Interference source identification method, device, medium, and product

Abstract

This application discloses an interference source identification method, device, medium, and product, belonging to the field of wireless network communication technology. The interference source identification method provided by this application includes: acquiring performance data and alarm data of the interfered cell; performing feature analysis on the interfered cell based on the performance data to identify and cluster interference areas caused by clock synchronization anomalies; performing cluster analysis on alarm sources related to clock synchronization anomalies based on the alarm data of the interfered cell to obtain at least one candidate alarm source region; matching the interference region with the at least one candidate alarm source region, and determining the final interference source location based on the matching result.

Description

Technical Field

[0001] This application belongs to the field of wireless network communication technology, and specifically relates to a method, device, medium and product for identifying interference sources. Background Technology

[0002] With the rapid development of wireless communication networks, especially the widespread deployment of TDD (Time Division Duplex) systems, network quality has become a key focus for telecom operators. TDD systems have stringent requirements for clock synchronization, and base stations typically rely on external clock sources such as GPS for synchronization. However, when a base station experiences problems such as GPS lock-out or satellite card malfunctions, leading to clock out-of-synchronization, the downlink transmission signal from the out-of-synchronization base station can severely interfere with the uplink reception of surrounding synchronized base stations, creating widespread regional interference and significantly impacting user experience in voice calls and data services.

[0003] Currently, traditional troubleshooting methods for interference caused by clock synchronization failure mainly rely on the personal experience of engineers. The specific process is as follows: engineers first sift through a large amount of call statistics data and alarm information from the network management system, and then perform manual analysis and judgment. After initially determining that the interference may originate from GPS synchronization failure, engineers need to carry equipment such as frequency sweepers and directional antennas to the site to conduct frequency sweep tests at multiple high points in the affected area, and determine the specific location of the interference source through cross-location.

[0004] However, due to the limitations of manual analysis and judgment of interference sources, related technologies suffer from low efficiency and poor accuracy in interference source identification. This is because manually processing massive amounts of network management data is time-consuming and labor-intensive. From data screening and analysis to on-site troubleshooting and location, the entire process typically takes several days, resulting in slow response and inability to quickly respond to and eliminate interference, leading to inefficiency and prolonged degradation of the wireless communication network. Furthermore, the identification of interference types and the judgment of interference sources are highly dependent on the individual experience of engineers. Engineers of different skill levels may produce vastly different analysis results, leading to poor accuracy in locating interference sources and even misjudgments. Summary of the Invention

[0005] This application provides a method, device, medium, and product for identifying interference sources, which can solve the problems of low efficiency and poor accuracy in interference source identification in related technologies.

[0006] In a first aspect, embodiments of this application provide a method for identifying interference sources, including: Obtain performance and alarm data of the affected cell; Based on the performance data of the affected cells, feature analysis is performed on the affected cells to identify and cluster the interference areas caused by clock synchronization anomalies. Based on the alarm data of the affected cells, cluster analysis is performed on alarm sources related to clock synchronization anomalies to obtain at least one candidate alarm source region. The interference area is matched with at least one candidate alarm source area, and the final location of the interference source is determined based on the matching result.

[0007] Secondly, embodiments of this application provide an electronic device, which includes a processor and a memory. The memory stores a computer program, and when the computer program is executed by the processor, it implements the steps of the method described in the first aspect.

[0008] Thirdly, embodiments of this application provide a computer-readable storage medium storing a program or instructions that, when executed by a processor, implement the steps of the method described in the first aspect.

[0009] Fourthly, embodiments of this application provide a computer program product, which is stored in a storage medium and, when executed by at least one processor, implements the steps of the method described in the first aspect.

[0010] In some embodiments of this application, performance data and alarm data of the interfered cell are acquired; based on the performance data of the interfered cell, feature analysis is performed on the interfered cell to identify and cluster interference areas caused by clock synchronization anomalies; based on the alarm data of the interfered cell, cluster analysis is performed on alarm sources related to clock synchronization anomalies to obtain at least one candidate alarm source region; the interference region is matched with at least one candidate alarm source region, and the final location of the interference source is determined according to the matching result. Thus, this embodiment of the application, on the one hand, identifies the range of the interfered area from the dimension of interference characteristics by analyzing the performance data of the interfered cell, and on the other hand, tracks the region of possible interference sources from the dimension of alarm sources by analyzing the alarm data of the interfered cell, and then achieves accurate location of the interference source through region matching, significantly improving the accuracy of interference source location. Furthermore, compared with related technologies, it eliminates the need for manual screening of interference source locations one by one, and also improves the efficiency of interference source location. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A schematic flowchart illustrating an interference source identification method provided in some embodiments of this application; Figure 2A schematic flowchart illustrating an interference source identification method provided in some embodiments of this application; Figure 3 A schematic flowchart illustrating an interference source identification method provided in some embodiments of this application; Figure 4 A schematic flowchart illustrating an interference source identification method provided in some embodiments of this application; Figure 5 A schematic flowchart illustrating an interference source identification method provided in some embodiments of this application; Figure 6 A schematic flowchart illustrating an interference source identification method provided in some embodiments of this application; Figure 7 A schematic diagram of an electronic device provided for some embodiments of this application. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0014] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0015] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0016] Currently, with the continuous development of wireless communication networks, mobile users' demands for anytime, anywhere service experiences are constantly growing, and basic telecommunications operators are increasingly emphasizing the provision of high-quality services to mobile users. Wireless communication network interference is one of the key factors affecting network quality, significantly impacting call quality, dropped calls, handover, and throughput; therefore, it is also a key issue that operators are focusing on addressing.

[0017] The most common interference encountered by LTE / NR TDD systems can be divided into several categories: intra-system interference and extra-system interference. Intra-system interference is mainly co-channel interference, including TDD frame out-of-synchronization (GPS loss of lock), TDD long-range interference, interference caused by data configuration errors, and interference caused by cross-coverage. Extra-system interference mainly refers to the impact of illegal use of LTE frequency bands by other systems, and spurious, blocking, or intermodulation interference from other systems on the system.

[0018] Because TDD base stations are strictly time-division duplex clock synchronization systems, their daily operation requires high clock synchronization. If a base station A in a network is out of sync with the clocks of other surrounding base stations, its DL signal will be received by surrounding base stations, thus interfering with their uplink reception. The out-of-sync signal transmitted by base station A will also interfere with the uplink reception of base station B. LTE / NR TDD base stations typically use external clock sources, lacking a reference for detection or comparison. This makes them highly susceptible to problems such as GPS lock-up, satellite card malfunctions, and clock source lock-up. When clock quality deviations become too large and the base station loses synchronization, it continues to operate using a clock with excessive deviation, ultimately causing service interference. Many surrounding base stations may also be affected, resulting in widespread radio frequency interference.

[0019] Currently, when wireless communication interference occurs, wireless network optimization engineers need to screen and identify the affected communication devices. They then go to the site and use a frequency sweeper and directional antenna to sweep the frequency in the area where the affected communication devices are located. The location of the interference source is determined by the intersection of the directions of the strongest interference signals at multiple points within that area. When determining the location of the interference source, due to the influence of signals reflected by surrounding buildings, it is necessary to select a relatively high location for frequency sweeping to determine the direction.

[0020] Related technologies (such as Chinese invention patent CN102083090A) disclose a method and apparatus for locating interference sources. This involves a network of interference source acquisition points formed by adding smart antennas to base stations to collect signals. By sampling data at each acquisition point, the path trajectory of the interference source can be determined. Interference source paths corresponding to different acquisition points will intersect at points, and these intersections represent the location of the interference source. However, current analysis and identification of out-of-sync interference and interference sources in TDD wireless communication networks relies on manual analysis of network management data, depending on the engineer's personal experience. This requires manually processing large amounts of data, resulting in poor efficiency and accuracy, and cannot guarantee rapid and effective identification and elimination of interference sources. Furthermore, current base stations generate GPS out-of-sync alarms when detecting GPS out-of-sync, but GPS out-of-sync can cause strong interference across a large area of ​​base stations in a short period, making problem localization difficult.

