5g-r multi-monitoring data fusion detection network situation awareness analysis method and system

Abstract

The application discloses a 5G-R multi-monitoring data fusion detection network situation awareness analysis method and system, through a 5G-R multi-monitoring data fusion detection network situation awareness analysis scheme, the multi-element monitoring data is fused around the communication process element information (namely three dimensions of time, position and event) which is centered on 5G-R service performance, a normalized label system is established, and according to the actual demand in network operation, it is divided into two parallel analysis routes of real-time and non-real-time; at the same time, in order to cope with the different requirements of real-time and non-real-time process, processing flow, key technology and resource requirement, a series of methods such as label normalization, deviation maintenance, multi-dimensional label association, feature storage and degradation protection are also provided to ensure the comprehensiveness, accuracy and efficiency of the analysis. The technical scheme is considered at the beginning of the design, which is convenient and fast to deploy, and introduces the data interface, label adaptation and model optimization process when AI is introduced.

Description

Technical Field

[0001] This invention relates to the field of railway 5G-R wireless network communication analysis technology, and in particular to an analysis method and system for 5G-R multi-monitoring data fusion detection network situational awareness. Background Technology

[0002] With the rapid development of high-speed rail, high-speed trains, and bullet trains, the mileage of the railway network has increased rapidly. To enhance passengers' wireless network communication experience, mobile operators are constantly striving to optimize network services along the routes. In the future, with the widespread adoption of autonomous driving technology, the demands on the network will be even higher, especially for dedicated communication networks (5G-R) for high-speed rail and bullet trains. Their role in train scheduling and control information transmission will become increasingly important, directly impacting the safety and reliability of railway operations.

[0003] Currently, both domestically and internationally, assessments of wireless network coverage, quality, and incidents along linear routes such as highways and railways primarily rely on several statistical indicators to summarize the overall wireless status of the test line. While these indicators provide a macro-level network overview, specific anomalies and quality defects are still handled through manual analysis. This approach suffers from several problems: the analysts' abilities and work attitudes are uncertain, and the lack of objective, quantifiable standards leads to insufficient analytical accuracy and even potential misjudgments. Furthermore, existing industry performance evaluation systems are insufficient for the demands of refined management. Although these systems provide basic standards, they still fall short in practical application for sophisticated management.

[0004] In light of the aforementioned issues, it is necessary to conduct in-depth research on existing network analysis methods and develop a highly accurate and efficient solution to achieve comprehensive perception and real-time analysis of the 5G-R network situation. This will not only help improve the quality and stability of network services but also provide stronger guarantees for the safety and efficiency of railway operations. Summary of the Invention

[0005] The purpose of this invention is to provide an analysis method and system for 5G-R multi-monitoring data fusion detection network situational awareness, which can accurately and efficiently achieve comprehensive perception and real-time analysis of 5G-R network situation.

[0006] The objective of this invention is achieved through the following technical solution:

[0007] An analytical method for 5G-R multi-monitoring data fusion detection network situational awareness includes:

[0008] Step 1: Acquire multi-source data from the 5G-R communication network and divide it into real-time monitoring data and non-real-time monitoring data;

[0009] Step 2: Store the real-time monitoring data and non-real-time monitoring data as raw data and manage them throughout their entire lifecycle, as well as use them for AI model training;

[0010] Step 3: Real-time monitoring data processing: Extract data from multiple dimensions of the real-time monitoring data, classify and label it, and then provide it to Step 5. Then, perform data analysis based on the normalized multidimensional data labels fed back from Step 5 to obtain the analysis results of the real-time monitoring data.

[0011] Step 4: Non-real-time monitoring data processing: After preprocessing the non-real-time monitoring data, the corresponding multidimensional data is provided to Step 5. The data is segmented and stored in combination with the normalized multidimensional data labels fed back from Step 5. The association between different data sources is established based on the normalized multidimensional data labels. The data is then divided according to different time scales. The data at each time scale is analyzed by combining the empirical model and the AI ​​model to obtain the analysis results of the non-real-time monitoring data.

[0012] Step 5: Establish a normalized tagging system based on 5G-R services and multidimensional data, and transfer the tags to Step 3 and Step 4 respectively.

[0013] Step 6: Combine the analysis results of the real-time monitoring data and the analysis results of the non-real-time monitoring data to output the analysis results of network situational awareness.

[0014] An analysis system for situational awareness detection networks based on 5G-R multi-monitoring data fusion, used to implement the aforementioned method, includes:

[0015] The multi-source data acquisition and segmentation unit is used to acquire multi-source data from the 5G-R communication network and segment it into real-time monitoring data and non-real-time monitoring data.

[0016] The raw data lifecycle management unit is used to store the real-time monitoring data and non-real-time monitoring data as raw data and manage them throughout their lifecycle, as well as for AI model training.

[0017] The real-time monitoring data processing unit is used for real-time monitoring data processing, including: extracting real-time monitoring data from multiple dimensions, classifying and labeling it and providing it to step 5, and then performing data analysis based on the normalized multi-dimensional data labels fed back from step 5 to obtain the analysis results of the real-time monitoring data.

[0018] The non-real-time monitoring data processing unit is used for non-real-time monitoring data processing, including: preprocessing the non-real-time monitoring data and providing the corresponding multidimensional data to step 5; combining the normalized multidimensional data labels fed back from step 5 to perform data segmentation and storage; establishing the association between different data sources based on the normalized multidimensional data labels; dividing the data according to different time scales; and combining the empirical model and the AI ​​model to perform data analysis on the data at each time scale to obtain the analysis results of the non-real-time monitoring data.

[0019] The tag adaptation unit is used to establish a normalized tag system based on 5G-R services and multidimensional data, and to transfer tags in steps 3 and 4 respectively.

[0020] The conclusion report output unit is used to integrate the analysis results of the real-time monitoring data and the analysis results of the non-real-time monitoring data to output the analysis results of network situational awareness.

[0021] As can be seen from the technical solution provided by the present invention, the analysis scheme for network situational awareness detection based on 5G-R multi-monitoring data fusion integrates multi-dimensional monitoring data (i.e., time, location, and event dimensions) centered on 5G-R service performance to establish a normalized tag system. Furthermore, it divides the analysis into two parallel paths—real-time and non-real-time—based on actual network operation and maintenance needs. Simultaneously, to address the differences in target requirements, processing procedures, key technologies, and resource requirements between real-time and non-real-time processes, it provides a series of methods such as tag normalization, deviation maintenance, multi-dimensional tag association, feature storage, and degradation protection to ensure the comprehensiveness, accuracy, and efficiency of the analysis. This technical solution was designed from the outset with programming convenience in mind, facilitating rapid deployment, and addressing the data interface, tag adaptation, and model optimization processes when introducing AI. Attached Figure Description

[0022] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 A flowchart illustrating an analysis method for 5G-R multi-monitoring data fusion detection network situational awareness provided in an embodiment of the present invention;

[0024] Figure 2 A schematic diagram of the overall framework of an analysis method for 5G-R multi-monitoring data fusion detection network situational awareness provided in an embodiment of the present invention;

[0025] Figure 3 This is a schematic diagram of data fusion in a 5G-R communication multi-monitoring system provided in an embodiment of the present invention;

[0026] Figure 4 This is a schematic diagram of the normalized coordinate relationship of tags in a 5G-R communication multi-monitoring system provided in an embodiment of the present invention;

[0027] Figure 5 A schematic diagram of multi-level situational awareness of a 5G-R communication network provided in an embodiment of the present invention;

[0028] Figure 6 This is a three-dimensional protection diagram of 5G-R services in various scenarios provided by embodiments of the present invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present invention.

