CN115037689A - Method and system for intelligently scheduling network traffic - Google Patents
Method and system for intelligently scheduling network traffic Download PDFInfo
- Publication number
- CN115037689A CN115037689A CN202210631312.7A CN202210631312A CN115037689A CN 115037689 A CN115037689 A CN 115037689A CN 202210631312 A CN202210631312 A CN 202210631312A CN 115037689 A CN115037689 A CN 115037689A
- Authority
- CN
- China
- Prior art keywords
- network
- scheduling
- training
- flow
- training body
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000012549 training Methods 0.000 claims abstract description 63
- 238000013135 deep learning Methods 0.000 claims abstract description 14
- 230000002452 interceptive effect Effects 0.000 claims abstract description 14
- 238000001514 detection method Methods 0.000 claims description 20
- 238000005259 measurement Methods 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 7
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 230000002159 abnormal effect Effects 0.000 claims description 4
- 239000000523 sample Substances 0.000 claims description 4
- 230000005540 biological transmission Effects 0.000 abstract description 11
- 230000004069 differentiation Effects 0.000 description 6
- 230000002787 reinforcement Effects 0.000 description 6
- 238000004088 simulation Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 239000010410 layer Substances 0.000 description 5
- 239000012792 core layer Substances 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000003631 expected effect Effects 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 2
- 230000002411 adverse Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 206010033799 Paralysis Diseases 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
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
- H04L47/12—Avoiding congestion; Recovering from congestion
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/50—Reducing energy consumption in communication networks in wire-line communication networks, e.g. low power modes or reduced link rate
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention provides a method and a system for intelligently scheduling network flow, which belong to the technical field of computer networks and comprise the following steps: selecting networks in different flow states to form a network set serving as a network training body; based on a deep learning method, utilizing the network training body to carry out interactive training to obtain a scheduling training body; measuring network flow information of the SDN network equipment, and acquiring a network state of the SDN network equipment according to a result of measuring the network flow information of the SDN network equipment; scheduling the acquired network through a scheduling training body; if not, storing the acquired network state in a network set serving as a network training body, forming a new network set, carrying out interactive training again to obtain an updated scheduling training body, and then scheduling the acquired network through the updated scheduling training body; the invention realizes real-time, dynamic and self-adaptive flow scheduling, and effectively improves the utilization rate of network resources and the data transmission quality.
Description
Technical Field
The invention relates to the technical field of computer networks, in particular to a method and a system for intelligently scheduling network traffic.
Background
In recent years, with the rapid development of 5G, cloud computing and the Internet of things, various network applications emerge endlessly, the network scale is continuously enlarged, and the network traffic is increased explosively. How to avoid Network congestion, improve Network resource utilization rate and guarantee user experience quality through reasonable scheduling of Network traffic becomes more and more a problem Software Defined Network (SDN) that needs key research in the Network field as a new Network design idea, and brings a new architecture for a traditional Network. The SDN not only realizes transfer control separation and centralized control, but also opens an interface, so that a third-party application can define a new network function only in a programming mode. SDN, as a Network technology with the greatest future development prospect, is a new paradigm for optimizing Network resource management, and therefore is also increasingly applied to Data Center Networks (DCNs) with higher Network performance. On the other hand, artificial intelligence technology is changing day by day, and deep learning and reinforcement learning are important technical supports thereof, and have shown strong vitality in recent years. Deep learning can explore the internal rules and high-level attributes among samples through learning training, so that the deep learning has good performance in the fields of processing character pictures, network traffic prediction and the like. The reinforcement learning has good and wide application prospect in the optimization work of the network routing strategy due to the reward mechanism and the iterative learning capability, and provides possibility for the realization of a dynamic and intelligent traffic scheduling method capable of adapting to network changes in real time along with the proposal of an SDN network architecture, the enhancement of network data collection and analysis capability and the breakthrough progress of deep learning and deep reinforcement learning in the fields of adaptive learning, automatic control and the like.
