Base station operation and maintenance method and device, electronic equipment and storage medium

Abstract

The invention relates to the technical field of communication, in particular to a base station operation and maintenance method and device, electronic equipment and a storage medium, and the method comprises the steps: according to one aspect of the disclosure, providing a base station operation and maintenance method based on a knowledge graph, and the method comprises the steps: collecting health degree data of a base station; determining the health state of the base station based on the health degree data and the knowledge graph of the operation and maintenance rules of the base station; determining a target fault problem of the base station based on the health degree data, the health state and the knowledge graph; based on the target fault problem and the knowledge graph, determining an operation and maintenance method corresponding to the target fault problem; based on an operation and maintenance method, operation and maintenance of the target fault problem are guided; wherein the knowledge graph is used for representing the incidence relation of the health state, the target fault problem and the operation and maintenance method. According to the method, the cause and influence of the fault and the recommended operation and maintenance method can be clearly pointed out, so that the interpretability of the decision is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of communication technology, and in particular to a base station operation and maintenance method, device, electronic device and storage medium. Background Art

[0002] With the continuous development of mobile communication technology, the development trend of mobile communication has brought huge challenges to network operation and maintenance. As the key infrastructure of mobile communication networks, the stable operation of base stations guarantees network quality and user experience, so it is very important to achieve highly reliable and efficient active operation and maintenance of base stations.

[0003] However, existing base station operation and maintenance mainly use artificial intelligence technologies such as machine learning (ML) models to select appropriate operation and maintenance strategies. The models have "black box characteristics", so the explainability of the operation and maintenance process is poor and the credibility of operation and maintenance decisions is relatively poor. Summary of the invention

[0004] Based on the above technical problems, a base station operation and maintenance method, device, electronic device and storage medium are provided.

[0005] According to one aspect of the present disclosure, a base station operation and maintenance method based on a knowledge graph is provided, including: collecting health data of the base station; determining the health status of the base station based on the health data and the knowledge graph of the base station operation and maintenance rules; determining the target fault problem of the base station based on the health data, the health status and the knowledge graph; determining the operation and maintenance strategy corresponding to the target fault problem based on the target fault problem and the knowledge graph; guiding the operation and maintenance of the target fault problem based on the operation and maintenance strategy; wherein the knowledge graph is used to characterize the association between the health status, the target fault problem and the operation and maintenance strategy.

[0006] In addition, according to a knowledge graph-based base station operation and maintenance method according to one aspect of the present disclosure, the health status of a base station is determined based on the knowledge graph of health data and base station operation and maintenance rules, including: Determine the health assessment conditions of the base station based on the health data and knowledge graph; Determine the health of the base station based on the evaluation condition and the health data; Determine the health status of the base station based on health and knowledge graph; Among them, the knowledge graph includes the relationship between evaluation conditions, health level and health status.

[0007] In addition, according to a knowledge graph-based base station operation and maintenance method according to one aspect of the present disclosure, based on health data, health status and knowledge graph, a target fault problem of a base station is determined, including: Based on the health status and knowledge graph, determine the fault judgment conditions corresponding to the health status; Based on the fault judgment condition and health data, determine the probability of the fault judgment condition; Based on probability, the fault problem corresponding to the fault judgment condition with the highest probability is selected as the target fault problem of the base station; Among them, the knowledge graph includes the relationship between health status, fault judgment conditions and target fault problems.

[0008] In addition, according to a base station operation and maintenance method based on a knowledge graph in one aspect of the present disclosure, an operation and maintenance strategy corresponding to the target fault problem is determined based on the target fault problem and the knowledge graph, including: Based on the target fault problem and the knowledge graph, determine the optimization algorithm for the target fault problem; Determine the operation and maintenance strategy corresponding to the target fault problem based on the optimization algorithm; Among them, the knowledge graph includes the relationship between target fault problems, optimization algorithms and operation and maintenance strategies.

[0009] According to another aspect of the present disclosure, a base station operation and maintenance device based on a knowledge graph includes: A collection module is used to collect health data of base stations; A status determination module is used to determine the health status of a base station based on the health data and the knowledge graph of base station operation and maintenance rules; A problem determination module is used to determine the target fault problem of the base station based on health data, health status and knowledge graph; The operation and maintenance method determination module is used to determine the operation and maintenance strategy corresponding to the target fault problem based on the target fault problem and the knowledge graph; The operation and maintenance module is used to provide guidance on the operation and maintenance of target fault problems based on the operation and maintenance strategy; Among them, the knowledge graph is used to characterize the relationship between health status, target fault problems and operation and maintenance strategies.

[0010] In addition, according to another aspect of the present disclosure, the base station operation and maintenance device based on the knowledge graph, the state determination module is also used to: Determine the health assessment conditions of the base station based on the health data and knowledge graph; Determine the health of the base station based on the evaluation condition and the health data; Determine the health status of the base station based on health and knowledge graph; Among them, the knowledge graph includes the relationship between evaluation conditions, health level and health status.

[0011] In addition, according to another aspect of the present disclosure, the base station operation and maintenance device based on the knowledge graph, the problem determination module is also used to: Based on the health status and knowledge graph, determine the fault judgment conditions corresponding to the health status; Based on the fault judgment condition and health data, determine the probability of the fault judgment condition; Based on probability, the fault problem corresponding to the fault judgment condition with the highest probability is selected as the target fault problem of the base station; Among them, the knowledge graph includes the relationship between health status, fault judgment conditions and target fault problems.

