WO2020119610A1 - 一种用于故障根因的识别方法、装置和设备 - Google Patents
一种用于故障根因的识别方法、装置和设备 Download PDFInfo
- Publication number
- WO2020119610A1 WO2020119610A1 PCT/CN2019/123841 CN2019123841W WO2020119610A1 WO 2020119610 A1 WO2020119610 A1 WO 2020119610A1 CN 2019123841 W CN2019123841 W CN 2019123841W WO 2020119610 A1 WO2020119610 A1 WO 2020119610A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- flow
- failure
- success
- target
- root cause
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/06—Management of faults, events, alarms or notifications
- H04L41/0631—Management of faults, events, alarms or notifications using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis
- H04L41/065—Management of faults, events, alarms or notifications using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis involving logical or physical relationship, e.g. grouping and hierarchies
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/06—Management of faults, events, alarms or notifications
- H04L41/0631—Management of faults, events, alarms or notifications using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/088—Non-supervised learning, e.g. competitive learning
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/09—Supervised learning
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/092—Reinforcement learning
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/06—Management of faults, events, alarms or notifications
- H04L41/0654—Management of faults, events, alarms or notifications using network fault recovery
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/16—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using machine learning or artificial intelligence
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/04—Processing captured monitoring data, e.g. for logfile generation
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N7/00—Computing arrangements based on specific mathematical models
- G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
Definitions
- the network it includes: a server and a forwarding node.
- Various forwarding nodes can be used to access various protocols between the servers, for example: Transmission Control Protocol (English: Transmission Control Protocol, TCP) access, Internet Protocol (English: Internet Protocol (IP: Access) and User Datagram Protocol (English: User Datagram Protocol, UDP: Access) etc.
- TCP Transmission Control Protocol
- IP Internet Protocol
- UDP User Datagram Protocol
- the specific process includes: the source server sends a TCP connection request to the destination server through each forwarding node, and the destination server sends a TCP response message to the source server through each forwarding node, thereby establishing a TCP access connection.
- a target success flow that is similar to the failure flow from multiple success flows in the network that is, a target success flow that differs from the failure flow by only a small number of characteristic indicators is determined, and a large number of less successful differences are used in combination
- the first machine learning model repeatedly trained on the success and failure flows can accurately learn the difference between the characteristic indexes between the failure flow to be analyzed and its related target success flow, thus based on the small number of different characteristic indexes,
- the target root cause of the failure flow can be accurately output. In this way, the root cause of the failure that causes the connectivity failure of the failure flow is realized. Accurate identification, thereby saving the maintenance cost of the network and improving the user experience of using the network.
- the streams mentioned in the embodiments of the present application for example: a first failure stream, a first target success stream, multiple first success streams, multiple second failure streams, and a second related to each second failure stream
- the target success flow and the plurality of second success flows are all TCP flows, IP flows, or UDP flows. It can be understood that the first failure flow, the first target success flow, and multiple first success flows are the same type of flow. Similarly, multiple second failure flows, and a second target related to each second failure flow The success flow and multiple second success flows must also be the same type of flow.
- the first target success flow and the first failure flow have a high similarity, specifically: the similarity between the first target success flow and the first failure flow is greater than a preset similarity threshold, or, the first target
- the similarity between the success flow and the first failure flow belongs to the largest N in the similarity between the plurality of first success flows and the first failure flow in the first network, and N is a preset value.
- an embodiment of the present application provides a computer-readable storage medium, where the computer-readable storage medium stores instructions, and when it runs on a computer, causes the computer to execute the method described in the first aspect.
- the machine learning model 100 since the information of the flow records generated in the network is complex and numerous, even if the convolutional neural network module 110 and the fully connected module 120 are reasonably constructed, and the training samples are large enough Set training, the machine learning model 100 cannot accurately obtain the true root cause of the failure of the input stream, and the machine learning model 100 is only suitable for identifying the root cause of the failure of the stream generated by the network corresponding to the training sample. It cannot be generalized and applied. Therefore, the machine learning model 100 cannot be widely used online no matter how many trainings are performed regardless of whether the input is all streams in the network or all failed streams in the network.