[0021] To address the issues of low efficiency and poor accuracy in interference source identification, the interference source identification method proposed in this application identifies the affected area from the perspective of interference characteristics by analyzing the performance data of the affected cell. On the other hand, it tracks the region of possible interference sources from the perspective of alarm sources by analyzing the alarm data of the affected cell. Then, it achieves accurate location of the interference source through region matching, which significantly improves the accuracy of interference source location. Furthermore, compared with related technologies, it eliminates the need for manual screening of interference source locations one by one, thus improving the efficiency of interference source location.

[0022] It should be noted that the interference source identification method provided in this application focuses on the accurate location of large-scale regional interference caused by abnormal base station clock synchronization in wireless communication networks. It is applicable to clock synchronization failure scenarios in TDD mobile communication systems (including but not limited to LTE and NR networks) caused by external clock source synchronization anomalies such as GPS loss of lock, BeiDou loss of synchronization, and IEEE 1588 clock source anomalies. Specifically, the interference source identification method provided in this application adopts a "dual data source analysis + spatial matching verification" architecture: the first analysis performs multi-dimensional feature analysis in the time, frequency, and spatial domains based on performance data to identify the interference area affected by the clock synchronization anomaly; the second analysis performs time series clustering analysis based on alarm data to track alarm source areas that may cause interference; finally, spatial matching is used to cross-verify the two to accurately locate the interference source. The core of the interference source identification method provided in this application is to achieve accurate location of the interference source through a dual verification mechanism that integrates performance data and alarm data. This aims to improve the accuracy and efficiency of interference source identification, significantly shorten network interference investigation time, and ensure the stable operation of the communication network and user experience. This application belongs to the field of wireless communication network optimization technology, specifically relating to interference source identification and localization technology in mobile communication networks.

[0023] The interference source identification method, device, medium, and product provided in this application will be described in detail below with reference to the accompanying drawings and through specific embodiments and application scenarios.

[0024] Figure 1 This is a flowchart illustrating an interference source identification method provided in some embodiments of this application.

[0025] like Figure 1 As shown in the embodiments of this application, an interference source identification method may include the following steps: Step 110: Obtain performance data and alarm data of the affected cell; Step 120: Based on the performance data of the affected cells, perform feature analysis on the affected cells, identify and cluster the interference areas caused by clock synchronization anomalies; Step 130: Based on the alarm data of the interfered cell, perform cluster analysis on the alarm sources related to clock synchronization anomalies to obtain at least one candidate alarm source region; Step 140: Match the interference area with at least one candidate alarm source area, and determine the final location of the interference source based on the matching results.

[0026] In step 110, the performance data of the interfered cell includes, but is not limited to, statistical indicators reflecting network operating status and channel quality, such as uplink interference level, physical resource block (PRB) utilization, PRB interference power, symbol-level interference power, channel quality indicator (CQI), reference signal received power (RSRP), and signal-to-interference-plus-noise ratio (SINR). Specifically, the performance data of the interfered cell can be obtained in real time through the northbound interface of the Operation and Maintenance Center (OMC) by parsing and resolving call statistics data. The collection granularity can be selected as 15 minutes or hourly. When the collection granularity is 15 minutes, the value of a single cell at time t throughout the day is 1-96; when the collection granularity is hourly, the value of a single cell at time t throughout the day is 1-24.

[0027] In step 110, the alarm data of the affected cell includes, but is not limited to, various alarm data related to clock synchronization anomalies, such as insufficient satellite lock alarms, abnormal clock reference source alarms, and unavailable system clock alarms. Each alarm data can be associated with the basic information of the base station that generated the alarm data, including but not limited to base station identifier, base station name, geographical location (longitude, latitude), alarm generation time, and alarm clearing time. For example, various alarm information reported in the network can be collected in real time through the northbound interface of the Operation and Maintenance Center (OMC) and aggregated to form an alarm dataset.

[0028] In step 120, the interference area affected by the clock synchronization anomaly is determined from the perspective of the interference impact range. Step 120 can be understood as the first analysis for interference source identification, which may include sub-steps such as time-domain and frequency-domain feature identification, spatial clustering identification, and interference area marking.

[0029] Among them, time-domain and frequency-domain feature identification includes identifying cells that meet preset time-domain features and preset frequency-domain features, and initially marking them as disturbed cells with clock synchronization anomalies.

[0030] Among them, spatial clustering identification includes density-based spatial clustering of cells initially marked as clock synchronization anomalies, which will identify the set of cells that are spatially clustered and meet the density conditions as the clock synchronization anomaly (loss of synchronization) interference area at the current time granularity.

[0031] The interference area marking includes clustering data within the current time granularity, integrating clock out-of-synchronization interference cells, marking interference cells and affected location areas, and generating a data table containing fields such as time, base station name, base station identifier, cell name, cell identifier, longitude, latitude, azimuth, cell uplink interference level, out-of-synchronization interference identifier, and affected area number, which serves as input for subsequent step 140.

[0032] In step 130, starting from the possible location of the interference source, the alarm source area that may cause interference is traced. Step 130 can be understood as the second analysis of interference source identification, which may include sub-steps such as alarm data aggregation, alarm source clustering, and time series closed-loop construction.

[0033] The alarm data aggregation includes acquiring multiple alarm data related to clock reference source anomalies generated within a preset time window, including but not limited to insufficient satellite lock alarms, clock reference source anomaly alarms, and system clock unavailability alarms. Each alarm is associated with information such as the geographical location (longitude and latitude) of the base station that generated it and the alarm generation time.

[0034] Among them, alarm source clustering includes performing cluster analysis on the geographical locations of multiple aggregated alarms to obtain one or more alarm source geographical clusters.

[0035] The time-series closed-loop construction involves sequentially connecting the geographical locations of each alarm source according to their generation time order within each geographical cluster, forming a closed-loop region. The number of alarm source connections is limited to alarms generated within a 30-minute period to ensure temporal correlation. The closed-loop region formed by the alarm source time series intersects with the geographic space to form the final clustered clock synchronization alarm source region, i.e., the candidate alarm source region.

[0036] Step 140 is the fusion verification stage for interference source identification. By spatially matching the interference region obtained in step 120 with the candidate alarm source region obtained in step 130, the precise location of the interference source is achieved. Compared with the traditional method of manually screening interference source locations one by one, the interference source identification method provided in this application embodiment, on the one hand, identifies the range of the interference area from the dimension of interference characteristics by analyzing the performance data of the interfered cell, and on the other hand, tracks the region of possible interference sources from the dimension of alarm sources by analyzing the alarm data of the interfered cell. Then, the precise location of the interference source is achieved through region matching, which significantly improves the accuracy of interference source location. Moreover, compared with related technologies, it eliminates the need for manual screening of interference source locations one by one, and also improves the efficiency of interference source location.

[0037] According to the interference source identification method provided in this application, the following steps are taken: First, performance data and alarm data of the affected cell are acquired. Based on the performance data, feature analysis is performed on the affected cell to identify and cluster interference areas caused by clock synchronization anomalies. Second, based on the alarm data of the affected cell, cluster analysis is performed on alarm sources related to clock synchronization anomalies to obtain at least one candidate alarm source region. Third, the interference region is matched with at least one candidate alarm source region, and the final interference source location is determined based on the matching result. Thus, this application, on the one hand, identifies the affected area from the interference feature dimension by analyzing the performance data of the affected cell; on the other hand, it tracks the regions of possible interference sources from the alarm source dimension by analyzing the alarm data of the affected cell, and then achieves accurate location of the interference source through region matching. This significantly improves the accuracy of interference source location. Furthermore, compared with related technologies, it eliminates the need for manual screening of interference source locations, thus improving the efficiency of interference source location.

