A radio frequency performance analysis and automatic decision method and system based on true value fusion

Abstract

The application provides a radio frequency performance analysis and automatic decision method and system based on true value fusion, and relates to the technical field of global navigation satellite system testing. True value information of a test scene including configuration true value and execution true value is obtained, the execution true value including channel calibration state; radio frequency data of at least an output end of a measured module is collected, target signal components and residual interference components are extracted and quality marks are generated; based on joint association keys, the true value information is aligned and fused with radio frequency analysis results, input validity check and confidence evaluation are performed; when alignment is successful, radio frequency performance indexes are obtained based on the fused data, hierarchical automatic decision and abnormal attribution are performed according to externally configured decision rules, and a decision result containing an evidence chain is output. The problems of inaccurate evaluation, low efficiency and difficult tracing caused by the fact that true values and analysis results cannot be automatically associated in the prior art are solved.

Description

Technical Field

[0001] This application relates to the field of global navigation satellite system testing technology, and in particular to a radio frequency performance analysis and automatic decision-making method and system based on truth fusion. Background Technology

[0002] In the RF verification process of GNSS array anti-jamming modules, testers typically need to construct target signals, interference signals, spatial direction of arrival, and dynamically changing scenarios, and collect and analyze the RF signals from the output of the module under test. Common analysis methods include observing the output spectrum, statistically analyzing output power, analyzing residual interference, comparing carrier-to-noise ratio changes, or connecting the module output to a GNSS receiver and judging the anti-jamming effect through the receiver's positioning, tracking, and timing results. Under laboratory conditions, testers have access to true-value information about the scenario, such as the target signal power, interference signal power, interference type, spatial direction of arrival, channel output status, and test time series set by the simulated source. However, this information often exists only as a manual configuration record and is not automatically correlated with the RF acquisition and analysis results. The analysis equipment can only acquire the output spectrum or output waveform, but cannot automatically identify the specific target signal, interference source, spatial direction of arrival status, channel parameters, and test stage corresponding to the current moment.

[0003] This situation leads to several technical shortcomings: First, analyzing only the output spectrum is insufficient to determine whether the target signal has been effectively preserved, because the lack of the true value of the target signal input and the reference power makes it impossible to assess whether the target signal has experienced excessive attenuation, amplitude fluctuations, or frequency response impairment. Second, analyzing only residual interference is insufficient to determine whether interference suppression is adequate, because the magnitude of residual interference at the output end must be correlated with the input interference power, interference type, interference bandwidth, and interference direction to accurately determine the interference suppression ratio and spatial suppression effect. Third, the lack of spatial direction truth correlation makes it difficult to verify whether the output results of the array anti-interference module are consistent with the preset spatial scenario, while the core capabilities of the module are closely related to the spatial direction. Fourth, the lack of channels... The correlation between state and truth values ​​makes it impossible to distinguish between problems with the module under test and abnormalities in the input channel, because the amplitude, phase, delay, and calibration status of each channel in a multi-channel RF environment simulation will affect the spatial equivalent accuracy. Fifth, manual interpretation is inefficient and inconsistent. Testers need to manually integrate simulation source settings, test scripts, spectrum analysis results, output records, and receiver status for comprehensive judgment. Different people may reach inconsistent conclusions, which is not conducive to batch testing and automated regression testing. Sixth, test results are difficult to trace. Without a truth fusion mechanism, test results usually only record output data and judgment conclusions, and it is impossible to trace back to the specific target power, interference power, incoming parameters, channel status, and test events corresponding to the conclusions.

[0004] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0005] This application provides a method and system for radio frequency performance analysis and automatic decision-making based on truth value fusion. It has the advantages of realizing the automatic correlation between truth information and radio frequency analysis results, improving the accuracy and automation of radio frequency performance analysis, and facilitating the traceability of test results.

[0006] Firstly, the radio frequency performance analysis and automatic decision-making method based on truth value fusion provided in this application adopts the following technical solution: A method for radio frequency performance analysis and automatic decision-making based on truth fusion, comprising: Obtain test scenario truth information, which includes target signal truth value, interference signal truth value, spatial direction of arrival truth value, and channel status truth value; the truth information includes configuration truth value and execution truth value, the configuration truth value is the planned setting value of the test scenario, the execution truth value is the actual execution feedback value of the test equipment, and the execution truth value includes channel calibration status; Collect radio frequency (RF) data from the module under test (DUT), including at least RF or intermediate frequency (IF) data from the output of the DUT; analyze the RF data to extract the target signal component and residual interference component, and generate a quality indicator characterizing the reliability of each component extraction. Based on the joint association key composed of event identifier, time window, frequency object and channel mapping, the truth information is aligned and fused with the radio frequency analysis results, and the fusion result is subjected to input validity verification and confidence evaluation; when the alignment fails or the confidence is lower than the preset confidence threshold, the output data is invalid or uncertain and does not enter the automatic decision. Based on the successful alignment and fusion, radio frequency performance indicators are obtained from the fused data. These radio frequency performance indicators are obtained through structured indicators input from external consumers or by calling an independent indicator calculation module. The radio frequency performance indicators are automatically judged in a hierarchical manner according to the externally configured judgment rules, and the judgment results and judgment evidence chain records are output. The judgment results include pass, fail, and uncertain.

[0007] Optionally, in the joint association key, the event identifier includes a test event number and an interference source identifier, the time window has a preset window tolerance, the frequency object includes a frequency point, bandwidth, signal type, and spatial orientation label or known waveform label, and the channel mapping includes an array element number, an RF output channel, a module input channel, and a channel calibration status.

[0008] Optionally, when the deviation between the execution truth value and the configuration truth value exceeds a preset deviation tolerance, the execution truth value is used as the standard for alignment and fusion, and a truth value replacement event is marked; when there is a measured input truth value at the input end of the module under test, the execution truth value is verified using the measured input truth value.