- a method for automatically and accurately identifying the root cause of the failure flow failure is provided That is, from a plurality of first success flows of the first network, a first target success flow related to the first failure flow is determined for the first failure flow generated in the first network, and introduced for the first failure flow Its first target success flow with high similarity; then, through the first failure flow and the first target success flow and the trained first machine learning model, the trained first machine learning model can be compared Learning two streams with high similarity, it is easier to find a small difference between the two streams, then you can effectively analyze the root cause of the target failure corresponding to the difference, that is, the first failure occurs
- the root cause of the connectivity failure thus, there is no need to manually analyze and determine the root cause of the fault through technical personnel, which also makes up for the problem that the root cause of the fault determined by using the method shown in FIG. 1 is not accurate and cannot be widely
- Step 201 Determine a first target success flow related to the first failure flow from multiple first success flows according to the first failure flow in the first network, where the first target success flow and the first failure flow Streams have high similarity.
- the request direction status flag indicates failure
- the TCP flow may be regarded as a failed TCP flow; in another case, when the characteristic indicator of the response recorded in the TCP flow also indicates a failure, then the TCP flow may be regarded as a failed TCP flow.
- the similarity between the first failure flow and each first success flow can be calculated by the following formula (1):
- Step 303 Record the first success flow corresponding to the highest similarity among the calculated similarities as the first target success flow.
- the first successful stream corresponding to the ear and the largest corresponding can be selected according to the following formula (2) as the corresponding to the first failed stream
- I E argmax (E t ) Formula (2) where argmax () is the Si corresponding to the maximum value obtained, I B is the Si corresponding to the maximum value of the determined Bi, recorded as the first target success flow .
- step 201 may also be implemented in other ways, for example: inputting multiple first success streams and first failure streams in the first network into the trained third machine learning model, According to the output result of the third machine learning model, the first target success flow is determined.
- a plurality of first success streams and first failure streams may be converted into data formats to obtain a plurality of first success streams and first failure streams having the same data format.
- the pre-processing function for performing data format conversion on multiple first success streams and first failure streams may be implemented by a pre-processing module independent of the functional unit implementing step 201; or it may be integrated in the Among the functional units that implement step 201, the functional unit that implements step 201 is implemented.
- Step 202 Determine the root cause of the target failure of the first failure flow according to the characteristic index of the first failure flow, the characteristic index of the first target success flow, and the trained first machine learning model.
- the first machine learning model is used to learn the input first target success flow and the first failure flow, and determine and output an output result corresponding to the target failure root cause of the first failure flow.
- the first machine learning model is a trained model obtained by training the constructed first machine learning model with a large number of training sample sets, where each training sample in the training sample set may specifically include multiple A success flow and a failure flow.
- the first machine learning model is trained to obtain the trained first machine learning model
- the training samples in the training sample set used may include:
- the known failure root in the second network is due to the first failure root cause.
- the second failure flow and the second target success flow related to the second failure flow, each training sample, the process of training the initially constructed first machine learning model may specifically include: In the first step, the second failure flow and The second target is successfully streamed into the first machine learning model, and the first learning failure root cause is determined according to the output result; the second step is to determine whether the first learning failure root cause is consistent with the first known failure root cause, if not, then Adjust the parameters of the first machine learning model, re-use the adjusted first machine learning model as the first machine learning model, and return to the first step of execution; until the first If the root cause of the learning fault is consistent with the first known root cause of the fault, the current first machine learning model is determined to be the trained first machine learning model mentioned in step 202.
- the structure of the first machine learning model is shown in FIG. 4, and the first machine learning model 400 may specifically include: a first neural network module 410, a second neural network module 420, and a third neural network module 420 o
- the connection relationship and signal transmission direction of each module in the first machine learning model 400 are specifically:
- the input of the first neural network module 310 may be the first failed stream itself or related data after the first failed stream has been processed; second
- the input of the neural network module 420 may be the first target success stream itself or related data after the first target success stream has been processed;
- the first neural network module 410 and the second neural network module 420 connect the output end to the third neural network module
- the output of the third neural network module 430 is the output of the first machine learning model 400 ⁇
- the first failure flow and the first target success flow itself or related data after processing may be input to the trained first machine learning model
- an output result corresponding to the target failure root cause of the first failure stream is output.
