A Closed-Loop Operation and Maintenance Management Method for 6G Networks Based on Network Operation and Maintenance Ontology and Knowledge Reasoning
By constructing a network operation and maintenance ontology model and knowledge graph, the semantic inconsistency problem of multi-source data in 6G networks was solved, enabling automated location of fault root causes and accurate generation of strategies. This formed a continuously optimized closed-loop operation and maintenance management system, enhancing the autonomous operation and maintenance capabilities of 6G networks.
Patent Information
- Application Number
- CN202611095493.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-22
- Publication Date
- 2026-11-10
AI Technical Summary
The semantics of multi-source operation and maintenance data in 6G networks are inconsistent, making it difficult to model them in a unified manner. There is a lack of inferable correlations between alarms, logs, performance indicators, topology and work orders. Fault root cause location relies on human experience, resulting in insufficient efficiency and accuracy. The handling strategies lack a sorting mechanism based on historical success rate, risk level and business impact. The effects of strategy execution are difficult to write back into the knowledge system, resulting in insufficient closed-loop optimization capabilities.
A network operation and maintenance ontology model and a 6G network operation and maintenance knowledge graph are constructed. Standardized operation and maintenance knowledge units are generated through entity extraction, relation extraction, attribute mapping and semantic normalization. Knowledge reasoning rules are used to determine the causes of faults and their impact range, generate and execute disposal strategies, and update the knowledge graph through execution effect feedback.
It improves the automation level of fault location and the accuracy of handling strategies, forms a continuously optimized closed-loop operation and maintenance management, and enhances the autonomous operation and maintenance capabilities of the 6G network.
Smart Images

Figure CN123075503A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the fields of intelligent operation and maintenance of 6G networks and communication network management technology, and in particular to a closed-loop operation and maintenance management method for 6G networks based on network operation and maintenance ontology and knowledge reasoning. Background Technology
[0002] As 6G networks evolve towards integrated air-space-ground communication, cloud-network-edge-device collaboration, network slicing, computing-network convergence, and intelligent autonomy, network structure, service types, and resource forms exhibit highly dynamic and heterogeneous characteristics. Complex bearing, dependency, and influence relationships exist between access nodes, core network functions, transmission links, edge computing nodes, virtualized network functions, and various services. Single-dimensional alarms or performance indicators are no longer sufficient to accurately reflect the network's operational status and cannot provide a scientific characterization of it.
[0003] In real-world operation and maintenance scenarios, 6G networks continuously generate multi-source data, including network topology data, network element configuration data, performance indicator data, alarm data, log data, service quality data, historical fault work order data, and operation and maintenance record data. The sources, formats, and semantic expressions of these data differ, easily creating information silos and resulting in a lack of unified semantic relationships between fault phenomena, fault objects, fault causes, and handling actions.
[0004] Existing network operation and maintenance methods typically rely on fixed thresholds, manual experience-based rules, or decentralized alarm correlation logic. When an anomaly occurs, operations and maintenance personnel often need to cross-query topology, alarms, logs, work orders, and operation records across multiple systems, resulting in low efficiency in root cause location and difficulty in accurately assessing the scope of business impact. Especially in the highly dynamic environment of 6G networks, traditional static rules struggle to adapt to changes in network status in a timely manner, easily leading to misjudgments, missed judgments, or duplicate handling.
[0005] Meanwhile, existing operation and maintenance platforms lack sufficient accumulation and reuse of historical fault experience. Historical work orders contain a large number of fault phenomena, root causes, handling actions, and handling results, but without unified ontology constraints and knowledge graph organization methods, the relevant experience is difficult to express in a structured way and used for automated reasoning. After the fault is handled, the execution effect is also difficult to be fed back into the knowledge base and reasoning rules in a timely manner, and the operation and maintenance process cannot form a closed loop of continuous optimization.
[0006] Based on the above analysis, the existing technologies have the following problems and shortcomings: the semantics of multi-source operation and maintenance data are inconsistent, making it difficult to model them in a unified manner; there is a lack of inferable correlations between alarms, logs, performance indicators, topology and work orders; the root cause of faults mainly relies on human experience, resulting in insufficient efficiency and accuracy; the handling strategies lack a sorting mechanism based on historical success rate, risk level and business impact; and the effects of strategy execution are difficult to write back into the knowledge system, resulting in insufficient closed-loop optimization capabilities.
[0007] The difficulty in solving the above problems lies in the fact that 6G network objects are large in scale, have many layers, and complex relationships. There are inconsistencies in naming, field meanings, and time status between different data sources. The causes of abnormal events may span multiple levels, including network entities, resource entities, business entities, and performance indicators. The selection of handling strategies also needs to take into account business priorities, execution risks, resource consumption, and rollback conditions.
[0008] Therefore, there is an urgent need in this field for a 6G network operation and maintenance management method that can unify semantic expression, perform knowledge reasoning, and support closed-loop feedback updates. Summary of the Invention
[0009] In view of this, embodiments of this application propose a closed-loop operation and maintenance management method for 6G networks based on network operation and maintenance ontology and knowledge reasoning. By constructing a network operation and maintenance ontology model and a 6G network operation and maintenance knowledge graph, the scattered operation and maintenance data is transformed into understandable, associative, and reasonable knowledge units. Through root cause confidence calculation and impact range identification, the automation level of fault location and the accuracy of handling strategies are improved. Through the feedback of execution effect, operation and maintenance experience is continuously accumulated, thereby improving the autonomous operation and maintenance capability of the 6G network.
[0010] To achieve the above technical objectives, embodiments of this application propose a closed-loop operation and maintenance management method for 6G networks based on network operation and maintenance ontology and knowledge reasoning. The method includes the following steps: acquiring multi-source operation and maintenance data from the 6G network, including at least one of network topology data, network element configuration data, performance indicator data, alarm data, log data, service quality data, historical fault work order data, and operation and maintenance operation record data; constructing a network operation and maintenance ontology model, which provides a unified semantic description of the operation and maintenance objects, operation and maintenance status, abnormal events, fault causes, and operation and maintenance actions of the 6G network; based on the network operation and maintenance ontology model, sequentially performing entity extraction, relation extraction, attribute mapping, entity alignment, and semantic normalization processing on the multi-source operation and maintenance data to generate standardized operation and maintenance knowledge units; and based on the standardized operation and maintenance knowledge units... Construct and update a 6G network operation and maintenance knowledge graph; respond to network anomalies, based on the 6G network operation and maintenance knowledge graph and preset knowledge reasoning rules, determine a set of candidate fault causes associated with the network anomalies; determine the target fault cause and its impact range based on the graph association path, alarm matching degree, performance indicator anomaly degree, historical fault similarity, and topology impact consistency between each candidate fault cause and the network anomaly; match candidate operation and maintenance handling strategies from the 6G network operation and maintenance knowledge graph based on the target fault cause and its impact range, sort the candidate operation and maintenance handling strategies, and generate the target operation and maintenance handling strategy; execute or output the target operation and maintenance handling strategy, and update the 6G network operation and maintenance knowledge graph and knowledge reasoning rules based on the network state change results after execution, so as to form a closed-loop operation and maintenance management of the 6G network.
