WO2012149748A1

WO2012149748A1 - Network optimization traffic control method, device and system

Info

Publication number: WO2012149748A1
Application number: PCT/CN2011/079827
Authority: WO
Inventors: 项炎平; 文刘飞
Original assignee: 华为技术有限公司
Priority date: 2011-09-19
Filing date: 2011-09-19
Publication date: 2012-11-08
Also published as: CN102388585B; CN102388585A

Abstract

Provided are a network optimization traffic control method, device and system. The method includes: a data source node obtaining a sub-data stream transmission path between the data source node and a server node according to network topology information, and loading first data stream traffic on the sub-data stream transmission path; a router node aggregating the link prices of several links in the sub-data stream transmission path to obtain a link aggregation price, and feeding back the link aggregation price to the data source node; the data source node obtaining second data stream traffic according to the link aggregation price and the server weight of the server node, and replacing the first data stream traffic with the second data stream traffic and loading the same to the sub-data stream transmission path. The embodiments in the present invention greatly reduce the calculation amount, improving the timeliness of network optimization.

Description

Network optimization flow control method, device and system

Technical field

The present invention relates to communication technologies, and in particular, to a network optimization flow control method, apparatus, and system. Background technique

The traditional Internet Service Provider (ISP) mainly provides Internet connection. It mainly solves the problem of traffic engineering (TE) in network optimization functions, specifies the optimal path for traffic, and minimizes the congestion degree of the network. The performance of the network is the optimization goal; the content Provider (CP) mainly provides the user with the required content, provides the user with reasonable resources, solves the problem of server selection (SS), and the user experience is optimize the target. If the ISP and CP optimization are combined, a joint optimization (JO) of TE and SS can be performed to obtain an optimal solution for the entire network, minimizing link congestion while minimizing end-to-end delay. Achieving a globally optimal balance of network performance can significantly improve network performance. In the prior art network optimization flow control method, an optimization processing server is usually set separately, and the server needs to collect network state information (such as link capacity, link traffic, user request information, and server status information) required for optimization problem solving. Etc.), obtain the optimal solution of the network in a centralized manner, and apply the optimal solution to the network. The centralized method refers to that the network has multiple data source nodes and multiple server nodes as data transmission destinations, and correspondingly has multiple sub-data stream transmission paths, and data flows of all sub-data stream transmission paths are It is uniformly processed by the optimization processing server, and then the obtained optimal data traffic is loaded into the sub-data stream transmission path in the network, and the network optimal solution is usually a sub-data stream between the data source node and the server node in the network. The data traffic of the transmission path.

However, the above prior art has the following technical defects: Since the solution is solved in a centralized manner, in the actual network, more network state information is needed to solve the optimal problem, especially in the case of When the network scale is expanded, the information demand increases with the number of nodes, the time spent collecting information and calculations will increase sharply, and the calculation amount is too large, and the network nodes are difficult to react in real time, which brings great difficulty to the implementation of the optimization scheme. Summary of the invention

The embodiment of the invention provides a network optimization flow control method, device and system, so as to realize joint optimization of ISP and CP, and the calculation amount is small, and the speed and real-time performance of the optimization processing are improved. An embodiment of the present invention provides a network optimization flow control method, where the network includes a data source node, a router node, and a server node, where two adjacent router nodes are a link, and the two adjacent nodes The router node constitutes a link control module; the method includes: the data source node obtaining, according to network topology information, a sub-data stream transmission path between the data source node and the server node, and in the sub- Loading the first data stream traffic on the data stream transmission path; one of the sub-data stream transmission paths includes a plurality of the links;