[0038] In some embodiments of this application, in order to accurately identify interference characteristics caused by clock synchronization anomalies from massive performance data and determine the geographical scope of their impact, such as... Figure 2 As shown, in step 120 above, based on performance data, feature analysis is performed on the interfered cells to identify and cluster the interference areas caused by clock synchronization anomalies, including: Step 1201: Obtain uplink interference performance data of multiple affected cells at a continuous time granularity; Step 1202: For each time granularity, identify the target cell that meets the preset time domain characteristics and preset frequency domain characteristics from multiple interfered cells, and mark the target cell as a clock synchronization anomaly type of interfered cell; Among them, the preset time domain characteristics include the symbol power of the uplink pilot time slot being continuously higher than the first threshold, and the preset frequency domain characteristics include the noise floor of all resource blocks within the system bandwidth being higher than the second threshold. Step 1203: Perform density-based spatial clustering on the disturbed cells with clock synchronization anomalies. The set of cells that are spatially clustered and meet the preset density conditions will be identified as the interference area caused by clock synchronization anomalies at the current time granularity.

[0039] In step 1201, uplink interference performance data refers to various indicators reflecting the interference level in the uplink receiving direction of a cell, including but not limited to: Physical Resource Block (PRB) interference level, symbol-level interference power, and uplink noise floor level. The continuous time granularity can be selected as 15 minutes or hours. For example, when selecting a 15-minute granularity, the value at time t for a single cell throughout the day ranges from 1 to 96; when selecting an hourly granularity, the value at time t for a single cell throughout the day ranges from 1 to 24. For instance, traffic statistics can be obtained in real-time through the northbound interface of the Operation and Maintenance Center (OMC), and the uplink interference performance indicators for each cell at each time granularity can be extracted from the traffic statistics. This data constitutes the basic input data for subsequent feature recognition.

[0040] In step 1202, target cells that meet the characteristics of clock synchronization abnormal interference are selected by performing feature analysis in both the time domain and frequency domain on the uplink interference performance data of each cell at each time granularity.

[0041] Based on the preset time-domain characteristics, interference caused by clock synchronization anomalies manifests in the time domain as a continuous and stable downlink pilot signal interfering with uplink symbols. Specifically, the preset time-domain characteristics include the symbol power of the uplink pilot time slot being consistently higher than a first threshold. More specifically, for the symbol power (SYMBn) of the first uplink subframe, n∈[0,6], the following condition is satisfied: The minimum symbol power (SYMBn) of the first uplink subframe is greater than the first threshold (e.g., the first threshold is -110dBm). The minimum value of the symbol power (SYMBn) of the first uplink subframe is greater than the daily uplink noise floor level A of the cell (the daily uplink noise floor level A of the cell can be determined through machine learning training). The symbol power waveform exhibits specific interval characteristics, namely, satisfying m(i+1)-m(i)=2 or 3, where m is the symbol index that satisfies the difference between the average value of SYMB[0,n] and SYMBn is greater than 10dB.

[0042] Based on the preset frequency domain characteristics, interference caused by clock synchronization anomalies manifests in the frequency domain as an increase in the noise floor of all resource blocks within the system bandwidth. Specifically, the preset frequency domain characteristics include the noise floor of all resource blocks within the system bandwidth being higher than the second threshold. More specifically, for the cell uplink interference level value PRBn (n∈[0,99], corresponding to 100 PRBs), the following conditions must be met: The minimum uplink interference level PRBn in the cell is greater than the second threshold (for example, the second threshold can be -110dBm). The minimum uplink interference level PRBn of the cell is greater than the cell's normal uplink noise floor level A.

[0043] It should be noted that the first and second thresholds can be the same (e.g., both -110dBm), or different values ​​can be set according to the network standard and actual scenario. For example, for LTE networks, exceeding -110dBm / PRB indicates a moderate interference level; for NR networks, exceeding -107dBm / PRB indicates interference.

[0044] For cells that simultaneously meet the above-mentioned preset time-domain characteristics and preset frequency-domain characteristics, they are marked as "clock synchronization anomaly disturbed cells" and used as input for subsequent spatial clustering.

[0045] In step 1203, since interference caused by clock synchronization anomalies usually exhibits regional contiguous characteristics, that is, multiple disturbed cells are geographically adjacent to each other and appear in clusters, it is necessary to perform spatial clustering analysis on the marked disturbed cells of the clock synchronization anomaly type.

[0046] Specifically, the density-based DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering algorithm is used for big data analysis to cluster problem cells by latitude and longitude. The clustering parameters are set as follows: Eps (neighborhood radius): Set to 0.1 (approximately 10 kilometers), indicating that two cells are considered adjacent when the distance between them is less than or equal to 0.1. Epsmin (minimum number of samples in the neighborhood of a core point): Set to 10, it means that the neighborhood of a core point must contain at least 10 small cells to form a cluster.

[0047] The advantages of using the DBSCAN algorithm are: it does not require pre-specifying the number of clusters, can discover clusters of arbitrary shapes, and can effectively identify and eliminate noise points (isolated disturbed cells). After cluster analysis, the set of cells that meet the above density conditions, that is, a group of cells that are spatially adjacent to each other and reach a certain size, is identified as the interference area caused by clock synchronization anomalies at the current time granularity.

[0048] After identifying the interference area, the interference area and affected cell information are marked. For example, data within the current time granularity is integrated and marked to generate a data table containing the following fields: time, base station name, base station ID, cell name, cell ID, longitude, latitude, azimuth, cell uplink interference level, out-of-synchronization interference identifier, and affected area number. This marked data will serve as input for subsequent steps (such as matching with alarm source areas).

[0049] Thus, in this embodiment, uplink interference performance data at a continuous time granularity is obtained in step 1201, providing a data foundation for feature analysis; then, feature filtering is performed in both the time and frequency domains in step 1202 to initially identify disturbed cells that meet the characteristics of clock synchronization anomalies; finally, density-based spatial clustering is performed in step 1203 to filter out isolated and scattered disturbed cells, retaining disturbed cell groups that exhibit regional contiguous characteristics in space, thereby determining the true clock synchronization anomaly interference area. Ultimately, the accurate identification of clock synchronization anomaly interference areas is achieved through a three-level filtering mechanism of time domain feature identification, frequency domain feature identification, and spatial clustering.

[0050] Furthermore, in some embodiments of this application, in order to improve the accuracy and comprehensiveness of interference type identification and achieve effective differentiation of different interference sources, step 1202 above, identifying the target cell from multiple cells that meets preset time-domain characteristics and preset frequency-domain characteristics, may further include: The pre-established interference feature library is invoked to perform multi-dimensional feature matching on multiple interfered cells, and the results of the multi-dimensional feature matching are obtained. Based on the results of multidimensional feature matching, the type of interference in the affected cell is determined.

[0051] The pre-established interference feature library is a database storing the characteristics of various typical interference signals. It includes time-domain, frequency-domain, and spatial-domain feature templates extracted from historical interference events and verified in the field. The interference feature library can be continuously trained and updated using machine learning algorithms to adapt to changes in the network environment and the emergence of new types of interference. It is worth noting that the interference feature library covers feature templates for multiple typical interference types, enabling it to identify and distinguish different types of interference, providing accurate type information for subsequent processing.

[0052] Specifically, for the uplink interference performance data of each affected cell, its time-domain waveform, frequency-domain distribution, and spatial distribution features are extracted and compared with various interference feature templates stored in the interference feature database. The similarity or matching degree is calculated to obtain the multi-dimensional feature matching result. In this way, by comprehensively matching the three dimensions of time, frequency, and spatial domains, the misjudgment that may occur due to single-dimensional identification is avoided, and the accuracy of interference type identification is significantly improved.

[0053] Based on the obtained multi-dimensional feature matching results, the interfered cells are classified into the interference type with the highest matching degree. After determining the interference type, different subsequent processing strategies can be adopted for different types of interference. For example, for clock synchronization anomaly interference, the subsequent alarm source matching process is initiated; for other types of interference, the data is directly output to maintenance personnel for processing. For example, after determining the interference type, targeted processing strategies can be adopted for different types of interference. For example, clock synchronization anomaly interference requires combining alarm data to locate the interference source, while intermodulation interference can directly locate the specific faulty device, achieving differentiated processing.

[0054] Multidimensional feature matching includes: time-domain feature matching, frequency-domain feature matching, and spatial-domain feature matching.

