CN-122027054-A - GSM-R vehicle-mounted front-end wireless interference test system and method thereof

CN122027054ACN 122027054 ACN122027054 ACN 122027054ACN-122027054-A

Abstract

The invention relates to the technical field of railway wireless communication, in particular to a GSM-R vehicle-mounted front-end wireless interference test system and a method thereof. The system comprises a high-resolution frequency spectrum acquisition unit, an integrated multi-dimensional testing unit, a dynamic scene adaptation unit and an intelligent interference diagnosis unit. The intelligent interference diagnosis unit constructs a vehicle-mounted adaptive multipath frequency offset estimation and interference diagnosis model according to the scene prior information generated by the multipath channel modeling and Doppler frequency offset estimation under the vehicle-mounted high-speed mobile environment by the dynamic scene adaptation unit, and realizes the full-dimensional interference feature extraction. The method comprises the steps of adopting a dual mode of known interference accurate recognition and unknown interference clustering to detect burst interference, decoupling source internal and external source interference through feature analysis and hypothesis testing, quantifying hazard degree and generating a disposal instruction. Meanwhile, the system is based on a diagnosis-acquisition closed-loop scheduling mechanism, acquisition parameters are adjusted in a targeted mode, and an interference model is optimized in an iterative mode, so that the system is adaptive to a vehicle-mounted dynamic scene.

Inventors

HONG WEI
QIN XIAODONG
ZHANG LUBIN
QIU FENGYING
WANG GUOYONG
DU YUEPING
ZHANG XIAOLAN
SONG JIACHUN
WU LI
ZHANG YAN
SHI WEIFANG
Ying Shufang
LIAO JUN

Assignees

中国铁路南昌局集团有限公司南昌通信段
北京响盈科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260226

Claims (10)