The router node aggregates traffic of the first data stream on the link into link traffic, and the link control module obtains a link price of the link according to the link traffic; Link price aggregation of several links in the sub-data stream transmission path to obtain a link aggregation price, and feedback the link aggregation price to the data source node; the data source node aggregates the price according to the link And the server weight of the server node, obtaining the second data stream traffic, and loading the second data stream traffic to the first data stream traffic to the sub-data stream transmission path. An embodiment of the present invention provides a network optimization flow control apparatus, including: an initial loading module, configured to obtain a sub-data stream transmission path between a data source node and a server node according to network topology information, and Loading the first data stream traffic on the sub-data stream transmission path; one of the sub-data stream transmission paths includes a plurality of the links; a feedback receiving module, configured to receive a link aggregation price of the sub-data stream transmission path fed back by the router node, where the link aggregation price is a chain of the sub-data stream transmission path by the router node The traffic price adjustment module is configured to: obtain, according to the link aggregation price, the server weight of the server node, the second data flow, and replace the second data flow with the first data Stream traffic is loaded to the sub-data stream transmission path. Another aspect of the present invention provides a network optimization flow control system, including: a data source node, a server node, and a routing node, where two adjacent router nodes are a link, and the two adjacent nodes The router node constitutes a link control module, and the data source node is configured to obtain, according to the network topology information, a sub-data stream transmission path between the data source node and the server node, and in the sub-data stream Loading a first data stream traffic on the transmission path; one of the sub-data stream transmission paths includes a plurality of the links; and further configured to obtain, according to a link aggregation price fed back by the router node, and a server weight of the server node, a second data stream traffic, and the second data stream traffic is replaced by the first data stream traffic to the sub-data stream transmission path; the router node is configured to use the first data on the link Flow traffic is aggregated into link traffic, and the link control module obtains a link price of the link according to the link traffic; Some price link streaming link path polymerization link aggregation price, and the price of the link aggregation is fed back to the data source node. The network optimization flow control method, device and system according to the embodiment of the present invention respectively solve the problem that the network optimization traffic calculation information demand is large by decomposing the traffic calculation in the network optimization into each node and the link in the network. It greatly reduces the amount of calculation and improves the real-time performance of network optimization. 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 to be used in the description of the embodiments will be briefly made. It is obvious that the drawings in the following description are the present invention. For some embodiments, other drawings may be obtained from those of ordinary skill in the art without departing from the drawings.

1 is a schematic flowchart of Embodiment 1 of a network optimization flow control method according to the present invention;

2 is a schematic flowchart of Embodiment 2 of a network optimization flow control method according to the present invention;

3 is a schematic diagram of dividing a sub-data stream transmission path in Embodiment 2 of the network optimization flow control method according to the present invention;

4 is a top view of a simulation example in the second embodiment of the network optimized flow control method of the present invention; FIG. 5 is a flow state change diagram obtained by solving the exponential function in the second embodiment of the network optimized flow control method of the present invention;

6 is a diagram showing a state of link utilization state obtained by solving an exponential function in the second embodiment of the network optimization flow control method of the present invention;

7 is a flow state change diagram obtained by solving a power function in the second embodiment of the network optimization flow control method according to the present invention;

8 is a diagram showing a state change of a link utilization state obtained by solving a power function in the second embodiment of the network optimization flow control method of the present invention;

9 is a schematic structural diagram of Embodiment 1 of a network optimization flow control apparatus according to the present invention;

FIG. 10 is a schematic structural diagram of Embodiment 2 of a network optimized flow control apparatus according to the present invention; FIG. 11 is a schematic structural diagram of an embodiment of a network optimized flow control system according to the present invention. detailed description

The technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Some embodiments, rather than all of the embodiments, are invented. Based on the embodiments of the present invention, those obtained by those skilled in the art obtained without creative labor All other embodiments are within the scope of the invention. The main technical solution of the embodiment of the present invention is to decompose the traffic calculation in the network optimization into each node and the link in the network, for example, calculate the sub-data flow of each sub-data stream transmission path in the network separately. Moreover, the parameters required in the flow calculation are also distributed in the network node in the sub-stream transmission path, thereby greatly reducing the calculation amount and improving the network optimization compared with the centralized processing method of the optimization processing server in the prior art. Real time. The technical solution of the present invention will be further described in detail below through the accompanying drawings and embodiments. Embodiment 1 FIG. 1 is a schematic flowchart of Embodiment 1 of a network optimization flow control method according to the present invention, wherein a network source node, a router node, and a server node may be included in the network; and, a chain between two adjacent router nodes The two adjacent router nodes form a link control module. As shown in FIG. 1 , the flow control method in this embodiment may include the following steps: Step 101: A data source node obtains a sub-data stream transmission path between the data source node and the server node according to network topology information. And loading the first data stream traffic on the sub-data stream transmission path; for example, compared with the prior art, in the prior art, the network processing traffic is calculated by using an optimization processing server separately set outside the network, but not in this embodiment. Then, the above optimization processing server is set, and the traffic calculation is performed directly through the data source node in the network. Moreover, the network includes a plurality of data source nodes, and the plurality of data source nodes respectively calculate traffic on the sub-data stream transmission path to which they are connected. The source node of the data stream transmission is a data source node, the destination node of the data stream transmission is a server node, and the path between the data source node and the server node is a data stream transmission path. In this embodiment, each data source node may divide, according to network topology information, a plurality of sub-data stream transmission paths to the server node according to possible transmission paths of the data streams. And, the data stream is from the data source node There may be multiple router nodes passing between the server nodes, and correspondingly, there may be several links on a sub-data stream transmission path. The foregoing network topology information may be obtained by the data source node from the router node. The data source node loads the first data stream traffic on each of the sub-data stream transmission paths obtained as the initial traffic, and the initial traffic may be a preset starting value, which is a smaller constant. For example, suppose there are three sub-stream transmission paths between the data source node A and the server node B, then the first data stream traffic is loaded in the three sub-stream data transmission paths, but the number of each sub-data stream transmission path is A data stream traffic may be the same or different. Step 102: The router node acquires link prices of several links in the sub-data stream transmission path, and aggregates them to obtain a link aggregation price, and feeds back the link aggregation price to the data source node; for example, a router node The first data stream traffic obtained in step 101 can be loaded onto each link in the sub-data stream transmission path according to the routing matrix. Wherein, it may be that multiple sub-data stream transmission paths pass through the same link, and corresponding data may be loaded on the link; for example, the link between the router node A and the router node B is simultaneously The first sub-stream transmission path and the second sub-data stream transmission path are loaded with the traffic a on the first sub-stream transmission path and the traffic b on the second sub-stream transmission path. Router node A can aggregate traffic a and traffic b into link traffic.

The link control module can calculate the link price based on the link traffic obtained above. The calculation of the link price can be performed in a variety of ways, depending on the performance requirements of the network optimization target; for example, in network optimization, the link utilization is more focused on, and it is desirable to make the link utilization ratio better. If it is high, the link price can be set to reflect the link utilization index, such as "link price = link traffic / link maximum capacity". There may be multiple links on a sub-stream transmission path, and each link can obtain the link price according to the above manner. The router node may aggregate the link price of the links in the sub-stream transmission path to obtain the link aggregation price, and feed back the link aggregation price to the data source. Node. The link aggregation price may be calculated in multiple ways. For example, the maximum of the multiple link prices may be selected as the link aggregation price, or the multiple link prices may be added to obtain the link aggregation price, or The multiple link prices are separately weighted to obtain the link aggregation price and the like.

Step 103: The data source node obtains the second data flow according to the link aggregation price and the server weight of the server node, and loads the second data flow to replace the first data flow to the sub data flow. For example, the link price is an indicator reflecting the link utilization. The data source node can adjust the initial traffic loaded in step 101 according to the link price. If the link price is lower, the link utilization rate is used. It is still low, and the link traffic can be increased. Then, the data source node can continue to increase the traffic based on the initial traffic of step 101 according to a certain growth factor. The data source node may also adjust the initial traffic loaded in step 101 according to the server weight of the server node. The server weight may reflect the processing capability of the server. For example, the server may be weighted according to performance parameters such as server load and service average delay. Value, if the server weight is high, it indicates that the server has high processing capacity (such as small load, small delay, etc.), then the data source node can continue to increase traffic based on the initial traffic based on a certain growth factor. . After the foregoing adjustment to the traffic, the data source node can obtain the second data flow, and the second data flow is the adjusted traffic, and the second data flow can be replaced by the first data flow in the step 101, and the traffic is loaded. To the sub-streaming path. In the method of the embodiment, the link performance requirement (link price) and the server performance requirement (server weight) are considered at the same time, which is a joint optimization between the ISP and the CP, and the global optimal balance of the network performance can be obtained.