[0030] First, the following explanations are provided for the terms that may be used in this article:

[0031] The terms “including,” “comprising,” “containing,” “having,” or other similar semantic descriptions should be interpreted as non-exclusive inclusion. For example, “including a technical feature element (such as raw material, component, ingredient, carrier, dosage form, material, size, part, component, mechanism, device, step, process, method, reaction conditions, processing conditions, parameter, algorithm, signal, data, product or article of manufacture, etc.)” should be interpreted as including not only the expressly listed technical feature element, but also other technical feature elements that are not expressly listed and are well-known in the art.

[0032] The term "composed of" excludes any technical features not expressly listed. When used in a claim, it closes the claim to exclude all technical features other than those expressly listed, except for associated conventional impurities. If the term appears only in a clause of a claim, it limits the claim to the elements expressly listed in that clause; elements recited in other clauses are not excluded from the overall claim.

[0033] The following is a detailed description of the analysis method and system for 5G-R multi-monitoring data fusion detection network situational awareness provided by this invention. Contents not described in detail in the embodiments of this invention belong to prior art known to those skilled in the art. Where specific conditions are not specified in the embodiments of this invention, they shall be performed according to conventional conditions in the art or conditions recommended by the manufacturer. Reagents or instruments used in the embodiments of this invention, unless otherwise specified by the manufacturer, are all conventional products that can be purchased commercially.

[0034] Example 1

[0035] This invention provides an analysis method and system for 5G-R multi-monitoring data fusion detection network situational awareness, such as... Figure 1 As shown, it mainly includes the following steps:

[0036] Step 1: Obtain multi-source data from the 5G-R communication network and divide it into real-time monitoring data and non-real-time monitoring data.

[0037] In this embodiment of the invention, the multi-source data of the 5G-R communication network includes: 5G network northbound alarm data, 5G interface monitoring system data, 5G wireless probe data, 5G transmission network management system data, 5G network environmental monitoring data, antenna and tower monitoring data, 5G-R network northbound performance data, 5G-R road test equipment data, 5G-R network configuration parameters, 5G-R equipment ledger, MDT minimum road test terminal measurement and acquisition system data, CIR system on-board equipment monitoring system data, and MCX railway business dedicated processing service data. After obtaining the above multi-source data of the 5G-R communication network, it can be divided into real-time monitoring data and non-real-time monitoring data according to data characteristics.

[0038] Preferably, considering that different measurement schemes will have certain deviations, for example, different terminals can be used to measure signal strength RSRP, but due to differences in terminal sensitivity, antenna interface attenuation, and antenna gain, deviations will occur, and this deviation amount Δp will change due to equipment aging or failure. Therefore, after acquiring multi-dimensional data from the 5G-R communication network, deviation corrections are made for each type of data. For example, at the same location, multiple measurement terminals collect wireless signals, and comparisons are made between them. Since the number of train control terminals on a railway line is limited, comparisons are made between all of them, thereby tracking and correcting deviations.

[0039] Step 2: Store the real-time monitoring data and non-real-time monitoring data as raw data and manage them throughout their entire lifecycle, as well as use them for AI model training.

[0040] In this embodiment of the invention, the data in step 1 is managed over its lifecycle, and can also be used for training AI models.

[0041] Step 3: Real-time monitoring data processing: Extract data from multiple dimensions of the real-time monitoring data, classify and label it, and then provide it to Step 5. Then, perform data analysis based on the normalized multidimensional data labels fed back from Step 5 to obtain the analysis results of the real-time monitoring data.

[0042] Preferably, the real-time monitoring data processing includes: multi-dimensional data label extraction, anomaly and weight identification, multi-source data adaptation and correlation, real-time event analysis, and maintenance of a real-time event experience database.

[0043] (1) The multi-dimensional data extraction includes: extracting the time, location and event dimensions from various real-time monitoring data respectively.

[0044] (2) The identification of anomalies and their weights includes: determining the weights based on the importance of 5G-R communication and the services it carries, classifying and labeling them accordingly, matching the labels based on the classification and labeling results, and determining whether there are any anomalies.

[0045] In this embodiment of the invention, the classification identification information may include measurement type, signaling type, service type, etc. The information in each category may differ in terms of time reference, granularity, and precision; and the location reference, granularity, and precision may also differ. The classification identification results can be used for subsequent label adaptation, thereby unifying all different classification identification information onto a label normalized coordinate system (see the following description for details). Figure 4 This is to allow data from different sources to corroborate each other.

[0046] Those skilled in the art will understand that abnormal situations can be classified according to the actual situation, for example, from minor to severe, such as: weak signal level (duration and affected mileage), poor signal quality (duration and affected mileage), high bit error rate, rescue handover, handover failure, communication interruption or interruption, service interruption or interruption, etc.

[0047] (3) The multi-source data adaptation and association includes: establishing associations between different data sources based on the normalized multidimensional data labels fed back in step 5.

[0048] (4) The real-time event analysis includes: on the basis of multi-source data adaptation and association, the event analysis is realized, that is, when an abnormal event occurs in the monitoring data of a certain data source, the data of the same time period and location interval in the monitoring data of other data sources are analyzed, and the real-time event analysis results are obtained by combining the analysis results of different monitoring data, that is, the analysis results of real-time monitoring data.

[0049] (5) The maintenance of the real-time event experience base includes: recording and classifying the analysis results of each real-time event.

[0050] Step 4: Non-real-time monitoring data processing: After preprocessing the non-real-time monitoring data, the corresponding multidimensional data is provided to Step 5. The data is then segmented and stored using the normalized multidimensional data labels fed back from Step 5. The association between different data sources is established based on the normalized multidimensional data labels. The data is then divided at different time scales. The data at each time scale is analyzed using both empirical and AI models to obtain the analysis results of the non-real-time monitoring data.

[0051] Preferably, the non-real-time monitoring data processing includes: preprocessing and data segmentation, multi-source data adaptation and correlation, short-term data organization, short-term data analysis, long-term data organization, long-term data analysis, and ultra-long-term data organization and ultra-long-term data analysis.

[0052] (1) The preprocessing and data segmentation include: preprocessing the non-real-time monitoring data and providing the corresponding multidimensional data to step 5, and performing data segmentation and storage in combination with the normalized multidimensional data labels fed back from step 5.

[0053] Those skilled in the art will understand that the process of transmitting information in a communication network is a layer-by-layer packaging process from the physical layer to the application layer (usually up to seven layers). The lower layers provide services to the higher layers, and each layer has its own identification and encapsulation rules. Preprocessing mainly involves parsing the required data parts to ensure the efficiency of the program. The data segmentation methods involved are relatively flexible, and users can choose according to their actual needs.

[0054] (2) The multi-source data adaptation and association includes: establishing associations between different data sources based on the normalized multidimensional data labels fed back in step 5.

[0055] (3) Short-term data organization, long-term data organization and ultra-long-term data organization, including: organizing data according to its time scale; among which, the time scale is divided into three categories, from smallest to largest, namely short-term, long-term and ultra-long-term.

[0056] (4) Short-term data analysis, long-term data analysis and ultra-long-term data analysis, including: analyzing data of the corresponding time scales through empirical models and AI models respectively, and combining the outputs of the two types of models to determine the final analysis results, and combining the final analysis results of all time scale data to obtain the analysis results of non-real-time monitoring data.