[0012] In addition, according to a base station operation and maintenance device based on a knowledge graph according to another aspect of the present disclosure, the device further includes: Knowledge graph management module, used to store and optimize knowledge graphs.

[0013] The present disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements any of the above-mentioned knowledge graph-based base station operation and maintenance methods.

[0014] The present disclosure also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, it implements any of the above-mentioned knowledge graph-based base station operation and maintenance methods.

[0015] The present disclosure also provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned knowledge graph-based base station operation and maintenance methods.

[0016] The base station operation and maintenance method, device, electronic device and storage medium provided by the present disclosure can quickly and accurately determine the health status of the base station by collecting the health data of the base station and based on the knowledge graph of the constructed base station operation and maintenance rules. After determining the health status of the base station, the health data, health status and related information in the knowledge graph are further used to accurately locate the specific target fault problem, so that the target fault problem can be guided based on the operation and maintenance method stored in the knowledge graph. When a base station fails, the method based on the knowledge graph can clearly point out the cause, impact and recommended operation and maintenance method of the fault, thereby improving the interpretability of the decision, and due to the transparency of the knowledge graph, the operation and maintenance personnel can more easily verify and verify the correctness of the decision, thereby enhancing the credibility of the operation and maintenance decision. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present disclosure or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 It is a schematic diagram illustrating the B5G base station intelligent active operation and maintenance process in the prior art provided according to an embodiment of the present disclosure.

[0019] Figure 2 It is a flowchart illustrating a base station operation and maintenance method based on a knowledge graph provided according to an embodiment of the present disclosure.

[0020] Figure 3 It illustrates the construction process of the knowledge graph provided according to the embodiment of the present disclosure.

[0021] Figure 4 It is a flowchart that further illustrates the base station operation and maintenance method based on the knowledge graph provided according to an embodiment of the present disclosure.

[0022] Figure 5 It is a flowchart that further illustrates the base station operation and maintenance method based on the knowledge graph provided according to an embodiment of the present disclosure.

[0023] Figure 6 It is a functional block diagram of a base station operation and maintenance device based on a knowledge graph according to an embodiment of the present disclosure.

[0024] Figure 7 It is a schematic diagram illustrating the structure of an electronic device provided according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solutions and advantages of the present disclosure clearer, the technical solutions in the present disclosure will be clearly and completely described below in conjunction with the drawings in the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, rather than all the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0026] With the continuous development of technologies such as the Internet of Things, artificial intelligence, and cloud computing, 5G technology is also constantly improving. 5G has realized applications such as telemedicine, autonomous driving, smart cities, and smart homes in three major application scenarios, including enhanced mobile broadband, ultra-high reliability and low-latency communications, and massive machine-type communications, which has a profound impact on society. In order to meet the communication needs in more complex scenarios in the future and further support the key transformation of economic industries to digitalization, networking, and intelligence, more efficient, reliable, flexible, and intelligent communication networks are needed. In this regard, research on B5G in academia and industry has become increasingly active.

[0027] However, the development trend of mobile communications has brought huge challenges to network operation and maintenance. As the key infrastructure of mobile communication networks, the stable operation of base stations guarantees network quality and user experience. Therefore, it is crucial to achieve highly reliable and efficient active operation and maintenance of base stations. Intelligent active operation and maintenance of base stations uses artificial intelligence technologies such as machine learning (ML) to select appropriate operation and maintenance strategies, maximize system performance, and achieve automated and intelligent base station operation and maintenance and network optimization. It solves the high cost and low efficiency of traditional optimization models that rely solely on expert experience when facing massive data, complex online network environments, and high-agility demand services.

[0028] Figure 1 It is a schematic diagram of the B5G base station intelligent active operation and maintenance process in the existing technology.

[0029] like Figure 1 The B5G base station intelligent active operation and maintenance process shown is mainly divided into two key stages: model training stage and model application stage.

[0030] Model training phase: This phase involves collecting base station historical data from various sources such as measurement reports, statistics, and alarm records. The collected data forms a training dataset for training ML models such as Convolutional Neural Network (CNN), Graph Neural Network (GNN), Generative Adversarial Network (GAN), and Long Short-Term Memory (LSTM). The trained models will be stored in the model library for future applications.

[0031] Model application phase: This phase uses the collected current base station health data as input, applies the trained model to the following three steps, and finally outputs the optimization strategy. The next step is base station health detection, base station fault location, and base station operation and maintenance decision-making.

[0032] The first is the base station health detection. The system will evaluate the current health status of the base station based on the collected base station health data and determine whether it needs further maintenance. In the base station fault location stage, the system will accurately locate the fault based on the current health status and combined with the base station data. Figure 1 In the example shown, the health status of a base station is detected as degraded. According to the model, it is carrying too many services, resulting in excessive load. The last step is base station operation and maintenance decision-making. According to the detected fault, the most appropriate operation and maintenance decision is selected, such as task migration, to improve base station availability and optimize network performance.

[0033] The above-mentioned B5G base station intelligent proactive operation and maintenance process mainly uses machine learning (ML) models to select appropriate operation and maintenance strategies. The model has "black box characteristics", so the explainability of the operation and maintenance process is poor, and the credibility of operation and maintenance decisions is relatively poor.

[0034] With the continuous development of 5G and future B5G technologies, the complexity and challenges of base station operation and maintenance are increasing. Traditional operation and maintenance methods based on machine learning models have problems such as poor interpretability and low decision credibility. Therefore, this paper proposes a base station operation and maintenance method based on knowledge graph to improve the interpretability and credibility of operation and maintenance decisions.