The rapid development of network technology and the wide deployment of network infrastructure promote the development of social digital transformation, and the calculation, transmission and storage of mass data also promote the construction and intelligent upgrading transformation of large-scale data centers. Network traffic is the amount of data transmitted over the network. Traffic scheduling actually achieves the purpose of adjusting network traffic by adjusting network parameters such as BGP routing policies and IGP Metric values. The flow scheduling is divided into analog scheduling and actual issuing scheduling. Because when the flow exception occurs, it may be that an enterprise is producing, at this time, it is impossible to modify the network parameters at will, and because if the parameters are modified incorrectly, other services may be affected to cause the whole network to be paralyzed, it is necessary to perform a simulation scheduling on the existing network before the network device parameters are issued. The aim of the simulation scheduling is to ensure that the flow components on the relevant links achieve the expected effect after the parameters of the network equipment are modified, and the normal operation of the actual network is not influenced. The simulation scheduling brings various network parameters of the analyzed and processed routing path into a network model of the system, simulates the routing condition and various service flow trends of the network after adjustment, and shows the flow direction of the network flow after adjustment, so as to determine whether the expected effect can be achieved. The simulated scheduling does not actually change the network flow trend, the actual issuing scheduling is to perform network parameter configuration issuing on the real network equipment after the network parameters of the simulated scheduling can meet the expected effect of network adjustment, and the configuration issuing can be issued to the equipment in a one-click manner in a ssh/telnet mode through related commands of BGP routing strategies. After the transmission, the network parameters of the network equipment are changed, and the flow direction is changed accordingly. Thus, the purpose of traffic scheduling is achieved.
Traffic scheduling, a technique that effectively utilizes network resources, can optimize network performance and help networks adapt to business changes quickly. In the network scheduling process, network flow measurement is an important basic part, and the purpose of network flow measurement is to improve the service quality, improve the resource utilization rate, start to diagnose or solve problems before a user reports the problems, and improve the reliability and the availability of the network. The running state of the network is judged by monitoring the state of the network. Traffic information measurement is therefore indispensable in traffic scheduling.
The traditional data center architecture and the traffic scheduling mode can not meet the requirements of low-delay connection and high-quality transmission, the traffic scheduling method lacks dynamic adaptability, and in addition, the problems of traffic and resource scheduling such as unbalanced network traffic load and low link resource utilization rate of the data center exist.
Disclosure of Invention
In view of this, the present invention provides a method and a system for intelligently scheduling network traffic, which implement real-time, dynamic, adaptive traffic scheduling, effectively improve the utilization rate of network resources and the quality of data transmission, optimize the network performance, improve the convergence rate of the traffic scheduling method, enable the traffic scheduling method to have a certain generalization capability, and have good convergence and robustness under different network architectures, thereby achieving the purpose of fully utilizing resources in the network.
In order to solve the above technical problem, the present invention provides a method for intelligently scheduling network traffic, which comprises the following steps:
selecting networks in different flow states to form a network set serving as a network training body;
based on a deep learning method, utilizing the network training body to carry out interactive training to obtain a scheduling training body;
measuring network flow information of the SDN network equipment, and acquiring a network state of the SDN network equipment according to a result of measuring the network flow information of the SDN network equipment;
judging whether the collected network state exists in the network set,
if the judgment result is yes, the user can judge that the operation is right,
scheduling the acquired network through a scheduling training body;
if the answer is no, the method further comprises the following steps of,
storing the acquired network state in a network set serving as a network training body, forming a new network set, carrying out interactive training again to obtain an updated scheduling training body, and then scheduling the acquired network through the updated scheduling training body.
Furthermore, based on a deep learning method, when the network training body is used for interactive training, a flow demand matrix is used as an input state.
Furthermore, when the flow matrix is used as the input state, the link load information is added in the network state representation, so that the accuracy of the feature representation learning of the key data stream is improved, and the training speed is further accelerated.
Further, the interactive training by using the network training entity based on the deep learning method is specifically based on the principle of the A3C algorithm.
Further, the network flow information measurement of the SDN network device is specifically realized by a NetFlow system.
Furthermore, when network flow information measurement is realized through a NetFlow system, the acquisition point is arranged at a core layer of the network, the NetFlow does not need to be started at an interconnection port between core layer routers, and the external Internet port of the core node router starts NetFlow inflow flow acquisition.
Furthermore, when network flow information measurement is realized through a NetFlow system, the acquisition point is arranged at an edge layer of the network, an external interconnection port of the edge layer router starts NetFlow inflow flow acquisition, and flow entering the network from other AS is analyzed.
Further, the network traffic information includes the number of input bytes, the number of input non-broadcast packets, the number of input broadcast packets, the number of output packet drops, the number of input packet errors, the number of input unknown protocol packets, the number of output bytes, the number of output non-broadcast packets, the number of output packet drops, the number of output packet errors, and the output queue length.