[0055] The time-domain feature matching includes: identifying the time-domain waveform features of the interference signal; the time-domain waveform features include at least one of the following: continuous stability features; ramp features; pilot interference features; periodic interval features.

[0056] Specifically, examples of various time-domain waveform characteristics are as follows: The characteristic of sustained stability refers to the continuous and stable existence of the interference signal in the time domain, without obvious fluctuations or intermittents. For example, clock synchronization anomaly interference manifests as continuous and stable downlink pilot interference uplink symbols in the time domain.

[0057] The slope characteristic refers to the gradual increase or decrease in the intensity of the interference signal in the time domain, resembling a slope. For example, due to its long propagation distance and large time delay, TDD ultra-long-range interference exhibits a slope characteristic in the time domain, with the interference intensity gradually increasing and then gradually decreasing over time.

[0058] Pilot interference characteristics refer to the fact that interference signals mainly appear at specific pilot time slot locations. For example, clock synchronization anomaly interference and air interface synchronization loss interference both manifest as downlink pilot signals interfering with uplink pilot time slots, exhibiting obvious pilot interference characteristics in the time domain.

[0059] Periodic interval characteristic refers to the occurrence of interference signals at fixed time intervals. For example, due to the time slot structure of the GSM system (approximately 0.577ms per time slot), 2G intermodulation interference exhibits a periodic interval characteristic in the time domain, with an interval of approximately 0.577ms.

[0060] Frequency domain feature matching includes: identifying the frequency domain distribution characteristics of interference signals; the frequency domain distribution characteristics include at least one of the following: full bandwidth noise floor rise characteristics; narrowband spacing characteristics.

[0061] Specifically, examples of various frequency domain distribution characteristics are as follows: The characteristic of full-bandwidth noise floor elevation refers to the fact that interference signals cause a general increase in the noise floor level across the entire bandwidth of the system, without any obvious selectivity. For example, clock synchronization anomaly interference, ultra-long-range interference, and air interface out-of-synchronization interference all exhibit full-bandwidth noise floor elevation in the frequency domain, with the interference level values ​​of all resource blocks within the system bandwidth being higher than normal.

[0062] Narrowband spacing characteristic refers to the fact that the interference signal manifests as multiple narrowband interferences in the frequency domain, with regular intervals between the narrowbands. For example, 2G intermodulation interference manifests as multiple narrowband interferences in the frequency domain, with the intervals between the narrowbands being integer multiples of 200kHz (corresponding to the channel bandwidth of the GSM system).

[0063] Among them, spatial feature matching includes: identifying the spatial distribution characteristics of the interfered cells; the spatial distribution characteristics include at least one of the following: regional contiguous features; strip-shaped distribution features; sporadic distribution features.

[0064] Specifically, examples of various spatial distribution characteristics are as follows: Regional contiguous characteristics refer to the situation where the affected cells are geographically adjacent to each other, distributed in patches, and covering a continuous area. For example, clock synchronization anomaly interference typically exhibits regional contiguous characteristics because the interference source (such as out-of-synchronization base stations) has a large impact range.

[0065] The zonal distribution characteristic refers to the geographical distribution of affected cells in a zonal or linear pattern, extending along a specific direction. For example, due to the influence of propagation paths and terrain, the affected cells of ultra-long-range interference may exhibit a zonal distribution characteristic, especially along rivers, valleys, and other similar directions.

[0066] The sporadic distribution characteristic refers to the fact that the affected cells are geographically dispersed and isolated, without obvious clustering characteristics. For example, intermodulation interference is usually caused by local device failures, with a small impact range, and the affected cells exhibit a sporadic distribution characteristic without regional clustering.

[0067] The interference types include at least one of the following: clock synchronization anomaly interference; ultra-long-range interference; intermodulation interference.

[0068] Specifically, examples of various interference types are shown below: Clock synchronization anomaly interference refers to interference in which the affected cell simultaneously meets the following characteristics: it has continuous stability and pilot interference characteristics in the time domain; it has full-bandwidth noise floor increase characteristics in the frequency domain; and it has regional contiguous characteristics in the spatial domain.

[0069] Ultra-long-range interference refers to interference occurring when the affected cell simultaneously meets the following characteristics: it exhibits a ramp characteristic in the time domain; it exhibits a full-bandwidth noise floor increase characteristic in the frequency domain; and it exhibits a banded distribution characteristic in the spatial domain.

[0070] Intermodulation interference is defined as interference occurring when an affected cell simultaneously meets the following characteristics: periodic intervals in the time domain (approximately 0.577ms period); narrow-band intervals in the frequency domain (intervals that are multiples of 200kHz); and sporadic distribution in the spatial domain.

[0071] Thus, this embodiment of the application invokes a pre-established interference feature library, which gathers multi-dimensional feature templates in the time, frequency, and spatial domains for various typical interferences. Then, it performs time-domain feature matching, frequency-domain feature matching, and spatial-domain feature matching on the interfered cell, comprehensively characterizing the interference signal from three dimensions. Finally, based on the results of the multi-dimensional feature matching, the interfered cell is accurately classified into specific types such as clock synchronization anomaly interference, ultra-long-range interference, and intermodulation interference. Therefore, by first invoking the interference feature library, then performing multi-dimensional feature matching, and finally determining the interference type, the technical approach achieves accurate identification and classification of interference types for the interfered cell.

[0072] In some embodiments of this application, in order to accurately track the sources of interference that may cause clock synchronization anomalies from massive amounts of alarm data and determine their possible geographical location range, such as Figure 3 As shown, in step 130 above, based on the alarm data, cluster analysis is performed on alarm sources related to clock synchronization anomalies to obtain at least one candidate alarm source region, including: Step 1301: Obtain multiple alarm data related to clock synchronization anomalies generated within a preset time window; wherein, each alarm data is associated with the geographical location of the base station that generated the alarm data; Step 1302: Perform density-based spatial clustering on the geographic locations of multiple alarm data to obtain at least one alarm source geographic cluster; Step 1303: Based on the generation time order of alarm data in the geographic cluster of each alarm source, connect the geographic locations of each alarm source in sequence to form a closed loop region, and determine the closed loop region as the candidate alarm source region.

[0073] In step 1301, the preset time window can be set according to actual needs, such as 30 minutes, 1 hour, or longer, to ensure that continuous alarms related to the same interference event can be captured. The selection of the time window needs to balance the amount of data and the correlation: a time window that is too short may miss relevant alarms, while a time window that is too long may introduce irrelevant noise data.

[0074] Alarm data related to clock synchronization anomalies includes, but is not limited to, the following types: Insufficient satellite lock alarm: For example, alarm ID 26122 "Insufficient satellite lock alarm" indicates that the base station receives insufficient satellite signals (less than 4 satellites) and cannot synchronize with the satellite clock; Clock reference source anomaly alarm: For example, alarm ID 26262 "Clock reference source anomaly alarm" indicates that the base station cannot synchronize with the configured clock reference source (such as GPS, BeiDou, IEEE 1588, etc.); System clock unavailable alarm: For example, alarm ID 26260 "System clock unavailable alarm" indicates that the base station system clock is unavailable, which may cause abnormal service processing.

[0075] Each alarm data entry is associated with the geographical location information of the base station that generated the alarm, including the base station's longitude and latitude coordinates. In addition, the alarm data also includes information such as the base station name, base station ID, alarm generation time, and alarm clearing time. This information can be obtained in real time from the OMC northbound interface or exported from the alarm management system.

[0076] Specifically, when a base station fails to receive valid signals from at least four satellites for more than 200 consecutive seconds, a "star card insufficient satellite lock alarm" will be triggered. If this state continues to cause clock synchronization failure, it may further trigger a "clock reference source abnormal alarm" and a "system clock unavailable alarm". The sequence and spatiotemporal distribution of these alarms provide important clues for tracing the source of interference.