[0035] Figure 2 It is a flowchart illustrating a base station operation and maintenance method based on a knowledge graph according to an embodiment of the present disclosure.

[0036] like Figure 2 As shown, the base station operation and maintenance method based on the knowledge graph according to the embodiment of the present disclosure specifically includes the following steps.

[0037] Step 201: Collect health data of base stations.

[0038] In one embodiment of the present disclosure, health data refers to performance indicators (such as transmit power, receiving sensitivity, transmission speed), hardware indicators (such as device temperature, voltage, current), and alarm information (such as device failure alarm, performance alarm) of the base station. When performing base station operation and maintenance, the health data of the base station can be collected through channels such as sensors and network management systems.

[0039] Step 202, based on the health data and the knowledge graph of base station operation and maintenance rules, determine the health status of the base station.

[0040] In one embodiment of the present disclosure, the collected health data is combined with the knowledge graph to determine the health status of the base station. Specifically, the collected health data is compared and analyzed with the knowledge graph to determine the health status of the base station. This process can judge the health status of the base station equipment through machine learning algorithms and expert experience. Here, the health status can be divided into different levels, such as healthy, sub-healthy, pathological, degraded, etc., to reflect the operating status and maintenance needs of the base station. Exemplarily, the collected health data is compared and analyzed with the preset rules in the knowledge graph, and the health status of the base station equipment is judged through machine learning algorithms and expert experience.

[0041] Expert experience refers to the knowledge and experience accumulated in the field of base station operation and maintenance, which comes from Structured data sources such as base station alarm information and measurement reports, semi-structured data sources such as encyclopedia knowledge, and unstructured data sources such as 3GPP protocol documents and professional literature in the field of base station operation and maintenance. The knowledge graph of base station operation and maintenance rules integrates the above structured data and is a structured knowledge base that represents and stores knowledge in the field of base station operation and maintenance by integrating data from different sources and types into a relationship network. The base station operation and maintenance rules in the knowledge graph enable operation and maintenance personnel to understand and handle complex operation and maintenance problems more effectively, thereby improving the efficiency and service quality of operation and maintenance.

[0042] Step 203: Determine the target fault problem of the base station based on the health data, health status and knowledge graph.

[0043] In one embodiment of the present disclosure, the target fault problem refers to the fault cause that is most likely to affect the performance and stability of the base station, which is obtained through knowledge graph analysis based on the current health status and health data. This fault problem may be the main reason for the current poor health status. Specifically, the collected health data, combined with the health status determined in step 102 and the knowledge graph are analyzed and compared, so that the fault problem that best matches the current health status can be identified, and it is determined that the best matching fault problem is the target fault problem of the base station. Exemplarily, in the knowledge graph, the health status, fault problems, etc. of the equipment are represented as entities. By analyzing and reasoning about the associations between these entities, the fault problem that best matches the current health status, that is, the target fault problem, can be determined.

[0044] Step 204: Determine the operation and maintenance strategy corresponding to the target fault problem based on the target fault problem and the knowledge graph.

[0045] In one embodiment of the present disclosure, an operation and maintenance strategy refers to a set of action plans and measures formulated to solve or prevent specific fault problems. Using the information in the knowledge graph, query the operation and maintenance strategies related to the target fault problem. Thus, the most appropriate strategy or solution for different fault problems stored in the knowledge graph is selected as the corresponding operation and maintenance strategy. Exemplarily, the operation and maintenance strategies related to the target fault problem are found through the relationship network in the knowledge graph, and the most appropriate operation and maintenance strategy is selected based on factors such as the effectiveness and feasibility of the strategy.

[0046] Step 205: Provide guidance on the operation and maintenance of the target fault problem based on the operation and maintenance strategy.

[0047] Among them, the knowledge graph is used to characterize the relationship between health status, target fault problems and operation and maintenance strategies.

[0048] It should be understood that the knowledge graph can store and represent the health status of the base station. In the knowledge graph, the health status of the device can be represented as an attribute of the entity. For example, for a certain base station, its health status can be used as an attribute of the base station entity. By real-time monitoring and collecting the health data of the equipment, this attribute can be dynamically updated to reflect the current health status of the equipment. Then, based on the health status, the most likely target fault problem with the health status can be determined through the relationship network in the knowledge graph. Once the target fault problem is determined, the corresponding operation and maintenance strategy or solution can be found through the knowledge graph. These strategies or solutions can be based on the summary of historical experience, or they can be based on the recommendation of expert experience and ML models, etc. Exemplarily, the operation and maintenance personnel can correctly implement the operation and maintenance strategy according to the specific steps and instructions of the operation and maintenance strategy provided in the knowledge graph, and perform troubleshooting and repair work.

[0049] In one embodiment of the present disclosure, the most appropriate strategy or solution is matched and selected by the target fault problem. After the operation and maintenance strategy is determined, the information in the knowledge graph is used to guide the operation and maintenance personnel to perform operations for the target fault problem, helping the operation and maintenance personnel to correctly implement the operation and maintenance strategy. In the knowledge graph, these operation and maintenance strategies or solutions can be represented as entities or relationships related to the target fault problem. By matching and selecting the most appropriate strategy or solution, the operation and maintenance personnel can be guided to perform troubleshooting and repair work.