[0011] To achieve the above objectives, embodiments of this application also propose a 6G network closed-loop operation and maintenance management system based on network operation and maintenance ontology and knowledge reasoning, used to implement the 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning as described above. The system includes: a data acquisition module for acquiring multi-source operation and maintenance data in the 6G network; wherein the multi-source operation and maintenance data includes at least one of network topology data, network element configuration data, performance indicator data, alarm data, log data, service quality data, historical fault work order data, and operation and maintenance operation record data; an ontology management module for constructing and maintaining a network operation and maintenance ontology model; wherein the network operation and maintenance ontology model is used to provide a unified semantic description of the operation and maintenance objects, operation and maintenance status, abnormal events, fault causes, and operation and maintenance actions of the 6G network; a knowledge extraction and standardization processing module for performing entity extraction, relation extraction, attribute mapping, entity alignment, and semantic normalization processing on the multi-source operation and maintenance data based on the network operation and maintenance ontology model to generate standardized operation and maintenance knowledge units; and a knowledge graph construction module. The system comprises the following modules: a construction and update module for building and updating a 6G network operation and maintenance knowledge graph based on standardized operation and maintenance knowledge units; an anomaly detection module for detecting network anomalies; a knowledge reasoning module for determining a set of candidate fault causes associated with detected network anomalies based on the 6G network operation and maintenance knowledge graph and preset knowledge reasoning rules; a root cause localization module for determining the target fault cause and its impact range based on the graph association path, alarm matching degree, performance indicator anomaly degree, historical fault similarity, and topology impact consistency between each candidate fault cause in the candidate fault cause set and the network anomaly; a target operation and maintenance handling strategy generation module for matching candidate operation and maintenance handling strategies from the 6G network operation and maintenance knowledge graph based on the target fault cause and its impact range, ranking the candidate operation and maintenance handling strategies, and generating the target operation and maintenance handling strategy; and a closed-loop feedback module for executing or outputting the target operation and maintenance handling strategy and updating the 6G network operation and maintenance knowledge graph and knowledge reasoning rules based on the network state changes after execution.
[0012] To achieve the above objectives, embodiments of this application also propose an electronic device, including: a processor and a memory, wherein the memory stores instructions executable by the processor, and the processor is configured to execute the instructions such that the electronic device can implement the 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning as described above.
[0013] To achieve the above objectives, embodiments of this application also propose a computer-readable storage medium storing a computer program that, when executed by a processor, enables a 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning as described above.
[0014] In some optional embodiments, the network operation and maintenance ontology model specifically consists of a four-layer structure: a concept layer, an attribute layer, a relationship layer, and a rule layer. The concept layer consists of multiple concept classes, including network entity classes, resource entity classes, business entity classes, performance indicator classes, alarm event classes, log event classes, fault cause classes, operation and maintenance strategy classes, and operation and maintenance result classes; Network entities include at least one of access nodes, core network functions, transmission links, edge computing nodes, network control nodes, and virtualized network functions; resource entities include at least one of computing resources, storage resources, spectrum resources, link resources, cache resources, and instance resources; performance metrics include at least one of latency, throughput, packet loss rate, jitter, connection success rate, session establishment success rate, resource utilization, and service experience quality metrics. The attribute layer is used to define the attribute fields of each concept class; The relationship layer is used to define the relationships between different concept classes, including carrier relationships, dependency relationships, influence relationships, cause relationships, occurrence relationships, association relationships, application relationships, execution relationships, and feedback relationships; The rule layer is used to define anomaly identification rules, root cause reasoning rules, impact scope identification rules, policy matching rules, and closed-loop feedback update rules.
[0015] In some optional embodiments, multi-source operation and maintenance data are sequentially processed through entity extraction, relation extraction, attribute mapping, entity alignment, and semantic normalization to generate standardized operation and maintenance knowledge units, including: Field mapping is performed on structured operation and maintenance data from multi-source operation and maintenance data, mapping the original fields to standard attributes in the network operation and maintenance ontology model; Template parsing is performed on semi-structured log data from multi-source operation and maintenance data to extract log event types, event times, object identifiers, and exception parameters; Text recognition is performed on unstructured historical fault work order data in multi-source operation and maintenance data to extract fault objects, fault phenomena, fault causes, handling actions, and handling results; Based on entity identifier, entity name, topology location, resource ownership relationship, and attribute similarity, perform entity alignment for the same operation and maintenance object from different data sources; The aligned entities, attributes, and relationships are converted into standardized operational knowledge units.
[0016] In some optional embodiments, a 6G network operation and maintenance knowledge graph is constructed and updated based on standardized operation and maintenance knowledge units, including: Standardized operation and maintenance knowledge units are converted into triplet data or attribute graph data consisting of entity nodes, attribute information, and relation edges; Set timestamps, data sources, confidence levels, and status identifiers for entity nodes and relationship edges; When real-time operation and maintenance data is received, it is converted into incremental knowledge units. If the entity node corresponding to the incremental knowledge unit already exists, then update the attribute information, relation edges, and confidence of the corresponding entity node; If the entity node corresponding to the incremental knowledge unit does not exist, then add the entity node and its relation edges.
[0017] In some optional embodiments, in response to a network anomaly event, based on the 6G network operation and maintenance knowledge graph and preset knowledge reasoning rules, a set of candidate fault causes associated with the network anomaly event is determined, including: Identify the abnormal object, abnormal indicator, abnormal time window, and abnormal level corresponding to the abnormal network event; Starting with abnormal objects and abnormal indicators, we search for related paths in the 6G network operation and maintenance knowledge graph along the bearer relationship, dependency relationship, influence relationship, occurrence relationship and cause relationship. Retrieve alarm events, log events, abnormal performance metrics, historical failure cases, and failure cause nodes related to the associated path; Fault cause nodes that meet the preset association conditions are identified as the candidate fault cause set.