[0009] Optionally, the alignment and fusion processing rules include: matching records within the time window tolerance according to the nearest time principle; matching frequency objects according to frequency band overlap; detecting channel substitution in channel mapping and outputting substitution alarms; distinguishing duplicate events according to event identifiers; marking missing records according to the missing type and reducing the confidence of the corresponding fusion result; and outputting invalid or uncertain data when the above processing fails.

[0010] Optionally, the time alignment includes: aligning by timestamp, event number, scenario stage, or sampling window so that each segment of radio frequency data corresponds to a specific test event and true value state.

[0011] Optionally, the channel alignment includes: associating the RF analysis results of each channel with the corresponding array element number, RF output channel, module input channel and channel calibration status to establish a traceable channel mapping relationship.

[0012] Optionally, the association of the frequency objects includes: associating the target signal component, the interference signal component, and the residual interference component with their corresponding truth objects in terms of frequency point, bandwidth, and signal type, and combining the interference source identifier, event identifier, spatial orientation label or known waveform label to distinguish multiple interference sources at the same frequency; when there is one component corresponding to multiple truth objects or multiple components corresponding to one truth object, the components are assigned according to a preset one-to-many or many-to-one matching rule.

[0013] Optionally, the radio frequency performance indicators are obtained through structured indicators input from external consumers or by calling an independent indicator calculation module; the radio frequency performance indicators include, for example, one or more of the following: target signal retention, interference rejection ratio, signal-to-interference ratio improvement, noise floor rise, frequency response impairment, and dynamic response indicators, the definition and calculation methods of which are not limited to this method.

[0014] Optionally, the hierarchical automatic decision is performed according to the levels of input validity, target protection, interference suppression, signal-to-interference ratio improvement, side effect control, and dynamic response, forming an interpretable decision conclusion; the threshold system and indicator hierarchical sources for each level are provided by external configuration; the decision conclusion is only for radio frequency domain performance, and whether to admit the global navigation satellite system closed-loop verification is determined by the external hierarchical verification process.

[0015] The method further includes: recording the truth information and its configuration version on which the judgment is based, the collected data or its storage reference, the results of radio frequency performance indicators, the judgment rule identifiers and judgment rule versions of each level of judgment, and the rule hit records and confidence levels of abnormal attribution, forming a traceable judgment evidence chain.

[0016] Optionally, when the judgment result is abnormal, the abnormality is attributed according to the deviation relationship between the true value and the radio frequency analysis result, and a predefined attribution rule table is used. The attribution rule table defines necessary conditions and exclusion conditions for each risk item. The risk items include target damage, interference residue, channel abnormality, incoming mismatch, noise floor increase, and dynamic convergence abnormality. When multiple risk items occur concurrently, they are sorted and output according to priority or confidence level. When there is evidence conflict, the confidence level of the attribution conclusion is reduced and the conflict event is recorded. When the confidence level is insufficient, an uncertain attribution result is output.

[0017] Secondly, this application provides a radio frequency performance analysis and automatic decision-making system based on truth value fusion, comprising: The information acquisition module is used to acquire test scenario truth information, which includes configuration truth and execution truth, and the execution truth includes channel calibration status. The data acquisition module is used to acquire radio frequency (RF) data from the module under test (DUT), including at least RF or intermediate frequency (IF) data from the output of the DUT; the RF data is analyzed to extract target signal components and residual interference components, and a quality indicator characterizing the reliability of each component extraction is generated. The fusion module is used to align and fuse the truth information with the radio frequency analysis results based on a joint association key composed of event identifiers, time windows, frequency objects, and channel mappings, and to perform input validity verification and confidence evaluation on the fusion results; when alignment fails or the confidence level is lower than a preset confidence threshold, the output data is invalid or in an uncertain state. The calculation module is used to obtain radio frequency performance indicators based on the fused data after successful alignment and fusion. The output module is used to perform hierarchical automatic judgment on the radio frequency performance indicators according to the externally configured judgment rules, and output the judgment result and the judgment evidence chain record. The judgment result includes pass, fail, and uncertain.

[0018] In summary, this application obtains test scenario truth information containing configuration truth values ​​and execution truth values, aligns and fuses the truth values ​​with RF analysis results based on a joint association key composed of event identifiers, time windows, frequency objects, and channel mappings, and obtains RF performance indicators based on the fused data after input validity gating, implements hierarchical automatic decision-making and anomaly attribution, and outputs a decision result containing a chain of evidence. When alignment fails or confidence is insufficient, the output data is invalid or indeterminate rather than a forced decision. This solves the problems of inaccurate evaluation, low efficiency, and difficulty in traceability caused by the inability to automatically associate truth values ​​with analysis results in the prior art, and has the advantages of improving test accuracy and efficiency and achieving result traceability. Attached Figure Description

[0019] Figure 1This is a flowchart illustrating the first embodiment of the radio frequency performance analysis and automatic decision-making method based on truth value fusion in this application; Figure 2 This is a structural block diagram of the first embodiment of the radio frequency performance analysis and automatic decision system based on truth value fusion of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0021] In traditional RF verification of existing GNSS array anti-jamming modules, there is a lack of automatic correlation between the true information of the test scenario (such as target signal, interference signal, spatial direction of arrival, and channel status) and the RF acquisition and analysis results. As a result, it is difficult to accurately determine the retention of the target signal and the effect of interference suppression by analyzing only the output spectrum or residual interference. Furthermore, the lack of correlation between the true values ​​of spatial direction of arrival and channel status makes manual interpretation inefficient, inconsistent, and difficult to trace test results.

