[go: up one dir, main page]

WO2013007180A1 - Procédé, dispositif et système de gestion de l'encombrement du réseau - Google Patents

Procédé, dispositif et système de gestion de l'encombrement du réseau Download PDF

Info

Publication number
WO2013007180A1
WO2013007180A1 PCT/CN2012/078375 CN2012078375W WO2013007180A1 WO 2013007180 A1 WO2013007180 A1 WO 2013007180A1 CN 2012078375 W CN2012078375 W CN 2012078375W WO 2013007180 A1 WO2013007180 A1 WO 2013007180A1
Authority
WO
WIPO (PCT)
Prior art keywords
congestion
network
rate
data stream
network node
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2012/078375
Other languages
English (en)
Chinese (zh)
Inventor
张登银
程春玲
李正
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huawei Technologies Co Ltd
Nanjing Post and Telecommunication University
Nanjing University of Posts and Telecommunications
Original Assignee
Huawei Technologies Co Ltd
Nanjing Post and Telecommunication University
Nanjing University of Posts and Telecommunications
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from CN201110191495.7A external-priority patent/CN102868671B/zh
Application filed by Huawei Technologies Co Ltd, Nanjing Post and Telecommunication University, Nanjing University of Posts and Telecommunications filed Critical Huawei Technologies Co Ltd
Publication of WO2013007180A1 publication Critical patent/WO2013007180A1/fr
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H04L47/24—Traffic characterised by specific attributes, e.g. priority or QoS
    • H04L47/2425—Traffic characterised by specific attributes, e.g. priority or QoS for supporting services specification, e.g. SLA
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H04L47/11—Identifying congestion