[0018] In some optional embodiments, the cause of the target failure is determined by the following steps: For each candidate fault cause in the candidate fault cause set, the following are calculated in sequence: the graph path correlation degree between the candidate fault cause and the network abnormal event, the alarm matching degree between the preset alarm mode corresponding to the candidate fault cause and the current alarm set, the indicator anomaly matching degree between the associated performance index of the candidate fault cause and the current performance index anomaly state, the historical fault similarity between the historical fault cases corresponding to the candidate fault cause and the current anomaly state, and the topological influence consistency between the influence propagation direction of the candidate fault cause in the current network topology and the actual anomaly distribution. Based on the path correlation degree of the graph, alarm matching degree, indicator anomaly matching degree, historical fault similarity, and topological impact consistency, the root cause confidence of the candidate fault causes is calculated. Candidate fault causes whose root cause confidence meets the preset conditions are identified as target fault causes; The formula for calculating root cause confidence is: ; in, Indicate the cause of the candidate fault The root cause confidence level, Indicate the cause of the candidate fault The degree of correlation of the graph path, Indicate the cause of the candidate fault Alarm matching degree Indicate the cause of the candidate fault The degree of abnormal matching of indicators Indicate the cause of the candidate fault Historical fault similarity, Indicate the cause of the candidate fault The topological impact on consistency, , , , , These are the preset weighting coefficients.
[0019] In some optional embodiments, the scope of influence of the target failure cause is determined by the following steps: Starting with the target fault cause, a search is conducted in the 6G network operation and maintenance knowledge graph along the influence relationships, dependency relationships, bearer relationships, and resource consumption relationships; Obtain network entities, resource entities, service entities, and performance metrics associated with the cause of the target failure; Based on the topological hierarchy and service carrying relationships among network entities, resource entities, and business entities, a set of affected objects is generated. Based on the business priority, degree of abnormality, and dependency level of the affected objects, determine the impact level and scope of the target failure cause.
[0020] In some optional embodiments, candidate operation and maintenance (O&M) handling strategies are matched from the 6G network O&M knowledge graph based on the cause of the target fault and its impact scope, and the candidate O&M handling strategies are ranked to generate the target O&M handling strategy, including: Query the operation and maintenance strategy nodes that are applicable to the target fault cause; Candidate maintenance and handling strategies that meet the applicable conditions are selected based on the impact scope of the target failure cause; Obtain the historical success rate, execution risk level, estimated execution time, resource consumption, business impact, and rollback conditions for each candidate operation and maintenance handling strategy; Candidate operation and maintenance strategies are ranked based on historical success rate, execution risk level, expected execution time, resource consumption, business impact, and rollback conditions. Candidate operation and maintenance handling strategies that meet the preset conditions in the ranking results are determined as target operation and maintenance handling strategies.
[0021] In some optional embodiments, the 6G network operation and maintenance knowledge graph and knowledge reasoning rules are updated based on the network state changes after execution to form a closed-loop operation and maintenance management system for the 6G network, including: Collect network status data after implementing the target operation and maintenance strategy; The network status data after implementing the target operation and maintenance strategy is compared with the network status data before implementing the target operation and maintenance strategy to obtain the strategy execution effect. Based on the strategy execution effect, generate operation and maintenance result knowledge units, write the operation and maintenance result knowledge units into the 6G network operation and maintenance knowledge graph, and adjust the historical success rate, confidence level or applicable conditions corresponding to the target operation and maintenance handling strategy according to the strategy execution effect. When the strategy execution effect does not meet the preset recovery conditions, knowledge reasoning is triggered again based on the updated network status data to redetermine the set of candidate fault causes or rematch candidate operation and maintenance handling strategies.
[0022] This application proposes a closed-loop operation and maintenance management method for 6G networks based on network operation and maintenance ontology and knowledge reasoning. Addressing the challenges of 6G network operation and maintenance, it improves the semantic consistency of multi-source heterogeneous data by using a unified network operation and maintenance ontology model to describe the operation and maintenance objects, states, abnormal events, fault causes, and handling actions of the 6G network. By constructing a 6G network operation and maintenance knowledge graph, it transforms topology, alarms, logs, performance data, work orders, and operation and maintenance records into understandable, associative, and reasonable knowledge units. Through multi-dimensional root cause confidence calculation and impact scope identification, it improves the accuracy of fault root cause localization. By ranking strategies based on historical success rates, risk levels, and business impact, it improves the rationality of handling strategy generation. By writing execution results back to the knowledge graph and reasoning rules, it continuously accumulates operation and maintenance experience, forming a continuously optimized closed-loop operation and maintenance management system, effectively enhancing the autonomous operation and maintenance capabilities of the 6G network. Attached Figure Description
[0023] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies of this application will be briefly introduced below. Obviously, the following drawings are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. The drawings described herein are only used to explain this application and are not intended to limit this application.
[0024] Figure 1 This is a flowchart of a 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning provided in one embodiment of this application; Figure 2 This is a detailed illustration of a closed-loop operation and maintenance management method for 6G networks based on network operation and maintenance ontology and knowledge reasoning, provided in one embodiment of this application. Figure 3 This is a diagram of the overall architecture for closed-loop operation and maintenance management of a 6G network provided in another embodiment of this application; Figure 4 This is a schematic diagram illustrating the construction and updating of an operation and maintenance knowledge graph provided in one embodiment of this application; Figure 5 This is a schematic diagram of knowledge reasoning and root cause localization provided in one embodiment of this application; Figure 6 This is a schematic diagram of the structure of a 6G network closed-loop operation and maintenance management system based on network operation and maintenance ontology and knowledge reasoning, provided in another embodiment of this application; Figure 7 This is a schematic diagram of the structure of an electronic device provided in another embodiment of this application. Detailed Implementation
[0025] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the various embodiments of this application will be described in detail below with reference to the accompanying drawings. Those skilled in the art will understand that many technical details have been provided in the embodiments of this application to facilitate better understanding. However, the technical solutions claimed in this application can be implemented even without these technical details and various variations and modifications based on the following embodiments. The division of the following embodiments is for ease of description and should not constitute any limitation on the specific implementation of this application. The following embodiments can be combined with and referenced by each other without contradiction.
[0026] One embodiment of this application proposes a closed-loop operation and maintenance management method for 6G networks based on network operation and maintenance ontology and knowledge reasoning. The implementation details of the 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning proposed in this embodiment are described in detail below. The following implementation details are provided for ease of understanding and are not necessary for implementing this solution.