[0077] In step 1302, the alarm data obtained in step 1301 is clustered according to geographical location. The purpose is to group spatially adjacent alarm sources together to form a geographical cluster of alarm sources. Since the impact range of clock synchronization anomaly interference sources (such as out-of-sync base stations) is usually large, multiple base stations around it may trigger alarms one after another. These alarm sources should exhibit clustering characteristics in terms of geographical location.

[0078] Specifically, the density-based DBSCAN clustering algorithm is used for cluster analysis. The DBSCAN algorithm can discover clusters of arbitrary shapes and does not require pre-specifying the number of clusters, making it suitable for scenarios where the distribution of alarm sources is uncertain. Through the density conditions of the DBSCAN clustering algorithm, isolated and scattered alarm sources (which may be due to local faults rather than global interference) are effectively filtered out, preventing them from interfering with the localization results and effectively filtering noisy data.

[0079] For example, different clustering strategies can be used depending on the number of alarm sources: If there are more than 3 alarm sources: use the DBSCAN clustering algorithm for density-based spatial clustering, and set the clustering parameters to Eps=0.1 (corresponding to about 10 kilometers) and Epsmin=3 (meaning that at least 3 alarm sources must be included in the neighborhood of a core point to form a cluster). If there are two alarm sources: the density condition of DBSCAN clustering is not met, and the center position of the line connecting the two base stations corresponding to the alarm sources can be directly determined as the candidate alarm source area. If there is only one alarm source, the location of that alarm source will be directly designated as a candidate alarm source area.

[0080] After DBSCAN clustering, alarm sources that meet the density criteria are grouped into the same alarm source geographical cluster, while isolated alarm sources that do not meet the density criteria are considered noise points and do not participate in the subsequent candidate region construction. Each alarm source geographical cluster represents a group of base stations that may have clock synchronization anomalies.

[0081] In this way, different processing strategies are adopted for different situations such as more than 3, 2, or 1 alarm sources, to ensure that reasonable candidate regions can be obtained in various scenarios and to adaptively handle different numbers of alarm sources.

[0082] In step 1303, for each alarm source geographical cluster, time-dimensional information is further introduced. The geographical locations of each alarm source are connected according to the order in which the alarms occurred, forming a closed-loop region. The core idea of ​​this step is that the impact of interference sources typically propagates from near to far; the base station closest to the interference source is the first to issue an alarm, and the alarm range gradually expands over time. Therefore, connecting the alarm source locations in chronological order can outline the trajectory of interference propagation, and the area where the earliest alarm occurs is often closer to the actual interference source.

[0083] Specifically, the geographic clustering for each alarm source includes: Sort all alarm sources within a cluster according to their alarm generation time, from earliest to latest. Connect the geographical locations of adjacent alarm sources with straight lines in chronological order; Connect the last alarm source to the first alarm source to form a closed loop area; The closed-loop area is geospatial corrected (e.g., taking into account actual terrain, roads, and other factors) to form the final candidate alarm source area.

[0084] It should be noted that the number of alarm source connections is limited to alarms generated within a preset time window to ensure temporal relevance. For example, connections can be limited to alarms generated within a 30-minute period; alarms generated more than 30 minutes ago may belong to different events and can be excluded from connection.

[0085] Through the processing in step 1303, each alarm source geographical cluster is transformed into a geographically closed-loop region, i.e., a candidate alarm source region. This region represents the geographical range where interference sources that may cause clock synchronization anomalies are located, with base stations within the region and at its boundaries being the focus of attention. Step 1303 considers both the geographical location (spatial dimension) and the alarm generation order (temporal dimension) of the alarm sources, which better reflects the pattern of interference propagation than simple spatial clustering, thus improving the accuracy of candidate region location. Furthermore, by limiting the number of alarm source connections through a preset time window, it ensures that alarm sources connected together belong to the same time series event, avoiding erroneous connections across events.

[0086] Finally, the candidate alarm source regions determined in step 1303 are output, including the geographical boundary information of the region (such as the coordinates of polygon vertices), the list of base stations in the region, the earliest alarm time, and other information. These candidate alarm source regions will be used as input for step 140, and spatial matching will be performed with the interference region determined in step 120.

[0087] Thus, in this embodiment, step 1301 acquires clock synchronization anomaly-related alarm data within a preset time window, ensuring the temporal and type relevance of the data. Then, step 1302 performs density-based spatial clustering on the geographical locations of the alarm sources, grouping spatially adjacent alarm sources into alarm source groups and filtering out isolated noise points. Finally, step 1303 introduces a time dimension, constructing a closed-loop region according to the time sequence of alarm generation, converting the alarm source groups into geographically candidate alarm source regions. Therefore, through a three-stage processing flow of alarm data acquisition, spatial clustering, and time-series closed-loop construction, accurate regional positioning of clock synchronization anomaly alarm sources is achieved.

[0088] In some embodiments of this application, in order to achieve effective fusion and cross-validation of the interference region and the candidate alarm source region, and to accurately locate the final interference source position, such as... Figure 4 As shown, in step 140 above, the interference area is matched with at least one candidate alarm source area, and the final location of the interference source is determined based on the matching result, including: Step 1401: Determine whether there is spatial intersection between the candidate alarm source area and the interference area; Step 1402: When the candidate alarm source area and the interference area intersect, the candidate alarm source area or the geographical location of the alarm source in the intersecting area is determined as the final interference source location. Step 1403: When there is no spatial intersection between the candidate alarm source area and the interference area, obtain multiple consecutive interference areas at different time granularities. Based on the temporal continuity characteristics of the interference areas, filter out the interference cells that meet the preset time aggregation conditions, and determine the final interference source investigation area based on the filtered cell list.

[0089] In step 1401, the interference area determined in step 120 and the candidate alarm source area determined in step 130 are subjected to spatial overlay analysis to determine whether there is a spatial intersection relationship between the two.

[0090] Specifically, the determination of spatial intersection can include the following cases: Point and surface intersection: If the candidate alarm source region is a single alarm source location point (when the number of alarm sources is 1), determine whether the location point falls within the polygon boundary of the interference region; Line and surface intersection: If the candidate alarm source area is the center point of the line connecting two alarm sources, determine whether the center point falls within the interference area; Intersection of surfaces: If the candidate alarm source region is a closed loop region formed by connecting multiple alarm sources, determine whether there is an overlap between the region and the interference region, that is, whether the intersection of the two polygons is not empty.

[0091] Determining spatial intersections can employ classic GIS spatial analysis algorithms, such as the ray casting method to determine if a point is inside a polygon, or the polygon clipping algorithm to determine if two polygons have overlapping areas.

[0092] In step 1402, when step 1401 determines that there is spatial intersection, it means that the alarm source area is within the interference influence range. The two corroborate each other, and the candidate alarm source area or the geographical location of the alarm source in the intersection area can be directly determined as the final interference source location.

[0093] Specifically, the final location of the interference source is determined based on the different types of candidate alarm source regions: Alarm source is a single point: If the candidate alarm source area is a single alarm source location point, and the point falls within the interference area, then the point is directly determined as the final interference source location; Alarm source is the center of the line connecting two points: If the candidate alarm source area is the center point of the line connecting two alarm sources, and the center point falls within the interference area, then the center point is determined as the final interference source location. Alarm source is a multi-point closed-loop area: If the candidate alarm source area is a closed-loop area formed by multiple alarm sources connected together, and this area overlaps with the interference area, then the overlapping area (i.e., the intersection of the two areas) is determined as the final location of the interference source. If the overlapping area is too large, the time sequence of the alarm sources can be further analyzed, and the location of the alarm source that first alarmed can be used as the priority investigation point.

[0094] Once the final location of the interference source is determined, its geographical coordinates and related information are output for use by on-site investigators.

[0095] In step 1403, if step 1401 determines that there is no spatial intersection, it means that the interference area at the current time granularity cannot match the alarm source area. This could be due to the following reasons: the interference source has not yet generated an alarm, the alarm information is delayed, or there is a deviation in the identification of the interference area. In this case, it is necessary to introduce multi-time granularity analysis to further narrow down the investigation scope by utilizing the temporal continuity characteristics of the interference area.