- the output result may be the target failure root cause of the first failure flow, so that the output result may be directly determined as the target failure root cause of the first failure flow; in another case, the The output result may be the identifier corresponding to the target failure root cause of the first failure stream.
- the first network, the second network, and the third network may be the same network or different networks; similarly, the first success flow, The second success flow and the third success flow may be multiple same success flows in the same network, or may be different multiple success flows in the same network, or may be different multiple successes in different networks.
- Flows; the first failure flow, the second failure flow, and the third failure flow may be different failure flows in the same network or failure flows in different networks; they are not specifically limited in the embodiments of the present application.
- the initial coefficient set to be reinforced is determined, it can be based on the randomly selected initial coefficient set, and the characteristic index of the third failure flow and the characteristic index of each third success flow obtained in step 601, According to the above formula (1), the similarity between each third successful flow and the third failed flow under the initial coefficient set is calculated.
- Step 604 Determine the second learning corresponding to the third failure flow under each initial coefficient set according to the characteristic index of the third failure flow, the characteristic index of the third target success flow corresponding to the multiple initial coefficient sets and the second machine learning model The difference between the root cause of the failure and the second known root cause of failure corresponding to the third failure flow.
- the third target success stream and the third failure stream corresponding to the multiple initial coefficient sets are input to the second machine learning model, and according to the output result of the second machine learning model, the first The difference between the second learning failure root cause corresponding to the three failure flows and the second known failure root cause corresponding to the third failure flow.
- the process of performing step 604 may specifically include:
- step 202 reference may be made to the relevant description in step 202.
- the sum of multiple first differences may be used as the difference corresponding to the initial coefficient set; as another example, the multiple first differences may also be averaged, and the average value obtained as the initial coefficient set Corresponding difference; as yet another example, you can also determine the contribution of the target coefficient set to the whole according to each round of learning, set corresponding weights for multiple first differences, and use the weighted values of the first differences and their corresponding weights as The difference corresponding to this initial coefficient set.
- the streams mentioned in the apparatus of the embodiments of the present application are, for example, a first failure stream, a first target success stream, multiple first success streams, multiple second failure streams, and each second failure stream
- the second target success flow and multiple second success flows are all TCP flows, IP flows, or UDP flows. It can be understood that the first failure flow, the first target success flow, and multiple first success flows are the same type of flow. Similarly, multiple second failure flows, and a second target related to each second failure flow The success flow and multiple second success flows must also be the same type of flow.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Databases & Information Systems (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