Compared with the prior art, the network optimization flow control method in this embodiment is not only a distributed computing method, that is, the traffic calculation is distributed to each data source node and the network node for calculation, and the centralized processing method is compared with the prior art. The network optimization flow control method in this embodiment is a process of dynamically adjusting the traffic, and the traffic can be adjusted in real time according to the network link state and the server state. Relative to the prior art Intraoperative static optimization calculation method can dynamically adapt to network changes, quickly respond to network abnormal states, and the system is robust. The network optimization flow control method of the embodiment is implemented by decomposing the traffic calculation in the network optimization into each node and the link in the network, thereby solving the problem that the network optimization traffic calculation information demand is large, and the calculation amount is greatly reduced. , improve the real-time performance of network optimization. The second embodiment of the present invention is basically the same as the first embodiment, but the steps in the first embodiment are described in more detail. As shown in FIG. 2, the flow control method of this embodiment may include the following steps: Step 201: A data source node divides, according to network topology information, a sub-data stream transmission path between the data source node and the server node. And loading the first data stream traffic on the sub-data stream transmission path; for example, as shown in FIG. 3, FIG. 3 is a schematic diagram of the sub-data stream transmission path division in the second embodiment of the network optimization flow control method according to the present invention. There may be multiple data transmission paths between the data source node R1 and the server node S1, and two are shown in FIG. 3, which are xl l and xl2, respectively. The data source node performs the division of the sub-data stream transmission path according to the network topology information, and the network topology information may be received by the router node. Specifically, the router node broadcasts the collected network topology information to the neighboring node, and after receiving the information, the data source node may locally generate a network topology table; the data source node further according to the content of the network topology table, The possible transmission path division of the data stream to the server node results in several independent sub-stream transmission paths. The data source node can calculate and generate the first data stream traffic, as the initial traffic, and load it onto the obtained sub-data stream transmission path. The intermediate router needs to perform a corresponding forwarding policy when the data stream is loaded, and may include loading based on MPLS and IP-based. For example, based on MPLS, since the router node supports the MPLS data pipe function, the first number can be directly According to the flow rate, the flow is loaded to the MPLS pipe, and the path is forwarded according to the path when the pipeline is established. When the IP is based, the path identifier may be added to the data packet corresponding to the first data flow, and the router node may forward the data packet according to the path identifier.

Step 202: The router node aggregates the traffic on the link to obtain the link traffic. The router node may load the first data flow obtained in step 201 into each chain in the sub-data transmission path according to the routing matrix. On the road. Among them, it is possible that multiple sub-data stream transmission paths pass through the same link, and corresponding data may be loaded on the link. For example, the link between router node A and router node B is passed by xl l and xl2 at the same time, then the traffic on a1 l and the traffic on xl2 are loaded on the link 1). Router node A can aggregate traffic a and traffic b into link traffic.

Step 203: The link control module obtains a link price of the link according to the link traffic. The link price is a parameter that reflects link performance, and the link control module can directly set according to performance requirements of the link. . For example, in network optimization, focusing on link utilization and hoping to make the link utilization higher, the link price can be set to reflect the link utilization index, such as "link price = chain Road traffic / link maximum capacity". Link based on a predetermined price utility function ^{/ (setting} ^ linear utility function associated with the link, the utility function is a nonlinear function, for example, monotonic convex function; link control module can link the utility function according ^ Update the link price, which can be used as the price obtained on a certain link. Among them, the utility function ^{/ (} calculated from the link state information, such as the link traffic, link delay, chain obtained in step 202, The length of the control cache queue required by the road, etc. Among them, setting the utility function as a nonlinear function can make the optimization method have a faster response speed and optimize the traffic to approach the target balance point flow faster.