[0096] Specifically, it includes the following sub-steps: First, obtain interference regions at multiple consecutive time granularities. For example, obtain interference region identification results for multiple consecutive time granularities (e.g., 4-6 consecutive 15-minute granularities, or 3-4 consecutive hourly granularities). These interference regions reflect how the range of interference influence changes over time.

[0097] Secondly, interfering cells are screened based on temporal continuity characteristics. For example, the performance of cells within each geographical cluster at a continuous temporal granularity is analyzed to screen out interfering cells that meet preset temporal clustering conditions.

[0098] Finally, the interference source investigation area is output. For example, based on the filtered cell list and combined with big data analysis of the cumulative number and intensity of interference occurrences, the final interference source investigation area is output. Specifically, this may include: spatially clustering the filtered interfering cells according to their geographical location to form new candidate areas; statistically analyzing the number of times each interfering cell is identified as interference and its average interference intensity as a basis for priority ranking; identifying areas where cells with high cumulative occurrences and high interference intensity are located as priority investigation areas; and outputting a results report containing information such as the cell list, area boundaries, and investigation priorities for on-site investigation personnel to refer to.

[0099] Thus, in this embodiment, spatial intersection judgment is first performed in step 1401 to establish the association between the interference area and the alarm source area. When spatial intersection exists, the location of the interference source is directly determined in step 1402, achieving rapid and accurate positioning. When spatial intersection does not exist, multi-time granularity analysis is introduced in step 1403 to further filter interfering cells using time continuity characteristics, ensuring that a reliable investigation area can be obtained under various complex scenarios. Therefore, through the branching processing flow of spatial intersection judgment, direct positioning, and iterative positioning, accurate positioning and multiple verification of the interference area are achieved.

[0100] In some embodiments of this application, in order to further verify the persistence and stability of interference from a time dimension, eliminate misjudgments caused by instantaneous interference or accidental factors, and accurately locate the interference source area that truly needs attention, step 1403 above, based on the temporal continuity characteristics of the interference area, filters out interference cells that meet preset time aggregation conditions, including: If interfering cells within a geographic cluster consistently appear at at least two consecutive time granularities, and the overlap of cells within the geographic cluster at consecutive time granularities exceeds a preset threshold, then the interfering cells within the geographic cluster will be included in the cell list.

[0101] Among them, the characteristics of temporal continuity include: continuous occurrence condition and overlap condition.

[0102] The condition of persistent occurrence refers to the fact that interfering cells within a geographic cluster persist across at least two consecutive time granularities. The core idea behind this condition is that genuine interference sources typically exert a continuous influence, causing repeated interference phenomena in the same area over multiple consecutive time periods; while transient interference (such as sudden external signals, brief equipment failures, etc.) often only occurs within a single time granularity and subsequently disappears automatically.

[0103] Specifically, for each geographical cluster, the results of interfering cell identification at the continuous time granularity are analyzed: If a cell is identified as an interfering cell at time t, but is not identified at time t+1, then the condition of continuous occurrence is not met, and it may be a transient interference or a false alarm. If a cell is identified as an interfering cell at both time t and time t+1, then the condition of continuous occurrence is met, indicating that the interference persists in the area. If a cell appears continuously at three or more consecutive time granularities (e.g., it is identified at times t, t+1, and t+2), its credibility as a stable disturbance is further enhanced.

[0104] For example, if 50 interfering cells are identified in a certain geographic cluster at time t and 45 interfering cells are identified at time t+1, and 40 of these cells are identified at both times t and t+1, then these 40 cells meet the condition of continuous occurrence.

[0105] The overlap condition refers to the fact that the overlap of cells within a cluster at a continuous time granularity is higher than a preset threshold. The core idea behind this condition is that a true source of interference will lead to a relatively stable disturbed area, meaning that the set of disturbed cells should maintain a high degree of consistency within a continuous time period. If the disturbed cells change drastically and the overlap is very low within a continuous time period, it may involve multiple different sources of interference, or the interference itself may be unstable.

[0106] The formula for calculating overlap is: Overlap = (Number of interfering cells shared by the two time granularities) / (Smaller value of the number of interfering cells in the two time granularities) × 100% In this formula, the numerator is the number of cells identified as interference at both consecutive time granularities; the denominator is the smaller value of the number of interference cells at both time granularities (choosing the smaller value as the denominator can avoid the problem of falsely high overlap due to the expansion of the interference range).

[0107] The preset threshold can be set according to the actual network conditions, for example, 70%. When the overlap is higher than 70%, it indicates that the two sets of interfering cells at different time granularities are highly consistent and point to the same stable interference source; when the overlap is lower than 70%, it indicates that the interference area changes significantly, and further analysis or expansion of the investigation scope may be required.

[0108] Interfering cells that simultaneously meet both the persistent occurrence condition and the overlap condition are included in the interfering cell list. This list contains the following information: Cell identifiers, including but not limited to base station name, base station ID, cell name, and cell ID; Geographical location, including but not limited to longitude and latitude; Time information, including but not limited to the time of first appearance, duration, and number of appearances; Interference intensity, including but not limited to average interference level and peak interference level; Priority scoring includes, but is not limited to, investigation priorities calculated based on indicators such as duration, frequency of occurrence, and intensity of interference.

[0109] Finally, the final area for investigating interference sources is determined based on the cell list. For example, further analysis of the selected interference cell list can determine the final area for investigating interference sources: Spatial re-clustering: The interfering cells in the inventory are re-clustered based on density (the DBSCAN algorithm can be used again) to form new geographical clusters. These new clusters reflect the truly noteworthy interference areas after time continuity screening.

[0110] Priority ranking: Based on factors such as the frequency of occurrence of each cell, interference intensity, and location within the cluster (central or peripheral area), priority is assigned to each cluster area and the cells within each area. Generally, cells with longer duration of occurrence, higher interference intensity, and those located in the central area of ​​the cluster have higher priority.

[0111] Region boundary determination: For each clustered region, the convex hull algorithm can be used to calculate its minimum bounding polygon, which serves as the geographical boundary of the interference source investigation area. Alternatively, the boundary can be appropriately adjusted according to the actual terrain and road conditions to form a navigation area that facilitates on-site investigation.

[0112] Output Investigation Report: Generate an interference source investigation report, which includes the following: an investigation area map, a list of key communities, and investigation recommendations. The investigation area map is used to mark the boundaries and priority levels of each cluster area. The list of key communities is a list of communities sorted by priority and their detailed information. The investigation recommendations are used to suggest areas and communities that on-site personnel should prioritize for investigation, as well as possible directions and routes.

[0113] Thus, this embodiment first filters out interfering cells that stably appear across multiple consecutive time granularities by using the continuous occurrence condition, eliminating transient interference and occasional false alarms; then, it calculates the similarity of the interfering cell sets across consecutive time granularities using the overlap condition, ensuring the stability and consistency of the interference region; finally, cells that meet the conditions are included in the list, and spatial re-clustering and priority sorting are performed to determine the final interference source investigation area. Therefore, through the technical path of continuous occurrence condition judgment, overlap condition judgment, cell list filtering, and investigation area determination, accurate screening of interfering cells based on temporal continuity characteristics is achieved.

[0114] In some embodiments of this application, after step 140, regardless of whether the location is directly determined in step 1402 or the investigation area is output in step 1403, on-site investigation and verification are ultimately required. On-site personnel, equipped with portable spectrum analyzers, frequency sweepers, filters, directional antennas, and other equipment, conduct frequency sweep tests in the target area. Through multi-point cross-location, the specific location and nature of the interference source are ultimately determined. After the on-site investigation results are fed back, they are used to verify and update the interference feature library: if the location is accurate, the feature data of this interference is added to the feature library as a positive sample to enhance the model's ability to identify this type of interference; if the location is inaccurate, the cause is analyzed and relevant algorithm parameters are adjusted to prevent similar problems from recurring. Based on the on-site investigation results, the interference feature library can be continuously enriched and improved, enabling the system to have self-learning capabilities, adapt to the emergence of new types of interference, and support continuous updates to the feature library.

