Intelligent networking communication method and system for rail train

Abstract

The application discloses an intelligent networking communication method and system for a rail train, and relates to the technical field of train wireless communication. The method comprises the following steps: constructing a virtual transmission cone according to the positions of a train and ground access points, and generating an interference noise cloud through spectrum scanning. Then, the method performs geometric superposition calculation on the virtual transmission cone and the interference noise cloud to determine a radio frequency conflict area. Based on the spatial position of the radio frequency conflict area in the transmission path, the method distinguishes different interference conflict types such as near-field flooding, mid-far-field attenuation and receiver sensitivity, and performs differentiated anti-interference strategies such as beamforming null avoidance, time-frequency domain offset or redundant retransmission. The application can implement precise defense against interference sources at different spatial positions.

Description

Technical Field

[0001] This application relates to the field of train wireless communication technology, and in particular to an intelligent networking communication method and system for rail trains. Background Technology

[0002] Modern urban rail transit systems rely heavily on stable wireless data links to transmit train control commands and status information. Maintaining a continuous connection between the onboard terminal and the ground access point during train operation is fundamental to ensuring safe operation.

[0003] With the widespread adoption of mobile internet, the number of smart devices carried by passengers has increased dramatically. When trains stop at stations during the evening rush hour, the density of people in the carriages is extremely high, and many passengers' mobile devices activate hotspot functions, emitting radio waves that overlap with the train's communication frequency band.

[0004] Because vehicle antennas are typically mounted on the roof, the physical distance between the antenna and passengers' heads is very close. This close proximity causes strong interference signals generated by passenger hotspots to directly cover the space surrounding the vehicle antenna.

[0005] Existing communication anti-interference technologies are typically based on signal-to-noise ratio (SNR) assessment or simple spectrum monitoring. When interference is detected, existing technologies often employ methods such as uniformly increasing transmit power or switching channels across the entire network. However, when faced with strong near-field interference sources located close to the antenna, increasing transmit power cannot overcome the signal shielding blind spots created by physical obstructions.

[0006] Meanwhile, when faced with random signal collisions on the transmission path, traditional channel switching strategies are unable to match the time-frequency characteristics of interference signals, resulting in control signals being submerged or distorted, which seriously affects the transmission reliability of critical train data. Summary of the Invention

[0007] In view of the aforementioned problems, this application is hereby filed.

[0008] Therefore, this application provides an intelligent networking communication method and system for rail trains, which can solve the problem of unstable vehicle-to-ground communication links and data transmission failures caused by passenger hotspot interference when the train stops at the platform.

[0009] To solve the above-mentioned technical problems, this application provides the following technical solution: In a first aspect, this application provides an intelligent networking communication method for rail trains, including: obtaining the train stopping position and the ground access point position, and constructing a virtual transmission cone based on the train stopping position and the ground access point position, wherein the virtual transmission cone represents the electromagnetic wave coverage space range of the train antenna sending control signals to the ground access point; Spectrum scanning is performed inside the carriage and around the train to obtain spatial location information of interference sources. Based on the spatial location information of interference sources, interference noise cloud is generated in three-dimensional space. The interference noise cloud represents the energy distribution pattern of radio frequency signals generated by passenger hotspot devices. Spatial geometric superposition is performed on the virtual transmission cone and the interference noise cloud to determine the overlapping volume of the virtual transmission cone and the interference noise cloud, and the overlapping volume is marked as the radio frequency conflict zone; Based on the spatial location of the radio frequency conflict zone in the virtual transmission cone, the target interference conflict type is determined, and an intelligent networking anti-interference strategy is executed based on the target interference conflict type.

[0010] Preferably, the construction of the virtual transmission cone based on the train stopping position and the ground access point position includes: Using the phase center of the train antenna as the origin and the antenna center of the ground access point as the target point, a main transmission vector is constructed connecting the origin and the target point. The beamforming configuration parameters built into the vehicle-mounted antenna are retrieved, and a conical geometric model extending outward along the main transmission vector is generated based on the direction parameters of the main transmission vector and the phase excitation distribution in the beamforming configuration parameters. A boundary truncation determination is performed on the conical geometric model. When any one of the physiological geometric truncation condition, energy threshold truncation condition, or line-of-sight truncation condition is triggered, the extension of the conical geometric model is trunculated to generate the virtual transmission cone.

[0011] Preferably, the step of performing a spectrum scan of the interior of the carriage and the surrounding area of ​​the train to obtain spatial location information of the interference source, and generating an interference noise cloud in three-dimensional space based on the spatial location information of the interference source, includes: Unlicensed frequency band signal sources inside the carriage are identified by spectrum scanning, and the spatial azimuth and elevation angles of each unlicensed frequency band signal source relative to the train antenna are determined by multi-antenna angle of arrival estimation technology. Multiple unlicensed frequency band signal sources that are spatially close and have a frequency overlap higher than a preset overlap threshold are grouped into interference clusters, and the geometric centroid of each interference cluster is calculated and marked as a noise anchor point. A spherical expansion scan is performed with the noise anchor point as the center. When it is determined that adjacent expanded spheres are in contact or overlap, the spheres are merged. A Boolean subtraction operation on the carriage structure is performed on the merged sphere to generate the interference noise cloud.

[0012] Preferably, the step of grouping multiple unlicensed frequency band signal sources that are spatially adjacent and have a frequency overlap higher than a preset overlap threshold into interference clusters, calculating the geometric centroid of each interference cluster, and marking the geometric centroid as a noise anchor point includes: Monitor the radio frequency signal strength of each unlicensed frequency band signal source at the receiving end, and estimate the local energy density parameters of each unlicensed frequency band signal source around the vehicle antenna based on the radio frequency signal strength and propagation distance model; High-energy interference sources with local energy density parameters higher than a preset density threshold are selected, and high-energy interference sources that are spatially close are marked as candidate aggregation objects; Calculate the frequency overlap between the candidate aggregation objects, and when the frequency overlap is determined to be higher than the preset overlap threshold, merge the candidate aggregation objects into the interference cluster; Based on the spatial coordinates of the candidate aggregation objects contained in the interference cluster, the mass center coordinates of the interference cluster are calculated, and the mass center coordinates are determined as the noise anchor point.

[0013] Preferably, the step of performing spatial geometric superposition of the virtual transmission cone and the interference noise cloud to determine the overlapping volume of the virtual transmission cone and the interference noise cloud, and marking the overlapping volume as the radio frequency conflict region, includes: The interference noise cloud and the virtual transmission cone are superimposed in a three-dimensional spatial coordinate system. The overlapping volume block inside the virtual transmission cone that is eroded by the interference noise cloud is extracted using a three-dimensional geometric intersection algorithm. The overlapping volume block is marked as a spatial conflict sub-body. Listen to the preamble of the channel signal and analyze the transmission period and duty cycle of the unlicensed frequency band signal source; The transmission time of the control data packet to be sent by the train is compared with the active time window of the unlicensed frequency band signal source. When it is determined that the transmission time is within the active time window and the spatial conflict sub-body is covered by the interference noise cloud, the radio frequency conflict zone is activated.

[0014] Preferably, determining the target interference type based on the spatial location of the radio frequency conflict zone in the virtual transmission cone includes: Using the physical outer edge of the train antenna as the first cutting surface, the near-field boundary as the second cutting surface, and the effective receiving range boundary of the ground access point antenna as the third cutting surface, the virtual transmission cone is divided into a near-field flooding area, a mid-to-far-field attenuation area, and a receiving end sensitive area. Determine whether the radio frequency conflict zone and the near-field flooding zone overlap. If they overlap, mark the target interference conflict type as a near-field conflict type. Determine whether the radio frequency conflict region and the mid-to-far field attenuation region overlap. If they overlap, mark the target interference conflict type as a mid-to-far field conflict type. Determine whether the radio frequency conflict area overlaps with the receiver's sensitive area. If there is overlap, mark the target interference conflict type as a far-field conflict type.

[0015] Preferably, the step of executing the intelligent networking anti-interference strategy based on the target interference conflict type includes: In response to the target interference conflict type being the near-field conflict type, the interference cluster anchor point corresponding to the radio frequency conflict zone is obtained, and the interference azimuth angle and interference elevation angle of the interference cluster anchor point relative to the train antenna are calculated. The weighting coefficients of each radiating element are adjusted by calling the antenna array control interface. In the direction directly opposite the interference azimuth and the interference elevation angle in the antenna radiation pattern, a gain null is artificially created by the phase cancellation principle. By compressing the main lobe width of the virtual transmission cone, the control signal energy is concentrated in the direction of the ground access point, forming a signal transmission corridor that avoids the radio frequency conflict zone.

[0016] Preferably, the step of executing the intelligent networking anti-interference strategy based on the target interference conflict type includes: In response to the target interference conflict type being the mid-to-far field conflict type, the time-domain and frequency-domain features of the interference signal inside the radio frequency conflict zone are extracted, and the center frequency point and transmission period of the interference signal are identified. Time offset processing is performed on the control data packets to be sent by the train, delaying the transmission time of the control data packets to the quiet interval within the transmission cycle; Alternatively, frequency switching processing can be performed on the control data packets to be sent by the train, switching the transmission frequency band of the control data packets to a spare frequency band that does not overlap with the center frequency point.

[0017] Preferably, the step of executing the intelligent networking anti-interference strategy based on the target interference conflict type includes: In response to the target interference conflict type being the far-field conflict type, the diversity reception logic of the ground access terminal is activated, multiple spatially separated receiving antennas are used to receive control signals, and the one with the optimal signal-to-noise ratio is selected as the input signal. Redundant check bytes are added to the control data packets to be sent by the train, and the modulation mode of the control data packets is downgraded to a low-order modulation mode; Alternatively, the transmission polarization of the control signal can be changed to counteract co-frequency interference using polarization isolation.

[0018] Secondly, this application also provides an intelligent networking communication system for rail trains, including: a transmission cone construction module, used to obtain the train stopping position and the ground access point position, and construct a virtual transmission cone based on the train stopping position and the ground access point position, wherein the virtual transmission cone represents the electromagnetic wave coverage space range of the train antenna sending control signals to the ground access point; The interference cloud generation module is used to perform spectrum scanning on the interior of the carriage and the surrounding area of ​​the train to obtain spatial location information of the interference source. Based on the spatial location information of the interference source, an interference noise cloud is generated in three-dimensional space. The interference noise cloud represents the energy distribution pattern of radio frequency signals generated by passenger hotspot devices. The conflict zone determination module is used to perform spatial geometric superposition of the virtual transmission cone and the interference noise cloud, determine the overlapping volume of the virtual transmission cone and the interference noise cloud, and mark the overlapping volume as the radio frequency conflict zone; The strategy execution module is used to determine the target interference conflict type based on the spatial location of the radio frequency conflict zone in the virtual transmission cone, and to execute an intelligent networking anti-interference strategy based on the target interference conflict type.

[0019] Implementing this application will have the following beneficial effects: This application provides an intelligent networking communication method and system for rail trains. By constructing a geometric model of a virtual transmission cone and an interference noise cloud, this application transforms abstract electromagnetic interference into a visualized spatial positional relationship, enabling the system to identify the specific location of the interference source relative to the antenna. For strong near-field interference adjacent to the antenna, beamforming null technology is used to actively avoid the interference direction, overcoming the physical shielding limitations that traditional power increases cannot solve. For interference along the transmission path, a time-frequency offset strategy is employed to stagger the transmission time or frequency band, reducing the probability of signal collisions. By distinguishing interference types at different spatial locations and implementing differentiated handling, this application effectively improves the communication stability of trains in densely populated station environments, ensuring the reliable transmission of train control data. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart of a smart networking communication method for rail trains, which relates to this application. Figure 2This is a schematic diagram illustrating the construction process of a virtual transmission cone for an intelligent networking communication method for rail trains, as described in this application. Figure 3 This is a schematic diagram illustrating the generation logic of an interference noise cloud in an intelligent networking communication method for rail trains, as described in this application. Figure 4 This is a schematic diagram of the spatial geometric superposition and collision zone activation mechanism of an intelligent networking communication method for rail trains involved in this application; Figure 5 This is a schematic diagram of the area determination rules for interference conflict types in an intelligent networking communication method for rail trains, which is involved in this application. Figure 6 This is a schematic diagram of the anti-interference handling process for near-field collisions in an intelligent networking communication method for rail trains, which is involved in this application. Figure 7 This is a schematic diagram of a time-frequency domain avoidance strategy for mid-to-far field collision types in an intelligent networking communication method for rail trains, as described in this application. Figure 8 This is a schematic diagram of a link margin compensation mechanism for far-field collision types in an intelligent networking communication method for rail trains, which is involved in this application. Figure 9 This is a schematic diagram illustrating the execution endpoints of various differentiated anti-interference strategies in an intelligent networking communication method for rail trains, as described in this application. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with 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.

[0023] Example 1, referring to Figures 1-9 This is the first embodiment of the present application, which provides an intelligent networking communication method for rail trains, including: This application provides a method that can effectively solve the problems mentioned above. The following will describe in detail how to implement the intelligent networking communication method for rail trains with reference to several embodiments. Figure 1 A flowchart of a method for intelligent networking communication for rail trains is shown, including: Step S1: Obtain the train stopping position and the ground access point position. Construct a virtual transmission cone based on the train stopping position and the ground access point position. The virtual transmission cone represents the electromagnetic wave coverage area of ​​the train antenna sending control signals to the ground access point.

[0024] It should be noted that during train stops at the platform, the spatial configuration of the communication link between the onboard antenna and the ground access point determines the energy coverage range of the control signal in three-dimensional space. The virtual transmission cone extends outward from the phase center of the train antenna along the direction pointing to the ground access point. The phase center refers to the equivalent synthesis starting point of the signal components of each radiating element of the antenna array during transmission. The virtual transmission cone is used to geometrically delineate the main energy coverage area of ​​the control signal, providing a unified spatial reference for subsequent superposition with interference noise clouds.

[0025] Specifically, the train's stopping position is determined by matching the stopping coordinates reported by the onboard positioning device with the electronic map of the track line. The matching method involves comparing the stopping coordinates with the coordinate range of each platform segment on the electronic map one by one. The platform segment within which the stopping coordinates fall is identified as the stopping platform, and the specific value of the stopping coordinates in the coordinate system is recorded. The locations of ground access points are pre-entered into the installation coordinate register of each platform access point along the line. During the train's entry into the station, after the onboard terminal completes wireless association with the access point, it reads the access point's device number from the association response message and then uses the device number to query the installation coordinate register to obtain the access point's coordinates.

[0026] Furthermore, after the virtual transmission cone is constructed, the three parameters—the cone's axial direction, angle, and axial extension length—are written into a unified spatial parameter record. The axial direction is determined by both the train's stopping coordinates and the access point's installation coordinates. The angle and axial extension length are calculated and truncated by subsequent steps S1.2 and S1.3. This spatial parameter record also serves as a shared coordinate reference for S2 (generating interference noise clouds), S3 (performing superposition), and S4 (cutting sub-regions), ensuring spatial consistency across all processes. For example, after a train stops at a platform, the stopping coordinates match and lock the corresponding platform section. The access point equipment number is read from the associated response message, and its installation coordinates are retrieved from the ledger. These two parameters determine the main transmission direction. The angle and extension length are then calculated item by item in S1.2 and S1.3 and written into the spatial parameter record.

[0027] In some preferred embodiments, constructing a virtual transmission cone based on the train's stopping location and the location of the ground access point may further include the following steps: S1.1, with the phase center of the train antenna as the origin of the coordinate system and the antenna center of the ground access point as the target point, construct the main transmission vector connecting the origin of the coordinate system and the target point; S1.2, retrieve the beamforming configuration parameters built into the vehicle antenna, and generate a conical geometric model extending outward along the main transmission vector based on the direction parameters of the main transmission vector and the phase excitation distribution in the beamforming configuration parameters. S1.3, perform boundary truncation judgment on the conical geometric model. When any one of the physiological geometric truncation condition, energy threshold truncation condition or line-of-sight truncation condition is triggered, the extension of the conical geometric model is trunculated and a virtual transmission cone is generated.

[0028] Specifically, in S1.1, a local three-dimensional Cartesian coordinate system is established with the phase center of the train antenna as the origin. The three coordinate axes are denoted as the horizontal east-west direction, the horizontal north-south direction, and the vertical direction, respectively. The coordinates of the target point in the coordinate system are read from the center of the ground access point antenna. The components of the target point coordinates in the three coordinate axes are subtracted from the corresponding components at the origin. The three differences are combined in the order of the coordinate axes to form the main transmission vector pointing from the origin to the target point. The main transmission vector is a spatial vector defined in this application to indicate the main transmission direction of the control signal. The main transmission vector is then normalized by dividing the components of the main transmission vector in the three coordinate axes by the length of the main transmission vector, resulting in a unit direction vector that only represents directional information. The direction parameter of the main transmission vector is represented by the components of the unit direction vector in the three coordinate axes.

[0029] It can be explained that the vehicle-mounted antenna is a multi-element array antenna. Each radiating element is given its own phase excitation during transmission. The phase excitation distribution refers to the set of parameters formed by arranging the phase excitation of all radiating elements in the array according to the element positions. The beamforming configuration parameters refer to the total table of amplitude and phase distribution parameters that are fixed at the factory when the antenna leaves the factory. The signal components transmitted by each element are superimposed in phase in a specific direction and weakened by each other in other directions. Therefore, the phase excitation distribution determines the direction of the synthesized beam and the main lobe angle.

[0030] Furthermore, in S1.2, the phase excitation of each radiating element in the beamforming configuration parameters is read along the main transmission vector direction. Based on the installation position coordinates of each radiating element, the main lobe pointing direction and the main lobe half-angle of the synthesized beam are calculated. The main lobe half-angle is denoted as... , representing the angle from the main lobe axis to the edge of the main lobe. A conical cross-section is set at preset axial steps along the axis, and the axial distance from the cross-section to the apex of the cone is denoted as . The cross-sectional radius is denoted as Cross-sectional radius according to Calculation, where The radius of the corresponding cross section. This represents the axial distance from the cross-section to the apex of the cone. Using the main lobe half-angle as a guide, the edge points of each cross-section are connected sequentially to obtain a conical geometric model extending outward along the main transmission vector. The preset axial step size is obtained based on an empirical trade-off between the accuracy of the cone model volume calculation and the computational load, with a preferred range of 0.2 meters to 1 meter. In the specific operation of this application, the straight-line distance between the train and the access point is read. If the straight-line distance is less than 300 meters, the preset axial step size is set to 0.2 meters; if the straight-line distance is not less than 300 meters, the preset axial step size is set to 0.5 meters.

[0031] Furthermore, in S1.3, for the physiological geometric truncation condition, the relative position of the outer surface of the cone and the surrounding entities is checked segment by segment along the cone axis. The checking method is to discretize the outer surface of the cone into surface sampling points at preset surface sampling intervals, and calculate the shortest distance between each surface sampling point and the entity surface in the carriage shell model, the roof equipment compartment model, and the platform canopy model. When the shortest distance is less than or equal to zero, it is determined that the surface sampling point touches the entity. The location where there is a touching point is determined to trigger the physiological geometric truncation condition, and the cone part after the touching position is truncated. The preset surface sampling interval is obtained based on the trade-off experience between the size of the entity model surface and the truncation determination accuracy. The preferred setting range is 0.1 meters to 0.3 meters. In the specific operation of this application, the side length of the platform canopy model surface is read. If the side length of the surface is greater than 0.5 meters, the preset surface sampling interval is set to 0.3 meters. If the side length of the surface is not greater than 0.5 meters, the preset surface sampling interval is set to 0.1 meters.

[0032] Simultaneously, regarding the energy threshold cutoff condition, the arriving power of the control signal at each position along the cone's axis is calculated according to the free-space propagation attenuation law. The arriving power is obtained by subtracting the free-space attenuation from the transmitted power. The free-space attenuation is calculated cumulatively based on the propagation distance and the corresponding wavelength of the operating frequency band, with an attenuation of 6dB for every doubling of distance. When the arriving power attenuates below the preset energy threshold, the energy threshold cutoff condition is triggered, and the cone extension is cut off at the corresponding position. The preset energy threshold is obtained by superimposing the receiving sensitivity of the ground access point with the uplink margin, and the preferred setting range is 60dB to 80dB below the transmitted power. In the specific operation of this application, the receiving sensitivity parameter of the ground access point is read. If the access point uses a high-sensitivity receiving module, the preset energy threshold is set to leave an 8dB margin above the receiving sensitivity; if the access point uses a conventional receiving module, the preset energy threshold is set to leave a 12dB margin.

[0033] For example, after the train stops at the platform, the phase center of the onboard antenna is used as the origin of the coordinate system, and the center of the ground access point antenna is located 120 meters away in the direction of 45 degrees azimuth and 10 degrees elevation, forming the main transmission vector. The main lobe half-angle is calculated according to the phase excitation distribution. The degree is 6 degrees, and the cross-sections are arranged in 0.2-meter increments along the axial direction and according to... After calculating the radius of each cross section, the edge points are connected to obtain the cone model. When the cone extends to 45 meters, the surface sampling point touches the edge of the platform canopy, triggering geometric truncation. The calculation continues until 95 meters, when the power drops below the preset energy threshold, triggering energy truncation. Finally, the virtual transmission cone covers the cone-shaped space from 0 meters to 95 meters along the axis, and the three parameters of the cone are written into the spatial parameter record.

[0034] Step S2: Perform spectrum scanning on the interior of the carriage and the surrounding area of ​​the train to obtain spatial location information of the interference source. Based on the spatial location information of the interference source, generate an interference noise cloud in three-dimensional space. The interference noise cloud represents the energy distribution pattern of radio frequency signals generated by passenger hotspot devices.

[0035] It should be noted that the passenger hotspot devices are distributed throughout the seating areas of the carriage, and the resulting radio frequency signals are not uniformly distributed in space, but rather form localized energy-rich areas around the signal source. The interference noise cloud is used to characterize this energy distribution pattern in three-dimensional space, enabling S3's superposition to locate the specific spatial area where the interference energy overlaps with the virtual transmission cone, rather than merely sensing interference at the frequency level.

[0036] Specifically, the spectrum scanning is performed by the on-board scanning receiving channel. This channel operates within the local three-dimensional Cartesian coordinate system established during the construction of the virtual transmission cone in S1. The scanning range covers all seating areas inside the carriage and a pre-defined perimeter scanning circle around the train. The scanning results include three raw data points for each signal source: frequency, received power, and direction of arrival. The spatial location information of the interference source is resolved in the local three-dimensional Cartesian coordinate system, sharing a coordinate reference with the virtual transmission cone. This ensures that the cloud generated in S2, the superposition results in S3, and the sub-region segmentation in S4 can be directly compared in spatial location.

[0037] Furthermore, the interference noise cloud is represented by a spherical ensemble model, that is, the energy distribution boundary is approximated by the union of several spheres. The center coordinates and radius of each sphere in the spherical ensemble model are recorded in the cloud parameter table. The cloud parameter table is passed down step by step in the superposition and cutting stages. The specific generation method is shown in S2.1 to S2.3.

[0038] Preferably, the process of performing a spectrum scan of the interior of the carriage and the surrounding area of ​​the train to obtain spatial location information of the interference source, and generating an interference noise cloud in three-dimensional space based on the spatial location information of the interference source, may also include the following steps: S2.1, identify unlicensed frequency band signal sources inside the carriage through spectrum scanning, and use multi-antenna angle of arrival estimation technology to determine the spatial azimuth and elevation angles of each unlicensed frequency band signal source relative to the train antenna; S2.2, group multiple unlicensed frequency band signal sources that are spatially close and have a frequency overlap higher than a preset overlap threshold into an interference cluster, calculate the geometric centroid of each interference cluster, and mark the geometric centroid as a noise anchor point; S2.3 Perform spherical expansion scanning with the noise anchor point as the center. When it is determined that adjacent expanded spheres are in contact or overlap, merge the spheres and perform Boolean subtraction operation on the merged spheres to generate interference noise cloud.

[0039] Specifically, in S2.1, the scanning receiving channel stays at fixed frequency intervals within a preset scanning frequency band for measurement. The stay measurement refers to the receiving channel staying at each frequency point for a preset stay time and recording the received power at that frequency point. Adjacent frequency points with received power higher than the noise floor threshold are connected to form a segment to obtain candidate signal segments. The occupied frequency bands of the candidate signal segments are then compared one by one with the list of authorized frequency bands. The list of authorized frequency bands includes licensed frequency bands such as vehicle-to-ground communication bands and public mobile communication bands. Any candidate signal segments not included in the list of authorized frequency bands are determined to be unlicensed frequency band signal sources.

[0040] The preset scanning frequency band is obtained by calculating the protection bandwidth extended to both sides of the train communication frequency band. The preferred setting range is 20MHz to 60MHz extended to both sides of the communication frequency band. In the specific operation of this application, the center frequency point and occupied bandwidth of the vehicle communication module are read. If the occupied bandwidth is less than 10MHz, the single-sided protection bandwidth is set to 60MHz. If the occupied bandwidth is not less than 10MHz, the single-sided protection bandwidth is set to 20MHz.

[0041] The noise floor threshold is obtained by measuring the noise floor power of the scanning frequency band during a period when there is no passenger hotspot activity in the depot. The preferred setting range is 5dB to 8dB above the noise floor power. In the specific operation of this application, the noise floor power value of the most recent depot self-test record is retrieved. If the difference between the ambient temperature of the depot self-test and the current scan exceeds 10°C, the noise floor threshold is raised by 2dB. Otherwise, the value is taken as the measured noise floor power plus 6dB.

[0042] Furthermore, in S2.1, multiple antenna elements in the train antenna array simultaneously receive signals from the same signal source. The path lengths of the incoming waves to different antenna elements are different, and this path length difference causes a phase difference between the received signals of each element. The phase difference between the received signals of each antenna element is calculated pairwise, and then the spatial azimuth and elevation angles of the signal source relative to the train antenna are deduced based on the installation spacing between the antenna elements. The spatial azimuth angle represents the direction of the incoming wave in the horizontal plane, and the elevation angle represents the tilt angle of the incoming wave in the vertical plane. The two angles, combined with the received power obtained from the scanning, constitute the spatial azimuth information of each unlicensed frequency band signal source. Spectrum scanning and multi-antenna angle of arrival estimation are existing measurement methods in the field of communications. This application unifies the outputs of both into the coordinate calculation stage of the local three-dimensional rectangular coordinate system, so that the measurement results directly participate in subsequent geometric calculations in coordinate form.

[0043] It can be explained that in S2.2, the merging criterion requires two conditions to be met simultaneously: first, the two signal sources are spatially adjacent, meaning the straight-line distance between their coordinates is less than a preset proximity distance; second, the frequency overlap of the two signal sources is higher than a preset overlap threshold. Calculate using the following formula: ; in, The frequency overlap of the frequency bands occupied by the two signal sources. The intersection width of the frequency bands occupied by the two signal sources. The width of the union of the frequency bands occupied by the two signal sources is defined as follows: The preset overlap threshold is obtained by statistically analyzing the relative positional relationship between the channel bandwidth of the hotspot device and the train communication frequency band, with a preferred setting range of 40% to 70%. In the specific operation of this application, the number of identified hotspot devices in the carriage is counted. If the number of hotspot devices exceeds a preset threshold, the preset overlap threshold is lowered to 40% to expand the merging range; otherwise, the preset overlap threshold is set to 60%. The preset quantity threshold is obtained by statistically analyzing the carriage capacity and the ratio of hotspot device activation, with a preferred setting range of 10% to 20% of the carriage capacity. In the specific operation of this application, the rated passenger capacity of the carriage is read. If it is during peak hours, the preset quantity threshold is set to 10% of the rated passenger capacity; if it is during off-peak hours, the preset quantity threshold is set to 20% of the rated passenger capacity. The preset proximity distance is obtained by statistical analysis of the spacing between adjacent seats in the carriage, with a preferred setting range of 1 to 2 meters. In the specific operation of this application, the spacing parameters of the carriage seats are read. If the seats are arranged facing each other, the preset proximity distance is set to 2 meters. If the seats are arranged in the same direction, the preset proximity distance is set to 1 meter.

[0044] Furthermore, in S2.2, for each interference cluster, the average value of the spatial coordinates of all signal sources within the cluster along the three coordinate axes is taken to obtain the centroid coordinates representing the geometric center of the cluster. The geometric centroid is marked as a noise anchor point. The noise anchor point is a reference point defined in this application to represent the position of the entire interference cluster in space. The coordinates of the noise anchor point, along with the list of signal sources within the cluster, are written into the cloud parameter table.

[0045] Meanwhile, in S2.3, a spherical expansion scan is performed with the noise anchor point as the center and the sphere radius increasing sequentially with a preset expansion step size. Each time the radius is increased, the straight-line distance between the centers of adjacent expanded spheres is calculated. When the straight-line distance is less than the sum of the radii of the two spheres, contact or overlap is determined. Spheres that have been determined to be in contact or overlap are merged into a single sphere. The radius of the merged sphere is the maximum range enclosed by the outer boundaries of the two spheres. The expansion and merging operations are repeated until no new contact occurs between all spheres. The preset expansion step size is obtained based on a trade-off between the directional error estimated by the angle of arrival and the cloud cluster's fit, with a preferred range of 0.3 meters to 0.8 meters. In the specific operation of this application, the estimated directional error value of the angle of arrival is obtained. If the estimated directional error value is greater than 5 degrees, the preset expansion step size is set to 0.8 meters; if the estimated directional error value is not greater than 5 degrees, the preset expansion step size is set to 0.3 meters.

[0046] On the other hand, in S2.3, the fused set of spheres is subjected to Boolean subtraction with the carriage structure model, including the inner wall, seats, and equipment cabinets. Boolean subtraction refers to removing the space overlapping with the carriage structure model from the space enclosed by the set of spheres. The remaining space is the interference noise cloud. The outer surface of the cloud represents the distribution boundary of radio frequency signal energy inside and outside the carriage. Finally, the center coordinates and radius of each sphere in the cloud are written into the cloud parameter table. For example, if three hotspot devices are found in the seating area, and the ratio of the intersection width to the union width of their occupied frequency bands is 65%, which is higher than the preset overlap threshold, and the distance between each pair is less than the preset proximity distance, they are merged into an interference cluster. The geometric centroid is located 0.5 meters above the center of the seating area and marked as a noise anchor point. After expanding three times with the anchor point as the center and stepping 0.3 meters, the cluster of spheres contacts and merges with the adjacent cluster of spheres. After performing Boolean subtraction with the seat model, the interference noise cloud above the seating area is generated.

[0047] In this embodiment of the application, multiple unlicensed frequency band signal sources that are spatially adjacent and have a frequency overlap higher than a preset overlap threshold are grouped into interference clusters, and the geometric centroid of each interference cluster is calculated and marked as a noise anchor point. The method may also include the following steps: S2.21, Monitor the radio frequency signal strength of each unlicensed frequency band signal source at the receiving end, and estimate the local energy density parameters of each unlicensed frequency band signal source around the vehicle antenna based on the radio frequency signal strength and propagation distance model; S2.22, filter out high-energy interference sources whose local energy density parameters are higher than the preset density threshold, and mark high-energy interference sources that are spatially close as candidate collection objects; S2.23, calculate the frequency overlap between candidate aggregation objects, and when the frequency overlap is determined to be higher than the preset overlap threshold, merge the candidate aggregation objects into an interference cluster; S2.24, Based on the spatial coordinates of the candidate aggregation objects contained in the interference cluster, calculate the mass center coordinates of the interference cluster and determine the mass center coordinates as the noise anchor point.

[0048] Specifically, in S2.21, a stationary power measurement is performed on each unlicensed frequency band signal source at the receiving end. Multiple power values ​​are repeatedly sampled according to a preset measurement period. Abnormal samples whose deviation from adjacent sampled values ​​exceeds the jump threshold are removed, and the average value is taken to obtain the radio frequency signal strength of each unlicensed frequency band signal source at the receiving end. The preset measurement period is obtained based on the trade-off experience between platform dwell time and sampling stability, and the preferred setting range is 100 milliseconds to 500 milliseconds. In the specific operation of this application, the train door opening signal is read. If it is during the peak passenger boarding and alighting period, the preset measurement period is set to 100 milliseconds; if it is during the off-peak period, the preset measurement period is set to 500 milliseconds. The jump threshold is obtained by statistically analyzing the power jump amplitude caused by the start-up and shutdown of the hotspot device in the adjacent seat. The preferred setting range is 6dB to 10dB of the average measurement value. In the specific operation of this application, the average value of all sampled values ​​of this measurement is calculated. If the average value is more than 10dB higher than the normal received power, the jump threshold is set to 10dB of the average value; otherwise, it is set to 6dB.

[0049] It can be noted that in S2.21, the propagation distance model adopts the inverse square law of free space propagation of radio frequency signals, that is, the signal power decreases according to a square law as the propagation distance increases. Solve from the following formula: ; in, This refers to the propagation distance from an unlicensed frequency band signal source to the receiver. To control the wavelength corresponding to the operating frequency band of the signal. The equivalent transmission power is taken from the typical transmission level of hotspot equipment. The signal strength is measured at the receiver. The propagation distance is then calculated. Then, combining the spatial azimuth and elevation angles obtained from S2.1, the spatial coordinates of each signal source are located in a local three-dimensional Cartesian coordinate system with the phase center of the train antenna as the origin. The spatial coordinates of each signal source are then used as the center of a sphere, and the propagation distance is... The slant range determines the spatial location of the signal source.

[0050] Furthermore, in S2.21, a local space is delineated using the spatial location of each signal source as the center and a preset density evaluation radius, and the local energy density parameter is... Calculate using the following formula: ;

[0051] in, This refers to the local energy density parameters of unlicensed frequency band signal sources around the vehicle-mounted antenna. The radio frequency signal strength measured at the receiving end. For propagation distance, the local energy density parameter characterizes the interference power level per unit volume. The preset density assessment radius is obtained based on statistical analysis of typical distributions of signal source spacing within the carriage, with a preferred range of 0.5 meters to 1.5 meters. In the specific operation of this application, the average spacing between identified signal sources within the carriage is obtained. If the average spacing is less than 1 meter, the preset density assessment radius is set to 0.5 meters; if the average spacing is not less than 1 meter, the preset density assessment radius is set to 1 meter.

[0052] Meanwhile, in S2.22, the local energy density parameter is... Each signal source is compared with a preset density threshold. Signal sources exceeding the preset density threshold are marked as high-energy interference sources. The spatial distance between each pair of high-energy interference sources is then checked, and high-energy interference sources with a spatial distance less than a preset proximity distance are marked as candidate aggregation objects. The preset density threshold is obtained by statistically analyzing the median value between the carriage noise energy density and the near-field energy density of hotspot devices. The preferred setting range is 20dB to 30dB above the noise energy density. In the specific operation of this application, the noise measurement value of the current platform period is obtained. If the noise measurement value is more than 5dB higher than the normal noise level, the preset density threshold is increased by 5dB accordingly; otherwise, the normal value is maintained.

[0053] On the other hand, in S2.23, the intersection width of the occupied frequency bands is calculated for each pair of candidate aggregation objects. with union width ,according to Obtain frequency overlap ,when When the overlap exceeds a preset threshold, the corresponding candidate aggregation objects are merged into the same interference cluster. Merging is performed pairwise until the merging relationship between all candidate aggregation objects no longer changes. In S2.24, based on the spatial coordinates of the candidate aggregation objects contained in the interference cluster, the average value is calculated in the three coordinate axes to obtain the mass center coordinates of the interference cluster. The mass center coordinates are determined as noise anchor points, and the list of high-energy interference sources within the interference cluster to which the noise anchor point belongs is recorded and written into the cloud parameter table for use in the strategy execution phase of S4.5.

[0054] For example, the propagation distances of the three hotspot devices in the middle of the carriage, as determined by the propagation distance model, are 0.8 meters, 1.2 meters, and 1.5 meters, respectively. The calculated local energy densities are all higher than the preset density threshold, and the distance between each pair is less than the preset proximity distance. They are marked as candidate aggregation objects. The ratio of the intersection width to the union width of the frequency bands occupied by the three is 65%, which is higher than the preset overlap threshold. They are merged into an interference cluster, and the coordinates of the mass center are obtained by the average coordinate value and determined as the noise anchor point.

[0055] Step S3: Perform spatial geometric superposition of the virtual transmission cone and the interference noise cloud to determine the overlapping volume of the virtual transmission cone and the interference noise cloud, and mark the overlapping volume as the radio frequency conflict zone.

[0056] It should be noted that the purpose of S3's spatial geometric overlay is to spatially compare the energy distribution characterized by the interference noise cloud with the signal coverage area defined by the virtual transmission cone, and to find the three-dimensional region occupied by both simultaneously. Interference only actually threatens the vehicle-to-ground link when the interference energy falls within the transmission path of the control signal. Therefore, the overlay result of S3 directly determines the type determination and the scope of the policy action in S4.

[0057] Specifically, the superposition of S3 is performed in the local three-dimensional Cartesian coordinate system established by S1. The input is the virtual transmission cone parameters in the spatial parameter record and the sphere set parameters in the cloud parameter table. The output is the volume block that overlaps with the two, as well as the axial start and end range and volume size of the overlapping volume block within the cone. The output results are written into the conflict record. The conflict record serves as the input condition for the time dimension comparison of S3.2 to S3.3 and the region cutting and type labeling of S4, forming a complete data chain from spatial superposition to type partitioning and then to strategy execution.

[0058] Step S3.1: Superimpose the interference noise cloud and the virtual transmission cone in a three-dimensional spatial coordinate system, and use a three-dimensional geometric intersection algorithm to extract the overlapping volume block inside the virtual transmission cone that is eroded by the interference noise cloud, and mark the overlapping volume block as a spatial conflict sub-body.

[0059] Specifically, in S3.1, the interference noise cloud and the virtual transmission cone are superimposed in a three-dimensional spatial coordinate system. The cloud is represented by a set of spheres in the cloud parameter table, and the cone is represented by cross-sectional regions segmented along the axial direction in the spatial parameter record. Both types of models are in the same coordinate system with the phase center of the train antenna as the origin. After superposition, the cloud spheres and cone segments can be directly compared one by one in the coordinate system.

[0060] Furthermore, a polar coordinate sampling grid is established for the cross-sectional area of ​​each segment along the cone's axial direction. Checkpoints are placed with the center of the cross-section as the pole and at preset angular and radial intervals. For each checkpoint, the straight-line distance from the checkpoint to the center of each sphere in the cloud is calculated. When the straight-line distance is less than the radius of the corresponding sphere, the checkpoint is determined to be inside the cloud. The proportion of checkpoints inside the cloud within the cross-section is statistically analyzed. If the proportion exceeds a preset threshold, the overlapping area within the segment is fitted using the outer boundary of the checkpoints inside the cloud. The overlapping area is multiplied by the axial segment length to obtain the overlapping volume block eroded by the cloud within that segment. All volume blocks from each segment are then aggregated along the axial direction to obtain the complete overlapping volume block. The preset angular interval is obtained based on a trade-off between the cross-sectional radius and the intersection accuracy, with a preferred range of 3 to 10 degrees. In the specific operation of this application, the radius of the current segment's cross-section is read. If the radius is greater than 2 meters, the preset angular interval is set to 3 degrees; if the radius is not greater than 2 meters, the preset angular interval is set to 5 degrees. The preset percentage threshold is obtained by calculating the statistical value of the edge checkpoint misjudgment rate, and the preferred setting range is 10% to 20%. In the specific operation of this application, the estimated value of the direction error of the current scanning period is read. If the estimated value of the direction error is greater than 5 degrees, the preset percentage threshold is set to 20%; otherwise, it is set to 10%.

[0061] It can be explained that in S3.1, the aggregated overlapping volume blocks are marked as spatial conflict sub-bodies. Spatial conflict sub-bodies are geometric bodies defined in this application to refer to the overlapping spatial portion of the cloud and the cone. The start and end ranges and volume of the spatial conflict sub-bodies along the cone's axis are recorded, and the recording results are written into the conflict record for use in the time dimension comparison in S3.3 and the sub-region cutting in S4.1. For example, if the cloud contains two spheres, a polar coordinate grid is established meter by meter along the cone's axis to check each cross-section. Within the range of 8 meters to 15 meters, the proportion of internal checkpoints in each cross-section is higher than a preset proportion threshold. The overlapping area is aggregated into a spatial conflict sub-bodies with a volume of approximately 2 cubic meters, and the axial start and end range is recorded as 8 meters to 15 meters and written into the conflict record.

[0062] In some embodiments, spatial geometric superposition is performed on the virtual transmission cone and the interference noise cloud to determine the overlapping volume of the virtual transmission cone and the interference noise cloud, and the overlapping volume is marked as the radio frequency conflict region. This may also include the following steps: S3.2, Monitor the preamble of the channel signal and analyze the transmission period and duty cycle of the unlicensed frequency band signal source; S3.3 compares the transmission time of the control data packet to be sent by the train with the active time window of the unlicensed frequency band signal source. When it is determined that the transmission time is within the active time window and the spatial conflict sub-body is covered by the interference noise cloud, the radio frequency conflict zone is activated.

[0063] Specifically, in S3.2, the preamble of the signal in the train communication receiving channel is continuously monitored. The preamble is a known signal with a fixed pattern at the beginning of a data frame. The identification method is to compare the received signal segment with the preamble pattern stored locally point by point. When the matching ratio of point-by-point comparison is higher than a preset matching threshold, it is determined that a preamble has been captured, and the time of each preamble capture is recorded. The preset matching threshold is obtained based on the statistical analysis of the pattern distortion caused by channel noise, and the preferred setting range is 90% to 95%. In the specific operation of this application, the current channel bit error rate measurement value is obtained. If the bit error rate is higher than one-thousandth, the preset matching threshold is set to 90%; otherwise, it is set to 95%.

[0064] Furthermore, in S3.2, the interval between two adjacent preamble acquisition moments is taken as a transmission interval. The average of multiple transmission intervals within a preset observation period is used to obtain the transmission period of the interference source. Simultaneously, the duration of each frame of interference data from the preamble acquisition moment to the signal power falling back to the noise floor threshold is recorded. The ratio of the sum of all durations to the preset observation period is used as the duty cycle. The preset observation period is obtained empirically by covering multiple transmission periods to smooth random jitter, with a preferred setting range of 2 to 10 seconds. In the specific operation of this application, a statistically obtained transmission period estimate is obtained. If the transmission period estimate is less than 0.5 seconds, the preset observation period is set to 2 seconds; if the transmission period estimate is not less than 0.5 seconds, the preset observation period is set to 5 seconds.

[0065] It can be explained that in S3.3, the active time window of the unlicensed frequency band signal source is calculated based on the transmission period and duty cycle. The active time window refers to the period during which the interference signal is in the transmission state within each transmission period. The starting point of the window is calculated by recursively pushing forward according to the transmission period based on the latest preamble acquisition time. The window width is obtained by multiplying the duty cycle by the transmission period. All windows are arranged in chronological order to form a window sequence.

[0066] Furthermore, in S3.3, the transmission time of the control data packet to be sent by the train is numerically compared with each active time window in the window sequence. When the transmission time falls within any active time window, it is determined that the transmission time is in the period of active interference, and the comparison result is transmitted with the conflict record.

[0067] Meanwhile, in S3.3, based on the fact that the transmission time falls within the active time window, the spatial conditions are further verified. The verification method is to extract the center point of the spatial conflict sub-body in the conflict record, calculate the straight-line distance from the center point to the center of each sphere in the cloud, and determine that the spatial conflict sub-body is covered by the interference noise cloud when the straight-line distance from the center point to the center of any sphere is less than the radius of the corresponding sphere.

[0068] On the other hand, in S3.3, when both conditions are met—the transmission time being within the active time window and the spatial conflict sub-body being covered by the interference noise cloud—the radio frequency conflict zone is activated. The radio frequency conflict zone marker in the conflict record is set to valid, and the activation time and the axial start and end range of the spatial conflict sub-body are written into the conflict record for sub-region segmentation in S4.1 and strategy execution in S4.5. For example, if the matching ratio of the preamble captured by the monitoring is 93%, which is higher than the preset matching threshold, the transmission period of the interference source is 40 milliseconds and the duty cycle is 30%, the calculated active time window width is 12 milliseconds, the moment when the train is about to send the control data packet falls within the next window, and the straight-line distance from the center point of the spatial conflict sub-body to the center of the cloud sphere is less than the radius of the sphere, both conditions are met simultaneously, the radio frequency conflict zone is activated, and written into the conflict record.

[0069] Step S4: Based on the spatial location of the radio frequency conflict zone in the virtual transmission cone, determine the target interference conflict type, and execute the intelligent networking anti-interference strategy based on the target interference conflict type.

[0070] It should be noted that the type of interference in S4 depends on the spatial location of the RF conflict zone within the virtual transmission cone. Different locations mean different coupling strengths between the interference source and the vehicle-mounted antenna, and consequently, different countermeasures can be taken. Interference close to the antenna is so strong that simply changing the transmission parameters is ineffective; it is necessary to address the shape of the radiation pattern. Interference near the ground access point manifests more as a decrease in the signal-to-noise ratio at the receiver, which can be remedied at the receiver and coding levels.

[0071] Specifically, the execution process of S4 is divided into three stages: spatial division, region comparison, and policy execution. Spatial division is performed by S4.1, which divides the virtual transmission cone into three sub-regions according to three cutting planes with different physical locations. Region comparison is performed by S4.2 to S4.4, which checks the overlap between the radio frequency conflict area and each sub-region and gives a type label. Policy execution is performed by S4.5 to S4.13, which selects the processing branch according to the type label.

[0072] Furthermore, for situations where the same spatial conflict sub-body is marked with multiple target interference conflict types, the handling sequence is executed according to the following rules: When the type marking simultaneously includes near-field and mid-to-far-field conflict types, the near-field handling branches S4.5 to S4.7 are executed first to reshape the radiation pattern. After the pattern reshaping is completed, the superposition in S3.1 is executed again. If the overlapping volume block still exists and falls within the mid-to-far-field attenuation region, the mid-to-far-field handling branches S4.8 to S4.10 are executed next. When the type marking includes far-field conflict types, the far-field handling branches S4.11 to S4.13 are executed at the ground access end, acting on the antenna end and the channel end respectively, along with the near-field and mid-to-far-field branches. The three do not conflict with each other and can be combined according to the marking results. The above sequence rules ensure that spatial coupling is eliminated first and then time-frequency collisions are eliminated, avoiding mutual cancellation of handling actions.

[0073] Preferably, determining the type of target interference conflict based on the spatial location of the radio frequency conflict zone in the virtual transmission cone may further include the following steps: S4.1, using the physical outer edge of the train antenna as the first cutting surface, the near-field boundary as the second cutting surface, and the effective receiving range boundary of the ground access point antenna as the third cutting surface, the virtual transmission cone is divided into a near-field flooding area, a mid-to-far-field attenuation area, and a receiver-sensitive area. S4.2 Determine whether the radio frequency conflict zone and the near-field flooding zone overlap. If they overlap, mark the target interference conflict type as the near-field conflict type. S4.3 Determine whether the radio frequency conflict zone and the mid-to-far field attenuation zone overlap. If they overlap, mark the target interference conflict type as the mid-to-far field conflict type. S4.4 Determine whether the radio frequency conflict area and the receiver's sensitive area overlap. If they overlap, mark the target interference conflict type as the far-field conflict type.

[0074] Specifically, in S4.1, the terms "first," "second," and "third" in the first, second, and third cutting surfaces are merely identifiers to distinguish three cutting surfaces with different physical locations. The first cutting surface is the curved surface where the physical outer edge of the train antenna is located, used to peel off the extremely close region attached to the antenna body. The second cutting surface is the spherical surface where the near-field boundary is located, and the near-field boundary is at a distance... Determine using the following formula: ;

[0075] in, The distance value is taken from the near-field boundary. This refers to the aperture size of the train antenna array. To control the wavelength corresponding to the signal operating frequency band, the area within the near-field boundary is the non-far-field of the antenna field. Interference enters the coupling range in this area without spatial diffusion. The third cutting surface is the curved surface where the effective receiving range boundary of the ground access point antenna is located. The effective receiving range boundary is calculated based on the receiving sensitivity of the ground access point and the preset link margin. That is, the curved surface formed by connecting the positions where the arriving power equals the receiving sensitivity plus the preset link margin with the access point as the center. The preset link margin is obtained by statistically analyzing the fading fluctuations of the vehicle-to-ground link, and the preferred setting range is 8dB to 15dB. In the specific operation of this application, the historical link measurement records of the station section are read. If the fading depth in the record exceeds 12dB, the preset link margin is set to 15dB; otherwise, it is set to 8dB.

[0076] Furthermore, in S4.1, the virtual transmission cone is spatially divided into three sections along its axial direction using three cutting surfaces. The method of execution is to calculate the intersection points of the three cutting surfaces with the cone's axis, and then arrange them sequentially along the axial direction according to the intersection points to obtain three sub-regions arranged sequentially. The section near the train antenna is the near-field flooding region, the middle section is the mid-to-far-field attenuation region, and the section near the ground access point is the receiver sensitive region. The near-field flooding region, the mid-to-far-field attenuation region, and the receiver sensitive region are the names of the three spatial segments defined in this application. The axial range of the three sub-regions, together with the total axial length of the cone, is written into the spatial parameter record.

[0077] It can be explained that the distance between the interference source and the antenna in the near-field flooding zone is extremely close. The interference signal enters the antenna coupling range without spatial attenuation, creating a flooding situation for the entire receiving link. Simply increasing the transmission power cannot penetrate this near-field shielding. The control signal in the sensitive area of ​​the receiver has already undergone full attenuation, and the remaining energy is limited. Co-channel interference superimposed in this area will directly reduce the signal-to-noise ratio at the ground access point. The difference in the physical properties of the two types of areas determines the difference in the focus of subsequent treatment.

[0078] Furthermore, in S4.2 to S4.4, the method for determining whether there is overlap is to extract the axial start and end range of the spatial conflict sub-body from the conflict record, compare the axial range of each of the three sub-regions with it one by one, and determine that there is overlap when two ranges have overlapping segments and the ratio of the length of the overlapping segment to the total axial length of the spatial conflict sub-body is higher than the preset overlap ratio threshold. All three types are checked one by one without mutual exclusion processing. When there is overlap with the near-field flooding area, the target interference conflict type is marked as near-field conflict type. When there is overlap with the mid-far-field attenuation area, it is marked as mid-far-field conflict type. When there is overlap with the receiver sensitive area, it is marked as far-field conflict type. Different axial segments of the same conflict sub-body can fall into different sub-regions and correspond to multiple type markings. The preset overlap ratio threshold is obtained by statistical analysis based on the relationship between the axial segment calculation error and the segment step length, with a preferred setting range of 5% to 15%. In the specific operation of this application, the axial segment step length of the cone is read. If the segment step length is 0.2 meters, the preset overlap ratio threshold is set to 10%. If the segment step length is 0.5 meters, the preset overlap ratio threshold is set to 5%.

[0079] Meanwhile, in S4.2 to S4.4, the division results of the three sub-regions and all type tags are written into the conflict record. The type tags and the processing order rules specified in S4 together serve as the input conditions for branch matching in S4.5 to S4.13. For example, the total axial length of the cone is 95 meters, and the antenna aperture size... The distance is 0.4 meters, and the corresponding wavelength in the operating frequency band is... The distance to the near-field boundary is 0.125 meters. according to The solution is approximately 2.56 meters. The effective receiving range boundary of the access point is located at 80 meters. The axial range of the spatial conflict sub-body is from the 6th to the 15th meter. The overlap ratios with the near-field flooding zone and the mid-to-far-field attenuation zone are both higher than the preset overlap ratio threshold. The target interference conflict type is simultaneously marked as near-field conflict type and mid-to-far-field conflict type. According to the processing order rule of S4, the near-field processing branch is executed first and then the superposition calculation is reviewed.

[0080] In this application, the intelligent networking anti-interference strategy based on the target interference conflict type may further include the following steps: S4.5, in response to the target interference conflict type being near-field conflict type, obtain the interference cluster anchor point corresponding to the radio frequency conflict zone, and calculate the interference azimuth angle and interference elevation angle of the interference cluster anchor point relative to the train antenna. S4.6, call the antenna array control interface to adjust the weighting coefficients of each radiating element, and artificially create a gain null in the direction of the interference azimuth and elevation angles in the antenna radiation pattern through the phase cancellation principle. S4.7 compresses the main lobe width of the virtual transmission cone, concentrating the control signal energy in the direction of the ground access point, thus forming a signal transmission corridor that avoids radio frequency conflict zones.

[0081] Specifically, in S4.5, the spatial conflict sub-body written when activating the radio frequency conflict zone is read from the conflict record, the interference cluster to which the cloud sphere covering the spatial conflict sub-body belongs is extracted from the cloud parameter table, and the noise anchor point of the interference cluster is read as the interference cluster anchor point. The interference cluster anchor point is a reference point defined in this application to represent the position of the interference cluster corresponding to the radio frequency conflict zone. The coordinates of the interference cluster anchor point and the phase center coordinates of the train antenna are in the same local three-dimensional rectangular coordinate system, which can directly participate in the angle calculation.

[0082] Furthermore, in S4.5, the coordinate components of the interference cluster anchor point in the three coordinate axes are denoted as... , , The coordinate components of the phase center of the train antenna are denoted as... , , Interference azimuth With interference pitch angle Calculate according to the following formulas: ; ;

[0083] in, The interference azimuth angle is the angle of the interference cluster anchor point relative to the train antenna. To interfere with the pitch angle, , , The coordinate components of the anchor point of the interference cluster in the three coordinate axes. , , The coordinate components of the phase center of the train antenna in the three coordinate axes are represented by two angles, which together uniquely indicate the direction of the incoming wave from the interference source relative to the train antenna.

[0084] It can be explained that in S4.6, the physical meaning of gain null is that the signals emitted by each radiating element in the antenna array are synthesized in space according to the superposition law of waves. When two signal components reach the same spatial point, they are out of phase and cancel each other out, and the combined field strength approaches zero. By setting an appropriate phase excitation amount for each radiating element, the combined gain of the array in a specified direction can be reduced to an extremely low level, which is represented as a notch on the radiation pattern. After aligning the notch with the direction of the incoming interference wave, the radiated energy of the train antenna toward the direction of interference is greatly reduced, and the coupling path in the direction of interference is actively cut off. This method does not rely on increasing the transmission power, and therefore can cope with strong near-field interference close to the antenna.

[0085] Furthermore, in S4.6, the antenna array control interface is invoked, and the amplitude coefficient and phase compensation amount of each radiating element are written to the interface. The amplitude coefficient maintains the in-phase superposition relationship of each radiating element in the main lobe direction, and the phase compensation amount... Solve using the following formula: ;

[0086] in, For the first Phase compensation amount of each radiating element This refers to the sequence number of the radiating element in the antenna array. For the first The propagation path difference between each radiating element and the reference element to the anchor point of the interference cluster is used. The reference element is the radiating element in the antenna array with the smallest straight-line distance from the anchor point of the interference cluster. To control the wavelength corresponding to the operating frequency band of the signal, the array radiation pattern is written after the signal is written, i.e., at the interference azimuth angle. With interference pitch angle A gain null trap is formed in the direction.

[0087] Meanwhile, in S4.7, the specific method for compressing the main lobe width is to increase the attenuation amplitude of each radiating element's amplitude coefficient along the array edge. The attenuation amplitude is determined by a cosine law based on the distance from each radiating element to the array center. The greater the decrease in amplitude coefficient of the edge elements, the larger the main lobe half-angle. The smaller the width of the main lobe, the more concentrated the control signal energy becomes within the narrow beam pointing towards the ground access point. The gap between the direction of the interference cluster anchor point and the direction of the main lobe forms a signal transmission corridor. This signal transmission corridor, as defined in this application, refers to the angular domain range between the main lobe and the null that is free from interference energy coverage. After compression, the narrowed main lobe half-angle is written back to the spatial parameter record to ensure that the virtual transmission cone model is consistent with the actual beam, thus avoiding distortion of the cone parameters used subsequently.

[0088] On the other hand, in S4.7, after narrowing, the array radiation pattern is reread, and the angular interval between any direction in the signal transmission corridor and the direction of the interference cluster anchor point is checked. When the minimum angular interval is higher than the preset isolation angle threshold, it is confirmed that the signal transmission corridor avoids the radio frequency conflict zone. The preset isolation angle threshold is obtained by statistically analyzing the sum of the null width and the main lobe width, and the preferred setting range is 10 degrees to 20 degrees. In the specific operation of this application, the current null width is read. If the null width is greater than 10 degrees, the preset isolation angle threshold is set to 20 degrees; if the null width is not greater than 10 degrees, the preset isolation angle threshold is set to 10 degrees.

[0089] For example, the interference cluster anchor point is located at an azimuth angle of 110 degrees and an elevation angle of -15 degrees for the train antenna, obtained by calculating the coordinate difference and arctangent. 110 degrees After setting the value to -15 degrees, write the command to the antenna array control interface. The solved phase compensation values ​​for each radiating element result in a notch in the radiation pattern at a depth exceeding 25 dB in the 110-degree direction. Simultaneously, the amplitude of the edge elements decays according to a cosine law, causing the main lobe half-angle to... The angle between the main lobe and the null is narrowed from 6 degrees to 4 degrees and the spatial parameter record is written back. The minimum angular interval between the main lobe and the null is 18 degrees, which is higher than the preset isolation angle threshold, and the signal transmission corridor is established.

[0090] In some embodiments, the specific operation of adjusting the weighting coefficients of each radiating element in S4.6 to create a gain null can be as follows: A1. Obtain the three-dimensional coordinates of the interference cluster anchor point, and retrieve the phase center coordinates of the train antenna and the position coordinates of each radiating element in the antenna array, then proceed to step A2.

[0091] A2. Based on the three-dimensional coordinates of the interference cluster anchor point and the position coordinates of each radiating unit, calculate the straight-line distance from each radiating unit to the interference cluster anchor point to obtain the interference propagation distance corresponding to each radiating unit, and then proceed to step A3.

[0092] A3. Select the radiating element with the smallest straight-line distance from the anchor point of the interference cluster in the antenna array as the reference element. Calculate the difference in interference propagation distance between each of the remaining radiating elements and the reference element to obtain the difference in interference propagation distance between each radiating element and the reference element. Proceed to step A4.

[0093] A4, each The integer wavelength portion is removed, leaving only the remaining distance less than one wavelength, and so on. Calculate the phase compensation for each radiating element so that the signal components emitted by each radiating element cancel each other out in the directions of interference azimuth and interference elevation, and proceed to step A5.

[0094] A5, superimpose the corresponding phase compensation amount on the phase excitation distribution of the original beamforming configuration parameters for the emission phase of each radiation element, and proceed to step A6.

[0095] A6. Determine whether the combined gain in the interference azimuth and pitch directions of the radiation pattern after phase compensation is lower than the preset null depth threshold. If yes, proceed to step A7; otherwise, proceed to step A8. The preset null depth threshold is obtained by statistically analyzing the gain reduction that still needs to be countered after near-field coupling attenuation of the typical transmit power of the passenger hotspot device. The preferred setting range is 20dB to 35dB lower than the main lobe gain. In the specific operation of this application, the interference signal strength measured at the interference cluster anchor point is obtained. If the interference signal strength is more than 10dB higher than the conventional hotspot power, the preset null depth threshold is set to 35dB; otherwise, it is set to 20dB.

[0096] A7: Determine the current phase compensation amount as the final weighting coefficient, write it into the antenna array control interface, and end the process.

[0097] Step A8 involves successively adjusting the phase compensation of each radiating element in both positive and negative directions with a preset fine-tuning step size. After adjustment, the process returns to step A6 for re-evaluation until the synthesized gain meets the preset null depth threshold. The final phase compensation amount is then determined as a weighting coefficient and written to the antenna array control interface, ending the process. The preset fine-tuning step size is obtained based on an empirical trade-off between phase compensation calculation accuracy and convergence speed, with a preferred range of 1 to 5 degrees. In the specific operation of this application, the wavelength corresponding to the working frequency band is read. If the wavelength is less than 10 cm, the preset fine-tuning step size is set to 1 degree to improve compensation accuracy; if the wavelength is not less than 10 cm, the preset fine-tuning step size is set to 3 degrees.

[0098] In some preferred embodiments, the intelligent networking anti-interference strategy based on the target interference conflict type may further include the following steps: S4.8, in response to the target interference conflict type being a mid-to-far field conflict type, extracts the time-domain and frequency-domain characteristics of the interference signal inside the radio frequency conflict zone, and identifies the center frequency point and transmission period of the interference signal; S4.9 performs time offset processing on the control data packets to be sent by the train, delaying the transmission time of the control data packets to the silent interval within the transmission cycle; S4.10, or perform frequency switching processing on the control data packets to be sent by the train, switching the transmission frequency band of the control data packets to a spare frequency band that does not overlap with the center frequency point.

[0099] Specifically, in S4.8, the interference cluster corresponding to the spatial conflict sub-body is read from the conflict record, the received power distribution sequence of the signal source in the interference cluster recorded in S2.1 is retrieved, the frequency point with the highest power is determined as the center frequency point, and the transmission period, duty cycle and active time window recorded in S3.2 and S3.3 are retrieved as time domain features. The time domain features and frequency domain features together constitute a complete profile of the interference signal.

[0100] Furthermore, in S4.9, the remaining time period outside the active time window within each transmission cycle is the silent gap. The silent gap is defined in this application as the time period during which the interference signal is in a silent state within the transmission cycle. The first silent gap after the current moment is found by recursively counting backwards from the latest preamble capture time, and the starting time and width of the silent gap are recorded.

[0101] It can be explained that in S4.9, a time limit check is performed before time offset processing. The control data packet carries a preset transmission time limit, which refers to the latest time that the control data packet must be transmitted as specified by the vehicle control logic. The start time of the silent gap is compared with the preset transmission time limit. Time offset processing is selected only when the start time of the silent gap is earlier than the preset transmission time limit and the width of the silent gap is greater than the transmission time required for the control data packet. The transmission time required is calculated based on the number of bytes in the control data packet and the number of bits carried by each symbol under the current modulation method. The preset transmission time limit is obtained by an engineering convention based on an integer fraction of the vehicle control cycle, with a preferred setting range of 50 milliseconds to 200 milliseconds. In the specific operation of this application, the configuration value of the vehicle control cycle is read. If the current control data packet is an emergency control command, the preset transmission time limit is set to half of the control cycle and not greater than 50 milliseconds. If it is a status control command, the preset transmission time limit is set to one-quarter of the control cycle and not greater than 200 milliseconds. When the start time of the silent gap is later than the preset transmission time limit or the width of the silent gap is insufficient, the time offset processing is abandoned and the frequency jump processing of S4.10 is rolled back.

[0102] Furthermore, in S4.10, the selection of backup frequency bands involves checking each candidate band in a pre-defined set. Several backup frequency bands are pre-defined within the train communication frequency band neighborhood. The frequency interval between each backup frequency band and the center frequency of the interference signal is checked one by one. Simultaneously, the noise floor power obtained from the S2.1 scan within each backup frequency band is checked. When the frequency interval is higher than a preset frequency interval threshold and the noise floor power is lower than a preset noise floor threshold, the corresponding backup frequency band is determined as the target switching frequency band. The preset frequency interval threshold is obtained based on statistics of the receiving equipment's ability to suppress adjacent channel interference, with a preferred setting range of 10MHz to 25MHz. In the specific operation of this application, the adjacent channel suppression parameters of the on-board communication module are read. If the adjacent channel suppression parameters are lower than 40dB, the preset frequency interval threshold is set to 25MHz; if the adjacent channel suppression parameters are not lower than 40dB, the preset frequency interval threshold is set to 10MHz. The preset noise floor threshold is obtained by calculating based on the demodulation margin reserved for the receiving sensitivity, and the preferred setting range is 3dB to 6dB below the receiving sensitivity. In the specific operation of this application, the receiving sensitivity of the ground access point on the spare frequency band is read. If there is no historical interference record in the spare frequency band, the preset noise floor threshold is set to 3dB below the receiving sensitivity. If there is a historical interference record, the preset noise floor threshold is set to 6dB below the receiving sensitivity.

[0103] Meanwhile, in S4.10, the specific method of frequency hopping processing is to switch the transmission frequency band of the control data packet from the current frequency band to the target switching frequency band. The switching is completed within the silence interval. The switching process includes sending a frequency switching command to the RF front end, waiting for the local oscillator frequency to stabilize, and sending the control data packet on the target switching frequency band. The stabilization of the local oscillator frequency is determined based on the ready status bit returned by the RF front end. After the switching is completed, the original frequency band no longer carries control data packets.

[0104] On the other hand, in S4.9 and S4.10, time offset processing and frequency hopping processing are performed one or the other according to the processing order rules specified in S4. When the time limit verification of S4.9 passes, time offset processing is performed first to avoid the link re-establishment overhead introduced by changing the frequency band. For example, if the occupied bandwidth corresponding to the center frequency point of the interference signal is 20MHz, the transmission period is 40ms, the duty cycle is 30%, the silence gap width is 28ms, the time required for controlling the transmission of data packets is 8ms and the preset transmission time limit is 100ms, and the start of the next silence gap is 35ms from the current time, which is earlier than the preset transmission time limit, the time limit verification passes, and the transmission time is postponed to the start of the next silence gap to perform time offset processing. If the start of the silence gap is 150ms from the current time, which is later than the preset transmission time limit, the transmission frequency band is switched to a spare frequency band 25MHz away from the center frequency point to perform frequency hopping processing.

[0105] In this embodiment of the application, the intelligent networking anti-interference strategy based on the target interference conflict type may further include the following steps: S4.11, In response to the target interference conflict type being far-field conflict type, activate the diversity reception logic of the ground access end, utilize multiple spatially separated receiving antennas to receive control signals, and select the one with the optimal signal-to-noise ratio as the input signal; S4.12, add redundant check bytes to the control data packets to be sent by the train, and downgrade the modulation mode of the control data packets to a low-order modulation mode; S4.13, or change the transmission polarization mode of the control signal, using polarization isolation to counteract co-channel interference.

[0106] Specifically, in S4.11, the diversity reception logic of the ground access terminal is activated. Multiple receiving antennas at the ground access terminal are spaced apart by a preset distance at their installation positions. The same control signal reaches each receiving antenna via different spatial paths, and the interference fading state received by each antenna is independent of each other, thus providing a physical basis for optimal selection. The preset distance is obtained based on empirical values ​​of the minimum interval for vehicle-to-ground link fading decorrelation, with a preferred setting range of 2 to 5 meters. In the specific operation of this application, each receiving antenna is deployed according to the installation interval of the platform canopy columns. If the column interval is not less than 3 meters, the preset distance is set to 3 meters; if the column interval is less than 3 meters, the preset distance is set to 2 meters.

[0107] Furthermore, in S4.11, the signal-to-noise ratio is calculated for the signal output from each receiving antenna. Signal-to-noise ratio Calculate using the following formula: ;

[0108] in, The signal-to-noise ratio of a single received signal. This refers to the received power of the pilot band in the received signal path. The pilot band refers to a pre-defined signal segment with a known pattern within the control data packet. The signal-to-noise ratios of each channel are compared based on the noise power measured in the received signal within the same time period. The signal-to-noise ratio (SNR) of the highest channel is selected as the input signal and sent to the demodulation channel, while the other channels are either disabled or put into continuous monitoring mode.

[0109] It can be explained that in S4.12, the method of adding redundant check bytes to the control data packets to be sent by the train is as follows: all bytes in the control data packet are performed bit by bit in a modulo-2 operation to obtain a check sequence. Modulo-2 operation refers to the operation of adding bits and taking the remainder as 0 or 1. The check sequence is appended to the end of the control data packet and sent with the packet. After receiving it, the ground access end recalculates the check sequence of all bytes in the packet according to the same rule and compares it bit by bit with the appended check sequence. If the two are inconsistent, it is determined that the data in the packet is damaged and a retransmission is requested. The redundant check bytes thus provide a basis for the detection and retransmission of damaged data packets. Cyclic redundancy check is an existing error detection method in the field of communication. This application embeds it into the far-field collision type handling branch and performs it in conjunction with reduced-order modulation. The method for obtaining the preset check bit length is based on the trade-off experience between the number of bytes in the control data packet and the required probability of missed detection. The preferred setting range is 8 bits to 16 bits. In the specific operation of this application, the number of bytes in the control data packet is read. If the number of bytes is less than 64 bytes, the preset check bit length is set to 8 bits. If the number of bytes is not less than 64 bytes, the preset check bit length is set to 16 bits.

[0110] Furthermore, in S4.12, the modulation method of the control data packet is downgraded to a low-order modulation method, that is, the number of bits carried by each modulation symbol is reduced. For example, the high-order method with 6 bits per symbol is reduced to the low-order method with 2 bits per symbol. After the number of bits is reduced, the amplitude difference and phase difference between adjacent modulation states increase accordingly. The judgment margin of the receiver in distinguishing adjacent states under the same noise conditions is correspondingly expanded, and the probability of misjudgment decreases.

[0111] Meanwhile, in S4.13, the transmit polarization of the control signal is changed, and the antenna feed port is switched to a polarization mode orthogonal to the current polarization, such as switching from vertical polarization to horizontal polarization, or from linear polarization to circular polarization. The difference in polarization state between the co-channel interference source and the control signal constitutes the polarization isolation. Polarization isolation refers to the difference in response of the receiving antenna to two orthogonal polarization signals. The higher the polarization isolation, the lower the residual energy of the interference signal after polarization matching by the receiving antenna, and the higher the degree of separation between the control signal and the interference signal.

[0112] On the other hand, in S4.11 to S4.13, diversity reception, redundancy check, and down-order modulation are superimposed according to the processing order rules specified in S4. Polarization change and the aforementioned processing can be performed individually or in combination. The common purpose of all processing is to counteract co-channel interference at the end of the transmission path with margin compensation at the receiver and link layers in far-field collision scenarios. For example, four receiving antennas at the ground access end are installed 3 meters apart, and each channel is configured according to... The measured signal-to-noise ratios were 12dB, 18dB, 9dB, and 15dB, respectively. The channel with the highest signal-to-noise ratio was selected as the input demodulation channel. The control data packet was sent after being appended with a 16-bit check sequence. The ground access end recalculated the check sequence and compared it bit by bit to confirm that the received packet was correct. The modulation method was downgraded from 6 bits per symbol to 2 bits per symbol. At the same time, the transmit polarization was switched from vertical polarization to horizontal polarization. After the co-channel interference was attenuated by polarization mismatch, the demodulation accuracy of the ground access end recovered.

[0113] Figure 2 The process of constructing the virtual transmission cone begins with establishing a coordinate system with the train antenna phase center as the origin, and generating the main transmission vector based on the ground access point location. After generating a conical geometric model according to beamforming configuration parameters, boundary truncation is performed segment by segment along the axial direction. When the physiological geometric boundary is reached, energy attenuation reaches a threshold, or line-of-sight is limited, the model extension is truncated, generating the virtual transmission cone. This establishes the energy coverage benchmark for the control signal in three-dimensional space, after which the process ends and proceeds to the next step. Figure 3 The process of generating interference noise clouds is shown.

[0114] Figure 3 To disrupt the logic of generating noise clouds, the following is implemented: Figure 2 The spatial location of unlicensed frequency band signal sources is obtained through spectrum scanning, and the local energy density is estimated. Signal sources that meet the energy threshold and are spatially adjacent are grouped into interference clusters, and their geometric centroids are calculated as noise anchor points. Subsequently, a spherical expansion scan is performed with the anchor points as the sphere centers. Overlapping spheres are merged and the solid structures of the carriage are removed to accurately characterize the radio frequency energy distribution pattern generated by passenger hotspot devices. The process ends after generating interference noise clouds and then proceeds to the next step. Figure 4 The spatial geometry overlay process is shown.

[0115] Figure 4 For spatial geometric overlay and conflict zone activation mechanism, it undertakes Figure 3 The interference noise cloud and the virtual transmission cone are superimposed in a three-dimensional coordinate system, and the overlapping volume block is extracted as the spatial conflict sub-body. The transmission period and duty cycle of the interference source are analyzed based on the channel preamble monitoring results. When the train data packet transmission time falls within the interference active window, and the spatial conflict sub-body is covered by the cloud, the radio frequency conflict zone is activated. Figure 4 The process is complete, and proceeds to... Figure 5 The conflict type determination process shown provides triggering conditions for subsequent differentiated anti-interference strategies; when the above conditions are not met or the radio frequency conflict zone is not activated, the process proceeds to... Figure 9 The process shown executes the standard communication sending procedure and ultimately merges into... Figure 9 The execution endpoint is shown.

[0116] Figure 5 To determine the area of ​​interference conflict type, the following rules are adopted. Figure 4Using the antenna's physical outer edge, near-field boundary, and effective receiving range boundary of the access point, the virtual transmission cone is divided into a near-field flooding region, a mid-to-far-field attenuation region, and a receiver-sensitive region. By comparing the axial range of the spatial conflict sub-regions with the overlap of each sub-region, the interference conflict type is marked: those overlapping with the near-field flooding region are marked as near-field conflict types, and the process proceeds to... Figure 6 The near-field anti-interference handling procedure is shown; the markings that overlap with the mid-to-far-field attenuation region indicate mid-to-far-field conflict types, and proceed to... Figure 7 The following is a mid-to-far field anti-interference handling procedure; areas overlapping with the receiver's sensitive area are marked as far-field conflict types, and the process proceeds to... Figure 8 The far-field anti-interference processing procedure shown enables accurate classification of the spatial location of interference sources.

[0117] Figure 6 The process for handling interference in near-field collisions is detailed, and it undertakes... Figure 5 The branch line deals with near-field interference types. It acquires the anchor points of the interference clusters and calculates the interference azimuth and elevation angles. By adjusting the weighting coefficients of each radiating element in the antenna array, a gain null is artificially created in the direction of the incoming interference wave. Simultaneously, the main lobe width is compressed, concentrating the control signal energy in the signal transmission corridor pointing towards the ground access point, effectively cutting off the coupling path of strong near-field interference. After near-field anti-interference processing is completed, it merges into the network. Figure 9 The execution endpoint is shown.

[0118] Figure 7 For time-frequency domain avoidance strategies of mid-to-far field collision types, it inherits... Figure 5 For mid-to-far field interference types, extract the center frequency and transmission period of the interference signal, and verify the duration and width of the silence gap. If the conditions are met, delay the data packet transmission time to the silence gap; otherwise, perform a frequency hopping, switching the transmission frequency band to a non-overlapping backup frequency band, thereby reducing the probability of signal collisions on the transmission path. After mid-to-far field anti-interference processing is completed, merge into the network. Figure 9 The execution endpoint is shown.

[0119] Figure 8 For link margin compensation mechanisms in the case of far-field collisions, it undertakes... Figure 5 Mid-to-far field interference types. Activate the diversity reception logic at the ground access end, selecting the input signal with the highest signal-to-noise ratio. Simultaneously, add redundant check bytes to the control data packets and downgrade to a lower-order modulation scheme, or change the transmit polarization to utilize polarization isolation to counteract co-channel interference, ensuring demodulation reliability at the end of the transmission path. After far-field anti-interference processing is completed, the signal is merged into the network. Figure 9 The execution endpoint is shown.

[0120] Figure 9 As the execution endpoint of each differentiated anti-interference strategy, it respectively undertakes Figure 6 The corresponding near-field beam reshaping, Figure 7Corresponding mid- and far-field time-frequency shift, Figure 8 Corresponding far-field link compensation and Figure 4 The corresponding normal communication process that does not trigger the conflict zone has four inputs, all of which complete state convergence here, marking the closed loop of intelligent networking anti-interference handling for the current communication cycle.

[0121] Example 2: This example also provides an intelligent networking communication system for rail trains, including: a transmission cone construction module, used to obtain the train stopping position and the ground access point position, and construct a virtual transmission cone based on the train stopping position and the ground access point position, the virtual transmission cone representing the electromagnetic wave coverage area of ​​the train antenna sending control signals to the ground access point; The interference cloud generation module is used to perform spectrum scanning inside the carriage and around the train to obtain spatial orientation information of the interference source. Based on the spatial orientation information of the interference source, it generates an interference noise cloud in three-dimensional space. The interference noise cloud represents the energy distribution pattern of radio frequency signals generated by passenger hotspot devices. The conflict zone determination module is used to perform spatial geometric superposition of the virtual transmission cone and the interference noise cloud, determine the overlapping volume of the virtual transmission cone and the interference noise cloud, and mark the overlapping volume as the radio frequency conflict zone. The strategy execution module is used to determine the target interference conflict type based on the spatial location of the radio frequency conflict zone in the virtual transmission cone, and to execute intelligent networking anti-interference strategies based on the target interference conflict type.

[0122] The above-mentioned unit modules can be embedded in the processor of the electronic device in hardware form or independent of it, or they can be stored in the memory of the electronic device in software form, so that the processor can call and execute the corresponding operations of the above modules.

[0123] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0124] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A smart networking communication method for rail trains, characterized in that, include: The train stopping position and the ground access point position are obtained. A virtual transmission cone is constructed based on the train stopping position and the ground access point position. The virtual transmission cone represents the electromagnetic wave coverage area of ​​the train antenna sending control signals to the ground access point. Spectrum scanning is performed inside the carriage and around the train to obtain spatial location information of interference sources. Based on the spatial location information of interference sources, interference noise cloud is generated in three-dimensional space. The interference noise cloud represents the energy distribution pattern of radio frequency signals generated by passenger hotspot devices. Spatial geometric superposition is performed on the virtual transmission cone and the interference noise cloud to determine the overlapping volume of the virtual transmission cone and the interference noise cloud, and the overlapping volume is marked as the radio frequency conflict zone; Based on the spatial location of the radio frequency conflict zone in the virtual transmission cone, the target interference conflict type is determined, and an intelligent networking anti-interference strategy is executed based on the target interference conflict type.

2. The intelligent networking communication method for rail trains as described in claim 1, characterized in that: The construction of the virtual transmission cone based on the train's stopping position and the ground access point's position includes: Using the phase center of the train antenna as the origin and the antenna center of the ground access point as the target point, a main transmission vector is constructed connecting the origin and the target point. The beamforming configuration parameters built into the vehicle-mounted antenna are retrieved, and a conical geometric model extending outward along the main transmission vector is generated based on the direction parameters of the main transmission vector and the phase excitation distribution in the beamforming configuration parameters. A boundary truncation determination is performed on the conical geometric model. When any one of the physiological geometric truncation condition, energy threshold truncation condition, or line-of-sight truncation condition is triggered, the extension of the conical geometric model is trunculated to generate the virtual transmission cone.

3. The intelligent networking communication method for rail trains as described in claim 2, characterized in that: The step of performing a spectrum scan of the interior of the carriage and the surrounding area of ​​the train to obtain spatial location information of the interference source, and generating an interference noise cloud in three-dimensional space based on the spatial location information of the interference source, includes: Unlicensed frequency band signal sources inside the carriage are identified by spectrum scanning, and the spatial azimuth and elevation angles of each unlicensed frequency band signal source relative to the train antenna are determined by multi-antenna angle of arrival estimation technology. Multiple unlicensed frequency band signal sources that are spatially close and have a frequency overlap higher than a preset overlap threshold are grouped into interference clusters, and the geometric centroid of each interference cluster is calculated and marked as a noise anchor point. A spherical expansion scan is performed with the noise anchor point as the center. When it is determined that adjacent expanded spheres are in contact or overlap, the spheres are merged. A Boolean subtraction operation on the carriage structure is performed on the merged sphere to generate the interference noise cloud.

4. The intelligent networking communication method for rail trains as described in claim 3, characterized in that: The step of grouping multiple unlicensed frequency band signal sources that are spatially adjacent and have a frequency overlap higher than a preset overlap threshold into interference clusters, calculating the geometric centroid of each interference cluster, and marking the geometric centroid as a noise anchor point includes: Monitor the radio frequency signal strength of each unlicensed frequency band signal source at the receiving end, and estimate the local energy density parameters of each unlicensed frequency band signal source around the vehicle antenna based on the radio frequency signal strength and propagation distance model; High-energy interference sources with local energy density parameters higher than a preset density threshold are selected, and high-energy interference sources that are spatially close are marked as candidate aggregation objects; Calculate the frequency overlap between the candidate aggregation objects, and when the frequency overlap is determined to be higher than the preset overlap threshold, merge the candidate aggregation objects into the interference cluster; Based on the spatial coordinates of the candidate aggregation objects contained in the interference cluster, the mass center coordinates of the interference cluster are calculated, and the mass center coordinates are determined as the noise anchor point.

5. The intelligent networking communication method for rail trains as described in claim 4, characterized in that: The step of performing spatial geometric superposition of the virtual transmission cone and the interference noise cloud to determine the overlapping volume of the virtual transmission cone and the interference noise cloud, and marking the overlapping volume as the radio frequency conflict region, includes: The interference noise cloud and the virtual transmission cone are superimposed in a three-dimensional spatial coordinate system. The overlapping volume block inside the virtual transmission cone that is eroded by the interference noise cloud is extracted using a three-dimensional geometric intersection algorithm. The overlapping volume block is marked as a spatial conflict sub-body. Listen to the preamble of the channel signal and analyze the transmission period and duty cycle of the unlicensed frequency band signal source; The transmission time of the control data packet to be sent by the train is compared with the active time window of the unlicensed frequency band signal source. When it is determined that the transmission time is within the active time window and the spatial conflict sub-body is covered by the interference noise cloud, the radio frequency conflict zone is activated.

6. The intelligent networking communication method for rail trains as described in claim 5, characterized in that: The determination of the target interference type based on the spatial location of the radio frequency conflict zone in the virtual transmission cone includes: Using the physical outer edge of the train antenna as the first cutting surface, the near-field boundary as the second cutting surface, and the effective receiving range boundary of the ground access point antenna as the third cutting surface, the virtual transmission cone is divided into a near-field flooding area, a mid-to-far-field attenuation area, and a receiving end sensitive area. Determine whether the radio frequency conflict zone and the near-field flooding zone overlap. If they overlap, mark the target interference conflict type as a near-field conflict type. Determine whether the radio frequency conflict region and the mid-to-far field attenuation region overlap. If they overlap, mark the target interference conflict type as a mid-to-far field conflict type. Determine whether the radio frequency conflict area overlaps with the receiver's sensitive area. If there is overlap, mark the target interference conflict type as a far-field conflict type.

7. The intelligent networking communication method for rail trains as described in claim 6, characterized in that: The execution of the intelligent networking anti-interference strategy based on the target interference conflict type includes: In response to the target interference conflict type being the near-field conflict type, the interference cluster anchor point corresponding to the radio frequency conflict zone is obtained, and the interference azimuth angle and interference elevation angle of the interference cluster anchor point relative to the train antenna are calculated. The weighting coefficients of each radiating element are adjusted by calling the antenna array control interface. In the direction directly opposite the interference azimuth and the interference elevation angle in the antenna radiation pattern, a gain null is artificially created by the phase cancellation principle. By compressing the main lobe width of the virtual transmission cone, the control signal energy is concentrated in the direction of the ground access point, forming a signal transmission corridor that avoids the radio frequency conflict zone.

8. The intelligent networking communication method for rail trains as described in claim 7, characterized in that: The step of executing the intelligent networking anti-interference strategy based on the target interference conflict type further includes: In response to the target interference conflict type being the mid-to-far field conflict type, the time-domain and frequency-domain features of the interference signal inside the radio frequency conflict zone are extracted, and the center frequency point and transmission period of the interference signal are identified. Time offset processing is performed on the control data packets to be sent by the train, delaying the transmission time of the control data packets to the quiet interval within the transmission cycle; Alternatively, frequency switching processing can be performed on the control data packets to be sent by the train, switching the transmission frequency band of the control data packets to a spare frequency band that does not overlap with the center frequency point.

9. The intelligent networking communication method for rail trains as described in claim 8, characterized in that: The step of executing the intelligent networking anti-interference strategy based on the target interference conflict type further includes: In response to the target interference conflict type being the far-field conflict type, the diversity reception logic of the ground access terminal is activated, multiple spatially separated receiving antennas are used to receive control signals, and the one with the optimal signal-to-noise ratio is selected as the input signal. Redundant check bytes are added to the control data packets to be sent by the train, and the modulation mode of the control data packets is downgraded to a low-order modulation mode; Alternatively, the transmission polarization of the control signal can be changed to counteract co-frequency interference using polarization isolation.

10. An intelligent networking communication system for rail trains, employing the intelligent networking communication method for rail trains as described in any one of claims 1 to 9, characterized in that, include: The transmission cone construction module is used to obtain the train stopping position and the ground access point position, and construct a virtual transmission cone based on the train stopping position and the ground access point position. The virtual transmission cone represents the electromagnetic wave coverage space range of the train antenna sending control signals to the ground access point. The interference cloud generation module is used to perform spectrum scanning on the interior of the carriage and the surrounding area of ​​the train to obtain spatial location information of the interference source. Based on the spatial location information of the interference source, an interference noise cloud is generated in three-dimensional space. The interference noise cloud represents the energy distribution pattern of radio frequency signals generated by passenger hotspot devices. The conflict zone determination module is used to perform spatial geometric superposition of the virtual transmission cone and the interference noise cloud, determine the overlapping volume of the virtual transmission cone and the interference noise cloud, and mark the overlapping volume as the radio frequency conflict zone; The strategy execution module is used to determine the target interference conflict type based on the spatial location of the radio frequency conflict zone in the virtual transmission cone, and to execute an intelligent networking anti-interference strategy based on the target interference conflict type.

Images