In addition, the link control module can synchronize the topology table of the network and the link calculation parameters to synchronize the distributed calculation errors between the links. Step 204: The router node aggregates link price of several links in the sub-data stream transmission path to obtain a link aggregation price, and feeds back the link aggregation price to the data source node; for example, a sub-data There may be multiple links on the streaming path, and each link can obtain the link price according to the above method. The router node may aggregate the link prices of the links in the sub-stream transmission path to obtain a link aggregation price, and feed back the link aggregation price to the data source node. The router node aggregates the link prices of the links on the sub-data stream transmission path according to the selected fairness rule; the fairness rule generally has a maximum and minimum fairness criterion and a proportional fairness criterion, respectively, corresponding to different Link aggregation method; for example, if the maximum and minimum fairness criterion is selected, the maximum of the multiple link prices can be selected as the link aggregation price, or the proportional fairness criterion is selected, and the multiple link prices are added to obtain the link. Aggregate the price, or select the weighted proportional fairness criterion, then add the weights of the multiple linkes separately to obtain the link aggregation price and so on. Step 205: The data source node obtains the second data flow according to the link aggregation price and the server weight of the server node, and loads the second data flow to replace the first data flow to the sub data flow. The transmission path; for example, the server weight may be an indicator reflecting the processing capability of the server node, wherein the link price represents an optimization requirement of the ISP, and the server weight represents the optimization requirement of the CP, that is, the parameter selection of the server node is determined according to the server node. Performance information dynamically assigns weights to server nodes. In this embodiment, the server weight is calculated according to the following formula (1):

w = H(y, D) (1)

The above formula (1) can be, for example: = 0, 0) = ; ^/^. ) + (1 _ (^/3⁄4) _; where: is a constant, and o ≤ r ≤ i;

^. Data traffic that the server can process per unit of time; For reference delay constant; in the above formula (1), y is the total load of the server node, and D is the service average delay of the server node. The server performance information (such as load, delay, etc.) may be sent by the server node to the data source node, and the data source node calculates the server weight of the server node according to the server performance information. During the process of sending a server node to a data source node, the router nodes in the network can forward. Since each sub-data stream is independently calculated, when a plurality of data streams pass through a link, the plurality of data streams will compete, and the capability of competing can pass the server weight of the target server node corresponding to each sub-data stream. For example, if the server has a higher weight, it indicates that the server has higher processing capacity (such as smaller load, less delay, etc.), and the sub-data stream corresponding to the server node can appropriately increase the traffic. Through the adaptive adjustment between the above sub-data streams, the goals of TE and SS can be achieved simultaneously. The above calculation of the server weight has no time step limitation, as long as it can be obtained before the flow calculation. After obtaining the server weight, the data source node can calculate the traffic of the sub-stream transmission path according to the link aggregation price and the server weight of the server node. Specifically, the data source node can perform traffic calculation according to the following formula (2):

among them,

Is a constant, and 0 < ;

w is the weight of the connected server; the result of the above formula (2) is the derivative of the flow, and the flow can be calculated according to the derivative value. The formula (2) calculates the flow is a greedy trial calculation method, the formula includes two parts, namely the rate increase part and the rate limit part. Among them, the rate increase part is a greedy expansion mode, the expansion rate is related to the price and the server weight; the ' ^w ' ( ) is added per unit time; The rate limiting part means limiting greedy expansion. The higher the rate of data flow, the higher the expansion constraint, and the rate is limited by ' ^X <. Wherein, the utility function in the formula (2) has a heuristic factor, and the value is based on the following formula (3):

a is a normal number;

q is the aggregate price obtained for each sub-stream; through the above formula, a small oscillation can be realized, which can make the calculated flow finally oscillate slightly near the equilibrium point, realizing the dynamic approximation of the flow, thus requiring less network information. Achieve higher control accuracy, that is, to improve the degree of approximation. Among them, the balance point is a target state of global optimization of the network, which is equivalent to the ideal optimal solution solved by static optimization in the prior art. In this embodiment, the balance point corresponds to x; = H(_y, D)./( ), which is consistent with the equilibrium point of the NBS (Nash bargaining solution) static solution, and the flow calculation method of the embodiment can be automatically approximated. The best of the presets, the optimization of the entire network in dynamic competition. The traffic calculated according to the above formula in the step 205 is the traffic adjusted according to the network state information and the server state information, and may be referred to as the second data flow. The data source node may replace the second data stream traffic with the first data stream traffic in step 201 and load onto the sub-data stream transmission path. Then, after loading the second data stream traffic, the router node can perform traffic aggregation and subsequent steps as in step 202. The entire network automatically obtains the network optimized traffic control result in the dynamic feedback adjustment shown in FIG. 2.