Furthermore, when the acquired network is scheduled through the scheduling training body, the network flow information is firstly subjected to differentiation and classification detection through the classification detection module.
A system for intelligently scheduling network flow comprises SDN network equipment, a NetFlow system, a network set storage module, a debugging module and a regulation and control platform;
the SDN network device is used for transmitting network traffic of a public network to an application server,
the NetFlow architecture is used for measuring network traffic information of the SDN network device,
the network set storage module is used for storing network flow data in different states, namely a network set used for a network training body,
the debugging module generates a scheduling instruction according to a scheduling training body obtained in advance and issues the scheduling instruction to the regulation and control platform;
the regulation and control platform is used for dynamically controlling network flow transmitted to the application terminal by the SDN network device according to the scheduling instruction.
Further, the NetFlow system comprises a detector, a collector and a reporting system;
the probes are used to listen to the network data,
the collector is used for collecting the data transmitted by the detector,
the reporting system is used to generate legible reports from the data collected by the collector.
Furthermore, the system for intelligently scheduling the network traffic also comprises a grading detection module, wherein the grading detection module carries out differentiation grading detection on the network traffic information, so that the debugging module can conveniently execute a differentiation scheduling strategy.
The technical scheme of the invention has the following beneficial effects:
1. the invention utilizes a deep reinforcement learning method, takes a flow demand matrix as state input, and trains an Agent in continuous interaction with different network states. The method learns how to select the key data flow set which has the greatest influence on the network congestion in different network states, and then achieves the aims of reducing the network congestion, reducing the network delay and improving the link utilization rate by scheduling the data flows.
2. The invention optimizes the flow transmission in the SDN (software Defined network) network by utilizing the deep reinforcement learning network technology, realizes the real-time, dynamic and self-adaptive flow scheduling, effectively improves the network resource utilization rate and the data transmission quality, has good convergence and robustness under different network architectures and achieves the purpose of fully utilizing the resources in the network.
3. According to the invention, through the hierarchical detection module, differential scheduling strategies are executed for different types of flows according to the result of the flow hierarchical detection, so that the network congestion condition is avoided, the network performance such as the network link utilization rate is improved, and the network service quality is guaranteed. Simulation results show that the differentiated flow scheduling can dynamically execute a flow scheduling strategy according to the current network state, and the good performance of the network is guaranteed.
4. The NetFlow system is used for measuring the network flow information of the SDN network device, so that adverse effects of abnormal flow on a network main body, such as network congestion caused by occupied bandwidth resources, network packet loss and time delay increase, and network unavailability caused by serious conditions or network device system resources (CPU, memory and the like) occupied, are avoided, and the network cannot provide normal services.
Drawings
FIG. 1 is a flow chart of a method for intelligent scheduling of network traffic in accordance with the present invention;
FIG. 2 is a block diagram of a system for intelligent scheduling of network traffic according to the present invention;
FIG. 3 is a second block diagram of the system for intelligently scheduling network traffic according to the present invention;
fig. 4 is a block diagram of the NetFlow system of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to fig. 1 to 3 of the embodiments of the present invention. It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the described embodiments of the invention, are within the scope of the invention.
The traditional data center architecture and the traffic scheduling mode can not meet the requirements of low-delay connection and high-quality transmission, the traffic scheduling method lacks dynamic adaptability, and in addition, the problems of traffic and resource scheduling such as unbalanced network traffic load and low link resource utilization rate of the data center exist. The method learns how to select the key data flow set which has the greatest influence on the network congestion under different network states, and then achieves the aims of reducing the network congestion, reducing the network delay and improving the link utilization rate by scheduling the data flows.
The A3C algorithm is called Asynchronous advertisement Actor-Critic, and the Actor-Critic is put into a plurality of threads for synchronous training, so that computer resources can be effectively utilized, and training effectiveness is improved. In short, each core of the server is a thread, that is, a parallel world, and the same program runs in the parallel world simultaneously, so that the running speed can be increased by multiple times. The running result in each thread is fed back to the main network, and the latest parameter update is obtained from the main network, so that a plurality of threads are combined together, the correlation of events is further weakened, and the convergence of programs is facilitated. The principle of the A3C algorithm belongs to the prior art and is not described in detail herein.
As shown in fig. 1: a method for intelligently scheduling network traffic comprises the following steps:
s1: selecting networks in different flow states to form a network set serving as a network training body;
s2: based on a deep learning method, utilizing the network training body to carry out interactive training to obtain a scheduling training body;
and based on a deep learning method, when the network training body is used for interactive training, a flow demand matrix is used as an input state.