Definitions

  • the present invention relates to network communication technologies, and in particular, to a network congestion control method, device, and system. Background technique
  • Network congestion is a prominent problem in today's networks. It is extremely important to use congestion control to ensure network quality of service (QoS). Network congestion can be manifested by increased packet delay, increased packet loss rate, and reduced performance of upper-layer applications.
  • QoS quality of service
  • the existing congestion control technology can be classified into two types according to the location of the congestion control at the network layer, and the first is based on the source-based Transport Control Protocol (TCP) congestion control, which is executed in the host and the network edge device;
  • TCP Transport Control Protocol
  • One type is network-based IP congestion control, which is implemented in network devices.
  • Network congestion control may also be referred to as admission control.
  • Corresponding admission control algorithms include: parameter-based admission control algorithm, measurement-based admission control algorithm, bandwidth-based admission control algorithm, and policy-based admission control algorithm.
  • the existing congestion control technology can only reflect the congestion of a single node in the network, and cannot effectively alleviate the congestion of the entire link.
  • the existing admission control algorithm cannot reflect the user's will, and the network utilization rate is not high. Summary of the invention
  • Embodiments of the present invention provide a network congestion control method, device, and system, which effectively alleviate the whole Congestion of the link, and can reflect the user's will and high network utilization.
  • the embodiment of the invention provides a network congestion control method, including:
  • the ingress network node obtains a congestion judging parameter corresponding to the data stream and an input reference rate, where the congestion judging parameter is determined according to the accumulation, the input reference rate is an excellent solution for ensuring that the network is not congested; and the ingress network node determines the parameter according to the congestion. Determining whether network congestion occurs, if network congestion occurs, performing network congestion control on the data flow according to the input reference rate and SLA information corresponding to the data flow.
  • the embodiment of the invention provides a network congestion control device, including:
  • An obtaining module configured to obtain a congestion judging parameter and an input reference rate corresponding to the data stream, where the congestion judging parameter is determined according to an accumulation, and the input reference rate is an excellent solution for ensuring that the network is not congested;
  • a control module configured to determine, according to the congestion determination parameter, whether network congestion occurs, and if network congestion occurs, perform network congestion control on the data flow according to the input reference rate and SLA information corresponding to the data flow.
  • the embodiment of the invention provides a network congestion control system, including:
  • a network proxy configured to acquire an input rate of the ingress network node, and obtain an output rate of the egress network node, determine an accumulation according to the input rate and the output rate, determine a congestion determination parameter according to the accumulation, and determine to ensure that the network is not congested
  • the excellent solution is used as the input reference rate
  • an ingress network node configured to determine whether network congestion occurs according to the congestion determination parameter, and if network congestion occurs, perform network congestion control on the data flow according to the input reference rate and SLA information corresponding to the data flow.
  • the embodiment of the present invention determines whether the network is congested according to the congestion determination parameter, and the congestion determination parameter is determined according to the accumulation, and the accumulation may reflect the congestion of the entire network. Therefore, the embodiment of the present invention can reflect the whole Congestion of the network is not just a congestion of a node, which effectively alleviates the congestion of the entire link.
  • the embodiment of the present invention can reflect the user. Willingness;
  • the input reference rate is an excellent solution to ensure that the entire network is not congested, and can maximize the network utilization in ensuring the network is not congested.
  • FIG. 1 is a schematic flow chart of a method according to a first embodiment of the present invention
  • FIG. 2 is a schematic structural diagram of a system according to a second embodiment of the present invention.
  • FIG. 3 is a schematic flow chart of a method corresponding to a second embodiment of the present invention.
  • FIG. 4 is a schematic structural diagram of a system according to a third embodiment of the present invention.
  • FIG. 5 is a schematic flowchart of processing of a congestion judging module according to an embodiment of the present invention.
  • FIG. 6 is a schematic structural diagram of a network topology in an intra-domain according to an embodiment of the present invention.
  • FIG. 7 is a schematic flowchart of a method for dynamic admission control according to an embodiment of the present invention.
  • FIG. 8 is a schematic flow chart of a method according to a third embodiment of the present invention.
  • FIG. 9 is a schematic structural diagram of a device according to a fourth embodiment of the present invention.
  • FIG. 10 is a schematic structural diagram of a system according to a fifth embodiment of the present invention. detailed description
  • FIG. 1 is a schematic flowchart of a method according to a first embodiment of the present invention, including:
  • Step 11 The ingress network node obtains a congestion determination parameter and an input reference rate corresponding to the data flow, where the congestion determination parameter is determined according to an accumulation, and the input reference rate is an excellent solution for ensuring that the network is not congested;
  • the ingress network node may obtain the congestion judgment parameter and the input parameter rate from the network proxy.
  • the network agent can obtain the input rate of the ingress network node and obtain the output rate of the egress network node, and determine the accumulation based on the input rate and the output rate. Further, the congestion determination parameter is determined based on the accumulation, and a function indicating the network congestion condition is determined, and then the excellent solution of the function is obtained as the input reference rate.
  • the formula for calculating the accumulation, the congestion determination parameter, and the input reference rate can be found in the following examples.
  • Step 12 The ingress network node determines whether network congestion occurs according to the congestion determination parameter. If network congestion occurs, network congestion control is performed on the data flow according to the input reference rate and the SLA information corresponding to the data flow.
  • the output rate of the data flow at the ingress network node is adjusted to a threshold corresponding to the static degraded and a smaller value of the input reference rate.
  • the SLA information corresponding to the data flow indicates that dynamic degradation needs to be performed, adjusting an output rate of the data flow at the ingress network node to the input reference rate; or, if the SLA information corresponding to the data flow Indicates that the current service needs to be stopped, and then the link of the data flow is stopped.