The GSM-R vehicle-mounted front-end wireless interference test system is characterized by comprising: the high-resolution frequency spectrum acquisition unit (1), the high-resolution frequency spectrum acquisition unit (1) adopts high sampling rate signal processing and narrow-band signal analysis technology to carry out waveform level fine acquisition on the GSM-R signal and output original waveform data containing time domain and frequency domain details; the integrated multi-dimensional testing unit (2), the integrated multi-dimensional testing unit (2) collects multi-dimensional service data through a unified time reference and data fusion engine; The dynamic scene adaptation unit (3), the dynamic scene adaptation unit (3) obtains core parameters representing the running state of the train and the line communication environment, carries out channel modeling and frequency offset estimation aiming at multipath effect and Doppler frequency shift under the high-speed mobile environment, and generates scene priori information comprising multipath incidence angle, frequency offset boundary and scene label; The intelligent interference diagnosis unit (4) is respectively connected with the high-resolution frequency spectrum acquisition unit (1), the integrated multi-dimensional testing unit (2), the dynamic scene adaptation unit (3) and the terminal performance perception evaluation unit (5) to establish a bidirectional real-time communication link, constructs a multipath frequency offset self-adaptive estimation model and an interference diagnosis model based on scene prior information provided by the dynamic scene adaptation unit (3), performs full-dimensional interference characteristic extraction on original waveform data output by the high-resolution frequency spectrum acquisition unit (1), adopts a dual-mode detection architecture parallel to known interference precise identification and unknown interference anomaly clustering to complete real-time detection of sudden interference in operation, realizes decoupling and traceability judgment of endogenous interference and external environment exogenous interference of the vehicle-mounted system through multi-dimensional feature association analysis and hypothesis inspection, completes interference safety influence evaluation and hierarchical treatment instruction generation based on the business-oriented interference hazard quantization model, simultaneously constructs a diagnosis-acquisition bidirectional closed-loop self-adaptive scheduling mechanism, transmits a target acquisition parameter adjustment instruction to the high-resolution frequency spectrum acquisition unit (1), and completes adaptation increment updating and scene based on actual measurement data; The terminal performance perception evaluation unit (5), the terminal performance perception evaluation unit (5) obtains core performance parameters of the vehicle-mounted terminal in the signal receiving and transmitting links through terminal performance test and baseline analysis, and forms terminal performance baseline data.
2. The GSM-R vehicle-mounted front-end wireless interference test system according to claim 1, wherein the high resolution spectrum acquisition unit (1) comprises a signal receiving preprocessing module (11), a high speed sampling processing module (12), a narrowband signal parsing module (13) and a waveform data output module (14), wherein: The signal receiving preprocessing module (11) receives the GSM-R signal and performs preliminary filtering processing on the received GSM-R signal; the high-speed sampling processing module (12) is used for carrying out high-speed sampling processing by adopting a high-sampling rate signal processing technology based on the GSM-R signal processed by the signal receiving preprocessing module (11); the narrow-band signal analysis module (13) analyzes time domain and frequency domain details by adopting a narrow-band signal analysis technology based on the signals sampled by the high-speed sampling processing module (12) to finish waveform level fine acquisition; The waveform data output module (14) is used for sorting the signals analyzed by the narrowband signal analysis module (13) into original waveform data containing time domain and frequency domain details and outputting the original waveform data to the intelligent interference diagnosis unit (4).
3. The GSM-R vehicle-mounted front-end radio interference test system according to claim 2, wherein the dynamic scene adaptation unit (3) comprises a parameter acquisition module (31), a channel modeling module (32), a frequency offset estimation module (33) and a scene information generation module (34), wherein: the parameter acquisition module (31) acquires core parameters representing the running state of the train and the line communication environment; the channel modeling module (32) performs channel modeling on the multipath effect in the high-speed mobile environment based on the core parameters acquired by the parameter acquisition module (31); The frequency offset estimation module (33) carries out frequency offset estimation on Doppler frequency shift under a high-speed mobile environment based on the core parameters acquired by the parameter acquisition module (31); The scene information generation module (34) generates scene prior information containing a multipath incidence angle, a frequency offset boundary and a scene label based on a channel modeling result of the channel modeling module (32) and a frequency offset estimation result of the frequency offset estimation module (33).
4. A GSM-R vehicle-mounted front-end wireless interference testing system according to claim 3, wherein said intelligent interference diagnostic unit (4) comprises a communication interaction module (41), a model building module (42), a dual mode detection module (43), an interference tracing and evaluation module (44) and a closed loop scheduling iteration module (45), wherein: The communication interaction module (41) respectively establishes a bidirectional real-time communication link with the high-resolution frequency spectrum acquisition unit (1), the integrated multi-dimensional testing unit (2), the dynamic scene adaptation unit (3) and the terminal performance perception evaluation unit (5); The model construction module (42) constructs a multipath frequency offset self-adaptive estimation model and an interference diagnosis model based on scene prior information provided by the dynamic scene adaptation unit (3), and performs full-dimension interference feature extraction on the original waveform data output by the high-resolution frequency spectrum acquisition unit (1); The dual-mode detection module (43) adopts a dual-mode detection architecture which accurately identifies the known interference and is clustered with the unknown interference abnormality to complete the real-time detection of the bursty interference in operation; the interference tracing and evaluating module (44) realizes decoupling and tracing judgment of in-vehicle system endogenous interference and external environment exogenous interference through multidimensional feature association analysis and hypothesis test, and completes interference safety influence evaluation and hierarchical treatment instruction generation based on a business-oriented interference hazard quantification model; The closed-loop scheduling iteration module (45) constructs a diagnosis-acquisition bidirectional linkage closed-loop self-adaptive scheduling mechanism, issues a target acquisition parameter adjustment instruction to the high-resolution frequency spectrum acquisition unit (1), and completes light incremental iteration and scene adaptation updating of the interference diagnosis model based on measured data.
