WO2019228190A1 - 网络故障分析方法及装置 - Google Patents

网络故障分析方法及装置 Download PDF

Info

Publication number
WO2019228190A1
WO2019228190A1 PCT/CN2019/087100 CN2019087100W WO2019228190A1 WO 2019228190 A1 WO2019228190 A1 WO 2019228190A1 CN 2019087100 W CN2019087100 W CN 2019087100W WO 2019228190 A1 WO2019228190 A1 WO 2019228190A1
Authority
WO
WIPO (PCT)
Prior art keywords
network element
fault
detected
policy control
log data
Prior art date
Application number
PCT/CN2019/087100
Other languages
English (en)
French (fr)
Inventor
宋平
凌强
陈仁爱
计炜梁
Original Assignee
华为技术有限公司
中国科学技术大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司, 中国科学技术大学 filed Critical 华为技术有限公司
Priority to EP19811179.1A priority Critical patent/EP3787233A4/en
Publication of WO2019228190A1 publication Critical patent/WO2019228190A1/zh
Priority to US17/105,229 priority patent/US20210083925A1/en

Links

Images

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
    • 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
    • H04L41/06—Management of faults, events, alarms or notifications
    • H04L41/0654—Management of faults, events, alarms or notifications using network fault recovery
    • H04L41/0659—Management of faults, events, alarms or notifications using network fault recovery by isolating or reconfiguring faulty entities
    • H04L41/0661—Management of faults, events, alarms or notifications using network fault recovery by isolating or reconfiguring faulty entities by reconfiguring faulty entities
    • 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
    • H04L41/0677—Localisation of faults
    • 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
    • H04L41/069—Management of faults, events, alarms or notifications using logs of notifications; Post-processing of notifications
    • 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
    • H04L41/08—Configuration management of networks or network elements
    • H04L41/0894—Policy-based network configuration management
    • 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
    • H04L41/14—Network analysis or design
    • H04L41/147—Network analysis or design for predicting network behaviour
    • 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
    • H04L41/12—Discovery or management of network topologies
    • 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
    • H04L41/14—Network analysis or design
    • H04L41/142—Network analysis or design using statistical or mathematical methods