[0050] In summary, according to the technical solution provided by the embodiment of the present disclosure, by collecting the health data of the base station and based on the knowledge graph of the constructed base station operation and maintenance rules, the health status of the base station can be quickly and accurately determined. On the basis of determining the health status of the base station, the health data, health status and related information in the knowledge graph can be further utilized to accurately locate the specific target fault problem, so that the target fault problem can be guided based on the operation and maintenance strategy stored in the knowledge graph. When a base station fails, the knowledge graph-based method can clearly point out the cause, impact and recommended operation and maintenance strategy of the fault, thereby improving the interpretability of the decision, and due to the transparency of the knowledge graph, the operation and maintenance personnel can more easily verify and verify the correctness of the decision, thereby enhancing the credibility of the operation and maintenance decision.

[0051] Further, according to a base station operation and maintenance method based on a knowledge graph in one aspect of the present disclosure, the health status of a base station is determined based on the health data and the knowledge graph of the base station operation and maintenance rules, including: Determine the health assessment conditions of the base station based on the health data and knowledge graph; Determine the health of the base station based on the evaluation condition and the health data; Determine the health status of the base station based on health and knowledge graph; Among them, the knowledge graph includes the relationship between evaluation conditions, health level and health status.

[0052] In one embodiment of the present disclosure, the knowledge graph includes an association between evaluation conditions, health, and health status. By comparing and analyzing the collected health data and the knowledge graph, the health evaluation conditions of the base station can be determined. The evaluation conditions are the basis for evaluating the health status of the base station, and may refer to the definition of what "health data" is used to evaluate the health of the base station. The evaluation conditions may also refer to weighting the collected health data. The health of the base station is calculated using the weighted health data. The calculated health is then compared with the knowledge graph. The current health status of the base station is thereby determined.

[0053] Exemplarily, the knowledge graph defines the relationship between the triplet data of the base station health evaluation condition, the base station health score range and the base station health status. The evaluation condition may refer to the weight distribution of the collected health data. By inputting the weighted health data into the trained health prediction model, outputting the health value, and using the health value to find the health score range corresponding to the evaluation condition in the knowledge graph, the health status of the base station is determined according to the corresponding health score range. Among them, the health prediction model is an ML model trained using health data collected from the historical data of the base station. Exemplarily, the health data collected from the historical data of the base station is input into the LSTM model, the model is trained, and a health prediction model is constructed.

[0054] In summary, according to the technical solution provided by the embodiment of the present disclosure, the knowledge graph is used to integrate the base station health data and operation and maintenance rules, and an accurate assessment of the health status of the base station is achieved through comparison and analysis. The knowledge graph contains the relationship between the evaluation conditions, health, and health status, making the evaluation process more intelligent and automated. At the same time, by training a health prediction model (such as an LSTM model), the health status of the base station can be predicted, which helps to detect potential problems in advance and take preventive measures. This method makes full use of the advantages of data-driven decision-making, improves the efficiency and accuracy of operation and maintenance, and provides a strong guarantee for the stable operation of the base station.

[0055] Furthermore, based on the health data, health status and knowledge graph, the target fault problems of the base station are determined, including: Based on the health status and knowledge graph, determine the fault judgment conditions corresponding to the health status; Based on the fault judgment condition and health data, determine the probability of the fault judgment condition; Based on probability, the fault problem corresponding to the fault judgment condition with the highest probability is selected as the target fault problem of the base station; Among them, the knowledge graph includes the relationship between health status, fault judgment conditions and target fault problems.

[0056] It should be understood that in the knowledge graph, the health state is usually represented as an entity of a base station or other device. This entity describes the current health status of the device. The determination of the health state is based on the collected health data and the base station health level evaluation conditions. Fault judgment conditions are rules and conditions for evaluating whether a device has a fault and the type of fault. In the knowledge graph, these conditions are represented as entities or relationships associated with the health state. For example, a certain health state may correspond to a set of specific fault judgment conditions. When these conditions are met, it can be inferred that the device may have a certain fault. Or a certain health state may correspond to multiple sets of specific fault judgment conditions. When the probability of one set of specific fault judgment conditions occurring is high, the target fault problem corresponding to the fault judgment condition can be inferred. The target fault problem refers to the most likely cause of the fault that affects the performance and stability of the device, which is obtained through knowledge graph analysis based on the current health state and health data. In the knowledge graph, the target fault problem is represented as an entity or relationship associated with the health state and the fault judgment condition. Exemplarily, when a device has a certain health state, the knowledge graph can determine the possible target fault problem by analyzing the fault judgment conditions related to it. This process may involve reasoning and calculation of multiple entities and relationships, but ultimately it can provide one or more most likely causes of the failure.

[0057] In the knowledge graph, the relationship between health status, fault judgment conditions, and target fault problems is represented by entities and relationships. These entities and relationships together form a complex information network that can monitor and evaluate the health status of the equipment in real time. Specifically, the health status reflects the current health status of the equipment, the fault judgment conditions are the rules and conditions for evaluating the health status of the equipment based on health data, and the target fault problem is the most likely cause of the fault inferred based on these rules and conditions. Based on the relationship between these three, potential fault problems can be predicted and diagnosed based on the real-time health status of the equipment.