[0022] To address this, embodiments of this application provide a method for radio frequency performance analysis and automatic decision-making based on truth fusion, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the radio frequency performance analysis and automatic decision-making method based on truth fusion of this application.

[0023] In this embodiment, the radio frequency performance analysis and automatic decision-making method based on truth value fusion includes the following steps: Step S10: Obtain test scenario truth information. Truth information includes target signal truth value, interference signal truth value, spatial direction of arrival truth value, and channel status truth value; truth information includes configuration truth value and execution truth value. Configuration truth value is the planned setting value of the test scenario, and execution truth value is the actual execution feedback value of the test equipment. Execution truth value includes channel calibration status.

[0024] For ease of understanding, the following explains some key terms in this embodiment: Test scenario truth information refers to the actual state data of test events recorded in the RF test environment, including two levels: configuration truth and execution truth. Configuration truth represents the planned settings of the test scenario, such as the planned power, frequency, and modulation scheme of the target signal, the planned power, frequency, type, and bandwidth of the interference signal, as well as the planned parameters for spatial direction of arrival and channel status. Execution truth represents the actual feedback values ​​executed by the test equipment, such as the readback value of the actual output power of the signal source, the actual status feedback of channel switching, and the channel calibration status. It may also include the measured input truth values ​​of the input terminals of the module under test. This truth information provides a benchmark for RF analysis.

[0025] Radio frequency (RF) data refers to raw electromagnetic wave signal data acquired from the input or output of the module under test (DUT) using RF acquisition equipment. This data is typically stored digitally and includes time-domain and frequency-domain information such as signal amplitude, phase, and frequency.

[0026] The target signal component refers to the portion of the signal that matches the characteristics of the expected target signal, separated from the acquired radio frequency data through signal processing techniques. This component reflects the processing and retention of the target signal by the module under test.

[0027] Residual interference component refers to the portion of interference signal that remains after processing by the module under test (DUT) from the acquired radio frequency data using signal processing techniques. This component reflects the DUT's effectiveness in suppressing interference signals.

[0028] Alignment fusion refers to establishing a precise correspondence between test scenario ground truth information and RF analysis results based on a joint association key composed of event identifiers, time windows, frequency objects, and channel mappings. This process ensures that each segment of RF data analysis result can be accurately associated with its corresponding test scenario ground truth state; output data is invalid or uncertain when alignment fails or confidence is insufficient.

[0029] Radio frequency (RF) performance metrics are parameters used to quantitatively evaluate the RF performance of a module under test, calculated based on aligned and fused data. These metrics are used to measure the module's performance in areas such as target signal processing and interference suppression.

[0030] The decision rules refer to the rules, thresholds, and level ranges provided by an external configuration for evaluating radio frequency performance indicators. The configuration version is saved along with the decision records.

[0031] The judgment result refers to the conclusion output by the system after automatically judging the radio frequency performance indicators in a hierarchical manner according to the externally configured judgment rules. The conclusion includes pass, fail, and uncertain, and is accompanied by a record of the judgment evidence chain.

[0032] Step S20: Collect radio frequency (RF) data from the module under test (DUT). The RF data includes at least the RF or intermediate frequency (IF) data from the output of the DUT, and may also include the measured data from the input of the DUT. Analyze the RF data, extract the target signal component and residual interference component, and generate a quality flag characterizing the reliability of each component extraction. The quality flag is used for subsequent input validity verification.

[0033] Step S30: Based on the joint association key consisting of event identifier, time window, frequency object and channel mapping, the truth information is aligned and fused with the radio frequency analysis results, and the fusion result is checked for input validity and evaluated for confidence. When the alignment fails or the confidence is lower than the preset confidence threshold, the output data is invalid or uncertain and does not enter the automatic decision.

[0034] In practice, the alignment and fusion includes time alignment, channel alignment, and frequency alignment.

[0035] Specifically, time alignment aims to ensure that each data point or segment in the RF data stream is accurately associated with the corresponding truth information at that moment. This is particularly critical for dynamically changing test scenarios, such as when the target signal, interference signal, or channel state changes over time; the RF analysis results must be synchronized with the truth in the time dimension. Time alignment can be achieved by synchronizing clock signals, recording a uniform timestamp during data acquisition, or matching using sequence numbers or event markers in data packets. For example, during data acquisition, a high-precision timestamp is appended to each RF data frame, while the truth information also carries its generation or activation timestamp. Accurate time correspondence is achieved by comparing and matching these timestamps.

[0036] By introducing three refined alignment mechanisms—time alignment, channel alignment, and frequency alignment—this application effectively solves the matching problem between truth information and RF analysis results across different dimensions. Time alignment ensures the time synchronization of data in dynamic scenarios, avoiding analysis errors caused by time misalignment; channel alignment establishes an accurate mapping between RF data and physical channels, guaranteeing the accuracy of multi-channel system analysis; and frequency alignment enables the system to accurately identify and associate specific signal components in the spectrum. These alignment measures work together to greatly improve the accuracy and reliability of fusing truth and RF analysis results, thus providing a solid foundation for the accurate calculation of subsequent RF performance indicators and making the automatic judgment results based on these indicators more credible and interpretable, effectively avoiding misjudgments caused by data mismatch.

[0037] However, in practice, ensuring that the acquired RF data accurately corresponds to the truth information of complex test scenarios in the time dimension is a key challenge for achieving effective fusion. Inaccurate time alignment may lead to RF data being associated with irrelevant truth states, thereby affecting the accuracy of subsequent RF performance index calculations and ultimately the reliability of automatic decision results.