[0115] In order to continuously enrich and improve the interference feature library, such as Figure 5 As shown, after step 140, the interference source identification method provided in this application embodiment may further include: Step 150: Based on the final on-site investigation results of the interference source location, verify and update the pre-established interference feature library, which is used to identify interfering cells caused by clock synchronization anomalies.

[0116] For example, after determining the final location of the interference source based on the matching results, on-site investigators, equipped with portable spectrum analyzers, frequency sweepers, filters, directional antennas, and other equipment, analyze the changes in noise floor within the affected frequency range to ultimately confirm the specific location and nature of the interference source. Simultaneously, based on the on-site investigation results, the interference feature database is verified and updated to continuously improve the system's identification accuracy.

[0117] In this way, the on-site investigation results of the embodiments of this application can continuously enrich and improve the interference feature library, enabling the system to have self-learning capabilities, adapt to the emergence of new interference, and support continuous updates to the feature library.

[0118] In practical applications, traffic statistics and key alarm information of 4G and 5G networks can be obtained through OMC to identify and handle GPS out-of-synchronization interference sources. The main steps include: identifying out-of-synchronization interference types based on interference characteristics, identifying GPS interference sources based on alarm sequence characteristics, and coordinating the removal of interference sources on-site. At the same time, the interference feature database will be continuously improved based on the results of on-site investigation to continuously improve the timeliness of this solution.

[0119] For example, such as Figure 6 As shown, the interference source identification method provided in this application embodiment may include the following steps: Step 601: Obtain performance data and alarm data of the affected cell.

[0120] In this embodiment, traffic statistics and key alarm information can be obtained in real time through the OMC northbound interface. Interference analysis can be performed at a granularity of 15 minutes or hours. If 15 minutes is selected, the value of time t for a single cell throughout the day is 1-96. If hourly granularity is selected, the value of time t for a single cell throughout the day is 1-24.

[0121] Step 602: For each time granularity, identify the target cell that meets the preset time domain characteristics and preset frequency domain characteristics from multiple interfered cells, and mark the target cell as a clock synchronization anomaly type of interfered cell; The performance data of the interfered cells obtained in step 601 may include uplink interference performance data of multiple interfered cells at a continuous time granularity; for each time granularity, target cells that meet preset time domain characteristics and preset frequency domain characteristics are identified based on the uplink interference performance data of the interfered cells, and they are marked as clock synchronization anomaly type of interfered cells.

[0122] Among them, the preset time-domain characteristics include the symbol power of the uplink pilot time slot being continuously higher than the first threshold, and the preset frequency-domain characteristics include the noise floor of all resource blocks within the system bandwidth being higher than the second threshold.

[0123] For example, if the interference meets the preset time-domain and frequency-domain characteristics, it can be preliminarily identified as TDD system out-of-sync interference (i.e., clock synchronization anomaly interference). In the time domain, there is continuous and stable downlink pilot interference with uplink symbols, and in the frequency domain, the noise floor within the entire system bandwidth is raised. Based on the collective characteristics of the performance changes of the disturbed cells, an atmospheric waveguide cell identification algorithm was developed. Machine learning algorithms were used to train relevant discrimination thresholds to achieve feature recognition.

[0124] One approach is to use cell performance data for uplink noise floor machine learning training to determine the daily uplink noise floor level A of the cell. Furthermore, according to the requirements of wireless communication network experience, LTE noise levels exceeding -110dBm / PRB and NR noise levels exceeding -107dBm / PRB are considered to indicate moderate interference.

[0125] If the interfered cell is determined to meet the preset time-domain and frequency-domain characteristics, then the cell data at time t can be aggregated as input for the fourth step of the TDD system's out-of-sync interference geographic location group characteristic judgment. The cell data includes, but is not limited to, sequence number, time, base station name, base station ID, cell name, cell ID, and cell uplink interference level. Otherwise, the cell data at time t is discarded, and the next data item is processed, i.e., processing of the data at time t+1 begins from step 601.

[0126] Step 603: Perform density-based spatial clustering on the disturbed cells with clock synchronization anomalies. The set of cells that are spatially clustered and meet the preset density conditions will be identified as the interference area caused by clock synchronization anomalies at the current time granularity.

[0127] Among them, out-of-sync interference in TDD systems exhibits obvious regional clustering characteristics, with interfering cells appearing in clusters. An algorithm for identifying the geographical location cluster characteristics of out-of-sync interference was developed, and cells meeting the above conditions were identified: a. The density-based DBSCAN clustering algorithm is used for big data analysis to cluster problem cells by latitude and longitude, with clustering parameters set as Eps=0.1 and Epsmin=10. b. If a community is located within this type of cluster area, it is determined to meet the performance characteristics of a geographical area.

[0128] Subsequently, the clustering data at time t is integrated to identify clock out-of-synchronization interference cells, mark the interfering cells and the affected location areas, and output the following information: serial number, time, base station name, base station ID, cell name, cell ID, longitude, latitude, azimuth, cell uplink interference level, out-of-synchronization interference, and affected area number.

[0129] Step 604: Perform density-based spatial clustering on the geographic locations of multiple alarm data to obtain at least one alarm source geographic cluster.

[0130] The alarm data of the interfered cell obtained in step 601 may include: multiple alarm data related to clock synchronization anomalies generated within a preset time window, with each alarm data associated with the geographical location (longitude, latitude) of the base station that generated the alarm data. For example, a dataset of clock synchronization anomaly alarms from multiple consecutive time periods can be aggregated. The alarm dataset includes multiple alarm data, each of which includes a sequence number, alarm number, alarm name, base station name, base station ID, longitude, latitude, alarm generation time, and alarm clearing time. The communication equipment mainly connects to the GPS system ground receiving device via the satellite clock line to monitor the operation of space satellites and obtain information such as the latitude, longitude, and time of the user's geographical location, thus synchronizing with the GPS clock. The receiver must simultaneously receive valid signals from at least four satellites to successfully resolve the user's location information. If four satellites are not found, a "satellite loss" problem will occur. If this cumulative state exceeds 200 seconds, the base station may report the following fault alarm: Star Card Insufficient Star Lock Alarm (Alarm No. 26122): The base station cannot synchronize with the GPS clock, which will cause the base station system clock to be unavailable.

[0131] Clock reference source abnormality alarm (alarm number 26262): The base station cannot synchronize with the reference clock source, which will cause the base station system clock to be unavailable.

[0132] System clock unavailable alarm (alarm number 26260): Various anomalies may occur in base station service processing.

[0133] Step 605: Based on the generation time order of alarm data in the geographic cluster of each alarm source, connect the geographic locations of each alarm source in sequence to form a closed loop region, and determine the closed loop region as the candidate alarm source region.

[0134] Specifically, based on the alarm dataset of clock synchronization anomalies at multiple consecutive time points, candidate alarm sources causing clock synchronization anomalies are located using time series clustering.

[0135] First, based on the number of alarm sources, if there are more than 3 alarm sources, the DBSCAN clustering algorithm based on density is used for big data analysis to cluster the alarm sources by latitude and longitude, and the clustering parameters are set to Eps=0.1 and Epsmin=3; if there are no more than 3 alarm sources, two types of results are generated to select the base station connection area corresponding to the alarm source (if there are 2 alarm sources, it is the center position of the base station connection, and if there is 1 alarm source, it is the location of the alarm source).

[0136] Secondly, based on the generation time of the clustered alarm sources, they are sequentially connected to form a closed-loop area, and the number of alarm sources connected is limited to alarms generated within 30 minutes.

[0137] Finally, the closed-loop region formed by the alarm source time series intersects with the geographic space to form the final clustered clock synchronization alarm source region.

[0138] Step 606: Determine whether there is spatial intersection between the candidate alarm source area and the interference area; Step 607: When the candidate alarm source area and the interference area intersect, the candidate alarm source area or the geographical location of the alarm source in the intersecting area is determined as the final interference source location.