[0058] In one embodiment of the present disclosure, according to the health status of the base station, the fault judgment conditions associated with it are searched in the knowledge graph. These conditions can be derived based on historical data, expert knowledge or statistical analysis, and are used to identify the characteristics of specific faults. After determining the applicable fault judgment conditions, the health data needs to be used to calculate the probability of these conditions being met. To determine the possibility of each fault judgment condition in the current health state. Finally, according to the calculated probability value, the fault problem corresponding to the fault judgment condition with the highest probability is selected as the target fault problem of the base station. This step is based on the principle of probability theory, that is, the higher the probability value, the greater the possibility of the fault problem occurring in the current health state. Therefore, selecting the fault problem with the highest probability as the target fault problem can provide more accurate fault location and analysis results. Using the health data and the determined fault judgment conditions, the probability of each fault judgment condition being met is calculated, which can be obtained according to the machine learning model. Exemplarily, historical health data is collected from the base station, including various performance indicators, alarm information, etc. At the same time, the corresponding fault records are collected to establish the association between the fault problem and the health data. The models that can be selected include decision trees, random forests, support vector machines (SVM), neural networks, etc. The data set is divided into training sets, validation sets, and test sets. The training set is used to train the model, the validation set is used to adjust the model parameters, and the test set is used to evaluate the model performance. The model is trained using the training set to train a model that can predict the probability of different faults in the base station based on the health data.

[0059] In summary, according to the technical solution provided by the embodiment of the present disclosure, through the introduction of the knowledge graph, the result of fault prediction is no longer a simple number or probability value, but explanatory information associated with specific fault judgment conditions. This enables operation and maintenance personnel to understand the prediction results more intuitively and improve the credibility of decision-making. And because the knowledge graph integrates historical data, expert knowledge, statistical analysis and other information, and the machine learning model has been rigorously trained and verified, the accuracy and reliability of the prediction results are improved. This helps operation and maintenance personnel make more informed and effective decisions.

[0060] Furthermore, based on the target fault problem and the knowledge graph, the operation and maintenance strategy corresponding to the target fault problem is determined, including: Based on the target fault problem and the knowledge graph, determine the optimization algorithm for the target fault problem; Determine the operation and maintenance strategy corresponding to the target fault problem based on the optimization algorithm; Among them, the knowledge graph includes the relationship between target fault problems, optimization algorithms and operation and maintenance strategies.

[0061] It is important to understand that in the knowledge graph, the target fault problem is usually represented as a node or entity. An optimization algorithm is a mathematical method or calculation process used to solve a specific problem. In the field of operation and maintenance, an optimization algorithm can help find the best operation and maintenance strategy or solution. In the knowledge graph, the optimization algorithm is usually represented as a node or relationship associated with the target fault problem. These relationships may describe how the algorithm is applied to a specific fault problem, as well as the output or result of the algorithm. The knowledge graph may also contain information such as the performance indicators and applicable scenarios of the algorithm so that the operation and maintenance personnel can choose the appropriate algorithm. The operation and maintenance strategy refers to the specific operation steps or strategies used to solve or prevent faults. These methods may include adjusting equipment parameters, replacing faulty parts, upgrading software versions, etc. In the knowledge graph, the operation and maintenance strategy is usually represented as a node or relationship associated with the target fault problem and the optimization algorithm. These relationships may describe the specific steps, required resources, expected effects, and other information of the operation and maintenance strategy. In the knowledge graph, a close association is formed between the target fault problem, the optimization algorithm, and the operation and maintenance strategy. These relationships not only describe the interactions and dependencies between them, but also provide rich information and context.

[0062] In one embodiment of the present disclosure, once the target fault problem is determined, the optimization algorithm associated with the fault problem can be searched in the knowledge graph. These optimization algorithms may be derived based on historical data, expert experience, or advanced algorithms (such as genetic algorithms, particle swarm optimization, etc.), and are designed to solve specific types of fault problems. According to the nature and characteristics of the target fault problem, the most appropriate optimization algorithm is selected. After the optimization algorithm is determined, the operation and maintenance strategy for the target fault problem is determined based on the output or recommendation of the algorithm. The operation and maintenance strategy may include specific operation steps, required resources, expected results, etc.

[0063] In summary, according to the technical solution provided by the embodiment of the present disclosure, by quickly searching for the optimization algorithm associated with the target fault problem in the knowledge graph, the operation and maintenance personnel can quickly locate the solution that best suits the current fault, thereby significantly reducing the time for fault handling. And the selection of the operation and maintenance strategy is based on the output or suggestion of the optimization algorithm, so the operation and maintenance decision is more scientific and reasonable, reducing the subjectivity and uncertainty of human judgment, and has higher accuracy and reliability.

[0064] It is important to understand that the knowledge graph of base station operation and maintenance rules is a pre-constructed knowledge graph. The knowledge graph is a relational network that connects all different types of information. Each node represents an "entity" that exists in the real world, and each edge is a "relationship" between entities. The core of the knowledge graph is actually a triple, which is composed of an entity, an entity, and a relationship. The knowledge graph of the above base station operation and maintenance rules can be constructed according to the following method.

[0065] Obtain data sources for constructing the knowledge graph of base station operation and maintenance rules. The data sources include: base station alarm information, measurement reports, encyclopedia knowledge, 3GPP protocol documents, and professional literature in the field of base station operation and maintenance; Perform entity extraction and relationship extraction on the data source to obtain triple data; Fuse triple data to construct a knowledge graph.

[0066] Specifically, see Figure 3 ,The process of building a knowledge graph mainly includes three steps: knowledge modeling, knowledge extraction, knowledge fusion and knowledge graph construction.

[0067] Knowledge modeling: This includes identifying relevant domain knowledge in the data source used to construct the knowledge graph for building base station operation and maintenance rules and constructing it into a structured format (i.e., structured data sources such as base station alarm information and measurement reports, semi-structured data sources such as encyclopedia knowledge, and unstructured data sources such as 3GPP protocol documents and professional literature in the field of base station operation and maintenance). For active base station operation and maintenance, this includes defining key entities such as base station health status, base station fault problems, operation and maintenance decisions, etc., and the relationships between them.

