instantaneous time delay of the previous control period, indicating that the current network delay is in an upward trend, the control is reduced. In the next control period of the period, the traffic of the service first message is allowed to be sent to control the network congestion, and the traffic of the service first message of the next control period may be controlled to be the service head of the next control period. The current allowed number of traffic of the message is ( l_m% ) times, where m is a positive real number less than 100, and m can be set to 5. In fact In applications, m can be obtained through a large number of experiments, usually a pre-configured fixed value.

Further, in this embodiment, in the initial stage of flow control, after the monitoring of the network status is started, the initial value of the allowable transmission traffic of the service first message may be set. That is, after the foregoing step 201, the method may specifically include the following steps: setting an initial value of the allowed transmission traffic of the service first message to a maximum value of the following two values: the actual successful traffic volume and the service first message in the previous control cycle. The minimum amount of traffic allowed to be sent. The minimum value of the allowed traffic to be sent is the first message of the service, and the minimum value of the allowed traffic is used to prevent the flow control parameters such as the network delay and the service success rate from being lost when there is no traffic at all. Happening.

Step 206: Control the allowed transmission traffic of the service first message in the next control period of any one of the control periods after the start of the monitoring to the current allowed transmission service volume of the service first message in the next control period ( l+n%) times.

The average delay of any one of the control periods after the start monitoring is greater than or equal to a lower limit value of the preset target delay interval (target delay - Δ), and less than or equal to the preset target delay interval When the limit value (target delay + Δ) is used, if the instantaneous delay of any one of the control periods after the start monitoring is smaller than the immediate delay of the previous control period, it indicates that the current network delay is in a downward trend, and the control period is increased. The traffic of the service first message in the next control period is allowed to be sent, so as to fully utilize the network resources such as the link, specifically, the transmission traffic of the next control cycle of the control cycle is controlled to be the next control cycle. The current first message of the service is allowed to transmit (l+n%) times of the traffic, where n is a positive real number less than 100, and specifically, n can be set to 2. In practical applications, n can be obtained by a large number of experiments, usually a fixed value, but the value of m is usually greater than the value of n.

Step 207: Determine whether an average delay of any one of the control periods after the start monitoring is greater than an upper limit of the preset target delay interval. If yes, execute step 208; otherwise, execute step 209.

When the average delay of any one of the control periods after the monitoring is started is not within the preset target delay interval, it is specifically determined whether the average delay of the control period is greater than the upper limit of the preset target delay interval, that is, the judgment Whether the average delay is greater than (target delay + Δ), if yes, indicating that the current network congestion is serious, step 208 is performed; otherwise, step 209 is performed.

Step 208: Control the allowed transmission traffic of the service first message in the next control period of any one of the control periods after the startup monitoring to be the service first cancellation in the next control period. The current amount of interest is allowed to be transmitted (lp%) times.

When the average delay of any one of the control periods after the start of monitoring is greater than the upper limit of the preset target delay interval (target delay + Δ), the transmission traffic is reduced, which may specifically be the next control of the control period. The allowable transmission traffic of the periodic service first message is controlled to be (lp%) times the current allowed traffic volume of the service first message of the next control period, where p is a positive real number less than 100, and specifically, p may be taken The value is 10. In practice, p can be obtained through a large number of experiments, usually a pre-configured fixed value.

Step 209: Control the allowed transmission traffic of the service first message in the next control period of any one of the control periods after the start of the monitoring to the current allowed service transmission amount of the service first message in the next control period ( l+q%) times.

When the average delay of any one of the control periods after the start of monitoring is greater than the lower limit of the preset target delay interval (target delay - Δ), the transmission traffic is increased, which may specifically be the next control of the control cycle. The allowable transmission traffic of the service first message in the cycle is controlled as (1+q%) times of the currently allowed traffic volume of the service first message of the next control cycle, where q is a positive real number less than 100, specifically Let q be 5. In practical applications, q can be obtained through a large number of experiments, usually a pre-configured fixed value, but the value of p is usually greater than the value of q, because the increase in traffic is to ensure the maximum use of bandwidth, so that Telecommunication equipment is effectively utilized. However, in order to achieve smooth service, it is necessary to ensure that the network is not congested. Therefore, if the increased traffic is always less than the reduced traffic, the bandwidth can be effectively utilized to improve efficiency. Success rate.

