Flow management method and device based on SDN (Software Defined Network) and electronic equipment

Abstract

The invention provides an SDN (Software Defined Network)-based traffic management method and device and electronic equipment. The method comprises the following steps: acquiring cache information, exit bandwidth and traffic data of each network equipment of a target network through an SDN controller; processing the cache information and the traffic data to obtain cache features and traffic features; identifying the congestion condition of the target network based on the cache feature and the flow feature to obtain a congestion identification result of the target network; and determining a bandwidth quota adjustment scheme of the target network based on the congestion identification result and the exit bandwidth. According to the invention, the congestion condition of the target network is identified based on the multi-dimensional feature information, the network congestion can be accurately identified, and the bandwidth quota adjustment scheme can be determined based on the congestion identification result and the exit bandwidth, so that the efficiency, adaptability and reliability of flow management can be improved.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a traffic management method, device and electronic device based on SDN. Background Art

[0002] With the rapid development of Internet technology, network traffic has exploded. Especially with the popularization of cloud computing, Internet of Things, high-definition video streaming and real-time communication, users have higher and higher requirements for communication quality. How to efficiently manage network traffic, reasonably allocate network resources and avoid network congestion are key issues that need to be solved. At present, traditional networks rely on fixed-configuration routers and switches, and forward traffic based on static routing protocols or predefined policies. They are inefficient, poorly adaptable and unreliable, and affect user experience. Summary of the invention

[0003] The present invention provides a traffic management method, device and electronic device based on SDN, which are used to solve the defects of low efficiency, poor adaptability and reliability in the prior art that traffic is forwarded based on static routing protocols or predefined strategies and relies on fixed-configuration routers and switches.

[0004] The present invention provides a traffic management method based on SDN, comprising: Obtain cache information, egress bandwidth, and traffic data of each network device in the target network through the SDN controller; Processing the cache information and the traffic data to obtain cache characteristics of the cache information and traffic characteristics of the traffic data; Based on the cache characteristics and traffic characteristics, identifying the congestion situation of the target network, and obtaining a congestion identification result of the target network; Based on the congestion identification result and the egress bandwidth, a bandwidth quota adjustment scheme for the target network is determined.

[0005] In some embodiments, based on the cache characteristics and traffic characteristics, identifying the congestion situation of the target network to obtain the congestion identification result of the target network includes: Inputting the cache characteristics and traffic characteristics into a pre-built identification model to obtain a congestion identification result of the target network output by the identification model; The identification model is obtained by training based on sample cache features and sample traffic features of each sample network device of the sample network and congestion identification result labels of the sample network.

[0006] In some embodiments, the recognition model includes a feature fusion layer and a recognition layer; Correspondingly, inputting the cache feature and the traffic feature into a pre-constructed recognition model to obtain the congestion recognition result of the target network output by the recognition model includes: Inputting the cache feature and the traffic feature into the feature fusion layer to obtain a fused feature vector output by the feature fusion layer; Inputting the fused feature vector into the recognition layer to obtain the congestion recognition result of the target network output by the recognition layer.

[0007] In some embodiments, recognizing the congestion situation of the target network based on the cache feature and the traffic feature to obtain the congestion recognition result of the target network includes: Grouping the traffic data packets of the target network based on the traffic feature and calculating the average traffic value of each group of traffic data packets; Determining the congestion level of each group of traffic data packets based on the cache feature of the network device corresponding to each group of traffic data packets and the average traffic value of each group of traffic data packets.

[0008] In some embodiments, the congestion recognition result includes the congestion level of each group of traffic data packets of the target network. Determining the bandwidth quota adjustment scheme of the target network based on the congestion recognition result and the egress bandwidth includes: Determining the egress bandwidth of the network device corresponding to each group of traffic data packets; Calculating the bandwidth quota of each group of traffic data packets based on the congestion level of each group of traffic data packets and the egress bandwidth of the network device corresponding to each group of traffic data packets; Determining the bandwidth quota adjustment scheme of the target network according to the bandwidth quota of each group of traffic data packets.

[0009] In some embodiments, the method further includes: Comparing the bandwidth quota of each group of traffic data packets with a corresponding preset egress bandwidth threshold, and in the case where the bandwidth quota of each group of traffic data packets is greater than the preset egress bandwidth threshold, allocating each group of traffic data packets to a network device with low load.