[0139] Specifically, it is determined that the candidate alarm source area and the interference area have spatial intersection; the intersecting area is the interference source. For example, if the candidate alarm source area is the location point of a candidate alarm source, and the location point falls within the interference area, it is determined to be an intersection, and the location point of the candidate alarm source is the location of the interference source.

[0140] Step 608: When there is no spatial intersection between the candidate alarm source area and the interference area, obtain multiple consecutive interference areas at different time granularities. Based on the temporal continuity characteristics of the interference areas, filter out the interference cells that meet the preset time aggregation conditions, and determine the final interference source investigation area based on the filtered cell list.

[0141] When the candidate alarm source area and the interference area do not spatially intersect, multiple problematic interference cells are aggregated. Interference cells occurring at multiple consecutive time points are aggregated and output. Time continuity is used to determine clock synchronization loss clustering characteristics, ultimately forming a list of problematic cells. The occurrence time and duration of interference in the affected cells are analyzed, and the temporal group characteristics and identification algorithm of the interfering cells are extracted. For cells meeting the following conditions, temporal grouping is determined: the clustered area must appear at least twice in consecutive time granularities, and the consistency of cells in the consecutive time granularity clusters must be >70%. Based on big data analysis of the cumulative number and intensity of clock synchronization loss interference, the interference area and the corresponding list of problematic cells are output.

[0142] Step 609: Based on the final on-site investigation results of the interference source location, verify and update the pre-established interference feature library, which is used to identify interfering cells caused by clock synchronization anomalies.

[0143] Both the interference source location output in step 607 and the interference area and corresponding list of problematic cells output in step 608 can be used for on-site troubleshooting. On-site troubleshooting of interference sources requires equipment including portable spectrum analyzers, frequency sweepers, filters, and directional antennas to analyze changes in noise floor within the affected frequency range. Compared to traditional manual detection methods, this approach offers higher accuracy and efficiency, allowing for the location and elimination of interference sources in the shortest possible time, thereby ensuring the stability and reliability of the GPS system. While manual troubleshooting typically takes at least 4-5 days, the method provided in this application can shorten the troubleshooting time to 1.5-2 days, preventing large-scale base station outages and improving the stability and reliability of the GPS system.

[0144] Thus, this application embodiment, on the one hand, identifies the area of ​​interference from the dimension of interference characteristics by analyzing the performance data of the interfered cell, and on the other hand, tracks the area of ​​possible interference sources from the dimension of alarm source by analyzing the alarm data of the interfered cell. Then, it achieves accurate positioning of interference sources through spatial region matching, which significantly improves the accuracy of interference source positioning. Moreover, compared with related technologies, it does not require manual screening of interference source locations one by one, which improves the efficiency of interference source positioning.

[0145] In addition, such as Figure 7 As shown in the illustration, this application also provides an electronic device 700, which may include a processor 710 and a memory 720. The memory 720 stores a computer program, which, when executed, implements the various processes of the above-described method embodiments.

[0146] The memory 720 is used to store programs or data. The memory may be, but is not limited to, random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), etc.

[0147] This application also provides a computer-readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described method embodiments and achieve the same technical effects. To avoid repetition, these will not be described again here.

[0148] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0149] This application also provides a computer program product, which is stored in a storage medium and executed by at least one processor to implement the various processes of the above method embodiments and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0150] This application also provides a chip, which includes a processor and a communication interface. The communication interface and the processor are coupled. The processor is used to run programs or instructions to implement the various processes of the above method embodiments and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0151] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0152] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0153] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A method for identifying interference sources, characterized in that, include: Obtain performance and alarm data of the affected cell; Based on the performance data, feature analysis is performed on the interfered cells to identify and cluster the interference areas caused by clock synchronization anomalies. Based on the alarm data, cluster analysis is performed on alarm sources related to clock synchronization anomalies to obtain at least one candidate alarm source region. The interference area is matched with the at least one candidate alarm source area, and the final location of the interference source is determined based on the matching result.

2. The method according to claim 1, characterized in that, Based on the performance data, feature analysis is performed on the interfered cells to identify and cluster the interference areas caused by clock synchronization anomalies, including: Acquire uplink interference performance data of multiple affected cells at a continuous time granularity; For each time granularity, target cells that meet preset time domain characteristics and preset frequency domain characteristics are identified from multiple interfered cells, and the target cells are marked as clock synchronization anomaly-type interfered cells; wherein, the preset time domain characteristics include the symbol power of the uplink pilot time slot being continuously higher than a first threshold, and the preset frequency domain characteristics include the noise floor of all resource blocks within the system bandwidth being higher than a second threshold; Density-based spatial clustering is performed on the disturbed cells of the clock synchronization anomaly type. The set of cells that are spatially clustered and meet the preset density conditions is determined as the interference area caused by clock synchronization anomaly at the current time granularity.

3. The method according to claim 2, characterized in that, The step of identifying target cells that satisfy preset time-domain features and preset frequency-domain features from multiple cells further includes: The pre-established interference feature library is invoked to perform multi-dimensional feature matching on multiple interfered cells, and the results of the multi-dimensional feature matching are obtained. The multidimensional feature matching includes: time-domain feature matching, frequency-domain feature matching, and spatial-domain feature matching; The time-domain feature matching includes: identifying the time-domain waveform features of the interference signal; the time-domain waveform features include at least one of the following: continuous stability feature; ramp feature; pilot interference feature; periodic interval feature; The frequency domain feature matching includes: identifying the frequency domain distribution characteristics of the interference signal; the frequency domain distribution characteristics include at least one of the following: full bandwidth noise floor rise characteristics; narrowband spacing characteristics; The spatial feature matching includes: identifying the spatial distribution characteristics of the interfered cells; the spatial distribution characteristics include at least one of the following: regional contiguous features; strip-shaped distribution features; sporadic distribution features; Based on the results of multidimensional feature matching, the interference type of the affected cell is determined; the interference type includes at least one of the following: clock synchronization anomaly interference; ultra-long-range interference; intermodulation interference.

4. The method according to claim 1, characterized in that, Based on the alarm data, cluster analysis is performed on alarm sources related to clock synchronization anomalies to obtain at least one candidate alarm source region, including: Acquire multiple alarm data related to clock synchronization anomalies generated within a preset time window; wherein each alarm data is associated with the geographical location of the base station that generated the alarm data; Density-based spatial clustering is performed on the geographic locations of the multiple alarm data to obtain at least one alarm source geographic cluster; Based on the generation time order of alarm data in each alarm source geographic cluster, the geographical locations of each alarm source are sequentially connected to form a closed loop region, and the closed loop region is determined as the candidate alarm source region.

5. The method according to claim 1, characterized in that, The step of matching the interference region with the at least one candidate alarm source region and determining the final location of the interference source based on the matching result includes: Determine whether the candidate alarm source region and the interference region have spatial intersection; If the candidate alarm source area and the interference area spatially intersect, the candidate alarm source area or the geographical location of the alarm source within the intersecting area shall be determined as the final interference source location. When the candidate alarm source area and the interference area do not intersect spatially, multiple consecutive interference areas with different time granularities are obtained. Based on the temporal continuity characteristics of the interference areas, interference cells that meet the preset time aggregation conditions are selected, and the final interference source investigation area is determined based on the selected cell list.

6. The method according to claim 5, characterized in that, The process of filtering out interfering cells that meet preset time clustering conditions based on the temporal continuity characteristics of the interference region includes: If interfering cells within a geographic cluster continuously appear at at least two consecutive time granularities, and the overlap of cells within the geographic cluster at consecutive time granularities is higher than a preset threshold, then the interfering cells within the geographic cluster are included in the cell list.

7. The method according to any one of claims 1-6, characterized in that, The method further includes: Based on the on-site investigation results of the final interference source location, the pre-established interference feature library is verified and updated. The interference feature library is used to identify interfering cells caused by clock synchronization abnormalities.

8. An electronic device, characterized in that, The electronic device includes a processor and a memory, the memory storing a computer program that, when executed by the processor, implements the steps of the method as described in any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, The computer program product is stored in a storage medium, and when executed by at least one processor, the computer program product implements the steps of the method as described in any one of claims 1-7.

Images