Definitions

  • the present application relates to the field of communication technologies, and in particular, to a method and a device for network failure analysis.
  • NF Network Function
  • This application provides two types of network failure analysis methods and devices for implementing failure analysis of network element equipment based on NF services.
  • the present application provides a method for detecting a network fault.
  • the method includes: the fault detection network element obtains log data generated within a set time period of the network element to be detected in the past. After acquiring the log data, the fault detection network element determines the log characteristics corresponding to the log data according to the acquired log data, and finds the log characteristic location in a fault relation database including an association relationship between different fault types and log characteristics. Corresponding fault type. After the fault detection network element detects the fault type, the fault control network element sends the found fault type to the policy control network element, so that the policy control network element can determine a corresponding recovery measure according to the fault type, and further makes the standby The detection network element can handle the fault that occurs through the recovery measures.
  • a fault detection function can be implemented in a 5G network by definition, and a network fault analysis architecture composed of a policy control network element and a fault detection network element is provided. Based on this architecture, the fault detection network element can obtain the log data Determine the fault type, and send the detected fault type to the policy control network element, so that the policy control network element determines a recovery measure, so that the network element to be detected can handle the fault by performing the recovery measure.
  • the network fault analysis method provided in the embodiment of the present application has no limitation on the format of log data, and does not need to input log data in a fixed format.
  • the input log data may be structured data such as ALM or unstructured log data. , Such as the log data generated by equipment produced by different manufacturers, which can reduce the dependence on the log data format, source, etc. during the network failure analysis process, thereby improving the efficiency of log processing.
  • the fault detection network element is further configured to receive a recovery measure sent by the policy control network element, and send the recovery measure to the network element to be detected.
  • the policy control network element may send recovery measures to the fault detection network element. The manner in which the fault detection network element sends to the network element to be detected instructs the network element to be detected to handle the fault that occurs through the recovery measures.
  • the fault detection network element obtains a past design of the network element to be detected.
  • the log data may be requested directly from the NE to be detected, and then the log data sent by the NE to be detected may be received.
  • the fault detection network element obtains log data within a set period of time for the network element to be detected in the past. You can request log data from the policy control network element, and then receive log data from the policy control network element.
  • the fault detection network element may receive a fault analysis sent by the policy control network element.
  • An indication message, and the fault analysis indication message carries at least one of the following information: a function identifier, a detection time, and a detection location.
  • the function identifier is used to identify a fault detection function
  • the detection time is used to indicate a time when the fault detection network element performs fault detection on the to-be-detected network element
  • the detection position is used to indicate the fault detection. The position where the network element performs fault detection on the network element to be detected.
  • the network element to be detected can be treated by the policy control network element as instructed Perform fault detection.
  • the fault detection network element may output the determined log characteristic and receive the fault corresponding to the output log characteristic. Type, learning the determined log feature and the received fault type into the fault relation database.
  • the fault relation database can improve the accuracy of fault detection by continuously learning new fault types and log characteristics.
  • the fault detection network element may further receive an update indication message sent by the policy control network element and update the fault relationship database, and the update indication message is used to instruct the fault detection network Meta updates the fault relation database.
  • the update instruction message may indicate update operations, such as updating the fault relation database, extracting new fault types, etc.
  • the update instruction message may also indicate update requirements, such as the judgment criteria when extracting new faults, and Parameters, etc., the fault detection network element can update the fault relation database under the instruction of the policy control network element.
  • the fault detection network element after the fault detection network element establishes a connection with the policy control network element, it can receive the fault knowledge sent by the policy control network element, and the fault knowledge may include recorded historical log data, and each log data corresponds to Log characteristics, and the type of failure corresponding to each log data.
  • the fault detection network element can be initialized based on the fault knowledge sent by the policy control network element, so that even if the fault detection network element can obtain a more accurate detection result during the initial detection.
  • the fault detection network element may broadcast its own fault type, so that the terminal device receives the After a network fault occurs, the fault detection network element broadcasts the fault type, records the fault type and reports it to the normal fault detection network element, so that the normal fault detection network element sends the fault type reported by the terminal device to the policy control network.
  • the policy control network element determines the recovery measures corresponding to the fault type, and instructs the fault detection network element that has a network fault to handle the network fault that occurs through the recovery measures.
  • the fault detection network element determines that it has a network fault, and the fault type cannot be reported to the policy control network element, the fault detection network element can broadcast its own fault type, so that the fault detection network works normally. After receiving the fault type broadcast by the fault detection network element where the network fault occurs, the element reports it to the policy control network element, so that the policy control network element determines the recovery measures corresponding to the fault type, and instructs the fault detection network element with the network fault to pass This recovery measure deals with network failures that occur.
  • the present application provides a network fault detection method, which includes: a policy control network element receiving a fault type sent by the fault detection network element, where the fault type is that the fault detection network element is faulty on the network element to be detected Time, it is determined according to the log data of the network element to be detected within a set duration in the past.
  • the policy control network element determines a corresponding recovery measure according to the received failure type, and sends the recovery measure to the network element to be detected, to instruct the network element to be detected to handle a fault that occurs through the recovery measure.
  • a fault detection function can be implemented in a 5G network by definition, and a network fault analysis architecture composed of a policy control network element and a fault detection network element is provided. Based on this architecture, the policy control network element can detect the network element based on the fault.
  • the type of the fault sent determines the recovery measures, so that the NE to be detected can handle the faults by performing the recovery measures.
  • the policy control network element when the policy control network element sends the recovery measures to the network element to be detected, if the fault detection network element does not directly communicate with the network element to be detected, but instead The network element to be detected performs communication, and the policy control network element may directly send the recovery measure to the network element to be detected.
  • the policy control network element sends the recovery measures to the network element to be detected, if the policy control network element does not directly communicate with the network element to be detected, but communicates with the network element to be detected through the fault detection network element , The policy control network element may send the recovery measure to the network element to be detected through the fault detection network element.
  • the policy control network element when the fault detection network element does not directly communicate with the network element to be detected, but communicates with the network element to be detected through the policy control element, the policy control network element receives the data sent by the fault detection network element. Prior to the fault type, the policy control network element may further receive a log data request sent by the fault detection network element and send the log data of the network element to be detected to the fault detection network element, so that the fault detection network element obtains Log data to the NE to be detected. The log data request is used to request log data recorded in advance by the network element to be detected within a set period of time in the past.
  • a fault analysis indication message may also be sent to the fault detection network element, and the fault analysis indication message carries at least one of the following information: a function identification, a detection time, and a detection location.
  • the function identifier is used to identify a fault detection function
  • the detection time is used to indicate a time when the fault detection network element performs fault detection on the to-be-detected network element
  • the detection position is used to indicate the fault detection. The position where the network element performs fault detection on the network element to be detected.
  • the network element to be detected can be treated by the policy control network element Perform fault detection.
  • the policy control network element may further send an update indication message to the fault detection network element, and the update indication message is used to instruct the fault detection network element to update the fault relationship database.
  • the update instruction message can indicate update operations, such as updating the fault relationship database, extracting new fault types, and so on.
  • the update instruction message can also indicate update requirements, such as judgment criteria when extracting new faults, parameters used in updating, etc.
  • the policy control network element may instruct the fault detection network element to update the fault relation database by sending the update instruction message.
  • the policy control network element may send fault knowledge to the fault detection network element.
  • the fault knowledge includes recorded historical log data and a log corresponding to each log data. Characteristics, and the type of failure corresponding to each log data.
  • the fault detection network element can be initialized based on the fault knowledge sent by the policy control network element, so that even if the fault detection network element can obtain a more accurate detection result during the initial detection.
  • the present application provides a method for predicting a network fault.
  • the method includes: the fault prediction network element obtains first log data of a network element to be detected within a set period of time in the past.
  • the fault prediction network element predicts the second log data of the network element to be detected after a preset period of time based on the obtained first log data.
  • the fault prediction network element determines a corresponding log feature according to the predicted second log data, and searches for a fault type corresponding to the log feature in a fault relation database, which includes different fault types and log characteristics. connection relation.
  • the fault prediction network element After the predicted fault type is obtained, the fault prediction network element sends the fault type to a policy control network element, so that the policy control network element can determine a corresponding preventive measure according to the fault type, and then the standby The detection network element may be processed through the preventive measures to avoid a failure that may occur after a preset duration.
  • the fault prediction function can be implemented in the 5G network by definition, and a network fault analysis architecture composed of a policy control network element and a fault prediction network element is provided.
  • the fault prediction network element can be based on the acquired past segment
  • the log data in the time period predicts the log data after a period of time in the future, and then determines the type of failure based on the predicted log data, and sends the predicted type of failure to the policy control network element, so that the policy control network element determines the recovery measures, and thus Yuan can execute the recovery measures to avoid possible failures after a preset period of time.
  • the fault prediction network element is further configured to receive a prediction measure sent by the policy control network element, and send the precautionary measure to the network element to be detected.
  • the policy control network element may send recovery measures to the fault prediction network element. The manner in which the fault prediction network element sends to the network element to be detected instructs the network element to be detected to handle the fault that occurs through the recovery measures.
  • the fault prediction network element obtains the pre-recorded network element to be detected
  • the first log data may be requested directly from the network element to be detected, and then the first log data sent by the network element to be detected may be received.
  • the fault prediction network element obtains a pre-recorded time period for the network element to be detected for a set period
  • the first log data may be requested from the policy control network element, and then the first log data sent by the policy control network element will be received.
  • the fault prediction network element may further receive A fault analysis indication message sent by the policy control network element, the fault analysis indication message carrying the preset duration, or the fault analysis indication message carrying at least one of the preset duration and the following information: function Identification, prediction time, prediction location.
  • the function identifier is used to identify a fault prediction function
  • the prediction time is used to indicate a time when the fault prediction network element performs a fault prediction on the network element to be detected
  • the predicted position is used to indicate the fault prediction. The position where the network element performs fault prediction on the network element to be detected.
  • the network element to be detected by the failure prediction may be directed to the network element to be detected under the instruction of the policy control network element Make failure predictions.
  • the fault prediction network element cannot find the fault type corresponding to the log feature in the fault relation database, the determined log feature may be output, and the output log feature corresponding to the For the fault type, the determined log characteristics and the received fault type are learned into the fault relation database.
  • the fault relation database can improve the accuracy of fault prediction by continuously learning new fault types and log characteristics.
  • the fault prediction network element may further receive an update indication message sent by the policy control network element, where the update indication message is used to instruct the fault prediction network element to update the fault relationship database, And update the fault relationship database.
  • the update instruction message may indicate update operations, such as updating the fault relation database, extracting new fault types, etc., and may also indicate the update requirements, such as judgment criteria when extracting new faults, parameters used when updating the fault relation database, and the like
  • the fault prediction network element may update the fault relation database under the instruction of the policy control network element.
  • the fault prediction network element after the fault prediction network element establishes a connection with the policy control network element, it can receive the fault knowledge sent by the policy control network element, and the fault knowledge includes recorded historical log data, and each log data corresponds to Log characteristics, and the type of failure corresponding to each log data.
  • the fault prediction network element can be initialized based on the fault knowledge sent by the policy control network element, so that even if the fault prediction network element can obtain a more accurate prediction result during the initial prediction.
  • the fault detection network element may broadcast its own fault type, so that the terminal device receives the After a network fault occurs, the fault detection network element broadcasts the fault type, records the fault type and reports it to the normal fault detection network element, so that the normal fault detection network element sends the fault type reported by the terminal device to the policy control network.
  • the policy control network element determines the recovery measures corresponding to the fault type, and instructs the fault detection network element that has a network fault to handle the network fault that occurs through the recovery measures.
  • the fault detection network element determines that it has a network fault, and the fault type cannot be reported to the policy control network element, the fault detection network element can broadcast its own fault type, so that the fault detection network works normally. After receiving the fault type broadcast by the fault detection network element where the network fault occurs, the element reports it to the policy control network element, so that the policy control network element determines the recovery measures corresponding to the fault type, and instructs the fault detection network element with the network fault to pass This recovery measure deals with network failures that occur.
  • the present application provides a method for predicting a network fault, the method comprising: a policy control network element receiving a fault type sent by the fault prediction network element, the fault type being the fault prediction network element according to the network to be detected Yuan is determined by log data within a set period of time in the past.
  • the policy control network element determines a corresponding preventive measure according to the type of the fault received, and sends the preventive measure to the to-be-detected network element to instruct the to-be-detected network element to process through the preventive measure to avoid a preset duration Possible failures later.
  • a fault prediction function can be implemented in a 5G network by definition, and a network fault analysis architecture composed of a policy control network element and a fault prediction network element is provided. Based on the architecture, the policy control network element can predict the network element based on the fault.
  • the type of the fault sent determines preventive measures, so that the network element to be detected can perform processing by executing the recovery measure to avoid a fault that may occur after a preset period of time.
  • the policy control network element when the policy control network element sends the precautionary measures to the network element to be detected, if the fault prediction network element does not directly communicate with the network element to be detected, but instead The network element to be detected performs communication, and the policy control network element may directly send the preventive measure to the network element to be detected.
  • the policy control network element sends the restoration measures to the network element to be detected, if the policy control network element does not directly communicate with the network element to be detected, but communicates with the network element to be detected through the fault prediction network element , The policy control network element may send the recovery measure to the network element to be detected through the fault prediction network element.
  • the policy control network element when the fault prediction network element does not directly communicate with the network element to be detected, but instead communicates with the network element to be detected through the policy control element, the policy control network element receives the Prior to the type of failure, the policy control network element may also receive a log data request sent by the fault prediction network element, where the log data request is used to request the network element to be detected in advance for a set period of time in the past Log data, and the policy control network element sends the log data of the network element to be detected to the fault prediction network element, so that the fault prediction network element obtains the log data of the network element to be detected.
  • a fault analysis indication message may also be sent to the fault prediction network element, where the fault analysis indication message carries the preset duration, or the fault analysis indication message carries the preset duration and the following At least one of the information: function identification, prediction time, prediction location.
  • the function identifier is used to identify a fault prediction function
  • the prediction time is used to indicate a time when the fault prediction network element performs a fault prediction on the network element to be detected
  • the predicted position is used to indicate the fault prediction.
  • the network element to be detected by the failure prediction may be directed to the network element to be detected under the instruction of the policy control network element. Make failure predictions.
  • the policy control network element may further send an update indication message to the fault prediction network element, and the update indication message is used to instruct the fault prediction network element to update the fault relationship database.
  • the update instruction message can indicate update operations, such as updating the fault relationship database, extracting new fault types, etc., and can also indicate update requirements, such as judgment criteria when extracting new faults, parameters used during update, etc., policy control The network element may instruct the fault prediction network element to update the fault relation database by sending the update instruction message.
  • the policy control network element after the policy control network element establishes a connection with the fault prediction network element, it can send fault knowledge to the fault prediction network element.
  • the fault knowledge includes recorded historical log data, and a log corresponding to each log data. Characteristics, and the type of failure corresponding to each log data.
  • the fault prediction network element can be initialized based on the fault knowledge sent by the policy control network element, so that even if the fault prediction network element can obtain a more accurate prediction result during the initial prediction.
  • the present application provides a fault detection network element, including: a first communication interface, for sending and receiving messages between the fault detection network element and a policy control network element.
  • Memory for storing computer programs.
  • the processing module is configured to call a computer program stored in the memory and execute: receiving, through the first communication interface or the second communication interface, pre-recorded log data of a network element to be detected within a set period of time in the past.
  • Two communication interfaces are used to send and receive messages between the fault detection network element and the network element to be detected. And, according to the obtained log data, a corresponding log characteristic is determined, and a fault type corresponding to the log characteristic is found in a fault relation database, where the fault relation database includes association relationships between different fault types and log characteristics. .
  • the fault type is sent to the policy control network element through the first communication interface, so that the policy control network element determines according to the fault type.
  • Corresponding recovery measures and instruct the network element to be detected to handle the faults that occur through the recovery measures.
  • the fault relation database may be stored in a memory included in the fault detection network element, or may be stored in other storage areas, such as a cloud storage space and the like.
  • the processing module when the processing module is configured to obtain pre-recorded log data of a network element to be detected within a set period of time in the past, the processing module may be specifically configured to:
  • the network element to be detected sends a log data request, and receives the log data sent by the network element to be detected through the second communication interface, and the log data request is used to request the log data.
  • the log data request is sent to the policy control network element through the first communication interface, and the log data sent by the policy control network element is received through the first communication interface.
  • the processing module may be further configured to receive the policy control network element through the first communication interface before looking up the fault type corresponding to the log characteristic in the fault relation database.
  • the fault analysis indication message sent, the fault analysis indication message carries at least one of the following information: function identification, detection time, and detection location.
  • the function identifier is used to identify a fault detection function
  • the detection time is used to indicate a time when the fault detection network element performs fault detection on the to-be-detected network element
  • the detection position is used to indicate the fault detection. The position where the network element performs fault detection on the network element to be detected.
  • the fault detection network element may further include an input-output module.
  • the processing module is further configured to output the determined log characteristics through the input / output module when the fault type corresponding to the log characteristics cannot be found in the fault relation database. And receiving a corresponding fault type input from the input-output module and determined according to the output log characteristics. And learning the log characteristics output by the input-output module and corresponding fault types received from the input-output module into the fault relation database.
  • the processing module may be further configured to receive an update instruction message sent by the policy control network element through the first communication interface, where the update instruction message is used to instruct to update the fault relationship database; And updating the fault relationship database according to the update instruction message.
  • the present application provides a policy control network element, including: a first communication interface, configured to send and receive messages between the policy control network element and a fault detection network element.
  • Memory for storing computer programs.
  • a processing module for invoking a computer program stored in the memory and executing: receiving, through the first communication interface, a fault type sent by a fault detection network element, where the fault type is that the fault detection network element is faulty on the network element to be detected Time, it is determined according to the log data of the network element to be detected within a set duration in the past. And, a corresponding recovery measure is determined according to the failure type received by the first communication interface.
  • the second communication The interface is used for receiving and sending messages between the fault detection network element and the network element to be detected.
  • the first communication interface may be further configured to receive a log data request sent by the fault detection network element before receiving a fault type sent by the fault detection network element, where the log data request is used for Request log data of the network element to be detected within a set period of time in the past. And sending log data of the network element to be detected to the fault detection network element.
  • the first communication interface may be further configured to send a fault analysis instruction message to the fault detection network element before receiving the fault type sent by the fault detection network element.
  • the function identifier is used to identify a fault detection function
  • the detection time is used to indicate a time when the fault detection network element performs fault detection on the to-be-detected network element
  • the detection position is used to indicate the fault detection. The position where the network element performs fault detection on the network element to be detected.
  • the first communication interface is further configured to send an update indication message to the fault detection network element, where the update indication message is used to instruct the fault detection network element to update the fault relationship database .
  • the present application provides a fault prediction network element, including: a first communication interface, for sending and receiving messages between the fault prediction network element and a policy control network element.
  • Memory for storing computer programs.
  • a processing module for invoking a computer program stored in the memory and executing: receiving, through the first communication interface or the second communication interface, first log data of a previously recorded network element to be detected within a set period of time in the past;
  • the second communication interface is used for receiving and sending messages between the fault detection network element and the network element to be detected.
  • the fault relation database includes association relationships between different fault types and log features. And sending the fault type to the policy control network element through the first communication interface, so that the policy control network element determines a corresponding preventive measure according to the fault type, and instructs the network element to be detected to pass The precautionary measures are taken to avoid malfunctions that may occur after a preset time period.
  • the fault relation database may be stored in a memory included in the fault detection network element, or may be stored in other storage areas, such as a cloud storage space and the like.
  • the processing module when the processing module is configured to obtain pre-recorded log data of a network element to be detected within a set period of time in the past, the processing module may be specifically configured to:
  • the network element to be detected sends a log data request, and receives the first log data sent by the network element to be detected through the second communication interface, and the log data request is used to request the first log data.
  • the log data request is sent to the policy control network element through the first communication interface, and the first log data sent by the policy control network element is received through the first communication interface.
  • the processing module is further configured to pass the first communication interface before predicting the second log data of the network element to be detected after a preset duration based on the first log data.
  • Receiving a fault analysis indication message sent by the policy control network element where the fault analysis indication message carries the preset duration, or the fault analysis indication message carries the preset duration and at least one of the following information: Function identification, prediction time, prediction location.
  • the function identifier is used to identify a fault prediction function
  • the prediction time is used to indicate a time when the fault prediction network element performs a fault prediction on the network element to be detected
  • the predicted position is used to indicate the fault prediction. The position where the network element performs fault prediction on the network element to be detected.
  • the fault prediction network element may further include an input-output module.
  • the processing module is further configured to output the determined log characteristics through the input / output module when the fault type corresponding to the log characteristics cannot be found in the fault relation database. And, receiving a corresponding failure type determined according to the output log characteristics input from the input-output module. And, the log characteristics output by the input-output module and corresponding fault types received from the input-output module are learned into the fault relation database.
  • the processing module is further configured to receive an update instruction message sent by the policy control network element through the first communication interface, where the update instruction message is used to instruct to update the fault relationship database. And updating the fault relationship database according to the update instruction message.
  • the present application provides a policy control network element, including: a first communication interface for receiving and sending messages between the policy control network element and a fault prediction network element.
  • Memory for storing computer programs.
  • a processing module configured to call a computer program stored in the memory, and execute: receiving, through the first communication interface, a fault type sent by a fault prediction network element, where the fault type is the fault prediction network element according to the network to be detected Yuan is determined by log data within a set period of time in the past.
  • a corresponding preventive measure is determined according to the type of failure received by the first communication interface.
  • the second communication The interface is used to send and receive messages between the fault prediction network element and the network element to be detected.