[0027] The specific process of the 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning proposed in this embodiment can be described as follows: Figure 1 As shown, its visual details are as follows Figure 2 As shown, the method includes: S101, acquire multi-source operation and maintenance data in the 6G network. The multi-source operation and maintenance data includes at least one of the following: network topology data, network element configuration data, performance index data, alarm data, log data, service quality data, historical fault work order data, and operation and maintenance record data.
[0028] In practical implementation, to achieve autonomous operation and maintenance of 6G networks, it is first necessary to acquire multi-source operation and maintenance data from the 6G network. The acquired multi-source operation and maintenance data includes at least one of the following: network topology data, network element configuration data, performance indicator data, alarm data, log data, service quality data, historical fault work order data, and operation and maintenance record data.
[0029] In one example, the overall architecture of 6G network closed-loop operation and maintenance management is as follows: Figure 3 As shown, its architecture includes a data acquisition layer, an ontology and knowledge layer, a knowledge graph layer, a reasoning and decision-making layer, and a handling and feedback layer. The acquisition of multi-source operation and maintenance data is achieved by the data acquisition layer. The data acquisition layer can obtain operation and maintenance data through network management systems, telemetry acquisition interfaces, configuration management databases, log platforms, alarm platforms, service quality monitoring platforms, and work order systems. For performance indicators and alarm data with high real-time requirements, streaming access can be used. For historical fault work orders and operation and maintenance operation records, batch import and periodic update methods can be used.
[0030] S102, Construct a network operation and maintenance ontology model. The network operation and maintenance ontology model is used to provide a unified semantic description of the operation and maintenance objects, operation and maintenance status, abnormal events, fault causes and operation and maintenance actions of 6G networks.
[0031] In practical implementation, after completing the collection of multi-source operation and maintenance data in the 6G network, a network operation and maintenance ontology model can be constructed. The network operation and maintenance ontology model is used to provide a unified semantic description of the operation and maintenance objects, operation and maintenance status, abnormal events, fault causes, and operation and maintenance actions of the 6G network.
[0032] In one example, please refer to Figure 3 The network operation and maintenance ontology model consists of four layers: concept layer, attribute layer, relationship layer, and rule layer.
[0033] The concept layer consists of multiple concept classes, including network entity classes, resource entity classes, business entity classes, performance indicator classes, alarm event classes, log event classes, fault cause classes, operation and maintenance strategy classes, and operation and maintenance result classes.
[0034] Network entities include at least one of access nodes, core network functions, transmission links, edge computing nodes, network control nodes, and virtualized network functions; resource entities include at least one of computing resources, storage resources, spectrum resources, link resources, cache resources, and instance resources; performance metrics include at least one of latency, throughput, packet loss rate, jitter, connection success rate, session establishment success rate, resource utilization, and service experience quality metrics.
[0035] The attribute layer is used to define the attribute fields of each concept class.
[0036] The network operations and maintenance ontology model not only defines conceptual classes such as network entities, resource entities, business entities, performance metrics, alarm events, log events, fault causes, operations and maintenance strategies, and operations and maintenance results, but also defines key attribute fields (attribute classes) for each conceptual class. For example, network entities can have attributes such as entity identifier, entity name, topology location, domain, running status, and version information; performance metrics can have attributes such as metric name, sampling time, metric value, threshold range, and anomaly level.
[0037] The relationship layer is used to define the relationships between different concept classes, including carrier relationships, dependency relationships, influence relationships, cause relationships, occurrence relationships, association relationships, application relationships, execution relationships, and feedback relationships.
[0038] For example, there can be a bearing relationship between business entities and network entities, an occupation relationship or a dependency relationship between network entities and resource entities, a causal relationship or an association relationship between alarm events and fault causes, an applicability relationship between operation and maintenance strategies and fault causes, and a feedback relationship between operation and maintenance results and operation and maintenance strategies.
[0039] The rule layer is used to define anomaly identification rules, root cause reasoning rules, impact scope identification rules, policy matching rules, and closed-loop feedback update rules.
[0040] S103, based on the network operation and maintenance ontology model, sequentially performs entity extraction, relation extraction, attribute mapping, entity alignment and semantic normalization processing on multi-source operation and maintenance data to generate standardized operation and maintenance knowledge units.
[0041] In practical implementation, after the network operation and maintenance ontology model is constructed, entity extraction, relation extraction, attribute mapping, entity alignment and semantic normalization processing can be performed on multi-source operation and maintenance data based on the network operation and maintenance ontology model to generate standardized operation and maintenance knowledge units.
[0042] In one example, the generation of standardized operations and maintenance knowledge units is crucial to the construction and updating of the operations and maintenance knowledge graph. Please refer to [reference needed]. Figure 3 , Figure 4For structured O&M data from multi-source O&M data, this embodiment performs field mapping, mapping the original fields to standard attributes in the network O&M ontology model. For semi-structured log data from multi-source O&M data, this embodiment performs template parsing to extract log event type, event time, object identifier, and abnormal parameters. For unstructured historical fault work orders from multi-source O&M data, this embodiment performs text recognition to extract fault object, fault phenomenon, fault cause, handling action, and handling result. After processing structured, semi-structured, and unstructured data, entity alignment can be performed on the same O&M object from different data sources based on entity identifier, entity name, topology location, resource ownership relationship, and attribute similarity. Finally, the aligned entities, attributes, and relationships are converted into standardized O&M knowledge units.
[0043] During entity alignment, entity identifiers, entity names, topological locations, resource ownership relationships, and attribute similarity can be used to determine whether records from different data sources correspond to the same operation and maintenance object. If they correspond to the same object, they are merged into a single entity node, and their attributes and relationships are updated. If no existing entity can be matched, a new entity node and its relationship edges are added. This avoids the same object appearing repeatedly in the graph, improving the accuracy of subsequent knowledge reasoning.
[0044] S104, construct and update the 6G network operation and maintenance knowledge graph based on standardized operation and maintenance knowledge units.
[0045] In practice, after the standardized operation and maintenance knowledge units are constructed, the 6G network operation and maintenance knowledge graph can be constructed and updated based on the standardized operation and maintenance knowledge units.
[0046] In one example, please refer to Figure 3 , Figure 4 During the knowledge graph construction process, standardized operation and maintenance knowledge units are converted into triplet data or attribute graph data composed of entity nodes, attribute information, and relation edges. Timestamps, data sources, confidence levels, and status identifiers are set for entity nodes and relation edges. When real-time operation and maintenance data is received, it is converted into incremental knowledge units. If the entity node corresponding to the incremental knowledge unit already exists, the attribute information, relation edges, and confidence level of the corresponding entity node are updated. If the entity node corresponding to the incremental knowledge unit does not exist, a new entity node and its relation edges are added, enabling the 6G network operation and maintenance knowledge graph to continuously evolve with changes in network status.