  • the network determines whether the network is congested according to the congestion determination parameter, and the congestion determination parameter is determined according to the accumulation.
  • the accumulation may reflect the congestion of the entire network, and thus may reflect the congestion of the entire network, not just a node. Congestion, which effectively alleviates congestion on the entire link.
  • SLA in network congestion control
  • SLA is the embodiment of user's will, it can reflect the user's willingness
  • the input reference rate is an excellent solution to ensure that the entire network is not congested. Network utilization can be maximized while ensuring that the network is not congested.
  • 2 is a schematic structural diagram of a system according to a second embodiment of the present invention
  • FIG. 3 is a schematic flowchart of a method corresponding to a second embodiment of the present invention.
  • a source 21, an ingress network node 22, an egress network node 23, a destination 24, and a network proxy 25 are included.
  • the source 21 is configured to send a data request to the ingress network node 22.
  • the source 21 can perform source congestion control according to the congestion window size through the existing FAST TCP protocol.
  • the ingress network node 22 is a gateway or a network edge node in which the data stream enters the network from the source end, and completes control of the data stream to implement resource allocation.
  • the control of the data flow may include: a static traffic limiting based on a Service Level Agreement (SLA), a dynamic admission control based on an SLA and an input reference rate.
  • SLA Service Level Agreement
  • the ingress network node 22 determines that there is no congestion according to the congestion determination parameter, or does not have a congestion determination parameter, that is, under normal circumstances, according to the SLA information signed by the user and the operator, for each
  • the IP address carried in the data stream or the user name dynamically bound to the IP address is used to limit the flow of the front end to prevent an individual user from sending excessive traffic.
  • the dynamic admission control may include: static degradation, dynamic degradation, or stopping the current service. If it is statically degraded, the flow rate of the data stream is adjusted to a static threshold corresponding to the previously signed SLA information; if it is dynamic degraded, the flow rate of the data stream is adjusted according to the input reference rate; if the current service is stopped, the service is stopped. A link to the data stream. Further, if the flow rate is greater than the input reference rate after static degradation, the dynamic degradation can be performed again. In addition, the ingress network node 22 also sends its own input rate to the network proxy 25.
  • the egress network node 23 is the gateway or network edge node from which the data stream enters the destination 24 from the network. In addition, the egress network node 23 also transmits its own output rate to the network proxy 25.
  • the network proxy 25 is configured to acquire the input rate of the ingress network node 22 and the egress rate of the egress network node 23, and calculate the congestion judging parameter and the input reference rate accordingly, and then the congestion judging parameter. And the input reference rate is sent to the ingress network node.
  • the data stream can enter the network through the ingress network node 22 through the source 21, and then the ingress network node 22 sends the data stream through the network to the egress network node 23, which in turn is sent by the egress network node 23 to the destination end 24.
  • the method in this embodiment may include:
  • Step 31 The source sends a data stream to the network border ingress network node.
  • control based on the congestion window can be performed at the source end.
  • Step 32 The ingress network node forwards the data stream to the destination end through the network intermediate device and the egress network node.
  • the static current limiting is performed based on the SLA information, and the SLA information includes the corresponding flow rate according to the SLA information corresponding to the data flow, and then the data flow is processed according to the flow rate.
  • Step 33 The network proxy obtains the input rate of the ingress network node and the output rate of the egress network node.
  • the ingress network node actively collects the input rate and the egress network node actively collects the output rate, and then sends each network proxy separately.
  • the network proxy actively acquires the input rate and the output rate from the ingress network node and the egress network node, respectively.
  • Step 34 The network agent calculates a congestion judgment parameter and an input reference rate
  • the network proxy can calculate the congestion judgment parameter and the input reference rate of each data stream according to the input rate and the output rate of each data stream. For the specific calculation process, refer to the embodiment shown in Figure 5.
  • Step 35 The network proxy sends the congestion determination parameter and the input reference rate to the ingress network node.
  • the network proxy may actively send the congestion judgment parameter and the input reference rate to the ingress network node, or the network proxy may send the congestion judgment parameter and the input reference rate to the ingress network node after receiving the request of the ingress network node.
  • Step 36 The ingress network node performs dynamic admission control.
  • the ingress network node determines, according to the congestion judgment parameter, that congestion can be performed after congestion occurs State acceptance control. For example, according to the SLA information corresponding to the data flow, it is determined that the static degraded, dynamically degraded, or stopped the current service, and then the corresponding processing is performed. For the specific processing procedure, refer to the embodiment shown in FIG.
  • Step 37 The ingress network node sends the data flow after the dynamic admission control. Through admission control, data flow is reduced and congestion is relieved.
  • FIGS. 2 and 3 show an overall schematic diagram of an embodiment of the present invention.
  • the foregoing apparatus may specifically include the modules shown in the following embodiments.
  • the source end 41 may include a congestion window estimation module 411 and a congestion window control module 412.
  • the congestion window estimation module 411 measures or calculates congestion for determining.
  • the size of the window is used to calculate the size of the congestion window.
  • the congestion window control module 412 controls the packet forwarding rate according to the size of the congestion window calculated by the congestion window estimation module, so that the packet forwarding rate does not exceed the size of the congestion window.
  • the ingress network node 42 includes a static current limiting module 421, a dynamic admission control module 422, and a first flow rate acquisition module 423.
  • the static current limiting module 421 statically limits the data flow under normal conditions according to the SLA information obtained from the SLA database; the dynamic admission control module 422 performs dynamic admission control after congestion occurs.
  • the first flow rate acquisition module 423 is configured to send the input rate of the ingress network node to the network agent.
  • the ingress network node 42 may also include a billing database module 424 and a billing module 425 for billing customer consuming services.