The following is a simulation example to illustrate the flow control result and static result finally achieved by the method of this embodiment. The state optimal solution is consistent, and the method of this embodiment is effective:

Referring to FIG. 4, FIG. 4 is a topological diagram of a simulation example in Embodiment 2 of a network optimization flow control method according to the present invention. The source data node is connected to a sub-stream transmission path, and the sub-data stream transmission path connected to the source 1 includes a three-segment link, and the three-segment link is a source link, a source link, and a source node. Linkl, Link2 and Link3. Assume that the capacity of the above three links is c. = 10000; This capacity is the maximum transmission rate between router nodes and is the normalized capacity, which can be, for example, kb.

If solved in a static manner, the static optimal solution of the network is as follows:

The traffic distribution on the four sub-stream transmission paths in the network is represented. For example, the traffic on the four sub-stream transmission paths in the embodiment should be the link capacity C. 1/3, y represents the link utilization of the three-segment link. For example, the utilization rates of Linkl, Link2, and Link3 in this embodiment are 2/3*C, respectively. , C. And C ₀ .

The dynamic solution method in this embodiment is used to calculate the optimized flow. For example, if the utility function of the 殳 link is an exponential function, the corresponding formula (2) is ' ^{= Mw} — ^exp ( ) _ ' ^] , the test factor

The value refers to formula (3), the link price formula is ^c j , and the link aggregation price obtained by the data source node is = ^^. The calculated result can be seen in Figure 5 and Figure 6. In this embodiment, for convenience of description, the design of the utility function is simplified, and the utility function is adopted.

Directly as a price The calculation formula is a relatively simple implementation form of the utility function, and can also be other power functions and other functional forms. 5 is a flow state change diagram obtained by solving an exponential function in the second embodiment of the network optimization flow control method according to the present invention, and FIG. 6 is a link utilization state obtained by solving the exponential function in the second embodiment of the network optimized flow control method of the present invention. Change chart. As can be seen from FIG. 5, the data transmission rate of the four data source nodes in the network shown in FIG. 4 is 1/3*C on average. , that is, the traffic equivalent to the sub-stream transmission path is 1/3*C. It is consistent with ^x * in the static optimal solution; as can be seen from Figure 6, the utilization of the three-segment link in the network shown in Figure 4 is both Co and the other is

2/3*C. , consistent with the static optimal solution. Moreover, as can be seen from Figures 5 and 6, the flow and utilization are a process of dynamic adaptive approximation from the initial value to the final optimized value, and a small oscillation near the equilibrium point until equilibrium is reached. For another example, suppose the utility function of the link is a power function, and the corresponding formula (2) is

• P =— ^x ' ^{= k[w} - ² - ^χ ' ^] , the value of the probe factor is given by formula (3), the link price formula is ^JC J , and the link aggregation price obtained by the data source node is = 3⁄4^ The calculation results obtained can be seen in Figure 7 and Figure 8. 7 is a flow state change diagram obtained by solving a power function in the second embodiment of the network optimized flow control method according to the present invention, and FIG. 8 is a link utilization state obtained by solving a power function in the second embodiment of the network optimized flow control method according to the present invention; Change chart. The results obtained are consistent with those of Figures 5 and 6, and are also consistent with the static optimal solution, but the process of dynamic approximation is different.

