TDOA positioning base station selection method and device, equipment, storage medium and computer program
By extracting the amplitude and time diffusion characteristics of candidate base stations and evaluating them using the AdaBoost model, the problem of signal propagation uncertainty of TDOA positioning system in urban environments is solved, and the reliability and positioning accuracy of base station selection are improved.
Patent Information
- Application Number
- CN202411126529.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-12-03
Smart Images

Figure CN119071725A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of wireless technology, and in particular to a TDOA positioning base station selection method, apparatus, device, storage medium and computer program. Background Art
[0002] Time Difference of Arrival (TDOA) positioning is a passive positioning technology widely used in wireless communications, the Internet of Things, and navigation systems. It determines the location of a target by measuring the time difference between a signal from a transmitting source to multiple base stations. The accuracy and reliability of TDOA positioning largely depends on the layout and performance of the selected base stations and the signal propagation characteristics between them and the target device. Therefore, choosing a suitable base station combination is crucial to improving the overall performance of the TDOA positioning system.
[0003] At present, most existing TDOA positioning algorithms assume that the signal propagation path is clear and unobstructed, that is, the ideal LOS (Line of Sight) condition. However, in actual applications, that is, in urban environments, buildings, vegetation, and even weather conditions can cause serious multipath effects, forming NLOS (Non-Line of Sight) propagation, which increases the delay of signal propagation time and has uncertainty, resulting in a decrease in the reliability of the base station selected based on TDOA positioning, and thus reduces the positioning accuracy of the base station.
[0004] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention
[0005] The main purpose of this application is to provide a TDOA positioning base station selection method, which aims to solve the technical problem that the reliability of the base station selected according to TDOA positioning is reduced due to the increased and uncertain signal propagation time delay, thereby reducing the positioning accuracy of the base station.
[0006] To achieve the above purpose, the present application proposes a TDOA positioning base station selection method, the method comprising:
[0007] Extract amplitude dispersion features and time dispersion features from the collected RRS data and TRS data of each candidate base station;
[0008] Inputting the amplitude dispersion feature and the time dispersion feature into a positioning base station prediction model to obtain a prediction result of each candidate base station as a positioning base station, wherein the positioning base station prediction model is obtained by inputting the time dispersion feature and the amplitude dispersion feature extracted from the sample data into an AdaBoost model for training;
[0009] Performing confidence evaluation on the prediction result to obtain the confidence of the prediction result corresponding to each candidate base station;
[0010] Based on the confidence, a base station that meets a preset confidence threshold requirement is selected from the candidate base stations as a positioning base station.
[0011] In one embodiment, the step of extracting amplitude dispersion features and time dispersion features from the collected RRS data and TRS data of each candidate base station includes:
[0012] Preprocessing the collected RRS data and TRS data of each candidate base station;
[0013] Align the timestamps of the preprocessed RRS data with the TRS data;
[0014] Determine a cross-correlation value between the RRS data and the TRS data based on the acquisition time points of the RRS data and the TRS data after the timestamps are aligned;
[0015] Based on the cross-correlation value, the amplitude dispersion feature and the time dispersion feature are determined.
[0016] In one embodiment, the amplitude dispersion characteristics include amplitude dispersion energy, amplitude dispersion maximum value, amplitude dispersion mean value, amplitude dispersion variance, amplitude dispersion kurtosis and amplitude dispersion skewness, and the step of determining the amplitude dispersion characteristics based on the cross-correlation value includes:
[0017] Determining the amplitude dispersion energy of the signal power distribution in the RRS data and the TRS data based on the sum of the squares of the cross-correlation values at each of the acquisition time points;
[0018] Filtering the maximum amplitude dispersion value corresponding to the target time point with the maximum cross-correlation value from the RRS data and the TRS data;
[0019] Based on the cross-correlation values at each of the acquisition time points, averaging the signal power to determine an amplitude dispersion mean of the signal power;
[0020] Based on the amplitude dispersion mean and the corresponding cross-correlation value, performing a fluctuation degree analysis on the signal power to determine the amplitude dispersion variance of the signal power;
[0021] Based on the amplitude dispersion mean, the corresponding amplitude dispersion variance and the cross-correlation value, performing a kurtosis analysis on the signal power to determine the amplitude dispersion kurtosis of the signal power;
[0022] Based on the amplitude dispersion mean, the corresponding amplitude dispersion variance and the cross-correlation value, the signal power is analyzed for asymmetry to determine the amplitude dispersion deviation of the signal power.
