A data transmission and detection system for a Beidou reference station

Abstract

The application discloses a Beidou reference station data transmission and detection system and relates to the technical field of wireless communication networks.The application introduces a link stability score, a link prediction state and a network environment disturbance level driven adaptive weight mechanism in a Beidou reference station multi-link data backhaul scene, so that path selection no longer depends on transmission parameters at a single moment, but considers the fluctuation characteristics and future trends of parameters in the time dimension, thereby significantly reducing the misjudgment of link availability caused by instantaneous interference in a complex electromagnetic environment.The application first constructs a normalized stability score by using statistical quantities such as a trend benchmark, a deviation amplitude and a deviation frequency, so that the system can identify which links are more stable in a long time window and avoid blind switching due to short-term false recovery.

Description

Technical Field

[0001] This invention relates to the field of wireless communication network technology, and in particular to a BeiDou reference station data transmission and detection system. Background Technology

[0002] BeiDou satellite navigation and positioning reference stations are widely deployed in diverse geographical and electromagnetic environments, including densely populated urban areas, areas surrounding industrial facilities, and areas near important infrastructure. These areas contain high-voltage power transmission and transformation equipment, high-power communication radars, and dense radio transmitting devices, forming a complex and variable electromagnetic field environment. In order to provide users with continuous and high-precision positioning services, the reference stations need to continuously transmit real-time observation data to the data center through multiple candidate links in the wireless communication network.

[0003] In existing technologies, to improve the reliability of data transmission from BeiDou reference stations, some systems adopt a combination strategy of multiple heterogeneous links and centralized quality assessment. For example, observation data is sent to the data center using multiple wireless communication links such as 4G / 5G cellular network links and microwave leased line links. At the data center, statistics and comparisons are performed based on transmission quality parameters such as packet loss rate and latency of each link, so as to select the link with better quality for subsequent data forwarding, in order to improve the overall transmission reliability through link redundancy.

[0004] However, in complex environments with strong and intermittent electromagnetic interference, the effectiveness of the above strategies is significantly limited. Strong interference can cause rapid and drastic fluctuations in wireless channel quality over a very short timescale, resulting in sudden jumps in transmission parameters such as packet loss rate and latency. Link quality assessment based on short sampling windows is prone to capturing only the momentary good state during interference gaps or the momentary deterioration state during interference peaks, thus leading to a biased judgment of the actual availability of the link in the subsequent period. This assessment bias may not only cause the system to frequently trigger unnecessary link switching, increasing the overhead of connection reconstruction and cache reordering, but may also lead to the selection of a transmission channel that is about to deteriorate at critical moments, causing interruption of observation data transmission or a sudden increase in latency, affecting the continuity and stability of BeiDou high-precision positioning services. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] This invention provides a BeiDou reference station data transmission and detection system that solves the problem that existing multi-link transmission relies solely on short-term transmission parameters for routing, making it susceptible to misjudgments and frequent switching induced by transient electromagnetic interference.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0008] In a first aspect, embodiments of the present invention provide a BeiDou reference station data transmission and detection system, comprising:

[0009] Includes at least one BeiDou reference station and a central server;

[0010] The Beidou reference station is equipped with at least two wireless communication links for communicating with the central server, and each wireless communication link corresponds to a physical link identifier.

[0011] The central server is configured to perform link quality assessment for each physical link and select a data transmission path for transmitting BeiDou reference station data based on the link quality assessment results of each physical link.

[0012] The link quality assessment includes:

[0013] Obtain the real-time transmission parameter sequence of the physical link;

[0014] Based on the real-time transmission parameter sequence, a link stability score is calculated to characterize the volatility of transmission parameters;

[0015] Based on the real-time transmission parameter sequence, the link quality of the physical link in a predetermined future time period is predicted to obtain the link prediction status.

[0016] Based on the real-time transmission parameters, the link stability score, and the link prediction status, a comprehensive link evaluation value is generated for the physical link.

[0017] The central server is configured to determine the current data transmission path for each BeiDou reference station based on the comprehensive evaluation value of each physical link, and control the BeiDou reference station to send observation data to the central server through the current data transmission path.

[0018] As a preferred embodiment of the BeiDou reference station data transmission and detection system described in this invention, the real-time transmission parameter sequence includes a time series of at least one of the following parameters: round-trip confirmation delay, packet loss rate, jitter, effective throughput, and retransmission count.

[0019] As a preferred embodiment of the BeiDou reference station data transmission and detection system described in this invention, the calculation of link stability score includes:

[0020] Within a preset evaluation window, the distribution characteristics of data points in the real-time transmission parameter sequence are analyzed;

[0021] The distribution characteristics include at least the magnitude of the deviation of each data point from the corresponding trend benchmark and the frequency of such deviation.

[0022] A normalized link stability score is determined based on the aforementioned distribution characteristics.

[0023] As a preferred embodiment of the BeiDou reference station data transmission and detection system of the present invention, the link prediction state is generated by a time series analysis model, which processes the real-time transmission parameter sequence to output a prediction result indicating the direction and magnitude of the change in the link quality level of the physical link within the predetermined future time period.

[0024] As a preferred embodiment of the BeiDou reference station data transmission and detection system of the present invention, the link quality assessment further includes abnormal transmission mode detection, which includes:

[0025] The acknowledgment delay of consecutive data units transmitted through the physical link is monitored to form a delay sequence;

[0026] Based on the comparison between the time delay sequence and the historical time delay distribution, abnormal time delay clusters composed of time delay values ​​that continuously exceed a preset deviation range are identified;

[0027] When the abnormal latency cluster is detected, the overall link evaluation value of the corresponding physical link is reduced.

[0028] As a preferred embodiment of the BeiDou reference station data transmission and detection system described in this invention, the generation of a comprehensive link evaluation value based on real-time transmission parameters, link stability score, and link prediction status includes:

[0029] Assign evaluation weights to the real-time transmission parameters, the link stability score, and the link prediction status, respectively.

[0030] The evaluation weights are adaptively adjusted based on the network environment disturbance level determined by the central server. The higher the network environment disturbance level, the greater the evaluation weights corresponding to the link stability score and the link prediction state.

[0031] As a preferred embodiment of the BeiDou reference station data transmission and detection system of the present invention, the data transmission path selection process includes a link switching suppression step. When switching from the currently used link to the target link, the link switching suppression step verifies whether the difference between the comprehensive link evaluation value of the target link and the currently used link meets the switching condition. The switching condition requires that the difference between the evaluation values ​​exceeds an adaptive threshold within a continuous observation period, and the size of the adaptive threshold is negatively correlated with the link stability score of the currently used link.

[0032] As a preferred embodiment of the BeiDou reference station data transmission and detection system of the present invention, the system further includes a cooperative mechanism for avoiding multi-station path selection oscillations, the cooperative mechanism including:

[0033] The central server generates and broadcasts reference link cost information to each BeiDou reference station.

[0034] After receiving the reference link cost information, each BeiDou reference station combines the private disturbance factor generated based on the unique identifier information of the BeiDou reference station to form a local differentiated link cost view of the BeiDou reference station.

[0035] Each BeiDou reference station participates in the initial link selection or triggers a link switching request to the central server based on its own differentiated link cost view.

[0036] As a preferred embodiment of the BeiDou reference station data transmission and detection system described in this invention, the system maintains a security trust level for each physical link. When the trigger frequency of abnormal transmission mode detection for a certain physical link exceeds a preset threshold, or when the link prediction status continuously indicates that the quality of the physical link will enter a low reliability range within a preset time period, the central server triggers a downgrade operation on the security trust level of the physical link and excludes physical links with a security trust level lower than the permitted threshold during subsequent data transmission path selection.

[0037] Secondly, this invention provides a method for data transmission and detection of BeiDou reference stations, including,

[0038] Step S1: Obtain the real-time transmission parameter sequence corresponding to the data sent by at least one BeiDou reference station to the central server through at least two wireless communication links;

[0039] Step S2: For each wireless communication link, calculate the link stability score and predict the link prediction state based on the real-time transmission parameter sequence, and generate a comprehensive link evaluation value based on the real-time transmission parameters, the link stability score and the link prediction state.

[0040] Step S3: Based on the comprehensive evaluation value of each wireless communication link, determine the data transmission path for each of the at least one BeiDou reference stations, and implement abnormal transmission mode detection, link switching suppression and multi-station coordination mechanism as needed.

[0041] The beneficial effects of this invention are as follows: By introducing an adaptive weighting mechanism driven by link stability scoring, link prediction status, and network environment disturbance level in the multi-link data backhaul scenario of BeiDou reference stations, this invention enables path selection to no longer rely solely on transmission parameters at a single moment, but comprehensively considers the fluctuation characteristics and future trends of parameters over time, thereby significantly reducing misjudgments of link availability caused by instantaneous interference in complex electromagnetic environments. Firstly, this invention utilizes statistical quantities such as trend benchmarks, deviation magnitude, and deviation frequency to construct a normalized stability score, enabling the system to identify which links perform more stably over a longer time window, avoiding blind switching due to short-term false improvements. Simultaneously, a time-series analysis model provides predictions of the direction and magnitude of changes in link quality levels, prioritizing paths among candidate links that are more likely to maintain a good state in the future, reducing the risk of immediate deterioration after switching. Based on this, the weights of real-time parameter scores, stability scores, and prediction scores are dynamically adjusted using network environment disturbance level as a control variable. When disturbances are small, the sensitivity of instantaneous parameters to new changes is fully utilized; when disturbances are severe, the reliance on stability and prediction information is increased, improving the robustness of overall decision-making. In conjunction with abnormal latency cluster detection and security trust level management, the system can perform long-term downgrading or even exclusion of links exhibiting recurring abnormal patterns, reducing the probability of problematic links being repeatedly selected among different base stations. Through link switching suppression mechanisms and multi-station collaborative path selection mechanisms, this invention also effectively suppresses frequent switching caused by small differences in evaluation and oscillations and congestion caused by multiple base stations simultaneously switching the same link, maintaining a relative balance in network group behavior.

[0042] In summary, this invention, while ensuring reliable real-time data transmission from BeiDou reference stations, significantly improves the continuity and stability of high-precision positioning services in complex electromagnetic environments. Attached Figure Description

[0043] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation on the scope of this application.

[0044] Figure 1 This is a schematic diagram of the framework of the Beidou reference station data transmission and detection system in the embodiment.

[0045] Figure 2 This is a flowchart illustrating the BeiDou reference station data transmission and detection method in this embodiment.

[0046] Figure 3 This is a schematic diagram of the link quality assessment process in the embodiment. Detailed Implementation

[0047] 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.

[0048] All terms used in this application (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0049] For example, the terms “first” and “second” used in this application are only used to distinguish and describe similar objects, to differentiate the first object from another object, and are not used to describe a specific order or sequence, nor should they be interpreted as indicating or implying relative importance.

[0050] This application proposes a BeiDou reference station data transmission and detection system, combining... Figure 1 As shown, the system includes:

[0051] Includes at least one BeiDou reference station and a central server;

[0052] The Beidou reference station is equipped with at least two wireless communication links for communicating with the central server, and each wireless communication link corresponds to a physical link identifier.

[0053] The central server is configured to perform link quality assessment for each physical link and select the data transmission path for transmitting BeiDou reference station data based on the link quality assessment results of each physical link.

[0054] Link quality assessment includes:

[0055] Obtain the real-time transmission parameter sequence of the physical link;

[0056] Based on real-time transmission parameter sequences, a link stability score is calculated to characterize the volatility of transmission parameters.

[0057] Based on the real-time transmission parameter sequence, the link quality of the physical link in a predetermined future time period is predicted to obtain the link prediction status.

[0058] Based on real-time transmission parameters, link stability score, and link prediction status, a comprehensive link evaluation value is generated for this physical link.

[0059] The central server is configured to determine the current data transmission path for each BeiDou reference station based on the comprehensive evaluation value of each physical link, and control the BeiDou reference station to send observation data to the central server through the current data transmission path;

[0060] In one embodiment, the real-time transmission parameter sequence includes a time series of at least one of the following parameters: round-trip acknowledgment delay, packet loss rate, jitter, effective throughput, and number of retransmissions;

[0061] In this embodiment, the round-trip confirmation delay can be obtained by the central server or BeiDou reference station timestamping the confirmation process of periodically sent probe messages or service messages at the application layer or transport layer. Packet loss rate and retransmission count can be calculated by statistically analyzing the number of data units sent within the evaluation window and the number of unacknowledged or retransmitted data units. Jitter can be estimated by the absolute value or standard deviation of the difference between adjacent delay samples. Effective throughput can be obtained by dividing the total number of successfully transmitted data bytes within the window by the window duration. Specifically, the sampling period for real-time transmission parameters can be configured between hundreds of milliseconds and several seconds according to the link type and service real-time requirements. In engineering implementation, it can optionally be set to sample once per second by default to balance parameter smoothness and sensitivity to changes. For example, the number of samples within the evaluation window can be 30 to 300 sampling points, corresponding to a time span of approximately half a minute to several minutes. The specific values ​​can be adjusted in different industry applications through on-site trial operation or maintenance experience to ensure the representativeness of the statistical results. When some parameters are missing in a certain sampling period, this embodiment can use the valid values ​​of the previous sampling period and internally mark the sampling point as an interpolation sample. In the case of multiple consecutive missing measurements, the statistical results of the corresponding parameters will not be updated temporarily to avoid occasional collection anomalies causing significant disturbance to the link quality assessment.

[0062] In one embodiment, calculating the link stability score includes:

[0063] Within a preset evaluation window, analyze the distribution characteristics of data points in the real-time transmission parameter sequence;

[0064] The distribution characteristics include at least the magnitude of each data point's deviation from the corresponding trend benchmark and the frequency of that deviation.

[0065] Determine the normalized link stability score based on distribution characteristics;

[0066] In one implementation, the normalized link stability score is as follows:

[0067] Step a, within the evaluation window, for each physical link Each type of real-time transmission parameter (e.g., round-trip confirmation delay, packet loss rate, etc.) form an observation sequence. ; when the sample index satisfies The position is based on the trend benchmark calculated using a moving average:

[0068] ,

[0069] in, Indicates in physical link Above, regarding the type of transmission parameters In the sample index The trend benchmark value at that point, This indicates the number of samples in the moving average. Indicates in physical link Above, regarding the type of transmission parameters In the sample index Real-time transmission parameter observations at the location, Represents the physical link index. Indicates the index of the transmission parameter type. This indicates the sample index within the evaluation window. Indicates the sample index that participated in the trend benchmark calculation;

[0070] Based on the trend benchmark, define the deviation at this position:

[0071] ,

[0072] in, Indicates in physical link Above, regarding the type of transmission parameters In the sample index The parameter deviation at that point. This represents the real-time transmission parameter observation value at the corresponding location. This represents the trend baseline value for the corresponding position;

[0073] Step b, collect data within the evaluation window. After a sample, to eliminate the differences in the dimensions of different parameters, a method is introduced that targets the parameter type. Reference deviation scale Normalize the deviation and construct a variance-based index:

[0074] ,

[0075] in, Indicates in physical link Above, regarding the type of transmission parameters The normalized volatility index, Indicates the number of samples within the evaluation window. Indicates the deviation amount. Indicates the type of transmission parameter The reference deviation scale is used to normalize parameters of different dimensions to a comparable order of magnitude;

[0076] Step c: Based on the deviation, introduce a judgment threshold. Statistical analysis of the frequency of excessive deviations; definition of deviation indicator:

[0077] ,

[0078] And obtain the deviation frequency index within the evaluation window:

[0079] ,

[0080] in, Indicates in physical link Above, regarding the type of transmission parameters In the sample index The deviation indicator value at the sampling point has a value of 1, which indicates that the absolute value of the deviation at that sampling point exceeds the threshold, and a value of 0, which indicates that it does not exceed the threshold. Indicates the type of transmission parameter The deviation from the judgment threshold, Indicates in physical link Above, regarding the type of transmission parameters The deviation frequency index;

[0081] Step d: Based on the volatility index and the deviation frequency index, introduce parameter-level weighting coefficients. and A normalized stability score is constructed using exponential mapping:

[0082] ,

[0083] in, Indicates in physical link Above, regarding the type of transmission parameters The normalized stability score, with a value range of [value range missing]. The closer the value is to 1, the smaller the fluctuation range and the lower the frequency of deviation, indicating a more stable link performance. Indicates the type of transmission parameter The volatility weighting coefficient is used to adjust the degree of influence of the volatility index on the score. Indicates the type of transmission parameter The deviation frequency weighting coefficient is used to adjust the degree of influence of the deviation frequency index on the score;

[0084] Step e, when for the same physical link Collected When transmitting parameters in real time, a weighted fusion of stability scores for each parameter type is performed to obtain a link-level normalized stability score:

[0085] ,

[0086] in, Indicates physical link The link-level normalized stability score, with a value range of [value range missing]. , This indicates the number of real-time transmission parameter types participating in the evaluation. Indicates the type of transmission parameter The fusion weighting coefficient is used to reflect the proportion of the impact of different transmission parameters on the overall stability, and satisfies... , This represents the parameter-level normalized stability score;

[0087] Through step ae, a normalized link stability scoring algorithm can be formed on the basis of the original distribution feature description, from trend benchmark, deviation statistics to multi-parameter fusion, providing quantifiable input for the generation of subsequent comprehensive link evaluation values;

[0088] Specifically, the above implementation method is refined around the link stability score, and the data point distribution characteristics are specifically divided into three types of statistics: trend deviation, fluctuation amplitude, and deviation frequency.

[0089] In terms of implementation, the system first extracts observations from each physical link within a fixed time window. A trend benchmark is constructed using a moving average to characterize slowly changing parts. Then, the deviation of the observations from this benchmark is calculated, transforming the original time series into a metric space centered on deviation. Based on this, a reference scale matching the parameter type is introduced to normalize the deviation and accumulate it in mean square form, yielding an amplitude index that reflects the overall strength of fluctuations. Simultaneously, a deviation threshold is set, and sampling points exceeding the threshold are considered abnormal events. Their frequency within the window is statistically analyzed, and a frequency index is used to reflect the occurrence of abnormal patterns such as continuous jitter or sudden congestion. Subsequently, through weighted exponential mapping, the amplitude and frequency indices are integrated into a single normalized score, ensuring the score naturally falls within a fixed range for easy comparison with the predicted state and other parameters. Considering the different contributions of parameters such as latency and packet loss to link performance, the scheme further employs a weighted fusion mechanism to convert multi-dimensional scores into link-level stability scores, providing a quantitative basis for the system to comprehensively evaluate and select paths among multiple links.

[0090] In this embodiment, the length of the evaluation window can be configured according to the comprehensive requirements of the service for response time and statistical stability. In urban public network environments, 30 to 120 sampling points are typically selected, while in dedicated line or satellite link environments, this can be appropriately increased to 100 to 300 sampling points to more fully reflect the characteristics of slow-changing fluctuations. Specifically, the number of samples used to construct the moving average for the trend benchmark can be no more than half the length of the evaluation window. In typical scenarios, it can be set to 5-20 sampling points, giving the trend benchmark a certain filtering capability for short-term mutations without excessive lag. For example, the reference deviation scale used for normalizing deviations can be determined according to the root mean square value or percentile interval width of various real-time transmission parameters during historical stable operation phases. This ensures that parameters of different dimensions have roughly similar orders of magnitude after normalization, thereby preventing one type of parameter from dominating the stability score. Similarly, the deviation judgment threshold can be set to one to two times the reference scale of the corresponding parameter. In specific engineering projects, the threshold can be fine-tuned using simulation or trial operation data to prevent normal small fluctuations from being frequently counted as abnormal and significantly disturbing the score. Optionally, when a physical link has just been connected and has not yet accumulated enough window samples, an initial stability score can be calculated with a shorter window. During the period when the number of samples has not reached the expected window length, confidence decay or its weight in path selection can be applied to the score results. Once there are enough samples, the complete stability score can be calculated according to the normal window parameters to improve the reliability of the new link evaluation results.

[0091] In one embodiment, the link prediction state is generated by a time series analysis model, which processes the real-time transmission parameter sequence to output a prediction result indicating the direction and magnitude of changes in the link quality level of the physical link within a predetermined future time period.

[0092] Specifically, in this embodiment, the time-series analysis model can be established based on the statistical results of the most recent evaluation windows in the historical real-time transmission parameter sequence, used to capture trend changes and periodic fluctuations on a minute-level time scale. For example, the model can use the average latency, packet loss rate, and stability score of each physical link in each evaluation window as input features, and output the expected direction of change in the link quality level in one or more future evaluation windows, such as determining whether the quality will significantly improve, remain basically the same, or significantly deteriorate, and the corresponding range of change. In engineering implementation, the predetermined future time period can be set to ten to tens of seconds, and in the urban public network environment, it can be appropriately shortened according to mobility and interference intensity to ensure that the prediction results still have reference value during the path selection decision period. Optionally, the parameters of the time-series analysis model can be obtained offline through simulation data or historical operating data in the early stage of system deployment, and incrementally updated at a low frequency based on new data during field operation to balance prediction accuracy and computational cost; when the historical data of a single link is insufficient to support independent modeling, the model parameters of other links in the same operator or the same area can be temporarily used as initial values, and gradually replaced with the proprietary model of that link as data accumulates.

[0093] In one embodiment, the link quality assessment further includes abnormal transmission pattern detection, which includes:

[0094] Monitor the acknowledgment delay of continuous data units transmitted through physical links to form a delay sequence;

[0095] Based on the comparison between the time delay sequence and the historical time delay distribution, abnormal time delay clusters composed of time delay values ​​that continuously exceed the preset deviation range are identified;

[0096] When an abnormal latency cluster is detected, the overall link evaluation value of the corresponding physical link is reduced.

[0097] Furthermore, for ease of engineering deployment, "continuous" in an abnormal delay cluster can be understood as the simultaneous occurrence of confirmation delays exceeding a preset deviation range at least several adjacent sampling points within the current evaluation window. The preset deviation range can be determined by adding a certain number of standard deviations to the historical delay mean, or by using a high percentile threshold from the historical distribution. For example, the system can first statistically analyze the baseline delay distribution over a period of time during the initial operation phase, considering sampling points exceeding a certain number of standard deviations from the baseline delay as abnormal candidates, and then classifying 3-5 or more consecutive abnormal candidate samples as an abnormal delay cluster. In this embodiment, the reduction processing can be achieved by multiplying the current comprehensive evaluation value by an attenuation coefficient less than one or subtracting a certain amount of deduction. The attenuation intensity can be correlated with the length or frequency of occurrence of the abnormal delay cluster, thereby gradually reducing the competitiveness of links that frequently experience abnormal delays in subsequent path selection. Optionally, to avoid excessive drop in link scores due to occasional short-term spikes, when only one short-length abnormal delay cluster occurs within a preset observation period, the maximum magnitude of a single reduction can be limited, and the comprehensive evaluation value can be gradually adjusted back based on the delay recovery situation in subsequent periods.

[0098] In one embodiment, generating a comprehensive link evaluation value based on real-time transmission parameters, link stability score, and predicted link status includes:

[0099] Evaluation weights are assigned to real-time transmission parameters, link stability scores, and link prediction status, respectively.

[0100] The evaluation weights are adaptively adjusted based on the network environment disturbance level determined by the central server. The higher the network environment disturbance level, the greater the evaluation weights corresponding to the link stability score and the link prediction status.

[0101] In one implementation, the adaptive adjustment of evaluation weights based on the level of network environmental disturbance is specified as a set of weight functions. This implementation lists a set of adaptive rules in the form of continuous functions and explains how to use these weights in the overall link evaluation value, as follows:

[0102] Step f, on the central server side, can target the physical link The comprehensive link evaluation value is expressed as a weighted combination of real-time transmission parameter score, link stability score, and link predicted state score; in this implementation, the central server determines the link evaluation value based on the current network environment disturbance level. Select evaluation weights to generate a comprehensive evaluation value:

[0103] ,

[0104] in, Indicates a physical link The comprehensive evaluation value of the link, Represents the physical link index. This indicates a network environment disturbance level of [level missing]. The evaluation weights assigned to the real-time transmission parameter scores are as follows: Indicates a physical link Real-time transmission parameter scoring, This indicates a network environment disturbance level of [level missing]. The evaluation weights assigned to the link stability score at that time. Indicates a physical link Link-level normalized stability score, This indicates a network environment disturbance level of [level missing]. The evaluation weights assigned to the link prediction state score at that time. Indicates a physical link Link prediction status score, This indicates the level of network environment disturbance determined by the central server, and the value range can be... The larger the value, the more severe the network environment disturbance;

[0105] In this embodiment, the real-time transmission parameter score can be obtained by normalizing and weighted summarizing various real-time transmission parameters within the current evaluation window, ensuring that its value range is consistent with the stability score and the predicted state score. The link predicted state score can be mapped to the scoring interval based on the direction and magnitude of quality level changes output by the time series analysis model, using a lookup table or piecewise function. Furthermore, when some scores cannot be updated in the current period due to missing data or the model not yet outputting data, the scoring results from the previous period can be used, and the transition can be gradually smoothed in subsequent periods to avoid drastic changes in the overall evaluation value due to the absence of a single scoring component.

[0106] Step g: Given the network disturbance level, a set of unnormalized weights that monotonically change with the disturbance level can be constructed first. To reflect that the higher the disturbance level, the greater the evaluation weights corresponding to the link stability score and the predicted link state, a linear amplification and attenuation approach can be adopted.

[0107] ,

[0108] ,

[0109] ,

[0110] in, This indicates a network environment disturbance level of [level missing]. The unnormalized weights of the real-time transmission parameter channel. This indicates a network environment disturbance level of [level missing]. Unnormalized weights of the time-link stability scoring channel This indicates a network environment disturbance level of [level missing]. Unnormalized weights of the time-link prediction state scoring channel. This represents the baseline weighting coefficient assigned to the real-time transmission parameter score when the network environment disturbance level is zero. This represents the baseline weighting coefficient assigned to the link stability score when the network environment disturbance level is zero. This represents the baseline weighting coefficient assigned to the link prediction state score when the network environment disturbance level is zero. This represents the environmental disturbance sensitivity coefficient, used to adjust the intensity of the impact of the disturbance level on the magnitude of changes in the three weights;

[0111] Under this structure, when When it increases, and It increases accordingly, and This decreases accordingly, which aligns with the design intent of relying more on stability and prediction information in highly perturbed environments;

[0112] Step h: To facilitate comparability between different channels, the central server can normalize the unnormalized weights to make the sum of the three evaluation weights equal to 1; in a network environment with a disturbance level of... When calculating the weights, the sum of the weights can be calculated first:

[0113] ,

[0114] in, This indicates a network environment disturbance level of [level missing]. The sum of the three unnormalized weights;

[0115] After obtaining the weights, the central server can calculate the normalized evaluation weights according to the following formula:

[0116] ,

[0117] ,

[0118] ,

[0119] in, This indicates a network environment disturbance level of [level missing]. Normalized real-time transmission parameter evaluation weights This indicates a network environment disturbance level of [level missing]. Normalized link stability score evaluation weights at that time This indicates a network environment disturbance level of [level missing]. The normalized link prediction state score evaluation weights at that time;

[0120] Through the above normalization process, the sum of the three weights can be kept constant under various disturbance levels, while also satisfying the requirement that the stability score weight and prediction score weight increase as the disturbance level increases.

[0121] In resource-constrained deployment environments, disturbance levels can be divided into several discrete levels, such as low disturbance, medium disturbance, and high disturbance. The central server can pre-configure a lookup table to assign a fixed weight combination to each disturbance level. For example, in the low disturbance level, the real-time transmission score accounts for a higher proportion; in the medium disturbance level, the three types of scores have similar proportions; and in the high disturbance level, the stability score and prediction score account for a higher proportion. At runtime, the corresponding weight combination is only looked up in the table based on the current disturbance level, avoiding real-time calculation of continuous functions and reducing the computational burden.

[0122] In this embodiment, the network environment disturbance level can be comprehensively determined by the central server based on the statistical results of real-time transmission parameters reported by all BeiDou reference stations and physical links. For example, it comprehensively considers indicators such as the network's latency variance, packet loss rate distribution, and frequency of abnormal latency clusters, mapping them to dimensionless level values ​​between zero and one, and dividing them into several discrete levels. For instance, a disturbance level less than a preset first threshold can be defined as low disturbance, a disturbance between the first and second thresholds as medium disturbance, and a disturbance higher than the second threshold as high disturbance. The threshold values ​​can be pre-configured by offline analysis of historical operating data or based on operational experience. Similarly, the central server can update the network environment disturbance level every 10 to 60 seconds. During the update, it first calculates the statistics of key performance indicators of the entire network within the most recent evaluation window, and then obtains the new level value according to the preset mapping rules, thereby reflecting the changing trends of the current electromagnetic environment and network load without introducing excessive fluctuations. Optionally, when the judgment result remains at the same disturbance level in multiple consecutive update cycles, the judgment conditions required for gear switching can be increased. For example, the gear change can be allowed only after the disturbance level exceeds a certain margin, so as to avoid frequent gear jitter caused by statistical noise.

[0123] Specifically, the above implementation method revolves around adaptive weight adjustment logic, constructing a weight adjustment mechanism with disturbance level as the core driving force. The overall idea is to first establish a comprehensive evaluation path, unifying three information sources—real-time transmission performance, historical stability statistics, and future quality prediction—onto a weighted path, and then controlling the proportion relationship among the three by the network environment disturbance level. To this end, a set of benchmark weights is introduced when the disturbance level is zero to reflect the relative importance of each information source when the network state is stable. Based on this, unnormalized weights are constructed through linear amplification and weakening, so that when the disturbance level increases, the stability-related path and the prediction-related path... The weight of a path increases accordingly, while the weight of a pathway, which directly depends on instantaneous performance, gradually decreases, which better meets the needs for robustness and foresight in complex environments. To facilitate cooperation with other modules, a normalization step is further introduced to keep the sum of the three types of weights constant under any disturbance level. This is beneficial for comparison between different links and also facilitates configuration migration between different policy versions. For scenarios with limited computing resources, the implementation method also provides an alternative to a segmented lookup table. By pre-setting a weight combination of a finite number of disturbance levels, the weight update can be completed by querying only the disturbance level during the runtime phase, thus balancing adaptability and implementation cost.

[0124] In one embodiment, the data transmission path selection process includes a link switching suppression step. When switching from the currently used link to the target link, the link switching suppression step verifies whether the difference between the comprehensive link evaluation values ​​of the target link and the currently used link meets the switching conditions. The switching conditions require that the difference between the evaluation values ​​exceeds an adaptive threshold within a continuous observation period, and the size of the adaptive threshold is negatively correlated with the link stability score of the currently used link.

[0125] For example, the observation period can be an integer multiple of the sampling period of real-time transmission parameters or the update period of the evaluation window. In practical engineering, it can be set to a period of 1 to 10 seconds to make the changes in the link's comprehensive evaluation value comparable between adjacent periods. In this embodiment, the adaptive threshold can be linearly or piecewise monotonically mapped within a preset range using the stability score of the currently used link. A higher stability score corresponds to a larger threshold to ensure that switching is only triggered when the target link is significantly better than the current link; a lower stability score corresponds to a smaller threshold, allowing the system to migrate more sensitively from links with large quality fluctuations to relatively better links. Furthermore, to avoid repeatedly triggering switching in a short period of time, this embodiment can set a minimum hold time or a minimum number of consecutive satisfying periods. If the difference in evaluation values ​​does not continuously meet the adaptive threshold requirement or the current link has not yet reached the minimum hold time, a switching decision is not executed. Optionally, when multiple candidate target links are detected to meet the switching conditions, the link with the highest comprehensive evaluation value and a security trust level not lower than the permitted threshold can be selected as the switching target, thereby balancing short-term performance and long-term reliability.

[0126] In one embodiment, the system further includes a cooperative mechanism for avoiding multi-station path selection oscillations, the cooperative mechanism including:

[0127] The central server generates and broadcasts reference link cost information to each BeiDou reference station.

[0128] After receiving the reference link cost information, each BeiDou reference station combines it with the private perturbation factor generated based on the unique identifier information of the BeiDou reference station to form a local differentiated link cost view of the BeiDou reference station.

[0129] Each BeiDou reference station participates in the initial link selection or triggers a link switching request to the central server based on its own differentiated link cost view. In this embodiment, the differentiated link cost view can be understood as superimposing a small bias derived from a private perturbation factor on each physical link based on the reference link cost broadcast by the central server. This allows different BeiDou reference stations to form slightly different ranking results when faced with the same reference cost information, thereby reducing the probability of a large number of reference stations selecting the same link at the same time. For example, the private perturbation factor can be obtained by hashing or mapping the unique identification information of the BeiDou reference station, and remains constant during system operation or is updated once over a long period of time. This ensures that the same reference station maintains a stable cost bias after restarting, facilitating problem reproduction and operation and maintenance analysis. Similarly, the central server can periodically broadcast or update the reference link cost at intervals of several seconds to tens of seconds. After receiving the data, each BeiDou reference station maps the private perturbation factor to an offset range much smaller than the change in the reference cost according to preset rules, adds it to the cost value of the corresponding link, and then performs the initial link selection or decides whether to initiate a switching request based on this. Optionally, to prevent a certain link from being excessively neglected due to cost bias over a long period of time, each BeiDou reference station can also set a maximum offset ratio locally. When the reference cost changes significantly or a certain link is not selected as the primary path for a long time, the negative offset for that link can be appropriately reduced to ensure that all links can get a certain trial opportunity in a statistical sense.

[0130] In one embodiment, the system maintains a security trust level for each physical link. When the trigger frequency of abnormal transmission mode detection for a certain physical link exceeds a preset threshold, or when the link prediction status continuously indicates that the quality of the physical link will enter a low reliability range within a preset time, the central server triggers a downgrade operation for the security trust level of the physical link and excludes physical links with a security trust level lower than the permitted threshold in the subsequent data transmission path selection process.

[0131] In this embodiment, the security trust level can be represented as a discrete level or a continuous interval. In a typical implementation, it can be divided into several levels, decreasing sequentially from the highest trust to the lowest trust. Each physical link is assigned an initial level during system initialization based on operator type, historical fault statistics, or manual evaluation results. For example, when the frequency of abnormal transmission pattern detection exceeds a threshold multiple times within a set observation window, or when the link prediction status indicates that the link quality will enter a low reliability interval for multiple consecutive evaluation windows, its security trust level can be gradually reduced by a preset step size. A minimum interval time can be set between each downgrade to avoid excessively frequent changes in trust level during short-term fluctuations. Similarly, when a link is downgraded to the lowest level or below the permitted threshold, the central server will no longer include it in the candidate set during path selection calculation, only retaining heartbeat detection or low-volume monitoring, so that its performance can be re-evaluated after the link recovers. Optionally, to prevent a one-way decline in trust level from causing the link to become unrecoverable for a long time, this embodiment allows the security trust level to be gradually increased in steps opposite to the downgrade when the link continues to perform well within a certain observation window and no new abnormal patterns appear. This ensures overall security while also taking into account the reuse of link resources.

[0132] This application proposes a method for data transmission and detection of BeiDou reference stations, combining... Figure 2 As shown, the method includes:

[0133] Step S1: Obtain the real-time transmission parameter sequence corresponding to the data sent by at least one BeiDou reference station to the central server through at least two wireless communication links;

[0134] Step S2: For each wireless communication link, calculate the link stability score and predict the link prediction status based on the real-time transmission parameter sequence, and generate a comprehensive link evaluation value based on the real-time transmission parameters, link stability score and link prediction status.

[0135] Step S3: Based on the comprehensive evaluation value of each wireless communication link, determine the data transmission path for at least one BeiDou reference station, and implement abnormal transmission mode detection, link switching suppression and multi-station coordination mechanism as needed.

[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0137] Furthermore, those skilled in the art will understand that although some embodiments herein include certain features included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of this application and form different embodiments. For example, all the embodiments above can be used in any combination. The information disclosed in this background section is intended only to enhance the understanding of the general background of this application and should not be construed as an admission or in any way implying that such information constitutes prior art known to those skilled in the art.

Claims

1. A BeiDou reference station data transmission and detection system, characterized in that, Includes at least one BeiDou reference station and a central server; The Beidou reference station is equipped with at least two wireless communication links for communicating with the central server, and each wireless communication link corresponds to a physical link identifier. The central server is configured to perform link quality assessment for each physical link and select a data transmission path for transmitting BeiDou reference station data based on the link quality assessment results of each physical link. The link quality assessment includes: Obtain the real-time transmission parameter sequence of the physical link; Based on the real-time transmission parameter sequence, a link stability score is calculated to characterize the volatility of transmission parameters; Based on the real-time transmission parameter sequence, the link quality of the physical link in a predetermined future time period is predicted to obtain the link prediction status. Based on the real-time transmission parameters, the link stability score, and the link prediction status, a comprehensive link evaluation value is generated for the physical link. The central server is configured to determine the current data transmission path for each BeiDou reference station based on the comprehensive evaluation value of each physical link, and control the BeiDou reference station to send observation data to the central server through the current data transmission path. The calculation of the link stability score includes: Within a preset evaluation window, the distribution characteristics of data points in the real-time transmission parameter sequence are analyzed; The distribution characteristics include at least the magnitude of the deviation of each data point from the corresponding trend benchmark and the frequency of such deviation. A normalized link stability score is determined based on the aforementioned distribution characteristics; A comprehensive link evaluation value is generated based on real-time transmission parameters, link stability score, and predicted link status, including: Assign evaluation weights to the real-time transmission parameters, the link stability score, and the link prediction status, respectively. The evaluation weights are adaptively adjusted based on the network environment disturbance level determined by the central server. The higher the network environment disturbance level, the greater the evaluation weights corresponding to the link stability score and the link prediction state.

2. The BeiDou reference station data transmission and detection system as described in claim 1, characterized in that, The real-time transmission parameter sequence includes a time series of at least one of the following parameters: round-trip acknowledgment delay, packet loss rate, jitter, effective throughput, and number of retransmissions.

3. The BeiDou reference station data transmission and detection system as described in claim 2, characterized in that, The predicted link status is generated by a time-series analysis model, which processes the real-time transmission parameter sequence to output a prediction result indicating the direction and magnitude of the change in the link quality level of the physical link within the predetermined future time period.

4. The BeiDou reference station data transmission and detection system as described in claim 3, characterized in that, The link quality assessment also includes abnormal transmission pattern detection, which includes: The acknowledgment delay of consecutive data units transmitted through the physical link is monitored to form a delay sequence; Based on the comparison between the time delay sequence and the historical time delay distribution, abnormal time delay clusters composed of time delay values ​​that continuously exceed a preset deviation range are identified; When the abnormal latency cluster is detected, the overall link evaluation value of the corresponding physical link is reduced.

5. The BeiDou reference station data transmission and detection system as described in claim 4, characterized in that, The data transmission path selection process includes a link switching suppression step. When switching from the currently used link to the target link, the link switching suppression step verifies whether the difference between the comprehensive link evaluation value of the target link and the currently used link meets the switching conditions. The switching conditions require that the difference between the evaluation values ​​exceeds an adaptive threshold within a continuous observation period, and the size of the adaptive threshold is negatively correlated with the link stability score of the currently used link.

6. The BeiDou reference station data transmission and detection system as described in claim 5, characterized in that, The system also includes a cooperative mechanism for avoiding multi-station path selection oscillations, the cooperative mechanism including: The central server generates and broadcasts reference link cost information to each BeiDou reference station. After receiving the reference link cost information, each BeiDou reference station combines the private disturbance factor generated based on the unique identifier information of the BeiDou reference station to form a local differentiated link cost view of the BeiDou reference station. Each BeiDou reference station participates in the initial link selection or triggers a link switching request to the central server based on its own differentiated link cost view.

7. The BeiDou reference station data transmission and detection system as described in claim 6, characterized in that, The system maintains a security trust level for each physical link. When the trigger frequency of abnormal transmission mode detection for a certain physical link exceeds a preset threshold, or when the link prediction status continuously indicates that the quality of the physical link will enter a low reliability range within a preset time, the central server triggers a downgrade operation on the security trust level of the physical link and excludes physical links with a security trust level lower than the permitted threshold in the subsequent data transmission path selection process.

8. A method for data transmission and detection of BeiDou reference stations, based on the BeiDou reference station data transmission and detection system according to any one of claims 1 to 7, characterized in that, include: Step S1: Obtain the real-time transmission parameter sequence corresponding to the data sent by at least one BeiDou reference station to the central server through at least two wireless communication links; Step S2: For each wireless communication link, calculate the link stability score and predict the link prediction state based on the real-time transmission parameter sequence, and generate a comprehensive link evaluation value based on the real-time transmission parameters, the link stability score and the link prediction state. Step S3: Based on the comprehensive evaluation value of each wireless communication link, determine the data transmission path for each of the at least one BeiDou reference stations, and implement abnormal transmission mode detection, link switching suppression and multi-station coordination mechanism as needed.

Images