The network optimization flow control method of the embodiment is implemented by decomposing the traffic calculation in the network optimization into each node and the link in the network, thereby solving the problem that the network optimization traffic calculation information demand is large, and the calculation amount is greatly reduced. , improve the real-time performance of network optimization. Embodiment 3 FIG. 9 is a schematic structural diagram of Embodiment 1 of a network optimization flow control apparatus according to the present invention. The network optimization flow control apparatus of this embodiment may be a data source node according to any embodiment of the present invention, and may perform any implementation of the present invention. The network optimized flow control method described in the example. This embodiment briefly describes the structure of the device. For the specific function and execution principle of each module, refer to the method embodiment. As shown in FIG. 9, the apparatus may include an initial loading module 91, a feedback receiving module 92, and a flow conditioning module 93. The initial loading module 91 may obtain a sub-data stream transmission path between the data source node and the server node according to the network topology information, and load the first data stream traffic on the sub-data stream transmission path; The data stream transmission path includes a plurality of the links; the feedback receiving module 92 may receive a link aggregation price of the sub-data stream transmission path fed back by the router node, where the link aggregation price is determined by the router node a plurality of link price aggregations in the sub-streaming transmission path are obtained; the traffic adjustment module 93 can obtain the second data stream traffic according to the link aggregation price and the server weight of the server node, and the second data flow rate The data stream traffic replaces the first data stream traffic to the sub-data stream transmission path. The network optimization flow control device of the embodiment, by setting an initial loading module, a feedback receiving module, and a traffic adjusting module, respectively decomposes the traffic calculation in the network optimization into each node and link in the network, and solves the network optimization traffic. Calculating the problem of large amount of information demand greatly reduces the amount of calculation and improves the real-time performance of network optimization.

Embodiment 4 FIG. 10 is a schematic structural diagram of Embodiment 2 of a network optimization flow control apparatus according to the present invention. This embodiment refines the structure in Embodiment 1 of the apparatus. For the functions and execution principles of each module and unit, refer to the method. As described in the examples. As shown in Figure 10, in the device, The initial loading module 91 can include a first loading unit 911 and/or a second loading unit 912. The first loading unit 911 can load the first data stream traffic to the MPLS pipe, and the sub data stream transmission path is an MPLS pipe. The second loading unit 912 can correspond to the first data stream traffic. A path marker is added to the data packet to cause the router node to load the first data stream traffic onto the sub-data stream transmission path according to the path label. Wherein, in the above-mentioned "first loading unit 911 and/or second loading unit 912", "and/or" means that the network optimization flow control device can have both the first loading unit 911 and the second loading unit. 912, that is, having two corresponding functions at the same time; or, there may be only one of the first loading unit 911 and the second loading unit 912, that is, having only one of the functions. Further, the device of the embodiment further The weight calculation module 94 may include a server performance information sent by the server node, and obtain a server weight of the server node according to the server performance information. The network optimization flow control device of the embodiment, By setting the initial loading module, the feedback receiving module and the traffic conditioning module, the traffic calculation in the network optimization is decomposed into the nodes and links in the network respectively, which solves the problem that the network optimization traffic calculation information demand is large, and is greatly reduced. The amount of calculation increases the real-time performance of network optimization.

Embodiment 5 FIG. 11 is a schematic structural diagram of an embodiment of a network optimization flow control system according to the present invention. The system in this embodiment may perform a network optimization flow control method according to any embodiment of the present invention, and specifically, functions and execution principles of each module and unit. See the method embodiments for description. As shown in FIG. 11, the system of this embodiment may include a data source node 1101, a server node 1102, and a routing node 1103. The two adjacent router nodes are a link between the two adjacent router nodes. The data source node is equivalent to the network optimized flow control device described in the embodiment of the present invention. The data source node 1101 is configured to obtain, according to the network topology information, a sub-data stream transmission path between the data source node and the server node, and load the first on the sub-data stream transmission path. a data stream traffic; one of the sub-data stream transmission paths includes a plurality of the links; the router node 1103, configured to aggregate traffic of the first data stream on the link into link traffic, where the link control The module obtains a link price of the link according to the link traffic; aggregates link price of several links in the sub-stream transmission path to obtain a link aggregation price, and aggregates the link price feedback The data source node 1101 is further configured to obtain a second data flow according to a link aggregation price fed back by the router node and a server weight of the server node, and obtain the second data flow The second data stream traffic replaces the first data stream traffic to the sub-data stream transmission path. The network optimization flow control system of the embodiment is respectively implemented by decomposing the traffic calculation in the network optimization into each node and the link in the network, thereby solving the problem that the network optimization traffic calculation information demand is large, and the calculation amount is greatly reduced. , improve the real-time performance of network optimization. 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 method includes the steps of the foregoing method embodiments; 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. 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. The modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