  • the first communication interface may be further configured to receive a log data request sent by the fault prediction network element before receiving a fault type sent by the fault prediction network element, where the log data request is used for Request the first log data of the network element to be detected within a set duration in the past. And sending the first log data of the network element to be detected to the fault prediction network element.
  • the first communication interface may be further configured to send a fault analysis indication message to the fault prediction network element before receiving the fault type sent by the fault prediction network element.
  • Carry the preset duration, or the fault analysis indication message carries the preset duration and at least one of the following information: a function identifier, a prediction time, and a prediction location.
  • the function identifier is used to identify a fault prediction function
  • the prediction time is used to indicate a time when the fault prediction network element performs a fault prediction on the network element to be detected
  • the predicted position is used to indicate the fault prediction. The position where the network element performs fault prediction on the network element to be detected.
  • the first communication interface may be further configured to send an update indication message to the fault prediction network element, where the update indication message is used to instruct the fault prediction network element to update the fault relationship. Library.
  • the present application further provides a computer-readable storage medium for storing computer software instructions for executing a function designed in any one of the first to fourth aspects or any one of the aspects. , which includes a program designed to execute any one of the first to fourth aspects or the method of designing any one of the aspects.
  • an embodiment of the present application provides a computer program product including instructions, which when executed on a computer, causes the computer to execute any one of the first to fourth aspects or any one of the foregoing aspects. Design the method described.
  • the network fault prediction methods provided in the third and fourth aspects of the embodiments of the present application may be used alone as an independent solution, or may be based on the network fault detection methods provided in the first and second aspects of the embodiments of the present application. Further use.
  • FIG. 1 is a schematic architecture diagram of a 5G communication system provided by this application.
  • FIG. 2 is a schematic structural diagram of a network fault analysis system provided by the present application.
  • FIG. 3 is a schematic flowchart of a network fault detection method provided by the present application.
  • FIG. 4 is a schematic flowchart of updating a fault relationship database provided by the present application.
  • FIG. 5 is a schematic diagram of a communication method provided by this application.
  • FIG. 6 is a schematic flowchart of a network fault detection method provided by the present application.
  • FIG. 7 is a schematic diagram of a fault detection network element initialization process provided by the present application.
  • FIG. 8 is a schematic diagram of another communication manner provided by this application.
  • FIG. 9 is a schematic flowchart of another network fault detection method provided by this application.
  • FIG. 10 is a schematic flowchart of a fault detection network element handling a fault provided by the present application.
  • FIG. 11 is a schematic flowchart of another fault detection network element handling a fault provided by the present application.
  • FIG. 12 is a schematic flowchart of a network fault prediction method provided by the present application.
  • FIG. 13 is a schematic diagram of a communication manner provided by this application.
  • FIG. 14 is a schematic flowchart of a network fault prediction method provided by the present application.
  • FIG. 15 is a schematic diagram of another communication manner provided by this application.
  • FIG. 16 is a schematic flowchart of another network fault prediction method provided by the present application.
  • FIG. 17 is a schematic structural diagram of a fault detection network element provided by this application.
  • FIG. 18 is a schematic structural diagram of a fault prediction network element provided by this application.
  • 19 is a schematic structural diagram of a policy control network element provided by this application.
  • 20 is a schematic structural diagram of a network fault analysis network element provided by the present application.
  • FIG. 21 is a schematic structural diagram of a policy control network element provided by this application.
  • NF network function
  • the architecture of the 5G communication system is a service-based architecture (SBA).
  • the architecture of the 5G communication system may include, but is not limited to, network open function network elements, policy control function network elements, data management network elements, application function network elements, core network access and mobility management function network elements, and session management function network elements. , User plane function network elements, and so on. Each functional network element is described below:
  • the session management function network element may be used for session management (including establishment, modification, and release of the session) of the terminal device, selection and reselection of user plane function network elements, and Internet protocol of the terminal device. (IP) address allocation, quality of service (QoS) control, etc.
  • IP Internet protocol
  • QoS quality of service
  • the policy control function network element can be used for policy control decisions, providing functions such as service data flow and application detection, gating, QoS, and flow-based charging control.
  • the policy control function network element may be a PCF (policy control function) network element.
  • the policy control function network element may still be a PCF network element or have another name, which is not limited in this application.
  • the PCF network element may provide an Npcf service.
  • the main function of the application function network element is to communicate with the 3rd Generation Partnership Project (3GPP) core network to provide services to affect service flow routing, access network capability opening, and policy control.
  • 3GPP 3rd Generation Partnership Project
  • the application function network element may be an AF (application function) network element.
  • the data management network element may be used to manage subscription data of the terminal device, registration information related to the terminal device, and the like.
  • the network open function network element can be used to enable 3GPP to securely provide network service capabilities to a third-party server.
  • the core network access and mobility management function network element may be used to manage access control and mobility of the terminal device.
  • the user plane function network element mainly provides user plane service processing functions, including service routing, packet forwarding, anchoring functions, QoS mapping and execution, identification of the uplink and routing to the data network, downlink packet buffer, and downlink Link data arrival notification trigger, connection with external data network, etc.
  • Each of the above network elements can also be called a functional entity, which can be a network element implemented on dedicated hardware, or a software instance running on dedicated hardware, or an example of a virtualized function on an appropriate platform.
  • the above The virtualization platform can be a cloud platform.
  • Figure 1 shows a possible example of a 5G communication system architecture, which specifically includes: application functions (AF), network exposure functions (NEF), policy control functions (PCF), Core access and mobility management functions (AMF), session management functions (SMF), user plane functions (UPF).
  • AF and PCF can be connected through the N5 interface
  • AMF and PCF can be connected through the N15 interface
  • SMF and PCF can be connected through the N7 interface
  • SMF and UPF can be connected through the N4 interface
  • SMF and NEF are connected.
  • the interface name is only an example description, which is not specifically limited in the embodiment of the present application. It should be understood that the names of the network elements shown in FIG.
  • the architecture of the communication system shown in FIG. 1 is not limited to including only the network elements shown in the figure, and may also include other devices not shown in the figure, and specific applications are not listed here one by one.
  • the embodiment of the present application does not limit the distribution form of each network element in the 5G communication system.
  • the distribution form shown in FIG. 1 is only exemplary, and is not limited in the present application.
  • the embodiment of the present application provides a network fault analysis system based on the architecture of a 5G communication system.
  • the system includes a policy control network element, and may further include a fault detection network element and / or a fault prediction network element.
  • the fault detection network element and policy control The network elements are connected through a communication interface, and the fault prediction network element and the policy control network element are connected through a communication interface, as shown in FIG. 2.
  • the fault detection network element can be used to implement fault detection on the network element equipment.
  • a network element that implements fault detection on a network element device may be named another name, such as AA. It should be understood that if the AA can also implement the functions implemented by the fault detection network element in the embodiment of the present application, the AA can also be understood as the fault detection network element in the embodiment of the present application.
  • Fault prediction network element can be used to implement fault prediction for network element equipment.
  • a network element that implements the function of predicting a failure of a network element device may be named another name, such as BB. It should be understood that if the BB can also implement the functions implemented by the fault prediction network element in the embodiment of the present application, the BB can also be understood as the fault prediction network element in the embodiment of the present application.
  • the policy control network element is used to determine a recovery measure according to the fault type detected by the fault detection network element, so that the network element device processes the fault that occurs through the recovery measure, and / or is used to predict the fault obtained by the fault prediction network element
  • the type determines preventive measures so that the network element equipment can prevent future failures through this preventive measure.
  • the network fault analysis system shown in FIG. 2 may further include all or part of the network elements in the 5G system shown in FIG. 1, and may also include other network elements not shown in FIG. 1. Specifically, this application They are not listed here one by one.
  • the embodiment of the present application does not limit the distribution form of each network element in the network fault analysis system.
  • the distribution form shown in FIG. 2 is only exemplary, and is not limited in the present application.
  • the following description uses the network element shown in FIG. 2 as an example. It should be understood that the names of all network elements in this application are merely examples, and may be called other names in future communications, or in the future. The network elements involved in this application in communication may also be replaced by other entities or devices with the same function, which are not limited in this application. The description is unified here, and will not be repeated later.
  • the network failure analysis system shown in FIG. 2 does not constitute a limitation of the communication system applicable to the embodiment of the present application.
  • the communication system shown in FIG. 2 is based on a 5G system architecture.
  • the method in the embodiment of the present application is also applicable to various communication systems in the future.
  • the multiple involved in this application means two or more.
  • the embodiment of the present application provides a network failure detection method and a network failure prediction method based on the network failure analysis system shown in FIG. 2.
  • the network failure detection method and the network failure prediction method can be used alone as a standalone solution or can be combined. Use, such as further use of network failure prediction methods based on network failure detection methods, or further use of network failure detection methods based on network failure prediction methods.
  • the fault detection network element obtains log data recorded in advance by a network element to be detected within a set period of time in the past.
  • the fault detection network element determines a corresponding log feature according to the acquired log data, and searches for a fault type corresponding to the log feature in a fault relation database, where the fault relation database includes association relationships between different fault types and log features. .
  • the fault detection network element may input the acquired log data into a pre-trained neural network model to obtain a fault type corresponding to the log data.
  • the neural network model can be used to extract log characteristics corresponding to log data, and find fault types corresponding to the log characteristics in a fault relation database.
  • the fault detection network element finds the fault type corresponding to the log feature, it sends the found fault type to the policy control network element.
  • the fault detection network element may send a fault analysis result cell to the policy control network element, and the fault analysis result cell may include, but is not limited to, a function identification, an analysis time, an analysis location, a fault state, and specific fault information.
  • the function identifier is used to identify a fault detection function
  • the detection time is used to indicate a time when the fault detection network element performs fault detection on the network element to be detected.
  • the detection position is used to indicate a position where the fault detection network element performs fault detection on the network element to be detected.
  • the fault status is used to identify whether the NE to be detected is faulty.
  • the specific information of the failure includes the type of the failure occurring in the network element to be detected, and / or whether the type of the failure is a known type of failure, and / or whether the type of the failure is a software failure or a hardware failure, and so on.
  • the failure analysis result cell may also include suggested restoration measures and so on.
  • the policy control network element determines a corresponding recovery measure according to the received fault type.
  • the policy control network element instructs the network element to be detected to handle a fault that occurs through the recovery measures.
  • a network fault analysis architecture composed of a policy control network element and a fault detection network element
  • a network fault detection method is provided based on the architecture.
  • the fault detection network element may be based on the acquired log data. Determine the fault type and send the detected fault type to the policy control network element.
  • the policy control network element can determine the recovery measures based on the received fault type, and the network element to be detected can handle the fault by performing the recovery measure.
  • the log data obtained by the fault detection network element may be structured data such as ALM, or may be unstructured data such as log data generated by equipment produced by different manufacturers, etc.
  • the structure of the log data is not limited here. Since the format of the log data is not limited by the fault detection network element, the fault detection network element can perform fault detection on network element equipment produced by different manufacturers, thereby improving the universality of the network fault detection method.
  • step S302 when the fault detection network element searches for the fault type corresponding to the log feature in the fault relation database, if the fault detection network element cannot find the fault type corresponding to the log feature in the fault relation database, it may also Further output the determined log characteristics, for example, to the network management engineer for viewing, so that the network management engineer can artificially analyze the corresponding fault type based on experience, and then input the analyzed fault type to the fault detection network element, such as through a display interface. Method input and output can also be input and output by printing, or input through input devices such as keyboard, and output devices such as display. After the fault detection network element receives the fault type corresponding to the output log feature, it can further learn the determined log feature and the received fault type into the fault relation database, so as to continuously improve and update the fault relation database. The purpose is to make the fault relationship database more comprehensive.
  • the fault detection network element may also update the fault relation database when receiving the update instruction message sent by the policy control network element.
  • the update indication message may be referred to as a fault knowledge update cell.
  • the process of updating the fault relation database can be implemented through steps S401 to S405, as shown in FIG. 4.
  • the policy control network element sends a request update module signaling to the failure detection network element, where the request update module signaling is used to instruct the request failure detection network element to update the failure relation database.
  • the fault detection network element sends an update module response signaling to the policy control network element, and the update module response signaling is used to notify the policy control network element to agree to update the fault relationship database.
  • the policy control network element sends a fault knowledge update cell to the fault detection network element.
  • the fault knowledge update cell may carry an update operation identifier, an update requirement, and the like, where the update operation identifier is used to identify a fault relationship database that is updated for fault detection, or used to identify extraction of a new fault type, or used to Identifies the fault relation database used for fault prediction and so on.
  • the update requirement may be a judgment criterion when extracting a new fault type, parameters used when updating a fault relation database, and the like.
  • the fault detection network element updates the fault relation database based on the received fault knowledge update cell.
  • steps S401 and S402 may not be executed, and may be executed directly from S403.
  • step S405 may also be performed.
  • the fault detection network element sends a fault knowledge cell to the policy control network element.
  • the fault knowledge cell carries the log data accumulated by the fault detection network element, and the log characteristics corresponding to each log data. , And the type of failure corresponding to each log data, and the function ID corresponding to each log data, the function ID is used to identify that the log data was obtained during fault detection, or used to identify that the log data is undergoing failure Obtained during inspection.
  • the fault relation database can improve the accuracy of fault detection, and can improve the scalability of network fault detection methods, so that when the amount of log data is large, it can maintain a good Performance.
  • the fault relation database can be easily expanded.
  • the fault detection network element does not directly communicate with the network element to be detected, but communicates with the policy control network element and the network element to be detected, as shown in FIG. 5.
  • the specific process of the network fault detection method is described with reference to the communication architecture shown in FIG. 5, as shown in FIG. 6.
  • steps S601 and S602 may be performed before the fault detection network element performs fault detection on the network element to be detected.
  • a fault detection network element establishes a connection with a policy control network element.
  • the policy control network element may send status signaling to the fault detection network element, where the status signaling is used to query whether the fault detection network element is in an idle state.
  • the fault detection network element returns to the policy control network element to be idle when the fault detection network element is in the idle state, the fault detection network element establishes a connection with the policy control network element.
  • the fault detection network element is busy, it returns the current state to busy to the policy control network element.
  • the policy control network element sends status signaling to the fault detection network element every first preset time interval until the fault detection network element returns the current status as idle.
  • the first preset duration may be 5s, 3s, 12s, etc., or other time periods. The embodiment of the present application does not specifically limit the first preset duration here.
  • the policy control network element establishes a connection with the network element to be detected. Specifically, the policy control network element may send status signaling to the network element to be detected. When the network element to be detected returns to the policy control network element as idle when it is in an idle state, the network element to be detected establishes a connection with the policy control network element. When the NE to be detected is busy, the current state is returned to the policy control NE as busy.
  • the policy control network element sends the status signaling to the network element to be detected every second preset time interval when the network element to be detected returns the current state as busy, until the network element to be detected returns to the current state as idle.
  • the second preset duration may be 5s, 3s, 12s, etc., or other time periods.
  • the second preset duration may be the same as the first preset duration, and the second preset duration may be different from the first preset duration. In this embodiment of the present application, the second preset duration is not specifically limited herein.
  • step S601 and step S602 have no strict execution order.
  • Step S601 can be executed first, then step S602 can be executed, step S602 can be executed first, step S601 can be executed later, or step S601 and step S602 can be executed simultaneously. Examples are not specifically limited here.
  • the fault detection network element may be initialized under the instruction of the policy control network element.
  • the initialization process of the fault detection network element may be specifically implemented through steps S701 to S705, referring to FIG. 7 Show:
  • the policy control network element sends a request initialization module signaling to the failure detection network element, and the request initialization module signaling is used to request the failure detection network element to initialize.
  • the fault detection network element sends an initialization module response signaling to the policy control network element, and the initialization module response signaling is used to notify the policy control network element to agree to perform initialization.
  • S703 The policy control network element sends a fault knowledge cell to the fault detection network element.
  • the fault detection network element initializes based on the received fault knowledge cell.
  • the fault detection network element sends a module initialization completion signaling to the policy control network element, and the module initialization completion signaling is used to notify the policy control network element of the failure detection network element to complete the initialization.
  • the fault knowledge sent by the policy control network element may include the fault knowledge accumulated by multiple fault detection network elements. Therefore, the fault detection network element Initializing the fault knowledge can perform fault detection on the to-be-detected network element based on the fault knowledge accumulated by multiple fault-detection network elements, thereby improving the accuracy of the fault-detection network element when performing fault detection on the to-be-detected network element.
  • steps S603 to S605 may be performed.
  • the policy control network element sends fault detection signaling to the fault detection network element, and the fault detection signaling is used to instruct the fault detection network element to perform fault detection on the detected network element.
  • the fault detection network element sends a fault detection response signaling to the policy control network element, and the fault detection response signaling is used to instruct the policy control network element to send a fault analysis information cell.
  • the policy control network element sends a fault analysis indication message to the fault detection network element.
  • the fault analysis indication message may be referred to as a fault analysis information cell.
  • the fault analysis information cell may carry at least one of the following information: function identification, detection time, and detection position.
  • step S301 when a faulty network element fails, the fault detection network element obtains pre-recorded log data recorded by the network element to be detected within a set period in the past. Specifically, steps S606 to S609 may be performed. achieve.
  • the policy control network element sends a log data request to the network element to be detected.
  • the log data request may be referred to as a log information cell, and the log information cell is used to request the network element to be detected in advance to record.
  • the log data within a set time period, and the log information cell can carry a policy to control the network element's requirements for the log data, such as the set time, the size of the log data, and the frequency of sending Japanese-style data.
  • the to-be-detected network element sends pre-recorded log data to the policy control network element according to the log information cell.
  • the fault detection network element sends a log information cell to the policy control network element.
  • the log information cell can carry the log data requirements of the fault detection network element, such as the set time, the size of the log data, and the frequency of sending Japanese-style data. and many more.
  • the policy control network element sends the log data sent by the network element to be detected to the fault detection network element according to the received log information cell.
  • step S305 the policy control network element instructs the network element to be detected to handle a fault that occurs through the recovery measures, which may be specifically implemented in step S610.
  • the policy control network element sends a fault recovery cell to the network element to be detected, and the fault recovery cell may carry a function identifier and a recovery measure identifier.
  • the function identifier is used to identify a fault detection function
  • the recovery measure identifier is used to identify a recovery measure determined by a policy control network element.
  • step S611 may also be performed.
  • the network element to be detected sends a failure recovery result cell to the policy control network element, and the failure recovery result cell may carry a function identifier and a recovery result identifier.
  • the function identifier is used to identify a fault detection function.
  • the restoration result identifier is used to identify the restoration success.
  • the recovery result identifier is used to identify the recovery failure.
  • steps S612 and S613 may be performed.
  • S612 The policy control network element is disconnected from the network element to be detected.
  • the policy control network element is disconnected from the fault detection network element.
  • step S612 and step S613 do not have a strict sequential execution order.
  • Step S612 may be performed first, then step S613, or step S613 may be performed first, and then step S612 may be performed at the same time.
  • Step S612 and step S613 may be performed at the same time. Examples are not specifically limited here.
  • the policy control network element does not directly communicate with the network element to be detected, but communicates with the network element to be detected through the fault detection network element, as shown in FIG. 8.
  • the specific process of the network fault detection method is described with reference to the communication architecture shown in FIG. 8, as shown in FIG. 9.
  • step S901 Before the fault detection network element performs fault detection on the network element to be detected, step S901 may be performed.
  • the fault detection network element establishes a connection with the network element to be detected.
  • the network element to be detected may send status signaling to the fault detection network element, where the status signaling is used to query whether the fault detection network element is in an idle state.
  • the fault detection network element When the fault detection network element is in the idle state, it returns a current state to idle to the network element to be detected, and then the fault detection network element establishes a connection with the network element to be detected.
  • the fault detection network element is busy, the current state is returned to the network element to be detected as busy.
  • the NE to be detected is busy when the fault detection NE returns the current status, it sends status signaling to the fault detection NE every third preset interval until the fault detection NE returns the current status as idle.
  • the third preset duration may be 5s, 3s, 12s, etc., or other time periods. The embodiment of the present application does not specifically limit the third preset duration here.
  • step S902 Before executing step S301, when the faulty network element fails, before acquiring the log data recorded in advance by the network element to be detected within a set period of time in the past, step S902 may be performed first.
  • the to-be-detected network element sends a request for failure detection signaling to the failed detection network element, and the requested failure detection signaling is used to request the failed detection network element to perform failure detection on the to-be-detected network element.
  • step S301 when a faulty network element fails, the fault detection network element obtains the log data recorded in advance by the network element to be detected within a set period in the past. Specifically, steps S903 and S904 can be performed. achieve.
  • the fault detection network element sends a log information cell to the network element to be detected, and the log information cell can carry the log data requirements of the fault detection network element, such as the set duration, the size of the log data, and the sending of Japanese-style data Frequency and so on.
  • the to-be-detected network element sends pre-recorded log data to the policy control network element according to the received log information cell.
  • the fault detection network element may send the found fault type to the policy control network element, and then step S905 may be performed first.
  • the fault detection network element establishes a connection with the policy control network element. Specifically, the fault detection network element may send status signaling to the policy control network element. When the policy control network element is in the idle state, it returns the current state to idle to the fault detection network element, and then the fault detection network element establishes a connection with the policy control network element. When the policy control network element is busy, the current state is returned to the fault detection network element as busy. When the fault detection network element returns the current status as busy, the policy control network element sends status signaling to the policy control network element every fourth preset time interval until the policy control network element returns the current status as idle.
  • the fourth preset duration may be 5s, 3s, 12s, etc., or other time periods.
  • the fourth preset duration may be the same as the third preset duration, and the fourth preset duration may be different from the third preset duration. In the embodiment of the present application, the fourth preset duration is not specifically limited herein.
  • the fault detection network element may be initialized under the instruction of the policy control network element.
  • the initialization process of the fault detection network element may be specifically shown in FIG. 7. I will not repeat them here.
  • step S305 the policy control network element instructs the network element to be detected to handle the fault that occurs through the recovery measures, which may be specifically implemented in steps S906 and S907.
  • the policy control network element sends a fault recovery cell to the fault detection network element.
  • the fault detection network element forwards the fault recovery cell to the network element to be detected.
  • steps S908 and S909 may also be performed.
  • the undetected network element sends a fault recovery result cell to the fault detection network element.
  • the fault detection network element forwards the fault recovery result cell to the policy control network element.
  • steps S910 and S911 may be performed.
  • S910 The fault detection network element is disconnected from the network element to be detected.
  • the fault detection network element is disconnected from the policy control network element.
  • step S910 and step S911 do not have a strict execution order.
  • Step S910 can be performed first, then step S911, or step S911 can be performed first, and then step S910 can be performed at the same time.
  • Step S910 and step S911 can be performed simultaneously. Examples are not specifically limited here.
  • the fault detection network element can broadcast the type of fault that occurs, so that the policy control network element determines the corresponding recovery measures according to the fault type, and The fault detection network element is instructed to handle the fault that occurs by itself through this recovery measure.
  • the fault type corresponding to the fault can be sent to the policy control network element through the normal fault detection network element.
  • the fault type corresponding to the fault can be sent to the policy control network element through the normal fault detection network element.
  • the fault detection network element broadcasts the fault type corresponding to the fault that occurs.
  • a fault detection network element where a fault occurs is called an abnormal fault detection network element