[0068] Knowledge extraction: refers to the process of obtaining relevant information triplets from structured data sources such as base station alarm information, measurement reports, semi-structured data sources such as encyclopedia knowledge, and unstructured data sources such as 3GPP protocol documents and professional literature in the field of base station operation and maintenance. In active base station operation and maintenance, neural network decision trees (BiLSTM+CRF, LSTM-RNN, NN-DT) and other methods are usually used to extract base station health detection rules and their characteristic indicators, base station fault problem classification rules and their characteristic indicators, and base station operation and maintenance rules. The specific output triplets are as follows: (base station, adopted,base station health assessment conditions); (base station health assessment conditions, base station health score range, base station health status); (base station health status, using,base station fault judgment conditions); (Base station fault judgment conditions, probability of judging base station fault problems, specific base station faults); (Base station specific failures, optimization algorithms, base station operation and maintenance decisions).

[0069] Among them, entity refers to: Base station: represents a specific base station device and is one of the core entities in the knowledge graph.

[0070] Health assessment conditions: various conditions and indicators used to assess the health status of a base station.

[0071] Base station health status: indicates the current health status of the base station, such as "Normal", "Warning" or "Severe Warning".

[0072] Base station fault judgment conditions: conditions and rules used to judge whether a base station fault occurs and the type of fault.

[0073] Base station specific faults: indicates various types of faults that may occur in the base station, such as "signal coverage problem", "hardware failure", etc.

[0074] Operation and maintenance decisions: Operation and maintenance strategies and suggestions for specific base station failures.

[0075] A relationship is: Adoption: Indicates that an entity (such as a base station) adopts a condition or rule (such as a health assessment condition).

[0076] Base station health score range: indicates the correspondence between the health score range obtained by the evaluation conditions and a certain health status.

[0077] Probability of judging a base station fault problem: indicates the correlation between a certain judgment condition and a certain fault problem, and is used to calculate the probability of a specific fault occurring in a base station based on a certain judgment condition.

[0078] Optimization algorithm: It means finding the best operation and maintenance decision for a specific fault through calculation and analysis.

[0079] Among them, the base station health detection rules and their characteristic indicators are: Base station,adopts,base station health assessment conditions; Base station health assessment conditions, base station health score range, base station health status; Base station fault problem classification rules and their characteristic indicators are: Base station health status, using,base station fault judgment conditions; Base station fault judgment conditions, probability of judging base station fault problems, and specific base station faults; Base station operation and maintenance rules refer to: Base station specific faults (target fault problems), optimization algorithms, and base station operation and maintenance decisions.

[0080] Knowledge fusion and knowledge graph construction: After knowledge extraction is completed, the next step is knowledge fusion, which integrates the above triple data to create a conflict-free, unified and complete representation. The last step is to build a knowledge graph to structure the modeled knowledge into a graph-based format. Nodes represent entities, and edges represent relationships and dependencies. In this way, the base station operation and maintenance system can simultaneously utilize data-driven and knowledge-driven insights to make the decision-making process of active base station operation and maintenance more comprehensive and more explainable.

[0081] Figure 4 It is a flowchart that further illustrates the base station operation and maintenance method based on the knowledge graph provided according to an embodiment of the present disclosure.

[0082] like Figure 4 As shown, a base station operation and maintenance method based on knowledge graph is provided according to an embodiment of the present disclosure.

[0083] Expert experience, i.e. structured data sources such as base station alarm information / indicators, measurement reports, semi-structured data sources such as encyclopedia knowledge / operation and maintenance logs, and unstructured data sources such as 3GPP protocol documents and professional literature in base station operation and maintenance related fields (such as root causes of faults), are used as the knowledge basis to build a knowledge graph of base station operation and maintenance rules. The internal structure of the knowledge graph (such as the definition and association of entities, relationships, and attributes) is transparent to its builders and users, has the characteristics of a white box, can drive knowledge, and is reliable and explainable. Combining knowledge graphs with ML models helps to enhance the intelligent configuration optimization pipeline of base stations. First, base station health detection: collect the health data of base stations (such as alarm information), use the health data to determine the selected base station health assessment conditions from the operation and maintenance rule knowledge graph (i.e. the knowledge graph of the aforementioned base station operation and maintenance rules) to calculate the health of the base station, judge the health status of the base station in the current network based on the health, and judge the health status of the base station as healthy, degraded, dangerous or faulty based on the health of the base station. The result of the health calculation can be achieved by combining the ML model with the knowledge graph. The ML model input is the indicators related to the base station health assessment (i.e. the aforementioned health data), such as alarm information, transmission power, transmission rate, etc., and outputs the health score of the current base station.

[0084] Secondly, if there is a risk in the health status, the base station fault is located: Similarly, combined with the knowledge graph, the fault judgment condition with the largest probability value is selected according to the given fault judgment condition to locate the specific base station fault (that is, the aforementioned target fault problem). For example, based on the judgment condition and its probability, the problem of excessive base station load is located from degradation. The result of the probability value here can be calculated by relying on the ML model combined with the knowledge graph. The input of the ML model is base station related data, including one or more of health data, health score, base station switching success rate, base station frequency resource utilization, etc., and the specific fault category of the base station is output.