[0047] S105, in response to network anomaly events, determines a set of candidate fault causes associated with the network anomaly events based on the 6G network operation and maintenance knowledge graph and preset knowledge reasoning rules.
[0048] In its specific implementation, this embodiment detects network anomalies in real time and, in response to such anomalies, determines a set of candidate fault causes associated with the network anomalies based on the 6G network operation and maintenance knowledge graph and preset knowledge reasoning rules.
[0049] In one example, please refer to Figure 5 When a network anomaly is detected, the system first determines the anomaly object, anomaly indicator, anomaly time window, and anomaly level in response to the anomaly. For example, when the session establishment success rate of an access node in a certain area is consistently below a threshold and accompanied by core network function alarms, the system uses the access node, session establishment success rate indicator, corresponding time window, and anomaly level as inference inputs.
[0050] Next, knowledge reasoning is performed, using abnormal objects and abnormal indicators as starting nodes. A search for related paths is conducted within the 6G network operation and maintenance knowledge graph, following relationships of bearer relationships, dependencies, influence relationships, occurrence relationships, and causes. During the search, alarm events, log events, abnormal performance indicators, historical fault cases, and fault cause nodes related to the associated paths are retrieved. Fault cause nodes that meet preset association conditions are identified as the candidate fault cause set.
[0051] S106. Based on the graph association path, alarm matching degree, performance index anomaly degree, historical fault similarity and topology impact consistency between each candidate fault cause in the candidate fault cause set and network abnormal events, determine the target fault cause and its impact range.
[0052] In the specific implementation, after knowledge reasoning is completed, root cause localization can be performed. In this embodiment, the target fault cause and its scope of influence are determined based on the graph association path between each candidate fault cause in the candidate fault cause set and the network abnormal event, the alarm matching degree, the abnormality degree of performance indicators, the similarity of historical faults, and the consistency of topological impact.
[0053] In one example, please refer to Figure 5 For each candidate fault cause in the candidate fault cause set, the following calculations are performed sequentially: the graph path correlation degree between the candidate fault cause and the network anomaly event; the alarm matching degree between the preset alarm mode corresponding to the candidate fault cause and the current alarm set; the indicator anomaly matching degree between the associated performance indicator of the candidate fault cause and the current performance indicator anomaly state; the historical fault similarity between the historical fault cases corresponding to the candidate fault cause and the current anomaly state; and the topological influence consistency between the influence propagation direction of the candidate fault cause in the current network topology and the actual anomaly distribution. Then, based on the graph path correlation degree, alarm matching degree, indicator anomaly matching degree, historical fault similarity, and topological influence consistency, the root cause confidence of the candidate fault cause can be calculated.
[0054] In one example, the formula for calculating root cause confidence is: ; in, Indicate the cause of the candidate fault The root cause confidence level, Indicate the cause of the candidate fault The degree of correlation of the graph path, Indicate the cause of the candidate fault Alarm matching degree Indicate the cause of the candidate fault The degree of abnormal matching of indicators Indicate the cause of the candidate fault Historical fault similarity, Indicate the cause of the candidate fault The topological impact on consistency, , , , , The weighting coefficients are preset and can be adjusted based on business type, network domain, historical experience, or manual configuration.
[0055] When the root cause confidence of a candidate fault meets preset conditions, the system identifies it as the target fault cause. Subsequently, using the target fault cause as the starting node, the system searches along the influence relationships, dependencies, bearer relationships, and resource consumption relationships within the 6G network operation and maintenance knowledge graph. Next, it acquires the network entities, resource entities, service entities, and performance indicators associated with the target fault cause. Based on the topological hierarchy and service bearer relationships between these entities, a set of affected objects is generated. Finally, based on the service priority, anomaly degree, and dependency level of the affected objects, the impact level and scope of the target fault cause are determined.
[0056] S107. Based on the cause of the target fault and its impact range, match candidate operation and maintenance handling strategies from the 6G network operation and maintenance knowledge graph, sort the candidate operation and maintenance handling strategies, and generate the target operation and maintenance handling strategy.
[0057] In practice, after determining the cause of the target fault and its scope of impact, candidate operation and maintenance (O&M) strategies can be matched from the 6G network O&M knowledge graph based on the cause of the target fault and its scope of impact. The candidate O&M strategies are then sorted to generate the target O&M strategy.
[0058] In one example, during the strategy generation phase, the system queries operation and maintenance strategy nodes that are applicable to the target fault cause, and filters candidate operation and maintenance handling strategies that meet the applicable conditions based on the impact scope of the target fault cause.
[0059] In one example, if the target failure is caused by excessive resource utilization of edge computing nodes, candidate strategies could include resource expansion, service migration, load balancing, releasing abnormal instances, or adjusting scheduling weights. If the target failure is caused by degraded transmission link quality, candidate strategies could include link switching, route recalculation, bandwidth adjustment, or isolation of faulty links.
[0060] In one example, when ranking candidate operation and maintenance (O&M) handling strategies, it is necessary to obtain the historical success rate, execution risk level, expected execution time, resource consumption, business impact, and rollback conditions of each candidate O&M handling strategy. Based on the historical success rate, execution risk level, expected execution time, resource consumption, business impact, and rollback conditions, the candidate O&M handling strategies are ranked, and the candidate O&M handling strategies that meet the preset conditions are determined as the target O&M handling strategies.
[0061] In one example, for high-priority business, the system can prioritize strategies that have fast recovery speed, minimal business impact, and clear rollback conditions; for low-risk scenarios, the system can choose strategies with lower resource consumption or higher historical success rates.
[0062] S108 executes or outputs the target operation and maintenance handling strategy, and updates the 6G network operation and maintenance knowledge graph and knowledge reasoning rules based on the network status change results after execution, so as to form a closed-loop operation and maintenance management of 6G network.
[0063] In practical implementation, after generating the target operation and maintenance handling strategy, the target operation and maintenance handling strategy can be executed or output, and the 6G network operation and maintenance knowledge graph and knowledge reasoning rules can be updated based on the network status change results after execution, so as to form a closed-loop operation and maintenance management of 6G network.
[0064] In one example, after executing the target operation and maintenance (O&M) strategy, network status data is collected. This post-strategy data is then compared with the pre-strategy data to determine the strategy's effectiveness. If business metrics recover, alarms are cleared, or anomaly levels decrease, an O&M result knowledge unit is generated and written into the knowledge graph. Simultaneously, the historical success rate, confidence level, or applicable conditions of the corresponding strategy are updated.