When the flow matrix is used as input state, link load information is added in network state representation, accuracy of learning of key data flow feature representation is improved, and training speed is further increased.
Wherein, the deep learning method is specifically based on the principle of an A3C algorithm when the network training body is used for interactive training.
By utilizing a deep reinforcement learning network technology, traffic transmission is optimized in an SDN (software Defined network) network, real-time, dynamic and self-adaptive traffic scheduling is realized, the network resource utilization rate and the data transmission quality are effectively improved, good convergence and robustness can be realized under different network architectures, and the purpose of fully utilizing resources in the network is achieved.
S3: measuring network flow information of the SDN network equipment, and acquiring a network state of the SDN network equipment according to a result of measuring the network flow information of the SDN network equipment;
the network flow information measurement of the SDN network device is specifically realized through a NetFlow system.
When the network flow information measurement is realized through the NetFlow system, the acquisition point is arranged at the core layer of the network, the NetFlow does not need to be started by the interconnection port between the routers of the core layer, and the NetFlow inflow flow acquisition is started by the external Internet port of the router of the core node.
Through the scheme, the number of the collected routers is small, so that the management is simple, and the configuration workload is small.
In other embodiments, when network traffic information measurement is realized through a NetFlow system, an acquisition point is arranged at an edge layer of a network, an external interconnection port of an edge layer router starts NetFlow inflow traffic acquisition, and traffic entering the network from other AS is analyzed.
By the mode, the acquisition ports are dispersed on the plurality of routers on the edge layer, accordingly, the data acquisition amount on a single router and the burden increased by flow acquisition are reduced, and the influence of the open of the Netflow on a service network is reduced.
The network flow information comprises input byte number, input non-broadcast packet number, input broadcast packet number, output packet discarding number, input packet error number, input unknown protocol packet number, output byte number, output non-broadcast packet number, output packet discarding number, output packet error number and output queue length.
Through the hierarchical detection module, according to the result of the hierarchical detection of the flow, the differentiated scheduling strategies are executed for different types of flows, and the network congestion condition is avoided, so that the network performance such as the network link utilization rate is improved, and the network service quality is guaranteed. Simulation results show that the differentiated flow scheduling can dynamically execute a flow scheduling strategy according to the current network state, and the good performance of the network is guaranteed.
S4: judging whether the collected network state exists in the network set,
s41: if the judgment result is yes, the user can judge that the operation is abnormal,
scheduling the acquired network through a scheduling training body;
s42: if the answer is no, the method further comprises the following steps of,
storing the acquired network state in a network set serving as a network training body, forming a new network set, carrying out interactive training again to obtain an updated scheduling training body, and then scheduling the acquired network through the updated scheduling training body.
When the acquired network is scheduled through the scheduling training body, differentiation and grading detection are firstly carried out on the network flow information through the grading detection module.
A system for intelligently scheduling network flow comprises SDN network equipment, a NetFlow system, a network set storage module, a debugging module and a regulation and control platform;
the SDN network device is used for transmitting network traffic of a public network to an application server,
the NetFlow architecture is used for measuring network traffic information of the SDN network device,
the network set storage module is used for storing network flow data in different states, namely a network set used for a network training body,
the debugging module generates a scheduling instruction according to a scheduling training body obtained in advance and issues the scheduling instruction to the regulation and control platform;
the regulation and control platform is used for dynamically controlling network flow transmitted to the application terminal by the SDN network device according to the scheduling instruction.
Further, the NetFlow system comprises a detector, a collector and a reporting system;
the probes are used to listen to the network data,
the collector is used for collecting the data transmitted by the detector,
the reporting system is used to generate easily readable reports from the data collected by the collectors.
The NetFlow system is used for measuring the network flow information of the SDN network device, so that adverse effects of abnormal flow on a network main body, such as network congestion caused by occupied bandwidth resources, network packet loss and time delay increase, and network unavailability caused by serious conditions or network device system resources (CPU, memory and the like) occupied, are avoided, and the network cannot provide normal services.
The system for intelligently scheduling the network traffic further comprises a grading detection module, wherein the grading detection module carries out differentiation grading detection on the network traffic information, so that a debugging module can conveniently execute a differentiation scheduling strategy.