[0057] Step 5: Establish a normalized labeling system centered on 5G-R services and multidimensional data, and transfer the labels to Step 3 and Step 4 respectively.

[0058] Preferably, the multidimensional data includes three dimensions: time, location, and event. The three dimensions are normalized separately to establish a normalized labeling system.

[0059] (1) The time normalization processing method includes: selecting the time of the monitoring data in the specified data source as the reference time, calculating the time deviation Δt between the time of other data sources and the reference time, and performing time deviation compensation.

[0060] (2) The location normalization processing method includes: the location data label contains multiple types of location information, selects the specified type of location information as the reference location, calculates the distance deviation Δd between the location in each data source and the reference location, and performs distance deviation compensation; wherein, if the current data source does not contain a location, a synchronization algorithm is used to share the location in other data sources with the current data source.

[0061] In the aforementioned time and location normalization processes, a relatively reliable benchmark (i.e., reference time and reference location) can usually be determined first. For example, during time normalization, the system time of the 5G communication system can be used as the reference time. This allows for real-time monitoring of the deviation between the clocks of other systems and the 5G system time, compensating for cumulative or sudden deviations. During location normalization, the precise railway track circuit information (kilometer marker information) contained in the data interface PRI can be used as the reference location. Kilometer markers and latitude and longitude have a corresponding relationship. The latitude and longitude obtained by other monitoring systems from BeiDou or GPS fluctuate and cannot be used directly. They need to be corrected to the railway line in real time by combining the operating speed and kilometer markers. Alternatively, the base station location can be used as a backup reference location. Although the accuracy is lower than that of the kilometer markers, it can serve as a backup to improve the system's resilience. Of course, the above are only examples of reference time and reference location. In practical applications, users can adjust them according to their needs or experience.

[0062] (3) The event normalization processing method includes: defining various types of events, assigning a unique code to each defined event, identifying the event type through the unique code, and establishing the association between each event and time and location.

[0063] Preferably, the event normalization process includes the following steps:

[0064] (3.1) Event coding: Assign a unique code to each event.

[0065] (3.2) Event Attribute Description: Describe the specific attributes of the event.

[0066] (3.3) Event data standardization: Standardize the time data from different data sources, including: data format unification and data field standardization.

[0067] (3.4) Synchronize events with time.

[0068] (3.5) and associating with location.

[0069] Step 6: Combine the analysis results of the real-time monitoring data and the analysis results of the non-real-time monitoring data to output the analysis results of network situational awareness.

[0070] Preferably, the analysis results of the network situational awareness, which combine the analysis results of the real-time monitoring data and the analysis results of the non-real-time monitoring data, include: comprehensive scoring and comparison, report generation and push, prediction and alarm, and network optimization operation closed loop.

[0071] (1) Comprehensive scoring and comparison, including: multi-dimensional evaluation and comparison of the analysis results of the real-time monitoring data and the analysis results of the non-real-time monitoring data, that is: through the analysis results of the non-real-time monitoring data, a health range benchmark is established, the boundaries of the health status of each operating dimension of the 5G communication network under different scenarios are determined (i.e., different health statuses correspond to a set of boundary values), and the analysis results of the real-time monitoring data are compared with each boundary to obtain the comprehensive scoring result of the 5G communication network; and multi-dimensional evaluation and comparison of the analysis results of the real-time monitoring data and the analysis results of the non-real-time monitoring data in the 5G-R communication network before and after adjustment and optimization are also carried out to determine the effect after adjustment.

[0072] Those skilled in the art will understand that communication systems are limited by many factors inside and outside the system. Communication systems have scenario-based health ranges. During long-term operation, the health range is defined (generally described by a series of operating indicators). When certain performance characteristics of the communication system reach the health boundary, the system will compare the deviation amplitude, distribution characteristics, etc., to analyze the cause of the problem and trigger alarms.

[0073] (2) Report generation and delivery, including: generating and delivering analysis reports by combining the analysis results of real-time monitoring data, the analysis results of non-real-time monitoring data and the comprehensive scoring results.

[0074] (3) Prediction and alarm, including: combining the analysis results of non-real-time monitoring data to predict problems and analyze trends, and combining the comprehensive scoring results of 5G communication network to determine events in real-time monitoring data that exceed the health range benchmark and generate corresponding alarm prompts.

[0075] In this embodiment of the invention, the output of the above analysis reports, predictions and alarms are collectively referred to as the analysis results of network situational awareness; the above alarm prompts are hierarchical prompts, and the corresponding alarm level is determined according to the boundary crossing situation, thereby generating alarm prompts of the corresponding level.

[0076] (4) Network optimization operation closed loop, including: adjusting and optimizing the 5G-R communication network based on the analysis results of network situation awareness.

[0077] Preferably, the method further includes: a step of inputting external environmental data, wherein the external environmental data includes: geographic topography data, meteorological data and environmental disturbance data; the external environmental data is provided to steps 3 and 4 to correct the corresponding analysis results, and the external environmental data is also used for AI model training.

[0078] Preferably, the method further includes: the steps of AI model training and deployment; during training, a training set is constructed using the stored raw data, and the selected AI model is trained; during the training process, the AI ​​model is corrected by combining external environmental data; after training is completed, the performance of the AI ​​model is evaluated, and the AI ​​model is tuned based on the evaluation results; and the AI ​​model is deployed to the production environment for processing non-real-time monitoring data.

[0079] The above-mentioned solutions provided by the embodiments of the present invention are applicable to the comprehensive process evaluation of various quantitative changes leading to qualitative changes, the early analysis and warning of abnormal events, and the cause analysis of abnormal events (such as communication events, wireless signal coverage and quality, railway line scenarios, weather and noise, etc.). They are used to solve the automatic correlation and intelligent analysis of multiple data sources, greatly improving the analysis accuracy and work efficiency. In particular, they highlight the early warning of quantitative changes from shallow correlation to deep correlation and even qualitative changes in abnormal events, as well as the post-abnormal event correlation backtracking analysis of the cause of the event.

[0080] To more clearly demonstrate the technical solution and its effects provided by the present invention, the method provided by the embodiments of the present invention will be described in detail below with reference to specific examples.

[0081] like Figure 2 The diagram illustrates the overall framework of a 5G-R multi-monitoring data fusion detection network situational awareness analysis method provided by an embodiment of the present invention. This method framework utilizes various 5G-R dedicated communication network monitoring and detection data, such as... Figure 3 As shown, this includes: 5G communication network, 5G interface DPI, 5G northbound data, 5G transmission network management, 5G wireless probes, 5G drive test equipment, 5G RMS, antenna / tower monitoring, and data center environmental monitoring, etc.; and, it establishes relationships between data through three-dimensional labels (time, location, and event), such as... Figure 4 As shown; during the data analysis process, real-time and non-real-time data are processed and combined separately, such as... Figure 2 and Figure 5As shown, to adapt to the situational awareness of 5G-R multi-monitoring data fusion detection network accessed by AI, a data and label channel with the AI ​​module was designed, along with iterative optimization of the training model. From the outset, this invention considered ease of programming, rapid deployment, significantly improved system analysis accuracy and efficiency, and wide applicability. The following sections provide a detailed description of each part of the method framework.

[0082] Part A: Real-time monitoring data interface.