  • a normally working fault detection network element is called a normal fault detection network element.
  • S1002 When the normal fault detection network element receives the fault type broadcasted by the abnormal fault detection network element, it forwards the fault type broadcasted by the abnormal fault detection network element to the policy control network element.
  • the policy control network element determines a recovery measure based on the fault type.
  • the policy control network element sends fault recovery signaling to the normal fault detection network element, and the fault recovery signaling is used to instruct the abnormal fault detection network element to handle the fault that occurs based on the recovery measure.
  • S1005 The normal fault detection network element forwards the fault recovery signaling to the abnormal fault detection network element.
  • the abnormal fault detection network element processes the fault that occurs according to the recovery measures indicated by the fault recovery signaling.
  • the abnormal failure detection network element broadcasts a failure recovery result signaling, and the failure recovery signaling is used to notify a policy control network element of a recovery result, where the recovery result is a successful failure clearing, or the recovery result is a failure clearing failure, or The recovery result is success in clearing the risk of failure, or the recovery result is failure to clear the risk of failure, and so on.
  • the fault type corresponding to the fault may be sent to the policy control network element through the terminal device.
  • the fault type corresponding to the fault may be sent to the policy control network element through the terminal device.
  • the fault detection network element broadcasts a fault type corresponding to the fault that occurs.
  • a fault detection network element where a fault occurs is called an abnormal fault detection network element
  • a normally working fault detection network element is called a normal fault detection network element.
  • the terminal device When receiving the fault type broadcast by the abnormal fault detection network element, the terminal device records the fault type and sends the fault type to the normal fault detection network element.
  • S1103 The normal fault detection network element forwards the fault type to the policy control network element.
  • the policy control network element determines a recovery measure based on the fault type.
  • the policy control network element sends fault recovery signaling to the normal fault detection network element, and the fault recovery signaling is used to instruct the abnormal fault detection network element to handle the fault that occurs based on the recovery measure.
  • S1106 The normal fault detection network element forwards the fault recovery signaling to the terminal device.
  • S1107 The terminal device forwards the fault recovery signaling to the abnormal fault detection network element.
  • the abnormal fault detection network element processes the fault that occurs based on the recovery measures indicated by the fault recovery signaling.
  • the abnormal failure detection network element broadcasts the failure recovery result signaling.
  • the terminal device After receiving the fault recovery result signaling broadcast by the abnormal fault detection network element, the terminal device sends the fault recovery result signaling to the normal fault detection network element.
  • S1111 The normal fault detection network element forwards the fault recovery result signaling to the policy control network element.
  • a network fault prediction method provided in an embodiment of the present application, as shown in FIG. 12, a specific process of the network fault prediction method includes:
  • the fault prediction network element obtains log data recorded in advance by a network element to be detected within a set period of time in the past.
  • the fault prediction network element predicts, based on the first log data, the second log data of the network element to be detected after a preset time in the future.
  • the fault prediction network element may input the obtained first log data into a pre-trained sequence-to-sequence (Sequence to Sequence, Seq2Seq) model to obtain the second log data.
  • the Seq2Seq model can be used to predict the log data after a preset duration in the future based on the log data in the past period.
  • the fault prediction network element determines the corresponding log characteristic according to the second log data, and searches for a fault type corresponding to the log characteristic in a fault relation database, where the fault relation database includes association relationships between different fault types and log characteristics. .
  • the fault detection network element may input the second log data into a pre-trained long-term short-term memory (LSTM) model to obtain a fault type.
  • LSTM long-term short-term memory
  • the LSTM model can be used to extract log characteristics corresponding to log data, and find fault types corresponding to the log characteristics in a fault relation database.
  • the fault prediction network element sends the found fault type to the policy control network element.
  • the failure prediction network element may send a failure analysis result cell to the policy control network element
  • the failure analysis result cell may include, but is not limited to, function identification, analysis time, analysis location, fault status, fault specific information, and future forecast. Set the duration and so on.
  • the function identifier is used to identify a fault prediction function
  • the detection time is used to indicate a time when the fault prediction network element performs a fault prediction on the network element to be detected.
  • the detection position is used to indicate a position where the fault prediction network element performs fault prediction on the network element to be detected.
  • the fault status is used to identify whether the NE to be detected is faulty.
  • the specific information of the fault includes a possible fault type of the network element to be detected, and / or whether the fault type is a known fault type, and / or the fault type is a software fault or a hardware fault, and so on.
  • the failure analysis result cell may also include suggested precautions and so on.
  • the policy control network element determines a corresponding preventive measure according to the received fault type.
  • the policy control network element instructs the network element to be detected to perform processing through the precautionary measure to avoid a failure that may occur after a preset duration in the future.
  • a network fault analysis architecture composed of a policy control network element and a fault prediction network element
  • a network fault prediction method is provided based on the architecture.
  • the fault prediction network element may be based on the obtained past segment.
  • the log data in the time period predicts the log data after a period of time in the future, and then determines the fault type based on the predicted log data, and sends the detected fault type to the policy control network element.
  • the policy control network element may determine preventive measures according to the type of the received fault, and the network element to be detected may perform processing by executing the recovery measure to avoid a failure that may occur after a preset period of time.
  • the log data obtained by the fault prediction network element may be structured data such as ALM, or may be unstructured data such as log data generated by equipment produced by different manufacturers, etc.
  • the structure of the log data is not limited here. Since the failure prediction network element has no limitation on the format of the log data, the failure prediction network element can perform failure prediction on network element equipment produced by different manufacturers, thereby improving the universality of the network failure prediction method.
  • step S1203 when the fault prediction network element searches for the fault type corresponding to the log characteristic in the fault relation database, if the fault prediction network element cannot find the fault corresponding to the log characteristic in the fault relation database Type, the determined log characteristics may be output, and the fault type corresponding to the output log characteristics may be received, and the determined log characteristics and the received fault types may be learned into the fault relation database, thereby updating the fault relation database.
  • the fault prediction network element may update the fault relation database when receiving the update instruction message sent by the policy control network element.
  • the update indication message may be referred to as a fault knowledge update cell.
  • the process of updating the fault relation database by the fault prediction network element is similar to the process of updating the fault relation database by the fault detection network element. For details, refer to FIG. 4, which will not be repeated here in this application.
  • the fault prediction network element does not directly communicate with the network element to be detected, but communicates with the policy control network element, the network element to be detected, and the network element to be detected, as shown in FIG. 13.
  • the specific process of the network fault prediction method is described with reference to the communication architecture shown in FIG. 13, as shown in FIG. 14.
  • steps S1401 and S1402 may be performed before the fault prediction network element performs fault prediction on the network element to be detected.
  • the fault prediction network element establishes a connection with the policy control network element. Specifically, the policy control network element may send status signaling to the fault prediction network element. When the fault prediction network element returns to the policy control network element as idle when it is in the idle state, the fault prediction network element establishes a connection with the policy control network element. When the fault prediction NE is busy, it returns the current state to busy to the policy control NE. When the fault prediction network element returns the current status as busy, the policy control network element sends status signaling to the fault prediction network element every first preset time interval until the fault prediction network element returns the current status as idle.
  • the first preset duration may be 5s, 3s, 12s, etc., or other time periods. The embodiment of the present application does not specifically limit the first preset duration here.
  • the policy control network element establishes a connection with the network element to be detected. Specifically, the policy control network element may send status signaling to the network element to be detected. When the network element to be detected returns to the policy control network element as idle when it is in an idle state, the network element to be detected establishes a connection with the policy control network element. When the NE to be detected is busy, the current state is returned to the policy control NE as busy.
  • the policy control network element sends the status signaling to the network element to be detected every second preset time interval when the network element to be detected returns the current state as busy, until the network element to be detected returns to the current state as idle.
  • the second preset duration may be 5s, 3s, 12s, etc., or other time periods.
  • the second preset duration may be the same as the first preset duration, and the second preset duration may be different from the first preset duration. In this embodiment of the present application, the second preset duration is not specifically limited herein.
  • step S1401 and step S1402 do not have a strict sequential execution order.
  • Step S1401 and step S1402 may be performed first, and step S1402 and step S1401 may be performed first, and steps S1401 and S1402 may be performed at the same time. Examples are not specifically limited here.
  • the fault prediction network element can be initialized under the instruction of the policy control network element.
  • the initialization process of the fault prediction network element is similar to the initialization process of the fault detection network element. For details, see As shown in FIG. 7, the embodiment of the present application is not repeated here.
  • steps S1403 to S1405 may be performed.
  • the policy control network element sends fault prediction signaling to the fault prediction network element, and the fault prediction signaling is used to instruct the fault prediction network element to perform fault prediction on the detected network element.
  • the fault prediction network element sends fault prediction response signaling to the policy control network element, and the fault prediction response signaling is used to instruct the policy control network element to send a fault analysis information cell.
  • the policy control network element sends a fault analysis information cell to the fault prediction network element.
  • the fault analysis information cell may carry at least one of the following information: a preset time length in the future, a function identifier, a detection time, and a detection position.
  • step S301 when the faulty network element fails, the fault prediction network element obtains log data recorded in advance by the network element to be detected within a set period of time in the past. Specifically, steps S1406 to S1409 can be performed.
  • the policy control network element sends a log information cell to the network element to be detected.
  • the log information cell can carry the policy control network element's requirements for log data, such as the set time, the size of the log data, and the frequency of sending Japanese-style data. and many more.
  • the network element to be detected sends pre-recorded log data to the policy control network element according to the received log information cell.
  • the fault prediction network element sends a log information cell to the policy control network element, and the log information cell can carry the log data requirements of the fault prediction network element, such as the set duration, the size of the log data, and the frequency of sending Japanese-style data and many more.
  • the policy control network element sends the log data from the network element to be detected to the fault prediction network element according to the received log information cell.
  • the fault prediction network element obtains the first log data.
  • the policy control network element instructs the network element to be detected to handle the fault that occurs through the preventive measures, which may be specifically implemented by step S1410.
  • the policy control network element sends a fault recovery cell to the network element to be detected, and the fault recovery cell may carry a function identifier and a precautionary measure identifier.
  • the function identifier is used to identify a fault prediction function
  • the precautionary measure identifier is used to identify a preventive measure determined by a policy control network element.
  • step S1411 may also be performed.
  • the to-be-detected network element sends a failure recovery result cell to the policy control network element, and the failure recovery result cell may carry a function identifier and a recovery result identifier.
  • the function identification is used to identify a fault prediction function.
  • the recovery result identifier is used to identify that the recovery is successful.
  • the recovery result identifier is used to identify the recovery failure.
  • steps S1412 and S1413 may be performed.
  • the policy control network element is disconnected from the network element to be detected.
  • the policy control network element is disconnected from the fault prediction network element.
  • the policy control network element does not directly communicate with the network element to be detected, but communicates with the network element to be detected through the fault prediction network element, as shown in FIG. 15.
  • the specific process of the network fault prediction method is described with reference to the communication architecture shown in FIG. 15, as shown in FIG. 16.
  • step S1601 Before the fault prediction network element performs fault prediction on the network element to be detected, step S1601 may be performed.
  • the fault prediction network element establishes a connection with the network element to be detected.
  • the network element to be detected may send status signaling to the fault prediction network element, and the status signaling is used to query whether the fault prediction network element is in an idle state.
  • the fault prediction network element is in the idle state, the current state of the fault prediction network element is idle, and the fault prediction network element establishes a connection with the network element to be detected.
  • the fault prediction network element is in a busy state, the current state is returned to the to-be-detected network element as busy.
  • the undetected network element returns the current status of the fault prediction network element as busy, it sends status signaling to the fault prediction network element every third preset time interval until the fault prediction network element returns the current state as idle.
  • the third preset duration may be 5s, 3s, 12s, etc., or other time periods. The embodiment of the present application does not specifically limit the third preset duration here.
  • step S1602 Before executing step S1201, before the fault prediction network element obtains the log data recorded in advance by the network element to be detected within a set period in the past, step S1602 may be performed.
  • the network element to be detected sends a fault prediction request request to the fault prediction network element, and the request fault prediction signaling is used to request the fault prediction network element to perform fault prediction on the network element to be detected.
  • step S1201 the fault prediction network element obtains log data recorded in advance by a network element to be detected within a set period of time in the past, which may be specifically implemented through steps S1603 and S1604.
  • the fault prediction network element sends a log information cell to the network element to be detected, and the log information cell can carry the log data requirements of the fault prediction network element, such as the set duration, the size of the log data, and the sending of Japanese-style data Frequency and so on.
  • the network element to be detected sends the pre-recorded log data to the policy control network element according to the received log information cell.
  • the fault prediction network element obtains the first log data.
  • the fault prediction network element may send the found fault type to the policy control network element, and then step S1605 may be performed first.
  • the fault prediction network element establishes a connection with the policy control network element. Specifically, the fault prediction network element may send status signaling to the policy control network element. When the policy control network element is in the idle state, it returns the current state to idle to the fault prediction network element, and then the fault prediction network element establishes a connection with the policy control network element. The policy control network element returns the current status to busy to the fault prediction network element when it is busy. When the fault prediction network element returns the current status as busy, the policy control network element sends status signaling to the policy control network element every fourth preset time interval until the policy control network element returns the current status as idle.
  • the fourth preset duration may be 5s, 3s, 12s, etc., or other time periods.
  • the fourth preset duration may be the same as the third preset duration, and the fourth preset duration may be different from the third preset duration. In the embodiment of the present application, the fourth preset duration is not specifically limited herein.
  • the fault prediction network element can be initialized under the instruction of the policy control network element.
  • the initialization process of the fault prediction network element is similar to the initialization process of the fault detection network element. For details, see As shown in FIG. 7, the embodiment of the present application is not repeated here.
  • step S1206 the policy control network element instructs the network element to be detected to process through this precautionary measure to avoid a failure that may occur after a preset duration in the future, which may be specifically implemented through steps S1606 and S1607.
  • S1606 The policy control network element sends a fault recovery cell to the fault prediction network element.
  • S1607 The fault prediction network element forwards the fault recovery cell to the network element to be detected.
  • steps S1608 and S1609 may also be performed.
  • the network element to be detected sends a fault recovery result cell to the fault prediction network element.
  • steps S1610 and S1611 may be performed.
  • the fault prediction network element is disconnected from the policy control network element.
  • step S1610 and step S1611 do not have a strict sequential execution order.
  • Step S1610 and step S1611 may be performed first, and step S1611 and step S1610 may be performed first, and step S1610 and step S1611 may be performed simultaneously. Examples are not specifically limited here.
  • the fault prediction network element can broadcast the type of fault that occurs, so that the policy control network element determines the corresponding recovery measures according to the fault type, and The fault prediction network element is instructed to handle the fault that occurs by itself through the recovery measures.
  • the processing process when the fault prediction network element itself fails is similar to the processing process when the fault detection network element itself fails, as shown in FIG. 10 or shown in FIG. 11.
  • the embodiments of this application are not repeated here.
  • each network element such as a fault detection network element and a policy control network element, includes a hardware structure and / or a software module corresponding to each function.
  • a hardware structure and / or a software module corresponding to each function includes a hardware structure and / or a software module corresponding to each function.
  • the present invention can be implemented in the form of hardware or a combination of hardware and computer software by combining the units and algorithm steps of each example described in the embodiments disclosed herein. Whether a certain function is performed by hardware or computer software-driven hardware depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.
  • the fault detection network element may include an acquisition module 1701, a detection module 1702, and a sending module 1703, as shown in FIG.
  • the obtaining module 1701 is configured to perform the foregoing process of obtaining log data of a network element to be detected in FIG. 3 to FIG. 11.
  • the detection module 1702 is configured to perform the foregoing fault detection process based on the log data acquired by the acquisition module 1701 in FIG. 3 to FIG. 11.
  • the sending module 1703 is configured to execute the process of sending information to the network element to be detected and / or the policy control network element in the foregoing FIG. 3 to FIG. 11.
  • the network element may further include a receiving module 1704.
  • the receiving module 1704 is configured to execute the process of receiving the information sent by the network element to be detected and / or the policy control network element in the foregoing FIG. 3 to FIG. 11.
  • the network element may further include an output module 1705 and a learning module 1706.
  • the output module 1705 is configured to execute the above-mentioned process of outputting log features in FIG. 3 to FIG. 11.
  • the learning module 1706 is configured to perform the process of learning the new log feature and the new fault type into the fault relation database in the foregoing FIG. 3 to FIG. 11.
  • the network element may further include an update module 1707.
  • the receiving module 1704 may be further configured to execute the process of updating the fault relationship database in FIG. 3 to FIG. 11. For the method described in the embodiments shown in FIG. 3 to FIG. 11, details are not described herein again.
  • a fault detection function can be implemented in a 5G network by definition, and a network fault analysis architecture composed of a policy control network element and a fault detection network element is provided. Based on this architecture, the fault detection network element can obtain the log data Determine the fault type, and send the detected fault type to the policy control network element, so that the policy control network element determines a recovery measure, so that the network element to be detected can handle the fault by performing the recovery measure.
  • the fault prediction network element may include an acquisition module 1801, a first prediction module 1802, a second prediction module 1803, and a sending module 1804, as shown in FIG. 18.
  • the obtaining module 1801 is configured to perform the foregoing process of obtaining the first log data of the network element to be detected in FIG. 12 to FIG. 16.
  • the first prediction module 1802 is configured to perform the foregoing process of predicting the second log data in FIG. 12 to FIG. 16.
  • the second prediction module 1803 is configured to perform the foregoing process of predicting a fault type in FIG. 12 to FIG. 16.
  • the sending module 1804 is configured to execute the process of sending information to the network element to be detected and / or the policy control network element in the foregoing FIG. 12 to FIG.
  • the network element may further include a receiving module 1805.
  • the receiving module 1805 is configured to perform the foregoing process of receiving information sent by a network element to be detected and / or a policy control network element in FIG. 12 to FIG. 16. For details, reference may be made to the methods described in the embodiments shown in FIG. 12 to FIG. 16, which are not described herein again in this application.
  • the fault prediction function can be implemented in the 5G network by definition, and a network fault analysis architecture composed of a policy control network element and a fault prediction network element is provided. Based on the architecture, the fault prediction network element can be based on the acquired past segment The log data in the time period predicts the log data after a period of time in the future, and then determines the type of failure based on the predicted log data, and sends the predicted type of failure to the policy control network element, so that the policy control network element determines the recovery measures, and thus Yuan can execute the recovery measures to avoid possible failures after a preset period of time.
  • the policy control network element may include a receiving module 1901, a determining module 1902, and a sending module 1903, as shown in FIG. 19.
  • the receiving module 1901 is configured to perform the foregoing process of receiving information sent by a network element to be detected and / or a network element for fault detection in FIG. 3 to FIG. 11.
  • a determining module 1902 is configured to execute the process of determining a recovery measure corresponding to a fault type in the foregoing FIG. 3 to FIG. 11.
  • the sending module 1903 is configured to execute the process of sending information to the network element to be detected and / or the network element for fault detection in the foregoing FIG. 3 to FIG. 11. For the method described in the embodiments shown in FIG. 3 to FIG. 11, details are not described herein again.
  • a fault detection function can be implemented in a 5G network by definition, and a network fault analysis architecture composed of a policy control network element and a fault detection network element is provided. Based on this architecture, the policy control network element can detect the network element based on the fault. The type of the fault sent determines the recovery measures, so that the NE to be detected can handle the faults by performing the recovery measures.
  • the receiving module 1901 is configured to perform the foregoing process of receiving information sent by a network element to be detected and / or a fault prediction network element in FIG. 12 to FIG. 16.
  • a determining module 1902 is configured to execute the foregoing process of determining a preventive measure corresponding to a fault type in FIG. 12 to FIG. 16.
  • the sending module 1903 is configured to execute the process of sending information to the network element to be detected and / or the fault prediction network element in the foregoing FIG. 12 to FIG. 16.
  • a fault prediction function can be implemented in a 5G network by definition, and a network fault analysis architecture composed of a policy control network element and a fault prediction network element is provided. Based on the architecture, the policy control network element can predict the network element based on the fault.
  • the type of the fault sent determines preventive measures, so that the network element to be detected can perform processing by executing the recovery measure to avoid a fault that may occur after a preset period of time.
  • FIG. 20 shows another possible structure diagram of the network fault analysis network element involved in the foregoing embodiment.
  • the network fault analysis network element 200 may be a fault detection network element or a fault prediction network element.
  • the network failure analysis network element 200 includes a processor 2002, a first communication interface 2001A, and a memory 2003.
  • the processor 2002 may be a central processing module (CPU), or a digital processing module.
  • the first communication interface 2001A is used to transmit and receive information to and from the policy control network element, and the memory 2003 is used to store a program executed by the processor 2002.
  • the memory 2003 can also be used to store a fault relation database.
  • the memory 2003 may be a non-volatile memory, such as a hard disk (HDD) or a solid-state drive (SSD), etc., or a volatile memory, such as a random access memory (random -access memory, RAM).
  • the memory 2003 is any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.
  • the processor 2002 is configured to execute the program code stored in the memory 2003, and is specifically configured to perform processing other than the sending and receiving information of the fault detection network element in the methods described in the embodiments shown in FIG. 3 to FIG. Processing other than information. For the method described in the embodiments shown in FIG. 3 to FIG. 16, details are not described herein again.
  • the network failure analysis network element may further include a second communication interface 2001B, and the second communication interface 2001B is configured to transmit and receive information to and from the network element to be detected.
  • the embodiments of the present application are not limited to the specific connection medium between the first communication interface 2001A, the second communication interface 2001B, the processor 2002, and the memory 2003.
  • the memory 2003, the processor 2002, and the first communication interface 2001A and the second communication interface 2001B are connected by a bus 2004 in FIG. 20, and the bus is indicated by a thick line in FIG.
  • the method is only a schematic description, and is not limited thereto.
  • the bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one thick line is used in FIG. 20, but it does not mean that there is only one bus or one type of bus.
  • FIG. 21 shows another possible structure diagram of the policy control network element 210 involved in the foregoing embodiment.
  • the policy control network element 210 includes a processor 2102, a first communication interface 2101A, and a memory 2103.
  • the processor 2102 may be a CPU, or a digital processing module.
  • the first communication interface 2101A is used to send and receive information to and from the fault detection network element, and is used to send and receive information to and from the fault prediction network element.
  • the memory 2103 is used to store a program executed by the processor 2102.
  • the memory 2103 may be a non-volatile memory, such as an HDD or an SSD, and may also be a volatile memory, such as a RAM.
  • the memory 2103 is any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.
  • the processor 2102 is configured to execute the program code stored in the memory 2103, and is specifically configured to execute other processes of the policy control network element in the methods described in the embodiments shown in FIG. 3 to FIG. 16 in addition to sending and receiving information. For the method described in the embodiments shown in FIG. 3 to FIG. 16, details are not described herein again.
  • the policy control network element may further include a second communication interface 2101B, and the second communication interface 2101B is configured to send and receive information to and from the network element to be detected.
  • the embodiments of the present application are not limited to the specific connection medium between the first communication interface 2101A, the second communication interface 2101B, the processor 2102, and the memory 2103.
  • the memory 2103, the processor 2102, the first communication interface 2101A, and the second communication interface 2101B are connected by a bus 2104 in FIG. 21, and the bus is indicated by a thick line in FIG.
  • the method is only a schematic description, and is not limited thereto.
  • the bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only a thick line is used in FIG. 21, but it does not mean that there is only one bus or one type of bus.
  • An embodiment of the present invention further provides a computer-readable storage medium for storing computer software instructions to be executed to execute the processing module, which includes a program for executing the processing module.
  • An embodiment of the present application provides a computer program product containing instructions, which when executed on a computer, causes the computer to execute the network fault analysis method described in FIG. 3 to FIG. 16.
  • this application may be provided as a method, a system, or a computer program product. Therefore, this application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application may take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
  • computer-usable storage media including, but not limited to, disk storage, CD-ROM, optical storage, etc.
  • These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to work in a specific manner such that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device, the instructions
  • the device implements the functions specified in one or more flowcharts and / or one or more blocks of the block diagram.
  • These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of steps can be performed on the computer or other programmable device to produce a computer-implemented process, which can be executed on the computer or other programmable device.
  • the instructions provide steps for implementing the functions specified in one or more flowcharts and / or one or more blocks of the block diagrams.