Specifically, when allowed to control the transmission of traffic, to determine a service layer after each control period can be transmitted traffic ₇ supported by the service allows the first message transmitted traffic on the business first complete each control cycle service start message The amount of the message is specifically controlled, that is, the first message of the service is rejected, instead of the intermediate message of the service, so that the amount of traffic actually sent by the service layer is controlled in the amount of the transmitted traffic.

Further, the method provided in this embodiment may further include the following steps: stopping the service when the amount of traffic received in the consecutive M control periods after the startup monitoring is less than or equal to the allowed transmission traffic of the service first message. The control of the first message allows the transmission of traffic, where M is a pre-set positive integer greater than 1, where M can be specifically set to 4. That is, when the network condition satisfies the preset de-control condition, the de-control condition here is received by the service layer in four consecutive control periods. If the service volume is less than or equal to the allowed traffic volume of the service first message of each determined control period, it indicates that the amount of traffic that needs to be sent is small, and the network congestion control is not required, and the monitoring operation of this embodiment is stopped.

The embodiment provides a network congestion control method, if the service success rate in the continuous N control periods is less than the preset service success ratio threshold, and the network congestion notification is received, the monitoring of the network status is started, and the process starts. The network delay of the round-trip of the message in each control period after the monitoring is started is counted, and the allowed transmission of the service first message of the next control period of any one of the later control periods is reduced according to the average delay of any one of the obtained control periods. Since the average delay of the control period can reflect the current network condition, this embodiment avoids the problem that the existing HTR flow control is easy to cause error control due to only considering the service success rate, and the network congestion is more effective. The control improves the business success rate after network congestion.

A person skilled in the art can understand that all or part of the steps of implementing the above method embodiments may be completed by using hardware related to program instructions, and the foregoing program may be stored in a computer readable storage medium, and the program is executed when executed. The steps of the foregoing method embodiments are included; and the foregoing storage medium includes: a medium that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk.

3 is a structural diagram of Embodiment 1 of a network congestion control apparatus according to the present invention. As shown in FIG. 3, this embodiment provides a network congestion control apparatus, which may specifically perform the steps in Embodiment 1 of the foregoing method. Let me repeat. The network congestion control apparatus provided in this embodiment may specifically include a start control module 301, a statistics module 302, and a first control module 303. The start control module 301 is configured to start monitoring the network status if the service success rate in the consecutive N control periods is less than a preset service success ratio threshold, and the network congestion notification is received, where N is a pre- Set a positive integer greater than 1. The statistics module 302 is configured to perform statistics on the network delay of the round-trip of the messages in each control period after the monitoring is started, and obtain the average delay of each control period after the startup monitoring. The first control module 303 is configured to reduce the service head in the next control period of any one of the control periods when the average delay of any one of the control periods after the start of the monitoring is greater than the upper limit of the preset target delay interval The message is allowed to send traffic. The average delay of any one of the control periods after the start monitoring is an average of the network delays of the service success responses received within any one of the control periods after the start monitoring.

4 is a structural diagram of Embodiment 2 of a network congestion control apparatus according to the present invention. As shown in FIG. 4, this embodiment provides a network congestion control apparatus, which can specifically execute each of the foregoing Embodiment 2 of the foregoing method. Steps, no more details here. The network congestion control apparatus provided in this embodiment may further include a second control module 401 on the basis of the foregoing FIG. The second control module 401 is configured to increase the service head in the next control period of any one of the control periods when the average delay of any one of the control periods after the start of the monitoring is less than the lower limit of the preset target delay interval The message is allowed to send traffic.

Specifically, the network congestion control apparatus provided in this embodiment may further include a third control module 402, where the average delay of any one of the control periods after the monitoring is started is greater than or equal to a preset target delay. When the lower limit value of the interval is less than or equal to the upper limit value of the preset target delay interval, the instantaneous delay according to any one of the control cycles after the start monitoring is performed on any one of the control cycles after the start monitoring The allowed traffic of the service first message in the next control cycle is controlled. The instantaneous delay of any one of the control periods after the start of the monitoring is the network delay of the service success response received in the last unit time in any one of the control periods after the start of the monitoring.

Specifically, the third control module 402 in this embodiment may specifically include a first control unit 412 and a second control unit 422. The first control unit 412 is configured to reduce the start time of any one of the control periods after the start monitoring is greater than the immediate time delay of the last control period of any one of the control periods after the start of monitoring, and reduce the start The allowed traffic of the service first message in the next control cycle of any one of the control cycles after the monitoring. The second control unit 422 is configured to: when the start of the monitoring, the instantaneous delay of any one of the control periods is less than the instantaneous delay of the last control period of any one of the control periods after the start of the monitoring, and add any The traffic allowed for the first message of the service in the next control cycle of a control cycle.