[0085] Finally, after locating the specific fault (i.e. determining the target fault problem), combined with the knowledge graph, the corresponding path is selected from the fault entity to output the final operation and maintenance solution (i.e. the aforementioned base station operation and maintenance strategy). Figure 4The selection of the base station operation and maintenance decision in the graph is to select the most appropriate optimization algorithm according to the nature and characteristics of the target fault problem. After determining the optimization algorithm, the operation and maintenance method for the target fault problem is determined according to the output or suggestion of the algorithm. Among them, the optimization algorithm includes a reinforcement learning algorithm and specific parameters. The above-mentioned optimization algorithm is an ML model. During the optimization process, the parameter settings in the operation and maintenance rule knowledge graph are also used to adjust the base station, solve the current problems of the base station, improve the availability of the base station, and improve the network performance. Here, the input of the ML model is selected according to the reinforcement learning algorithm input parameters displayed in the graph, such as the number of LRU spare parts, base station health, maintenance cost, etc., and the operation and maintenance strategy is output. The operation and maintenance strategy can be detected by evaluating the status indicators after the base station operation and maintenance, such as base station health, resource utilization, etc.

[0086] The above-mentioned ML can be an ML model formed by collecting historical data of base stations, such as total throughput, number of base station spare parts, maximum transmission power, etc. through model training. The ML model has black box characteristics, which can drive data with accuracy and effectiveness.

[0087] The following is a use case for implementing proactive O&M for a base station with excessive load. Figure 5 As shown, a base station operation and maintenance method based on a knowledge graph provided according to an embodiment of the present disclosure is further illustrated.

[0088] Step 501, collect data of base stations in the cell network where the user is located.

[0089] Specifically, the real-time data of the current base station is collected, including transmission power, frequency bandwidth, etc. (ie, the aforementioned health data).

[0090] Step 502, construct a knowledge graph of base station operation and maintenance rules based on expert experience and domain knowledge (i.e., the aforementioned base station alarm information, measurement reports and other structured data sources, encyclopedia knowledge and other semi-structured data sources, and 3GPP protocol documents, base station operation and maintenance related professional literature and other unstructured data sources).

[0091] Step 503: Train the ML model based on historical cell network base station data and embed the knowledge graph.

[0092] Step 504, based on the base station operation and maintenance rule knowledge graph, complete the base station health detection, base station fault location and base station operation and maintenance decision-making process in the base station intelligent active operation and maintenance pipeline.

[0093] Specifically, the trained ML model is used to detect the health of all base stations in the current network. According to the operation and maintenance rule knowledge graph combined with the base station health assessment conditions, the health status of the current base station is judged to be degraded. After completing the base station health status assessment, the edge with the highest probability of the corresponding fault judgment condition is found from the knowledge graph, and the target fault problem is determined. When it is confirmed that it is a problem of excessive base station load, the task migration decision output last in the graph is executed. Specifically, the reinforcement learning algorithm and its parameters are used for optimization algorithm calculation, and the operation and maintenance strategy is finally output (i.e., the base station operation and maintenance decision 5 in the figure). The operation and maintenance strategy includes indicators of the base station status, including base station health, resource utilization, etc.

[0094] In the above-mentioned base station operation and maintenance method, the application of knowledge graphs enhances the base station health detection, base station target fault problem determination and the output of base station operation and maintenance strategies.

[0095] 1) Base station health detection: Knowledge graphs can guide the feature engineering of base station health detection and improve detection performance by selecting appropriate features to improve problems caused by less abnormal data or poor data quality.

[0096] 2) Target fault problem: Use the white-box characteristics of knowledge graphs to solve the black-box problems of neural networks. At the same time, when facing complex network environments, it helps classify base station faults by better capturing the relevance of problems.

[0097] 3) Base station operation and maintenance strategy: KG ensures strong interpretability, making the final output of operation and maintenance decisions trustworthy. In addition, it also helps to supplement reasoning, making decisions more comprehensive and achieving better base station optimization results.

[0098] Figure 6 It is a functional block diagram illustrating a base station operation and maintenance device based on a knowledge graph according to an embodiment of the present disclosure.

[0099] like Figure 6 As shown, the base station operation and maintenance device 600 based on the knowledge graph includes: a collection module 601, a state determination module 602, a problem determination module 603 and an operation and maintenance method determination module 604, an operation and maintenance module 605 and a knowledge graph management module 606.

[0100] Further, the collection module 601 is used to collect health data of the base station; Furthermore, the status determination module 602 is used to determine the health status of the base station based on the health data and the knowledge graph of the base station operation and maintenance rules.

[0101] Furthermore, the problem determination module 603 is used to determine the target fault problem of the base station based on the health data, health status and knowledge graph.

[0102] Furthermore, the operation and maintenance method determination module 604 is used to determine the operation and maintenance strategy corresponding to the target fault problem based on the target fault problem and the knowledge graph; Furthermore, the operation and maintenance module 605 is used to guide the operation and maintenance of the target fault problem based on the operation and maintenance strategy.

[0103] Among them, the knowledge graph is used to characterize the relationship between health status, target fault problems and operation and maintenance strategies.

[0104] Furthermore, the state determination module 602 is further configured to: Determine the health assessment conditions of the base station based on the health data and knowledge graph; Determine the health of the base station based on the evaluation condition and the health data; Determine the health status of the base station based on health and knowledge graph; Among them, the knowledge graph includes the relationship between evaluation conditions, health level and health status.

[0105] Furthermore, the problem determination module 603 is also used to: Based on the health state and the knowledge graph, determining a fault judgment condition corresponding to the health state; Determining the probability of the fault judgment condition based on the fault judgment condition and the health data; Based on the probability, selecting the fault problem corresponding to the fault judgment condition with the highest probability as the target fault problem of the base station; Among them, the knowledge graph includes the association between the health status, the fault judgment conditions and the target fault problem.