[0065] If the strategy execution effect does not meet the preset recovery conditions, knowledge reasoning can be retried based on the updated network status data to redetermine the set of candidate fault causes or rematch candidate operation and maintenance handling strategies. This avoids process interruption after a single handling failure and allows for root cause correction, strategy adjustment, and knowledge updates through continuous feedback.
[0066] Taking a scenario of declining session establishment success rate in the 6G core network as an example, the system detects a decrease in session establishment success rate and an increase in latency in a certain service area, accompanied by multiple alarm events. The system first extracts knowledge units related to this service area from topology, configuration, performance, alarms, and logs, and then searches the knowledge graph along bearer relationships and dependencies for relevant core network functions, edge computing nodes, transmission links, and service instances. Subsequently, the system combines historical fault cases and alarm patterns to include "insufficient core network function instance resources," "transmission link congestion," and "abnormal recent configuration changes" in the candidate fault cause set.
[0067] The system further calculates the root cause confidence of the three candidate fault causes. If, in addition to "insufficient core network functional instance resources," there is a high degree of graph path correlation, alarm matching, indicator anomaly matching, and historical fault similarity, and its impact propagation direction is consistent with the actual anomaly distribution, then the system identifies it as the target fault cause. Subsequently, the system identifies the affected service instances, access nodes, and performance indicators, and matches candidate handling strategies, including expanding instances, migrating services, adjusting load balancing weights, and rolling back recent configurations. The system ranks the strategies based on historical success rate, execution risk, and business impact, and outputs the target operation and maintenance handling strategy.
[0068] After the policy is executed, the system re-collects resource utilization, session establishment success rate, latency, alarm status, and service quality indicators. If the above indicators recover to the preset conditions, the system writes the current abnormal event, the cause of the target failure, the handling strategy, and the execution result into the knowledge graph, thus serving as a historical failure case for subsequent similar abnormal events; if the indicators do not recover, the system re-performs reasoning and policy matching based on the new network state.
[0069] This embodiment proposes a closed-loop operation and maintenance management method for 6G networks based on network operation and maintenance ontology and knowledge reasoning. Addressing the operational challenges of 6G networks, it unifies the description of 6G network operation and maintenance objects, states, abnormal events, fault causes, and handling actions through a network operation and maintenance ontology model, effectively improving the semantic consistency of multi-source heterogeneous data. By constructing a 6G network operation and maintenance knowledge graph, topology, alarms, logs, performance data, work orders, and operation and maintenance records are transformed into understandable, associative, and reasonable knowledge units. Multi-dimensional root cause confidence calculation and impact scope identification improve the accuracy of fault root cause localization. Strategy ranking based on historical success rate, risk level, and business impact enhances the rationality of handling strategy generation. By writing execution results back to the knowledge graph and reasoning rules, operation and maintenance experience is continuously accumulated, forming a continuously optimized closed-loop operation and maintenance management system, effectively improving the autonomous operation and maintenance capabilities of 6G networks.
[0070] The steps described above are merely for clarity in describing the technical solution. In actual implementation, they can be combined into one step, or certain steps can be broken down into multiple steps, as long as they involve the same logical relationship, they are all within the scope of protection of this application. Any insignificant modifications or designs added to the algorithm or process, as long as they do not change the core of the algorithm or process, are also within the scope of protection of this application.
[0071] Another embodiment of this application proposes a 6G network closed-loop operation and maintenance management system based on network operation and maintenance ontology and knowledge reasoning, used to implement the 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning as described in the above method embodiment. The details of the 6G network closed-loop operation and maintenance management system based on network operation and maintenance ontology and knowledge reasoning proposed in this embodiment are described below. The following implementation details are provided for ease of understanding and are not essential for implementing this solution. Figure 6 This is a schematic diagram of the structure of a 6G network closed-loop operation and maintenance management system based on network operation and maintenance ontology and knowledge reasoning proposed in this embodiment. It includes: a data acquisition module M201, an ontology management module M202, a knowledge extraction and standardization processing module M203, a knowledge graph construction and update module M204, an anomaly detection module M205, a knowledge reasoning module M206, a root cause localization module M207, a target operation and maintenance handling strategy generation module M208, and a closed-loop feedback module M209.
[0072] The data acquisition module M201 is used to acquire multi-source operation and maintenance data in the 6G network. The multi-source operation and maintenance data includes at least one of the following: network topology data, network element configuration data, performance index data, alarm data, log data, service quality data, historical fault work order data, and operation and maintenance record data.
[0073] The ontology management module M202 is used to build and maintain the network operation and maintenance ontology model. The network operation and maintenance ontology model is used to provide a unified semantic description of the operation and maintenance objects, operation and maintenance status, abnormal events, fault causes and operation and maintenance actions of the 6G network.
[0074] The knowledge extraction and standardization module M203 is used to perform entity extraction, relation extraction, attribute mapping, entity alignment and semantic normalization on multi-source operation and maintenance data based on the network operation and maintenance ontology model, and generate standardized operation and maintenance knowledge units.
[0075] The knowledge graph construction and update module M204 is used to construct and update the 6G network operation and maintenance knowledge graph based on standardized operation and maintenance knowledge units.
[0076] The M205 anomaly detection module is used to detect abnormal network events.
[0077] The knowledge reasoning module M206 is used to respond to detected network anomalies by determining a set of candidate fault causes associated with the network anomalies based on the 6G network operation and maintenance knowledge graph and preset knowledge reasoning rules.
[0078] The root cause localization module M207 is used to determine the target fault cause and its scope of influence based on the graph association path between each candidate fault cause in the candidate fault cause set and the network abnormal event, the degree of alarm matching, the degree of performance index abnormality, the similarity of historical faults, and the consistency of topology impact.
[0079] The target operation and maintenance handling strategy generation module M208 is used to match candidate operation and maintenance handling strategies from the 6G network operation and maintenance knowledge graph based on the cause of the target fault and its impact scope, sort the candidate operation and maintenance handling strategies, and generate the target operation and maintenance handling strategy.
[0080] The closed-loop feedback module M209 is used to execute or output the target operation and maintenance handling strategy, and update the 6G network operation and maintenance knowledge graph and knowledge reasoning rules based on the network status change results after execution.
[0081] It is worth noting that all modules involved in this embodiment are logical modules. In practical applications, a logical module can be a physical module, a part of a physical module, or an organic combination of multiple physical modules. Furthermore, to highlight the innovative aspects of this application, this embodiment does not introduce modules that are not closely related to solving the technical problems proposed in this application. However, this does not mean that other modules are absent from this embodiment.