Landscapes

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

Abstract

本申请提供网络故障分析方法及装置，用以实现基于NF服务对网元设备进行故障分析。该方法包括：故障检测网元获取待检测网元过去一个设定时长内产生的日志数据。在获取到日志数据后，故障检测网元根据获取的日志数据确定该日志数据对应的日志特征，并在故障关系库中查找日志特征所对应的故障类型。故障检测网元在检测到故障类型之后，向策略控制网元发送查找到的故障类型，从而使得策略控制网元可以根据所述故障类型确定对应的恢复措施，进而使得待检测网元可以通过恢复措施处理出现的故障。

Description

网络故障分析方法及装置
本申请要求在2018年05月29日提交中国专利局、申请号为201810526345.9、发明名称为“网络故障分析方法及装置”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及通信技术领域，尤其涉及网络故障分析方法及装置。
背景技术
随着网络的演进，网络管理的趋势也在随之变化，朝着综合化、分布式、智能化方向发展，通过各种网络功能，使其能通过网络分析实现网络资源的计划、建设与优化。在5G标准中明确提出了网络功能(Network Function，NF)服务的概念功能。3GPP文档中虽然对NF服务进行了介绍，但是并没有对NWDAF的网络分析功能进行标准定义，同样没有具体定义5G网络中应该实现什么样的网络功能，怎么实现这些功能，要如何控制和通信等等。因此，对于如何基于NF服务对网元设备进行故障分析，目前还没有解决方案。
发明内容
本申请提供两种网络故障分析方法及装置，用以实现基于NF服务对网元设备进行故障分析。
第一方面，本申请提供了一种网络故障检测方法，该方法包括：故障检测网元获取待检测网元过去一个设定时长内产生的日志数据。在获取到日志数据后，所述故障检测网元根据获取的日志数据确定该日志数据对应的日志特征，并在包括不同故障类型与日志特征的关联关系的故障关系库中查找所述日志特征所对应的故障类型。所述故障检测网元在检测到故障类型之后，向策略控制网元发送查找到的所述故障类型，从而使得策略控制网元可以根据所述故障类型确定对应的恢复措施，进而使得所述待检测网元可以通过所述恢复措施处理出现的故障。本申请实施例中通过定义5G网络中可以实现故障检测功能，并且提供一个由策略控制网元以及故障检测网元构成的网络故障分析架构，基于该架构，故障检测网元可以根据获取的日志数据确定故障类型，并向策略控制网元发送检测的故障类型，以使策略控制网元确定恢复措施，从而待检测网元可以通过执行该恢复措施来处理出现的故障。并且，本申请实施例提供的网络故障分析方法对日志数据的格式没有限制，不需要输入固定格式的日志数据，输入的日志数据可以是结构化的数据比如ALM，也可以是非结构化的日志数据，如不同厂商生产的设备所产生的日志数据，从而可以减少网络故障分析过程中对日志数据格式、来源等的依赖，进而可以提高日志处理的效率。
在一种可能的设计中，所述故障检测网元还用于接收所述策略控制网元发送的恢复措施，并将所述恢复措施发送给所述待检测网元。上述设计中，若策略控制网元不与待检测网元直接通信，而是通过故障检测网元与待检测网元进行通信，则策略控制网元可以通过将恢复措施发送给故障检测网元，然后由故障检测网元发送给待检测网元的方式，指示所 述待检测网元通过所述恢复措施处理出现的故障。
在一种可能的设计中，若策略控制网元不与待检测网元直接通信，而是通过故障检测网元与待检测网元进行通信，则故障检测网元获取待检测网元过去一个设定时长内的日志数据时，可以直接向待检测网元请求日志数据，然后接收待检测网元发送的日志数据。或者，若故障检测网元不与待检测网元直接通信，而是通过策略控制网元与待检测网元进行通信，则故障检测网元获取待检测网元过去一个设定时长内的日志数据时，可以向策略控制网元请求日志数据，然后接收策略控制网元发送的日志数据。
在一种可能的设计中，在所述故障检测网元在故障关系库中查找所述日志特征所对应的故障类型之前，所述故障检测网元可以接收所述策略控制网元发送的故障分析指示消息，所述故障分析指示消息携带如下信息中的至少一项：功能标识、检测时间、检测位置。其中，所述功能标识用于标识故障检测功能，所述检测时间用于指示所述故障检测网元对所述待检测网元进行故障检测的时间，所述检测位置用于指示所述故障检测网元对所述待检测网元进行故障检测的位置。上述设计中，若故障检测网元不与待检测网元直接通信，而是通过策略控制网元与待检测网元进行通信，故障检测网元可以在策略控制网元的指示下对待检测网元进行故障检测。
在一种可能的设计中，若所述故障检测网元在故障关系库中查找不到所述日志特征所对应的故障类型，可以将确定的日志特征输出，并接收输出的日志特征对应的故障类型，将确定的所述日志特征和接收的所述故障类型学习到所述故障关系库中。上述设计中，故障关系库通过不断学习新的故障类型以及日志特征，可以提高故障检测的准确性。
在一种可能的设计中，所述故障检测网元还可以接收所述策略控制网元发送的更新指示消息，并更新所述故障关系库，所述更新指示消息用于指示所述故障检测网元更新所述故障关系库。上述设计中，更新指示消息可以指示更新操作，如更新故障关系库、提取新故障类型等等，更新指示消息还可以指示更新要求，如提取新故障时的判断准则、更新故障关系库时采用的参数等等，故障检测网元可以在策略控制网元的指示下更新故障关系库。
在一种可能的设计中，故障检测网元在与策略控制网元建立连接之后，可以接收策略控制网元发送的故障知识，该故障知识可以包括记录的历史日志数据，以及每个日志数据对应的日志特征，以及每个日志数据对应的故障类型。上述设计中，故障检测网元可以基于策略控制网元发送的故障知识进行初始化，从而即使故障检测网元在初次检测时可以得到较准确的检测结果。
在一种可能的设计中，若故障检测网元确定自身出现网络故障，导致无法向策略控制网元上报故障类型时，该故障检测网元可以广播自身的故障类型，从而终端设备在接收到该出现网络故障的故障检测网元广播的故障类型后，记录该故障类型并上报给正常工作的故障检测网元，以使正常工作的故障检测网元将终端设备上报的故障类型发送给策略控制网元，使得策略控制网元确定该故障类型对应的恢复措施，并指示出现网络故障的故障检测网元通过该恢复措施处理出现的网络故障。
在一种可能的设计中，若故障检测网元确定自身出现网络故障，导致无法向策略控制网元上报故障类型时，该故障检测网元可以广播自身的故障类型，从而正常工作的故障检测网元在接收到该出现网络故障的故障检测网元广播的故障类型后上报给策略控制网元，使得策略控制网元确定该故障类型对应的恢复措施，并指示出现网络故障的故障检测网元 通过该恢复措施处理出现的网络故障。
第二方面，本申请提供了一种网络故障检测方法，该方法包括：策略控制网元接收故障检测网元发送的故障类型，所述故障类型是所述故障检测网元在待检测网元故障时，根据所述待检测网元在过去一个设定时长内的日志数据确定的。所述策略控制网元根据接收到的故障类型确定对应的恢复措施，并向所述待检测网元发送所述恢复措施，以指示所述待检测网元通过所述恢复措施处理出现的故障。本申请实施例中通过定义5G网络中可以实现故障检测功能，并且提供一个由策略控制网元以及故障检测网元构成的网络故障分析架构，基于该架构，策略控制网元可以根据故障检测网元发送的故障类型确定恢复措施，从而待检测网元可以通过执行该恢复措施来处理出现的故障。
在一种可能的设计中，所述策略控制网元向所述待检测网元发送所述恢复措施时，若故障检测网元不与待检测网元直接通信，而是通过策略控制网元与待检测网元进行通信，则策略控制网元可以直接向所述待检测网元发送所述恢复措施。或者，所述策略控制网元向所述待检测网元发送所述恢复措施时，若策略控制网元不与待检测网元直接通信，而是通过故障检测网元与待检测网元进行通信，则策略控制网元可以通过所述故障检测网元向所述待检测网元发送所述恢复措施。
在一种可能的设计中，当故障检测网元不与待检测网元直接通信，而是通过策略控制网元与待检测网元进行通信时，在策略控制网元接收故障检测网元发送的故障类型之前，所述策略控制网元还可以接收所述故障检测网元发送的日志数据请求，并向所述故障检测网元发送所述待检测网元的日志数据，从而故障检测网元获取到待检测网元的日志数据。所述日志数据请求用于请求所述待检测网元预先记录的在过去一个设定时长内的日志数据。
在一种可能的设计中，若故障检测网元不与待检测网元直接通信，而是通过策略控制网元与待检测网元进行通信，则所述策略控制网元在接收故障检测网元发送的故障类型之前，还可以向所述故障检测网元发送故障分析指示消息，所述故障分析指示消息携带如下信息中的至少一项：功能标识、检测时间、检测位置。其中，所述功能标识用于标识故障检测功能，所述检测时间用于指示所述故障检测网元对所述待检测网元进行故障检测的时间，所述检测位置用于指示所述故障检测网元对所述待检测网元进行故障检测的位置。上述设计中，若故障检测网元不与待检测网元直接通信，而是通过策略控制网元与待检测网元进行通信，故障检测网元可以在策略控制网元的指示下对待检测网元进行故障检测。
在一种可能的设计中，所述策略控制网元还可以向所述故障检测网元发送更新指示消息，所述更新指示消息用于指示所述故障检测网元更新所述故障关系库。上述设计中，更新指示消息可以指示更新操作，如更新故障关系库、提取新故障类型等等，更新指示消息还可以指示更新要求，如提取新故障时的判断准则、更新时采用的参数等等，策略控制网元可以通过发送该更新指示消息的方式指示故障检测网元更新故障关系库。
在一种可能的设计中，策略控制网元在与故障检测网元建立连接之后，可以向故障检测网元发送故障知识，该故障知识包括记录的历史日志数据，以及每个日志数据对应的日志特征，以及每个日志数据对应的故障类型。上述设计中，故障检测网元可以基于策略控制网元发送的故障知识进行初始化，从而即使故障检测网元在初次检测时可以得到较为准确的检测结果。
第三方面，本申请提供了一种网络故障预测方法，该方法包括：故障预测网元获取待 检测网元在过去一个设定时长内的第一日志数据。所述故障预测网元基于获取的第一日志数据预测所述待检测网元在预设时长后的第二日志数据。所述故障预测网元根据预测的第二日志数据，确定对应的日志特征，并在故障关系库中查找所述日志特征所对应的故障类型，所述故障关系库包括不同故障类型与日志特征的关联关系。在得到预测的故障类型之后，所述故障预测网元向策略控制网元发送所述故障类型，从而使得所述策略控制网元可以根据所述故障类型确定对应的预防措施，进而使得所述待检测网元可以通过所述预防措施进行处理以避免预设时长后可能出现的故障。本申请实施例中通过定义5G网络中可以实现故障预测功能，并且提供一个由策略控制网元以及故障预测网元构成的网络故障分析架构，基于该架构，故障预测网元可以根据获取的过去一段时间内的日志数据预测未来一段时间后的日志数据，之后根据预测的日志数据确定故障类型，并向策略控制网元发送预测的故障类型，以使策略控制网元确定恢复措施，从而待检测网元可以通过执行该恢复措施进行处理以避免预设时长后可能出现的故障。
在一种可能的设计中，所述故障预测网元还用于接收所述策略控制网元发送的预测措施，并将所述预防措施发送给所述待检测网元。上述设计中，若策略控制网元不与待检测网元直接通信，而是通过故障预测网元与待检测网元进行通信，则策略控制网元可以通过将恢复措施发送给故障预测网元，然后由故障预测网元发送给待检测网元的方式，指示所述待检测网元通过所述恢复措施处理出现的故障。
在一种可能的设计中，若策略控制网元不与待检测网元直接通信，而是通过故障预测网元与待检测网元进行通信，则故障预测网元获取预先记录的待检测网元在过去一个设定时长内的第一日志数据时，可以直接向待检测网元请求第一日志数据，然后接收待检测网元发送的第一日志数据。或者，若故障预测网元不与待检测网元直接通信，而是通过策略控制网元与待检测网元进行通信，则故障预测网元获取预先记录的待检测网元在过去一个设定时长内的第一日志数据时，可以向策略控制网元请求第一日志数据，然后将接收策略控制网元发送的第一日志数据。
在一种可能的设计中，在所述故障预测网元基于所述第一日志数据预测所述待检测网元在预设时长后的第二日志数据之前，所述故障预测网元还可以接收所述策略控制网元发送的故障分析指示消息，所述故障分析指示消息携带所述预设时长，或者，所述故障分析指示消息携带所述预设时长以及如下信息中的至少一项：功能标识、预测时间、预测位置。其中，所述功能标识用于标识故障预测功能，所述预测时间用于指示所述故障预测网元对所述待检测网元进行故障预测的时间，所述预测位置用于指示所述故障预测网元对所述待检测网元进行故障预测的位置。上述设计中，若故障预测网元不与待检测网元直接通信，而是通过策略控制网元与待检测网元进行通信，故障预测网元可以在策略控制网元的指示下对待检测网元进行故障预测。
在一种可能的设计中，若所述故障预测网元在故障关系库中查找不到所述日志特征所对应的故障类型时，可以将确定的日志特征输出，并接收输出的日志特征对应的故障类型，将确定的所述日志特征和接收的所述故障类型学习到所述故障关系库中。上述设计中，故障关系库通过不断学习新的故障类型以及日志特征，可以提高故障预测的准确性。
在一种可能的设计中，所述故障预测网元还可以接收所述策略控制网元发送的更新指示消息，所述更新指示消息用于指示所述故障预测网元更新所述故障关系库，并更新所述故障关系库。上述设计中，更新指示消息可以指示更新操作，如更新故障关系库、提取新 故障类型等等，还可以指示更新要求，如提取新故障时的判断准则、更新故障关系库时采用的参数等等，故障预测网元可以在策略控制网元的指示下更新故障关系库。
在一种可能的设计中，故障预测网元在与策略控制网元建立连接之后，可以接收策略控制网元发送的故障知识，该故障知识包括记录的历史日志数据，以及每个日志数据对应的日志特征，以及每个日志数据对应的故障类型。上述设计中，故障预测网元可以基于策略控制网元发送的故障知识进行初始化，从而即使故障预测网元在初次预测时可以得到较为准确的预测结果。
在一种可能的设计中，若故障检测网元确定自身出现网络故障，导致无法向策略控制网元上报故障类型时，该故障检测网元可以广播自身的故障类型，从而终端设备在接收到该出现网络故障的故障检测网元广播的故障类型后，记录该故障类型并上报给正常工作的故障检测网元，以使正常工作的故障检测网元将终端设备上报的故障类型发送给策略控制网元，使得策略控制网元确定该故障类型对应的恢复措施，并指示出现网络故障的故障检测网元通过该恢复措施处理出现的网络故障。
在一种可能的设计中，若故障检测网元确定自身出现网络故障，导致无法向策略控制网元上报故障类型时，该故障检测网元可以广播自身的故障类型，从而正常工作的故障检测网元在接收到该出现网络故障的故障检测网元广播的故障类型后上报给策略控制网元，使得策略控制网元确定该故障类型对应的恢复措施，并指示出现网络故障的故障检测网元通过该恢复措施处理出现的网络故障。
第四方面，本申请提供了一种网络故障预测方法，该方法包括：策略控制网元接收故障预测网元发送的故障类型，所述故障类型是所述故障预测网元根据所述待检测网元在过去一个设定时长内的日志数据确定的。所述策略控制网元根据接收到的故障类型确定对应的预防措施，并向待检测网元发送所述预防措施，以指示所述待检测网元通过所述预防措施进行处理以避免预设时长后可能出现的故障。本申请实施例中通过定义5G网络中可以实现故障预测功能，并且提供一个由策略控制网元以及故障预测网元构成的网络故障分析架构，基于该架构，策略控制网元可以根据故障预测网元发送的故障类型确定预防措施，从而待检测网元可以通过执行该恢复措施进行处理以避免预设时长后可能出现的故障。
在一种可能的设计中，所述策略控制网元向所述待检测网元发送所述预防措施时，若故障预测网元不与待检测网元直接通信，而是通过策略控制网元与待检测网元进行通信，则策略控制网元可以直接向所述待检测网元发送所述预防措施。或者，所述策略控制网元向所述待检测网元发送所述恢复措施时，若策略控制网元不与待检测网元直接通信，而是通过故障预测网元与待检测网元进行通信，则策略控制网元可以通过所述故障预测网元向所述待检测网元发送所述恢复措施。
在一种可能的设计中，当故障预测网元不与待检测网元直接通信，而是通过策略控制网元与待检测网元进行通信时，在策略控制网元接收故障预测网元发送的故障类型之前，所述策略控制网元还可以接收所述故障预测网元发送的日志数据请求，所述日志数据请求用于请求所述待检测网元预先记录的在过去一个设定时长内的日志数据，并所述策略控制网元向所述故障预测网元发送所述待检测网元的日志数据，从而故障预测网元获取到待检测网元的日志数据。
在一种可能的设计中，若故障预测网元不与待检测网元直接通信，而是通过策略控制网元与待检测网元进行通信，则所述策略控制网元在接收故障预测网元发送的故障类型之 前，还可以向所述故障预测网元发送故障分析指示消息，所述故障分析指示消息携带所述预设时长，或者，所述故障分析指示消息携带所述预设时长以及如下信息中的至少一项：功能标识、预测时间、预测位置。其中，所述功能标识用于标识故障预测功能，所述预测时间用于指示所述故障预测网元对所述待检测网元进行故障预测的时间，所述预测位置用于指示所述故障预测网元对所述待检测网元进行故障预测的位置。上述设计中，若故障预测网元不与待检测网元直接通信，而是通过策略控制网元与待检测网元进行通信，故障预测网元可以在策略控制网元的指示下对待检测网元进行故障预测。
在一种可能的设计中，所述策略控制网元还可以向所述故障预测网元发送更新指示消息，所述更新指示消息用于指示所述故障预测网元更新所述故障关系库。上述设计中，更新指示消息可以指示更新操作，如更新故障关系库、提取新故障类型等等，还可以指示更新要求，如提取新故障时的判断准则、更新时采用的参数等等，策略控制网元可以通过发送该更新指示消息的方式指示故障预测网元更新故障关系库。
在一种可能的设计中，策略控制网元在与故障预测网元建立连接之后，可以向故障预测网元发送故障知识，该故障知识包括记录的历史日志数据，以及每个日志数据对应的日志特征，以及每个日志数据对应的故障类型。上述设计中，故障预测网元可以基于策略控制网元发送的故障知识进行初始化，从而即使故障预测网元在初次预测时可以得到较为准确的预测结果。
第五方面，本申请提供了一种故障检测网元，包括：第一通信接口，用于所述故障检测网元与策略控制网元之间收发消息。存储器，用于存储计算机程序。处理模块，用于调用所述存储器存储的计算机程序，执行：通过所述第一通信接口或第二通信接口接收预先记录的待检测网元在过去一个设定时长内的日志数据，所述第二通信接口用于所述故障检测网元与所述待检测网元之间收发消息。以及，根据获取的所述日志数据，确定对应的日志特征，并在故障关系库中查找所述日志特征所对应的故障类型，其中所述故障关系库包括有不同故障类型与日志特征的关联关系。以及，在查找到所述日志特征所对应的故障类型时，通过所述第一通信接口向所述策略控制网元发送所述故障类型，以使所述策略控制网元根据所述故障类型确定对应的恢复措施，并指示所述待检测网元通过所述恢复措施处理出现的故障。故障关系库可以存储在故障检测网元包括的存储器中，也可以存储在其他存储区域，如云端存储空间等等。
在一种可能的设计中，所述处理模块在用于获取预先记录的待检测网元在过去一个设定时长内的日志数据时，可以具体用于：通过所述第二通信接口向所述待检测网元发送日志数据请求，并通过所述第二通信接口接收所述待检测网元发送的所述日志数据，所述日志数据请求用于请求所述日志数据。或者，通过所述第一通信接口向所述策略控制网元发送所述日志数据请求，并通过所述第一通信接口接收所述策略控制网元发送的所述日志数据。
在一种可能的设计中，所述处理模块，还可以用于在所述故障关系库中查找所述日志特征所对应的故障类型之前，通过所述第一通信接口接收所述策略控制网元发送的故障分析指示消息，所述故障分析指示消息携带如下信息中的至少一项：功能标识、检测时间、检测位置。其中，所述功能标识用于标识故障检测功能，所述检测时间用于指示所述故障检测网元对所述待检测网元进行故障检测的时间，所述检测位置用于指示所述故障检测网元对所述待检测网元进行故障检测的位置。
在一种可能的设计中，所述故障检测网元还可以包括输入输出模块。所述处理模块，还用于在故障关系库中查找不到所述日志特征所对应的故障类型时，将确定的日志特征通过所述输入输出模块输出。并接收从所述输入输出模块输入的根据输出的所述日志特征确定的对应故障类型。以及将所述输入输出模块输出的所述日志特征和从所述输入输出模块接收的对应故障类型学习到所述故障关系库中。
在一种可能的设计中，处理模块，还可以用于通过所述第一通信接口接收所述策略控制网元发送的更新指示消息，所述更新指示消息用于指示更新所述故障关系库；并根据所述更新指示消息，更新所述故障关系库。
第六方面，本申请提供了一种策略控制网元，包括：第一通信接口，用于所述策略控制网元与故障检测网元之间收发消息。存储器，用于存储计算机程序。处理模块，用于调用所述存储器存储的计算机程序，执行：通过所述第一通信接口接收故障检测网元发送的故障类型，所述故障类型是所述故障检测网元在待检测网元故障时，根据所述待检测网元在过去一个设定时长内的日志数据确定的。以及，根据所述第一通信接口接收到的所述故障类型确定对应的恢复措施。以及，通过所述第一通信接口或第二通信接口向所述待检测网元发送所述恢复措施，以指示所述待检测网元通过所述恢复措施处理出现的故障；所述第二通信接口用于所述故障检测网元与所述待检测网元之间收发消息。
在一种可能的设计中，所述第一通信接口，还可以用于接收故障检测网元发送的故障类型之前，接收所述故障检测网元发送的日志数据请求，所述日志数据请求用于请求所述待检测网元在过去一个设定时长内的日志数据。以及，向所述故障检测网元发送所述待检测网元的日志数据。
在一种可能的设计中，所述第一通信接口，还可以用于在接收故障检测网元发送的故障类型之前，向所述故障检测网元发送故障分析指示消息，所述故障分析指示消息携带如下信息中的至少一项：功能标识、检测时间、检测位置。其中，所述功能标识用于标识故障检测功能，所述检测时间用于指示所述故障检测网元对所述待检测网元进行故障检测的时间，所述检测位置用于指示所述故障检测网元对所述待检测网元进行故障检测的位置。
在一种可能的设计中，所述第一通信接口，还用于向所述故障检测网元发送更新指示消息，所述更新指示消息用于指示所述故障检测网元更新所述故障关系库。
第七方面，本申请提供了一种故障预测网元，包括：第一通信接口，用于所述故障预测网元与策略控制网元之间收发消息。存储器，用于存储计算机程序。处理模块，用于调用所述存储器存储的计算机程序，执行：通过所述第一通信接口或第二通信接口接收预先记录的待检测网元在过去一个设定时长内的第一日志数据；所述第二通信接口用于所述故障检测网元与所述待检测网元之间收发消息。以及，基于所述第一日志数据预测所述待检测网元在预设时长后的第二日志数据；根据所述第二日志数据，确定对应的日志特征，并在故障关系库中查找所述日志特征所对应的故障类型，所述故障关系库包括不同故障类型与日志特征的关联关系。以及，通过所述第一通信接口向所述策略控制网元发送所述故障类型，以使所述策略控制网元根据所述故障类型确定对应的预防措施，并指示所述待检测网元通过所述预防措施进行处理以避免预设时长后可能出现的故障。故障关系库可以存储在故障检测网元包括的存储器中，也可以存储在其他存储区域，如云端存储空间等等。
在一种可能的设计中，所述处理模块在用于获取预先记录的待检测网元在过去一个设定时长内的日志数据时，可以具体用于：通过所述第二通信接口向所述待检测网元发送日 志数据请求，并通过所述第二通信接口接收所述待检测网元发送的所述第一日志数据，所述日志数据请求用于请求所述第一日志数据。或者，通过所述第一通信接口向所述策略控制网元发送所述日志数据请求，并通过所述第一通信接口接收所述策略控制网元发送的所述第一日志数据。
在一种可能的设计中，所述处理模块，还用于在基于所述第一日志数据预测所述待检测网元在预设时长后的第二日志数据之前，通过所述第一通信接口接收所述策略控制网元发送的故障分析指示消息，所述故障分析指示消息携带所述预设时长，或者，所述故障分析指示消息携带所述预设时长以及如下信息中的至少一项：功能标识、预测时间、预测位置。其中，所述功能标识用于标识故障预测功能，所述预测时间用于指示所述故障预测网元对所述待检测网元进行故障预测的时间，所述预测位置用于指示所述故障预测网元对所述待检测网元进行故障预测的位置。
在一种可能的设计中，所述故障预测网元还可以包括输入输出模块。所述处理模块，还用于在故障关系库中查找不到所述日志特征所对应的故障类型时，将确定的日志特征通过所述输入输出模块输出。以及，接收从所述输入输出模块输入的根据输出的所述日志特征确定的对应故障类型。以及，将所述输入输出模块输出的所述日志特征和从所述输入输出模块接收的对应故障类型学习到所述故障关系库中。
在一种可能的设计中，处理模块，还用于通过所述第一通信接口接收所述策略控制网元发送的更新指示消息，所述更新指示消息用于指示更新所述故障关系库。并根据所述更新指示消息，更新所述故障关系库。