[0038] To address this, this embodiment further proposes that the time alignment includes: alignment using timestamps, event numbers, scenario stages, or sampling windows, so that each segment of RF data corresponds to a specific test event and true value state. Specifically, a timestamp is a mechanism for accurately recording the moment an event occurs. During implementation, a high-precision timestamp can be appended to each piece of acquired RF data and each recorded true value information (including the true value of the target signal, the true value of the interference signal, the true value of the spatial orientation, and the true value of the channel state). These timestamps are typically generated by a strictly synchronized clock source (e.g., a system clock synchronized via GPS or NTP protocol), ensuring that the time base of all data sources is consistent. By comparing and matching these timestamps, the system can accurately associate RF data at a specific moment with the true value state at the same moment, thereby achieving time synchronization alignment.

[0039] Furthermore, event numbers can be used to align discrete test events with clearly defined start and end points. During test scenario design, unique event numbers can be pre-defined for critical test events or system state changes (e.g., the appearance of a target signal, switching of interference signals, change of the operating mode of the module under test). When these events occur, the RF data acquisition system and the truth recording system synchronously record the corresponding event numbers. During the data processing phase, the system associates RF data acquired during or after a specific event with the corresponding truth state by matching the event numbers. This approach is particularly suitable for handling discontinuous or phased test processes.

[0040] Furthermore, scenario phases provide a macro-level time alignment method. A complex test scenario can be divided into multiple logically independent phases, each corresponding to a specific set of truth configurations and test conditions. For example, a "target search phase," "interference suppression phase," or "multi-target tracking phase" can be defined. During data acquisition and truth recording, the system identifies the current scenario phase. By associating RF data with its corresponding scenario phase, it ensures that the data acquired within a specific test phase matches the preset truth state of that phase. This method is suitable for holistic analysis of long-term, variable scenarios.

[0041] In addition, sampling windows are a strategy for periodic or on-demand data alignment. The system can preset fixed sampling time windows (e.g., acquiring a 10-millisecond data window every 100 milliseconds) or trigger sampling windows based on specific conditions. Within each sampling window, radio frequency (RF) data is acquired simultaneously, and the corresponding truth information is recorded. By matching the RF data with the truth information within the same sampling window, precise local time alignment can be achieved. This method is suitable for applications requiring periodic or batch data processing, ensuring a high degree of consistency between data and truth values ​​within each analysis cycle.

[0042] By employing the aforementioned technical solutions and utilizing various refined alignment mechanisms such as timestamps, event numbers, scene stages, or sampling windows, a precise and reliable correspondence can be established between each segment of acquired RF data and specific test events and true states. This precise time alignment significantly improves the accuracy and effectiveness of fusing true information with RF analysis results, thereby enabling RF performance indicators calculated based on the fused data to more accurately reflect the actual performance of the module under test. Therefore, automatic judgments based on these high-precision indicators will be more reliable and convincing, effectively avoiding misjudgments caused by time misalignment, and providing a solid data foundation for RF performance evaluation and fault attribution.

[0043] In some of the embodiments described above, the truth information is directly derived from the setting record of the analog source. However, the setting value of the analog source is not necessarily equal to the actual input signal received by the module under test. Link loss, channel status, and execution deviation may all cause the setting to deviate from the actual value.

[0044] In this embodiment, the truth information includes configuration truth and execution truth. The configuration truth is the planned setting value of the test scenario, and the execution truth is the actual execution receipt value of the test device. The execution truth includes the channel calibration status. When the deviation between the execution truth and the configuration truth exceeds a preset deviation tolerance, the execution truth is used as the standard for alignment and fusion, and a truth replacement event is marked. When there is a measured input truth at the input end of the module under test, the execution truth is verified using the measured input truth.

[0045] Specifically, the configuration truth value comes from the planned parameters during the test scenario design phase; the execution truth value comes from the actual effective parameters fed back by the test equipment during execution, such as the readback value of the actual output power of the signal source, the actual status feedback of channel switching, and the effective status and calibration version of channel calibration; the measured input truth value comes from the measured sampling at the input end of the module under test, such as measuring the actual power and spectrum of the input signal through the coupling port. The three layers of truth values ​​are verified step by step according to priority: with the configuration truth value as the benchmark, when the deviation between the execution truth value and the configuration truth value exceeds the deviation tolerance, the execution truth value is used and the truth value replacement event is recorded; the measured input truth value is used to verify the reliability of the execution truth value.

[0046] Through the above technical solution, the true value information is no longer equivalent to the simulated source setting value, but distinguishes three levels: planning, execution and actual measurement, and incorporates channel calibration status and deviation tolerance verification, so that the benchmark on which the fusion is based is closer to the actual input of the module under test, thereby improving the credibility of the indicators and judgment conclusions.

[0047] Specifically, the element number is a unique identifier for each individual radiating element in the antenna array. In practice, the test system assigns a unique numeric or alphanumeric identifier to each element. When acquiring RF data, the corresponding element number is recorded simultaneously. After RF analysis, the analysis results (e.g., signal strength, phase, or noise level of a specific element) are linked to that element number. This association can be achieved by adding specific fields to the data structure, using association tables in a database, or through metadata tags, ensuring that each analysis result can be clearly traced back to its physical origin.

[0048] An RF output channel refers to the physical path or port from which a radio frequency (RF) signal is output from the module under test (DUT) or test equipment. Correlating RF analysis results with RF output channels means that the analysis results can clearly identify which specific RF output path's performance is reflected. For example, in a multi-channel RF system, each RF output channel (such as Tx1, Tx2, etc.) has its unique RF characteristics, and correlation can distinguish the performance of different transmission paths. This is typically achieved by recording and transmitting unique identifiers for the RF output channels in the test configuration.

[0049] A module input channel refers to the physical path or port through which an RF signal enters the module under test (DUT). Correlating RF analysis results with module input channels helps trace the initial state of the signal before it is processed within the module. For example, when the DUT has multiple receive input ports, correlation can identify which input channel's signal caused a specific RF analysis result. This typically requires the test system to synchronously record and transmit the module input channel identification information during signal routing or switching.