[0082] It is not difficult to see that this embodiment is a system embodiment corresponding to the above method embodiments, and this embodiment can be implemented in conjunction with the above method embodiments. The relevant technical details and technical effects mentioned in the above method embodiments are still valid in this embodiment, and will not be repeated here to reduce repetition. Accordingly, the relevant technical details mentioned in this embodiment can also be applied to the above method embodiments.
[0083] Another embodiment of this application provides an electronic device, such as Figure 7 As shown, it includes: a processor C301 and a memory C302. The memory C302 stores instructions that the processor C301 can execute. When the processor C301 is configured to execute the instructions, the electronic device can realize a 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning as described in the above method embodiment.
[0084] The memory and processor are connected via a bus, which includes any number of interconnecting buses and bridges. The bus can connect various circuits of one or more processors and memories, as well as other circuits such as peripherals, voltage regulators, and power management circuits—all well-known in the art and therefore not described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single component or multiple components, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor is transmitted over the wireless medium via an antenna, which also receives and transmits data to the processor.
[0085] The processor manages the bus and handles general processing, providing various functions, including but not limited to timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory, on the other hand, is used to store data used by the processor during operation.
[0086] Another embodiment of this application proposes a computer-readable storage medium storing a computer program that, when executed by a processor, can implement a 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning as described in the above method embodiments.
[0087] That is, those skilled in the art will understand that all or part of the steps in the above method embodiments can be implemented by a program instructing related hardware. The program is stored in a storage medium and includes several instructions to cause a device (such as a microcontroller, chip, etc.) or processor to execute all or part of the steps of the method described in the method embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.
[0088] It will be understood by those skilled in the art that the above embodiments are specific implementations of this application, and various changes in form and detail can be made in practical applications without departing from the spirit and scope of this application. For those skilled in the art, several improvements and modifications can be made without departing from the principles of this application, and these improvements and modifications are also considered to be within the scope of protection of this application.
Claims
1. A closed-loop operation and maintenance management method for 6G networks based on network operation and maintenance ontology and knowledge reasoning, characterized in that, The method includes: Acquire multi-source operation and maintenance data in the 6G network. Multi-source operation and maintenance data includes at least one of the following: network topology data, network element configuration data, performance index data, alarm data, log data, service quality data, historical fault work order data, and operation and maintenance record data. A network operation and maintenance ontology model is constructed, which is used to provide a unified semantic description of the operation and maintenance objects, operation and maintenance status, abnormal events, fault causes and operation and maintenance actions of 6G networks. Based on the network operation and maintenance ontology model, entity extraction, relation extraction, attribute mapping, entity alignment and semantic normalization are performed on multi-source operation and maintenance data in sequence to generate standardized operation and maintenance knowledge units. A 6G network operation and maintenance knowledge graph is constructed and updated based on standardized operation and maintenance knowledge units. In response to network anomaly events, based on the 6G network operation and maintenance knowledge graph and preset knowledge reasoning rules, a set of candidate fault causes associated with the network anomaly events is determined; Based on the graph association path, alarm matching degree, performance index anomaly degree, historical fault similarity and topology impact consistency between each candidate fault cause and network anomaly event in the candidate fault cause set, determine the target fault cause and its impact range. Based on the cause of the target fault and its impact range, candidate operation and maintenance handling strategies are matched from the 6G network operation and maintenance knowledge graph, and the candidate operation and maintenance handling strategies are sorted to generate the target operation and maintenance handling strategy. Execute or output target operation and maintenance strategies, and update the 6G network operation and maintenance knowledge graph and knowledge reasoning rules based on the network status changes after execution, so as to form a closed-loop operation and maintenance management of 6G network.
2. The 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning according to claim 1, characterized in that, The network operations and maintenance ontology model consists of four layers: concept layer, attribute layer, relationship layer, and rule layer. The concept layer consists of multiple concept classes, including network entity classes, resource entity classes, business entity classes, performance indicator classes, alarm event classes, log event classes, fault cause classes, operation and maintenance strategy classes, and operation and maintenance result classes; Network entities include at least one of access nodes, core network functions, transmission links, edge computing nodes, network control nodes, and virtualized network functions; resource entities include at least one of computing resources, storage resources, spectrum resources, link resources, cache resources, and instance resources; performance metrics include at least one of latency, throughput, packet loss rate, jitter, connection success rate, session establishment success rate, resource utilization, and service experience quality metrics. The attribute layer is used to define the attribute fields of each concept class; The relationship layer is used to define the relationships between different concept classes, including carrier relationships, dependency relationships, influence relationships, cause relationships, occurrence relationships, association relationships, application relationships, execution relationships, and feedback relationships; The rule layer is used to define anomaly identification rules, root cause reasoning rules, impact scope identification rules, policy matching rules, and closed-loop feedback update rules.
3. The 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning according to claim 1, characterized in that, Multi-source operation and maintenance data are sequentially processed through entity extraction, relation extraction, attribute mapping, entity alignment, and semantic normalization to generate standardized operation and maintenance knowledge units, including: Field mapping is performed on structured operation and maintenance data from multi-source operation and maintenance data, mapping the original fields to standard attributes in the network operation and maintenance ontology model; Template parsing is performed on semi-structured log data from multi-source operation and maintenance data to extract log event types, event times, object identifiers, and exception parameters; Text recognition is performed on unstructured historical fault work order data in multi-source operation and maintenance data to extract fault objects, fault phenomena, fault causes, handling actions, and handling results; Based on entity identifier, entity name, topology location, resource ownership relationship, and attribute similarity, perform entity alignment for the same operation and maintenance object from different data sources; The aligned entities, attributes, and relationships are converted into standardized operational knowledge units.
4. The 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning according to claim 3, characterized in that, A 6G network operation and maintenance knowledge graph is constructed and updated based on standardized operation and maintenance knowledge units, including: Standardized operation and maintenance knowledge units are converted into triplet data or attribute graph data consisting of entity nodes, attribute information, and relation edges; Set timestamps, data sources, confidence levels, and status identifiers for entity nodes and relationship edges; When real-time operation and maintenance data is received, it is converted into incremental knowledge units. If the entity node corresponding to the incremental knowledge unit already exists, then update the attribute information, relation edges, and confidence of the corresponding entity node; If the entity node corresponding to the incremental knowledge unit does not exist, then add the entity node and its relation edges.
