Rail transit vehicle-ground wireless reliable communication method

Abstract

The invention relates to the technical field of communication, in particular to a rail transit train-ground wireless reliable communication method, which comprises the following steps of: based on a train position, a train current speed, a train prediction track and a base station coverage range, selecting a service base station and a specified number of front and back adjacent base stations according to the train prediction track; according to the method, the service base station and the front and back adjacent base stations are dynamically selected based on the train prediction track and the real-time channel gain to construct the cooperative cluster, and the target train channel vector is extracted from the cluster, so that the base station set participating in cooperative processing can be ensured to be most relevant to the current and future positions of the train all the time; the dynamic and prospective base station selection mechanism can actively adapt to the high-speed movement characteristic of the train, and base stations with remarkable signal contribution or potential strong interference are managed in a unified manner in advance, so that a foundation is laid for beam forming and interference suppression.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and in particular to a reliable wireless communication method for rail transit vehicles to the ground. Background Technology

[0002] The field of communication technology is a broad discipline encompassing information transmission, exchange, and processing. Specifically, this topic focuses on wireless communication, particularly dedicated mobile communication systems for high-speed mobile environments.

[0003] Traditional cellular network base station selection and handover mechanisms are mostly based on lagging signal quality measurements. For high-speed moving trains, this reactive approach is insufficient to effectively handle rapidly changing channel conditions and frequent cell edge crossings, often leading to handover delays, handover failures, or temporary degradation in communication quality. For example, when a train approaches a cell boundary at high speed, signal strength drops rapidly, but there is an inherent delay in identifying and executing a handover to a better base station, making it difficult to guarantee communication quality during this period. Secondly, in terms of interference management, existing technologies typically employ power control or static interference coordination strategies based on large areas. These methods have limited effectiveness in suppressing strong directional, dynamically changing co-channel interference generated by adjacent base stations under linear coverage in rail transit systems. Therefore, improvements are needed. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a reliable wireless communication method for rail transit vehicles to the ground.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a reliable wireless communication method for rail transit vehicles, comprising the following steps: Based on the train's location, current speed, predicted trajectory, and base station coverage, a serving base station and a specified number of adjacent base stations are selected according to the predicted trajectory. The corresponding gain value in the channel gain matrix of the base station for the target train is compared with a preset threshold. Base stations that exceed the preset threshold are included in the cooperative set to construct a dynamic cluster. The channel vector of the target train is extracted from each base station in the dynamic cluster to form a subset of channel state information, thus obtaining the initial cell cluster configuration information and the channel state subset. Based on the initial cell cluster configuration information and channel state subset, each cooperating base station within the cluster specified in the initial cell cluster configuration information and channel state subset uses the obtained channel vector of the target train to calculate beamforming weights, and performs distributed iterative calculations to adjust beamforming weights by exchanging necessary other user interference information, until the beamforming weights of each base station converge or reach a preset number of iterations, thereby generating an optimized cooperative beamforming weight set.

[0006] Preferably, the method further includes: For control signaling services, passenger information system services, on-board monitoring services, and passenger Wi-Fi services within the rail transit system, bandwidth requirements, latency tolerance, reliability requirements, and burst modeling are performed. Based on the recorded bandwidth requirements, latency tolerance, reliability threshold, and burst characteristic parameters of each service, as well as the preset service importance level, a service priority queue is established. At the same time, historical channel data sequences and current real-time channel measurements are collected. By weighted averaging of historical data and combining it with current measurements, channel quality parameter values ​​within future transmission time intervals are predicted, and the service priority queue and predicted channel quality indicators are obtained. Based on the optimized cooperative beamforming weight set, the service priority queue, and the predicted channel quality index, available time slot units, subcarrier units, power units, and spatial beam units are integrated into a resource pool. Then, according to the service priority queue and the service quality requirements of each service in the predicted channel quality index, fixed resource slices are allocated to the highest priority control signaling service. The remaining multi-dimensional resources are then dynamically divided and scheduled according to the real-time data requests of other services, service priorities, and the predicted channel quality of the corresponding links to establish a dynamic resource allocation scheme.

[0007] Preferably, the steps for obtaining the initial cell cluster configuration information and the channel state subset are as follows: Based on the train's location, current speed, predicted trajectory, and base station coverage, the system analyzes the future locations of multiple sampling points of the train, matches the sampling point locations with the geographical information of the coverage areas pre-stored by each base station, selects the serving base station and a specified number of adjacent base stations, and obtains a list of candidate base stations. Based on the candidate base station list and the channel gain matrix of the base stations to the target train, the real-time channel gain value to the target train is extracted for each base station in the candidate base station list, and the real-time channel gain value is compared with a preset signal strength threshold. Base stations with real-time channel gain values ​​greater than the preset signal strength threshold are selected, and a dynamic cluster is constructed to obtain a set of cooperative base stations. Based on the cooperative base station set, for each cooperative base station identified in the cooperative base station set, the channel state information between the cooperative base station and the target train is read from the real-time channel measurement results, and the amplitude and phase data of the multi-antenna channel vector are extracted to obtain the initial cell cluster configuration information and channel state subset.

[0008] Preferably, the step of obtaining the optimized cooperative beamforming weight set is as follows: Based on the initial cell cluster configuration information and channel state subset, each cooperating base station in the cluster independently retrieves the target train channel vector corresponding to the initial cell cluster configuration information and channel state subset, performs a conjugate transpose operation on the channel vector, and uses the operation result as the initial transmit beamforming weight to obtain the initial beamforming weight of each base station. Based on the initial beamforming weights of each base station and other user interference information, each cooperating base station announces to each other the estimated interference channel information for other users in the same cluster. Combined with the initial beamforming weights of each base station and the target channel vector, an iterative calculation input parameter set is established. Based on the set of input parameters for the iterative operation, weight update calculations are performed locally to maximize the target signal power while suppressing interference, until the change in beamforming weights calculated by all base stations between two consecutive iterations is less than the preset convergence criterion value, thereby generating an optimized collaborative beamforming weight set.

[0009] Preferably, the steps for obtaining the service priority queue and the predicted channel quality index are as follows: For control signaling services, passenger information system services, on-board monitoring services, and passenger Wi-Fi services within the rail transit system, the bandwidth requirements, latency tolerance limits, reliability guarantee thresholds, and data burst characteristics of each service record are reviewed. Weighted scores are then calculated based on preset service importance level coefficients, and the results are sorted in descending order to establish service priority ranking data. Based on historical channel data sequences and current real-time channel measurements, a sequence of historical channel quality indicators within a specified time period is read. A sliding window mean filter is applied to the data points in the historical channel quality indicator sequence. The filtered historical data is then concatenated with the latest collected real-time channel measurements to obtain a preprocessed channel dataset.

[0010] Preferably, the steps for obtaining the service priority queue and predicted channel quality index further include: based on the preprocessed channel dataset and the service priority ranking data, assigning a weight factor that decays exponentially over time to the smoothed historical data points in the preprocessed channel dataset to calculate the average value of the weighted data points, and correcting the average value using the current measurement value to obtain the predicted value of the future channel quality parameters; integrating the service priority ranking data to obtain the service priority queue and predicted channel quality index.

[0011] Preferably, the step of obtaining the dynamic resource allocation scheme is as follows: Based on the optimized cooperative beamforming weight set, the service priority queue and the predicted channel quality index, all schedulable time slot resource units, frequency subcarrier resource units, base station transmit power resource units and available spatial beam directions determined by the optimized cooperative beamforming weight set are uniformly assigned unique identifiers and their current availability is recorded to establish integrated multi-dimensional resource pool information. Based on the integrated multi-dimensional resource pool information, service priority queue, and predicted channel quality index, the control signaling service with the highest priority is identified from the service priority queue and predicted channel quality index. According to the predefined fixed bandwidth, latency, and reliability requirements, a target number of time-frequency resource blocks are directly selected from the integrated multi-dimensional resource pool information for marking and reservation, and the reserved resource allocation record and remaining resource pool are obtained.

[0012] Preferably, the step of obtaining the dynamic resource allocation scheme further includes: based on the reserved resource allocation record and the remaining resource pool, the service priority queue and the predicted channel quality index, and other real-time data requests, matching and allocating the remaining time slots, subcarriers, power, and beam resources marked as available in the reserved resource allocation record and the remaining resource pool one by one according to the real-time data transmission request volume of other services, the priority order in the service priority queue and the predicted channel quality index, and the predicted channel quality value of the corresponding link, to establish a dynamic resource allocation scheme.

[0013] Compared with the prior art, the advantages and positive effects of the present invention are as follows: This invention dynamically selects serving base stations and adjacent base stations based on the predicted train trajectory and real-time channel gain to construct a cooperative cluster. It then extracts the target train's channel vector from this cluster, ensuring that the set of base stations participating in the cooperative processing is always most relevant to the train's current and future location. This dynamic and forward-looking base station selection mechanism proactively adapts to the high-speed movement of trains, preemptively incorporating base stations with significant signal contributions or potential strong interference into unified management, laying the foundation for beamforming and interference suppression. Compared to traditional fixed cell division or reactive handover, this improves flexibility and robustness in dealing with rapidly changing channel environments. Furthermore, each cooperating base station within the cluster calculates beamforming weights using the acquired channel vectors and performs distributed iterative calculations to adjust the weights by exchanging necessary interference information from other users until convergence. This achieves coherent signal superposition at the target train location and effective suppression of interference from other users. This cooperative beamforming method not only enhances the target signal strength but also reduces co-channel interference within the system, thereby improving the target train's received signal-to-noise ratio, increasing the reliability of the communication link and data transmission rate, especially in traditionally weak coverage areas such as cell edges. Furthermore, modeling and prioritizing various service characteristics within the rail transit system, combined with historical and current channel data, predicts future channel quality, providing a basis for differentiated service assurance and making subsequent resource allocation more targeted. Finally, available time slots, subcarriers, power, and spatial beam resources are integrated into a multi-dimensional resource pool. Based on service priority and predicted channel quality, fixed resource slices are allocated to the highest priority control signaling services, while remaining resources are dynamically allocated and scheduled. This ensures the communication quality of critical services while improving the utilization efficiency of spectrum and power resources, meeting the differentiated communication service quality requirements of diverse rail transit services. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the steps of the present invention. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0016] Please see Figure 1 This invention provides a technical solution, a reliable wireless communication method for rail transit vehicles to the ground, comprising the following steps: Based on the train's location, current speed, predicted trajectory, and base station coverage, a serving base station and a specified number of adjacent base stations are selected according to the predicted trajectory. The corresponding gain value in the channel gain matrix of the base station for the target train is compared with a preset threshold. Base stations that exceed the preset threshold are included in the cooperative set to construct a dynamic cluster. The channel vector of the target train is extracted from each base station in the dynamic cluster to form a subset of channel state information, thus obtaining the initial cell cluster configuration information and the channel state subset. Based on the initial cell cluster configuration information and channel state subset, each cooperating base station within the cluster specified in the initial cell cluster configuration information and channel state subset uses the obtained channel vector of the target train to calculate beamforming weights, and performs distributed iterative calculations to adjust beamforming weights by exchanging necessary interference information from other users, until the beamforming weights of each base station converge or reach the preset number of iterations, thereby generating an optimized set of cooperative beamforming weights. For control signaling services, passenger information system services, on-board monitoring services, and passenger Wi-Fi services within the rail transit system, bandwidth requirements, latency tolerance, reliability requirements, and burst modeling are performed. Based on the recorded bandwidth requirements, latency tolerance, reliability threshold, and burst characteristic parameters of each service, as well as the preset service importance level, a service priority queue is established. At the same time, historical channel data sequences and current real-time channel measurements are collected. By weighted averaging of historical data and combining it with current measurements, channel quality parameter values ​​within future transmission time intervals are predicted, and the service priority queue and predicted channel quality indicators are obtained. Based on the optimized collaborative beamforming weight set, service priority queue, and predicted channel quality index, available time slot units, subcarrier units, power units, and spatial beam units are integrated into a resource pool. Then, according to the service priority queue and the service quality requirements of each service in the predicted channel quality index, fixed resource slices are allocated to the highest priority control signaling service. The remaining multi-dimensional resources are then dynamically divided and scheduled according to the real-time data requests of other services, service priorities, and the predicted channel quality of the corresponding links to establish a dynamic resource allocation scheme.

[0017] The steps for obtaining the initial cell cluster configuration information and channel state subset are as follows: Based on the train's location, current speed, predicted trajectory, and base station coverage, the system analyzes the future locations of multiple sampling points of the train, matches the sampling point locations with the geographical information of the coverage areas pre-stored by each base station, selects the serving base station and a specified number of adjacent base stations, and obtains a list of candidate base stations. Based on the candidate base station list and the channel gain matrix of the base station to the target train, the real-time channel gain value to the target train is extracted for each base station in the candidate base station list, and the real-time channel gain value is compared with a preset signal strength threshold. Base stations with real-time channel gain values ​​greater than the preset signal strength threshold are selected to construct a dynamic cluster and obtain a set of cooperative base stations. Based on the set of cooperative base stations, for each cooperative base station identified in the set, the channel state information between the base station and the target train is read from the real-time channel measurement results. The amplitude and phase data of the multi-antenna channel vector are extracted to obtain the initial cell cluster configuration information and the channel state subset.

[0018] Specifically, based on the train's location, current speed, predicted trajectory, and base station coverage, the system first predicts the train's trajectory for the next 10 seconds based on its current speed and a preset prediction time window, setting a sampling interval (e.g., 0.2 seconds). This process resolves 50 predicted sampling points for the future train, each containing longitude, latitude, and altitude information. Next, these 50 predicted sampling points are matched one by one with the pre-stored geographical information of the coverage areas of all base stations in the system. This geographical information is a polygonal region centered on the base station, formed by contour lines of signal strength obtained through actual road testing or simulation. For example, the coverage area of ​​base station A consists of a series of geographical coordinate points. The description involves determining whether a train's predicted sampling point is located within the coverage polygon of a certain base station. If a sampling point falls within the coverage of multiple base stations simultaneously, all of these base stations are recorded. Then, the serving base station that provides the primary communication connection to the train at the current moment is identified. This is usually the base station with the strongest signal at the train's current location. Based on the train's predicted trajectory and direction of travel, the serving base station, along with a specified number of base stations in front of it (e.g., 3) and a specified number of base stations behind it (e.g., 1), are selected from the matched base stations. The selection of these adjacent base stations is based on their coverage order on the train's predicted trajectory, thus obtaining a candidate base station list.

[0019] Based on the candidate base station list and the channel gain matrix of the base stations to the target train, for each base station in the candidate base station list, the real-time channel gain value from that base station to the target train is retrieved from the channel gain matrix. This channel gain matrix is ​​a dynamically updated two-dimensional table maintained by the system, where rows represent base stations and columns represent trains. The values ​​represent the gain values ​​under the combined effects of path loss, shadowing fading, and multipath effects, typically in dB. Subsequently, each extracted real-time channel gain value is compared with a preset signal strength threshold. This preset signal strength threshold is set to ensure that the base station can provide an effective signal contribution to the train, avoiding the introduction of base stations with weak signals that could lead to poor cooperation or unnecessary computational overhead. Its value can be obtained through statistical analysis of historical communication data, for example, by statistically analyzing data from similar track sections and environments. The threshold is set to -95dB, which is the average channel gain between the base station and the train that can successfully establish a high-quality communication link, minus one standard deviation, or an empirical value. For example, if the channel gain for generally good communication is above -80dB, a threshold of -95dB can be set to ensure a certain signal quality. For instance, if statistical analysis determines that the minimum channel gain for effective cooperation is -90dBm, then this threshold is -90dBm. If a candidate base station, such as base station B, has a real-time channel gain of -85dBm to the target train, then because -85dBm is greater than -90dBm, the base station will be selected. Conversely, if it is -98dBm, it will not be selected. In this way, all base stations with real-time channel gain values ​​greater than the preset signal strength threshold are selected. These selected base stations are combined to construct a dynamic cluster, resulting in a set of cooperating base stations.

[0020] Based on the set of cooperating base stations, for each identified cooperating base station in the set, the complete channel state information (CSI) between the cooperating base station and the target train is accurately read from the real-time channel measurement results maintained by the base station side and periodically reported to the target train or obtained through detection signal measurement. This channel state information is specifically represented as a channel matrix describing the characteristics of multiple-input multiple-output (MIMO) channels. Each element of the matrix Indicates the first root transmitting antenna to the first The complex channel gain between the root receiving antennas, and then from the channel matrix. For each cooperating base station, the multi-antenna channel vector to the target train is extracted. Specifically, if the target train is equipped with... The base station is equipped with a root receiving antenna. If the transmitting antenna is used, then one is extracted. The matrix or a specific simplified form such as The equivalent channel vector (if considering a single data stream) is obtained, and the amplitude and phase data constituting these complex channel gains are further analyzed, for example, for each element in the channel vector. It can be represented as ,in It's the amplitude. Phase refers to the aggregation of the identifiers of all cooperating base stations and their respective channel vectors (including amplitude and phase data) to the target train, resulting in the initial cell cluster configuration information and a subset of channel states.

[0021] The steps for obtaining the optimized cooperative beamforming weight set are as follows: Based on the initial cell cluster configuration information and channel state subset, each cooperating base station in the cluster independently retrieves the target train channel vector corresponding to the initial cell cluster configuration information and channel state subset, performs conjugate transpose operation on the channel vector, and uses the operation result as the initial transmit beamforming weight to obtain the initial beamforming weight of each base station. Based on the initial beamforming weights of each base station and other user interference information, each cooperating base station announces to each other the estimated interference channel information for other users in the same cluster. Combined with the initial beamforming weights of each base station and the target channel vector, an iterative calculation input parameter set is established. Based on the iterative calculation input parameter set, weight update calculations are performed locally to maximize the target signal power while suppressing interference, until the change in beamforming weights calculated by all base stations between two consecutive iterations is less than the preset convergence criterion value, thus generating an optimized collaborative beamforming weight set.

[0022] Specifically, based on the initial cell cluster configuration information and a subset of channel states, each selected cooperating base station within the dynamic cluster independently performs the operation. First, each cooperating base station accurately retrieves the channel vector specific to its own connection with the target train from the received initial cell cluster configuration information and channel state subset, denoted as... ,in The base station index is a vector describing the channel characteristics from each transmit antenna of the base station to each receive antenna of the target train. Then, each cooperating base station uses its acquired channel vector... Performing the conjugate transpose operation, we get ,in This represents the conjugate transpose. The purpose of this operation is to maximize the transmitted energy of the signal in its own channel direction. Then, the result of this operation is... This is directly used as the initial transmit beamforming weight for the base station during the beamforming process. ,in The Frobenius norm is represented, and normalization is performed to control the transmit power. This process is executed in parallel for all cooperating base stations within the cluster to obtain the initial beamforming weights for each base station.

[0023] Based on the initial beamforming weights obtained from each base station And other necessary user interference information: First, each cooperating base station communicates with each other via an inter-base station control channel (e.g., the X2 interface or its equivalent interface) regarding its estimated interference channel information for users served by other cooperating base stations in the same cluster. Specifically, the base stations... It will transmit signals to non-target users. (by other base stations) Interference channel vectors generated by the service Information, or based on its initial beamforming weights The resulting impact on users Estimated interference intensity These interference messages are broadcast or sent point-to-point to other relevant base stations within the cluster, and are transmitted to other base stations. Initial beamforming weights and base stations Channel vector to its target user Together, for each base station It will collect interference information from other base stations and its own target channel vector. Its own initial beamforming weights Together with the initial beamforming weights of other base stations within the cluster, they constitute the complete set of input parameters required for iterative computation.

[0024] Based on the iterative computation input parameter set, each cooperating base station independently performs beamforming weight update calculations on its local processor. The goal of this calculation is to maximize the signal power of its base station to its target train. Simultaneously, in the iterative calculation to maximize its target signal power, interference to other users served by other base stations within the same cluster is suppressed to an acceptable level. This is achieved, for example, through iterative optimization using the Maximum Signal-to-Interference-Ratio (Max-SINR) criterion or the Minimum Mean Square Error (MMSE) criterion. In the middle, base station Beamforming weights The update will take into account other base stations in the first stage. The weight of the next iteration The resulting interference, and the target channel For example, weight updates involve solving an optimization problem, such as... ,in It is noise power. From base station to base station The target user's interference channel, this iterative process is performed synchronously or asynchronously among all cooperating base stations, until the beamforming weights calculated by all base stations are within the range of two consecutive iterations (e.g., the first iteration). Second and third The magnitude of change between (times), for example, by calculating All are less than a preset convergence criterion value, which is set according to the requirements for convergence speed and accuracy, for example, set to... If the change range after 10 iterations is from Down to If the convergence is reached, or the preset maximum number of iterations is reached (e.g., 20 times) to prevent infinite iteration, then the latest beamforming weight set is the desired result, generating the optimized cooperative beamforming weight set.

[0025] The steps for obtaining the service priority queue and predicted channel quality indicators are as follows: For control signaling services, passenger information system services, on-board monitoring services, and passenger Wi-Fi services within the rail transit system, the bandwidth requirements, latency tolerance limits, reliability guarantee thresholds, and data burst characteristics of each service record are reviewed. Weighted scores are then calculated based on preset service importance level coefficients, and the results are sorted in descending order to establish service priority ranking data. Based on historical channel data sequences and current real-time channel measurements, the historical channel quality index sequence within a specified time period is read, and a sliding window mean filter is applied to the data points in the historical channel quality index sequence. The filtered historical data is then concatenated with the latest collected current real-time channel measurements to obtain a preprocessed channel dataset. Based on the preprocessed channel dataset and service priority ranking data, the average value of the weighted data points is calculated by assigning a weight factor that decays exponentially over time to the smoothed historical data points in the preprocessed channel dataset. The average value is then corrected using the current measurement value to obtain the predicted value of the future channel quality parameters. The service priority ranking data is then integrated to obtain the service priority queue and the predicted channel quality index.

[0026] Specifically, for the four typical services predefined within the rail transit system—control signaling services, Passenger Information System (PIS) services, Onboard Television (CCTV) services, and Passenger Wi-Fi services—the normalized bandwidth requirement values ​​for each service are first retrieved from the system configuration database or service requirement documentation. For example, the bandwidth requirement for control signaling services is 0.1 (unit, e.g., a percentage of total available bandwidth), for PIS services it is 0.2, for CCTV services it is 0.3, and for Passenger Wi-Fi it is 0.4. Simultaneously, the latency tolerance limits are also retrieved, such as 10 milliseconds for control signaling services, 100 milliseconds for PIS services, 500 milliseconds for CCTV services, and 1000 milliseconds for Passenger Wi-Fi, as well as the reliability guarantee threshold, such as 99 milliseconds for control signaling services. The data burst characteristics are assessed using statistical parameters, such as Peak-to-Average Ratio (PAR). For example, the PAR for control signaling services is 1.5, for PIS services 2.0, for CCTV services 3.0, and for passenger Wi-Fi 5.0. Next, an importance level coefficient is preset for each service. This coefficient is subjectively set by the operator based on the service's impact on driving safety and passenger experience. For example, the importance level coefficient for control signaling services is 10, for PIS services 7, for CCTV services 5, and for passenger Wi-Fi 3. Then, a weighted score is applied to each service. The scoring formula can be designed as a weighted sum of various indicators. ,in These are the weights of each indicator, and these weights also need to be set in advance based on business characteristics and operational needs. For example, This score is then multiplied by the importance level coefficient of the corresponding service to obtain the final comprehensive score. For example, the final score for the control signaling service is: After calculating the final comprehensive score of all services, the services are sorted in descending order of score to establish service priority ranking data.

[0027] Based on the historical channel data sequence stored in the system and the current real-time channel measurement value collected by the wireless communication module, the system first reads the historical channel quality index sequence within a specified time period from the historical channel database. For example, it selects the parameter sequence such as Channel Quality Indicator (CQI), Received Signal Strength Indicator (RSSI), or Signal-to-Noise Ratio (SNR) recorded every second within the past 10 minutes. For example, if SNR is selected as the indicator, 600 historical SNR data points are obtained. Then, a sliding window mean filter is performed on the data points in this historical SNR sequence containing 600 data points. The size of the sliding window is set according to the channel change rate and noise level. For example, the window size is 5 data points, that is, the value of the current data point is replaced by the average of the current data point and the two data points before and after it to smooth short-term fluctuations and noise, resulting in a filtered historical SNR data sequence. Next, this filtered historical SNR data sequence is concatenated with the latest collected current real-time channel measurement value, such as the SNR value at the current moment, to form a channel dataset that includes the smoothed historical trend and the latest state, thus obtaining a preprocessed channel dataset.

[0028] Based on the preprocessed channel dataset and established service priority ranking data obtained in the previous step, the smoothed historical data points contained in the preprocessed channel dataset are first assigned weight factors that decay exponentially over time. The calculation method for these weight factors is as follows: ,in It is the distance from the current time. The weight of historical data points It is a smoothing factor between 0 and 1, for example, set to 0.3, and its value can be adjusted according to the required decay rate of the importance of historical data. The larger the value, the higher the weight of recent data. These weighting factors are then used to calculate a weighted average of historical data points as a preliminary prediction of future channel quality. Next, this weighted average is corrected using current real-time measurements from the preprocessed channel dataset. Correction methods include... The correction factor It is also a preset value between 0 and 1, such as 0.7, used to balance the influence of historical trends and the current instantaneous state. Its specific value can be set based on prior knowledge of the channel change characteristics or through experimental optimization. For example, if the channel change is relatively gentle, A larger value can be taken, and a smaller value can be taken otherwise. This calculation yields the predicted values ​​of channel quality parameters for one or more future transmission time intervals, such as the predicted average SNR value for the next second. Finally, this predicted channel quality parameter value is integrated with the previously established service priority ranking data, that is, the priority information of each service is associated with its corresponding predicted link quality information to obtain the service priority queue and the predicted channel quality index.

[0029] The steps to obtain a dynamic resource allocation scheme are as follows: Based on the optimized cooperative beamforming weight set, service priority queue and predicted channel quality index, all schedulable time slot resource units, frequency subcarrier resource units, base station transmit power resource units and available spatial beam directions determined by the optimized cooperative beamforming weight set are uniformly assigned unique identifiers and their current availability is recorded to establish an integrated multi-dimensional resource pool information. Based on the integration of multi-dimensional resource pool information, service priority queues and predicted channel quality indicators, the control signaling service with the highest priority is found from the service priority queues and predicted channel quality indicators. According to the predefined fixed bandwidth, latency and reliability requirements, the target number of time-frequency resource blocks are directly selected from the integrated multi-dimensional resource pool information for marking and reservation, and the reserved resource allocation record and the remaining resource pool are obtained. Based on the reserved resource allocation record and remaining resource pool, service priority queue and predicted channel quality index, and other real-time data requests, the remaining time slots, subcarriers, power and beam resources marked as available in the reserved resource allocation record and remaining resource pool are matched and allocated one by one according to the real-time data transmission request volume of other services, the priority order in the service priority queue and predicted channel quality index, and the predicted channel quality value of the corresponding link, to establish a dynamic resource allocation scheme.

[0030] Specifically, based on the optimized cooperative beamforming weight set, service priority queue, and predicted channel quality index generated in previous steps, all schedulable radio resources in the system are first inventoried and identified. This includes physical layer-defined time slot resource units (e.g., a specific time slot number in a radio frame), frequency subcarrier resource units (e.g., a specific subcarrier index in an OFDM system), transmit power resource units available to each cooperating base station (e.g., discrete power levels or continuous power ranges in dBm), and available spatial beam directions pointing towards the target train determined by the optimized cooperative beamforming weight set (each beam direction is defined by a specific set of beamforming weights, and there can be multiple alternative beams). These resource units of different dimensions are uniformly encoded and given unique identifiers. For example, time slot T1, subcarrier F5, power level P3 of base station B2, and beam direction V4 can be combined into a unique identifier for a resource block. The current availability status (e.g., idle, allocated, reserved) of each such identified resource unit is recorded, and a dynamically updated integrated multi-dimensional resource pool information is established.

[0031] Based on the established integrated multi-dimensional resource pool information and the service priority queue and predicted channel quality indicators obtained in the aforementioned steps, the highest priority service is first identified from the service priority queue and predicted channel quality indicators. According to the settings, this is usually a control signaling service. Then, based on the system's predefined fixed Quality of Service (QoS) requirements for control signaling services, these requirements include a fixed minimum bandwidth guarantee (e.g., equivalent to 2 standard resource blocks), strict latency limits (e.g., resource allocation must be completed within the next available time slot to meet the 10ms latency requirement), and extremely high latency requirements. For high reliability requirements (e.g., allocated resources must show links with an SNR greater than or equal to 20dB in the predicted channel quality index), query available resources that meet these conditions from the integrated multidimensional resource pool information, and directly select the target number of time-frequency resource blocks. For example, select two consecutive resource blocks with good predicted channel quality, and mark these selected resource blocks as "reserved for control signaling services" in the integrated multidimensional resource pool information. At the same time, record the detailed information of this allocation, such as the allocated resource identifier, allocation time, allocation object, etc., and obtain the reserved resource allocation record and the updated remaining resource pool.

[0032] Based on the reserved resource allocation record and remaining resource pool generated in the previous step, as well as the service priority queue and predicted channel quality index, and combined with the real-time data transmission requests received at the current moment from other services such as passenger information system service, vehicle monitoring service, and passenger Wi-Fi service, which include the expected data rate or data volume, the remaining time slots, subcarriers, power, and beam resources marked as available in the reserved resource allocation record and remaining resource pool are first traversed and evaluated. Then, according to the priority order of each service in the service priority queue and predicted channel quality index, the real-time data transmission requests of other services are processed one by one from high to low. For the currently processed service, the requested data transmission volume, its specific priority value in the service priority queue, and the predicted channel quality value of the link between the service and the serving base station (e.g., predicted SNR or...) are considered. The system identifies and allocates the best combination of resources from the remaining resource pool to meet the target bandwidth (SNR). The matching principle can be to prioritize resources with good predicted channel quality or to prioritize the needs of high-priority services. For example, if a passenger information system (PIS) service (with higher priority than vehicle monitoring) requests 1Mbps bandwidth, and there are two sets of available resource blocks, one with a predicted SNR of 15dB that can provide 0.8Mbps and the other with a predicted SNR of 18dB that can provide 1.2Mbps, then the resource block with the SNR of 18dB will be allocated to the PIS service first. If resources are sufficient, they will continue to be allocated to the next priority service. If resources are insufficient, adjustments may be needed based on more detailed fairness or efficiency strategies, such as proportional allocation or rejection of some low-priority requests, until all remaining resources are allocated or all service requests are processed, thus establishing a dynamic resource allocation scheme.

[0033] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A reliable communication method for rail transit train-ground wireless communication, characterized in that, The method comprises the following steps: Based on the train position, the train current speed, the train predicted trajectory and the base station coverage range, the service base station and a specified number of adjacent base stations before and after the train predicted trajectory are selected, the corresponding gain values in the channel gain matrix of the base stations to the target train are compared with the preset threshold value, the base stations exceeding the preset threshold value are included in the cooperation set, a dynamic zone cluster is constructed, the channel vectors of each base station in the dynamic zone cluster to the target train are extracted to form a channel state information subset, and initial cell cluster configuration information and a channel state subset are obtained; Based on the initial cell cluster configuration information and the channel state subset, each cooperative base station in the cluster specified from the initial cell cluster configuration information and the channel state subset uses the obtained channel vector to the target train to calculate the beamforming weight value, and through the exchange of necessary other user interference information, distributed iterative operation is performed to adjust the beamforming weight value until the beamforming weight value of each base station converges or reaches the preset iteration number, and an optimized cooperative beamforming weight value set is generated.

2. The reliable train-ground communication method of claim 1, wherein, The method further comprises: The bandwidth demand, the time delay tolerance, the reliability requirement and the burst modeling are performed for the control signaling service, the passenger information system service, the vehicle-mounted monitoring service and the passenger Wi-Fi service in the rail transit system, the business priority queue is established according to the recorded business bandwidth demand value, the time delay tolerance value, the reliability threshold value and the burst characteristic parameter, and the preset business importance level, the historical channel data sequence and the current channel real-time measurement value are collected, the future transmission time interval channel quality parameter value is predicted through the weighted average of the historical data and the combination of the current measurement value, and the business priority queue and the predicted channel quality index are obtained; Based on the optimized cooperative beamforming weight value set, the business priority queue and the predicted channel quality index, the available time slot unit, the sub-carrier unit, the power unit and the spatial beam unit are integrated into a resource pool, then according to the business priority and the service quality demand in the business priority queue and the predicted channel quality index, a fixed resource piece is allocated to the highest priority control signaling service, the remaining multi-dimensional resources are dynamically divided and scheduled according to the real-time data request of other services, the business priority and the corresponding link predicted channel quality, and a dynamic resource allocation scheme is established.

3. The reliable train-ground communication method of claim 1, wherein, The initial cell cluster configuration information and the channel state subset are obtained by: Based on the train position, the train current speed, the train predicted trajectory and the base station coverage range, the future multiple sampling point positions of the train are analyzed, the sampling point positions are matched with the pre-stored coverage area geographic information of each base station, the service base station and a specified number of adjacent base stations before and after the train predicted trajectory are selected, and a candidate base station list is obtained; Based on the candidate base station list and the channel gain matrix of the base stations to the target train, the real-time channel gain value of each base station in the candidate base station list to the target train is extracted, the real-time channel gain value is compared with the preset signal strength threshold value, the base stations with the real-time channel gain value greater than the preset signal strength threshold value are screened, a dynamic zone cluster is constructed, and a cooperative base station set is obtained; Based on the set of cooperating base stations, read the channel state information between the target train from the real-time channel measurement results for each determined cooperating base station in the set of cooperating base stations, extract the amplitude and phase data of the multi-antenna channel vector, and obtain the initial cell cluster configuration information and channel state subset.

4. The reliable train-ground communication method of claim 1, wherein, The obtaining step of the optimized cooperative beamforming weight set is: Based on the initial cell cluster configuration information and channel state subset, each cooperating base station within the cluster independently retrieves the target train channel vector corresponding to the initial cell cluster configuration information and channel state subset. Each cooperating base station retrieves the channel vector specific to its own connection with the target train from the received initial cell cluster configuration information and channel state subset, denoted as... ,in For the base station index, the channel vector describes the channel characteristics from each transmit antenna of the base station to each receive antenna of the target train. Each cooperating base station obtains the channel vector. Performing the conjugate transpose operation, we get ,in This represents the conjugate transpose, and the result of this operation... This is directly used as the initial transmit beamforming weight for the base station during the beamforming process. ,in The Frobenius norm is used to represent the initial transmit beamforming weights, and the calculation results are used as the initial transmit beamforming weights to obtain the initial beamforming weights for each base station. Based on the initial beamforming weight of each base station and the interference information of other users, each cooperating base station informs each other of the estimated interference channel information of other users in the same cluster, and combines the initial beamforming weight of each base station and the target channel vector to establish an iteration operation input parameter set. Based on the iterative operation input parameter set, a weight update calculation is performed locally to maximize target signal power while suppressing interference until the variation amplitude of the beamforming weight values calculated by all base stations between two consecutive iterations is less than a preset convergence criterion value, specifically, in each iteration In the method, the base station updates the beamforming weight values of the base station by referring to the interference generated by the weight values of the other base stations in the first iteration, and the target channel , specifically , where is the noise power, is the interference channel from the base station to the target user of the base station , and the iteration process is performed synchronously or asynchronously among all cooperative base stations until the variation amplitude of the beamforming weight values calculated by all base stations between two consecutive iterations is less than a preset convergence criterion value, thereby generating an optimized cooperative beamforming weight value set.

5. The reliable ground-to-train communication method for rail transit vehicles according to claim 2, characterized in that, The obtaining step of the service priority queue and predicted channel quality index is: For control signaling services, passenger information system services, vehicle monitoring services and passenger Wi-Fi services in the rail transit system, the bandwidth demand values, upper limits of delay tolerance, reliability guarantee thresholds and data burst characteristic statistical parameters of each service record are consulted, and a weighted score is obtained by combining the preset service importance level coefficient; and the weighted score is multiplied by the importance level coefficient of the corresponding service to obtain a comprehensive score. The formula of the weighted score is: wherein are weights of each index, is a weighted score, , , , are respectively bandwidth demand value, upper limit of delay tolerance, reliability guarantee threshold and data burst characteristic statistical parameter, arranged in descending order according to the comprehensive score result, and service priority arrangement data is established. Based on the historical channel data sequence and the current channel real-time measurement value, the historical channel quality index sequence in the past specified time length is read, and the data points in the historical channel quality index sequence are subjected to sliding window mean filtering. The size of the sliding window is set according to the channel change rate and noise level. The filtered historical data and the latest collected current channel real-time measurement value are spliced to obtain a preprocessed channel data set.

6. The reliable ground-to-train communication method for rail transit vehicles according to claim 5, characterized in that, The service priority queue and the acquisition of the predicted channel quality indicator further comprises: based on the pre-processed channel data set and the service priority arrangement data, assigning a weight factor with exponential decay over time to the smoothed historical data points in the pre-processed channel data set, calculating the average value of the weighted data points, the weight factor is calculated in the manner that wherein is the weight of the historical data point at the time ago from the current time, is a smoothing factor between 0 and 1, and the average value is corrected using the current measurement value to obtain the predicted value of the future channel quality parameter, the service priority queue and the predicted channel quality indicator are acquired by integrating the service priority arrangement data.

7. The reliable ground-to-train communication method for rail transit vehicles of claim 2, wherein, The obtaining step of the dynamic resource allocation scheme is: Based on the optimized cooperative beamforming weight set, the service priority queue and the predicted channel quality index, all schedulable time slot resource units, frequency subcarrier resource units, base station transmit power resource units and available spatial beam directions determined by the optimized cooperative beamforming weight set are uniformly compiled with unique identifiers and recorded with current available states to establish integrated multi-dimensional resource pool information. Based on the integrated multi-dimensional resource pool information, the service priority queue and the predicted channel quality index, the control signaling service with the highest priority is found from the service priority queue and the predicted channel quality index, and a target number of time-frequency resource blocks are directly selected from the integrated multi-dimensional resource pool information according to the predefined fixed bandwidth, delay and reliability requirements to be marked and reserved, and a reserved resource allocation record and a remaining resource pool are obtained.

8. The method of claim 7, wherein, The obtaining step of the dynamic resource allocation scheme further includes: based on the reserved resource allocation record and the remaining resource pool, the service priority queue and the predicted channel quality index and other real-time data requests of services, the remaining time slots, subcarriers, powers and beam resources marked as available in the reserved resource allocation record and the remaining resource pool are matched and allocated one by one according to the real-time data transmission request amount of each service, the priority order in the service priority queue and the predicted channel quality value of the corresponding link to establish a dynamic resource allocation scheme.

Images