[0050] Channel calibration status refers to whether an RF channel has been calibrated before or during testing, along with the specific parameters and results of the calibration. Correlating RF analysis results with channel calibration status allows for the assessment of the impact of calibration on test results and the determination of the validity and accuracy of those results. Calibration status can be a simple Boolean value (e.g., "calibrated" or "not calibrated"), a calibration date and timestamp, a set of calibration parameters (such as gain compensation values, phase compensation values), or a reference to a detailed calibration report. This correlation ensures that the channel calibration status is fully considered when analyzing RF performance, avoiding misjudgments due to lack of calibration or improper calibration.

[0051] Establishing a traceable channel mapping relationship is the comprehensive goal of the above-mentioned associations. This means systematically associating RF analysis results with multi-dimensional information such as array element number, RF output channel, module input channel, and channel calibration status to form a clear and complete mapping chain. Any RF analysis result can be traced back to its specific physical source (which array element, which RF output / input channel) and its calibration status through this mapping relationship. This mapping relationship is usually implemented by building a unified data model, database structure, or metadata management system, ensuring the integrity, interpretability, and auditability of the test data.

[0052] Through the above technical solution, this embodiment effectively addresses the challenges of unclear channel information and difficulty in tracing the source of problems in RF performance analysis. When RF performance indicators are abnormal, this solution can accurately pinpoint the source of the anomaly, such as whether it is a problem with a specific array element, RF output channel, module input channel, or related to channel calibration status. This greatly improves the efficiency and accuracy of fault diagnosis, making the RF performance evaluation of the module under test more refined and reliable. Furthermore, this clear mapping relationship provides a solid data foundation for subsequent performance optimization and anomaly attribution, ensuring the integrity and interpretability of the test results.

[0053] The association of frequency objects includes: associating the target signal component, interference signal component, and residual interference component with their corresponding ground truth objects in terms of frequency, bandwidth, and signal type, and distinguishing multiple interference sources at the same frequency by combining interference source identifiers, event identifiers, spatial orientation tags, or known waveform tags; when one component corresponds to multiple ground truth objects or multiple components correspond to one ground truth object, they are assigned according to a preset one-to-many or many-to-one matching rule; the radio frequency performance indicators include, for example, one or more of the following: target signal retention, interference suppression ratio, signal-to-interference ratio improvement, noise floor rise, frequency response impairment, and dynamic response indicators, the definition and calculation method of which are not limited to the scope of this method.

[0054] Specifically, when multiple interference sources operate at the same or similar frequencies, frequency, bandwidth, and signal type alone are insufficient to distinguish which interference source corresponds to the spectral remnant. Therefore, joint identification is achieved using interference source identifiers, event identifiers, spatial direction tags, or known waveform tags: known waveform tags determine the source by correlating components with pre-stored interference waveform templates; spatial direction tags compare the estimated direction of a component with the true direction. One-to-many matching rules are used when a component may correspond to multiple true objects, such as sorting by waveform correlation and recording the highest candidate; many-to-one matching rules are used when multiple components correspond to the same true object, such as aggregating multiple spectral components of the same interference source and assigning them uniformly. RF performance indicators are defined and calculated by an independent indicator evaluation system. This method uses the fused data to call its calculation results or directly consumes externally input structured indicators without repeatedly limiting their formulas and definitions. Through the above joint identification and matching rules, the system can accurately identify the source and attributes of each frequency component, avoiding indicator calculation deviations caused by unclear signal attribution.

[0055] In some of the embodiments described above, the aligned and fused data is used for indicator acquisition and decision-making. However, if a decision is still forced when the truth record is missing, the time alignment exceeds the tolerance, or there is a permutation in the channel mapping, unreliable decision conclusions will be generated.

[0056] To address this, this embodiment further proposes that the hierarchical automatic decision-making is conducted according to the levels of input validity, target protection, interference suppression, signal-to-interference ratio (SIR) improvement, side effect control, and dynamic response, forming interpretable decision conclusions. The threshold system and index sources for each level are provided by external configuration. The decision conclusions only apply to radio frequency (RF) domain performance, and whether or not the system is admitted to the closed-loop verification of the Global Navigation Satellite System (GNSS) is determined by the external hierarchical verification process. Specifically, the hierarchical automatic decision-making mechanism decomposes the complex RF performance evaluation into multiple logically clear and interrelated sub-decision levels: the input validity level verifies the availability of fused data; the target protection level evaluates the degree of target signal retention; the interference suppression level evaluates the interference suppression effect; the SIR improvement level quantifies the improvement of the signal-to-interference ratio; the side effect control level detects negative effects such as increased noise floor; and the dynamic response level evaluates the response performance under dynamic scenarios. The thresholds, level ranges, and rule sets upon which each level of decision is based are loaded as external configurations, and the configuration version is saved along with the decision record. The decision conclusion only applies to the performance in the radio frequency domain and does not involve the admission decision for entering the closed-loop verification of the Global Navigation Satellite System. This admission decision is made by the external hierarchical verification process based on the decision results and structured data output by this method.

[0057] Simultaneously, the method also includes: recording the truth information and its configuration version on which the judgment is based, the acquired data or its storage reference, the RF performance index results, the judgment rule identifiers and versions of each level of judgment, and the rule hit records and confidence levels of anomaly attribution, forming a traceable chain of judgment evidence. Specifically, to ensure the integrity, reproducibility, and traceability of the test results, the system will record in detail all key information relied upon by each judgment, including the truth information and its configuration version of the test scenario, the original acquired data or its storage reference, the values ​​and units of each RF performance index, the specific rule identifiers of each level of judgment, the versions of the judgment rules, and the rule hit records and confidence levels of anomaly attribution. All records are associated with a unique test identifier, forming a complete traceable chain.