[0010] In some embodiments, the training process of the recognition model includes: Obtaining the sample cache information and sample traffic data of each sample network device of the sample network through an SDN controller; Determining the congestion recognition result label of the sample network; Processing the sample cache information and sample traffic data to obtain sample cache features and sample traffic features; Using the sample cache feature and the sample traffic feature as training samples, and using the congestion identification result label of the sample network as the sample label, train the initial identification model, and after training is completed, obtain the identification model.

[0011] In some embodiments, the training of the initial identification model includes: Input the sample cache feature and the sample traffic feature into the initial identification model to obtain the predicted congestion identification result of the sample network output by the initial identification model; Based on the predicted congestion identification result and the congestion identification result label, calculate the loss function value; According to the loss function value, iteratively optimize the parameters of the initial identification model to obtain the identification model.

[0012] The present invention also provides a traffic management device based on SDN, including: An acquisition unit, configured to obtain the cache information, egress bandwidth, and traffic data of each network device of the target network through an SDN controller; A feature extraction unit, configured to process the cache information and traffic data to obtain the cache feature of the cache information and the traffic feature of the traffic data; An identification unit, configured to identify the congestion situation of the target network based on the cache feature and the traffic feature to obtain the congestion identification result of the target network; A determination unit, configured to determine a bandwidth quota adjustment plan for the target network based on the congestion identification result and the egress bandwidth.

[0013] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the SDN-based traffic management method as described in any one of the above.

[0014] The SDN-based traffic management method, device, and electronic device provided by the present invention obtain the cache information, egress bandwidth, and traffic data of each network device of the target network through an SDN controller; process the cache information and traffic data to obtain the cache feature and the traffic feature; identify the congestion situation of the target network based on the cache feature and the traffic feature, and can accurately identify network congestion. Furthermore, based on the congestion identification result and the egress bandwidth, a bandwidth quota adjustment plan can be determined, thereby improving the efficiency, adaptability, and reliability of traffic management. Description of the Drawings

[0015] To more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0016] Figure 1 It is a schematic flowchart of the SDN-based traffic management method provided by an embodiment of the present invention.

[0017] Figure 2 It is a schematic flowchart of the training process of the recognition model provided by an embodiment of the present invention.

[0018] Figure 3 It is a schematic structural diagram of the SDN-based traffic management device provided by an embodiment of the present invention.

[0019] Figure 4 It is a schematic structural diagram of the electronic device provided by an embodiment of the present invention. Detailed implementation manners

[0020] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention with reference to the drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments in the present invention belong to the scope of protection of the present invention.

[0021] Figure 1 It is a schematic flowchart of the SDN-based traffic management method provided by an embodiment of the present invention. As Figure 1 shown, a SDN-based traffic management method is provided, including the following steps: Step 110, Step 120, Step 130, and Step 140. The process steps of this method are only a possible implementation manner of the present invention.

[0022] Step 110: Obtain the cache information, egress bandwidth, and traffic data of each network device in the target network through the SDN controller.

[0023] Software Defined Networking (SDN) is a network architecture. Its core idea is to achieve centralized and intelligent management of the network through the decoupling of the control layer and the data forwarding layer. As the core component, the SDN controller can obtain the cache information, egress bandwidth, and traffic data of network devices in real time through standardized interfaces.

[0024] Among them, the cache information includes but is not limited to: the queue status of switches / routers, buffer utilization, etc.; the egress bandwidth includes the real-time throughput and bandwidth capacity of each port; the traffic data at least includes: traffic table statistical information (such as the number of data packets, the number of bytes), traffic paths, and protocol distributions.

[0025] Step 120: Process the cache information and traffic data to obtain cache features and traffic features.

[0026] Among them, the cache features at least include: cache queue length, cache overflow times, cache hit rate, and other features; the traffic features at least include: traffic rate, latency, jitter, packet loss rate, traffic burstiness, and other features.

[0027] Optionally, perform preprocessing such as data cleaning, denoising, normalization, and alignment on the cache information and traffic data to obtain preprocessed cache information and traffic data; extract features from the preprocessed cache information and traffic data to obtain cache features and traffic features.

[0028] Step 130: Identify the congestion situation of the target network based on the cache features and traffic features to obtain the congestion identification result of the target network.

[0029] Optionally, determine the respective weights of the cache features and traffic features, fuse the cache features and traffic features to obtain fused features, and identify the congestion situation of the target network based on the fused features.

[0030] Optionally, the congestion identification result includes congestion time, congestion location, and congestion level, such as mild congestion, moderate congestion, and severe congestion.

[0031] Optionally, adaptively adjust the congestion level determination criteria according to the target network topology and service types (such as real-time video and file transfer).