  • the egress network node 43 can include a second flow rate acquisition module 431 for transmitting an output rate to the network agent.
  • the network proxy 44 may include an SLA database 441 and a congestion judging module 442; the SLA database 441 is configured to hold SLA information corresponding to the data stream, and the congestion judging module 442 is configured to obtain a congestion judging parameter and an input reference rate.
  • the congestion judging module is located on the network proxy and is used to calculate the congestion judging parameter and the input reference rate. At the same time, the calculated congestion judgment parameter and the input reference rate are provided to the dynamic admission control module. The congestion judging module needs to collect the flow from the first flow rate acquisition module and the second flow rate acquisition module. After the speed information, the accumulation of each data stream, the congestion judgment parameters, and the input reference rate are calculated.
  • the specific processing flow of the congestion judgment module can be seen in the following figure.
  • the data stream when the data stream is transmitted, the data stream is sent by the ingress network node to the egress network node. Therefore, there is a connection relationship between the ingress network node and the egress network node as shown in FIG. 2.
  • the calculation of each parameter and the corresponding control are mainly performed. Therefore, the connection relationship between the ingress network node and the egress network node is not reflected.
  • FIG. 5 is a schematic diagram of a process flow of a congestion judging module according to an embodiment of the present invention, including: Step 501: Acquire an input rate of an ingress network node and an output rate of an egress network node. Where, represents the input rate of the data stream, ⁇ '' represents the output rate of the data stream.
  • Step 502 Calculate the accumulation based on the input rate and the output rate.
  • the accumulation is a collection of data packets queued for a link, in all nodes through which the data flow passes.
  • the accumulation can be calculated as follows:
  • FIG. 6 is a schematic structural diagram of a network topology in an intra-domain according to an embodiment of the present invention, where an input rate of the data stream i at the bottleneck j is represented, and u represents an output rate of the data stream i at the bottleneck j.
  • M is the sequence number of the detection period, and r is the data sampling interval.
  • Step 503 Determine whether the accumulation is greater than the threshold. If yes, go to step 504. Otherwise, go to step 505.
  • the threshold value 6 can be set, for example, according to physical characteristics or human experience.
  • the network agent determines the congestion determination parameter according to the accumulation, and the ingress network node can determine whether the network is congested according to the congestion determination parameter to perform corresponding control.
  • Step 504 Set the congestion judgment parameter to r, and then perform step 507.
  • r can be any set value, usually r can be 2 or 3.
  • Step 505 Determine whether the congestion determination parameter is 0. If yes, go to step 508, otherwise, go to step 506.
  • Step 506 Decrease the congestion judgment parameter by 1.
  • ) indicates the congestion judgment parameter of the data stream in the period “, and the calculation formula can be: ⁇ b
  • a 1 (n) -a l (nY) + (X ( ⁇ ) - u (n)) x ⁇ (n) represents the input rate of the data stream in the nth detection period, M '(n) the nth detection The cycle represents the output rate of the data stream, and ⁇ is the data sampling time interval.
  • the threshold 6 in the equation represents the length of the queue that each stream attempts to maintain in each buffer of the network in the equilibrium state.
  • the value of 6 can be determined by default based on the physical characteristics of the device. For example, if the physical device has a memory size of 512K, then 6 can default to 512K:. Of course, the value of 6 can also be managed by the network.
  • the administrator is set, and the network administrator at this time can select a value smaller than the physical characteristic as the value of the 6 according to actual experience.
  • the Wu value is the number of rewards. r indicates that the current network condition is not congested when congestion does not occur in consecutive r periods ⁇ .
  • the value of r can be any set value, and can usually be taken as 2 or 3.
  • Step 507 Determine whether the congestion determination parameter is 0. If yes, go to step 508, otherwise go to step 511.
  • Step 508 Determine whether the actual input rate is less than or equal to the input reference rate. If yes, go to step 509, otherwise go to step 510.
  • the actual input rate refers to the input rate of the ingress network node.
  • the input reference rate has an initial value, which can be adjusted later.
  • the calculation formula can be as follows.
  • Step 509 Keep the input reference rate unchanged, and then repeat step 501 and subsequent steps.
  • Step 510 Increase the input reference rate, and then repeat step 501 and subsequent steps.
  • Step 511 The input reference rate is decreased, and then step 501 and subsequent steps are repeated. That is, the input reference rate is calculated as:
  • ⁇ is the sequence number of the detection period, which is the input reference rate of the data stream in the nth detection period, which is the value corresponding to the SLA of the data stream, and ( ⁇ ) represents the input rate of the data stream in the ⁇ th detection period, ⁇ '(n) represents the output rate of the data stream in the nth detection period, and T'(n) represents the congestion judgment parameter of the data stream in the nth detection period, and the sum is the set value.
  • V '' is the normal level determined in the SLA information signed by the user and the operator, that is, the corresponding flow rate at the time of static current limiting.
  • ' ⁇ ( «-1) 0 means that the network condition of the previous cycle is not congested
  • ⁇ ( «-1)> «'( «-1) means that the input rate of the previous cycle is greater than the reference input rate.
  • the network is not congested.
  • the reference input rate is too small and can be increased appropriately.
  • the added value is “, « is a static value preset for the system, indicating the rate value that can be appropriately increased.
  • ( «-1)>0 means that the network condition of the previous cycle is congestion, that is, the reference input rate of the previous cycle is too large and needs to be reduced.
  • the reduced value depends on ("-1), X M '(M_1) and ⁇ ' ⁇ ( «-1).
  • the input rate control method described above can correct the input rate of the previous detection period according to the current network condition, so that the input rate changes are relatively smooth, and the rate control is closer to the current network congestion.
  • the congestion judging module calculates the congestion judgment parameter and the input reference rate
  • the two parameters may be sent to the dynamic admission control module in the ingress network node for admission control.
  • the dynamic admission control module is located on the ingress network node, and obtains the SLA information and the congestion processing mode of the service according to the service requesting the user, and determines whether to accept the user request and accepts according to the current network resource condition;
  • the service performs resource scheduling.
  • Specific content can include:
  • the user's feedback operation on congestion can be stored in the SLA database in a regular form in advance.
  • the user and the operator agree in the signed SLA:
  • the processor for each type of service can be static downgrade, dynamic downgrade, stop the current business. That is, the foregoing congestion processing manner may include: static degrading, dynamic degrading, or stopping the current service. If the current network resource condition indicates that the network is congested, the SLA database may be queried and the SLA information and the congestion processing mode are obtained.
  • the SLA information includes the video service
  • the current service is a video service
  • the corresponding congestion processing mode is static degraded
  • the flow rate is limited to a static threshold determined when the SLA is signed in advance; if the corresponding congestion processing mode is dynamic degraded, the input reference rate is used as the threshold, and the dynamic speed limit is used; if the corresponding congestion processing mode is to stop the current service, the stop is stopped.
  • a link to a service data stream (such as a video service); when statically degraded, the flow rate is still greater than the input reference rate, and dynamic degradation can be performed again.
  • the obtained congestion processing mode is static degraded and dynamically degraded, it can be understood as accepting the user request.
  • the specific receiving manner can also be determined by the congestion processing method, for example, static degrading or dynamic degrading. After the static degrading or the dynamic degrading is determined, the corresponding processing may be performed according to the foregoing static degrading or dynamic degrading manner to complete resource scheduling for the accepted service.
  • the congestion judgment module of the network proxy provides a congestion judgment parameter for the dynamic admission control module of the ingress network node to determine whether the current service is congested.
  • the input reference rate is also provided.
  • the rate is used as the threshold; the SLA database provides User SLA information, based on this, performs corresponding operations, which can be static degraded, dynamically degraded, and stopped.
  • FIG. 7 is a schematic flowchart of a method for dynamic admission control according to an embodiment of the present invention, including: Step 701: Receive a congestion determination parameter and an input reference rate.
  • Step 702 Determine whether congestion occurs. If step 703 is performed, repeat the steps.
  • Step 703 Query the SLA information.
  • the SLA information corresponding to the data flow is queried from the SLA database.
  • Step 704 Determine whether a downgrade is required. If step 705 is performed, otherwise step 710 is performed. Step 705: Determine whether it is a static degradation. If the step 706 is performed, otherwise step 708 is performed.
  • the SLA information includes a processing manner, and the processing manner may include static degrading, dynamic degrading, or stopping the current service.
  • Step 706 Perform a static downgrade.
  • the SLA information may include a threshold corresponding to the static degraded.
  • the output rate of the data stream at the ingress network node is adjusted to the threshold.
  • Step 707 Determine whether the actual flow rate exceeds the input reference rate. If yes, go to step 708, otherwise go to step 709.
  • Step 708 Perform dynamic degradation.
  • the output rate of the data stream at the ingress network node is adjusted to the input reference rate.
  • Step 709 Perform resource allocation.
  • Step 710 Stop the current service.
  • the operator P signs an SLA agreement with the user A, and stores the SLA information of the user A in the SLA database.
  • the authentication module of ingress gateway C authenticates user A.
  • the authenticated user A requests the video on demand service from the streaming server D in the DiffServ domain.
  • the network agent G collects the flow rate information of the ingress gateway and the egress gateway, thereby calculating the accumulation, the congestion judging parameter and the input reference rate, wherein the network agent G may determine whether the current service is congested according to whether the current service is congested. For example, as described above, when the accumulation is larger than the threshold, it indicates that congestion has occurred, and the congestion judgment parameter can be set to .
  • the ingress gateway C performs static traffic limiting, flow rate collection, and charging processing on the video on demand service data stream; when congestion occurs, the ingress gateway C performs dynamic admission control, flow rate collection, and charging for the video on demand service data stream. deal with.
  • FIG. 8 is a schematic flowchart of a method according to a third embodiment of the present invention.
  • the embodiment includes: Step 801: User A sends a video-on-demand service data stream to the ingress gateway C.
  • Step 802 The static current limiting module of the ingress gateway C acquires the SLA information of the user A.
  • the user A's information is extracted from the data stream (assuming that the user A's ID number registered in the operator is 0001), and the SLA database is queried to obtain the user whose ID is 0001, which is a registered legal user and the current account has a balance of 200 yuan.
  • the user's SLA information is: Open video on demand service, the video on demand service bandwidth is 2M/s, agree to adopt dynamic degradation when congestion occurs, and the QoS parameter requires an average delay of 0.5s, an average packet loss rate of 0.5%, and arrive at 22 o'clock in the evening. The next day, 7:00, the total bandwidth of the user is 10M/s.
  • Step 803 The static current limiting module of the ingress gateway C performs static traffic limiting, so that the total traffic of the user A is 10 M/s.
  • Step 804 The network proxy G calculates an accumulation of the video on demand service data stream, a congestion determination parameter, and an input reference rate.
  • the input rate to the ingress gateway C is 2.5M/S; the calculated accumulation is 10M, and the current congestion is determined.
  • the calculated congestion judgment parameter is 5, and the input reference rate is reduced to 2M/s.
  • Step 805 The ingress gateway C receives the congestion judging parameter and the input reference rate from the network proxy G.
  • the congestion judgment parameter is 5
  • the input reference rate is 2M/s
  • the congestion judgment parameter is greater than
  • 0 can determine that the video on demand service is in a congested state.
  • Step 806 The dynamic admission control module of the ingress gateway C acquires the SLA information.
  • the SLA information of User A is queried to the SLA database, and the SLA information indicates that User A has signed a clause agreeing to dynamic downgrade.
  • Step 807 The dynamic admission control module of the ingress gateway C performs dynamic degradation.
  • the flow rate is adjusted from 2.2 M/S to 2 M/s.
  • Step 808 The first flow rate collection module of the portal gateway C writes the collected information into the billing database.
  • the collected information includes a flow rate, a QoS parameter, wherein the flow rate is 2.2 M/s, and the QoS parameters include: an average delay Is and an average packet loss rate of 1%.
  • the congestion judgment parameter is 3, the input reference rate is 2 M/s, data processing is performed, and the billing database is written;
  • the charging module retrieves the SLA information of the user A from the SLA database, and extracts the real-time rate from the billing database to 2.2 M/S and the time is 7200 s, and the calculated usage fee is 5 yuan and the access fee is 0.5 yuan; The total cost is 5.5 yuan.
  • the total cost of calculating the 30 days is 145 yuan.
  • the accumulation network as a network congestion signal