[0083] Real-time monitoring data mainly includes real-time dynamic data from data sources such as: 5G network northbound alarm data, 5G interface monitoring system data, 5G wireless probe data, 5G transmission network management system data, 5G network environmental monitoring data, antenna and tower monitoring data, 5G-R network northbound performance data, 5G-R road test equipment data, 5G-R network configuration parameters, 5G-R equipment ledger, MDT minimum road test terminal measurement and acquisition system data, CIR system vehicle-mounted equipment monitoring system data, and MCX railway business dedicated processing service data. Its main characteristic is that any anomaly in the communication network will inevitably lead to fluctuations in real-time dynamic data, making it the most sensitive area for situational awareness.

[0084] In this section, real-time monitoring data is quickly parsed, filtered, and partitioned for buffering. (1) For data that is encapsulated layer by layer, fast parsing is used, only parsing the layers and information that are directly related to network performance to ensure parsing efficiency. (2) Layers and information that are not directly related to network performance are filtered to simplify the data. (3) Partitioning for buffering is the format preparation and label preparation before entering the real-time monitoring data processing section E. Different data sources are processed through their own independent channels. The channels meet their respective data bandwidth and processing capabilities. The channels themselves can also check the connectivity status of the data sources. To ensure efficiency, this is generally done in memory.

[0085] In addition, the real-time monitoring data stream will be preserved losslessly as a file in Part C for data review or as input for AI training by combining labels.

[0086] Part B: Non-real-time monitoring data interface.

[0087] Non-real-time monitoring data mainly includes: 5G network northbound alarm data, 5G interface monitoring system data, 5G wireless probe data, 5G transmission network management system data, 5G network environmental monitoring data, antenna and tower monitoring data, 5G-R network northbound performance data, 5G-R road test equipment data, 5G-R network configuration parameters, 5G-R equipment ledger, MDT minimum road test terminal measurement and acquisition system data, CIR system vehicle-mounted equipment monitoring system data, MCX railway business dedicated processing service data, and other non-real-time dynamic data from various data sources. Its main characteristics are: the data is not collected and transmitted immediately during the collection, processing, and transmission process, or the data source is not the equipment involved in the work (indirect monitoring equipment), or the data fluctuations may not necessarily cause extreme events such as communication interruptions, but rather represent the potential for abnormal events in situational awareness.

[0088] This section describes interface adaptation for non-real-time dynamic data. Non-real-time monitoring data is typically parsed, cleaned, and stored within its respective system. It usually utilizes database sharing, with data packaged into shared directories or pushed via triggers based on fixed time intervals or file sizes. Different data sources utilize their own independent channels, each with its own bandwidth and read capabilities.

[0089] In addition, the data stream of non-real-time monitoring data will be preserved losslessly as a file in Part C for data review or as input for AI training by combining labels.

[0090] Part C: Full lifecycle management of raw data.

[0091] Raw data lifecycle management primarily involves backing up the raw data. Since no storage is unlimited, accessing massive amounts of data is extremely resource-intensive. This invention employs data source type, scope, and time-segmented storage; it can adapt to various data types, such as binary data and files; the data value assessment strategy includes: defining the weight of monitoring data for different scenarios, dynamically adjusting the importance weight of data through multiple data time sliding windows, and downgrading the weight of historical data through different levels of network adjustments.

[0092] Part D: Tag adaptation.

[0093] Tag adaptation requires establishing a three-dimensional, unified tag system centered on 5G-R services and based on time, location, and events (e.g., ...). Figure 4 To analyze and ensure the resources, signaling, and quality of services in various scenarios, such as... Figure 6As shown, Part E (real-time monitoring data processing) and Part F (non-real-time monitoring data processing) interact with Part D to transmit tags. Parts E and F provide the time, location, and event information in the data. Real-time and non-real-time monitoring data come from diverse sources, involving multiple stages of the 5G-R communication network. The clock start points, sample granularity, accuracy deviations, and formats vary across systems, and some monitoring data lacks location information. Events belong to different communication protocol layers. Therefore, the three dimensions of service-based 5G-R—time, location, and event—need to be normalized, such as... Figure 3 As shown.

[0094] (D1) Time Normalization Scheme. Numerous monitoring systems in a communication network operate with their own system clocks. These clocks typically have different reference points, transmission delays, accuracy, formats, and calibration methods. The first step in joint detection and monitoring is to establish time synchronization and maintain time deviation compensation between systems. This invention selects the time of monitoring data from a designated system data source as the reference time. The time of other systems deviates from the reference time by a time deviation Δt. Δt is a variable; it is calculated and monitored for deviation compensation. When Δt becomes too large (i.e., exceeds a set value), an alarm is triggered to reset the clock of the system with the excessive Δt deviation.

[0095] For example, the time of the 5G-R interface monitoring system (because the information of each communication link is configured with complete and high-precision system time) can be selected as the reference time.

[0096] (D2) Location Normalization Scheme. All communication links have corresponding locations. The location of wired networks in communication is generally very clear and accurate. The difficulty in positioning lies in the wireless network portion along the railway line. The types of location information collected along the line include: kilometer markers based on track circuits (high accuracy, large intervals), latitude and longitude reported by terminals (accuracy affected by terrain, such as mountain tunnels), and base station coverage (wireless signal fluctuations).

[0097] This invention selects a specific type of location information as a reference position. Other types of locations have a distance deviation Δd from the reference position. Δd is a variable. The vehicle's movement includes kilometer markers and speed information passing through kilometer markers. The position to fill the gaps between kilometer markers is calculated. For example, kilometer markers can be selected as reference positions (because the information from each communication link can be anchored to the relative position of the nearest kilometer marker).

[0098] Specifically: If the data source does not have location information, the location information is shared with systems that do not have location information through an inter-system synchronization algorithm (supplemented by referring to the time base).

[0099] Special: Figure 4The horizontal axis shown primarily includes the directions of communication and vehicle movement. For example, communication includes uplink and downlink, and the directions mainly include uplink and downlink. Similarly, the directions of vehicle movement also mainly include uplink and downlink. Wireless performance varies greatly in different directions.

[0100] (D3) Event Normalization Scheme. Communication networks contain a wide variety of events, including user behavior events, network fault events, and equipment maintenance events. To achieve unified event management, it is first necessary to clearly define and classify various events. For example, user behavior events can be categorized as dwell, access, hold, move, terminate, and release; network fault events can be categorized as network faults, wireless network interruptions, transmission network interruptions, and terminal faults. For each defined event, a unique code needs to be assigned to ensure the event's uniqueness and identifiability within the system, including the following steps:

[0101] (D3.1) Event coding: Assign a unique code to each event for fast retrieval and processing.

[0102] (D3.2) Event Attribute Description: Describe the specific attributes of the event, including the time, location, triggering conditions, and scope of influence of the event.

[0103] (D3.3) Event Data Standardization: Event data from different data sources needs to be standardized. Different systems may generate event data with variations in format and structure. To achieve unified event analysis, this data needs to be standardized. This mainly includes: (3.31) Data Format Standardization: Converting event data from different systems into a unified format, such as JSON or XML. (3.32) Data Field Standardization: Standardizing each field in the event data to ensure that each field has the same meaning and data type across different systems.

[0104] (D3.4) Synchronize events with time: To ensure the accuracy of event analysis, events from different data sources need to be synchronized in time. Here, we can refer to the established normalized time system to unify the timestamps of all events to the clock of the 5G-R interface monitoring system and perform time deviation compensation.