[0058] Through the above technical solution, this embodiment introduces a hierarchical automatic decision-making mechanism, refining the complex RF performance evaluation into multiple manageable levels. This ensures that the decision conclusion is no longer a simple pass or fail, but clearly identifies the specific link and cause of the performance anomaly. Simultaneously, the threshold system and hierarchical indicator sources are loaded as external configurations, decoupling the decision logic from the indicator evaluation system and hierarchical verification process. Furthermore, by comprehensively recording the decision evidence chain, engineers can quickly trace back to the original test conditions and data, avoiding blind troubleshooting and significantly improving the efficiency and accuracy of fault diagnosis. This interpretable and traceable decision-making mechanism not only enhances the depth and breadth of RF performance analysis but also strengthens the reliability and persuasiveness of test results.

[0059] Step S40: Based on successful alignment and fusion, obtain RF performance indicators from the fused data. RF performance indicators are obtained through structured indicators input from external consumers or by calling an independent indicator calculation module. The specific definition and calculation criteria of the indicators are maintained by an independent indicator evaluation system.

[0060] In practical implementation, after establishing a correspondence between truth information and radio frequency analysis results, the fused data is provided to an independent indicator calculation module, or the structured indicator results from external input are directly read. The specific calculation criteria for the indicators are maintained by an independent indicator evaluation system, and this embodiment does not limit its calculation details. For example, the fused data can be sent to the indicator calculation service through a data interface, and the returned indicator identifier, value, unit, and validity description can be received.

[0061] Step S50: Perform hierarchical automatic judgment on the radio frequency performance indicators according to the externally configured judgment rules, and output the judgment result and judgment evidence chain record. The judgment result includes pass, fail, and uncertain.

[0062] In some of the embodiments described above, the aligned and fused data is used for indicator acquisition and decision-making. However, if a decision is still forced when the truth record is missing, the time alignment exceeds the tolerance, or there is a permutation in the channel mapping, unreliable decision conclusions will be generated.

[0063] To address this, this embodiment further proposes to perform input validity verification and confidence assessment on the fusion results. When alignment fails or the confidence level falls below a preset confidence threshold, the output data is invalid or uncertain and does not proceed to automatic judgment. The alignment fusion processing rules include: matching records within the time window tolerance according to the closest time principle; matching frequency objects according to frequency band overlap; detecting channel substitution in channel mapping and outputting substitution alarms; distinguishing duplicate events according to event identifiers; and marking missing records according to the missing type and reducing the confidence level of the corresponding fusion result. Specifically, the closest time principle means matching RF data with the truth record whose timestamp is closest within the preset window tolerance; frequency band overlap matching means determining the attribution according to the overlap ratio between the component frequency band and the truth object frequency band; channel substitution detection means comparing the actual channel connection relationship with the channel mapping truth value, and outputting substitution alarms when reversed or misaligned connections are found; duplicate events refer to similar events that occur multiple times in the same test process, distinguished by event identifiers and serial numbers; missing records include missing truth values, missing acquired data, and missing timestamps, marked according to the missing type and reducing the confidence level of the corresponding fusion result.

[0064] When output data is invalid or in an uncertain state, the system simultaneously records the specific reasons for this state, such as alignment failures, missing record types, or confidence values ​​below the threshold. This allows for re-testing after reorganizing the data collection or supplementing the truth records. Through this technical solution, data quality issues are explicitly gated before judgment, and uncertain data is no longer subject to forced judgment, avoiding misjudgments caused by garbage input and improving the reliability of automatic judgment conclusions.

[0065] Through the above technical solution, this embodiment introduces a hierarchical automatic decision-making mechanism, refining the complex RF performance evaluation into multiple manageable levels. This makes the decision conclusion no longer a simple pass or fail, but can clearly point out the specific link and cause of the performance anomaly. For example, when the decision result shows a fail, it can immediately pinpoint whether it is insufficient "target protection," inadequate "interference suppression," or a "side effect control" problem, thereby greatly improving the interpretability of the decision result. Simultaneously, by comprehensively recording the truth information, collected data, indicator calculation results, and anomaly sources on which the decision is based, this embodiment constructs a complete traceable chain. This allows engineers to quickly trace back to the original test conditions and data when any performance problem occurs, conduct in-depth analysis and fault attribution, avoid blind troubleshooting, significantly improve the efficiency and accuracy of fault diagnosis, and provide a solid data foundation for subsequent performance optimization. This interpretable and traceable decision-making mechanism not only enhances the depth and breadth of RF performance analysis but also strengthens the reliability and persuasiveness of test results.

[0066] In response, this embodiment further proposes that when the judgment result is abnormal, the abnormality is attributed according to the deviation relationship between the true value and the radio frequency analysis result, and a predefined attribution rule table is used. The attribution rule table defines necessary conditions and exclusion conditions for each risk item. The risk items include target damage, interference residue, channel abnormality, incoming direction mismatch, noise floor rise, and dynamic convergence abnormality. When multiple risk items occur concurrently, they are output in order of priority or confidence. When there is a conflict of evidence, the confidence of the attribution conclusion is reduced and the conflict event is recorded. When the confidence is insufficient, an uncertain attribution result is output.

[0067] Specifically, each rule in the attribution rule table includes a risk item identifier, necessary conditions, exclusion conditions, and an output conclusion. Necessary conditions are the evidentiary conditions that must be met simultaneously for an attribution to be valid. For example, the necessary conditions for a target damage risk item are that the power of the target signal component is lower than the target signal true value by more than a preset amplitude, and the quality indicator of the target signal component is valid. Exclusion conditions are the evidentiary conditions that exclude the risk item once met. For example, when the measured input true value shows that the input target signal power is lower than the configured true value by more than the deviation tolerance, the target damage is excluded, and the input true value is marked as abnormal. By combining necessary and exclusion conditions, arbitrary attributions based solely on a single deviation phenomenon can be avoided.