第八方面，本申请提供了一种策略控制网元，包括：第一通信接口，用于所述策略控制网元与故障预测网元之间收发消息。存储器，用于存储计算机程序。处理模块，用于调用所述存储器存储的计算机程序，执行：通过所述第一通信接口接收故障预测网元发送的故障类型，所述故障类型是所述故障预测网元根据所述待检测网元在过去一个设定时长内的日志数据确定的。并根据所述第一通信接口接收到的故障类型确定对应的预防措施。以及，通过所述第一通信接口或第二通信接口向所述待检测网元发送所述恢复措施，以指示所述待检测网元通过所述恢复措施处理出现的故障，所述第二通信接口用于所述故障预测网元与所述待检测网元之间收发消息。
在一种可能的设计中，所述第一通信接口，还可以用于接收故障预测网元发送的故障类型之前，接收所述故障预测网元发送的日志数据请求，所述日志数据请求用于请求所述待检测网元在过去一个设定时长内的第一日志数据。并向所述故障预测网元发送所述待检测网元的第一日志数据。
在一种可能的设计中，所述第一通信接口，还可以用于在接收故障预测网元发送的故障类型之前，向所述故障预测网元发送故障分析指示消息，所述故障分析指示消息携带所述预设时长，或者，所述故障分析指示消息携带所述预设时长以及如下信息中的至少一项：功能标识、预测时间、预测位置。其中，所述功能标识用于标识故障预测功能，所述预测时间用于指示所述故障预测网元对所述待检测网元进行故障预测的时间，所述预测位置用于指示所述故障预测网元对所述待检测网元进行故障预测的位置。
在一种可能的设计中，所述第一通信接口，还可以用于向所述故障预测网元发送更新指示消息，所述更新指示消息用于指示所述故障预测网元更新所述故障关系库。
第九方面，本申请还提供了一种计算机可读存储介质，用于存储为执行上述第一方面 至第四方面中任一方面或任一方面的任意一种设计的功能所用的计算机软件指令，其包含用于执行上述第一方面至第四方面中任一方面或任一方面的任意一种设计的方法所设计的程序。
第十方面，本申请实施例提供了一种包含指令的计算机程序产品，当其在计算机上运行时，使得计算机执行上述第一方面至第四方面中任一方面或任一方面的任意一种设计所述的方法。
本申请实施例第三方面、第四方面所提供的网络故障预测方法可以作为一个独立方案单独使用，也可以在本申请实施例第一方面、第二方面所提供的网络故障检测方法的基础上进一步使用。
应理解，本申请实施例的第五至八方面分别与本申请实施例的第一至四方面的技术方案一致，各方面及对应的可实施的设计方式所取得的有益效果相似，不再赘述。
附图说明
图1为本申请提供的一种5G通信系统的架构示意图；
图2为本申请提供的一种网络故障分析系统的结构示意图；
图3为本申请提供的一种网络故障检测方法的流程示意图；
图4为本申请提供的一种更新故障关系库的流程示意图；
图5为本申请提供的一种通信方式示意图；
图6为本申请提供的一种网络故障检测方法的流程示意图；
图7为本申请提供的一种故障检测网元初始化过程的示意图；
图8为本申请提供的另一种通信方式示意图；
图9为本申请提供的另一种网络故障检测方法的流程示意图；
图10为本申请提供的一种故障检测网元处理自身故障的流程示意图；
图11为本申请提供的另一种故障检测网元处理自身故障的流程示意图；
图12为本申请提供的一种网络故障预测方法的流程示意图；
图13为本申请提供的一种通信方式示意图；
图14为本申请提供的一种网络故障预测方法的流程示意图；
图15为本申请提供的另一种通信方式示意图；
图16为本申请提供的另一种网络故障预测方法的流程示意图；
图17为本申请提供的一种故障检测网元的结构示意图；
图18为本申请提供的一种故障预测网元的结构示意图；
图19为本申请提供的一种策略控制网元的结构示意图；
图20为本申请提供的一种网络故障分析网元的结构示意图；
图21为本申请提供的一种策略控制网元的结构示意图。
具体实施方式
为了使本申请的目的、技术方案和优点更加清楚，下面将结合附图对本申请作进一步地详细描述。
随着网络的演进，网络管理的趋势也在随之变化，朝着综合化、分布式、智能化方向 发展，通过各种网络功能，使其能通过网络分析实现网络资源的计划、建设与优化。在5G标准中明确提出了网络功能(network function，NF)服务的概念功能，即将传统网元转换为NF，然后NF再被分解为多个NF服务。NF服务之间通过通信接口进行通信，且每个NF服务独立于其他NF服务进行部署、升级、扩展。
5G通信系统的构架为基于服务的架构(service based architecture，SBA)。5G通信系统的架构中可以但不限于包括网络开放功能网元、策略控制功能网元、数据管理网元、应用功能网元、核心网接入和移动性管理功能网元、会话管理功能网元、用户面功能网元等等。下面对各个功能网元进行描述：
所述会话管理功能网元，可用于负责所述终端设备的会话管理(包括会话的建立、修改和释放)，用户面功能网元的选择和重选、所述终端设备的互联网协议(internet protocol，IP)地址分配、服务质量(quality of service，QoS)控制等。
所述策略控制功能网元，可用于负责策略控制决策、提供基于业务数据流和应用检测、门控、QoS和基于流的计费控制等功能等。例如，在5G中，所述策略控制功能网元可以是PCF(policy control function)网元。在未来通信，如6G中，所述策略控制功能网元仍可以是PCF网元，或有其它的名称，本申请不做限定。当所述策略控制功能网元是PCF网元，所述PCF网元可以提供Npcf服务。
所述应用功能网元，主要功能是与第三代合作伙伴计划(the 3rd generation partnership project，3GPP)核心网通信来提供服务，来影响业务流路由、接入网能力开放、策略控制等。例如，在5G中，所述应用功能网元可以是AF(application function)网元。
所述数据管理网元，可用于管理所述终端设备的签约数据、与所述终端设备相关的注册信息等。
所述网络开放功能网元，可用于使3GPP能够安全地向第三方服务器提供网络业务能力。
所述核心网接入和移动性管理功能网元，可用于对所述终端设备的接入控制和移动性进行管理。
所述用户面功能网元，主要提供用户平面的业务处理功能，包括业务路由、包转发、锚定功能、QoS映射和执行、上行链路的标识识别并路由到数据网络、下行包缓存和下行链路数据到达的通知触发、与外部数据网络连接等。
以上各个网元也可以称为功能实体，既可以是在专用硬件上实现的网络元件，也可以是在专用硬件上运行的软件实例，或者是在适当平台上虚拟化功能的实例，例如，上述虚拟化平台可以为云平台。
图1示出了5G通信系统架构的一种可能的示例，具体包括：应用功能(application function，AF)、网络曝光功能(network exposure function，NEF)、策略控制功能(Policy Control function，PCF)、核心接入和移动性管理功能(Core Access and Mobility Management Function，AMF)、会话管理功能(session management function，SMF)、用户面功能(user plane function，UPF)。其中，AF与PCF之间可以通过N5接口相连，AMF与PCF之间可以通过N15接口相连，SMF与PCF之间可以通过N7接口相连，SMF与UPF之间可以通过N4接口相连，SMF与NEF之间可以通过PGw接口相连。接口名称只是一个示例说明，本申请实施例对此不作具体限定。应理解，图1中所示的网元的名称在这里仅作为一种示例说明，并不作为对本申请的方法适用的通信系统架构中包括的网元的限定。图1所示的 通信系统的架构中不限于仅包含图中所示的网元，还可以包含其它未在图中表示的设备，具体本申请在此处不再一一列举。本申请实施例并不限定5G通信系统中各个网元的分布形式，图1所示的分布形式只是示例性的，本申请不作限定。
本申请实施例基于5G通信系统的架构提供一种网络故障分析系统，该系统包括策略控制网元，还可以包括故障检测网元和/或故障预测网元，其中，故障检测网元与策略控制网元通过通信接口相连，故障预测网元与策略控制网元通过通信接口相连，如图2所示。
故障检测网元，可用于实现对网元设备进行故障检测。在5G系统或者未来通信系统中可能将实现对网元设备进行故障检测的功能网元命名为其他的名称，如AA。应理解，若该AA也可以实现本申请实施例中的故障检测网元所实现的功能，也可以将AA理解为本申请实施例中的故障检测网元。
故障预测网元，可用于实现对网元设备进行故障预测。在5G系统或者未来通信系统中可能将实现对网元设备进行故障预测的功能网元命名为其他的名称，如BB。应理解，若该BB也可以实现本申请实施例中的故障预测网元所实现的功能，也可以将BB理解为本申请实施例中的故障预测网元。
策略控制网元，用于根据故障检测网元检测得到的故障类型确定恢复措施，以使网元设备通过该恢复措施处理出现的故障，和/或，用于根据故障预测网元预测得到的故障类型确定预防措施，以使网元设备通过该预防措施预防未来可能出现的故障。
需要说明的是，图2所示的网络故障分析系统还可以包括图1所示的5G系统中的全部或部分网元，也可以包括包含其它未在图1中表示的网元，具体本申请在此处不再一一列举。
需要说明的是，本申请实施例并不限定网络故障分析系统中各个网元的分布形式，图2所示的分布形式只是示例性的，本申请不作限定。
为方便说明，本申请后续均以图2所示的网元为例进行说明，应理解，本申请中所有网元的名称仅仅作为示例，在未来通信中还可以称为其它名称，或者在未来通信中本申请涉及的网元还可以通过其它具有相同功能的实体或者设备等来替代，本申请对此均不作限定。这里做统一说明，后续不再赘述。
需要说明的是，图2所示的网络故障分析系统并不构成本申请实施例能够适用的通信系统的限定。图2所示的通信系统基于5G系统架构，可选的，本申请实施例的方法还适用于未来的各种通信系统等。
本申请中所涉及的多个，是指两个或两个以上。
另外，需要理解的是，在本申请的描述中，“第一”、“第二”等词汇，仅用于区分描述的目的，而不能理解为指示或暗示相对重要性，也不能理解为指示或暗示顺序。
下面结合附图对本申请实施例所提供的方案作具体说明。
本申请实施例基于图2所示的网络故障分析系统提供一种网络故障检测方法、以及一种网络故障预测方法，网络故障检测方法以及网络故障预测方法可以作为一个独立方案单独使用，也可以结合使用，如在网络故障检测方法的基础上进一步使用网络故障预测方法，或者，在网络故障预测方法的基础上进一步使用网络故障检测方法。
本申请实施例提供的一种网络故障分析方法，参阅图3所示，网络故障检测方法的具体流程包括：
S301，故障检测网元获取待检测网元预先记录的在过去一个设定时长内的日志数据。
S302，故障检测网元根据获取的日志数据，确定对应的日志特征，并在故障关系库中查找所述日志特征所对应的故障类型，所述故障关系库包括不同故障类型与日志特征的关联关系。
在一种可能的实现方式中，故障检测网元可以将获取的日志数据输入预先训练好的神经网络模型中，得到该日志数据对应的故障类型。该神经网络模型可以用于提取日志数据对应的日志特征，并在故障关系库中查找所述日志特征所对应的故障类型。
S303，故障检测网元在查找到所述日志特征所对应的故障类型时，向策略控制网元发送查找到的故障类型。
示例性的，故障检测网元可以向策略控制网元发送故障分析结果信元，该故障分析结果信元可以但不限于包括功能标识、分析时间、分析位置、故障状态、故障具体信息等等。其中，所述功能标识用于标识故障检测功能，所述检测时间用于指示所述故障检测网元对所述待检测网元进行故障检测的时间。所述检测位置用于指示所述故障检测网元对所述待检测网元进行故障检测的位置。故障状态用于标识待检测网元是否出现故障。故障具体信息包括待检测网元出现的故障类型、和/或该故障类型是否为已知故障类型、和/或该故障类型为软件故障还是硬件故障等等。故障分析结果信元还可以包括建议的恢复措施等等。
应理解，本申请中所有信元、信令的名称仅仅作为示例，在未来通信中还可以称为其它名称，或者在未来通信中本申请涉及的信元、信令还可以通过其它具有相同功能的信息、消息、信元或信令等来替代，本申请对此均不作限定。这里做统一说明，后续不再赘述。
S304，策略控制网元根据接收到的故障类型确定对应的恢复措施。
S305，策略控制网元指示所述待检测网元通过所述恢复措施处理出现的故障。
本申请实施例中通过提供一个由策略控制网元以及故障检测网元构成的网络故障分析架构，并且基于该架构提供一种网络故障检测方法，具体的，故障检测网元可以根据获取的日志数据确定故障类型，并向策略控制网元发送检测的故障类型，策略控制网元可以根据接收到的故障类型确定恢复措施，而待检测网元可以通过执行该恢复措施来处理出现的故障。通过本申请实施例提供的网络故障检测方法，可以实现基于NF服务实现对待检测网元进行故障检测。
在本申请实施例中，故障检测网元获取的日志数据可以是结构化的数据比如ALM，也可以是非结构化的数据如不同厂商生产的设备所产生的日志数据等等，本申请实施例在这里不对日志数据的结构进行限定。由于故障检测网元对日志数据的格式没有限制，因此故障检测网元可以对不同厂商生产的网元设备进行故障检测，从而可以提高网络故障检测方法的普适性。
在步骤S302中故障检测网元在故障关系库中查找所述日志特征所对应的故障类型时，若故障检测网元在故障关系库中查找不到所述日志特征所对应的故障类型，还可以进一步将确定的日志特征输出，比如输出给网管工程人员查看，以使网管工程人员人为根据经验分析对应的故障类型，并将分析出的故障类型再输入给故障检测网元，比如通过显示界面的方式输入和输出，也可以通过打印的方式输入和输出，或者通过键盘等输入设备进行输入，显示器等输出设备进行输出。故障检测网元在接收输出的日志特征对应的故障类型后，进而可以将确定的所述日志特征和接收的所述故障类型学习到所述故障关系库中，从而达到不断完善和更新故障关系库的目的，使得故障关系库信息量更为全面。
其中，故障检测网元还可以在接收到策略控制网元发送的的更新指示消息时更新故障 关系库。在本申请实施例中，可以将更新指示消息称为故障知识更新信元。
故障关系库的更新过程可以通过步骤S401至S405实现，参阅图4所示。
S401，策略控制网元向故障检测网元发送请求更新模块信令，该请求更新模块信令用于指示请求故障检测网元更新故障关系库。
S402，故障检测网元向策略控制网元发送更新模块回复信令，该更新模块回复信令用于通知策略控制网元同意更新故障关系库。
S403，策略控制网元向故障检测网元发送故障知识更新信元。故障知识更新信元可以携带更新操作标识、更新要求等等，其中，更新操作标识用于标识更新用于进行故障检测的故障关系库，或者，用于标识提取新的故障类型，或者，用于标识更新用于进行故障预测的故障关系库等等。更新要求可以为提取新的故障类型时的判定准则、更新故障关系库时采用的参数等等。
S404，故障检测网元基于接收到的故障知识更新信元更新故障关系库。
当然，上述步骤S401和S402也可以不必执行，直接从S403开始执行。
在执行完步骤S404之后，还可以执行S405，故障检测网元向策略控制网元发送故障知识信元，故障知识信元携带故障检测网元积累的日志数据，以及每个日志数据对应的日志特征，以及每个日志数据对应的故障类型，每个日志数据对应的功能标识，该功能标识用于标识该日志数据是在进行故障检测时所获取的，或者用于标识该日志数据是在进行故障检预测时所获取的。
故障关系库通过不断学习新的故障类型以及日志特征，可以提高故障检测的准确性，并且可以提高网络故障检测方法的可扩展性，使得当日志数据的数据量较大时，可以保持较好的性能，当故障种类增多时，可以轻松地扩展故障关系库。
在一种可能的实施方式中，故障检测网元不与待检测网元直接通信，而是通过与策略控制网元与待检测网元进行通信，如图5所示。结合图5所示的通信架构对网络故障检测方法的具体流程进行描述，如图6所示。
在故障检测网元对待检测网元进行故障检测之前，可以执行步骤S601与步骤S602。
S601，故障检测网元与策略控制网元建立连接。具体的，策略控制网元可以向故障检测网元发送状态信令，所述状态信令用于查询故障检测网元是否处于空闲状态。故障检测网元在处于空闲状态时向策略控制网元回复当前状态为空闲，则故障检测网元与策略控制网元建立连接。故障检测网元在处于忙碌状态时向策略控制网元回复当前状态为忙碌。策略控制网元在故障检测网元回复当前状态为忙碌时，每间隔第一预设时长向故障检测网元发送状态信令，直到故障检测网元回复当前状态为空闲。第一预设时长可以为5s、3s、12s等等，也可以为其他时间段，本申请实施例在这里不对第一预设时长进行具体限定。
S602，策略控制网元与待检测网元建立连接。具体的，策略控制网元可以向待检测网元发送状态信令。待检测网元在处于空闲状态时向策略控制网元回复当前状态为空闲，则待检测网元与策略控制网元建立连接。待检测网元在处于忙碌状态时向策略控制网元回复当前状态为忙碌。策略控制网元在待检测网元回复当前状态为忙碌时，每间隔第二预设时长向待检测网元发送状态信令，直到待检测网元回复当前状态为空闲。第二预设时长可以为5s、3s、12s等等，也可以为其他时间段，第二预设时长可以与第一预设时长相同，第二预设时长可以与第一预设时长不相同，本申请实施例在这里不对第二预设时长进行具体限定。
其中，步骤S601与步骤S602没有严格的先后执行顺序，可以先执行步骤S601，后执行步骤S602，也可以先执行步骤S602，后执行步骤S601，也可以同时执行步骤S601与步骤S602，本申请实施例在这里不做具体限定。
在策略控制网元与故障检测网元建立连接之后，故障检测网元可以在策略控制网元的指示下进行初始化，故障检测网元的初始化过程具体可以通过步骤S701至S705实现，参阅图7所示：
S701，策略控制网元向故障检测网元发送请求初始化模块信令，该请求初始化模块信令用于请求故障检测网元进行初始化。
S702，故障检测网元向策略控制网元发送初始化模块回复信令，该初始化模块回复信令用于通知策略控制网元同意进行初始化。
S703，策略控制网元向故障检测网元发送故障知识信元。
S704，故障检测网元基于接收到的故障知识信元进行初始化。
S705，故障检测网元向策略控制网元发送模块初始化完成信令，该模块初始化完成信令用于向策略控制网元通知故障检测网元完成初始化。
由于策略控制网元可以与多个故障检测网元通信，因此策略控制网元发送的故障知识可能包括多个故障检测网元积累的故障知识，因此故障检测网元通过基于策略控制网元发送的故障知识进行初始化，可以基于多个故障检测网元积累的故障知识对待检测网元进行故障检测，从而可以提高故障检测网元在对待检测网元进行故障检测时的准确性。
在策略控制网元分别与待检测网元、故障检测网元建立连接之后，可以执行步骤S603至S605。
S603，策略控制网元向故障检测网元发送故障检测信令，该故障检测信令用于指示故障检测网元对待检测网元进行故障检测。
S604，故障检测网元向策略控制网元发送故障检测回复信令，该故障检测回复信令用于指示策略控制网元发送故障分析信息信元。
S605，策略控制网元向故障检测网元发送故障分析指示消息，在本申请实施例中，可以将故障分析指示消息称为故障分析信息信元。所述故障分析信息信元携带可以如下信息中的至少一项：功能标识、检测时间、检测位置。
当然，上述步骤S603和S604也可以不必执行，在策略控制网元分别与待检测网元、故障检测网元建立连接之后，直接从S605开始执行。
在一种可能的实施方式中，步骤S301，故障检测网元在待检测网元故障时，获取待检测网元预先记录的在过去一个设定时长内的日志数据，具体可以通过步骤S606至S609实现。
S606，策略控制网元向待检测网元发送日志数据请求，在本申请实施例中，可以将日志数据请求称为日志信息信元，该日志信息信元用于请求待检测网元预先记录的在过去一个设定时长内的日志数据，且该日志信息信元可以携带策略控制网元对日志数据的要求，如设定时长、日志数据的大小、日式数据的发送频率等等。
S607，待检测网元将预先记录的日志数据根据日志信息信元发送给策略控制网元。
S608，故障检测网元向策略控制网元发送日志信息信元，该日志信息信元可以携带故障检测网元对日志数据的要求，如设定时长、日志数据的大小、日式数据的发送频率等等。
S609，策略控制网元根据接收到的日志信息信元将待检测网元发送的日志数据发送给 故障检测网元。
示例性的，步骤S305，策略控制网元指示所述待检测网元通过所述恢复措施处理出现的故障，具体可以通过步骤S610实现。
S610，策略控制网元向待检测网元发送故障恢复信元，该故障恢复信元可以携带功能标识以及恢复措施标识。其中，所述功能标识用于标识故障检测功能，恢复措施标识用于标识策略控制网元确定的恢复措施。
在执行完步骤S610之后，还可以执行步骤S611，待检测网元向策略控制网元发送故障恢复结果信元，该故障恢复结果信元可以携带功能标识以及恢复结果标识。其中，所述功能标识用于标识故障检测功能。当待检测网元通过策略控制网元确定的恢复措施成功处理出现的故障时，恢复结果标识用于标识恢复成功。当待检测网元通过策略控制网元确定的恢复措施未能成功处理出现的故障时，恢复结果标识用于标识恢复失败。
在执行完步骤S305之后，可以执行步骤S612和步骤S613。
S612，策略控制网元与待检测网元断开连接。
S613，策略控制网元与故障检测网元断开连接。
其中，步骤S612与步骤S613没有严格的先后执行顺序，可以先执行步骤S612，后执行步骤S613，也可以先执行步骤S613，后执行步骤S612，也可以同时执行步骤S612与步骤S613，本申请实施例在这里不做具体限定。
在另一种可能的实施方式中，策略控制网元不与待检测网元直接通信，而是通过与故障检测网元与待检测网元进行通信，如图8所示。结合图8所示的通信架构对网络故障检测方法的具体流程进行描述，如图9所示。
在故障检测网元对待检测网元进行故障检测之前，可以执行步骤S901。
S901，故障检测网元与待检测网元建立连接。具体的，待检测网元可以向故障检测网元发送状态信令，所述状态信令用于查询故障检测网元是否处于空闲状态。故障检测网元在处于空闲状态时向待检测网元回复当前状态为空闲，则故障检测网元与待检测网元建立连接。故障检测网元在处于忙碌状态时向待检测网元回复当前状态为忙碌。待检测网元在故障检测网元回复当前状态为忙碌时，每间隔第三预设时长向故障检测网元发送状态信令，直到故障检测网元回复当前状态为空闲。第三预设时长可以为5s、3s、12s等等，也可以为其他时间段，本申请实施例在这里不对第三预设时长进行具体限定。
在执行步骤S301，故障检测网元在待检测网元故障时，获取待检测网元预先记录的在过去一个设定时长内的日志数据之前，可以先执行步骤S902。
S902，待检测网元向故障检测网元发送请求故障检测信令，该请求故障检测信令用于请求故障检测网元对待检测网元进行故障检测。
在一种可能的实施方式中，步骤S301，故障检测网元在待检测网元故障时，获取待检测网元预先记录的在过去一个设定时长内的日志数据，具体可以通过步骤S903和S904实现。
S903，故障检测网元向待检测网元发送日志信息信元，且该日志信息信元可以携带故障检测网元对日志数据的要求，如设定时长、日志数据的大小、日式数据的发送频率等等。
S904，待检测网元根据接收到日志信息信元将预先记录的日志数据发送给策略控制网元。
可选的，在执行步骤S303，故障检测网元向策略控制网元发送查找到的故障类型之前， 可以先执行步骤S905。
S905，故障检测网元与策略控制网元建立连接。具体的，故障检测网元可以向策略控制网元发送状态信令。策略控制网元在处于空闲状态时向故障检测网元回复当前状态为空闲，则故障检测网元与策略控制网元建立连接。策略控制网元在处于忙碌状态时向故障检测网元回复当前状态为忙碌。故障检测网元在策略控制网元回复当前状态为忙碌时，每间隔第四预设时长向策略控制网元发送状态信令，直到策略控制网元回复当前状态为空闲。第四预设时长可以为5s、3s、12s等等，也可以为其他时间段，第四预设时长可以与第三预设时长相同，第四预设时长可以与第三预设时长不相同，本申请实施例在这里不对第四预设时长进行具体限定。
在策略控制网元与故障检测网元建立连接之后，故障检测网元可以在策略控制网元的指示下进行初始化，故障检测网元的初始化过程具体可以参阅图7所示，本申请实施例在这里不再重复赘述。
示例性的，步骤S305，策略控制网元指示所述待检测网元通过所述恢复措施处理出现的故障，具体可以通过步骤S906和步骤S907实现。
S906，策略控制网元向故障检测网元发送故障恢复信元。
S907，故障检测网元向待检测网元转发该故障恢复信元。
在执行完步骤S907之后，还可以执行步骤S908和步骤S909。
S908，待检测网元向故障检测网元发送故障恢复结果信元。
S909，故障检测网元向策略控制网元转发故障恢复结果信元。
在执行完步骤S909之后，可以执行步骤S910和步骤S911。
S910，故障检测网元与待检测网元断开连接。
S911，故障检测网元与策略控制网元断开连接。
其中，步骤S910和步骤S911没有严格的先后执行顺序，可以先执行步骤S910，后执行步骤S911，也可以先执行步骤S911，后执行步骤S910，也可以同时执行步骤S910和步骤S911，本申请实施例在这里不做具体限定。
若故障检测网元自身出现故障导致无法向策略控制网元发送故障类型时，故障检测网元可以通过广播自身出现的故障类型，以使策略控制网元根据该故障类型确定对应的恢复措施，并指示故障检测网元通过该恢复措施处理自身出现的故障。