  • the monitoring complexity is reduced, the control signal quantity is reduced, and the bandwidth is saved. It can promote each node in the network to maintain a relatively stable queue length, reduce the delay of network measurement, and realize simple implementation, and the amount of information to be managed is small.
  • the ingress network node, the egress network node, and the network proxy they coordinate with each other, interact with each other, and work together to achieve joint control congestion.
  • the user signs the SLA, and determines that when the congestion occurs, the current service processing method is static degraded, dynamically degraded or stopped the current service, and the user is fed back to the congestion.
  • the operation is written into the SLA database in the form of SLA rules, which not only reflects the user's wishes, but also reduces the complexity of the operation.
  • the input reference rate which is the optimal value of the input rate in the current network
  • the input reference rate is used as the flow threshold at the time of dynamic degradation, and the network utilization is maximized while avoiding congestion.
  • FIG. 9 is a schematic structural diagram of a device according to a fourth embodiment of the present invention.
  • the device may be an ingress network node, for example, an ingress gateway or an ingress network edge node.
  • the device includes an obtaining module 91 and a control module 92.
  • the obtaining module 91 is configured to obtain a congestion judging parameter and an input reference rate corresponding to the data stream, where the congestion judging parameter is determined according to an accumulation, and the input reference rate is to ensure that the network is not congested.
  • An optimal solution; the control module 92 is configured to determine, according to the congestion determination parameter, whether network congestion occurs, and if network congestion occurs, perform network congestion on the data flow according to the input reference rate and SLA information corresponding to the data flow. control.
  • control module is specifically configured to: if the SLA information corresponding to the data flow indicates that static degrading is required, adjust an output rate of the data flow at the ingress network node to a threshold corresponding to the static degraded And a smaller value of the input reference rate; or, if the SLA information corresponding to the data stream indicates that dynamic degradation needs to be performed, adjusting an output rate of the data stream at the ingress network node to the input reference rate; Alternatively, if the SLA information corresponding to the data flow indicates that the current service needs to be stopped, the link of the data flow is stopped.
  • the control module may be specifically the dynamic admission control module described above.
  • control module is further configured to: if it is determined that network congestion does not occur according to the congestion determination parameter, adjust an output rate of the data stream at an ingress network node to an SLA information corresponding to the data flow. Static current limit threshold.
  • control module may be specifically the static admission control module described above.
  • the acquiring module is configured to receive the congestion determination parameter and an input reference rate that are sent by the network proxy; or send a request to the network proxy, and receive the network proxy after receiving the request.
  • the congestion judgment parameter and the input reference rate are transmitted.
  • the network determines whether the network is congested according to the congestion determination parameter, and the congestion determination parameter is determined according to the accumulation.
  • the accumulation may reflect the congestion of the entire network, and thus may reflect the congestion of the entire network, not just a node. Congestion, which effectively alleviates congestion on the entire link.
  • SLA in network congestion control
  • SLA is the embodiment of user's will, it can reflect the user's willingness
  • the input reference rate is an excellent solution to ensure that the entire network is not congested. Network utilization can be maximized while ensuring that the network is not congested.
  • FIG. 10 is a schematic structural diagram of a system according to a fifth embodiment of the present invention, including a network proxy 101 and an ingress network node 102; the network proxy 101 is configured to acquire an input rate of an ingress network node, and obtain an output rate of an egress network node, according to the input. The rate and the output rate determine an accumulation, determine a congestion determination parameter based on the accumulation, and determine a superior solution that ensures that the network is not congested as an input reference rate; and the ingress network node 102 is configured to determine whether network congestion occurs according to the congestion determination parameter. Plugging, if network congestion occurs, performing network congestion control on the data stream according to the input reference rate and SLA information corresponding to the data stream.
  • the network proxy may determine the network congestion determination parameter by using the following formula:
  • « is the sequence number of the detection period
  • '( ⁇ ) indicates that the congestion r and 6 of the data stream in the nth detection period are preset values
  • ⁇ ' ( ⁇ ) indicates the integrated formula of the data stream in the nth detection period for:
  • ( ⁇ ) represents the input rate of the data stream in the nth detection period
  • ⁇ '( ⁇ ) represents the output rate of the data stream in the nth detection period
  • r is the data sampling time interval.
  • the input reference rate of the data stream in the nth detection period is a value corresponding to the SLA of the data stream
  • (n) represents the input rate of the data stream i in the nth detection period
  • ⁇ '' ( ⁇ ) represents the output rate of the data stream in the nth detection period
  • ⁇ ⁇ ( ⁇ ) represents the congestion judgment parameter of the data stream in the nth detection period
  • the above module for calculating the network congestion determination parameter and the input reference rate may be specifically the above congestion determination module.
  • the ingress network node may be specifically configured to: if the congestion determination parameter is greater than 0, determine that network congestion occurs, otherwise determine that network congestion does not occur.
  • the network proxy may be specifically configured to: receive the input rate and output rate that are respectively sent by the ingress network node and the egress network node, respectively; or send a request to the ingress network node and the egress network node, respectively, and receive the The ingress network node and the egress network node are received The input rate and output rate respectively sent after the request.
  • the network determines whether the network is congested according to the congestion determination parameter, and the congestion determination parameter is determined according to the accumulation.
  • the accumulation may reflect the congestion of the entire network, and thus may reflect the congestion of the entire network, not just a node. Congestion, which effectively alleviates congestion on the entire link.
  • SLA in network congestion control
  • SLA is the embodiment of user's will, it can reflect the user's willingness
  • the input reference rate is an excellent solution to ensure that the entire network is not congested. Network utilization can be maximized while ensuring that the network is not congested.
  • the foregoing program may be stored in a computer readable storage medium, and when executed, the program includes The foregoing steps of the method embodiment; 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.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