[0032] It can be understood that by identifying the congestion situation of the target network based on the cache features and traffic features to obtain the congestion identification result of the target network, multiple dimensions of features are comprehensively considered, which can avoid misjudgment by a single indicator and improve the accuracy of identification.

[0033] Step 140: Determine the bandwidth quota adjustment plan for the target network based on the congestion identification result and egress bandwidth.

[0034] Optionally, classify the services of the target network to determine critical services (such as video conferencing, core database access), high-bandwidth demand services (such as video live streaming, large file transfer).

[0035] Optionally, allocate fixed guaranteed bandwidth (such as a minimum quota of 30%) for critical services to ensure that they are not affected by congestion; for high-bandwidth demand services, allocate bandwidth dynamically as needed.

[0036] Optionally, determine the priorities of different services and limit the bandwidth of low-priority services.

[0037] In the embodiments of the present invention, the SDN controller is used to obtain the cache information, egress bandwidth, and traffic data of each network device in the target network; the cache information and traffic data are processed to obtain cache features and traffic features; based on the cache features and traffic features, the congestion situation of the target network is identified, and network congestion can be accurately identified. Furthermore, based on the congestion identification result and the egress bandwidth, a bandwidth quota adjustment scheme can be determined, thereby improving the efficiency, adaptability, and reliability of traffic management.

[0038] In some embodiments, based on the cache features and traffic features, identifying the congestion situation of the target network to obtain the congestion identification result of the target network includes: Inputting the cache features and traffic features into a pre-constructed identification model to obtain the congestion identification result of the target network output by the identification model; wherein, the identification model is trained based on the sample cache features and sample traffic features of each sample network device in the sample network, and the congestion identification result label of the sample network.

[0039] It can be understood that by inputting the cache features and traffic features into a pre-constructed identification model to obtain the congestion identification result of the target network output by the identification model, the efficiency, accuracy, and interpretability of network congestion identification can be improved, and it can dynamically adapt to complex scenarios.

[0040] In some embodiments, the identification model includes a feature fusion layer and an identification layer; Correspondingly, inputting the cache features and traffic features into a pre-constructed identification model to obtain the congestion identification result of the target network output by the identification model includes: Inputting the cache features and traffic features into the feature fusion layer to obtain a fused feature vector output by the feature fusion layer; Inputting the fused feature vector into the identification layer to obtain the congestion identification result of the target network output by the identification layer.

[0041] Optionally, determine the weights of the cache features and the weights of the traffic features based on the feature fusion layer, and fuse the cache features and the traffic features according to the weights of the cache features and the weights of the traffic features to obtain a fused feature vector.

[0042] Optionally, based on the identification layer, compare the fused feature vector with a preset congestion determination index to determine the congestion identification result of the target network.

[0043] Optionally, determine the preset congestion determination index based on user input or according to the characteristics of the target network.

[0044] In some embodiments, based on cache characteristics and traffic characteristics, the congestion situation of the target network is identified to obtain the congestion identification result of the target network, including: Based on the traffic characteristics, the traffic data packets of the target network are grouped, and the average traffic value of each group of traffic data packets is calculated; Based on the cache characteristics of the network device corresponding to each group of traffic data packets and the average traffic value of each group of traffic data packets, the congestion level of each group of traffic data packets is determined.

[0045] Optionally, the traffic characteristics are identified by deep packet inspection technology, and the traffic data packets of the target network are grouped.

[0046] Optionally, based on traffic characteristics (such as source IP, destination IP, source port, destination port, protocol type), the data packets are divided into logical groups, and the grouping examples are as follows: Video streaming media (destination port: 1935, 80, 443, protocol: RTMP / HTTP); Real-time communication (destination port: 5060, 5004, protocol: SIP / RTP); File transfer (destination port: 21, 22, protocol: FTP / SFTP).

[0047] Optionally, the average traffic value of each group of traffic (such as rate average) is associated with the cache characteristics of the corresponding network device (such as queue occupancy rate) to construct a feature matrix; based on the feature matrix, the congestion level of each group of traffic data packets is determined.

[0048] It can be understood that by grouping the traffic data packets of the target network based on traffic characteristics, it helps to distinguish the contribution of different service traffic to congestion and accurately locate the problem source.