[0068] When the necessary conditions for multiple risk items are met simultaneously, it is determined to be a concurrent multi-risk event. The risk items are output in order of priority or confidence level, and all hit risk items and their evidence are retained. When there is a conflict between the evidence of different rules, such as the same deviation simultaneously hitting the necessary and exclusion conditions of a risk item, the confidence level of the attribution conclusion is reduced and the conflict event is recorded. When the confidence level of any risk item is lower than the preset confidence threshold, an uncertain attribution result is output, and a prompt is made indicating the need for supplementary test data or manual review. For example, if the target signal component is significantly weaker than the target signal true value and the quality flag is valid, but the exclusion condition for channel anomaly in the attribution rule table is not met, it is attributed to target damage; if the residual interference component is much higher than the expected level corresponding to the interference signal true value, it is attributed to interference residue; if the performance of a specific RF channel deviates significantly from the channel state true value, it is attributed to channel anomaly; if the spatial characteristics of the component do not match the spatial direction true value, it is attributed to direction mismatch; if the noise floor unexpectedly increases, it is attributed to noise floor rise; if the expected performance state is not reached in time under dynamic scenarios, it is attributed to dynamic convergence anomaly.

[0069] Through the above technical solutions, the attribution of anomalies is upgraded from an enumeration-based judgment to a rule-based and interpretable reasoning process: the necessary and exclusion conditions clarify the boundaries of each risk item, and there are clear handling paths for multiple concurrent risks, evidence conflicts, and insufficient confidence. The attribution results are accompanied by rule hit records and confidence levels, which significantly improves the accuracy and auditability of fault diagnosis, enabling engineers to quickly identify and resolve the root causes of performance anomalies.

[0070] This embodiment acquires test scenario truth information containing configuration truth values ​​and execution truth values. Based on a joint association key composed of event identifiers, time windows, frequency objects, and channel mappings, it aligns and fuses the truth values ​​with RF analysis results. After input validity gating, it obtains RF performance indicators based on the fused data, implements hierarchical automatic decision-making and anomaly attribution, and outputs a decision result containing a chain of evidence. When alignment fails or confidence is insufficient, it outputs invalid or uncertain data rather than a forced decision. This solves the problems of inaccurate evaluation, low efficiency, and difficulty in traceability caused by the inability to automatically associate truth values ​​with analysis results in the prior art. It has the advantages of improving test accuracy and efficiency and achieving traceability of results.

[0071] Furthermore, embodiments of this application also propose a computer-readable storage medium storing a program for radio frequency performance analysis and automatic decision-making based on truth value fusion. When the program for radio frequency performance analysis and automatic decision-making based on truth value fusion is executed by a processor, it implements the steps of the method for radio frequency performance analysis and automatic decision-making based on truth value fusion as described above.

[0072] Reference Figure 2 , Figure 2This is a structural block diagram of the first embodiment of the radio frequency performance analysis and automatic decision system based on truth value fusion of this application.

[0073] like Figure 2 As shown in the embodiments of this application, the radio frequency performance analysis and automatic decision system based on truth fusion includes: Information acquisition module 10 is used to acquire test scenario truth information, the truth information including configuration truth and execution truth, the execution truth including channel calibration status; The data acquisition module 20 is used to acquire radio frequency (RF) data of the module under test, which includes at least RF or intermediate frequency (IF) data from the output of the module under test; analyze the RF data, extract the target signal component and residual interference component, and generate a quality indicator characterizing the reliability of the extraction of each component; The fusion module 30 is used to align and fuse the truth information with the radio frequency analysis results based on the joint association key composed of event identifier, time window, frequency object and channel mapping, and to perform input validity verification and confidence evaluation on the fusion result; when the alignment fails or the confidence is lower than the preset confidence threshold, the output data is invalid or uncertain. The calculation module 40 is used to obtain radio frequency performance indicators based on the fused data after successful alignment and fusion. The output module 50 is used to perform hierarchical automatic judgment on the radio frequency performance indicators according to the externally configured judgment rules, and output the judgment result and the judgment evidence chain record. The judgment result includes pass, fail, and uncertain.

[0074] It should be understood that the above are merely illustrative examples and do not constitute any limitation on the technical solution of this application. In specific applications, those skilled in the art can make settings as needed, and this application does not impose any restrictions on this.

[0075] This embodiment acquires test scenario truth information containing configuration truth values ​​and execution truth values. Based on a joint association key composed of event identifiers, time windows, frequency objects, and channel mappings, it aligns and fuses the truth values ​​with RF analysis results. After input validity gating, it obtains RF performance indicators based on the fused data, implements hierarchical automatic decision-making and anomaly attribution, and outputs a decision result containing a chain of evidence. When alignment fails or confidence is insufficient, it outputs invalid or uncertain data rather than a forced decision. This solves the problems of inaccurate evaluation, low efficiency, and difficulty in traceability caused by the inability to automatically associate truth values ​​with analysis results in the prior art. It has the advantages of improving test accuracy and efficiency and achieving traceability of results.

[0076] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of this application. In practical applications, those skilled in the art can select some or all of it to achieve the purpose of this embodiment according to actual needs, and no restrictions are imposed here.

[0077] In addition, for technical details not described in detail in this embodiment, please refer to the method for radio frequency performance analysis and automatic decision based on truth value fusion provided in any embodiment of this application, which will not be repeated here.