5. A 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning according to claim 1, characterized in that, In response to network anomaly events, based on the 6G network operation and maintenance knowledge graph and pre-defined knowledge reasoning rules, a set of candidate fault causes associated with the network anomaly events is determined, including: Identify the abnormal object, abnormal indicator, abnormal time window, and abnormal level corresponding to the abnormal network event; Starting with abnormal objects and abnormal indicators, we search for related paths in the 6G network operation and maintenance knowledge graph along the bearer relationship, dependency relationship, influence relationship, occurrence relationship and cause relationship. Retrieve alarm events, log events, abnormal performance metrics, historical failure cases, and failure cause nodes related to the associated path; Fault cause nodes that meet the preset association conditions are identified as the candidate fault cause set.
6. The 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning according to claim 1, characterized in that, The cause of the target failure is determined through the following steps: For each candidate fault cause in the candidate fault cause set, the following are calculated in sequence: the graph path correlation degree between the candidate fault cause and the network abnormal event, the alarm matching degree between the preset alarm mode corresponding to the candidate fault cause and the current alarm set, the indicator anomaly matching degree between the associated performance index of the candidate fault cause and the current performance index anomaly state, the historical fault similarity between the historical fault cases corresponding to the candidate fault cause and the current anomaly state, and the topological influence consistency between the influence propagation direction of the candidate fault cause in the current network topology and the actual anomaly distribution. Based on the path correlation degree of the graph, alarm matching degree, indicator anomaly matching degree, historical fault similarity, and topological impact consistency, the root cause confidence of the candidate fault causes is calculated. Candidate fault causes whose root cause confidence meets the preset conditions are identified as target fault causes; The formula for calculating root cause confidence is: ; in, Indicate the cause of the candidate fault The root cause confidence level, Indicate the cause of the candidate fault The degree of correlation of the graph path, Indicate the cause of the candidate fault Alarm matching degree Indicate the cause of the candidate fault The degree of abnormal matching of indicators Indicate the cause of the candidate fault Historical fault similarity, Indicate the cause of the candidate fault The topological impact on consistency, , , , , These are the preset weighting coefficients.
7. A 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning according to claim 6, characterized in that, The scope of impact of the target failure cause is determined through the following steps: Starting with the target fault cause, a search is conducted in the 6G network operation and maintenance knowledge graph along the influence relationships, dependency relationships, bearer relationships, and resource consumption relationships; Obtain network entities, resource entities, service entities, and performance metrics associated with the cause of the target failure; Based on the topological hierarchy and service carrying relationships among network entities, resource entities, and business entities, a set of affected objects is generated. Based on the business priority, degree of abnormality, and dependency level of the affected objects, determine the impact level and scope of the target failure cause.
8. The 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning according to claim 1, characterized in that, Based on the cause of the target fault and its impact scope, candidate operation and maintenance (O&M) handling strategies are matched from the 6G network O&M knowledge graph, and these strategies are ranked to generate the target O&M handling strategy, including: Query the operation and maintenance strategy nodes that are applicable to the target fault cause; Candidate maintenance and handling strategies that meet the applicable conditions are selected based on the impact scope of the target failure cause; Obtain the historical success rate, execution risk level, estimated execution time, resource consumption, business impact, and rollback conditions for each candidate operation and maintenance handling strategy; Candidate operation and maintenance strategies are ranked based on historical success rate, execution risk level, expected execution time, resource consumption, business impact, and rollback conditions. Candidate operation and maintenance handling strategies that meet the preset conditions in the ranking results are determined as target operation and maintenance handling strategies.
9. A 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning according to claim 1, characterized in that, The 6G network operation and maintenance knowledge graph and knowledge reasoning rules are updated based on the network state changes after execution to form a closed-loop operation and maintenance management system for the 6G network, including: Collect network status data after implementing the target operation and maintenance strategy; The network status data after implementing the target operation and maintenance strategy is compared with the network status data before implementing the target operation and maintenance strategy to obtain the strategy execution effect; Based on the strategy execution effect, generate operation and maintenance result knowledge units, write the operation and maintenance result knowledge units into the 6G network operation and maintenance knowledge graph, and adjust the historical success rate, confidence level or applicable conditions corresponding to the target operation and maintenance handling strategy according to the strategy execution effect. When the strategy execution effect does not meet the preset recovery conditions, knowledge reasoning is triggered again based on the updated network status data to redetermine the set of candidate fault causes or rematch candidate operation and maintenance handling strategies.
10. A 6G network closed-loop operation and maintenance management system based on network operation and maintenance ontology and knowledge reasoning, used to implement the 6G network closed-loop operation and maintenance management method based on network operation and maintenance ontology and knowledge reasoning as described in any one of claims 1 to 9, characterized in that, The system includes: The data acquisition module is used to acquire multi-source operation and maintenance data in the 6G network; wherein, the multi-source operation and maintenance data includes at least one of network topology data, network element configuration data, performance index data, alarm data, log data, service quality data, historical fault work order data, and operation and maintenance record data; The ontology management module is used to build and maintain the network operation and maintenance ontology model; the network operation and maintenance ontology model is used to provide a unified semantic description of the operation and maintenance objects, operation and maintenance status, abnormal events, fault causes and operation and maintenance actions of the 6G network. The knowledge extraction and standardization module is used to extract entities, extract relations, map attributes, align entities and normalize semantics from multi-source operation and maintenance data based on the network operation and maintenance ontology model, thereby generating standardized operation and maintenance knowledge units. The knowledge graph construction and update module is used to construct and update the 6G network operation and maintenance knowledge graph based on standardized operation and maintenance knowledge units; Anomaly detection module, used to detect abnormal network events; The knowledge reasoning module is used to respond to detected network anomalies by determining a set of candidate fault causes associated with the network anomalies based on the 6G network operation and maintenance knowledge graph and preset knowledge reasoning rules. The root cause localization module is used to determine the target fault cause and its scope of influence based on the graph association path between each candidate fault cause in the candidate fault cause set and network anomaly events, alarm matching degree, performance indicator anomaly degree, historical fault similarity and topology impact consistency. The target operation and maintenance handling strategy generation module is used to match candidate operation and maintenance handling strategies from the 6G network operation and maintenance knowledge graph based on the cause of the target fault and its impact scope, sort the candidate operation and maintenance handling strategies, and generate the target operation and maintenance handling strategy. The closed-loop feedback module is used to execute or output target operation and maintenance handling strategies, and update the 6G network operation and maintenance knowledge graph and knowledge reasoning rules based on the network status changes after execution.