在一种可能的实施方式中，故障检测网元自身出现故障时，可以将自身故障对应的故障类型通过正常工作的故障检测网元发送给策略控制网元，具体可以参阅步骤S1001至S1008，如图10所示。
S1001，故障检测网元广播自身出现的故障所对应的故障类型。为了方面描述，本申请实施例中将出现故障的故障检测网元称为非正常故障检测网元，将正常工作的故障检测网元称为正常故障检测网元。
S1002，正常故障检测网元在接收到非正常故障检测网元广播的故障类型时，将非正常故障检测网元广播的故障类型转发给策略控制网元。
S1003，策略控制网元基于该故障类型确定恢复措施。
S1004，策略控制网元向正常故障检测网元发送故障恢复信令，该故障恢复信令用于指示非正常故障检测网元基于该恢复措施处理自身出现的故障。
S1005，正常故障检测网元将该故障恢复信令转发给非正常故障检测网元。
S1006，非正常故障检测网元基于该故障恢复信令指示的恢复措施处理自身出现的故障。
S1007，非正常故障检测网元广播送故障恢复结果信令，该该故障恢复信令用于通知策略控制网元恢复结果，其中，恢复结果为清除故障成功，或者恢复结果为清除故障失败，或者恢复结果为清除故障风险成功，或者恢复结果为清除故障风险失败等等。
S1008，正常故障检测网元在接收到非正常故障检测网元广播的故障恢复结果信令时，将该故障恢复结果信令转发给策略控制网元。
在另一种可能的实施方式中，故障检测网元自身出现故障时，可以将自身故障对应的故障类型通过终端设备发送给策略控制网元，具体可以参阅S1101至S1111，如图11所示。
S1101，故障检测网元广播自身出现的故障所对应的故障类型。为了方面描述，本申请实施例中将出现故障的故障检测网元称为非正常故障检测网元，将正常工作的故障检测网元称为正常故障检测网元。
S1102，终端设备在接收到非正常故障检测网元广播的故障类型时记录该故障类型，并发送给正常故障检测网元。
S1103，正常故障检测网元将该故障类型转发给策略控制网元。
S1104，策略控制网元基于该故障类型确定恢复措施。
S1105，策略控制网元向正常故障检测网元发送故障恢复信令，该故障恢复信令用于指示非正常故障检测网元基于该恢复措施处理自身出现的故障。
S1106，正常故障检测网元将该故障恢复信令转发给终端设备。
S1107，终端设备将该故障恢复信令转发给非正常故障检测网元。
S1108，非正常故障检测网元基于该故障恢复信令指示的恢复措施处理自身出现的故障。
S1109，非正常故障检测网元广播送故障恢复结果信令。
S1110，终端设备在接收到非正常故障检测网元广播的故障恢复结果信令后，发送给正常故障检测网元。
S1111，正常故障检测网元将该故障恢复结果信令转发给策略控制网元。
本申请实施例提供的一种网络故障预测方法，参阅图12所示，网络故障预测方法的具体流程包括：
S1201，故障预测网元获取待检测网元预先记录的在过去一个设定时长内的日志数据。
S1202，所述故障预测网元基于所述第一日志数据预测所述待检测网元在未来预设时长后的第二日志数据。
在一种可能的实现方式中，故障预测网元可以将获取的第一日志数据输入预先训练好的序列到序列(sequence to sequence，Seq2Seq)模型中，得到第二日志数据。该Seq2Seq模型可以用于根据过去一个时段的日志数据预测未来预设时长后的日志数据。
S1203，故障预测网元根据第二日志数据，确定对应的日志特征，并在故障关系库中查找所述日志特征所对应的故障类型，所述故障关系库包括不同故障类型与日志特征的关联关系。
在一种可能的实现方式中，故障检测网元可以将第二日志数据输入预先训练好的长期短期记忆(long short term memory，LSTM)模型中，得到故障类型。该LSTM模型可以用于提取日志数据对应的日志特征，并在故障关系库中查找所述日志特征所对应的故障类 型。
S1204，故障预测网元向策略控制网元发送查找到的故障类型。
示例性的，故障预测网元可以向策略控制网元发送故障分析结果信元，该故障分析结果信元可以但不限于包括功能标识、分析时间、分析位置、故障状态、故障具体信息、未来预设时长等等。其中，所述功能标识用于标识故障预测功能，所述检测时间用于指示所述故障预测网元对所述待检测网元进行故障预测的时间。所述检测位置用于指示所述故障预测网元对所述待检测网元进行故障预测的位置。故障状态用于标识待检测网元是否出现故障。故障具体信息包括待检测网元可能出现的故障类型、和/或该故障类型是否为已知故障类型、和/或该故障类型为软件故障还是硬件故障等等。故障分析结果信元还可以包括建议的预防措施等等。
S1205，策略控制网元根据接收到的故障类型确定对应的预防措施。
S1206，策略控制网元指示待检测网元通过该预防措施进行处理以避免未来预设时长后可能出现的故障。
本申请实施例中通过提供一个由策略控制网元以及故障预测网元构成的网络故障分析架构，并且基于该架构提供一种网络故障预测方法，具体的，故障预测网元可以根据获取的过去一段时间内的日志数据预测未来一段时间后的日志数据，之后根据预测的日志数据确定故障类型，并向策略控制网元发送检测的故障类型。策略控制网元可以根据接收到的故障类型确定预防措施，而待检测网元可以通过执行该恢复措施进行处理以避免预设时长后可能出现的故障。通过本申请实施例提供的网络故障预测方法，可以实现基于NF服务实现对待检测网元进行故障预测。
在本申请实施例中，故障预测网元获取的日志数据可以是结构化的数据比如ALM，也可以是非结构化的数据如不同厂商生产的设备所产生的日志数据等等，本申请实施例在这里不对日志数据的结构进行限定。由于故障预测网元对日志数据的格式没有限制，因此故障预测网元可以对不同厂商生产的网元设备进行故障预测，从而可以提高网络故障预测方法的普适性。
可选的，在步骤S1203中故障预测网元在故障关系库中查找所述日志特征所对应的故障类型时，若故障预测网元在故障关系库中查找不到所述日志特征所对应的故障类型，可以将确定的日志特征输出，并接收输出的日志特征对应的故障类型，将确定的所述日志特征和接收的所述故障类型学习到所述故障关系库中，从而更新故障关系库。
其中，故障预测网元可以在接收到策略控制网元发送的的更新指示消息时更新故障关系库。在本申请实施例中，可以将更新指示消息称为故障知识更新信元。故障预测网元更新故障关系库的过程与故障检测网元更新故障关系库的过程类似，具体可以参阅图4所示，本申请在这里不再重复赘述。
在一种可能的实施方式中，故障预测网元不与待检测网元直接通信，而是通过与策略控制网元与待检测网元与待检测网元进行通信，如图13所示。结合图13所示的通信架构对网络故障预测方法的具体流程进行描述，如图14所示。
在故障预测网元对待检测网元进行故障预测之前，可以执行步骤S1401与步骤S1402。
S1401，故障预测网元与策略控制网元建立连接。具体的，策略控制网元可以向故障预测网元发送状态信令。故障预测网元在处于空闲状态时向策略控制网元回复当前状态为空闲，则故障预测网元与策略控制网元建立连接。故障预测网元在处于忙碌状态时向策略 控制网元回复当前状态为忙碌。策略控制网元在故障预测网元回复当前状态为忙碌时，每间隔第一预设时长向故障预测网元发送状态信令，直到故障预测网元回复当前状态为空闲。第一预设时长可以为5s、3s、12s等等，也可以为其他时间段，本申请实施例在这里不对第一预设时长进行具体限定。
S1402，策略控制网元与待检测网元建立连接。具体的，策略控制网元可以向待检测网元发送状态信令。待检测网元在处于空闲状态时向策略控制网元回复当前状态为空闲，则待检测网元与策略控制网元建立连接。待检测网元在处于忙碌状态时向策略控制网元回复当前状态为忙碌。策略控制网元在待检测网元回复当前状态为忙碌时，每间隔第二预设时长向待检测网元发送状态信令，直到待检测网元回复当前状态为空闲。第二预设时长可以为5s、3s、12s等等，也可以为其他时间段，第二预设时长可以与第一预设时长相同，第二预设时长可以与第一预设时长不相同，本申请实施例在这里不对第二预设时长进行具体限定。
其中，步骤S1401与步骤S1402没有严格的先后执行顺序，可以先执行步骤S1401，后执行步骤S1402，也可以先执行步骤S1402，后执行步骤S1401，也可以同时执行步骤S1401与步骤S1402，本申请实施例在这里不做具体限定。
在策略控制网元与故障预测网元建立连接之后，故障预测网元可以在策略控制网元的指示下进行初始化，故障预测网元的初始化过程与故障检测网元的初始化过程类似，具体可以参阅图7所示，本申请实施例在这里不再重复赘述。
在策略控制网元分别与待检测网元、故障预测网元建立连接之后，可以执行步骤S1403至S1405。
S1403，策略控制网元向故障预测网元发送故障预测信令，该故障预测信令用于指示故障预测网元对待检测网元进行故障预测。
S1404，故障预测网元向策略控制网元发送故障预测回复信令，该故障预测回复信令用于指示策略控制网元发送故障分析信息信元。
S1405，策略控制网元向故障预测网元发送故障分析信息信元。所述故障分析信息信元携带可以如下信息中的至少一项：未来预设时长、功能标识、检测时间、检测位置。
当然，上述步骤S1403和S1404也可以不必执行，在策略控制网元分别与待检测网元、故障预测网元建立连接之后，直接从S1405开始执行。
步骤S301，故障预测网元在待检测网元故障时，获取待检测网元预先记录的在过去一个设定时长内的日志数据，具体可以通过步骤S1406至S1409。
S1406，策略控制网元向待检测网元发送日志信息信元，该日志信息信元可以携带策略控制网元对日志数据的要求，如设定时长、日志数据的大小、日式数据的发送频率等等。
S1407，待检测网元根据接收到的日志信息信元将预先记录的日志数据发送给策略控制网元。
S1408，故障预测网元向策略控制网元发送日志信息信元，该日志信息信元可以携带故障预测网元对日志数据的要求，如设定时长、日志数据的大小、日式数据的发送频率等等。
S1409，策略控制网元根据接收到的日志信息信元将来自待检测网元的日志数据发送给故障预测网元。从而故障预测网元获取到第一日志数据。
示例性的，S1206，策略控制网元指示所述待检测网元通过所述预防措施处理出现的 故障，具体可以通过步骤S1410实现。
S1410，策略控制网元向待检测网元发送故障恢复信元，该故障恢复信元可以携带功能标识以及预防措施标识。其中，所述功能标识用于标识故障预测功能，预防措施标识用于标识策略控制网元确定的预防措施。
在执行完步骤S1410之后，还可以执行步骤S1411。
S1411，待检测网元向策略控制网元发送故障恢复结果信元，该故障恢复结果信元可以携带功能标识以及恢复结果标识。其中，所述功能标识用于标识故障预测功能。当待检测网元通过策略控制网元确定的预防措施进行处理成功避免未来预设时长后可能出现的故障时，恢复结果标识用于标识恢复成功。当待检测网元通过策略控制网元确定的预防措施进行处理未能成功避免未来预设时长后可能出现的故障时，恢复结果标识用于标识恢复失败。
在执行完步骤S1206之后，可以执行步骤S1412和步骤S1413。
S1412，策略控制网元与待检测网元断开连接。
S1413，策略控制网元与故障预测网元断开连接。
其中，执行步骤S1412与步骤S1413时没有严格的先后顺序，可以先执行步骤S1412，后执行步骤S1413，也可以先执行步骤S1413，后执行步骤S1412，也可以同时执行步骤S1412与步骤S1413，本申请实施例在这里不做具体限定。
在另一种可能的实施方式中，策略控制网元不与待检测网元直接通信，而是通过与故障预测网元与待检测网元进行通信，如图15所示。结合图15所示的通信架构对网络故障预测方法的具体流程进行描述，如图16所示。
在故障预测网元对待检测网元进行故障预测之前，可以执行步骤S1601。
S1601，故障预测网元与待检测网元建立连接。具体的，待检测网元可以向故障预测网元发送状态信令，所述状态信令用于查询故障预测网元是否处于空闲状态。故障预测网元在处于空闲状态时向待检测网元回复当前状态为空闲，则故障预测网元与待检测网元建立连接。故障预测网元在处于忙碌状态时向待检测网元回复当前状态为忙碌。待检测网元在故障预测网元回复当前状态为忙碌时，每间隔第三预设时长向故障预测网元发送状态信令，直到故障预测网元回复当前状态为空闲。第三预设时长可以为5s、3s、12s等等，也可以为其他时间段，本申请实施例在这里不对第三预设时长进行具体限定。
在执行步骤S1201，故障预测网元获取待检测网元预先记录的在过去一个设定时长内的日志数据之前，可以先执行步骤S1602。
S1602，待检测网元向故障预测网元发送请求故障预测信令，该请求故障预测信令用于请求故障预测网元对待检测网元进行故障预测。
在一种可能的实施方式中，步骤S1201，故障预测网元获取待检测网元预先记录的在过去一个设定时长内的日志数据，具体可以通过步骤S1603和S1604实现。
S1603，故障预测网元向待检测网元发送日志信息信元，且该日志信息信元可以携带故障预测网元对日志数据的要求，如设定时长、日志数据的大小、日式数据的发送频率等等。
S1604，待检测网元根据接收到日志信息信元将预先记录的日志数据发送给策略控制网元。从而故障预测网元获取到第一日志数据。
可选的，在执行步骤S1204，故障预测网元向策略控制网元发送查找到的故障类型之 前，可以先执行步骤S1605。
S1605，故障预测网元与策略控制网元建立连接。具体的，故障预测网元可以向策略控制网元发送状态信令。策略控制网元在处于空闲状态时向故障预测网元回复当前状态为空闲，则故障预测网元与策略控制网元建立连接。策略控制网元在处于忙碌状态时向故障预测网元回复当前状态为忙碌。故障预测网元在策略控制网元回复当前状态为忙碌时，每间隔第四预设时长向策略控制网元发送状态信令，直到策略控制网元回复当前状态为空闲。第四预设时长可以为5s、3s、12s等等，也可以为其他时间段，第四预设时长可以与第三预设时长相同，第四预设时长可以与第三预设时长不相同，本申请实施例在这里不对第四预设时长进行具体限定。
在策略控制网元与故障预测网元建立连接之后，故障预测网元可以在策略控制网元的指示下进行初始化，故障预测网元的初始化过程与故障检测网元的初始化过程类似，具体可以参阅图7所示，本申请实施例在这里不再重复赘述。
示例性的，步骤S1206，策略控制网元指示待检测网元通过该预防措施进行处理以避免未来预设时长后可能出现的故障，具体可以通过步骤S1606和步骤S1607实现。
S1606，策略控制网元向故障预测网元发送故障恢复信元。
S1607，故障预测网元向待检测网元转发该故障恢复信元。
在执行完步骤S1607之后，还可以执行步骤S1608和S1609。
S1608，待检测网元向故障预测网元发送故障恢复结果信元。
S1609，故障预测网元向策略控制网元转发故障恢复结果信元。
在执行完步骤S1609之后，可以执行步骤S1610和步骤S1611。
S1610，故障预测网元与待检测网元断开连接。
S1611，故障预测网元与策略控制网元断开连接。
其中，步骤S1610和步骤S1611没有严格的先后执行顺序，可以先执行步骤S1610，后执行步骤S1611，也可以先执行步骤S1611，后执行步骤S1610，也可以同时执行步骤S1610和步骤S1611，本申请实施例在这里不做具体限定。
若故障预测网元自身出现故障导致无法向策略控制网元发送故障类型时，故障预测网元可以通过广播自身出现的故障类型，以使策略控制网元根据该故障类型确定对应的恢复措施，并指示故障预测网元通过该恢复措施处理自身出现的故障。
在一种可能的实施方式中，故障预测网元自身出现故障时的处理过程与故障检测网元自身出现故障时的处理过程类似，具体可以参阅图10所示，或者具体可以参阅图11所示，本申请实施例在这里不再重复赘述。
上述主要从各个网元之间交互的角度对本发明实施例提供的方案进行了介绍。可以理解的是，各个网元，例如故障检测网元、策略控制网元等为了实现上述功能，其包含了执行各个功能相应的硬件结构和/或软件模块。本领域技术人员应该很容易意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，本发明能够以硬件或硬件和计算机软件的结合形式来实现。某个功能究竟以硬件还是计算机软件驱动硬件的方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本发明的范围。
当故障检测网元通过软件模块来实现相应的功能，故障检测网元可以包括获取模块1701、检测模块1702以及发送模块1703，如图17所示。其中，获取模块1701，用于执 行上述图3至图11中获取待检测网元的日志数据的过程。检测模块1702，用于执行上述图3至图11中基于获取模块1701获取的日志数据进行故障检测的过程。发送模块1703，用于执行上述图3至图11中向待检测网元和/或策略控制网元发送信息的过程。所述网元还可以包括接收模块1704。所述接收模块1704，用于执行上述图3至图11中接收待检测网元和/或策略控制网元发送的信息的过程。所述网元还可以包括输出模块1705以及学习模块1706。所述输出模块1705，用于执行上述图3至图11中输出日志特征的过程。所述学习模块1706，用于执行上述图3至图11中学习新的日志特征以及新的故障类型到所述故障关系库中的过程。所述网元还可以包括更新模块1707。所述接收模块1704，还可以用于执行上述图3至图11中更新故障关系库的过程。可以参见图3至图11所示实施例所述的方法，本申请在此不再赘述。
本申请实施例中通过定义5G网络中可以实现故障检测功能，并且提供一个由策略控制网元以及故障检测网元构成的网络故障分析架构，基于该架构，故障检测网元可以根据获取的日志数据确定故障类型，并向策略控制网元发送检测的故障类型，以使策略控制网元确定恢复措施，从而待检测网元可以通过执行该恢复措施来处理出现的故障。
当故障预测网元通过软件模块来实现相应的功能，故障预测网元可以包括获取模块1801、第一预测模块1802、第二预测模块1803以及发送模块1804，如图18所示。其中，获取模块1801，用于执行上述图12至图16中获取待检测网元的第一日志数据的过程。第一预测模块1802，用于执行上述图12至图16中预测第二日志数据的过程。第二预测模块1803，用于执行上述图12至图16中预测故障类型的过程。发送模块1804，用于执行上述图12至图16中向待检测网元和/或策略控制网元发送信息的过程。所述网元还可以包括接收模块1805。所述接收模块1805，用于执行上述图12至图16中接收待检测网元和/或策略控制网元发送的信息的过程。可以参见图12至图16所示实施例所述的方法，本申请在此不再赘述。
本申请实施例中通过定义5G网络中可以实现故障预测功能，并且提供一个由策略控制网元以及故障预测网元构成的网络故障分析架构，基于该架构，故障预测网元可以根据获取的过去一段时间内的日志数据预测未来一段时间后的日志数据，之后根据预测的日志数据确定故障类型，并向策略控制网元发送预测的故障类型，以使策略控制网元确定恢复措施，从而待检测网元可以通过执行该恢复措施进行处理以避免预设时长后可能出现的故障。
当策略控制网元通过软件模块来实现相应的功能，策略控制网元可包括接收模块1901、确定模块1902和发送模块1903，如图19所示。
在一个可能的实施例中，接收模块1901，用于执行上述图3至图11中接收待检测网元和/或故障检测网元发送的信息的过程。确定模块1902，用于执行上述图3至图11中确定故障类型对应的恢复措施的过程。发送模块1903，用于执行上述图3至图11中向待检测网元和/或故障检测网元发送信息的过程。可以参见图3至图11所示实施例所述的方法，本申请在此不再赘述。
本申请实施例中通过定义5G网络中可以实现故障检测功能，并且提供一个由策略控制网元以及故障检测网元构成的网络故障分析架构，基于该架构，策略控制网元可以根据故障检测网元发送的故障类型确定恢复措施，从而待检测网元可以通过执行该恢复措施来处理出现的故障。
在另一个可能的实施例中，接收模块1901，用于执行上述图12至图16中接收待检测网元和/或故障预测网元发送的信息的过程。确定模块1902，用于执行上述图12至图16中确定故障类型对应的预防措施的过程。发送模块1903，用于执行上述图12至图16中向待检测网元和/或故障预测网元发送信息的过程。可以参见图12至图16所示实施例所述的方法，本申请在此不再赘述。
本申请实施例中通过定义5G网络中可以实现故障预测功能，并且提供一个由策略控制网元以及故障预测网元构成的网络故障分析架构，基于该架构，策略控制网元可以根据故障预测网元发送的故障类型确定预防措施，从而待检测网元可以通过执行该恢复措施进行处理以避免预设时长后可能出现的故障。
图20示出了上述实施例中所涉及的网络故障分析网元的另一种可能的结构示意图，网络故障分析网元200可以为故障检测网元，也可以为故障预测网元。网络故障分析网元200包括处理器2002、第一通信接口2001A以及存储器2003。处理器2002，可以是一个中央处理模块(central processing unit，CPU)，或者为数字处理模块等等。第一通信接口2001A，用于与策略控制网元之间收发信息，存储器2003，用于存储处理器2002执行的程序。存储器2003，还可以用于存储故障关系库。存储器2003可以是非易失性存储器，比如硬盘(hard disk drive，HDD)或固态硬盘(solid-state drive，SSD)等，还可以是易失性存储器(volatile memory)，例如随机存取存储器(random-access memory，RAM)。存储器2003是能够用于携带或存储具有指令或数据结构形式的期望的程序代码并能够由计算机存取的任何其他介质，但不限于此。处理器2002用于执行存储器2003存储的程序代码，具体用于执行图3至图16所示实施例所述的方法中故障检测网元除收发信息之外的其他处理或者故障预测网元除收发信息之外的其他处理。可以参见图3至图16所示实施例所述的方法，本申请在此不再赘述。
网络故障分析网元还可以包括第二通信接口2001B，第二通信接口2001B用于与待检测网元之间收发信息。
本申请实施例中不限定上述第一通信接口2001A、第二通信接口2001B、处理器2002以及存储器2003之间的具体连接介质。本申请实施例在图20中以存储器2003、处理器2002以及第一通信接口2001A、第二通信接口2001B之间通过总线2004连接，总线在图20中以粗线表示，其它部件之间的连接方式，仅是进行示意性说明，并不引以为限。所述总线可以分为地址总线、数据总线、控制总线等。为便于表示，图20中仅用一条粗线表示，但并不表示仅有一根总线或一种类型的总线。
图21示出了上述实施例中所涉及的策略控制网元210的另一种可能的结构示意图。策略控制网元210包括处理器2102、第一通信接口2101A以及存储器2103。处理器2102，可以是一个CPU，或者为数字处理模块等等。第一通信接口2101A，用于与故障检测网元之间收发信息或用于与故障预测网元之间收发信息，存储器2103，用于存储处理器2102执行的程序。存储器2103可以是非易失性存储器，比如HDD或SSD等，还可以是volatile memory，例如RAM。存储器2103是能够用于携带或存储具有指令或数据结构形式的期望的程序代码并能够由计算机存取的任何其他介质，但不限于此。处理器2102用于执行存储器2103存储的程序代码，具体用于执行图3至图16所示实施例所述的方法中策略控制网元除收发信息之外的其他处理。可以参见图3至图16所示实施例所述的方法，本申请在此不再赘述。
策略控制网元还可以包括第二通信接口2101B，第二通信接口2101B用于与待检测网元之间收发信息。
本申请实施例中不限定上述第一通信接口2101A、第二通信接口2101B、处理器2102以及存储器2103之间的具体连接介质。本申请实施例在图21中以存储器2103、处理器2102以及第一通信接口2101A、第二通信接口2101B之间通过总线2104连接，总线在图21中以粗线表示，其它部件之间的连接方式，仅是进行示意性说明，并不引以为限。所述总线可以分为地址总线、数据总线、控制总线等。为便于表示，图21中仅用一条粗线表示，但并不表示仅有一根总线或一种类型的总线。
本发明实施例还提供了一种计算机可读存储介质，用于存储为执行上述处理模块所需执行的计算机软件指令，其包含用于执行上述处理模块所需执行的程序。
本申请实施例提供一种包含指令的计算机程序产品，当其在计算机上运行时，使得计算机执行图3至图16所述的网络故障分析方法。
本领域内的技术人员应明白，本申请的实施例可提供为方法、系统、或计算机程序产品。因此，本申请可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本申请可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本申请是参照根据本申请的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
显然，本领域的技术人员可以对本申请进行各种改动和变型而不脱离本申请的精神和范围。这样，倘若本申请的这些修改和变型属于本申请权利要求及其等同技术的范围之内，则本申请也意图包含这些改动和变型在内。