[0106] Furthermore, the knowledge graph management module 606 is used to store and optimize the knowledge graph.

[0107] Specifically, the knowledge graph management module is responsible for storing the knowledge graph of base station operation and maintenance rules, including all entities (such as base stations, health assessment conditions, health status, fault judgment conditions, etc.) and the relationships between them. Ensure that the data of the knowledge graph is structured and organized to facilitate access and use by other modules. And as time goes by and operation and maintenance practices accumulate, the knowledge graph needs to be continuously updated and optimized to reflect the latest operation and maintenance rules and experience.

[0108] Furthermore, the knowledge graph management module 606 is also used to maintain the knowledge graph, including adding new entities and relationships, updating existing information, and deleting outdated or no longer applicable rules.

[0109] Figure 7 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 7As shown, the electronic device may include: a processor 710, a communication interface 720, a memory 730 and a communication bus 740, wherein the processor 710, the communication interface 720 and the memory 730 communicate with each other through the communication bus 740. The processor 710 may call the logic instructions in the memory 730 to execute the above-mentioned base station operation and maintenance method based on the knowledge graph.

[0110] In addition, the logic instructions in the above-mentioned memory 730 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present disclosure is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present disclosure. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0111] On the other hand, the present disclosure also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the above-mentioned base station operation and maintenance method.

[0112] On the other hand, the present disclosure also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented to perform the above-mentioned base station operation and maintenance method when executed by a processor.

[0113] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0114] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0115] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them. Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. A base station operation and maintenance method based on knowledge graph, characterized in that: include: Collect health data of base stations; Determine the health status of the base station based on the health data and the knowledge graph of base station operation and maintenance rules; Determine a target fault problem of the base station based on the health data, the health status and the knowledge graph; Based on the target fault problem and the knowledge graph, determine an operation and maintenance strategy corresponding to the target fault problem; Based on the operation and maintenance strategy, provide guidance on the operation and maintenance of the target fault problem; The knowledge graph is used to characterize the relationship between health status, target fault problems and operation and maintenance strategies.

2. The base station operation and maintenance method based on knowledge graph according to claim 1, characterized in that: The determining the health status of the base station based on the knowledge graph of the health data and the base station operation and maintenance rules includes: Determining a health evaluation condition of the base station based on the health data and the knowledge graph; Determining the health of the base station based on the evaluation condition and the health data; Determining a health status of the base station based on the health degree and the knowledge graph; Among them, the knowledge graph includes the association between the evaluation conditions, health level and health status.

3. The base station operation and maintenance method based on knowledge graph according to claim 1, characterized in that: The determining a target fault problem of the base station based on the health data, the health state and the knowledge graph includes: Based on the health state and the knowledge graph, determining a fault judgment condition corresponding to the health state; Determining the probability of the fault judgment condition based on the fault judgment condition and the health data; Based on the probability, selecting the fault problem corresponding to the fault judgment condition with the highest probability as the target fault problem of the base station; Among them, the knowledge graph includes the association between the health status, the fault judgment conditions and the target fault problem.

4. The base station operation and maintenance method based on knowledge graph according to claim 1, characterized in that: The determining, based on the target fault problem and the knowledge graph, an operation and maintenance strategy corresponding to the target fault problem includes: Based on the target fault problem and the knowledge graph, determining an optimization algorithm for the target fault problem; Based on the optimization algorithm, determine the operation and maintenance strategy corresponding to the target fault problem; Among them, the knowledge graph includes the relationship between the target fault problem, the optimization algorithm and the operation and maintenance strategy.

5. A base station operation and maintenance device based on knowledge graph, characterized in that: include: A collection module is used to collect health data of base stations; A status determination module, configured to determine the health status of the base station based on the health data and the knowledge graph of base station operation and maintenance rules; A problem determination module, configured to determine a target fault problem of the base station based on the health data, the health status and the knowledge graph; An operation and maintenance method determination module, used to determine the operation and maintenance strategy corresponding to the target fault problem based on the target fault problem and the knowledge graph; An operation and maintenance module, used to guide the operation and maintenance of the target fault problem based on the operation and maintenance strategy; The knowledge graph is used to characterize the relationship between health status, target fault problems and operation and maintenance methods.

6. The base station operation and maintenance device based on knowledge graph according to claim 4, characterized in that: The state determination module is further used for: Determining a health evaluation condition of the base station based on the health data and the knowledge graph; Determining the health of the base station based on the evaluation condition and the health data; Determining a health status of the base station based on the health degree and the knowledge graph; Among them, the knowledge graph includes the association between the evaluation conditions, health level and health status.

7. The base station operation and maintenance device based on knowledge graph according to claim 4, characterized in that: The problem determination module is further used for: Based on the health state and the knowledge graph, determining a fault judgment condition corresponding to the health state; Determining the probability of the fault judgment condition based on the fault judgment condition and the health data; Based on the probability, selecting the fault problem corresponding to the fault judgment condition with the highest probability as the target fault problem of the base station; Among them, the knowledge graph includes the association between the health status, the fault judgment conditions and the target fault problem.

8. The base station operation and maintenance device based on knowledge graph according to claim 4, characterized in that: The device also includes: Knowledge graph management module, used to store and optimize knowledge graphs.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, it implements the base station operation and maintenance method based on the knowledge graph as described in any one of claims 1 to 6.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the base station operation and maintenance method based on the knowledge graph is implemented as described in any one of claims 1 to 6.

Images