Rights request

A network optimization flow control method, characterized in that the network comprises a data source node, a router node and a server node, wherein two adjacent router nodes are a link, and the two adjacent ones The router node constitutes a link control module; the method includes: the data source node obtaining, according to network topology information, a sub-data stream transmission path between the data source node and the server node, and in the sub-data Loading the first data stream traffic on the streaming path; one of the sub-data stream transmission paths includes a plurality of the links;

The router node aggregates traffic of the first data stream on the link into link traffic, and the link control module obtains a link price of the link according to the link traffic; Link price aggregation of several links in the sub-data stream transmission path to obtain a link aggregation price, and feedback the link aggregation price to the data source node;

The data source node obtains the second data stream traffic according to the link aggregation price and the server weight of the server node, and loads the second data stream traffic to replace the first data stream traffic to the Substream streaming path.

The network optimization flow control method according to claim 1, wherein the loading of the first data stream on the sub-stream transmission path comprises: the sub-data stream transmission path being an MPLS pipe, Loading the first data stream traffic onto the MPLS pipe; or adding a path marker to the data packet corresponding to the first data stream traffic, so that the router node marks the first according to the path identifier Data stream traffic is loaded onto the sub-stream stream transmission path.

The network optimization flow control method according to claim 1, further comprising: the data source node receiving server performance information sent by the server node, and obtaining the server node according to the server performance information Server weights.

The network optimization flow control method according to claim 1, wherein the link control module obtains a link price of the link according to the link traffic, and includes:

The link control module obtains a utility function of the link according to the link traffic, and obtains a link price of the link according to the utility function;

The utility function is a non-linear function.

The network optimization flow control method according to claim 1, wherein the utility function includes a heuristic factor, and the heuristic factor is a normal number when the link aggregation price is less than or equal to 1, in the A link aggregation price greater than 1 is a negative constant.

A network optimization flow control device, comprising: an initial loading module, configured to obtain a sub-data stream transmission path between a data source node and a server node according to network topology information, and in the sub-data Loading the first data stream traffic on the streaming path; one of the sub-data stream transmission paths includes a plurality of the links;

a feedback receiving module, configured to receive a link aggregation price of the sub-data stream transmission path fed back by the router node, where the link aggregation price is a chain of the sub-data stream transmission path by the router node The traffic price adjustment module is configured to: obtain, according to the link aggregation price, the server weight of the server node, the second data flow, and replace the second data flow with the first data Stream traffic is loaded to the sub-data stream transmission path.

The network optimization flow control device according to claim 6, wherein the initial loading module comprises:

a first loading unit, configured to load the first data stream to the MPLS pipe, where the sub-data stream transmission path is an MPLS pipe; or

a second loading unit, configured to add a path identifier to the data packet corresponding to the first data flow So that the router node loads the first data stream traffic onto the sub-data stream transmission path according to the path label.

The network optimization flow control device according to claim 6, further comprising: a weight calculation module, configured to receive server performance information sent by the server node, and obtain the server according to the server performance information The server weight of the node.

A network optimization flow control system, comprising: a data source node, a server node, and a routing node, wherein two adjacent router nodes are a link, and the two adjacent router nodes Forming a link control module; the data source node is configured to obtain, according to network topology information, a sub-data stream transmission path between the data source node and the server node, and on the sub-data stream transmission path Loading the first data stream traffic; one of the sub-data stream transmission paths includes a plurality of the links; and further configured to obtain the second data according to the link aggregation price fed back by the router node and the server weight of the server node Flowing traffic, and replacing the first data stream traffic with the first data stream traffic to the sub-data stream transmission path; the router node, configured to aggregate traffic of the first data stream on the link a link traffic, the link control module obtaining a link price of the link according to the link traffic; transmitting the sub data Some price link link path polymerization link aggregation price, and the price of the polymerization feedback link to the data source node.