Description
Claims
Priority Applications (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| EP19896877.8A EP3883185B1 (en) | 2018-12-11 | 2019-12-07 | Fault root cause identification method and apparatus and device |
| BR112021011097-6A BR112021011097A2 (pt) | 2018-12-11 | 2019-12-07 | Método, aparelho e dispositivo de identificação de causa raiz de falha |
| US17/342,659 US11956118B2 (en) | 2018-12-11 | 2021-06-09 | Fault root cause identification method, apparatus, and device |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811512160.9 | 2018-12-11 | ||
| CN201811512160.9A CN111385106B (zh) | 2018-12-11 | 2018-12-11 | 一种用于故障根因的识别方法、装置和设备 |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| US17/342,659 Continuation US11956118B2 (en) | 2018-12-11 | 2021-06-09 | Fault root cause identification method, apparatus, and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2020119610A1 true WO2020119610A1 (zh) | 2020-06-18 |
Family
ID=71077140
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2019/123841 Ceased WO2020119610A1 (zh) | 2018-12-11 | 2019-12-07 | 一种用于故障根因的识别方法、装置和设备 |
Country Status (5)
| Country | Link |
|---|---|
| US (1) | US11956118B2 (zh) |
| EP (1) | EP3883185B1 (zh) |
| CN (1) | CN111385106B (zh) |
| BR (1) | BR112021011097A2 (zh) |
| WO (1) | WO2020119610A1 (zh) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114285730A (zh) * | 2020-09-18 | 2022-04-05 | 华为技术有限公司 | 确定故障根因的方法,装置以及相关设备 |
| US11336507B2 (en) * | 2020-09-30 | 2022-05-17 | Cisco Technology, Inc. | Anomaly detection and filtering based on system logs |
| US11438251B1 (en) * | 2022-02-28 | 2022-09-06 | Bank Of America Corporation | System and method for automatic self-resolution of an exception error in a distributed network |
| CN114338415B (zh) * | 2022-03-08 | 2022-06-03 | 腾讯科技(深圳)有限公司 | 一种端口扫描方法、装置、计算机设备及存储介质 |
| US20250080395A1 (en) * | 2023-09-01 | 2025-03-06 | Dish Wireless L.L.C. | Extreme validation for fault detection |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104063609A (zh) * | 2014-07-01 | 2014-09-24 | 北京金控自动化技术有限公司 | 一种利用神经网络辅助判定污染源监测数据有效性的方法 |
| CN106201757A (zh) * | 2016-07-12 | 2016-12-07 | 网易(杭州)网络有限公司 | 一种异常处理方法和装置 |
| US20170102984A1 (en) * | 2015-10-13 | 2017-04-13 | Huawei Technologies Co., Ltd. | Fault Diagnosis Method and Apparatus for Big-Data Network System |
| CN106959662A (zh) * | 2017-05-10 | 2017-07-18 | 东北大学 | 一种电熔镁炉异常工况识别及控制方法 |
| CN108509975A (zh) * | 2018-01-26 | 2018-09-07 | 北京三快在线科技有限公司 | 一种异常在线聚类方法及装置,电子设备 |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014006692A1 (ja) * | 2012-07-03 | 2014-01-09 | 富士通株式会社 | 制御対象フロー特定プログラム、制御対象フロー特定方法および制御対象フロー特定装置 |
| CN103914735B (zh) * | 2014-04-17 | 2017-03-29 | 北京泰乐德信息技术有限公司 | 一种基于神经网络自学习的故障识别方法及系统 |
| CN106209404B (zh) * | 2015-04-30 | 2019-05-03 | 华为技术有限公司 | 网络异常流量分析方法及系统 |
| WO2017025773A1 (en) * | 2015-08-07 | 2017-02-16 | Telefonaktiebolaget Lm Ericsson (Publ) | Root cause analysis of call failures in a communication network |
| US10027530B2 (en) * | 2015-10-30 | 2018-07-17 | Telefonaktiebolaget Lm Ericsson (Publ) | System and method for troubleshooting SDN networks using flow statistics |
| US10079721B2 (en) * | 2016-04-22 | 2018-09-18 | Netsights360 | Integrated digital network management platform |
| WO2018059687A1 (en) * | 2016-09-29 | 2018-04-05 | Telefonaktiebolaget Lm Ericsson (Publ) | Handling of drop events of traffic flows |
| CN106453392B (zh) * | 2016-11-14 | 2019-04-09 | 中国人民解放军防空兵学院 | 基于流量特征分布的全网络异常流识别方法 |
| US10637715B1 (en) * | 2017-05-02 | 2020-04-28 | Conviva Inc. | Fault isolation in over-the-top content (OTT) broadband networks |
| CN107579956B (zh) * | 2017-08-07 | 2021-05-11 | 奇安信科技集团股份有限公司 | 一种用户行为的检测方法和装置 |