5. The GSM-R vehicle-mounted front-end wireless interference test system according to claim 4, wherein the process of constructing the multipath frequency offset adaptive estimation model and the interference diagnosis model by the model construction module (42) and performing full-dimensional interference feature extraction on the original waveform data comprises the steps of: S42.1, reading the real-time running speed, the multipath incidence angle and the line scene label of the train output by the dynamic scene adaptation unit (3), reading the sampling period of the high-resolution frequency spectrum acquisition unit (1), and determining the values of the carrier frequency of the GSM-R signal, the propagation speed of the electromagnetic wave and the frequency offset compensation coefficient; S42.2, carrying out dimension unification on carrier frequencies, and calculating cosine values of multipath incidence angles; S42.3, matching scene weights based on the line scene labels, and carrying out vehicle-mounted scene adaptation compensation on the basic Doppler frequency offset to obtain the Doppler frequency offset after dynamic compensation ; S42.4, will Blending the Kalman filtering state transition matrix, and updating to obtain the Kalman filtering state transition matrix adapting to the vehicle-mounted dynamic scene ; S42.5, will Substituting the filter equation to execute frequency offset estimation on the original waveform data to obtain a final frequency offset estimation value ; S42.6 based on And performing full-dimensional interference feature extraction of time domain, frequency domain and space domain on the original waveform data, and constructing an interference diagnosis model adapting to the GSM-R vehicle-mounted scene.
6. The GSM-R vehicle-mounted front-end wireless interference test system according to claim 5, wherein said dual-mode detection module (43) employs a dual-mode detection architecture that accurately recognizes known interference in parallel with unknown interference anomaly clustering to complete real-time detection of bursty interference in operation, comprising the steps of: s43.1, reading the full-dimensional interference characteristics output by the model construction module (42) to form a characteristic vector to be detected; S43.2, calling a preset GSM-R vehicle-mounted typical interference template feature vector, and calculating to obtain the matching degree through a similarity algorithm ; S43.3 based on Judging whether the feature to be detected is known interference or not, and outputting a preliminary judgment result; s43.4, reading the real-time running speed of the train output by the dynamic scene adaptation unit (3), and dynamically adjusting the DBSCAN cluster radius to obtain ; S43.5, in order to As a core parameter, performing unsupervised clustering on the characteristics which are not judged to be known interference, and completing unknown interference identification; S43.6, carrying out parallel mutual verification on the known interference recognition result and the unknown interference clustering result, and outputting a final interference detection result.
7. The GSM-R vehicle-mounted front-end wireless interference testing system according to claim 6, wherein the process of decoupling and tracing the source and determining by the interference tracing and evaluating module (44) and completing the interference security impact evaluation and hierarchical disposition instruction generation comprises the following steps: S44.1, reading interference related data output by the integrated multi-dimensional testing unit (2) and the terminal performance perception evaluation unit (5), wherein the interference related data at least comprises interference occurrence time, vehicle-mounted equipment start-stop time and interference positioning distance; s44.2, respectively calculating time association weights and space association weights of the interference and the vehicle-mounted equipment; S44.3, carrying out fusion processing on the time correlation weight and the space correlation weight to obtain the time-space correlation weight for judging the attribution of the interference source ; S44.4 based on Executing hypothesis testing, completing decoupling and tracing judgment of the in-vehicle system in-vivo interference and the external environment exogenous interference, and outputting a tracing result; s44.5, reading the performance baseline value and the actually measured value under interference of each GSM-R vehicle-mounted service, and analyzing the performance influence degree of interference on each service; s44.6, combining the traceability judgment result and the service performance influence degree, and obtaining the interference hazard degree through a quantization algorithm ; S44.7 based on A hierarchical disposition instruction is generated.
8. The GSM-R vehicle-mounted front-end radio interference test system of claim 7, wherein said closed-loop scheduling iteration module (45) comprises a scheduling trigger sub-module, a parameter adjustment sub-module, and a model iteration sub-module, wherein: the scheduling triggering submodule starts bidirectional linkage scheduling of interference diagnosis and spectrum acquisition according to the interference detection result output by the double-mode detection module (43); The parameter adjustment submodule adjusts acquisition parameters of the high-resolution frequency spectrum acquisition unit (1) according to the bidirectional linkage scheduling instruction so as to adapt to interference diagnosis requirements; The model iteration submodule executes scene pruning and increment updating on the interference diagnosis model based on the measured data acquired after the high-resolution frequency spectrum acquisition unit (1) is adjusted, and light iteration and scene adaptation of the interference diagnosis model are realized.
9. The GSM-R vehicle-mounted front-end radio interference test system according to claim 8, wherein said terminal performance awareness assessment unit (5) comprises a performance acquisition module (51), a parameter calibration module (52) and a baseline generation module (53), wherein: the performance acquisition module (51) acquires performance parameters of a signal receiving link and a signal transmitting link of the vehicle-mounted terminal in real time and outputs original performance parameter data of the vehicle-mounted terminal; the parameter calibration module (52) performs normalization calibration processing on the original performance parameter data output by the performance acquisition module (51) and outputs standardized performance parameter data; The base line generation module (53) generates and outputs terminal performance base line data of the vehicle-mounted terminal based on the standardized performance parameter data output by the parameter calibration module (52), wherein the terminal performance base line data is used for providing a base reference for vehicle-mounted terminal performance evaluation and system interference analysis.
A method for testing the wireless interference of a front end of a GSM-R vehicle, based on the system for testing the wireless interference of the front end of the GSM-R vehicle according to any one of claims 1 to 9, comprising the steps of: S1, data acquisition and preprocessing, namely acquiring signal data, service data, train operation and line environment parameters and vehicle terminal performance parameters of a GSM-R vehicle-mounted front end in real time, preprocessing and normalizing the acquired various data, and generating original waveform data, multidimensional service data, scene priori information and terminal performance baseline data; S2, multidimensional feature extraction and model construction, namely completing Doppler frequency offset dynamic compensation and frequency offset estimation based on various preprocessed data, executing time domain, frequency domain and space domain full-dimensional interference feature extraction on the original waveform data, and constructing a multipath frequency offset self-adaptive estimation model and an interference diagnosis model which are adaptive to a GSM-R vehicle-mounted scene; S3, interference detection, namely carrying out matching judgment on known interference based on the extracted full-dimensional interference characteristics through template matching and unsupervised clustering dual-mode parallel detection, carrying out cluster recognition on unknown interference, and outputting a final interference detection result through parallel mutual verification; S4, calculating space-time association weights of interference and vehicle-mounted equipment based on multi-dimensional service data, terminal performance baseline data and interference related data, completing interference source attribution judgment, analyzing the performance influence degree of interference on each service, and generating interference hazard degree and grading treatment instructions; S5, closed-loop scheduling and model iteration, namely starting a bidirectional linkage scheduling mechanism according to a final interference detection result, issuing a target acquisition parameter adjustment instruction, and executing scene pruning and incremental updating on the interference diagnosis model based on the actual measurement data acquired after adjustment, so as to realize light-weight iteration and scene adaptation of the model.