Further, when the instantaneous delay of any one of the control periods after the startup monitoring is greater than the instantaneous delay of the last control period of any one of the control periods after the startup monitoring, the first control unit 412 is further configured to determine Whether the difference between the instantaneous delay of any one of the control periods after the start of monitoring and the instantaneous delay of the last control period of any one of the control periods after the start of monitoring is greater than or equal to a preset delay interval, and if so, The step of reducing the allowable transmission traffic of the service first message in the next control period of any one of the control periods after the start of the monitoring is performed. The second control unit 422 is further configured to determine the startup monitoring when the instantaneous delay of any one of the control periods after the startup monitoring is less than the immediate delay of the previous control period of any one of the control periods after the startup monitoring. Immediate time of the last control cycle of any subsequent control cycle And if the difference between the instantaneous delay of any one of the control periods after the start monitoring is greater than or equal to a preset delay interval, if yes, performing the adding of any one of the control periods after the start monitoring A step of allowing traffic to be sent for a service first message in a control cycle.

Further, the network congestion control apparatus provided in this embodiment may further include a setting module 403, configured to set an initial value of the allowed transmission traffic of the service first message to the following after the monitoring of the network status is initiated. The maximum of the two values: the actual successful traffic in the previous control cycle, and the minimum allowed traffic of the service first message.

Further, the network congestion control apparatus provided in this embodiment may further include a decoupling module 404, configured to: when the amount of traffic received in consecutive M control periods is less than or equal to the permission of the service first message When the traffic is sent, the control of the allowed transmission traffic of the service first message is stopped, where M is a pre-set positive integer greater than 1.

The embodiment provides a network congestion control device, if the service success rate in the continuous N control periods is less than the preset service success ratio threshold, and the network congestion notification is received, the monitoring of the network status is started, and the process starts. The network delay of the round-trip of the message in each control period after the monitoring is started is counted, and the allowed transmission of the service first message of the next control period of any one of the later control periods is reduced according to the average delay of any one of the obtained control periods. Traffic volume; Since the average delay of the control period can reflect the current network congestion degree, this embodiment avoids the problem that the existing HTR flow control is easy to cause error control due to only considering the service success rate, and the network congestion is realized. Effectively control, improve the business success rate after network congestion.

It should be noted that the above embodiments are only for explaining the technical solutions of the present invention, and are not intended to be limiting; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those skilled in the art that: The technical solutions described in the foregoing embodiments are modified, or some of the technical features are equivalently replaced; and the modifications or substitutions do not deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

Claim

A network congestion control method, comprising:

If the service success rate in the consecutive N control periods is less than the preset service success ratio threshold, and the network congestion notification is received, the monitoring of the network status is started, where N is a pre-set positive integer greater than 1.

2. The method according to claim 1, further comprising:

When the average delay of any one of the control periods after the start of monitoring is less than the lower limit of the preset target delay interval, the allowed transmission service of the service first message in the next control period of the any one of the control periods is increased. the amount.

The method according to claim 1 or 2, further comprising:

When the average delay of any one of the control periods after the start monitoring is greater than or equal to the lower limit value of the preset target delay interval, and less than or equal to the upper limit value of the preset target delay interval, according to the The instantaneous delay of any one control period after the monitoring is started controls the allowed transmission traffic of the service first message;

The instantaneous delay of any one of the control periods after the start of monitoring is the network delay of the service success response received in the last unit time in any one of the control periods after the start of the monitoring.

The method according to claim 3, wherein the controlling the allowable transmission traffic of the service first message according to the instantaneous delay of any one of the control periods after the startup monitoring comprises:

When the instantaneous delay of any one of the control periods after the startup monitoring is greater than the instantaneous delay of the last control period of any one of the control periods after the startup monitoring, reducing any one of the control periods after the startup monitoring Allowable transmission of service first message in a control cycle Quantity; or

When the instantaneous delay of any one of the control periods after the start monitoring is less than the instantaneous time delay of the last control period of any one of the control periods after the start of monitoring, increase any one of the control periods after the start monitoring The amount of traffic allowed for a service first message in a control cycle.