| WO2020036110A1 (ja) * | 2018-08-15 | 2020-02-20 | ソニー株式会社 | ネットワーク監視システム、ネットワーク監視方法、及びプログラム |
-
2018
- 2018-12-11 CN CN201811512160.9A patent/CN111385106B/zh active Active
-
2019
- 2019-12-07 WO PCT/CN2019/123841 patent/WO2020119610A1/zh not_active Ceased
- 2019-12-07 EP EP19896877.8A patent/EP3883185B1/en active Active
- 2019-12-07 BR BR112021011097-6A patent/BR112021011097A2/pt not_active IP Right Cessation
-
2021
- 2021-06-09 US US17/342,659 patent/US11956118B2/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104063609A (zh) * | 2014-07-01 | 2014-09-24 | 北京金控自动化技术有限公司 | 一种利用神经网络辅助判定污染源监测数据有效性的方法 |
| US20170102984A1 (en) * | 2015-10-13 | 2017-04-13 | Huawei Technologies Co., Ltd. | Fault Diagnosis Method and Apparatus for Big-Data Network System |
| CN106201757A (zh) * | 2016-07-12 | 2016-12-07 | 网易(杭州)网络有限公司 | 一种异常处理方法和装置 |
| CN106959662A (zh) * | 2017-05-10 | 2017-07-18 | 东北大学 | 一种电熔镁炉异常工况识别及控制方法 |
| CN108509975A (zh) * | 2018-01-26 | 2018-09-07 | 北京三快在线科技有限公司 | 一种异常在线聚类方法及装置,电子设备 |
Non-Patent Citations (1)
| Title |
|---|
| See also references of EP3883185A4 |
Also Published As
| Publication number | Publication date |
|---|---|
| EP3883185A1 (en) | 2021-09-22 |
| CN111385106B (zh) | 2022-03-01 |
| CN111385106A (zh) | 2020-07-07 |
| EP3883185B1 (en) | 2025-03-26 |
| US20210297305A1 (en) | 2021-09-23 |
| US11956118B2 (en) | 2024-04-09 |
| BR112021011097A2 (pt) | 2021-08-31 |
| EP3883185A4 (en) | 2022-01-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020119610A1 (zh) | 一种用于故障根因的识别方法、装置和设备 | |
| JP7728944B2 (ja) | ブロックチェーンネットワークにおける確率的リレーのためのフロー制御 | |
| US11669751B2 (en) | Prediction of network events via rule set representations of machine learning models | |
| CN109768940B (zh) | 多业务sdn网络的流量分配方法及装置 | |
| CN104584483B (zh) | 用于自动确定服务质量降级的原因的方法和设备 | |
| US9300577B2 (en) | Application intelligent request management based on server health and client information | |
| WO2021052379A1 (zh) | 一种数据流类型识别方法及相关设备 | |
| CN116055324B (zh) | 一种用于数据中心网络自优化的数字孪生方法 | |
| US20200280519A1 (en) | Load Balancer Metadata Forwarding On Secure Connections | |
| CN111835579A (zh) | 一种网络流量调度仿真有效性的测试方法及系统 | |
| Ayan et al. | Quality of service management in telecommunication network using machine learning technique | |
| Tong et al. | Machine learning based root cause analysis for SDN network | |
| KR20220029142A (ko) | Sdn 컨트롤러 서버 및 이의 sdn 기반 네트워크 트래픽 사용량 분석 방법 | |
| Zhang et al. | Achieving high availability in inter-DC WAN traffic engineering | |
| US20250068438A1 (en) | Autonomous generation of network and device configurations | |
| CN114124660B (zh) | 一种修复网络故障的方法及系统 | |
| CN117081995B (zh) | 基于分布式限速策略的云网络服务级别协议保障方法及系统 | |
| CN116367223B (zh) | 基于强化学习的xr服务优化方法、装置、电子设备和存储介质 | |
| CN117579534A (zh) | 智能路由方法、装置、电子设备及存储介质 | |
| Sedaghat et al. | R2T-DSDN: reliable real-time distributed controller-based SDN: S. Sedaghat, AH Jahangir | |
| CN116723154A (zh) | 一种基于负载均衡的路由分发方法及系统 | |
| Xu et al. | Minimizing multi-controller deployment cost in software-defined networking | |
| TWI718778B (zh) | 聯網裝置及其流量預估方法 | |
| CN110365582A (zh) | 一种基于sdn网络的多约束路由方法、一种控制器 | |
| Akinola et al. | Performance Modeling of Software-Defined Networking Paradigm in a Public Health Management System. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 19896877 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| REG | Reference to national code |
Ref country code: BR Ref legal event code: B01A Ref document number: 112021011097 Country of ref document: BR |
|
| ENP | Entry into the national phase |
Ref document number: 2019896877 Country of ref document: EP Effective date: 20210617 |
|
| ENP | Entry into the national phase |
Ref document number: 112021011097 Country of ref document: BR Kind code of ref document: A2 Effective date: 20210609 |