Description

GSM-R vehicle-mounted front-end wireless interference test system and method thereof Technical Field The invention relates to the technical field of railway wireless communication, in particular to a GSM-R vehicle-mounted front-end wireless interference test system and a method thereof. Background The GSM-R digital mobile communication system is a special mobile communication core for railways, the vehicle-mounted front end is used as a key node for communication between a train and the ground, and the communication quality of the GSM-R digital mobile communication system directly relates to the driving safety of the railways. The multipath effect and Doppler frequency shift generated by the high-speed operation of the train can distort the signal characteristics, and in addition, the in-vehicle endogenous electromagnetic interference and the peripheral exogenous wireless interference of the line are easy to cause error codes and interruption of a communication link, so that the accurate test, detection and traceability treatment of the wireless interference are carried out on the front end of the GSM-R train, and the method is a core technical requirement for guaranteeing the stable operation of a railway communication system. In the prior art, related patents have been researched and explored in the field of GSM-R wireless interference identification and monitoring. The invention patent CN202411924347.5 discloses a GSM-R interference automatic identification method and device based on sweep frequency, the method obtains a railway line radio signal through a GSM-R special antenna, obtains corresponding data through decoding of a sweep frequency meter and spectrum analysis of a frequency spectrograph, and combines the information of a known district ledger to identify internal and external interference and risk of a GSM-R network, determine the interference type and related information and display in real time, thereby improving the efficiency, comprehensiveness and timeliness of interference identification. In another example, patent CN202210944220.4 discloses an automatic identification device and method for GSM-R interference source, the scheme uses uniform circular array antenna and special antenna for GSM-R to receive radio signal along railway, and after processing by multichannel receiver and power divider, respectively, the frequency sweep and real-time spectrum processing of correspondent signal are completed, the real-time processing unit combines space-time positioning information and cell carrier frequency configuration information to identify neighboring cell and monitor its multipath signal, finally combines frequency sweep, spectrum processing, multipath monitoring and carrier frequency configuration information to complete interference signal identification, and outputs interference identification result. The technical scheme has the advantages of corresponding design, but has the technical defects that firstly, the adaptability to a high-speed mobile vehicle-mounted scene is insufficient, the comprehensive performance, traceability accuracy and scheme self-optimization capability of interference detection are obviously limited, the invention patent CN202411924347.5 only relies on cell ledger information and preset interference types to identify internal and external interference, an adaptive analysis model is not constructed for multipath effects and Doppler frequency shifts generated by train high-speed movement, the signal characteristics of dynamic distortion cannot be analyzed, only known interference can be identified, a linkage mechanism of diagnosis and acquisition is not needed, and the model iterative optimization capability is also lacked, the invention patent CN202210944220.4 can monitor multipath signals of adjacent cells, but does not accurately estimate and compensate Doppler frequency offset, does not have quantized interference source attribution judging logic, cannot decouple vehicle-mounted internal interference and external exogenous interference, cannot quantitatively evaluate and classify interference hazards, and cannot complete scene adaptation of the model according to train running states and line scenes. The invention CN202411924347.5 only completes basic decoding and spectrum measurement of signals, does not realize waveform level fine acquisition of GSM-R signals and is easy to miss weak interference characteristics, the invention CN202210944220.4 does not acquire and standardize performance parameters of a receiving and signaling link of a vehicle-mounted terminal, does not construct a terminal performance baseline, only can qualitatively identify interference, can not quantitatively evaluate the hazard degree of the interference to railway key business, and lacks scientific decision basis of interference treatment. In view of this, we propose a system and method for testing the wireless interference of the front end of the GSM-R vehicle. Disclosure of Invention