5. The method of claim 4, wherein

When the instantaneous delay of any one of the control periods after the startup monitoring is greater than the instantaneous delay of the last control period of any one of the control periods after the startup monitoring, the method further includes: determining the after the startup monitoring Whether the difference between the instantaneous delay of any one of the control periods and the instantaneous delay of the last control period of any one of the control periods after the start of monitoring is greater than or equal to a preset delay interval, and if so, the reduction is performed The step of enabling the transmission of the traffic of the service first message in the next control cycle of any one of the control cycles after the monitoring is started;

When the instantaneous delay of any one of the control periods after the startup monitoring is less than the instantaneous delay of the last control period of any one of the control periods after the startup monitoring, the method further includes: determining the after the startup monitoring Whether the difference between the instantaneous delay of the last control period of any one of the control periods and the instantaneous delay of any one of the control periods after the start of monitoring is greater than or equal to a preset delay interval, and if so, the increase is performed And the step of initiating the transmission of the traffic volume of the service first message in the next control period of any one of the control periods after the monitoring.

The method according to any one of claims 1-5, further comprising: after the starting monitoring of the network condition, the method further comprising:

The initial value of the allowable sending traffic of the service first message is set to the maximum of the following two values: the actual successful traffic in the previous control period, and the minimum allowable traffic volume of the service first message.

The method according to claim 1 or 2, further comprising:

When the amount of traffic received in the consecutive M control periods is less than or equal to the allowed transmission traffic of the service first message, stopping the control of the allowed transmission traffic of the service first message, where M is preset A positive integer greater than one.

8. A network congestion control device, comprising:

The control module is configured to start monitoring the network status if the service success rate in the consecutive N control periods is less than a preset service success ratio threshold, and the network congestion notification is received, where N is preset a positive integer greater than one; a statistics module, configured to perform statistics on a network delay of a message round-trip in each control period after the monitoring is started, and obtain an average delay of each control period after the startup monitoring;

9. The device according to claim 8, further comprising:

a second control module, configured to: when the average delay of any one of the control periods after the startup monitoring is less than a lower limit value of the preset target delay interval, increase the next control period of the any one of the control periods The first message of the business is allowed to send traffic.

The device according to claim 8 or 9, further comprising:

a third control module, configured to: when an average delay of any one of the control periods after the start monitoring is greater than or equal to a lower limit value of the preset target delay interval, and less than or equal to a preset target delay interval The limit time, according to the instantaneous delay of any one of the control periods after the startup monitoring, controlling the allowable transmission traffic of the service first message in the next control period of any one of the control periods after the startup monitoring;

The device of claim 10, wherein the third control module comprises: a first control unit, configured to: when the start monitoring is performed, an immediate delay of any one of the control cycles is greater than the start monitoring The instantaneous time delay of the last control period of any subsequent control period, reducing the allowable transmission traffic of the service first message in the next control period of any one of the control periods after the startup monitoring;

a second control unit, configured to: when the start of the monitoring, the instantaneous delay of any one of the control periods is less than the instantaneous delay of the previous control period of any one of the control periods after the start of the monitoring, and add any The traffic allowed for the first message of the service in the next control cycle of a control cycle.

12. Apparatus according to claim 11 wherein: The first control unit is further configured to determine that the instantaneous delay of any one of the control periods after the start of the monitoring is greater than the instantaneous time delay of the previous control period of any one of the control periods after the start of the monitoring Whether the difference between the instantaneous delay of any one of the control periods after the start of monitoring and the instantaneous delay of the last control period of any one of the control periods after the start of the monitoring is greater than or equal to the preset delay interval, and if so, is performed And the step of reducing the allowable transmission traffic of the service first message in the next control period of any one of the control periods after the startup monitoring;

The second control unit is further configured to determine that the instantaneous delay of any one of the control periods after the startup monitoring is less than the instantaneous delay of the last control period of any one of the control periods after the startup monitoring Whether the difference between the instantaneous delay of the last control period of any one of the control periods after the start of monitoring and the instantaneous delay of any one of the control periods after the start of the monitoring is greater than or equal to the preset delay interval, and if so, The step of increasing the allowable traffic of the service first message in the next control cycle of any one of the control periods after the start of the monitoring is performed.

The device according to any one of claims 8 to 12, further comprising: a setting module, configured to: after the starting the monitoring of the network condition, setting the permission to send the traffic of the service first message The initial value is the maximum of the following two values: the actual successful traffic in the previous control cycle, and the minimum allowed traffic of the service first message.

The device according to claim 8 or 9, further comprising:

a demodulation module, configured to stop controlling the allowed transmission traffic of the service first message when the traffic received in the consecutive M control periods is less than or equal to the allowed transmission traffic of the service first message, where , M is a pre-set positive integer greater than 1.