Claims (27)

  1. 一种网络故障检测方法，其特征在于，所述方法包括：
    故障检测网元获取预先记录的待检测网元在过去一个设定时长内的日志数据；
    所述故障检测网元根据所述日志数据，确定对应的日志特征，并在故障关系库中查找所述日志特征所对应的故障类型，所述故障关系库包括不同故障类型与日志特征的关联关系；
    若所述故障检测网元查找到所述日志特征所对应的故障类型，则所述故障检测网元向策略控制网元发送查找到的所述故障类型，以使所述策略控制网元根据所述故障类型确定对应的恢复措施，并指示所述待检测网元通过所述恢复措施处理出现的故障。
  2. 如权利要求1所述的方法，其特征在于，所述故障检测网元还用于接收所述策略控制网元发送的恢复措施，并将所述恢复措施发送给所述待检测网元。
  3. 如权利要求1所述的方法，其特征在于，故障检测网元获取预先记录的所述待检测网元在过去一个设定时长内的日志数据，包括：
    所述故障检测网元向所述待检测网元发送日志数据请求，并接收所述待检测网元发送的所述日志数据，所述日志数据请求用于请求所述日志数据；或者
    所述故障检测网元向所述策略控制网元发送所述日志数据请求，并接收所述策略控制网元发送的所述日志数据。
  4. 如权利要求1所述的方法，其特征在于，在所述故障检测网元在故障关系库中查找所述日志特征所对应的故障类型之前，还包括：
    所述故障检测网元接收所述策略控制网元发送的故障分析指示消息，所述故障分析指示消息携带如下信息中的至少一项：功能标识、检测时间、检测位置；
    其中，所述功能标识用于标识故障检测功能，所述检测时间用于指示所述故障检测网元对所述待检测网元进行故障检测的时间，所述检测位置用于指示所述故障检测网元对所述待检测网元进行故障检测的位置。
  5. 如权利要求1至4任一项所述的方法，其特征在于，还包括：
    所述故障检测网元在故障关系库中查找不到所述日志特征所对应的故障类型时，将确定的日志特征输出；并
    接收输出的日志特征对应的故障类型，将确定的所述日志特征和接收的所述故障类型学习到所述故障关系库中。
  6. 如权利要求1至5任一项所述的方法，其特征在于，所述方法还包括：
    所述故障检测网元接收所述策略控制网元发送的更新指示消息，所述更新指示消息用于指示所述故障检测网元更新所述故障关系库；
    所述故障检测网元更新所述故障关系库。
  7. 一种网络故障检测方法，其特征在于，所述方法包括：
    策略控制网元接收故障检测网元发送的故障类型，所述故障类型是所述故障检测网元在待检测网元故障时，根据所述待检测网元在过去一个设定时长内的日志数据确定的；
    所述策略控制网元根据所述故障类型确定对应的恢复措施；
    所述策略控制网元向所述待检测网元发送所述恢复措施，以指示所述待检测网元通过所述恢复措施处理出现的故障。
  8. 如权利要求7所述的方法，其特征在于，所述策略控制网元向所述待检测网元发送所述恢复措施，包括：
    所述策略控制网元直接向所述待检测网元发送所述恢复措施；或者
    所述策略控制网元通过所述故障检测网元向所述待检测网元发送所述恢复措施。
  9. 如权利要求7或8所述的方法，其特征在于，在策略控制网元接收故障检测网元发送的故障类型之前，还包括：
    所述策略控制网元接收所述故障检测网元发送的日志数据请求，所述日志数据请求用于请求所述待检测网元预先记录的在过去一个设定时长内的日志数据；
    所述策略控制网元向所述故障检测网元发送所述待检测网元的日志数据。
  10. 如权利要求7所述的方法，其特征在于，在策略控制网元接收故障检测网元发送的故障类型之前，还包括：
    所述策略控制网元向所述故障检测网元发送故障分析指示消息，所述故障分析指示消息携带如下信息中的至少一项：功能标识、检测时间、检测位置；
    其中，所述功能标识用于标识故障检测功能，所述检测时间用于指示所述故障检测网元对所述待检测网元进行故障检测的时间，所述检测位置用于指示所述故障检测网元对所述待检测网元进行故障检测的位置。
  11. 如权利要求7至10任一项所述的方法，其特征在于，所述方法还包括：
    所述策略控制网元向所述故障检测网元发送更新指示消息，所述更新指示消息用于指示所述故障检测网元更新所述故障关系库。
  12. 一种网络故障预测方法，其特征在于，所述方法包括：
    故障预测网元获取预先记录的待检测网元在过去一个设定时长内的第一日志数据；
    所述故障预测网元基于所述第一日志数据预测所述待检测网元在预设时长后的第二日志数据；
    所述故障预测网元根据所述第二日志数据，确定对应的日志特征，并在故障关系库中查找所述日志特征所对应的故障类型，所述故障关系库包括不同故障类型与日志特征的关联关系；
    所述故障预测网元向策略控制网元发送所述故障类型，以使所述策略控制网元根据所述故障类型确定对应的预防措施，并指示所述待检测网元通过所述预防措施进行处理以避免预设时长后可能出现的故障。
  13. 如权利要求12所述的方法，其特征在于，所述故障预测网元还用于接收所述策略控制网元发送的预测措施，并将所述预防措施发送给所述待检测网元。
  14. 如权利要求12所述的方法，其特征在于，故障预测网元获取预先记录的待检测网元在过去一个设定时长内的第一日志数据，包括：
    所述故障预测网元向所述待检测网元发送日志数据请求，并接收所述待检测网元发送的所述第一日志数据，所述日志数据请求用于请求所述第一日志数据；或者
    所述故障预测网元向所述策略控制网元发送所述日志数据请求，并接收所述策略控制网元发送的所述第一日志数据。
  15. 如权利要求12所述的方法，其特征在于，在所述故障预测网元基于所述第一日志数据预测所述待检测网元在预设时长后的第二日志数据之前，还包括：
    所述故障预测网元接收所述策略控制网元发送的故障分析指示消息，所述故障分析指 示消息携带所述预设时长，或者，所述故障分析指示消息携带所述预设时长以及如下信息中的至少一项：功能标识、预测时间、预测位置；
    其中，所述功能标识用于标识故障预测功能，所述预测时间用于指示所述故障预测网元对所述待检测网元进行故障预测的时间，所述预测位置用于指示所述故障预测网元对所述待检测网元进行故障预测的位置。
  16. 一种故障检测网元，其特征在于，包括：
    第一通信接口，用于所述故障检测网元与策略控制网元之间收发消息；
    存储器，用于存储计算机程序；
    处理模块，用于调用所述存储器存储的计算机程序，执行：
    通过所述第一通信接口或第二通信接口接收预先记录的待检测网元在过去一个设定时长内的日志数据；所述第二通信接口用于所述故障检测网元与所述待检测网元之间收发消息；
    根据获取的所述日志数据，确定对应的日志特征，并在故障关系库中查找所述日志特征所对应的故障类型，其中所述故障关系库包括有不同故障类型与日志特征的关联关系；在查找到所述日志特征所对应的故障类型时，通过所述第一通信接口向所述策略控制网元发送所述故障类型，以使所述策略控制网元根据所述故障类型确定对应的恢复措施，并指示所述待检测网元通过所述恢复措施处理出现的故障。
  17. 如权利要求16所述的故障检测网元，其特征在于，所述处理模块在用于获取预先记录的待检测网元在过去一个设定时长内的日志数据时，具体用于：
    通过所述第二通信接口向所述待检测网元发送日志数据请求，并通过所述第二通信接口接收所述待检测网元发送的所述日志数据，所述日志数据请求用于请求所述日志数据；或者
    通过所述第一通信接口向所述策略控制网元发送所述日志数据请求，并通过所述第一通信接口接收所述策略控制网元发送的所述日志数据。
  18. 如权利要求16所述的故障检测网元，其特征在于，所述处理模块，还用于在所述故障关系库中查找所述日志特征所对应的故障类型之前，通过所述第一通信接口接收所述策略控制网元发送的故障分析指示消息，所述故障分析指示消息携带如下信息中的至少一项：功能标识、检测时间、检测位置；
    其中，所述功能标识用于标识故障检测功能，所述检测时间用于指示所述故障检测网元对所述待检测网元进行故障检测的时间，所述检测位置用于指示所述故障检测网元对所述待检测网元进行故障检测的位置。
  19. 如权利要求16至18任一项所述的故障检测网元，其特征在于，还包括输入输出模块；
    所述处理模块，还用于在故障关系库中查找不到所述日志特征所对应的故障类型时，将确定的日志特征通过所述输入输出模块输出；并接收从所述输入输出模块输入的根据输出的所述日志特征确定的对应故障类型；以及将所述输入输出模块输出的所述日志特征和从所述输入输出模块接收的对应故障类型学习到所述故障关系库中。
  20. 如权利要求16至19任一项所述的故障检测网元，其特征在于，处理模块，还用于通过所述第一通信接口接收所述策略控制网元发送的更新指示消息，所述更新指示消息用于指示更新所述故障关系库；并根据所述更新指示消息，更新所述故障关系库。
  21. 一种策略控制网元，其特征在于，包括：
    第一通信接口，用于所述策略控制网元与故障检测网元之间收发消息；
    存储器，用于存储计算机程序；
    处理模块，用于调用所述存储器存储的计算机程序，执行：
    通过所述第一通信接口接收故障检测网元发送的故障类型，所述故障类型是所述故障检测网元在待检测网元故障时，根据所述待检测网元在过去一个设定时长内的日志数据确定的；并
    根据所述第一通信接口接收到的所述故障类型确定对应的恢复措施；并
    通过所述第一通信接口或第二通信接口向所述待检测网元发送所述恢复措施，以指示所述待检测网元通过所述恢复措施处理出现的故障，所述第二通信接口用于所述故障检测网元与所述待检测网元之间收发消息。
  22. 如权利要求21所述的策略控制网元，其特征在于，所述第一通信接口，还用于接收故障检测网元发送的故障类型之前，接收所述故障检测网元发送的日志数据请求，所述日志数据请求用于请求所述待检测网元在过去一个设定时长内的日志数据；并向所述故障检测网元发送所述待检测网元的日志数据。
  23. 如权利要求21所述的策略控制网元，其特征在于，所述第一通信接口，还用于在接收故障检测网元发送的故障类型之前，向所述故障检测网元发送故障分析指示消息，所述故障分析指示消息携带如下信息中的至少一项：功能标识、检测时间、检测位置；
    其中，所述功能标识用于标识故障检测功能，所述检测时间用于指示所述故障检测网元对所述待检测网元进行故障检测的时间，所述检测位置用于指示所述故障检测网元对所述待检测网元进行故障检测的位置。
  24. 如权利要求21至23任一项所述的策略控制网元，其特征在于，所述第一通信接口，还用于向所述故障检测网元发送更新指示消息，所述更新指示消息用于指示所述故障检测网元更新所述故障关系库。
  25. 一种故障预测网元，其特征在于，包括：
    第一通信接口，用于所述故障预测网元与策略控制网元之间收发消息；
    存储器，用于存储计算机程序；
    处理模块，用于调用所述存储器存储的计算机程序，执行：
    通过所述第一通信接口或第二通信接口接收预先记录的待检测网元在过去一个设定时长内的第一日志数据；所述第二通信接口用于所述故障检测网元与所述待检测网元之间收发消息；
    基于所述第一日志数据预测所述待检测网元在预设时长后的第二日志数据；根据所述第二日志数据，确定对应的日志特征，并在故障关系库中查找所述日志特征所对应的故障类型，所述故障关系库包括不同故障类型与日志特征的关联关系；
    通过所述第一通信接口向所述策略控制网元发送所述故障类型，以使所述策略控制网元根据所述故障类型确定对应的预防措施，并指示所述待检测网元通过所述预防措施进行处理以避免预设时长后可能出现的故障。
  26. 如权利要求25所述的故障预测网元，其特征在于，所述处理模块在用于获取预先记录的待检测网元在过去一个设定时长内的日志数据时，具体用于：
    通过所述第二通信接口向所述待检测网元发送日志数据请求，并通过所述第二通信接口接收所述待检测网元发送的所述第一日志数据，所述日志数据请求用于请求所述第一日志数据；或者
    通过所述第一通信接口向所述策略控制网元发送所述日志数据请求，并通过所述第一通信接口接收所述策略控制网元发送的所述第一日志数据。
  27. 如权利要求25所述的故障预测网元，其特征在于，所述处理模块，还用于在基于所述第一日志数据预测所述待检测网元在预设时长后的第二日志数据之前，通过所述第一通信接口接收所述策略控制网元发送的故障分析指示消息，所述故障分析指示消息携带所述预设时长，或者，所述故障分析指示消息携带所述预设时长以及如下信息中的至少一项：功能标识、预测时间、预测位置；
    其中，所述功能标识用于标识故障预测功能，所述预测时间用于指示所述故障预测网元对所述待检测网元进行故障预测的时间，所述预测位置用于指示所述故障预测网元对所述待检测网元进行故障预测的位置。
PCT/CN2019/087100 2018-05-29 2019-05-15 网络故障分析方法及装置 WO2019228190A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP19811179.1A EP3787233A4 (en) 2018-05-29 2019-05-15 METHOD AND DEVICE FOR ANALYSIS OF NETWORK FAULTS
US17/105,229 US20210083925A1 (en) 2018-05-29 2020-11-25 Network fault analysis method and apparatus

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810526345.9 2018-05-29
CN201810526345.9A CN110545195A (zh) 2018-05-29 2018-05-29 网络故障分析方法及装置

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US17/105,229 Continuation US20210083925A1 (en) 2018-05-29 2020-11-25 Network fault analysis method and apparatus

Publications (1)

Publication Number Publication Date
WO2019228190A1 true WO2019228190A1 (zh) 2019-12-05

Family

ID=68698718

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/087100 WO2019228190A1 (zh) 2018-05-29 2019-05-15 网络故障分析方法及装置

Country Status (4)

Country Link
US (1) US20210083925A1 (zh)
EP (1) EP3787233A4 (zh)
CN (1) CN110545195A (zh)
WO (1) WO2019228190A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111817891A (zh) * 2020-07-07 2020-10-23 中国联合网络通信集团有限公司 网络故障处理方法、装置、存储介质及电子设备
CN112308126A (zh) * 2020-10-27 2021-02-02 深圳前海微众银行股份有限公司 故障识别模型训练方法、故障识别方法、装置及电子设备
CN112667426A (zh) * 2020-12-31 2021-04-16 北京奇艺世纪科技有限公司 日志分析方法及装置

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113572628B (zh) * 2020-04-28 2023-04-28 中国移动通信集团浙江有限公司 数据关联方法、装置、计算设备及计算机存储介质
CN111800289B (zh) * 2020-05-19 2021-12-10 华为技术有限公司 通信网络故障分析方法和装置
CN111782345B (zh) * 2020-07-07 2022-10-28 郑州迪维勒普科技有限公司 容器云平台日志收集及分析告警方法
CN111740868B (zh) * 2020-07-07 2023-12-15 腾讯科技（深圳）有限公司 告警数据的处理方法和装置及存储介质
CN116325685A (zh) * 2020-08-05 2023-06-23 上海诺基亚贝尔股份有限公司 促进对网络功能行为的分析的服务供应
CN112838950B (zh) * 2020-12-30 2022-05-06 浙江三维利普维网络有限公司 一种网络故障分析方法、装置、存储介质及电子装置
CN112882856A (zh) * 2021-02-09 2021-06-01 江苏云从曦和人工智能有限公司 系统维护方法、装置以及计算机可读存储介质
CN115514615A (zh) * 2021-06-07 2022-12-23 中国移动通信集团有限公司 用于宽带网络的故障检测方法及装置
CN115706695A (zh) * 2021-08-03 2023-02-17 华为技术有限公司 网络故障根因的确定方法、装置、设备及存储介质
CN113872802B (zh) * 2021-09-17 2024-01-19 支付宝(杭州)信息技术有限公司 检测网元的方法和装置
CN114095333A (zh) * 2021-11-23 2022-02-25 天翼数字生活科技有限公司 一种网络排障方法、装置、设备及可读存储介质
CN114826876B (zh) * 2022-01-11 2024-05-03 杭州金硕信息技术有限公司 一种基于日志分析和在线仿真的云服务故障检测系统及方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101719836A (zh) * 2008-10-09 2010-06-02 联想(北京)有限公司 一种故障检测方法及装置
CN105471659A (zh) * 2015-12-25 2016-04-06 华为技术有限公司 一种故障根因分析方法和分析设备
CN106452925A (zh) * 2016-12-02 2017-02-22 华为技术有限公司 在nfv系统中检测故障的方法、装置和系统
US20170104609A1 (en) * 2015-10-09 2017-04-13 Openet Telecom Ltd. System and Method for Enabling Service Lifecycle Based Policy, Licensing, and Charging in a Network Function Virtualization Ecosystem

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6353902B1 (en) * 1999-06-08 2002-03-05 Nortel Networks Limited Network fault prediction and proactive maintenance system
CN102196478B (zh) * 2010-03-01 2014-10-22 中兴通讯股份有限公司 一种网管系统故障的诊断方法和系统
CN105653444B (zh) * 2015-12-23 2018-07-13 北京大学 基于互联网日志数据的软件缺陷故障识别方法和系统
CN105743699A (zh) * 2016-01-27 2016-07-06 中国航空工业集团公司沈阳飞机设计研究所 一种虚拟化环境的故障预警方法及系统
CN105812177B (zh) * 2016-03-08 2019-10-18 华为技术有限公司 一种网络故障处理方法和处理设备
CN107248927B (zh) * 2017-05-02 2020-06-09 华为技术有限公司 故障定位模型的生成方法、故障定位方法和装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101719836A (zh) * 2008-10-09 2010-06-02 联想(北京)有限公司 一种故障检测方法及装置
US20170104609A1 (en) * 2015-10-09 2017-04-13 Openet Telecom Ltd. System and Method for Enabling Service Lifecycle Based Policy, Licensing, and Charging in a Network Function Virtualization Ecosystem
CN105471659A (zh) * 2015-12-25 2016-04-06 华为技术有限公司 一种故障根因分析方法和分析设备
CN106452925A (zh) * 2016-12-02 2017-02-22 华为技术有限公司 在nfv系统中检测故障的方法、装置和系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3787233A4

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111817891A (zh) * 2020-07-07 2020-10-23 中国联合网络通信集团有限公司 网络故障处理方法、装置、存储介质及电子设备
CN112308126A (zh) * 2020-10-27 2021-02-02 深圳前海微众银行股份有限公司 故障识别模型训练方法、故障识别方法、装置及电子设备
CN112667426A (zh) * 2020-12-31 2021-04-16 北京奇艺世纪科技有限公司 日志分析方法及装置
CN112667426B (zh) * 2020-12-31 2023-09-05 北京奇艺世纪科技有限公司 日志分析方法及装置

Also Published As

Publication number Publication date
US20210083925A1 (en) 2021-03-18
EP3787233A4 (en) 2021-06-23
EP3787233A1 (en) 2021-03-03
CN110545195A (zh) 2019-12-06

Similar Documents

Publication Publication Date Title
WO2019228190A1 (zh) 网络故障分析方法及装置
US10484265B2 (en) Dynamic update of virtual network topology
CN109219020A (zh) 一种网络切片方法及系统
US20200382360A1 (en) Cloud service transaction capsulation
US12074767B2 (en) Data processing method and device for cooperating with an artificial intelligence management plane, electronic device, and storage medium
US20130054635A1 (en) Procuring communication session records
CN115562879B (zh) 算力感知方法、装置、电子设备和存储介质
CN108696581A (zh) 分布式信息的缓存方法、装置、计算机设备以及存储介质
CN111953568B (zh) 丢包信息管理方法与装置
CN110248156B (zh) 视频分析方法、平台设备、智能相机和智能监控系统
EP3306471A1 (en) Automatic server cluster discovery
CN111294553B (zh) 处理视频监控业务信令的方法、装置、设备和存储介质
CN111371575B (zh) 一种通话问题定界的方法和装置
CN117042026A (zh) 业务可视化模型构建方法、装置、设备、介质和程序产品
CN116643937A (zh) 数据日志的图像分析
EP4344259A1 (en) Session binding relationship processing method and apparatus, electronic device, and readable medium
WO2023010823A1 (zh) 网络故障根因的确定方法、装置、设备及存储介质
Makai et al. Predicting Mobility Management Demands of Cellular Networks based on User Behavior
CN111124866B (zh) 一种语音交互的方法和装置
CN110896545B (zh) 在线计费漫游故障定位方法及相关装置、存储介质
WO2024093219A1 (zh) 业务路径的优化方法、装置及存储介质
CN113542069B (zh) 一种流量牵引方法、装置、电子设备及存储介质
CN113905105B (zh) 一种建立应用依赖关系的方法及装置
US20190058637A1 (en) Mapping of network device connectivity using performance data
WO2023198174A1 (en) Methods and systems for predicting sudden changes in datacenter networks

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: 19811179

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2019811179

Country of ref document: EP

Effective date: 20201124

NENP Non-entry into the national phase

Ref country code: DE