La présente invention porte sur un procédé, un dispositif et un système de gestion de l'encombrement du réseau. Le procédé comprend les étapes suivantes : un nœud de réseau d'entrée obtient un paramètre de détermination d'encombrement et un débit de référence d'entrée correspondant à des flux de données, le paramètre de détermination d'encombrement étant défini par des agrégats de réseau et le débit de référence d'entrée étant la valeur optimale permettant d'assurer que l'encombrement du réseau ne se produit pas ; le nœud de réseau d'entrée détermine si l'encombrement du réseau s'est produit ou non en fonction du paramètre de détermination d'encombrement, et si l'encombrement du réseau s'est produit, applique une gestion de l'encombrement du réseau aux flux de données conformément au débit de référence d'entrée et aux informations SLA correspondant aux flux de données. Les modes de réalisation de la présente invention peuvent efficacement réduire l'encombrement d'une liaison entière, et refléter le souhait d'utilisateurs et avoir un fort taux d'utilisation du réseau.
PCT/CN2012/078375 2011-07-08 2012-07-09 Procédé, dispositif et système de gestion de l'encombrement du réseau Ceased WO2013007180A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201110191495.7 2011-07-08
CN201110191495.7A CN102868671B (zh) 2011-07-08 网络拥塞控制方法、设备及系统