[0105] (D3.5) Associating with Location: Location information of an event is crucial for subsequent analysis and processing. By normalizing the location system, events are associated with specific location information. For example, a base station outage event can be associated with the specific location of the base station, thereby helping to locate and resolve network faults.

[0106] Based on the above scheme, the following can be achieved: (1) Multi-dimensional parallel adaptation: Parallel adaptation of multiple monitoring systems in terms of communication events, monitoring time, line location and railway business dimensions, and construction of an integrated set of railway communication services based on 5G-R network; (2) Dynamic maintenance of time variables: The time offset Δt monitored by multiple detection systems is a variable that needs a mechanism to continuously monitor its range of change and implement compensation for offset Δt; (3) Dynamic maintenance of location variables along the line: The location offset Δd monitored by multiple detection systems along the railway is a variable that needs a mechanism to continuously monitor its range of change and implement compensation for offset Δd; (4) Dynamic maintenance of multiple measurement deviations: The differences caused by the sensitivity of different monitoring systems, radio frequency connection methods or usage scenarios need a mechanism to continuously monitor their range of change and implement correction of deviation Δp.

[0107] Part E: Real-time monitoring data processing.

[0108] This section mainly includes: multi-dimensional data label extraction, anomaly and weight identification, multi-source data adaptation and association, real-time event analysis, and maintenance of the real-time event experience library.

[0109] (E1) Multidimensional data label extraction is the process of "extracting" time, location and event from the data. If a certain dimension is missing, it can be ignored.

[0110] (E2) Anomaly and its weight identification: This part requires classification and labeling based on the importance of 5G-R communication and the services it carries. Both of the above are sent to the tag adaptation module.

[0111] (E3) Multi-source data adaptation and association are based on the system label normalized coordinate system in Part D (e.g., ... Figure 3 As shown), calibrating data from different monitoring systems forms a data correlation across the entire system (e.g. Figure 2 (As shown).

[0112] (E4) Real-time event analysis is in the context of... Figure 2 and 3 Under the architecture of a system-labeled normalized coordinate system and data fusion from multiple monitoring systems, comprehensive event analysis can be completed quickly. For example, if a 5G wireless probe detects a sharp 7dB drop in wireless signal strength in a specific railway line section during a certain period, it will initiate a scan of relevant measurements and signaling from other systems during the same time period (and extended time periods before and after), and within the same location range (with appropriate extension of the range). This includes checking for alarms from relevant equipment, antenna azimuth shifts, alarms from the corresponding cell equipment room's environmental monitoring, and the signaling (abnormal releases and abnormal handovers) and measurements (abnormal strength and quality) of all terminals monitored by the 5G-R interface at that time period and location. Through cross-verification of monitoring data from different systems, a more definitive problem description and analysis result can be obtained.

[0113] (E5) The real-time event experience library maintenance can categorize and record each event, including: network device fault files, terminal device performance logs, noise characteristics, coverage and quality, signaling and triggering, and other historical descriptions, as well as effective handling methods.

[0114] Part F: Non-real-time monitoring data processing.

[0115] This section mainly includes: preprocessing and data segmentation, multi-source data adaptation and association, short-term data organization, short-term data analysis, long-term data organization, long-term data analysis, ultra-long-term data organization, and ultra-long-term data analysis.

[0116] (F1) Preprocessing and Data Segmentation: Non-real-time monitoring data comes from diverse sources and involves multiple aspects of the 5G-R communication network, such as... Figure 2 As shown, each stage is monitored by multiple monitoring systems, categorized according to the stage; there are network-side monitoring and terminal monitoring, categorized according to the reporting object; communication monitoring includes service content, communication events, and channel measurements, categorized according to the monitored content. Finally, tag addition and data segmentation and storage are completed through interaction with part D.

[0117] (F2) Multi-source data adaptation and correlation: Based on the 5G-R network communication logic, cross-data source connections are established using tags as anchors to comprehensively reproduce communication events and channel measurements in the time and space dimensions surrounding communication services. 5G-R is divided into wired and wireless communication. The wired part of 5G-R is fully IP-based, focusing on the rationality of possible node routing, bandwidth load, and latency. The wireless part of 5G-R needs to withstand complex scenarios along railway lines, continuously monitoring signal strength, quality, interference, and signal overlap while detecting bandwidth load.

[0118] (F3) To address the diversity and complexity of scenarios, the data is divided into multiple time scales and analyzed separately. The changes brought about by scenarios to wireless communication are essentially functions on the time axis; different scenario types correspond to different time scales. For example, wireless signal fading in space is divided into fast fading and slow fading: fast fading occurs in milliseconds, while slow fading may last for several seconds or tens of seconds; the impact of rain, snow, and dense fog on wireless signals depends on the meteorological cycle; the obstruction and attenuation of wireless signals by vegetation growth cycles and meteorological changes need to be considered in relation to seasonal cycles; changes in terrain and landforms often accumulate over years; and the lifespan of equipment varies depending on the working environment. Different time analysis periods are set in the non-real-time monitoring data processing to address the changing cycles of different scenarios.

[0119] Based on this, in this embodiment of the invention, multiple time periods are assigned to the data, generally divided into three categories: short-term data organization and analysis (minutes, hours, days), long-term data organization and analysis (weeks, months, quarters), and ultra-long-term data organization and analysis (years, many years). The analysis for different periods includes two parts: an empirical model and an AI model. The empirical model is a mathematical description of the distribution of multi-dimensional monitoring data during the stable operation of each link of the communication system in various scenarios across the entire timeframe. The AI ​​model is a specially trained "brain" capable of learning and predicting based on the input data.

[0120] AI models are data-driven: trained on massive amounts of data, they learn patterns and rules within that data. For example, language models learn numerous features such as grammar and semantics in text. They also exhibit automatic learning capabilities: automatically extracting features from data without requiring manual design of each feature. Furthermore, they possess highly complex structures: AI models like deep neural networks have many hidden layers, enabling them to handle highly complex nonlinear problems. AI models excel at handling complex tasks: performing exceptionally well in image recognition, speech recognition, and natural language processing. For instance, in medical image disease recognition, they can detect subtle features imperceptible to the human eye. They are also highly scalable: given sufficient data and computing resources, their performance can be continuously optimized and expanded. They have strong generalization capabilities: well-trained models can make reasonable predictions about new data. However, building AI models requires massive amounts of data: insufficient data significantly reduces model performance; they consume significant computing resources: training and deploying complex AI models requires powerful computing equipment and substantial energy; and they have poor interpretability: it is difficult to understand how the model makes decisions, especially black-box models in deep learning.

[0121] Empirical models are built upon experience: they are models constructed by people based on past experience, observations, and theoretical knowledge; parameters are often determined manually: the parameters of the model are usually determined through human experience and statistical methods. Empirical models are highly interpretable: because they are built upon experience and known theories, their principles and decision-making processes are relatively easy to understand; they are efficient in familiar domains: in domains with sufficient experiential knowledge, models can be quickly built and applied; and they have relatively flexible data requirements: they do not require the massive amounts of data required for AI models and can be built even with limited data. However, empirical models struggle to handle complex, high-dimensional data: their ability to process complex, high-dimensional data (such as massive signal distribution data) is limited; their generalization ability is limited: when encountering new situations that differ significantly from experience, the model's accuracy may decrease; and they are limited by human cognition: the quality of the model highly depends on human experiential knowledge and may overlook some unknown but important factors.

[0122] Therefore, it is necessary to combine the advantages and disadvantages of experience models and AI models in parallel installation and operation for railway communications with extremely high safety requirements.