[0078] Furthermore, it should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0079] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0080] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory (ROM) / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of this application. The above are only preferred embodiments of this application and do not limit the patent scope of this application. All equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for radio frequency performance analysis and automatic decision-making based on truth value fusion, characterized in that, include: Obtain the truth information of the test scenario, which includes the truth value of the target signal, the truth value of the interference signal, the truth value of the spatial direction of arrival, and the truth value of the channel state; The truth information includes configuration truth and execution truth. The configuration truth is the planned setting value of the test scenario, and the execution truth is the feedback value actually executed by the test device. The execution truth includes the channel calibration status. Collect radio frequency (RF) data from the module under test (DUT), including at least RF or intermediate frequency (IF) data from the output of the DUT; analyze the RF data to extract the target signal component and residual interference component, and generate a quality indicator characterizing the reliability of each component extraction. Based on the joint association key composed of event identifier, time window, frequency object and channel mapping, the truth information is aligned and fused with the radio frequency analysis results, and the fusion result is subjected to input validity verification and confidence evaluation; when the alignment fails or the confidence is lower than the preset confidence threshold, the output data is invalid or uncertain and does not enter the automatic decision. Based on the successful alignment and fusion, radio frequency performance indicators are obtained from the fused data; The radio frequency performance indicators are automatically judged in a hierarchical manner according to the externally configured judgment rules, and the judgment results and judgment evidence chain records are output. The judgment results include pass, fail, and uncertain.

2. The method according to claim 1, characterized in that, When the deviation between the execution truth value and the configuration truth value exceeds the preset deviation tolerance, the execution truth value is used as the standard for alignment and fusion, and the truth value replacement event is marked; when there is a measured input truth value at the input end of the module under test, the execution truth value is verified using the measured input truth value.

3. The method according to claim 1, characterized in that, In the joint association key, the event identifier includes the test event number and the interference source identifier, the time window has a preset window tolerance, the frequency object includes the frequency point, bandwidth, signal type and spatial orientation label or known waveform label, and the channel mapping includes the array element number, RF output channel, module input channel and channel calibration status.

4. The method according to claim 1, characterized in that, The alignment and fusion processing rules include: matching records within the time window tolerance according to the nearest time principle; matching frequency objects according to frequency band overlap; detecting channel substitution in channel mapping and outputting substitution alarms; distinguishing duplicate events according to event identifiers; marking missing records according to the missing type and reducing the confidence of the corresponding fusion result; and outputting invalid or uncertain data when the above processing fails.

5. The method according to claim 3, characterized in that, The association of frequency objects includes: associating the target signal component, interference signal component, and residual interference component with their corresponding truth objects in terms of frequency point, bandwidth, and signal type, and combining interference source identifier, event identifier, spatial orientation label or known waveform label to distinguish multiple interference sources at the same frequency; when there is one component corresponding to multiple truth objects or multiple components corresponding to one truth object, they are assigned according to a preset one-to-many or many-to-one matching rule.

6. The method according to claim 1, characterized in that, The radio frequency performance indicators are obtained through structured indicators input from external consumers or by calling independent indicator calculation modules. The radio frequency performance indicators include, for example, one or more of the following: target signal retention, interference rejection ratio, signal-to-interference ratio improvement, noise floor rise, frequency response impairment, and dynamic response indicators. Their definitions and calculation methods are not limited to the scope of this method.

7. The method according to claim 1, characterized in that, The hierarchical automatic decision-making is carried out according to the levels of input validity, target protection, interference suppression, signal-to-interference ratio improvement, side effect control, and dynamic response, forming an interpretable decision conclusion; the threshold system and indicator sources for each level are provided by external configuration; the decision conclusion is only for radio frequency domain performance, and whether to admit the global navigation satellite system closed-loop verification is determined by the external hierarchical verification process.

8. The method according to claim 1, characterized in that, When the judgment result is abnormal, the abnormality is attributed according to the deviation relationship between the true value and the radio frequency analysis result, and a predefined attribution rule table is used. The attribution rule table defines the necessary conditions and exclusion conditions for each risk item. The risk items include target damage, interference residue, channel abnormality, incoming direction mismatch, noise floor rise, and dynamic convergence abnormality. When multiple risk items occur concurrently, they are output in order of priority or confidence. When there is conflict in the evidence, the confidence of the attribution conclusion is reduced and the conflict event is recorded. When the confidence is insufficient, an uncertain attribution result is output.

9. The method according to claim 1, characterized in that, The judgment evidence chain record includes: the truth information and its configuration version on which the judgment is based, the collected data or its storage reference, the radio frequency performance index results, the judgment rule identifier and judgment rule version of each level of judgment, and the rule hit record and confidence level of the anomaly attribution; the judgment evidence chain record is associated with a unique test identifier to form a traceable test result.

10. A radio frequency performance analysis and automatic decision-making system based on truth value fusion, characterized in that, include: The information acquisition module is used to acquire test scenario truth information, which includes configuration truth and execution truth, and the execution truth includes channel calibration status. The data acquisition module is used to acquire radio frequency (RF) data from the module under test (DUT), including at least RF or intermediate frequency (IF) data from the output of the DUT; the RF data is analyzed to extract target signal components and residual interference components, and a quality indicator characterizing the reliability of each component extraction is generated. The fusion module is used to align and fuse the truth information with the radio frequency analysis results based on a joint association key composed of event identifiers, time windows, frequency objects, and channel mappings, and to perform input validity verification and confidence evaluation on the fusion results; when alignment fails or the confidence level is lower than a preset confidence threshold, the output data is invalid or in an uncertain state. The calculation module is used to obtain radio frequency performance indicators based on the fused data after successful alignment and fusion. The output module is used to perform hierarchical automatic judgment on the radio frequency performance indicators according to the externally configured judgment rules, and output the judgment result and the judgment evidence chain record. The judgment result includes pass, fail, and uncertain.

Images