Publications (1)

Publication Number Publication Date
WO2013007180A1 true WO2013007180A1 (fr) 2013-01-17

Family

ID=47447269

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2012/078375 Ceased WO2013007180A1 (fr) 2011-07-08 2012-07-09 Procédé, dispositif et système de gestion de l'encombrement du réseau

Country Status (1)

Country Link
WO (1) WO2013007180A1 (fr)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2000028706A1 (fr) * 1998-11-10 2000-05-18 Odyssia Systems, Inc. Procede et appareil pour limiter l'encombrement dans un reseau a commutation par paquets
CN1610287A (zh) * 2004-11-23 2005-04-27 北京航空航天大学 分布式仿真系统数据传输拥塞控制方法
CN1780465A (zh) * 2004-11-18 2006-05-31 华为技术有限公司 一种实现过载拥塞控制的方法
US7936694B2 (en) * 2006-04-03 2011-05-03 Hewlett-Packard Development Company, L.P. Sniffing-based network monitoring

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2000028706A1 (fr) * 1998-11-10 2000-05-18 Odyssia Systems, Inc. Procede et appareil pour limiter l'encombrement dans un reseau a commutation par paquets
CN1780465A (zh) * 2004-11-18 2006-05-31 华为技术有限公司 一种实现过载拥塞控制的方法
CN1610287A (zh) * 2004-11-23 2005-04-27 北京航空航天大学 分布式仿真系统数据传输拥塞控制方法
US7936694B2 (en) * 2006-04-03 2011-05-03 Hewlett-Packard Development Company, L.P. Sniffing-based network monitoring

Also Published As

Publication number Publication date
CN102868671A (zh) 2013-01-09

Similar Documents

Publication Publication Date Title
US11558302B2 (en) Data transmission method and apparatus
CN103999409B (zh) 用于带宽变化移动数据网络的链路缓冲器大小和队列长度估算的方法
Zhang et al. Novel quick start (QS) method for optimization of TCP
US9596281B2 (en) Transport accelerator implementing request manager and connection manager functionality
US6839767B1 (en) Admission control for aggregate data flows based on a threshold adjusted according to the frequency of traffic congestion notification
US8509074B1 (en) System, method, and computer program product for controlling the rate of a network flow and groups of network flows
CN112104562B (zh) 拥塞控制方法及装置、通信网络、计算机存储介质
WO2010000168A1 (fr) Procédé, appareil et système d'ajustement de rendement de codage multimédia
CN110730469B (zh) 一种基于扩展卡尔曼无线网络带宽预测方法及其拥塞控制
WO2011009304A1 (fr) Procédé et dispositif destinés à mesurer une largeur de bande disponible à l'aide de messages mélangés
CN101599965A (zh) 基于测量的自适应高速信息传输协议
EP3295612A1 (fr) Gestion de performances de liaison montante
CN103916329A (zh) 一种命名数据网络传输控制方法及系统
Pokhrel et al. Fair coexistence of regular and multipath TCP over wireless last-miles
CN103428104B (zh) 一种基于内容中心网络的拥塞控制方法
WO2013017001A1 (fr) Procédé, dispositif et système pour réaliser une auto-adaptation de performances d'application
US12348433B2 (en) Method and system for dynamic quota-based congestion management
JP5249966B2 (ja) スループットを向上させる通信制御方法および通信システム、ならびにそのためのプログラム
CN102739508B (zh) 一种高速网络数据传输的方法及系统
US20130308454A1 (en) Method and apparatus for improving http adaptive streaming performance using tcp modifications at content source
WO2013007180A1 (fr) Procédé, dispositif et système de gestion de l'encombrement du réseau
JP4238150B2 (ja) 通信網におけるレート制御方法およびレート制御システム
JP2008236709A (ja) 通信品質制御方法と通信品質制御装置および通信システムとプログラム
Seferoglu et al. Congestion state-based dynamic FEC algorithm for media friendly transport layer
KR100737678B1 (ko) 멀티미디어 스트리밍 서비스에 대한 지연시간 분석방법

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 12811058

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 12811058

Country of ref document: EP

Kind code of ref document: A1