[0049] In some embodiments, the congestion identification result includes the congestion level of each group of traffic data packets of the target network. Based on the congestion identification result and the egress bandwidth, a bandwidth quota adjustment plan for the target network is determined, including: Determine the egress bandwidth of the network device corresponding to each group of traffic data packets; Based on the congestion level of each group of traffic data packets and the egress bandwidth of the network device corresponding to each group of traffic data packets, calculate the bandwidth quota of each group of traffic data packets; According to the bandwidth quota of each group of traffic data packets, determine the bandwidth quota adjustment plan for the target network.

[0050] Optionally, traffic groups (such as video streams, access requests) are bound to the network devices passed through (such as core switches, edge routers) to construct a "traffic group - device" association table.

[0051] Optionally, determine the weights of different congestion levels and the corresponding bandwidth adjustment strategies.

[0052] For example, in the case of mild congestion, the adjustment strategy is on-demand allocation, with a small amount of speed limiting for low-priority traffic; in the case of moderate congestion, the adjustment strategy is to dynamically compress high-burst traffic to ensure critical services; in the case of severe congestion, the adjustment strategy is to force capacity expansion or strictly limit the flow.

[0053] In some embodiments, the above method further includes: Compare the bandwidth quota of each group of traffic data packets with the corresponding preset egress bandwidth threshold. When the bandwidth quota of each group of traffic data packets is greater than the preset egress bandwidth threshold, allocate each group of traffic data packets to a network device with low load.

[0054] It can be understood that by comparing the bandwidth quota of each group of traffic data packets with the corresponding preset egress bandwidth threshold and allocating each group of traffic data packets to a network device with low load when the bandwidth quota of each group of traffic data packets is greater than the preset egress bandwidth threshold, it is possible to respond to network state changes in real time, support burst traffic and business growth, and achieve efficient utilization of resources.

[0055] Figure 2 It is a schematic flow diagram of the training process of the recognition model provided by the embodiments of the present invention. As Figure 2 shown, in some embodiments, the training process of the recognition model includes: Step 210: Obtain the sample cache information and sample traffic data of each sample network device in the sample network through the SDN controller; Step 220: Determine the congestion recognition result label of the sample network; Step 230: Process the sample cache information and sample traffic data to obtain sample cache features and sample traffic features; Step 240: Use the sample cache features and sample traffic features as training samples, and the congestion recognition result label of the sample network as the sample label to train the initial recognition model. After the training is completed, the recognition model is obtained.

[0056] In some embodiments, training the initial recognition model includes: Input the sample cache features and sample traffic features into the initial recognition model to obtain the predicted congestion recognition result of the sample network output by the initial recognition model; Based on the predicted congestion recognition result and the congestion recognition result label, calculate the loss function value; According to the loss function value, iteratively optimize the parameters of the initial recognition model to obtain the recognition model.

[0057] Optionally, the initial recognition model includes an initial feature fusion layer and an initial recognition layer; Correspondingly, input the sample cache feature and the sample traffic feature into the pre-constructed initial recognition model, and obtain the congestion recognition result of the sample network output by the initial recognition model, including: Input the sample cache feature and the sample traffic feature into the initial feature fusion layer to obtain the sample fusion feature vector output by the initial feature fusion layer; Input the sample fusion feature vector into the initial recognition layer to obtain the predicted congestion recognition result of the sample network output by the initial recognition layer.

[0058] The traffic management device based on SDN provided by the embodiments of the present invention will be described below. The traffic management device based on SDN described below can be correspondingly referred to the traffic management method based on SDN described above.

[0059] Figure 3 It is a schematic structural diagram of the traffic management device based on SDN provided by the embodiments of the present invention, as Figure 3 shown. The traffic management device 300 based on SDN includes: An obtaining unit 310, configured to obtain the cache information, the egress bandwidth, and the traffic data of each network device of the target network through an SDN controller; A feature extraction unit 320, configured to process the cache information and the traffic data to obtain a cache feature and a traffic feature; An identification unit 330, configured to identify the congestion situation of the target network based on the cache feature and the traffic feature, and obtain the congestion recognition result of the target network; A determination unit 340, configured to determine a bandwidth quota adjustment plan for the target network based on the congestion recognition result and the egress bandwidth.

[0060] Optionally, identifying the congestion situation of the target network based on the cache feature and the traffic feature to obtain the congestion recognition result of the target network includes: Input the cache feature and the traffic feature into the pre-constructed recognition model, and obtain the congestion recognition result of the target network output by the recognition model; Wherein, the recognition model is trained based on the sample cache features and sample traffic features of each sample network device of the sample network, and the congestion recognition result label of the sample network.