By the technical scheme, real-time, dynamic and self-adaptive flow scheduling is realized, the utilization rate of network resources and the data transmission quality are effectively improved, the network performance is optimized, the convergence rate of the flow scheduling method can be improved, the flow scheduling method has certain generalization capability, and good convergence and robustness can be achieved under different network architectures, so that the purpose of fully utilizing the resources in the network is achieved, differential scheduling strategies are executed for different types of flows according to the result of flow classification detection, the network congestion condition is avoided, the network performance such as the utilization rate of network links is improved, and the network service quality is guaranteed. Simulation results show that the differentiated flow scheduling can dynamically execute a flow scheduling strategy according to the current network state, and the good performance of the network is guaranteed.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (10)
1. A method for intelligently scheduling network traffic is characterized by comprising the following steps:
selecting networks in different flow states to form a network set serving as a network training body;
based on a deep learning method, utilizing the network training body to carry out interactive training to obtain a scheduling training body;
measuring network flow information of the SDN network equipment, and acquiring a network state of the SDN network equipment according to a result of measuring the network flow information of the SDN network equipment;
judging whether the collected network state exists in the network set,
if the judgment result is yes, the user can judge that the operation is abnormal,
scheduling the acquired network through a scheduling training body;
if the answer is no, the method further comprises the following steps of,
storing the acquired network state in a network set serving as a network training body, forming a new network set, carrying out interactive training again to obtain an updated scheduling training body, and then scheduling the acquired network through the updated scheduling training body.
2. The method for intelligent scheduling of network traffic of claim 1, wherein: and based on a deep learning method, when the network training body is used for interactive training, a flow demand matrix is used as an input state.
3. The method for intelligent scheduling of network traffic of claim 2, wherein: when the flow matrix is used as input state, link load information is added in the network state representation, accuracy of feature representation learning of the key data stream is improved, and training speed is further accelerated.
4. The method of intelligent scheduling of network traffic of claim 3, wherein: the deep learning method is based on the principle of the A3C algorithm when the network training body is used for interactive training.
5. The method for intelligent scheduling of network traffic of claim 1, wherein: the network flow information measurement of the SDN network device is specifically realized through a NetFlow system.
6. The method for intelligent scheduling of network traffic of claim 5, wherein: the network flow information comprises input byte number, input non-broadcast packet number, input broadcast packet number, output packet discarding number, input packet error number, input unknown protocol packet number, output byte number, output non-broadcast packet number, output packet discarding number, output packet error number and output queue length.
7. The method for intelligent scheduling of network traffic of claim 1, wherein: when the acquired network is scheduled through the scheduling training body, the network flow information is firstly subjected to differential hierarchical detection through the hierarchical detection module.
8. A system for intelligently scheduling network flow is characterized by comprising SDN network equipment, a NetFlow system, a network set storage module, a debugging module and a regulation and control platform;
the SDN network device is used for transmitting network traffic of a public network to an application server,
the NetFlow architecture is used for measuring network traffic information of the SDN network device,
the network set storage module is used for storing network flow data in different states, namely a network set used for a network training body,
the debugging module generates a scheduling instruction according to a scheduling training body obtained in advance and issues the scheduling instruction to the regulation and control platform;
the regulation and control platform is used for dynamically controlling network flow transmitted to the application terminal by the SDN network device according to the scheduling instruction.
9. The system for intelligently scheduling network traffic according to claim 8, wherein the NetFlow architecture comprises a probe, a collector, a reporting system;
the probes are used to listen to the network data,
the collector is used for collecting the data transmitted by the detector,
the reporting system is used to generate easily readable reports from the data collected by the collectors.