[0123] Part G: Conclusion Report Output.

[0124] This section mainly includes: comprehensive scoring and comparison, report generation and delivery (routine reports, special analysis reports), prediction and alerts, and closed-loop network optimization operations.

[0125] (G1) Comprehensive scoring and comparison: 5G-R communication consists of multiple components and carries a variety of services, so comprehensive scoring requires a series of evaluation indicators for multi-dimensional evaluation.

[0126] (G2) Report generation and delivery (routine reports, special analysis reports): Various professional and management departments of railway 5G-R communication network operation and maintenance need a series of routine reports such as daily and weekly reports, and special analysis reports are for specific issues.

[0127] (G3) Prediction and Alerts: Predict and alert on potential problems using health experience models or AI models.

[0128] (G4) Network optimization operation closed loop: During the operation and maintenance of 5G-R communication network, the network is adjusted and optimized, and the indicators before and after the adjustment are iteratively tracked.

[0129] The data output in the conclusion report comes from real-time monitoring data processing and non-real-time monitoring data processing. The report results will be compared with the AI ​​model parameter optimization in AI model training.

[0130] Based on the above scheme, multi-level situational awareness can be achieved, such as... Figure 5 As shown: Based on the multi-level situational awareness and multi-dimensional evaluation of 5G-R services, 5G-R carries many services with different requirements for communication resources and quality, so each service has different situational awareness requirements.

[0131] In this embodiment of the invention, an emergency degradation handling mechanism is also introduced: when data is missing, data mismatch occurs, or some monitoring systems malfunction, the overall monitoring and analysis will not be completely rendered ineffective, and some monitoring tasks can still be completed, depending on the extent of the damage.

[0132] H section: External environment data input.

[0133] In this embodiment of the invention, the external environmental data mainly includes: geographical and geomorphological data, meteorological data, environmental disturbances, etc.

[0134] (H1) Geographic and geomorphological data: Generally connected to three-dimensional geographic information digital maps, online maps have high timeliness updates.

[0135] (H2) Meteorological data: It is necessary to connect with the national meteorological data release platform to obtain authoritative and accurate data.

[0136] (H3) Environmental interference data: Interference data mainly comes from dedicated wireless monitoring probes, frequency sweepers, and 5G-R communication operation and maintenance of subcarrier noise floor.

[0137] In this embodiment of the invention, external environmental data participates in the real-time monitoring data processing and non-real-time monitoring data processing parts to correct the analysis results, and the external environmental data enters the AI ​​model training part to enrich the AI ​​training data.

[0138] Part I: AI Model Training.

[0139] This section mainly includes: AI data interface, AI model selection, AI model training, AI model tuning and deployment, AI model evaluation, and correction of external factors.

[0140] (I1) AI Data Interface: To meet the needs of AI training with massive amounts of data, it is necessary to ensure lossless data transmission, high-speed transmission, and accurate labeling to feed the AI ​​model. This mainly includes:

[0141] (I1.1) Data interfaces include: RESTful API, which interacts with AI models via HTTP requests; common operations include data upload, model inference requests, and result retrieval. WebSocket, used for real-time data transmission and bidirectional communication, suitable for scenarios requiring low latency and high-frequency interaction. gRPC, a high-performance Remote Procedure Call (RPC) framework that supports multiple languages ​​and is suitable for AI model calls in large-scale distributed systems. GraphQL, a flexible API query language that allows clients to precisely request the data they need, reducing unnecessary data transmission.

[0142] (I1.2) Data formats include: JSON, a lightweight data exchange format widely used for data transfer in API calls; XML, a traditional data format, less popular than JSON but still used in some scenarios; Protocol Buffers (Protobuf), an efficient serialization format developed by Google, commonly used in gRPC communication; and CSV / TSV, simple text formats for structured data, commonly used for data import and export.

[0143] (I1.3) Data preprocessing includes: data cleaning (removing noisy data, handling missing values, and correcting erroneous data); data standardization / normalization (converting data to a uniform scale for easier model processing); feature engineering (extracting useful features from the raw data to enhance model performance); and data augmentation (expanding the training dataset by generating new data samples, such as image rotation and text synonym replacement).

[0144] (I1.4) Model inference includes: Online inference, which processes user requests in real time, typically through API calls; Batch inference, which processes large amounts of data at once, suitable for offline tasks or background processing; and Edge computing, which performs model inference locally on the device, reducing reliance on the cloud and suitable for low-latency requirements.

[0145] (I1.5) Security includes: authentication and authorization, ensuring that only authorized users or systems can access the AI ​​model; data encryption, using encryption technology during data transmission and storage to protect sensitive data; and model protection, preventing the model from being maliciously attacked or abused, such as adversarial sample attacks and model reverse engineering.

[0146] (I1.6) Performance optimizations include: model compression, which reduces model size and computational load through techniques such as quantization and pruning, thereby improving inference speed; distributed computing, which utilizes multiple machines to process data and model inference in parallel, thereby improving processing capacity; and caching mechanisms, which cache the results of frequently requested data, reducing redundant calculations and improving response speed.

[0147] (I1.7) Development tools and frameworks include: TensorFlow Serving: a tool for deploying and managing TensorFlow models. PyTorch Serve: a model serving framework for PyTorch that supports rapid deployment and inference. Kubernetes: used for containerized deployment and management of AI models, supporting elastic scaling and high availability. Docker: used for containerized AI models, simplifying deployment and environment management.

[0148] (I1.8) Data storage includes: databases, used to store training data, model parameters, inference results, etc. Common databases include relational databases (such as MySQL, PostgreSQL) and NoSQL databases (such as MongoDB, Cassandra). Object storage is used to store large-scale data files, such as training datasets, model files, etc. Common object storage services include AWS S3, Google Cloud Storage, etc.

[0149] (I1.9) Logging and monitoring include: logging, which records key information during API calls and model inference to facilitate troubleshooting and performance analysis; and monitoring, which monitors the system's operational status in real time, including API request volume and model activity.

[0150] (I2.1) AI Model Selection: Based on the business analysis objectives and the needs of the data source characteristics, select an appropriate AI model. Commonly used models include:

[0151] (I2.2) Classification Models: Classification models are used to assign data to predefined categories. Common tasks include image classification, text classification, and spam detection.

[0152] (I2.3) Regression Models: Regression models are used to predict continuous values. Common tasks include house price forecasting and sales forecasting.

[0153] (I2.4) Clustering Models: Clustering models are used to divide data into different groups. Common tasks include customer segmentation and image segmentation.

[0154] (I2.5) Recommendation Models: Recommendation models are used to provide personalized recommendations to users. Common tasks include product recommendations and video recommendations.

[0155] (I2.6) Natural Language Processing Models (NLP Models): NLP models are used to process and analyze text data. Common tasks include text classification, sentiment analysis, machine translation, etc.

[0156] (I2.7) Computer Vision Models: Computer vision models are used to process image and video data. Common tasks include image classification, object detection, and image segmentation.

[0157] (I2.8) Time Series Forecasting Models: Time series forecasting models are used to process time series data. Common tasks include stock price forecasting and weather forecasting.

[0158] (I2.9) Generative Models: Generative models are used to generate new data samples. Common tasks include image generation and text generation.

[0159] (I3) AI Training Process: The full dataset and the defined model are the inputs for model training. This is a core process in machine learning and deep learning, through which the model learns from the data and optimizes its performance to accomplish a specific task. The basic goal of training is to enable the model to make accurate predictions or decisions on new data. The training process includes the following five steps:

[0160] (I3.1) Forward Propagation: The input data is passed through the model to calculate the predicted value.

[0161] (I3.2) Calculate the loss: Use the loss function to calculate the difference between the predicted value and the true value.

[0162] (I3.3) Backward Propagation: Calculates the gradient of the loss function with respect to the model parameters.

[0163] (I3.4) Parameter update: Update the model parameters using optimization algorithms to reduce the loss function.

[0164] (I3.5) Iteration: Repeat the above steps until the model converges or reaches the predetermined number of training rounds.

[0165] (I4) AI Model Evaluation: After training, the model's performance needs to be evaluated periodically to ensure that it is not overfitting or underfitting. Commonly used evaluation metrics include:

[0166] (I4.1) Accuracy: The proportion of correctly predicted samples out of the total number of samples in a classification task.

[0167] (I4.2) Precision: The proportion of samples correctly predicted as positive in a classification task out of the total number of samples correctly predicted as positive.

[0168] (I4.3) Recall: The proportion of samples correctly predicted as positive in a classification task to the actual number of samples that are positive.

[0169] (I4.4) F1 Score: The harmonic mean of precision and recall.

[0170] (I4.5) Mean Squared Error (MSE): The average squared difference between the predicted and actual values ​​in a regression task.

[0171] (I4.6)R 2 R-squared score: The goodness of fit of the regression line in a regression task.

[0172] (I5) AI Model Tuning: Based on the evaluation results, adjust the model's hyperparameters, such as learning rate, batch size, and number of hidden layer nodes, to optimize model performance. Common tuning methods include:

[0173] (I5.1) Grid Search: Searches within a predefined hyperparameter grid.

[0174] (I5.2) Random Search: Randomly selects a hyperparameter space for the search.

[0175] (I5.3) Bayesian optimization: Optimize hyperparameters using Bayesian methods.

[0176] (I6) AI Model Deployment: After training, the model is deployed to the production environment for practical applications, and the AI ​​prediction results are compared with traditional analysis results. The deployment process needs to consider the model's performance, scalability, and maintainability.

[0177] (I7) External factor correction: External factor data is also used as part of AI training to correct the model.

[0178] Preferably, the embodiments of the present invention also establish an integrated tag: the events, time, location and services of the 5G-R network communication process are fully tagged to facilitate manual analysis or AI model training. The tag includes two parts: identification class and attribute class.

[0179] Example 2

[0180] This invention also provides an analysis system for 5G-R multi-monitoring data fusion detection network situational awareness. This system is used to implement the method provided in the foregoing embodiments, and can also be found in... Figure 2 The system mainly includes:

[0181] The multi-source data acquisition and segmentation unit is used to acquire multi-source data from the 5G-R communication network and segment it into real-time monitoring data and non-real-time monitoring data.

[0182] The raw data lifecycle management unit is used to store the real-time monitoring data and non-real-time monitoring data as raw data and manage them throughout their lifecycle, as well as for AI model training.

[0183] The real-time monitoring data processing unit is used for real-time monitoring data processing, including: extracting real-time monitoring data from multiple dimensions, classifying and labeling it and providing it to step 5, and then performing data analysis based on the normalized multi-dimensional data labels fed back from step 5 to obtain the analysis results of the real-time monitoring data.

[0184] The non-real-time monitoring data processing unit is used for non-real-time monitoring data processing, including: preprocessing the non-real-time monitoring data and providing the corresponding multidimensional data to step 5; combining the normalized multidimensional data labels fed back from step 5 to perform data segmentation and storage; establishing the association between different data sources based on the normalized multidimensional data labels; dividing the data according to different time scales; and combining the empirical model and the AI ​​model to perform data analysis on the data at each time scale to obtain the analysis results of the non-real-time monitoring data.

[0185] The tag adaptation unit is used to establish a normalized tag system based on 5G-R services and multidimensional data, and to transfer tags in steps 3 and 4 respectively.

[0186] The conclusion report output unit is used to integrate the analysis results of the real-time monitoring data and the analysis results of the non-real-time monitoring data to output the analysis results of network situational awareness.

[0187] Since the main technical details of the various parts involved in the system have been described in detail in the previous embodiments, they will not be repeated here.

[0188] Through the above description of the embodiments, those skilled in the art can clearly understand that the above embodiments can be implemented by software, or by using software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solutions of the above embodiments can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.), including several instructions to cause a computer device (such as a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0189] Those skilled in the art will understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the system can be divided into different functional modules to complete all or part of the functions described above.

[0190] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims. The information disclosed in the background section is intended only to enhance the understanding of the overall background technology of the present invention and should not be construed as an admission or implication in any way that such information constitutes prior art known to those skilled in the art.

Claims

1. A method for analyzing the situational awareness of a 5G-R multi-monitoring data fusion detection network, characterized in that, include: Step 1: Acquire multi-source data from the 5G-R communication network and divide it into real-time monitoring data and non-real-time monitoring data; Step 2: Store the real-time monitoring data and non-real-time monitoring data as raw data and manage them throughout their entire lifecycle, as well as use them for AI model training; Step 3: Real-time monitoring data processing: Extract data from multiple dimensions of the real-time monitoring data, classify and label it, and then provide it to Step 5. Then, perform data analysis based on the normalized multi-dimensional data labels fed back from Step 5 to obtain the analysis results of the real-time monitoring data. Among them, the classification and labeling results include: measurement, signaling and business. Step 4: Non-real-time monitoring data processing: After preprocessing the non-real-time monitoring data, the corresponding multidimensional data is provided to Step 5. The data is segmented and stored in combination with the normalized multidimensional data labels fed back from Step 5. The association between different data sources is established based on the normalized multidimensional data labels. The data is then divided according to different time scales. The data at each time scale is analyzed by combining the empirical model and the AI ​​model to obtain the analysis results of the non-real-time monitoring data. Step 5: Establish a normalized tagging system centered on 5G-R services and multidimensional data, and transfer the normalized multidimensional data tags from Steps 3 and 4 respectively. The multidimensional data includes three dimensions: time, location, and event. Normalize each of these three dimensions to establish a normalized tagging system containing three dimension tags: time, location, and event. The time normalization process includes: selecting a time from a specified data source as a reference time, calculating the time deviation between the time from other data sources and the reference time, and compensating for the time deviation. The location normalization process includes: selecting… Select a specified type of location information as a reference location, calculate the distance deviation between the location in each data source and the reference location, and perform distance deviation compensation; wherein, if the current data source does not contain a location, a synchronization algorithm is used to share the location in other data sources with the current data source; the event normalization processing method includes: defining various types of events, assigning a unique code to each defined event, identifying the event type through the unique code, and establishing the association between each event and time and location; step 6, combining the analysis results of the real-time monitoring data and the analysis results of the non-real-time monitoring data, and outputting the analysis results of network situational awareness.

2. The analysis method for 5G-R multi-monitoring data fusion detection network situational awareness according to claim 1, characterized in that, The 5G-R communication network multi-source data includes: 5G network northbound alarm data, 5G interface monitoring system data, 5G wireless probe data, 5G transmission network management system data, 5G network environmental monitoring data, antenna and tower monitoring data, 5G-R network northbound performance data, 5G-R road test equipment data, 5G-R network configuration parameters, 5G-R equipment ledger, MDT minimum road test terminal measurement and acquisition system data, CIR system on-board equipment monitoring system data, and MCX railway business dedicated processing service data.

3. The analysis method for 5G-R multi-monitoring data fusion detection network situational awareness according to claim 1, characterized in that, The real-time monitoring data processing includes: multi-dimensional data label extraction, anomaly and weight identification, multi-source data adaptation and correlation, real-time event analysis, and maintenance of a real-time event experience database; wherein: The multi-dimensional data extraction includes: extracting time, location, and event dimensions from various types of real-time monitoring data respectively; The anomaly and its weight identification includes: determining the weight based on the importance of 5G-R communication and the services it carries, classifying and labeling them accordingly, matching the labels based on the classification and labeling results, and determining whether there are any anomalies. The multi-source data adaptation and association includes: establishing associations between different data sources based on the normalized multidimensional data labels fed back in step 5; The real-time event analysis includes: based on multi-source data adaptation and correlation, event analysis is realized, that is, when an abnormal event occurs in the monitoring data of a certain data source, the data of the same time period and location range in the monitoring data of other data sources are analyzed, and the real-time event analysis results are obtained by combining the analysis results of different monitoring data, that is, the analysis results of real-time monitoring data. The maintenance of the real-time event experience base includes recording and classifying the analysis results of each real-time event.

4. The analysis method for 5G-R multi-monitoring data fusion detection network situational awareness according to claim 1, characterized in that, The non-real-time monitoring data processing includes: preprocessing and data segmentation, multi-source data adaptation and correlation, short-term data organization, short-term data analysis, long-term data organization, long-term data analysis, and ultra-long-term data organization and ultra-long-term data analysis; wherein: The preprocessing and data segmentation include: preprocessing the non-real-time monitoring data and providing the corresponding multidimensional data to step 5, and performing data segmentation and storage in combination with the normalized multidimensional data labels fed back from step 5. The multi-source data adaptation and association includes: establishing associations between different data sources based on the normalized multidimensional data labels fed back in step 5; The time scales are divided into three categories, from smallest to largest: short-term, long-term, and ultra-long-term. Short-term data organization, long-term data organization, and ultra-long-term data organization include: organizing data according to its time scale; Short-term data analysis, long-term data analysis, and ultra-long-term data analysis include: analyzing data at the corresponding time scales using empirical models and AI models respectively, combining the outputs of the two types of models to determine the final analysis results, and integrating the final analysis results of data at all time scales to obtain the analysis results of non-real-time monitoring data.

5. The analysis method for 5G-R multi-monitoring data fusion detection network situational awareness according to claim 1, characterized in that, The steps of event normalization include: Event coding: Assign a unique code to each event; Event Attribute Description: Describes the specific attributes of the event; Event data standardization: Standardize the time data from different data sources, including: unifying data format and standardizing data fields; Synchronize events with time and associate them with location.

6. The analysis method for 5G-R multi-monitoring data fusion detection network situational awareness according to claim 1, characterized in that, The analysis results of the combined real-time and non-real-time monitoring data output the network situational awareness analysis results, including: comprehensive scoring and comparison, report generation and push, prediction and alarm, and network optimization operation closed loop; wherein: The comprehensive scoring and comparison includes: multi-dimensional evaluation and comparison of the analysis results of the real-time monitoring data and the analysis results of the non-real-time monitoring data, namely: establishing a health range benchmark through the analysis results of the non-real-time monitoring data, determining the boundaries of the health status of each operational dimension of the 5G communication network under different scenarios, and comparing the analysis results of the real-time monitoring data with each boundary to obtain the comprehensive scoring result of the 5G communication network; and multi-dimensional evaluation and comparison of the analysis results of the real-time monitoring data and the analysis results of the non-real-time monitoring data in the 5G-R communication network before and after adjustment and optimization to determine the effect of the adjustment. Report generation and delivery include: generating and delivering analysis reports by combining analysis results from real-time monitoring data, analysis results from non-real-time monitoring data, and comprehensive scoring results; Prediction and alerting include: predicting problems and analyzing trends by combining the analysis results of non-real-time monitoring data, and identifying events in real-time monitoring data that exceed the health range benchmark by combining the comprehensive scoring results of 5G communication networks, and generating corresponding alert prompts; the above analysis reports, predictions and alerts are collectively referred to as the analysis results of network situational awareness. The network optimization operation loop includes: adjusting and optimizing the 5G-R communication network based on the analysis results of network situational awareness.

7. The analysis method for 5G-R multi-monitoring data fusion detection network situational awareness according to claim 1, characterized in that, The method further includes a step of inputting external environmental data, which includes geographical topography data, meteorological data, and environmental disturbance data; the external environmental data is provided to steps 3 and 4 to correct the corresponding analysis results; and the external environmental data is also used for AI model training.

8. The analysis method for 5G-R multi-monitoring data fusion detection network situational awareness according to claim 1 or 7, characterized in that, The method also includes: AI model training and deployment steps; during training, a training set is constructed using stored raw data to train the selected AI model; during training, the AI ​​model is corrected by combining external environmental data; after training, the performance of the AI ​​model is evaluated, and the AI ​​model is tuned based on the evaluation results; and the AI ​​model is deployed to the production environment for processing non-real-time monitoring data.

9. A 5G-R multi-monitoring data fusion detection network situational awareness analysis system, characterized in that, To implement the method according to any one of claims 1 to 8, comprising: The multi-source data acquisition and segmentation unit is used to acquire multi-source data from the 5G-R communication network and segment it into real-time monitoring data and non-real-time monitoring data. The raw data lifecycle management unit is used to store the real-time monitoring data and non-real-time monitoring data as raw data and manage them throughout their lifecycle, as well as for AI model training. The real-time monitoring data processing unit is used for real-time monitoring data processing, including: extracting real-time monitoring data from multiple dimensions, classifying and labeling it before providing it to step 5, and then performing data analysis based on the normalized multi-dimensional data labels fed back from step 5 to obtain the analysis results of the real-time monitoring data; among which, the classification and labeling results include: measurement category, signaling category and business category. The non-real-time monitoring data processing unit is used for non-real-time monitoring data processing, including: preprocessing the non-real-time monitoring data and providing the corresponding multidimensional data to step 5; combining the normalized multidimensional data labels fed back from step 5 to perform data segmentation and storage; establishing the association between different data sources based on the normalized multidimensional data labels; dividing the data according to different time scales; and combining the empirical model and the AI ​​model to perform data analysis on the data at each time scale to obtain the analysis results of the non-real-time monitoring data. The tag adaptation unit is used to establish a normalized tag system based on 5G-R services and multidimensional data, and to transmit the normalized multidimensional data tags in steps 3 and 4 respectively. The multidimensional data includes three dimensions: time, location, and event. The three dimensions are normalized to establish a normalized tag system containing three dimension tags: time, location, and event. The time normalization process includes: selecting a time from a specified data source as a reference time, calculating the time deviation between the time from other data sources and the reference time, and compensating for the time deviation. The location normalization process includes: selecting a specified type of location information as a reference location, calculating the distance deviation between the location in each data source and the reference location, and compensating for the distance deviation. If the current data source does not contain a location, a synchronization algorithm is used to share the location from other data sources with the current data source. The event normalization process includes: defining various types of events, assigning a unique code to each defined event, identifying the event type through the unique code, and establishing the association between each event and time and location. The conclusion report output unit is used to integrate the analysis results of the real-time monitoring data and the analysis results of the non-real-time monitoring data to output the analysis results of network situational awareness.

Images