[0061] Optionally, the recognition model includes a feature fusion layer and an identification layer; Correspondingly, input the cache feature and the traffic feature into the pre-constructed recognition model, and obtain the congestion recognition result of the target network output by the recognition model, including: Input the cache feature and traffic feature into the feature fusion layer to obtain the fused feature vector output by the feature fusion layer; Input the fused feature vector into the recognition layer to obtain the congestion recognition result of the target network output by the recognition layer.

[0062] Optionally, based on the cache feature and traffic feature, identify the congestion situation of the target network to obtain the congestion recognition result of the target network, including: Based on the traffic feature, group the traffic data packets of the target network and calculate the average traffic value of each group of traffic data packets; Based on the cache feature of the network device corresponding to each group of traffic data packets and the average traffic value of each group of traffic data packets, determine the congestion level of each group of traffic data packets.

[0063] Optionally, the congestion recognition result includes the congestion level of each group of traffic data packets of the target network. Based on the congestion recognition result and the egress bandwidth, determine the bandwidth quota adjustment scheme of the target network, including: Determine the egress bandwidth of the network device corresponding to each group of traffic data packets; Based on the congestion level of each group of traffic data packets and the egress bandwidth of the network device corresponding to each group of traffic data packets, calculate the bandwidth quota of each group of traffic data packets; According to the bandwidth quota of each group of traffic data packets, determine the bandwidth quota adjustment scheme of the target network.

[0064] Optionally, the traffic management device based on SDN further includes: A load balancing unit, configured to compare the bandwidth quota of each group of traffic data packets with the corresponding preset egress bandwidth threshold, and in the case where the bandwidth quota of each group of traffic data packets is greater than the preset egress bandwidth threshold, allocate each group of traffic data packets to a network device with low load.

[0065] Optionally, the training process of the recognition model includes: Obtain the sample cache information and sample traffic data of each sample network device of the sample network through the SDN controller; Determine the congestion recognition result label of the sample network; Process the sample cache information and sample traffic data to obtain the sample cache feature and sample traffic feature; Use the sample cache feature and sample traffic feature as training samples, and use the congestion recognition result label of the sample network as the sample label to train the initial recognition model. After the training is completed, obtain the recognition model.

[0066] Optionally, training the initial recognition model includes: Input the sample cache feature and sample traffic feature into the initial recognition model to obtain the predicted congestion recognition result of the sample network output by the initial recognition model; Calculate the loss function value based on the predicted congestion recognition result and the congestion recognition result label; Iteratively optimize the parameters of the initial recognition model according to the loss function value to obtain the recognition model.

[0067] It should be noted here that the SDN-based traffic management device provided in the embodiments of the present invention can implement all the method steps implemented in the above-mentioned embodiments of the SDN-based traffic management method, and can achieve the same technical effects. The same parts and beneficial effects as those in the method embodiments will not be specifically described in this embodiment.

[0068] Figure 4 The structural schematic diagram of the electronic device provided in the embodiments of the present invention is shown in Figure 4 As shown, the electronic device may include: a processor 410, a communication interface 420, a memory 430, and a communication bus 440. Among them, the processor 410, the communication interface 420, and the memory 430 complete communication with each other through the communication bus 440. The processor 410 can call the logical instructions in the memory 430 to execute the SDN-based traffic management method, and the method includes: obtaining the cache information, egress bandwidth, and traffic data of each network device in the target network through the SDN controller; processing the cache information and traffic data to obtain cache features and traffic features; based on the cache features and traffic features, identifying the congestion situation of the target network to obtain the congestion recognition result of the target network; and determining the bandwidth quota adjustment plan of the target network based on the congestion recognition result and the egress bandwidth.

[0069] In addition, when the logical instructions in the above-mentioned memory 430 are implemented in the form of software function units and sold or used as an independent product, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0070] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the SDN-based traffic management method provided by each of the above methods. The method includes: obtaining cache information, egress bandwidth, and traffic data of each network device in a target network through an SDN controller; processing the cache information and traffic data to obtain cache characteristics and traffic characteristics; identifying the congestion situation of the target network based on the cache characteristics and traffic characteristics to obtain a congestion identification result of the target network; and determining a bandwidth quota adjustment scheme for the target network based on the congestion identification result and the egress bandwidth.

[0071] In another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the SDN-based traffic management method provided by each of the above methods. The method includes: obtaining cache information, egress bandwidth, and traffic data of each network device in a target network through an SDN controller; processing the cache information and traffic data to obtain cache characteristics and traffic characteristics; identifying the congestion situation of the target network based on the cache characteristics and traffic characteristics to obtain a congestion identification result of the target network; and determining a bandwidth quota adjustment scheme for the target network based on the congestion identification result and the egress bandwidth.

[0072] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative efforts.

[0073] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disc, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A traffic management method based on SDN, characterized in that, Including: Obtaining the cache information, egress bandwidth, and traffic data of each network device in the target network through an SDN controller; Processing the cache information and traffic data to obtain the cache characteristics of the cache information and the traffic characteristics of the traffic data; Identifying the congestion situation of the target network based on the cache characteristics and traffic characteristics to obtain the congestion identification result of the target network; Determining a bandwidth quota adjustment plan for the target network based on the congestion identification result and egress bandwidth.

2. The SDN-based traffic management method according to claim 1, wherein Identifying the congestion situation of the target network based on the cache characteristics and traffic characteristics to obtain the congestion identification result of the target network, including: Inputting the cache characteristics and traffic characteristics into a pre-constructed identification model to obtain the congestion identification result of the target network output by the identification model; Wherein, the identification model is trained based on the sample cache characteristics and sample traffic characteristics of each sample network device in the sample network, and the congestion identification result label of the sample network.

3. The SDN-based traffic management method according to claim 2, characterized in that The identification model includes a feature fusion layer and an identification layer; Correspondingly, the step of inputting the cache characteristics and traffic characteristics into a pre-constructed identification model to obtain the congestion identification result of the target network output by the identification model includes: Inputting the cache characteristics and traffic characteristics into the feature fusion layer to obtain a fused feature vector output by the feature fusion layer; Inputting the fused feature vector into the identification layer to obtain the congestion identification result of the target network output by the identification layer.

4. The SDN-based traffic management method according to claim 1, characterized in that Identifying the congestion situation of the target network based on the cache characteristics and traffic characteristics to obtain the congestion identification result of the target network, including: Grouping the traffic data packets in the target network based on the traffic characteristics and calculating the average traffic value of each group of traffic data packets; Determining the congestion level of each group of traffic data packets based on the cache characteristics of the network device corresponding to each group of traffic data packets and the average traffic value of each group of traffic data packets.

5. The SDN-based traffic management method according to claim 1, characterized in that The congestion identification result includes the congestion level of each group of traffic data packets in the target network. The step of determining a bandwidth quota adjustment plan for the target network based on the congestion identification result and egress bandwidth includes: Determining the egress bandwidth of the network device corresponding to each group of traffic data packets; Calculating the bandwidth quota of each group of traffic data packets based on the congestion level of each group of traffic data packets and the egress bandwidth of the network device corresponding to each group of traffic data packets; Determining a bandwidth quota adjustment plan for the target network according to the bandwidth quota of each group of traffic data packets.

6. The SDN-based traffic management method according to claim 5, wherein The method further includes: Comparing the bandwidth quota of each group of traffic data packets with a corresponding preset egress bandwidth threshold, and in the case where the bandwidth quota of each group of traffic data packets is greater than the preset egress bandwidth threshold, allocating each group of traffic data packets to a network device with low load.

7. The SDN-based traffic management method according to claim 2, characterized in that The training process of the identification model includes: Obtaining the sample cache information and sample traffic data of each sample network device in the sample network through an SDN controller; Determining the congestion identification result label of the sample network; Process the sample cache information and sample traffic data to obtain the sample cache features of the sample cache information and the sample traffic features of the sample traffic data; Use the sample cache features and sample traffic features as training samples, and use the congestion identification result label of the sample network as the sample label to train the initial identification model. After the training is completed, obtain the identification model.

8. The SDN-based traffic management method according to claim 7, wherein The training of the initial identification model includes: Input the sample cache features and sample traffic features into the initial identification model to obtain the predicted congestion identification result of the sample network output by the initial identification model; Calculate the loss function value based on the predicted congestion identification result and the congestion identification result label; According to the loss function value, iteratively optimize the parameters of the initial identification model to obtain the identification model.

9. A traffic management device based on SDN, characterized in that It includes: An acquisition unit for acquiring the cache information, egress bandwidth, and traffic data of each network device of the target network through the SDN controller; A feature extraction unit for processing the cache information and traffic data to obtain the cache features of the cache information and the traffic features of the traffic data; An identification unit for identifying the congestion situation of the target network based on the cache features and traffic features to obtain the congestion identification result of the target network; A determination unit for determining the bandwidth quota adjustment plan of the target network based on the congestion identification result and the egress bandwidth.

10. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the SDN-based traffic management method according to any one of claims 1 to 8.

Images