10. The system according to claim 8, further comprising a hierarchical detection module, wherein the hierarchical detection module performs differentiated hierarchical detection on the network traffic information, so as to facilitate the debugging module to execute the differentiated scheduling policy.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210631312.7A CN115037689A (en) | 2022-06-06 | 2022-06-06 | Method and system for intelligently scheduling network traffic |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210631312.7A CN115037689A (en) | 2022-06-06 | 2022-06-06 | Method and system for intelligently scheduling network traffic |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115037689A true CN115037689A (en) | 2022-09-09 |
Family
ID=83123283
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210631312.7A Pending CN115037689A (en) | 2022-06-06 | 2022-06-06 | Method and system for intelligently scheduling network traffic |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115037689A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111756634A (en) * | 2020-07-15 | 2020-10-09 | 中国舰船研究设计中心 | Carrier-based network performance self-optimization method based on reinforcement learning |
| US20200403913A1 (en) * | 2019-06-21 | 2020-12-24 | Beijing University Of Posts And Telecommunications | Network Resource Scheduling Method, Apparatus, Electronic Device and Storage Medium |
| CN112564974A (en) * | 2020-12-08 | 2021-03-26 | 武汉大学 | Deep learning-based fingerprint identification method for Internet of things equipment |
| CN112949739A (en) * | 2021-03-17 | 2021-06-11 | 中国电子科技集团公司第二十九研究所 | Information transmission scheduling method and system based on intelligent traffic classification |
| CN113904842A (en) * | 2021-09-30 | 2022-01-07 | 中国信息通信研究院 | Method for detecting DDoS attack in IPv6 network based on condition generation countermeasure network under SDN |
| CN114189937A (en) * | 2021-11-10 | 2022-03-15 | 中国科学院计算技术研究所 | Real-time centralized wireless network scheduling method and device based on deep reinforcement learning |
-
2022
- 2022-06-06 CN CN202210631312.7A patent/CN115037689A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20200403913A1 (en) * | 2019-06-21 | 2020-12-24 | Beijing University Of Posts And Telecommunications | Network Resource Scheduling Method, Apparatus, Electronic Device and Storage Medium |
| CN111756634A (en) * | 2020-07-15 | 2020-10-09 | 中国舰船研究设计中心 | Carrier-based network performance self-optimization method based on reinforcement learning |
| CN112564974A (en) * | 2020-12-08 | 2021-03-26 | 武汉大学 | Deep learning-based fingerprint identification method for Internet of things equipment |
| CN112949739A (en) * | 2021-03-17 | 2021-06-11 | 中国电子科技集团公司第二十九研究所 | Information transmission scheduling method and system based on intelligent traffic classification |
| CN113904842A (en) * | 2021-09-30 | 2022-01-07 | 中国信息通信研究院 | Method for detecting DDoS attack in IPv6 network based on condition generation countermeasure network under SDN |
| CN114189937A (en) * | 2021-11-10 | 2022-03-15 | 中国科学院计算技术研究所 | Real-time centralized wireless network scheduling method and device based on deep reinforcement learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108259367B (en) | Service-aware flow strategy customization method based on software defined network | |
| CN109995583B (en) | Delay-guaranteed NFV cloud platform dynamic capacity expansion and contraction method and system | |
| CN112491714B (en) | Intelligent QoS route optimization method and system based on deep reinforcement learning in SDN environment | |
| Hui et al. | Digital twin for networking: A data-driven performance modeling perspective | |
| CN110324260B (en) | Network function virtualization intelligent scheduling method based on flow identification | |
| Hu et al. | EARS: Intelligence-driven experiential network architecture for automatic routing in software-defined networking | |
| CN114884895A (en) | Intelligent traffic scheduling method based on deep reinforcement learning | |
| CN113518012B (en) | Distributed cooperative flow simulation environment construction method and system | |
| Wang et al. | xnet: Improving expressiveness and granularity for network modeling with graph neural networks | |
| US20230145097A1 (en) | Autonomous traffic (self-driving) network with traffic classes and passive and active learning | |
| Huang et al. | Intelligent traffic control for QoS optimization in hybrid SDNs | |
| CN116599904A (en) | Parallel transmission load balancing device and method | |
| Sun et al. | Accelerating convergence of federated learning in mec with dynamic community | |
| WO2023045565A1 (en) | Network management and control method and system thereof, and storage medium | |
| Le et al. | An ai-based traffic matrix prediction solution for software-defined network | |
| Zheng et al. | Enabling robust DRL-driven networking systems via teacher-student learning | |
| CN108280018A (en) | A kind of node workflow communication overhead efficiency analysis optimization method and system | |
| CN114189433A (en) | Intention-driven network system | |
| Alkenani et al. | Enhance work for java based network analyzer tool used to analyze network simulator files | |
| CN115037689A (en) | Method and system for intelligently scheduling network traffic | |
| CN114050984B (en) | Intelligent power distribution and utilization business communication bandwidth prediction method for intelligent park | |
| Zhou et al. | Tsengine: Enable efficient communication overlay in distributed machine learning in wans | |
| CN110601897A (en) | Network resource configuration method and device | |
| Aykurt et al. | Autonomous Network Management in Multi-Domain 6G Networks based on Graph Neural Networks | |
| Zhao et al. | Power-Efficient Software-Defined Data Center Network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |