An AI interference adaptive coordination method for room sharing construction and sharing

Abstract

The application discloses an AI interference adaptive coordination method for room-sharing construction and sharing, and relates to the technical field of wireless communication network optimization. The method comprises the following steps: acquiring field interference observation data and room-sharing configuration data, merging to generate room-sharing observation data; identifying intermodulation, adjacent frequency leakage and load congestion interference states, and generating hierarchical interference situation data; backtracking abnormal interference to form interference contribution records, constructing a permission-constrained coordination action space, generating parameter coordination actions through a state-action filtering model, and writing the updated results into a permission chain. By converting the field interference observation data into hierarchical interference situation data with mechanism distinction and hierarchical correlation, backtracking the source object, disturbed object and intervenable parameters of abnormal interference states, and combining the permission constraints to generate parameter coordination actions, the closed-loop processing of interference identification, contribution positioning and cross-operator parameter coordination in the room-sharing construction and sharing scene is realized.

Description

Technical Field

[0001] This invention relates to the field of wireless communication network optimization technology, specifically to an AI interference adaptive coordination method for indoor distributed antenna systems (DAS) co-construction and sharing. Background Technology

[0002] Indoor distributed antenna systems (DAS) are used to address insufficient wireless signal coverage within buildings. With the increasing adoption of multi-operator network co-construction and sharing models, the same building or the same DAS area may simultaneously carry frequency band resources, coverage units, and service loads from different operators, creating complex coupling relationships between DAS links, combining devices, antenna locations, and shared coverage areas. In such scenarios, intermodulation interference, adjacent channel leakage, and load congestion can easily occur across shared areas, frequency band resources, and operator affiliations, making it difficult to directly reflect the true source of interference from a single frequency band or area.

[0003] Current indoor distribution system interference optimization methods largely rely on on-site frequency scanning, manual parameter verification, and single-operator parameter adjustments. They typically determine interference strength based on local test results and then optimize by adjusting power, frequency points, or coverage boundaries. This approach struggles to correlate on-site observations with shared area relationships, frequency band affiliations, and operator affiliations. It also fails to distinguish the different formation mechanisms of intermodulation interference, adjacent channel leakage, and load congestion, making it impossible to effectively trace abnormal interference states back to the source, affected object, and modifiable parameters.

[0004] Furthermore, in indoor distributed antenna system (DAS) co-construction and sharing environments, parameter coordination actions typically involve equipment ownership, frequency band permissions, and adjustable parameter ranges from different operators. Existing optimization methods lack action constraint processing oriented towards the boundaries of co-construction and sharing responsibilities, which easily leads to problems such as inconsistencies between candidate adjustment actions and operator authorization scope, unclear parameter scope, and lack of retesting, updates, and rollback basis after adjustment. It is difficult to form a closed-loop processing flow from interference identification and parameter coordination to authorization recording and execution traceability. Summary of the Invention

[0005] Based on the shortcomings of the existing technology described above, the purpose of this invention is to provide an AI interference adaptive coordination method for indoor distributed antenna systems (DAS) co-construction and sharing, so as to solve the above-mentioned technical problems.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an AI interference adaptive coordination method for indoor distributed antenna system (DAS) co-construction and sharing, comprising: Acquire field interference observation data and indoor distribution sharing configuration data of the co-construction and sharing area of ​​indoor distribution system; and merge the field interference observation data according to the sharing area and frequency band affiliation based on the indoor distribution sharing configuration data to generate indoor distribution sharing observation data. Extract frequency band combination relationships, adjacent frequency coupling relationships, and load quality synchronization relationships from indoor distributed monitoring data; identify interference states corresponding to intermodulation interference, adjacent frequency leakage, and load congestion; and map the interference states to the shared area, frequency band resources, and operator affiliation space to generate hierarchical interference situation data. Extract abnormal interference states from the hierarchical interference situation data, call the corresponding mechanism relationship according to the interference type of the abnormal interference state, and trace the abnormal interference state back to the interference contribution record between the source object, the disturbed object and the interventionable parameters. Based on the interference type and interventionable parameters in the interference contribution record, candidate coordination actions are matched, and permission boundaries and parameter ranges of the candidate coordination actions are verified based on the indoor distributed antenna system (DAS) configuration data to generate a permission-constrained coordination action space. A state-action filtering model is used to process hierarchical interference situation data and interference contribution records within the permission-constrained coordinated action space to generate parameters for coordinated actions and action execution ranges. Authorization records are generated based on parameter coordination actions and action execution ranges. After executing parameter coordination actions, retest data is obtained. Layered interference situation data is updated based on the retest data. Rollback status is generated based on the layered interference situation data before and after the update. Authorization records, parameter coordination actions, action execution ranges, retest data, and rollback status are written into the permission chain to generate AI interference adaptive coordination results.

[0007] The present invention is further configured such that, in obtaining the field interference observation data and indoor distributed antenna system (DAS) sharing configuration data of the shared indoor area, and merging the field interference observation data according to the shared area and frequency band affiliation based on the DAS sharing configuration data to generate indoor DAS sharing observation data, the following steps are taken: Field interference observation data is generated by acquiring field test records, operation observation records, and acquisition location information within the shared indoor distribution system area through field acquisition terminals. The indoor distributed antenna system (DAS) resource configuration library is used to read the indoor distributed antenna system (DAS) sharing configuration data corresponding to the indoor distributed antenna system co-construction and sharing area. The configuration fields of the indoor distributed antenna system (DAS) sharing configuration data are parsed to extract the shared area identifier, indoor distributed antenna system (DAS) point identifier, frequency band affiliation identifier, operator affiliation identifier and permission configuration fields, and a configuration association index is established. The source of the field interference observation data is verified and time window is marked. The field interference observation data with time window marking is divided into observation segments corresponding to the indoor distribution point location. Based on the configuration association index, the observation segments are bound to the corresponding shared area and frequency band affiliation to generate regional frequency band observation entries; Based on the operator's affiliation identifier, the regional frequency band observation entries are affixed and completed. For the regional frequency band observation entries with completed affiliation, duplicate records are removed and missing affiliation markers are removed to generate indoor distributed antenna system (DAS) shared observation data.

[0008] The present invention is further configured such that the extraction of frequency band combination relationships, adjacent channel coupling relationships, and load quality synchronization relationships from indoor distributed antenna system (DAS) shared observation data, and the identification of interference states corresponding to intermodulation interference, adjacent channel leakage, and load congestion, includes: Frequency band field parsing is performed on indoor distributed antenna system (DAS) shared observation data to extract frequency band records that are in the same shared area or have indoor DAS link coupling. Frequency band combination relationships are established based on the source frequency band, target frequency band, and combining connection relationship in the frequency band records. Frequency adjacency analysis is performed on indoor distributed monitoring data to extract the spectral position relationship, coverage overlap relationship and adjacent frequency observation intensity between the target frequency band and adjacent frequency bands, and to generate adjacent frequency coupling relationship. Time window alignment is performed on indoor distributed monitoring data, and load change records and network quality change records under the same shared area and frequency band are extracted to establish load quality synchronization relationship. Intermodulation interference state is formed by matching intermodulation interference mechanism based on frequency band combination relationship; adjacent channel leakage state is formed by matching adjacent channel leakage mechanism based on adjacent channel coupling relationship; and load congestion state is formed by matching congestion interference mechanism based on load quality synchronization relationship.

[0009] The present invention is further configured such that the step of mapping the interference state to the shared area, frequency band resources and operator affiliation space to generate hierarchical interference situation data includes: The intermodulation interference status, adjacent channel leakage status and load congestion status are classified, and the shared area identifier, frequency band resource identifier and operator affiliation identifier corresponding to the interference status are extracted to generate an interference status index. Based on the interference status index, the intermodulation interference status, adjacent channel leakage status, and load congestion status are bound to the corresponding shared area, frequency band resource, and operator affiliation, respectively, generating association records between the area layer, frequency band layer, and affiliation layer; The associated records are merged for similar interference and cross-layer association to generate hierarchical interference situation data that includes the location of shared areas, frequency band resource ownership, operator ownership, and interference type.

[0010] The present invention is further configured such that extracting the abnormal interference state from the hierarchical interference situation data includes: The layered interference situation data is sliced ​​according to the shared area, frequency band resources and operator affiliation, and situation slice records corresponding to the shared area and frequency band resources are generated. Based on the type of interference in the situation slice record, the mechanism discrimination conditions corresponding to the frequency band combination relationship, adjacent frequency coupling relationship or load quality synchronization relationship are invoked to perform anomaly matching on the situation slice record; The situation slice records that meet the abnormal matching results are marked as abnormal interference states. The shared area identifier, frequency band resource identifier, operator affiliation identifier and interference type corresponding to the abnormal interference state are retained to generate an abnormal interference state set. The abnormal interference state set is merged within the same region and frequency band to generate abnormal interference state records for interference backtracking processing.

[0011] The present invention is further configured such that, according to the interference type of the abnormal interference state, the corresponding mechanism relationship is invoked to trace the abnormal interference state back to the interference contribution record between the source object, the disturbed object, and the interventionable parameters, including: Based on the interference type recorded in the abnormal interference status record, the corresponding mechanism relationship is selected from the frequency band combination relationship, adjacent frequency coupling relationship and load quality synchronization relationship to generate the backtracking path corresponding to the abnormal interference status. Based on the backtracking path and indoor distributed antenna system (DAS) configuration data, source objects are generated by filtering objects that have frequency band association, coverage association, or affiliation association with abnormal interference states. The shared area, frequency band resources, and operator affiliation of the abnormal interference state are used as the affiliation information of the disturbed object to generate a disturbed object corresponding to the source object. Based on the interference type of the abnormal interference state and the configuration attributes of the source object, adjustable frequency band parameters, power parameters, or time domain parameters are matched to generate interventionable parameters; The interference contribution record is generated by binding the source object, the object being disturbed, the type of interference, and the parameters that can be intervened.

[0012] The present invention is further configured such that the matching of candidate coordination actions based on the interference type and the interveneable parameters in the interference contribution record includes: The interference contribution record is parsed to extract the interference type, source object, disturbed object and interventionable parameters to generate an action matching index; Based on the action matching index, the correspondence between the interference type and the parameter action is invoked, and the interventionable parameters are matched with the corresponding coordinated action type to generate the initial coordinated action; Bind the source object and the disturbed object to the initial coordination action to generate a candidate coordination action that includes the action object, action type and target object; Candidate coordination actions are merged and mutually exclusive based on their origin, and the set of candidate coordination actions corresponding to the interference contribution record is retained.

[0013] The present invention is further configured such that the step of generating a permission-constrained coordination action space by verifying the permission boundaries and parameter ranges of candidate coordination actions based on indoor distributed antenna system (DAS) configuration data includes: Extract operator affiliation information, authorization boundary information, and adjustable parameter range corresponding to candidate coordination actions from indoor distributed antenna system (DAS) configuration data to generate action verification basis; Based on the action verification criteria, the action objects of the candidate coordinated actions are verified for attribution, the objects affected by the candidate coordinated actions are verified for boundaries, and the parameter values ​​of the candidate coordinated actions are verified for range to generate action verification results. The action verification results are matched with the candidate coordination action set, and candidate coordination actions that cross the authorization boundary or exceed the adjustable range of parameters are eliminated to generate an executable coordination action set. Based on the set of executable coordinated actions, an action selection range corresponding to the hierarchical interference situation data and interference contribution records is established, generating a permission-constrained coordinated action space.

[0014] The present invention is further configured such that the state-action filtering model processes hierarchical interference situation data and interference contribution records within the permission-constrained coordinated action space to generate parameter coordination actions and action execution ranges, including: The hierarchical interference situation data is sorted out to extract the shared area, frequency band resources, operator affiliation and interference type corresponding to the abnormal interference status, and a coordinated status record is generated. Organize the contributing objects of the interference contribution records, extract the correspondence between the source object, the disturbed object, the interventionable parameters and the candidate coordinated actions, and generate action association records; Bind the coordination state record and action association record to the permission constraint coordination action space to generate a set of optional actions corresponding to the abnormal interference state; The state-action filtering model coordinates the action space based on the coordination state record, action association record, and permission constraints, and generates parameter coordination actions corresponding to abnormal interference states by filtering the set of optional actions. The action execution scope is generated by coordinating the scope of action in the action space based on the source object, the disturbed object, and the permission constraints associated with the coordinated action.

[0015] The present invention is further configured such that, in the step of generating an authorization record based on parameter coordination actions and action execution range, obtaining retest data after executing parameter coordination actions, updating hierarchical interference situation data based on the retest data, generating a rollback state based on the hierarchical interference situation data before and after the update, and writing the authorization record, parameter coordination actions, action execution range, retest data, and rollback state into the permission chain to generate an AI interference adaptive coordination result, the following steps are included: The execution field of parameter coordination actions and action execution scope is parsed to extract the action object, action area, frequency band resources, parameter adjustment items and operator affiliation. Based on the indoor distributed antenna system (DAS) configuration data, the authorized subject and authorized boundary are matched to generate the authorized record corresponding to the parameter coordination action. Based on the authorization record, the parameter coordination action is converted into an execution task, and the execution status of the execution task is registered to generate the execution record of the parameter coordination action; Within the scope of action execution, acquire on-site retest data after execution, and merge the on-site retest data according to the shared area and frequency band affiliation to generate retest data corresponding to the parameter coordinated action; Based on the retest data, the hierarchical interference situation data within the action execution range is updated to generate post-execution hierarchical interference situation data; The state comparison between the layered interference situation data before and after execution generates a rollback state corresponding to the parameter coordination action. Authorization records, parameter coordination actions, action execution scope, execution records, retest data, and rollback status are written into the permission chain in the order of execution time to generate AI interference adaptive coordination results.

[0016] This invention provides an AI-based adaptive interference coordination method for indoor distributed antenna system (DAS) co-construction and sharing. The method acquires on-site interference observation data and DAS configuration data for the shared area. Based on the DAS configuration data, the on-site interference observation data is merged according to the shared area and frequency band affiliation to generate DAS shared observation data. Frequency band combination relationships, adjacent frequency coupling relationships, and load quality synchronization relationships are extracted from the DAS shared observation data. Intermodulation interference, adjacent frequency leakage, and load congestion-related interference states are identified. These interference states are mapped to the shared area, frequency band resources, and operator affiliation space to generate hierarchical interference situation data. Abnormal interference states are extracted from the hierarchical interference situation data. The corresponding mechanism relationships are invoked according to the interference type of the abnormal interference state, and the abnormal interference state is traced back to the source object, the affected object, and the intervened parameters. Interference contribution records are generated; candidate coordination actions are matched with the interference type and interventionable parameters in the interference contribution records; permission boundaries and parameter ranges of the candidate coordination actions are verified based on the indoor distributed antenna system (DAS) configuration data to generate a permission-constrained coordination action space; a state-action screening model is used to process hierarchical interference situation data and interference contribution records within the permission-constrained coordination action space to generate parameter coordination actions and action execution ranges; authorization records are generated based on parameter coordination actions and action execution ranges; after executing parameter coordination actions, retest data is obtained; hierarchical interference situation data is updated based on the retest data; a rollback state is generated based on the hierarchical interference situation data before and after the update; the authorization records, parameter coordination actions, action execution ranges, retest data, and rollback state are written into the permission chain to generate AI interference adaptive coordination results. The beneficial effects include: 1. Improve the accuracy of interference situation analysis in co-constructed and shared indoor distribution systems: By forming interference status through frequency band combination relationships, adjacent frequency coupling relationships and load quality synchronization relationships, the field observation results are mapped to the shared area, frequency band resources and operator affiliation space, so that the interference situation has the basis for mechanism differentiation and hierarchical association; 2. Enhance the ability to trace back the source of abnormal interference: Extract abnormal interference status from the hierarchical interference situation data, and trace back the source object, the disturbed object and the interventionable parameters according to the interference type, so that parameter coordination no longer depends on the single point observation results, and enhances the pertinence of interference source location and coordination object selection; 3. Improve the controllability of cross-carrier parameter coordination: Generate candidate coordination actions based on interference contribution records, and then combine them with indoor distributed antenna system (DAS) sharing configuration data to form a space for permission-constrained coordination actions, so that parameter coordination actions are consistent with carrier affiliation, authorization boundaries and parameter adjustable range, and support retesting, updating and rollback processing.

[0017] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 The flowchart illustrates an AI interference adaptive coordination method for indoor distributed antenna system (DAS) co-construction and sharing, as an exemplary embodiment of the present invention. Detailed Implementation

[0019] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention and not for limiting the scope of protection of the present invention.

[0020] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0021] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention. Example

[0022] An AI-based adaptive interference coordination method for indoor distributed antenna system (DAS) co-construction and sharing, such as... Figure 1 As shown, it includes: Acquire field interference observation data and indoor distribution sharing configuration data of the co-construction and sharing area of ​​indoor distribution system; and merge the field interference observation data according to the sharing area and frequency band affiliation based on the indoor distribution sharing configuration data to generate indoor distribution sharing observation data. Extract frequency band combination relationships, adjacent frequency coupling relationships, and load quality synchronization relationships from indoor distributed monitoring data; identify interference states corresponding to intermodulation interference, adjacent frequency leakage, and load congestion; and map the interference states to the shared area, frequency band resources, and operator affiliation space to generate hierarchical interference situation data. Extract abnormal interference states from the hierarchical interference situation data, call the corresponding mechanism relationship according to the interference type of the abnormal interference state, and trace the abnormal interference state back to the interference contribution record between the source object, the disturbed object and the interventionable parameters. Based on the interference type and interventionable parameters in the interference contribution record, candidate coordination actions are matched, and permission boundaries and parameter ranges of the candidate coordination actions are verified based on the indoor distributed antenna system (DAS) configuration data to generate a permission-constrained coordination action space. A state-action filtering model is used to process hierarchical interference situation data and interference contribution records within the permission-constrained coordinated action space to generate parameters for coordinated actions and action execution ranges. Authorization records are generated based on parameter coordination actions and action execution ranges. After executing parameter coordination actions, retest data is obtained. Layered interference situation data is updated based on the retest data. Rollback status is generated based on the layered interference situation data before and after the update. Authorization records, parameter coordination actions, action execution ranges, retest data, and rollback status are written into the permission chain to generate AI interference adaptive coordination results.

[0023] The present invention is further configured such that, in obtaining the field interference observation data and indoor distributed antenna system (DAS) sharing configuration data of the shared indoor area, and merging the field interference observation data according to the shared area and frequency band affiliation based on the DAS sharing configuration data to generate indoor DAS sharing observation data, the following steps are taken: Field interference observation data is generated by acquiring field test records, operational observation records, and acquisition location information within the shared indoor distribution system (NDS) area using field acquisition terminals. Specifically, in the shared indoor DDS area of ​​a commercial complex, the field acquisition terminals are configured as handheld acquisition devices or mobile acquisition programs with frequency sweep acquisition, location marking, and operational observation import functions. The acquisition objects include frequency sweep records, operational observation records, and acquisition location information for different test points within each floor. Frequency sweep records include the test frequency band, received signal level, interference level, and acquisition time; operational observation records include changes in service load, network quality, and cell identifiers within the same test time window; and acquisition location information includes floor number, area number, test point number, and adjacent markers for indoor DDS locations. The acquisition frequency can be set to acquire data 3 to 5 times consecutively at each test point, with an interval of 5 to 10 seconds between each acquisition. Operational observation records are imported in 1-minute or 5-minute performance statistics windows. After acquisition, consecutive records at the same test point are timestamped. Records with missing acquisition time, missing frequency band fields, or missing location fields are marked as records to be verified. Records with complete fields and consecutive timestamps are combined to form the field interference observation data. For example, test point P03 in area A of layer B2 forms frequency sweep records in the 1800MHz, 2100MHz and 3500MHz frequency bands. At the same time, the SINR change and PRB occupancy record of this test point within 5 minutes are correlated to form the field interference observation data corresponding to P03. The system reads the indoor distributed antenna system (DAS) sharing configuration data corresponding to the shared indoor DAS area from the indoor DAS resource configuration library. It then parses the configuration fields of this data to extract the shared area identifier, indoor DAS location identifier, frequency band affiliation identifier, operator affiliation identifier, and permission configuration fields, establishing a configuration association index. Specifically, the indoor DAS resource configuration library stores engineering configuration records for the shared indoor DAS area, with data sources including design parameter tables, as-built drawings, frequency planning tables, and operator shared resource tables. When parsing the configuration fields of the shared indoor DAS configuration data, it first reads the building, floor, area, antenna location, combiner, feeder branch, frequency band affiliation, and operator affiliation fields, then establishes a configuration association index based on the engineering correspondence between these fields. The shared area identifier distinguishes the physical coverage area within a floor; the indoor DAS location identifier corresponds to the antenna or test reference point; the frequency band affiliation identifier records the standard or carrier frequency resource to which the frequency band belongs; the operator affiliation identifier records the operator to which the frequency band or equipment belongs; and the permission configuration field records whether the frequency band or equipment is allowed to be adjusted, the types of parameters allowed to be adjusted, and the adjustment boundaries. The underlying logic for configuring the association index is as follows: using the shared area identifier as the spatial index, the indoor distribution point identifier as the equipment index, the frequency band affiliation identifier as the spectrum index, and the operator affiliation identifier as the affiliation index, these four types of indexes are used to create a queryable association table. For example, if antenna point ANT-03 in area A simultaneously carries the 1800MHz frequency band of operator A and the 2100MHz frequency band of operator B, the association index records are configured as "Area A—ANT-03—1800MHz—Operator A" and "Area A—ANT-03—2100MHz—Operator B". Subsequent field observation records can be assigned to the corresponding shared area and frequency band affiliation based on this index. Source verification and time window marking are performed on the field interference observation data. Field interference observation data with completed time window marking is divided into observation segments corresponding to indoor distributed antenna system (DAS) location identifiers. Specifically, when verifying the source of field interference observation data, consistency checks are performed based on the acquisition terminal number, acquisition personnel account, acquisition task number, and acquisition location information. If the acquisition terminal number does not match the acquisition task, or the acquisition location does not belong to the shared area in the indoor DAS shared configuration data, the corresponding record is marked as a source abnormal record. Time window marking is performed on the data that passes source verification, using a preset statistical period as the time window, for example, a 5-minute running observation window. Scanning records and running observation records from the same test point within this window form the same observation segment. The rules for forming observation segments are: records under the same indoor DAS location identifier, the same acquisition time window, and the same frequency band are grouped into the same segment; records spanning two time windows are assigned to the corresponding window according to the acquisition time; for repeatedly uploaded identical test records, the more recent record is retained. Through this process, continuous, discrete, and repeatedly uploaded acquisition data can be organized into mergeable observation segments. For example, test point P03 in area A generates three 1800MHz sweep frequency records and one 5-minute running observation record within the window from 10:00 to 10:05. After being marked with the time window, the observation segment “P03—10:00 / 10:05—1800MHz” is generated. Based on the configuration association index, observation segments are bound to the corresponding shared area and frequency band affiliation to generate regional frequency band observation entries. Specifically, the acquisition location information in the observation segment is matched with the indoor distributed antenna system (DAS) point identifier in the configuration association index. First, the shared area corresponding to the observation segment is determined, and then the frequency band affiliation identifier is matched according to the frequency band field in the observation segment. If the observation segment only contains acquisition location information but lacks an indoor DAS point identifier, the nearest indoor DAS point or preset test reference point is matched according to the floor number, area number, and test point number. If there are multiple matchable points in the same area, the point consistent with the test route specified in the acquisition task is selected first. After completing the location matching and frequency band matching, the observation segment is bound to the shared area-frequency band affiliation entry structure to generate regional frequency band observation entries. Regional frequency band observation entries must retain at least the shared area identifier, frequency band affiliation identifier, observation time window, frequency sweep record summary, running observation summary, and original observation segment reference. This entry serves as the basic data for subsequent extraction of frequency band combination relationships, adjacent frequency coupling relationships, and load quality synchronization relationships. For example, after configuring the associated index matching, the observation segment “P03—10:00 / 10:05—1800MHz” generates the regional frequency band observation entry “Area A—1800MHz—10:00 / 10:05”; The regional frequency band observation entries are completed by assigning ownership based on the operator affiliation identifier. Duplicate records are removed and missing affiliation markers are added to these completed entries, generating indoor distributed antenna system (DAS) shared observation data. Specifically, the regional frequency band observation entries are completed based on the operator affiliation identifier in the configuration association index. If the frequency band affiliation in a regional frequency band observation entry is already registered in the configuration association index, the corresponding operator affiliation is written. If the same frequency band is shared or overlapped by multiple operators within the shared area, the corresponding affiliation record is determined based on the indoor DAS location identifier, frequency planning table, and operator affiliation identifier. When removing duplicate records from completed regional frequency band observation entries, the shared area, frequency band affiliation, operator affiliation, and time window are used as the criteria for duplicate judgment, retaining entries with high field completeness and more recent acquisition times. Entries that cannot match operator or frequency band affiliation are marked as missing affiliation, but the original observation segment reference is retained for subsequent manual verification or configuration updates. After the above processing, indoor DAS shared observation data is formed, whose data structure includes regional frequency band observation entries, operator affiliation, time window, observation record summary, and missing markers. Taking Area A as an example, the 1800MHz observation entries are completed to be assigned to Operator A, and the 2100MHz observation entries are completed to be assigned to Operator B. Duplicate 1800MHz old records are removed, and 2300MHz records without assignments are marked as pending verification. This generates indoor distributed observation data that can be used for subsequent interference status identification.

[0024] The present invention is further configured such that the extraction of frequency band combination relationships, adjacent channel coupling relationships, and load quality synchronization relationships from indoor distributed antenna system (DAS) shared observation data, and the identification of interference states corresponding to intermodulation interference, adjacent channel leakage, and load congestion, includes: Frequency band field parsing is performed on the indoor distributed antenna system (DAS) shared observation data to extract frequency band records located in the same shared area or with indoor DAS link coupling. Frequency band combination relationships are established based on the source frequency band, target frequency band, and combining connection relationship in the frequency band records. Specifically, in Area A of Floor B2 of a commercial building where three operators share the same indoor DAS combining link, the indoor DAS shared observation data already includes the shared area identifier, frequency band affiliation, operator affiliation, indoor DAS location, frequency sweep records, operational observation records, and time window. When parsing the frequency band field in this data, frequency band records located in the same area, under the same combiner, or under the same feeder branch are first filtered according to the shared area and indoor DAS link identifier. Then, the source frequency band, target frequency band, operator affiliation, and combining connection relationship in the frequency band records are read. The frequency band combination relationship refers to the engineering correspondence between two source frequency bands and one target interfered frequency band located in the same shared area or with a combining propagation path. Its source is the frequency planning table, combiner connection table, and on-site frequency sweep records. Taking Area A as an example, if Operator A's 1800MHz band and Operator B's 2100MHz band enter the ANT-03 antenna location through the same combiner, and an abnormal rise occurs near the target frequency band in the frequency sweep record, then "1800MHz—2100MHz—Target Frequency Band—ANT-03—Area A" is registered as a frequency band combination relationship. This processing organizes the scattered frequency band observation records into a structured relationship that can be matched for intermodulation mechanisms, avoiding subsequent intermodulation judgments relying solely on single-point interference intensity. Frequency adjacency analysis is performed on indoor distributed antenna system (DAS) shared observation data to extract the spectral positional relationship, coverage overlap relationship, and adjacent frequency observation intensity between the target frequency band and adjacent frequency bands, generating adjacent frequency coupling relationships. Specifically, frequency adjacency analysis uses a frequency planning table and regional frequency band observation entries as input, filters according to the spectral positional relationship between the target frequency band and adjacent frequency bands, and extracts the target frequency band, adjacent frequency bands, adjacent frequency observation intensity, and coverage overlap relationship. The coverage overlap relationship is derived from the coverage configuration of the DAS points, the location of the test points, and the on-site frequency sweep intensity distribution, and is used to indicate whether adjacent frequency band signals arrive at the disturbed location simultaneously within the same shared area. The adjacent frequency coupling relationship refers to the potential leakage relationship formed by adjacent frequency bands within the same shared area due to spectral adjacency and coverage overlap, and its recorded content includes the source of the adjacent frequency band, the target frequency band, the overlapping area, and the adjacent frequency observation intensity. Taking area A as an example, if the target frequency band is the 1800MHz band of operator A, and the adjacent frequency band is the carrier frequency near 1810MHz of operator B, and the intensity of the adjacent frequency band is higher than the stable observation level in the area at both test points P03 and P04, then the adjacent frequency band and the target frequency band form an adjacent-channel coupling relationship in area A. This step provides two criteria for determining adjacent-channel leakage: spectral adjacency and spatial coverage. Subsequently, it can distinguish adjacent-channel leakage from ordinary background noise increases. Time window alignment is performed on the indoor distributed antenna system (DAS) shared observation data to extract load change records and network quality change records under the same shared area and frequency band, establishing a load quality synchronization relationship. Specifically, time window alignment uses the observation time in the area and frequency band observation entries as the benchmark, grouping operational observation records and network quality change records under the same shared area and frequency band into the same statistical window. Operational observation records can be imported from the network management performance statistics interface or field acquisition terminals, with a sampling period of 1 minute or 5 minutes. Network quality change records can include indicators corresponding to the scenario, such as SINR, RSRQ, call drop rate, access failure rate, or scheduling success rate. The load quality synchronization relationship refers to the associated records where load increases and network quality decreases simultaneously within the same time window. Its formation process includes timestamp alignment, missing window removal, and outlier filtering. Outlier filtering can use a continuous window confirmation method. For example, if two consecutive 5-minute windows within the same shared area show loads higher than the historical stable range and network quality lower than the corresponding stable range, the synchronization relationship is retained. Taking the 1800MHz band in Area A as an example, if the PRB occupancy rate increases from 55% to 82% in two consecutive windows from 10:00 to 10:10, and the SINR decreases from 18dB to 9dB, then a "Area A—1800MHz—Operator A—Load Quality Synchronization Relationship" is formed. This step establishes the load congestion judgment on the synchronous fact of load changes and quality degradation, providing a time dimension basis for the subsequent generation of congestion interference status. Intermodulation interference states are formed based on frequency band combination relationships; adjacent channel leakage states are formed based on adjacent channel coupling relationships; and load congestion states are formed based on load quality synchronization relationships. Specifically, after extracting the three types of relationships, state matching is performed according to preset engineering mechanism rules. When matching intermodulation interference mechanisms, the two source frequency bands and the target frequency band in the frequency band combination relationship are read. The intermodulation products are then combined with the combining connection relationship to determine whether they fall within the target frequency band protection range. The target frequency band protection range is derived from the frequency planning table or network standard bandwidth configuration. An intermodulation interference state is formed when the frequency combination is close to the target frequency band and there is a combining or coverage coupling path. When matching adjacent channel leakage mechanisms, the adjacent frequency band, target frequency band, and adjacent channel observation strength in the adjacent channel coupling relationship are read. The adjacent channel leakage is then combined with the equipment's adjacent channel suppression requirements or the field stability observation level to determine whether the adjacent channel leakage enters the target frequency band reception range. An adjacent channel leakage state is formed when spectrum adjacency, coverage overlap, and abnormal adjacent channel strength are satisfied. When matching congestion interference mechanisms, the load changes and network quality changes in the load quality synchronization relationship are read. Combined with the historical stability window of the shared area, it is determined whether there is a continuous change in load increase accompanied by quality decrease. When the continuous window synchronization change is satisfied, a load congestion state is formed. The output of this step corresponds to three interference states. The state record retains the interference type, shared area, frequency band affiliation, operator affiliation, and triggering basis. This can be directly mapped to hierarchical interference situational data. In the B2 layer A area embodiment, the field acquisition terminal obtains frequency sweep records at test points P03 and P04. The network management performance interface synchronously provides 5 minutes of operational observation records, and the indoor distribution resource configuration library provides combiner connection, frequency band affiliation, and operator affiliation. Frequency band analysis revealed that 1800MHz and 2100MHz form a frequency band combination relationship under the ANT-03 merging path, and the sweep noise near the target frequency band increases, generating an intermodulation interference state. Frequency adjacency analysis showed that the adjacent frequency band near 1810MHz overlapped with 1800MHz in area A, and the adjacent channel observation intensity was continuously higher than the stable record, generating an adjacent channel leakage state. Time window alignment showed that the 1800MHz frequency in area A experienced a synchronous change in load increase and SINR decrease from 10:00 to 10:10, generating a load congestion state. After outputting the three states, they can be verified with the manual investigation results: the original manual investigation only marked the presence of interference in area A, but this embodiment further obtains the three states of intermodulation, adjacent channel leakage, and load congestion and their triggering basis, so that subsequent layered interference situation data can retain the specific interference mechanism, regional location, and frequency band affiliation. The verification results show that after the indoor distributed observation data is extracted and the mechanism is matched, a traceable record of the interference state can be formed, which provides a structured input for subsequent abnormal interference extraction and interference contribution backtracking.

[0025] The present invention is further configured such that the step of mapping the interference state to the shared area, frequency band resources and operator affiliation space to generate hierarchical interference situation data includes: Intermodulation interference, adjacent channel leakage, and load congestion states are categorized, and the shared area identifier, frequency band resource identifier, and operator affiliation identifier corresponding to each interference state are extracted to generate an interference state index. Specifically, in the shared indoor distribution network area on floor B2 of a commercial complex, the indoor distribution network management platform has received the intermodulation interference, adjacent channel leakage, and load congestion states generated in the previous steps. Each interference state carries a shared area identifier, frequency band resource identifier, operator affiliation identifier, interference type, trigger time window, and reference to the original observation entry. Among them, the shared area identifier comes from the floor and area division in the indoor distribution resource configuration database, the frequency band resource identifier comes from the frequency planning table and frequency band affiliation record, the operator affiliation identifier comes from the shared resource registration table, and the interference type comes from the previous mechanism matching results. When categorizing the above interference states, state records lacking area identifiers, frequency band resource identifiers, or operator affiliation identifiers are first removed and marked as pending verification states; then, the interference state index is generated in the order of "shared area identifier—frequency band resource identifier—operator affiliation identifier—interference type—time window". This index is used to uniquely locate the position of a specific interference state within a shared area, frequency band resource, and operator affiliation. For example, if the 1800MHz band in Area A belongs to Operator A, and a load congestion state occurs between 10:00 and 10:05, then an interference state index of "Area A—1800MHz—Operator A—Load Congestion—10:00 / 10:05" is generated; if the 2100MHz band in Area A belongs to Operator B, and an adjacent channel leakage state occurs between 10:00 and 10:05, then a corresponding adjacent channel leakage index is generated. Through this process, scattered interference states are transformed into queryable, bindable, and traceable index records, providing a unified entry point for subsequent hierarchical mapping. Based on the interference state index, intermodulation interference state, adjacent channel leakage state, and load congestion state are bound to the corresponding shared area, frequency band resource, and operator affiliation, respectively, generating association records between the area layer, frequency band layer, and affiliation layer. Specifically, after the interference state index is generated, each index is mapped to the area layer, frequency band layer, and affiliation layer. The area layer represents the spatial unit of the indoor distributed antenna system (DAS) co-construction and shared area, such as a floor, zone, or antenna coverage area; the frequency band layer represents the frequency band resources carried within this area, such as 1800MHz, 2100MHz, or 3500MHz; and the affiliation layer represents the operator entity corresponding to the frequency band resource or equipment resource. During mapping, the area layer node is first determined based on the shared area identifier in the interference state index, then the frequency band layer node is determined based on the frequency band resource identifier, and finally the affiliation layer node is determined based on the operator affiliation identifier. The interference state is used as the associated event connecting the area layer, frequency band layer, and affiliation layer. If different operators share or overlap deployments of the same frequency band within the same area, association records are established separately according to the operator affiliation identifier in the configured association index to avoid interference states from different operators being mixed into the same affiliation link. Taking Area A of Layer B2 as an example, the load congestion status of Operator A at 1800MHz is bound to "Area A—1800MHz—Operator A", and the adjacent channel leakage status of Operator B at 2100MHz is bound to "Area A—2100MHz—Operator B". If there is an adjacent channel coupling relationship between the two, the adjacent channel coupling source reference is retained in the association record. The association record generated in this step includes regional layer nodes, frequency band layer nodes, home layer nodes, interference type, trigger window, and mechanism source reference, enabling the subsequent system to view the interference status according to spatial, spectral, or operator dimensions. The process involves merging similar interference records and cross-layer correlation analysis to generate hierarchical interference situation data, including shared area location, frequency band resource ownership, operator ownership, and interference type. Specifically, when organizing related records, records are first merged based on the same shared area, frequency band resource, operator ownership, and interference type. The merging condition can be set to the same interference type appearing in two consecutive statistical windows or at adjacent test points. The statistical window can be 5 minutes, derived from the unified time granularity of the field acquisition terminal and network management performance statistics interface. During merging, the earliest trigger time, most recent trigger time, associated observation entries, and mechanism source references are retained, and the original records are not deleted for subsequent backtracking. Subsequently, cross-layer correlation analysis is performed to write the intermodulation correlations and adjacent frequency correlations between different frequency bands within the same area, as well as the shared impact relationships between different operator ownership within the same frequency band, into the hierarchical interference situation data. The hierarchical interference situation data includes shared area location, frequency band resource ownership, operator ownership, interference type, trigger window, correlation status, and original record references. Taking the implementation example, in area A of layer B2, from 10:00 to 10:10, operator A experienced 1800MHz load congestion for two consecutive windows, while from 10:00 to 10:05, operator B experienced 2100MHz adjacent channel leakage. After merging similar data, two situation records were generated: one for "Area A - 1800MHz - Operator A - Load Congestion" and another for "Area A - 2100MHz - Operator B - Adjacent Channel Leakage". After cross-layer processing, both situation records were attached to the area layer of area A, retaining the frequency band layer and the attribution layer path. Compared with only outputting single-point interference alarms, this layered interference situation data can directly reflect the interference occurrence area, affected frequency band, and operator affiliation, providing a structured basis for subsequent abnormal interference extraction, interference contribution tracing, and generation of permission constraint coordination actions.

[0026] The present invention is further configured such that extracting the abnormal interference state from the hierarchical interference situation data includes: The layered interference situation data is sliced ​​according to shared area, frequency band resources, and operator affiliation, generating situation slice records corresponding to the shared area and frequency band resources. Specifically, in the shared indoor distribution network area on the B2 floor of a commercial complex, the indoor distribution network management platform receives the previously generated layered interference situation data. This data includes the location of the shared area, frequency band resource affiliation, operator affiliation, interference type, trigger time window, and references to the original observation records. A situation slice refers to the smallest situation analysis unit formed by a shared area, a frequency band resource, an operator affiliation, and a time window. Its source is the associated records of the area layer, frequency band layer, and affiliation layer in the layered interference situation data. During implementation, a 5-minute situation slice time window is first used to group interference status records under the same shared area, the same frequency band resource, and the same operator affiliation into the same slice. Then, records lacking time windows, area identifiers, or frequency band identifiers are removed or marked as pending verification. Subsequently, the interference type, observation intensity, quality changes, load changes, and mechanism source references corresponding to each slice are retained to generate situation slice records. For example, "Area A—1800MHz—Operator A—10:00 / 10:05" can be formed into one situational slice, and "Area A—2100MHz—Operator B—10:00 / 10:05" can be formed into another situational slice. This processing ensures that subsequent anomaly detection is based on clear spatial, frequency band, and attribution targets, avoiding the mixing of data from different areas or different operators; Based on the interference type in the situational awareness slice records, the mechanism discrimination conditions corresponding to the frequency band combination relationship, adjacent frequency coupling relationship, or load quality synchronization relationship are invoked to perform anomaly matching on the situational awareness slice records. Specifically, the interference type of each situational awareness slice record is read, and the corresponding mechanism discrimination conditions are invoked according to the interference type. If the interference type is intermodulation interference, the discrimination conditions corresponding to the frequency band combination relationship are invoked to check whether the source frequency band combination in the slice comes from the same combining path, whether the intermodulation landing point is close to the protection range of the target frequency band, and whether the observation intensity of the target frequency band increases within the continuous window. If the interference type is adjacent frequency leakage, the discrimination conditions corresponding to the adjacent frequency coupling relationship are invoked to check whether there is spectral adjacency between the target frequency band and adjacent frequency bands, whether the coverage area overlaps, and whether the adjacent frequency observation intensity exceeds the stable observation limit of the area. If the interference type is load congestion, the discrimination conditions corresponding to the load quality synchronization relationship are invoked to check whether the service load increases within the same time window, whether the network quality decreases, and whether this change is maintained within the continuous window. The stable observation limit can be formed by the normal observation records of the same time period in the same shared area over the past 7 days, and the continuous window can be set to two 5-minute windows. Through the above-mentioned categorized matching, each situation slice record generates an abnormal matching result of "meeting abnormal conditions" or "not meeting abnormal conditions", and retains the mechanism source that triggered the result for easy backtracking later. Situational slice records that meet the abnormal matching results are marked as abnormal interference states. The shared area identifier, frequency band resource identifier, operator affiliation identifier, and interference type corresponding to the abnormal interference state are retained, generating an abnormal interference state set. Specifically, after completing the abnormal matching, situational slice records that meet the corresponding mechanism discrimination conditions are marked as abnormal states. An abnormal interference state refers to an interference state that simultaneously possesses area affiliation, frequency band affiliation, operator affiliation, and interference type, and is confirmed by the mechanism discrimination conditions. During marking, the shared area identifier, frequency band resource identifier, operator affiliation identifier, interference type, trigger time window, abnormal matching result, and original situational slice reference are retained, generating an abnormal interference state set. Taking Area A of Layer B2 as an example, if, within two consecutive windows from 10:00 to 10:10, the 1800MHz operator band A in Area A experiences increased load accompanied by a decrease in SINR, and this change satisfies the load quality synchronization relationship discrimination condition, then the corresponding slice is marked as "load congestion abnormal interference state." If, in the 2100MHz operator band B in Area A, there is both an increase in intensity of adjacent bands and coverage overlap, then the corresponding slice is marked as "adjacent channel leakage abnormal interference state." This step converges the normal situation records into abnormal objects that can enter the interference backtracking process. The system merges abnormal interference state sets within the same region and frequency band to generate abnormal interference state records for interference backtracking. Specifically, when merging abnormal interference state sets, the shared region identifier and frequency band resource identifier are used as the primary merging conditions, while the operator affiliation identifier, interference type, and trigger time window are used as secondary merging conditions. For the same interference type occurring consecutively within the same region, frequency band, and operator affiliation, they are merged into a single abnormal interference state record, retaining the first trigger window, last trigger window, relevant situational slice references, and mechanism identification source. For cases where intermodulation interference and adjacent channel leakage coexist within the same region and frequency band, separate abnormal interference state records for different interference types are generated, while retaining common region and frequency band information in the records. Taking a case study, if Operator A (1800MHz) in Area A meets the load congestion criteria for three consecutive windows from 10:00 to 10:15, the system will merge the three abnormal interference states into a single abnormal interference state record: "Area A—1800MHz—Operator A—Load Congestion". If Operator B (2100MHz) in Area A meets the adjacent channel leakage criteria from 10:00 to 10:05, then a separate abnormal interference state record will be formed. The number of records after merging is reduced, but the original slice references and mechanistic basis are still retained, which can be directly used as input for subsequent backtracking of source objects, disturbed objects, and modifiable parameters.

[0027] The present invention is further configured such that, according to the interference type of the abnormal interference state, the corresponding mechanism relationship is invoked to trace the abnormal interference state back to the interference contribution record between the source object, the disturbed object, and the interventionable parameters, including: Based on the interference type in the abnormal interference status record, corresponding mechanism relationships are selected from frequency band combination relationships, adjacent channel coupling relationships, and load quality synchronization relationships to generate a backtracking path corresponding to the abnormal interference status. Specifically, in the indoor distributed antenna system (DAS) co-construction and sharing scenario of Area A on the B2 floor of a commercial complex, the abnormal interference status record originates from the previous anomaly extraction results. Its data content includes shared area identifier, frequency band resource identifier, operator affiliation identifier, interference type, trigger time window, and situational slice reference. When parsing the abnormal interference status record, the interference type field is read first. If the interference type is intermodulation interference, the previously established frequency band combination relationship is invoked; if the interference type is adjacent channel leakage, the adjacent channel coupling relationship is invoked; if the interference type is load congestion, the load quality synchronization relationship is invoked. The backtracking path refers to the relationship chain that searches backward from the abnormal interference status to the possible interference source. Its source is the frequency band combination, adjacent channel coupling, or load quality synchronization record retained during the previous mechanism identification process. For example, when a carrier's 1800MHz frequency band in Area A is marked as having an adjacent channel leakage anomaly, the system reads the situational slice reference corresponding to the anomaly record, searches for adjacent frequency bands that have spectrum adjacency and coverage overlap with 1800MHz, and generates a backtracking path of "Area A—1800MHz disturbed frequency band—adjacent frequency band—corresponding indoor distribution point—carrier affiliation". Through this path, subsequent screening targets have a clear data source and engineering relationship basis; Based on the backtracking path and indoor distributed antenna system (DAS) sharing configuration data, source objects are generated by filtering objects that have frequency band associations, coverage associations, or affiliation associations with abnormal interference states. Specifically, source objects refer to frequency bands, DAS locations, combining links, or operator-affiliated objects that may contribute to abnormal interference states. During implementation, the backtracking path is matched with the frequency band affiliation, location coverage, combining connection, and operator affiliation in the DAS sharing configuration data. For intermodulation interference, source frequency bands and combining locations that share the same combining path as the disturbed frequency band and whose frequency band combination falls within the intermodulation discrimination range are selected. For adjacent channel leakage, adjacent frequency bands and their affiliated locations that are adjacent to the disturbed frequency band, have overlapping coverage in the same shared area, and whose adjacent channel observation intensity is higher than the stable observation upper limit for two consecutive 5-minute windows are selected. For load congestion, bearer objects with continuously increasing load and synchronously decreasing network quality under the same frequency band affiliation in the same area are selected. Taking the 1800MHz adjacent channel leakage anomaly in Area A as an example, the configuration data shows that the frequency band near Operator B's 1810MHz overlaps with Operator A's 1800MHz frequency band in the ANT-03 and ANT-04 coverage areas, and the adjacent channel observation intensity is continuously higher than the stable upper limit by 3dB from 10:00 to 10:10. Therefore, the corresponding frequency band of Operator B, the ANT-03 coverage point, and the corresponding attribution record are selected as the source objects. This processing ensures that the source objects are jointly limited by the mechanism path and configuration relationship, reducing the inclusion of irrelevant frequency bands or irrelevant areas in the backtracking scope. The shared area, frequency band resources, and operator affiliation of the abnormal interference state are used as the attribution information of the disturbed object to generate a disturbed object corresponding to the source object. Specifically, the disturbed object refers to the combination of area, frequency band, and operator affiliation actually affected by the abnormal interference state. During implementation, the shared area identifier, frequency band resource identifier, and operator affiliation identifier in the abnormal interference state record are read and verified for consistency with the area layer, frequency band layer, and attribution layer records in the hierarchical interference situation data. If the frequency band resources in the abnormal record are consistent with the frequency band affiliation registered in the configuration association index, a disturbed object is generated. If different operators share or overlap deployments in the same area and frequency band, disturbed objects are formed according to the operator affiliation identifier in the abnormal state record. Taking the 1800MHz adjacent channel leakage anomaly of operator A in area A as an example, the disturbed object record is "Area A—1800MHz—Operator A—ANT-03 / ANT-04 coverage area", and the corresponding situation slice, frequency sweep record, and operation observation record reference are retained. This disturbed object establishes a one-to-one or one-to-many correspondence with the source object selected in the previous step, providing a clear boundary for the subsequent determination of interventionable parameters. Based on the interference type and configuration attributes of the source object in the abnormal interference state, adjustable frequency band parameters, power parameters, or time domain parameters are matched to generate interventionable parameters. Specifically, interventionable parameters refer to frequency band, power, or time domain parameters that can be adjusted within the permitted range of the permission configuration fields. These parameters originate from the permission configuration fields and equipment operating parameter records in the indoor distributed antenna system (DAS) configuration data. During implementation, the configuration attributes of the source object are first read, including frequency band affiliation, equipment affiliation, power configuration, carrier frequency configuration, and time slot configuration. Then, parameter categories are selected according to the interference type. For intermodulation interference, priority is given to matching carrier frequency adjustment, power limiting, or combining link isolation configurations related to the source frequency band combination. For adjacent frequency leakage, priority is given to matching adjacent frequency band power limiting, guard band adjustment, or coverage boundary control. For load congestion, priority is given to matching service peak shifting, carrier frequency resource switching, or capacity sharing parameters. Taking the 1800MHz adjacent channel leakage anomaly in Area A as an example, the source objects are the adjacent frequency band of Operator B (1810MHz) and the ANT-03 coverage point. Configuration data indicates that downlink power reduction is allowed for this frequency band, with an allowable adjustment range of 1dB to 3dB. Therefore, "adjacent frequency band downlink power limiting" is considered an operable parameter. If the permission configuration field shows that carrier frequency changes are not allowed for this frequency band, then carrier frequency reconfiguration is not included in the operable parameters. This process limits the intervention parameters to the actual configuration attributes and permission boundaries of the source objects, providing a feasible basis for subsequent candidate coordination actions. Interference contribution records are generated by binding the source object, the affected object, the interference type, and the interventionizable parameters. Specifically, the interference contribution record is a structured record describing "who caused the impact, who was affected, what type of interference it is, and what parameters can be adjusted" in an abnormal interference state. During implementation, the source object identifier, affected object identifier, interference type, backtracking path number, interventionizable parameters, and original observation reference are written into the same record. For cases where there are multiple source objects for the same affected object, records are generated separately according to the mechanism matching results between the source object and the affected object, and the corresponding trigger windows and discrimination criteria are retained. An implementation example is used for verification: Operator A in Area A's 1800MHz band experienced an adjacent channel leakage anomaly from 10:00 to 10:10. The backtracking path points to the adjacent 1810MHz band of Operator B. The source object is "Operator B's 1810MHz band—ANT-03 coverage point," the affected object is "Operator A's 1800MHz band—Area A coverage area," and the interventionizable parameter is "adjacent band downlink power limiting of 1dB to 3dB." The system generates an interference contribution record based on this and passes it to the subsequent candidate coordination action matching process. After on-site verification, this record can directly locate the source of adjacent frequency bands, the interfered frequency band, and the range of parameters that can be adjusted. Compared with the alarm method that only records "interference exists in area A", the generation of subsequent coordination actions has a clear target and permission basis.

[0028] The present invention is further configured such that the matching of candidate coordination actions based on the interference type and the interveneable parameters in the interference contribution record includes: Interference contribution records are parsed to extract interference type, source object, affected object, and modifiable parameters to generate an action matching index. Specifically, in the shared indoor distribution network (ADN) area of ​​Zone A on floor B2 of a commercial complex, the ADN management platform receives previously generated interference contribution records. These records include the source object, affected object, interference type, modifiable parameters, backtracking path number, and original observation reference. The source object can be an adjacent frequency band, a combination of source and frequency bands, a combined link, or a load-bearing object; the affected object can be a shared area experiencing abnormal interference, a combination of frequency band resources, and operator affiliation; the modifiable parameters originate from the permission configuration fields and equipment parameter records in the ADN shared configuration data. When parsing the interference contribution records, the interference type field is first checked to ensure it belongs to intermodulation interference, adjacent channel leakage, or load congestion. Then, the modifiable parameters are checked to ensure they match the configuration attributes of the source object. After the fields are complete, the interference type, source object, affected object, and modifiable parameters are organized into an action matching index. The action matching index describes the correspondence between "what kind of interference, which object caused it, which object it affects, and which parameter can be adjusted." For example, if the record shows that "Operator A in Area A at 1800MHz is affected by adjacent channel leakage from Operator B at 1810MHz, and the intervention parameter is downlink power limiting in the adjacent frequency band," then an action matching index of "adjacent channel leakage—Operator B 1810MHz—Operator A 1800MHz—downlink power limiting" is generated. This index provides a clear entry point for subsequent action matching, ensuring that coordinated actions do not deviate from the interference contribution record; Based on the action matching index, the correspondence between interference types and parameter actions is invoked to match the interventionable parameters with the corresponding coordinated action types to generate initial coordinated actions. Specifically, the correspondence between interference types and parameter actions is pre-stored in the coordinated action rule table. The rule table is jointly formed by indoor distribution optimization experience, a list of adjustable equipment parameters, and operator permission configurations, and is maintained along with the indoor distribution resource configuration library. When processing the action matching index, the interference type is read first, then the category of interventionable parameters is read, and then the corresponding action type is searched in the coordinated action rule table. For intermodulation interference, when the interventionable parameter is source band power or carrier frequency configuration, the action that can be matched is source band power limiting, carrier frequency staggering, or combined link isolation verification. For adjacent channel leakage, when the interventionable parameter is adjacent band power or coverage boundary, the action that can be matched is adjacent channel power limiting, coverage boundary contraction, or guard band adjustment. For load congestion, when the interventionable parameter is time domain resources or bearer frequency band, the action that can be matched is service period staggering, carrier frequency resource sharing, or load migration. Taking adjacent channel leakage records in Area A as an example, the interference type in the action matching index is adjacent channel leakage, and the interventionable parameter is downlink power limiting at 1810MHz for Operator B. The rule table matches it as the initial coordination action of "adjacent frequency band power limiting" and records the action triggering basis, parameter direction, and allowed adjustment category. This process converts the mechanism type and adjustable parameters into specific action types, avoiding the formation of candidate actions by manual ad-hoc judgment. The process involves binding the source object and the affected object to the initial coordination action to generate candidate coordination actions that include the action object, action type, and target object. Specifically, after generating the initial coordination action, the source object in the interference contribution record is bound as the action object, and the affected object is bound as the target object. The action object represents the object for which parameter adjustments or status checks need to be performed, and the target object represents the affected area, frequency band resources, and operator affiliation corresponding to the action. During binding, the indoor distribution point, frequency band affiliation, and operator affiliation of the source object are read, and then the shared area, affected frequency band, and operator affiliation of the affected object are read. These two information, along with the initial coordination action, are combined to form candidate coordination actions. Candidate coordination actions must include at least the action object, action type, target object, interveneable parameters, triggered interference type, and reference to the original contribution record. Taking the implementation example, the initial coordination action is "power limiting in adjacent frequency bands," the source object is "Operator B's 1810MHz band—ANT-03 coverage point," and the affected object is "Operator A's 1800MHz band—Area A coverage area." Then, the candidate coordination action is generated as "perform downlink power limiting on Operator B's 1810MHz adjacent frequency band, with the affected object being Operator A's 1800MHz affected frequency band in Area A." This candidate coordination action preserves the contribution relationship between the source and affected objects, and subsequent permission checks can directly determine the action subject, scope, and parameter boundaries. Candidate coordination actions are subject to source merging and mutual exclusion sorting. The set of candidate coordination actions corresponding to the interference contribution record is retained. Specifically, after candidate coordination actions are generated, actions from the same source object, of the same action type, and within the same or consecutive time window are subject to source merging. Source merging means combining repeated actions from the same source object that point to nearby disturbed objects into a single action record, while retaining the list of disturbed objects involved and the original contribution record reference. For example, when Operator B's 1810MHz band simultaneously affects Operator A's 1800MHz band corresponding to test points P03 and P04 in Area A, the two "adjacent frequency band power limiting" actions are merged into a single candidate coordination action covering the corresponding points in Area A. Mutual exclusion sorting is used to handle actions with opposite parameter directions or that cannot be executed simultaneously within the same time window, such as simultaneous power increase and decrease in the same frequency band, or simultaneous retention and migration actions on the same carrier frequency. During the data processing, the retention order was determined based on interference type, mechanism triggering duration window, and permission configuration fields. Actions with longer duration windows, complete mechanism sources, and clearly defined parameter permissions were retained first, while the remaining actions were marked as actions awaiting review. For example, the system generated three candidate actions between 10:00 and 10:10. Two of these were power limiting actions in the adjacent 1810MHz band of Operator B, which were merged into one candidate action after source aggregation. The third action was carrier frequency adjustment in the same band, but the permission configuration field did not allow carrier frequency adjustment, so it was not included in the candidate coordination action set. The final candidate coordination action set included the "Power Limiting Action in the Adjacent 1810MHz Band of Operator B" and carried the corresponding interference contribution record number for subsequent permission boundary and parameter range verification.

[0029] The present invention is further configured such that the step of generating a permission-constrained coordination action space by verifying the permission boundaries and parameter ranges of candidate coordination actions based on indoor distributed antenna system (DAS) configuration data includes: The system extracts operator affiliation information, authorized boundary information, and adjustable parameter ranges corresponding to candidate coordinated actions from the indoor distributed antenna system (DAS) sharing configuration data to generate action verification criteria. Specifically, in a multi-operator indoor DAS co-construction and sharing scenario in Zone A of Floor B2 of a commercial complex, the DAS management platform receives a set of candidate coordinated actions generated previously. These candidate actions include the action object, action type, target object, operable parameters, and corresponding interference contribution record number. The DAS sharing configuration data originates from the indoor DAS resource configuration library, which is maintained through the engineering parameter management interface, frequency planning table, and operator shared resource table. Fields include shared area, indoor DAS location, frequency band affiliation, operator affiliation, adjustable parameter items, and authorized boundaries. During implementation, the system first queries the corresponding operator affiliation based on the action object in the candidate coordinated actions, then queries the shared area and frequency band range that the action is allowed to affect based on the target object, and finally reads the adjustable parameter range. Authorized boundaries refer to the boundaries of the area, frequency band, location, and parameter type that the operator or management entity is allowed to adjust. Adjustable parameter ranges refer to the allowed value range of executable parameters or the allowed action categories. For example, if the candidate coordination action is "Power limiting in adjacent frequency bands of Operator B's 1810MHz band", and the configuration database records that this frequency band belongs to Operator B, allowing downlink power limiting within the coverage area of ​​ANT-03 to ANT-04 in Area A, with an allowable adjustment range of 1dB to 3dB, then the attribution, area boundary, and adjustment range are extracted to generate the action verification basis. This basis retains the configuration record number and update time, enabling subsequent verification processes to trace back to the specific configuration source. Based on the action verification criteria, the action objects of candidate coordinated actions are verified for attribution, the affected objects for boundary verification, and the parameter values ​​for candidate coordinated actions are verified for range, generating action verification results. Specifically, when verifying candidate coordinated actions, the attribution of the action object is first verified to check whether the frequency band, indoor distribution point, or equipment object in the action object is consistent with the operator attribution in the indoor distribution sharing configuration data; then, the boundary of the affected object is verified to check whether the shared area, disturbed frequency band, and coverage point affected by the candidate action are within the authorized boundary; finally, the parameter value range is verified to check whether the power limiting, carrier frequency adjustment, or time-domain peak shifting parameters in the candidate action are within the adjustable range. Each verification generates a pass, fail, or pending review status, and records the triggering reason. Taking Area A as an example, if the candidate action is "Operator B 1810MHz downlink power limiting by 2dB", the action object belongs to the same operator B in the configuration library, the effective area ANT-03 to ANT-04 in Area A is within the authorized boundary, and 2dB is within the allowed range of 1dB to 3dB, then this candidate action generates a verification pass result. If another candidate action is "Operator B 1810MHz carrier frequency migration", but the permission configuration field does not enable carrier frequency migration, then the parameter range verification generates a failure result. Through this item-by-item verification process, the executing subject, impact boundary, and parameter values ​​of the candidate coordination actions all form recordable judgment results. The action verification results are matched with the candidate coordination action set. Candidate coordination actions that cross the authorization boundary or exceed the adjustable parameter range are eliminated to generate an executable coordination action set. Specifically, after verification, the verification result of each action is written back to the candidate coordination action set according to the candidate coordination action number. For candidate coordination actions that pass the attribution verification, boundary verification, and parameter range verification, they are retained as executable coordination actions. For candidate coordination actions that cross the authorization boundary, have inconsistent action object attribution, or have parameter values ​​exceeding the adjustable range, they are eliminated and the reason for elimination is marked. For candidate coordination actions with missing configuration fields or outdated authorization records, they are marked as actions pending review and are not included in the executable coordination action set. The executable coordination action set retains the action object, action type, target object, allowed parameter range, verification basis number, and interference contribution record number. For example, the preceding candidate coordination action set contains three actions: Operator B 1810MHz power limiting by 2dB, Operator B 1810MHz carrier frequency migration, and Operator A 1800MHz power increase. After verification, the 2dB power limiting action passed; the carrier frequency migration action was rejected because the permission was not granted; the power increase action was rejected because it would amplify the impact on adjacent channels and exceeded the authorized parameter direction. Ultimately, an executable coordination action record was formed, which directly corresponds to the interference contribution record of the 1800MHz adjacent channel leakage anomaly in Area A. Based on the set of executable coordinated actions, an action selection range corresponding to the hierarchical interference situation data and interference contribution records is established to generate a permission-constrained coordinated action space. Specifically, after the set of executable coordinated actions is formed, the executable coordinated actions are bound to the hierarchical interference situation data and interference contribution records to establish the action selection range. The action selection range refers to the action boundary allowed by the state action screening model under the current abnormal interference state, which is jointly limited by the executable action object, the area of ​​action, the frequency band of action, the range of allowed parameters, and the corresponding interference contribution record. During implementation, firstly, the executable coordinated actions are associated with the source object and the disturbed object according to the interference contribution record number. Then, the spatial layer, frequency band layer, and affiliation layer on which the action can be applied are determined according to the shared area, frequency band resources, and operator affiliation in the hierarchical interference situation data. Finally, the action verification basis and verification results are written into the action space record. Taking Area A as an example, the permission-constrained coordinated action space record is "Area A—Operator B 1810MHz—Adjacent frequency band power limiting—Allowed 1dB to 3dB—Affected object of Operator A 1800MHz—Corresponding adjacent channel leakage interference contribution record". The subsequent status action filtering model can only select parameter coordination actions within this action space, and uses the records in this space as the basis for generating authorization records and execution scope. Compared with the candidate action set without permission constraints, this action space removes unauthorized and out-of-scope actions, while retaining the correspondence between interference contribution, configuration permissions, and action boundaries, thus providing a foundation for cross-carrier parameter coordination to be executable, traceable, and rollbackable.

[0030] The present invention is further configured such that the state-action filtering model processes hierarchical interference situation data and interference contribution records within the permission-constrained coordinated action space to generate parameter coordination actions and action execution ranges, including: The layered interference situation data is processed to organize its status, extracting the shared area, frequency band resources, operator affiliation, and interference type corresponding to abnormal interference states, and generating coordination status records. Specifically, in the shared indoor distribution network area of ​​Zone A on the B2 floor of a commercial complex, the indoor distribution network management platform receives the previously generated layered interference situation data. The data sources include frequency scanning records uploaded by on-site acquisition terminals, 5-minute operation observation records provided by the network management performance interface, and frequency band affiliation and operator affiliation records provided by the indoor distribution resource configuration library. When organizing the layered interference situation data, the completeness of the fields is first checked according to the shared area, frequency band resources, and operator affiliation, and records lacking area or frequency band identifiers are removed. Then, the abnormal interference states are time-aligned according to a unified 5-minute statistical window, and abnormal states under the same area, frequency band, and operator affiliation within consecutive windows are grouped into the same coordination status. The coordination status record is used to characterize the interference object that needs to be coordinated and processed. Its fields include shared area, frequency band resources, operator affiliation, interference type, duration window, and situation record reference. For example, if Operator A in Area A experiences adjacent channel leakage anomalies for two consecutive windows from 10:00 to 10:10, a coordinated status record of "Area A—1800MHz—Operator A—Adjacent Channel Leakage—Two Windows" is generated. This record organizes the situational data into a model-readable state object, providing a clear basis for subsequent action selection in terms of region, frequency band, and affiliation. The interference contribution records are organized by contributing objects, and the correspondence between source objects, affected objects, interventionable parameters, and candidate coordination actions is extracted to generate action association records. Specifically, the interference contribution records come from previous backtracking processing and already contain source objects, affected objects, interference types, interventionable parameters, and candidate coordination actions. When organizing the interference contribution records, it is first checked whether the source object exists in the indoor distributed antenna system (DAS) configuration data, and then it is checked whether the interventionable parameters belong to the open parameter items of that source object. Subsequently, the source object, affected object, interventionable parameters, and candidate coordination actions are bound one-to-one to generate action association records. Action association records are used to describe the interference contribution source and affected objects corresponding to a candidate action. The formation rules are as follows: records where the same source object affects the same affected object and has the same interventionable parameters are merged into one record; records where the same source object corresponds to different interference types are retained separately. Taking an implementation example, Operator A in Area A's 1800MHz band is affected by adjacent channel leakage from Operator B's 1810MHz band. The intervention parameter is downlink power limiting, and the candidate coordination action is adjacent band power limiting. This generates an action association record of "Operator B 1810MHz — Operator A 1800MHz — Adjacent Channel Leakage — Power Limiting". This record establishes a traceable correspondence between the candidate action and the interference contribution record, preventing action selection from deviating from the source and affected objects. The coordination status record and action association record are bound to the permission-constrained coordination action space to generate a set of optional actions corresponding to the abnormal interference state. Specifically, the permission-constrained coordination action space is formed by prior permission verification and records the area, frequency band, operator affiliation, and parameter range that each candidate coordination action is allowed to operate in. During implementation, the shared area, frequency band resources, and operator affiliation in the coordination status record are matched with the disturbed object in the action association record. After a successful match, the source object and interventionable parameters in the action association record are compared with the permission-constrained coordination action space, and actions within the authorized boundaries and adjustable parameter ranges are retained. The set of optional actions is the set of actions allowed to enter the model screening under the current abnormal interference state. Its fields include coordination status number, action object, action type, affected object, allowed parameter range, and contribution record reference. For example, the 1800MHz adjacent channel leakage anomaly in Area A corresponds to two action association records. Among them, "Operator B 1810MHz downlink power limiting 1dB to 3dB" is within the permission-constrained coordination action space, while the other "Operator B 1810MHz carrier frequency migration" has not been granted permission. Therefore, the set of optional actions only retains the power limiting action. Through this binding process, the actions entering the model have already undergone permission constraints and object constraints; The state-action filtering model, based on coordination state records, action association records, and a permission-constrained coordination action space, filters the set of selectable actions to generate parameter-coordinated actions corresponding to abnormal interference states. Specifically, this embodiment employs an interpretable state-action filtering model. The model's inputs are coordination state records, action association records, and a permission-constrained coordination action space. The model's filtering criteria include the degree of abnormality persistence, the degree of contribution matching, and the execution constraint margin. The degree of abnormality persistence is derived from the number of consecutive 5-minute windows of abnormal interference states, with two or more consecutive windows considered priority processing objects. The degree of contribution matching is derived from the completeness of the backtracking path between the source object and the disturbed object in the action association records, with actions having a clear mechanistic relationship and original observation references being prioritized. The execution constraint margin is derived from the adjustable range of parameters in the permission-constrained coordination action space, with actions whose parameters are in the middle of the allowable range and do not touch the boundary being prioritized. The model first eliminates actions that do not match the current coordination state, then prioritizes the remaining actions and outputs a parameter-coordinated action. Taking the example, an abnormal leakage in the 1800MHz adjacent channel of Area A lasted for two windows, originating from the 1810MHz adjacent frequency band of Operator B. Permissions allowed for downlink power limiting of 1dB to 3dB. The model reads records of similar actions in the same area from the historical retest feedback table and selects "downlink power limiting 2dB" as the parameter coordination action. The generation process of this action is jointly defined by three types of records: status, contribution, and permissions. The action selection process is interpretable. The action execution scope is generated based on the source object, the affected object, and the scope of effect in the permission constraint coordination action space associated with the parameter coordination action. Specifically, after generating the parameter coordination action, the source object, the affected object, and the scope of effect in the permission constraint coordination action space associated with the action are read to determine the action execution scope. The action execution scope includes the execution area, execution frequency band, execution object, point of effect, and parameter boundary, and its source is the indoor distributed antenna system (DAS) configuration data and the permission constraint coordination action space. Taking the implementation example, if the parameter coordination action is "Operator B 1810MHz downlink power limit 2dB", the source object is Operator B 1810MHz frequency band and ANT-03 to ANT-04 coverage points, and the affected object is Operator A 1800MHz frequency band in Area A, then the action execution scope is generated as "Area A—ANT-03 to ANT-04—Operator B 1810MHz—Downlink power limit 2dB—Affected object of Operator A 1800MHz". After the execution scope is generated, it is passed to the subsequent authorization record and retesting process. In the sample verification, there was an adjacent channel leakage anomaly in two consecutive windows of 1800MHz in area A before execution. After executing the parameter coordination action and completing the retest, the adjacent channel observation intensity of the target frequency band decreased, and the SINR recovered from 9dB to 14dB. The abnormal interference state did not continue to be triggered in the subsequent two 5-minute windows, indicating that the parameter coordination action and the action execution range can support the subsequent authorization, execution, retest and rollback judgment.

[0031] The present invention is further configured such that, in the step of generating an authorization record based on parameter coordination actions and action execution range, obtaining retest data after executing parameter coordination actions, updating hierarchical interference situation data based on the retest data, generating a rollback state based on the hierarchical interference situation data before and after the update, and writing the authorization record, parameter coordination actions, action execution range, retest data, and rollback state into the permission chain to generate an AI interference adaptive coordination result, the following steps are included: The execution fields of parameter coordination actions and their execution scope are parsed to extract the action object, effective area, frequency band resources, parameter adjustment items, and operator affiliation. Based on the indoor distributed antenna system (DAS) shared configuration data, the authorized subject and authorized boundaries are matched to generate the authorization record corresponding to the parameter coordination action. Specifically, in the indoor DAS co-construction and sharing area of ​​Zone A on the B2 floor of a commercial complex, the previous state action screening model has generated the parameter coordination action "Operator B 1810MHz adjacent frequency band downlink power limit 2dB", with the execution scope being "Area A ANT-03 to ANT-04 coverage points". When parsing the execution fields of this parameter coordination action and its execution scope, the action object, effective area, frequency band resources, parameter adjustment items, and operator affiliation are read. The action object comes from the source object in the interference contribution record, the effective area comes from the permission constraint coordination action space, and the parameter adjustment items come from the interventionable parameters. Subsequently, based on the permission configuration fields in the indoor DAS shared configuration data, the authorized subject and authorized boundaries are matched. The authorized subject can be the corresponding operator's network optimization management account or the authorized account of the shared indoor DAS management platform, and the authorized boundaries include the allowed adjustment area, frequency band, location, and parameter amplitude. If the action object belongs to Operator B, the area of ​​effect is within the authorized range of Area A, and the 2dB power limiting is within the allowable range of 1dB to 3dB, then an authorization record is generated. The authorization record retains the action number, authorizing subject, authorization time, authorization boundary, parameter adjustment items, and configuration record references, so that parameter adjustments have a clear source of authority and execution basis; Based on the authorization record, parameter coordination actions are converted into execution tasks, and the execution status of these tasks is registered to generate execution records for the parameter coordination actions. Specifically, after the authorization record is generated, the parameter coordination actions are converted into executable tasks that can be issued. The process of forming an execution task includes action format conversion, device object matching, and execution status registration. Action format conversion converts "downlink power limiting 2dB" into a parameter change task that can be recognized by the network management system or indoor distribution control platform; device object matching locates the remote unit, radio frequency unit, or frequency band configuration object corresponding to the coverage points ANT-03 to ANT-04 in Area A based on the action execution range; execution status registration records the status of the task as created, pending execution, in execution, completed execution, or failed execution. During implementation, before issuing the task, the consistency between the authorization record number and the device object number is verified. After verification, an execution task number is generated, and the execution task number, action object, parameter value before adjustment, parameter value after adjustment, execution time, and execution result are written into the execution record. Taking this embodiment as an example, the system reduces the downlink power of operator B's 1810MHz band by 2dB from the original configuration value, and records the execution start time 10:15, execution completion time 10:16 and task completion status, providing factual basis for subsequent retesting and rollback judgment; Within the action execution range, on-site retest data is acquired after execution. This data is then merged according to shared area and frequency band affiliation to generate retest data corresponding to the parameter coordination action. Specifically, after the task is completed, on-site retesting is performed within the action execution range. Retest data sources include frequency sweep records from the on-site acquisition terminal, operational observation records from the network management performance interface, and acquisition location information. The on-site acquisition terminal selects test points P03 and P04 within the ANT-03 to ANT-04 coverage area to perform retesting on the disturbed frequency band 1800MHz and the source frequency band 1810MHz. Each test point is continuously sampled three times, with a sampling interval of 10 seconds. The network management performance interface simultaneously acquires SINR, RSRQ, access failure rate, and load records within two 5-minute windows after execution. Before retesting, the collected data undergoes timestamp correction, duplicate record removal, and location consistency verification to ensure consistency between the retest data and the action execution range. Subsequently, data was merged according to shared area and frequency band affiliation. The records of the 1800MHz disturbed frequency band and the 1810MHz source frequency band of P03 and P04 within Area A were combined into retest data corresponding to the coordinated action for this parameter. This retest data retains the action number and execution record number, allowing direct tracing back to this parameter adjustment. Based on the retest data, the hierarchical interference situation data within the action execution range is updated to generate post-execution hierarchical interference situation data. Specifically, after the retest data is generated, the retest records are remapped to shared areas, frequency band resources, and operator affiliations according to the construction rules of the previous hierarchical interference situation data. In the specific processing, the shared area identifier in the retest data is first matched with the area layer node in the original hierarchical interference situation data; then, the retest frequency band is assigned to the corresponding frequency band layer; and finally, it is written to the affiliation layer based on operator affiliation. For intermodulation, adjacent channel leakage, or load congestion states that still occur after retesting, they are re-marked according to the same mechanism discrimination conditions; for states that do not meet the abnormal discrimination conditions after retesting, the corresponding abnormal state is updated to a resolved state. Taking this embodiment as an example, before execution, there was an adjacent channel leakage anomaly in the 1800MHz operator A frequency band in area A. After execution, the retest showed that the observation intensity of the adjacent frequency band decreased, and the SINR of the 1800MHz frequency band recovered from 9dB to 14dB. Moreover, the adjacent channel leakage discrimination condition was not met in the subsequent two 5-minute windows. Therefore, the abnormal status in the hierarchical interference situation data after execution was updated to be resolved, while the situation record and retest record before execution were retained. The system compares the pre- and post-execution layered interference situation data to generate a rollback status corresponding to the parameter coordination action. Specifically, the status comparison uses the pre- and post-execution layered interference situation data as input to compare the interference type, trigger window, and network quality changes within the same shared area, frequency band, and operator. The rollback status refers to the determination record of whether the parameter coordination action needs to be withdrawn or retained. Its generation rules can be set as follows: if the original abnormal state is not triggered again within two consecutive 5-minute windows after execution, and no new intermodulation interference, adjacent channel leakage, or load congestion states appear, then a "no rollback required" status is generated; if the original abnormal state continues to trigger after execution, or the quality of the disturbed object does not recover, or a new cross-frequency band abnormal state appears, then a "rollback recommended" status is generated; if the retest data is missing or the collection range is inconsistent with the action execution range, a "pending review" status is generated. Taking this embodiment as an example, after execution, the adjacent channel leakage state of Area A at 1800MHz did not continue to be triggered, the target frequency band quality recovered, and no new coverage quality anomalies appeared in the source frequency band of Operator B at 1810MHz. Therefore, a "no rollback required" state was generated. This state retains the comparison basis and retest record number, which can be used for auditing and accountability in the future. The authorization record, parameter coordination action, action execution scope, execution record, retest data, and rollback status are written into the permission chain in chronological order of execution time to generate the AI ​​interference adaptive coordination result. Specifically, the permission chain is an audit chain record that stores the authorization, execution, retest, and rollback status in chronological order of execution time. Each record contains a current record summary and a previous record summary. The record summary can be generated by a one-way digest algorithm using the authorization record number, action number, execution record number, retest data number, rollback status, and timestamp. During implementation, the authorization record is written first, followed by the parameter coordination action and action execution scope, and then the execution record, retest data, and rollback status. Each write checks whether the previous record summary is consistent; if the check passes, a new record is generated on the chain. As a sample verification, the system sequentially forms a chain record of "Authorization passed—Power limiting 2dB—Execution of ANT-03 to ANT-04 in Area A—Retest completed—No rollback required," ultimately generating the AI ​​interference adaptive coordination result. The results include interference type, source object, affected object, action performed, retest changes, and rollback status, which can fully reflect the processing process of a cross-operator indoor distribution interference coordination from authorization to retest closed loop.

[0032] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An AI interference adaptive coordination method for indoor distributed antenna system (DAS) co-construction and sharing, characterized in that, include: Acquire field interference observation data and indoor distribution sharing configuration data of the co-construction and sharing area of ​​indoor distribution system; and merge the field interference observation data according to the sharing area and frequency band affiliation based on the indoor distribution sharing configuration data to generate indoor distribution sharing observation data. Extract frequency band combination relationships, adjacent frequency coupling relationships, and load quality synchronization relationships from indoor distributed monitoring data; identify the interference states corresponding to intermodulation interference, adjacent frequency leakage, and load congestion; and map the interference states to the shared area, frequency band resources, and operator affiliation space to generate hierarchical interference situation data. Extract abnormal interference states from the hierarchical interference situation data, call the corresponding mechanism relationship according to the interference type of the abnormal interference state, and trace the abnormal interference state back to the interference contribution record between the source object, the disturbed object and the interventionable parameters. Based on the interference type and interventionable parameters in the interference contribution record, candidate coordination actions are matched, and permission boundaries and parameter ranges of the candidate coordination actions are verified based on the indoor distribution sharing configuration data to generate a permission-constrained coordination action space. A state-action filtering model is used to process hierarchical interference situation data and interference contribution records within the permission-constrained coordinated action space to generate parameters for coordinated actions and action execution ranges. Authorization records are generated based on parameter coordination actions and action execution ranges. After executing parameter coordination actions, retest data is obtained. Layered interference situation data is updated based on the retest data. Rollback status is generated based on the layered interference situation data before and after the update. Authorization records, parameter coordination actions, action execution ranges, retest data, and rollback status are written into the permission chain to generate AI interference adaptive coordination results.

2. The AI ​​interference adaptive coordination method for indoor distributed antenna system (DAS) co-construction and sharing as described in claim 1, characterized in that, Acquire field interference observation data and indoor distribution system (IDS) sharing configuration data for the shared indoor distribution system area. Based on the IDS sharing configuration data, merge the field interference observation data according to the sharing area and frequency band affiliation to generate indoor IDS sharing observation data, including: Field interference observation data is generated by acquiring field test records, operation observation records, and acquisition location information within the shared indoor distribution system area through field acquisition terminals. The indoor distributed antenna system (DAS) resource configuration library is used to read the indoor distributed antenna system (DAS) sharing configuration data corresponding to the indoor distributed antenna system co-construction and sharing area. The configuration fields of the indoor distributed antenna system (DAS) sharing configuration data are parsed to extract the shared area identifier, indoor distributed antenna system (DAS) location identifier, frequency band affiliation identifier, operator affiliation identifier and permission configuration fields, and a configuration association index is established. The source of the field interference observation data is verified and time window is marked. The field interference observation data with time window marking is divided into observation segments corresponding to the indoor distribution point location. Based on the configuration association index, the observation segments are bound to the corresponding shared area and frequency band affiliation to generate regional frequency band observation entries; Based on the operator's affiliation identifier, the regional frequency band observation entries are affixed and completed. For the regional frequency band observation entries with completed affiliation, duplicate records are removed and missing affiliation markers are removed to generate indoor distributed antenna system (DAS) shared observation data.

3. The AI ​​interference adaptive coordination method for indoor distributed antenna system (DAS) co-construction and sharing as described in claim 1, characterized in that, Extracting frequency band combination relationships, adjacent-channel coupling relationships, and load quality synchronization relationships from indoor distributed antenna system (DAS) shared observation data, and identifying interference states corresponding to intermodulation interference, adjacent-channel leakage, and load congestion, including: Frequency band field parsing is performed on indoor distributed antenna system (DAS) shared observation data to extract frequency band records that are in the same shared area or have indoor DAS link coupling. Frequency band combination relationships are established based on the source frequency band, target frequency band, and combining connection relationship in the frequency band records. Frequency adjacency analysis is performed on indoor distributed monitoring data to extract the spectral position relationship, coverage overlap relationship and adjacent frequency observation intensity between the target frequency band and adjacent frequency bands, and to generate adjacent frequency coupling relationship. Time window alignment is performed on indoor distributed monitoring data, and load change records and network quality change records under the same shared area and frequency band are extracted to establish load quality synchronization relationship. Intermodulation interference state is formed by matching intermodulation interference mechanism based on frequency band combination relationship; adjacent channel leakage state is formed by matching adjacent channel leakage mechanism based on adjacent channel coupling relationship; and load congestion state is formed by matching congestion interference mechanism based on load quality synchronization relationship.

4. The AI ​​interference adaptive coordination method for indoor distributed antenna system (DAS) co-construction and sharing as described in claim 3, characterized in that, The layered interference situation data is generated by mapping interference status to shared areas, frequency band resources, and operator affiliation spaces, including: The intermodulation interference status, adjacent channel leakage status and load congestion status are classified, and the shared area identifier, frequency band resource identifier and operator affiliation identifier corresponding to the interference status are extracted to generate an interference status index. Based on the interference status index, the intermodulation interference status, adjacent channel leakage status, and load congestion status are bound to the corresponding shared area, frequency band resource, and operator affiliation, respectively, generating association records between the area layer, frequency band layer, and affiliation layer; The associated records are merged for similar interference and cross-layer association to generate hierarchical interference situation data that includes the location of shared areas, frequency band resource ownership, operator ownership, and interference type.

5. The AI ​​interference adaptive coordination method for indoor distributed antenna system (DAS) co-construction and sharing as described in claim 1, characterized in that, The abnormal interference states extracted from the layered interference situation data include: The layered interference situation data is sliced ​​according to the shared area, frequency band resources and operator affiliation, and situation slice records corresponding to the shared area and frequency band resources are generated. Based on the type of interference in the situation slice record, the mechanism discrimination conditions corresponding to the frequency band combination relationship, adjacent frequency coupling relationship or load quality synchronization relationship are invoked to perform anomaly matching on the situation slice record; The situation slice records that meet the abnormal matching results are marked as abnormal interference states. The shared area identifier, frequency band resource identifier, operator affiliation identifier and interference type corresponding to the abnormal interference state are retained to generate an abnormal interference state set. The abnormal interference state set is merged within the same region and frequency band to generate abnormal interference state records for interference backtracking processing.

6. The AI ​​interference adaptive coordination method for indoor distributed antenna system (DAS) co-construction and sharing as described in claim 5, characterized in that, Based on the type of abnormal interference state, the corresponding mechanism relationship is invoked, and the abnormal interference state is traced back to the interference contribution record between the source object, the disturbed object, and the interventionizable parameters, including: Based on the interference type recorded in the abnormal interference status record, the corresponding mechanism relationship is selected from the frequency band combination relationship, adjacent frequency coupling relationship and load quality synchronization relationship to generate the backtracking path corresponding to the abnormal interference status. Based on the backtracking path and indoor distributed antenna system (DAS) configuration data, source objects are generated by filtering objects that have frequency band association, coverage association, or affiliation association with abnormal interference states. The shared area, frequency band resources, and operator affiliation of the abnormal interference state are used as the affiliation information of the disturbed object to generate a disturbed object corresponding to the source object. Based on the interference type of the abnormal interference state and the configuration attributes of the source object, adjustable frequency band parameters, power parameters, or time domain parameters are matched to generate interventionable parameters; The interference contribution record is generated by binding the source object, the object being disturbed, the type of interference, and the parameters that can be intervened.

7. The AI ​​interference adaptive coordination method for indoor distributed antenna system (DAS) co-construction and sharing as described in claim 6, characterized in that, Based on the interference type in the interference contribution record and the interventionable parameters, candidate coordination actions are matched, including: The interference contribution record is parsed to extract the interference type, source object, disturbed object and interventionable parameters to generate an action matching index; Based on the action matching index, the correspondence between the interference type and the parameter action is invoked, and the interventionable parameters are matched with the corresponding coordinated action type to generate the initial coordinated action; Bind the source object and the disturbed object to the initial coordination action to generate a candidate coordination action that includes the action object, action type and target object; Candidate coordination actions are merged and mutually exclusive based on their origin, and the set of candidate coordination actions corresponding to the interference contribution record is retained.

8. The AI ​​interference adaptive coordination method for indoor distributed antenna system (DAS) co-construction and sharing as described in claim 7, characterized in that, Based on the indoor distributed antenna system (DAS) configuration data, the candidate coordination actions are validated for permission boundaries and parameter ranges to generate a permission-constrained coordination action space, including: Extract operator affiliation information, authorization boundary information, and adjustable parameter range corresponding to candidate coordination actions from indoor distributed antenna system (DAS) configuration data to generate action verification basis; Based on the action verification criteria, the action objects of the candidate coordinated actions are verified for attribution, the objects affected by the candidate coordinated actions are verified for boundaries, and the parameter values ​​of the candidate coordinated actions are verified for range to generate action verification results. The action verification results are matched with the candidate coordination action set, and candidate coordination actions that cross the authorization boundary or exceed the adjustable range of parameters are eliminated to generate an executable coordination action set. Based on the set of executable coordinated actions, an action selection range corresponding to the hierarchical interference situation data and interference contribution records is established, generating a permission-constrained coordinated action space.

9. The AI ​​interference adaptive coordination method for indoor distributed antenna system (DAS) co-construction and sharing as described in claim 8, characterized in that, The state-action filtering model is used to process hierarchical interference situational data and interference contribution records within the permission-constrained coordinated action space to generate parameters. The coordinated actions and the scope of action execution include: The hierarchical interference situation data is sorted out to extract the shared area, frequency band resources, operator affiliation and interference type corresponding to the abnormal interference status, and a coordinated status record is generated. Organize the contributing objects of the interference contribution records, extract the correspondence between the source object, the disturbed object, the interventionable parameters and the candidate coordinated actions, and generate action association records; Bind the coordination state record and action association record to the permission constraint coordination action space to generate a set of optional actions corresponding to the abnormal interference state; The state-action filtering model coordinates the action space based on the coordination state record, action association record, and permission constraints, and generates parameter coordination actions corresponding to abnormal interference states by filtering the set of optional actions. The action execution scope is generated by coordinating the scope of action in the action space based on the source object, the disturbed object, and the permission constraints associated with the coordinated action.

10. The AI ​​interference adaptive coordination method for indoor distributed antenna system (DAS) co-construction and sharing as described in claim 9, characterized in that, Authorization records are generated based on parameter coordination actions and action execution ranges. After executing parameter coordination actions, retest data is obtained. Layered interference situation data is updated based on the retest data. A rollback status is generated based on the layered interference situation data before and after the update. The authorization records, parameter coordination actions, action execution ranges, retest data, and rollback status are written into the permission chain to generate AI interference adaptive coordination results, including: The execution field of parameter coordination actions and action execution scope is parsed to extract the action object, action area, frequency band resources, parameter adjustment items and operator affiliation. Based on the indoor distributed antenna system (DAS) configuration data, the authorized subject and authorized boundary are matched to generate the authorized record corresponding to the parameter coordination action. Based on the authorization record, the parameter coordination action is converted into an execution task, and the execution status of the execution task is registered to generate the execution record of the parameter coordination action; Within the scope of action execution, acquire on-site retest data after execution, and merge the on-site retest data according to the shared area and frequency band affiliation to generate retest data corresponding to the parameter coordinated action; Based on the retest data, the hierarchical interference situation data within the action execution range is updated to generate post-execution hierarchical interference situation data; The state comparison between the layered interference situation data before and after execution generates a rollback state corresponding to the parameter coordination action. Authorization records, parameter coordination actions, action execution scope, execution records, retest data, and rollback status are written into the permission chain in the order of execution time to generate AI interference adaptive coordination results.

Images