[0023] In one embodiment, the time dispersion feature includes a time dispersion mean, a time dispersion variance, a time dispersion kurtosis, and a time dispersion skewness, and the step of determining the time dispersion feature based on the cross-correlation value includes:
[0024] Normalizing the cross-correlation values at each acquisition time point to obtain a normalized cross-correlation at each acquisition time point;
[0025] Based on the normalized cross-correlation, averaging the signal arrival time differences in the RRS data and the TRS data to obtain a time dispersion mean of the arrival time differences;
[0026] Based on the time dispersion mean and each of the normalized cross-correlations, performing a fluctuation degree analysis on the arrival time difference to determine the time dispersion variance of the arrival time difference;
[0027] Based on the time dispersion mean, the time variance and each of the normalized cross-correlations, performing a kurtosis analysis on the arrival time difference to determine the time dispersion kurtosis of the arrival time difference;
[0028] Based on the time dispersion mean, the time variance and each of the normalized cross-correlations, an asymmetry degree analysis is performed on the arrival time difference to determine the time dispersion skewness of the arrival time difference.
[0029] In one embodiment, the positioning base station prediction model is obtained based on the following steps:
[0030] Collect RRS data and TRS data of each sample base station;
[0031] Extracting time dispersion features and amplitude dispersion features from the RRS data and TRS data of each sample base station;
[0032] Iteratively training the preselected weak classifiers based on the time diffusion feature and the amplitude diffusion feature to obtain the weights of each of the weak classifiers;
[0033] Based on the weights, the weak classifiers are combined into a strong classifier to obtain the positioning base station prediction model.
[0034] In one embodiment, the step of performing confidence evaluation on the prediction result to obtain the confidence of the prediction result corresponding to each candidate base station includes:
[0035] Calling the weights corresponding to the weak classifiers in the positioning base station prediction model and the reference results of the weak classifiers predicting each candidate base station;
[0036] Based on the reference results and the corresponding weights, the confidence level of the prediction results is evaluated to obtain a final decision value of the prediction results corresponding to each candidate base station;
[0037] The final decision value is mapped and converted to obtain the confidence level of the prediction result corresponding to each candidate base station.
[0038] In addition, to achieve the above purpose, the present application also proposes a TDOA positioning base station selection device, the TDOA positioning base station selection device comprising:
[0039] An extraction module is used to extract amplitude dispersion features and time dispersion features from the collected RRS data and TRS data of each candidate base station;
[0040] A prediction module, used for inputting the amplitude dispersion feature and the time dispersion feature into a positioning base station prediction model to obtain a prediction result of each candidate base station as a positioning base station, wherein the positioning base station prediction model is obtained by inputting the time dispersion feature and the amplitude dispersion feature extracted from the sample data into an AdaBoost model for training;
[0041] An evaluation module, used to perform confidence evaluation on the prediction result to obtain the confidence of the prediction result corresponding to each candidate base station;
[0042] The selection module is used to select a base station that meets a preset confidence threshold requirement from each of the candidate base stations based on the confidence as a positioning base station.
[0043] In addition, to achieve the above-mentioned purpose, the present application also proposes a TDOA positioning base station selection device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, and the computer program is configured to implement the steps of the TDOA positioning base station selection method as described above.
[0044] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the TDOA positioning base station selection method described above are implemented.
[0045] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the TDOA positioning base station selection method as described above.
[0046] One or more technical solutions proposed in this application have at least the following technical effects:
[0047] This application extracts amplitude dispersion features and time dispersion features from RRS data and TRS data to form a standardized feature vector, and then uses a positioning base station prediction model to select and predict the base station based on the feature vector. The positioning base station prediction model is obtained by inputting the time dispersion features and amplitude dispersion features extracted from sample data into the AdaBoost model for training. Since the AdaBoost model effectively improves the prediction accuracy in complex scenarios by integrating weak classifiers, and combines the amplitude dispersion and time dispersion features, as well as the base station working status parameters, the generated standardized feature vector more comprehensively reflects the characteristics of the TDOA base station. The positioning base station prediction model obtained by training with the AdaBoost model also has the characteristics of the AdaBoost model. The algorithm in the AdaBoost model is used for classification prediction. The strong learning ability and low complexity of the algorithm make the base station selection in complex environments more intelligent and accurate. Compared with traditional algorithms, the prediction effect is greatly improved and the calculation cost is greatly reduced. In order to ensure that the contribution of each candidate base station can be combined when finally selecting the base station, after the positioning base station prediction model outputs the prediction result, the prediction result is evaluated for confidence, so as to identify and give priority to those base stations with high positioning contribution and reliable prediction results, so as to make more reliable base station selection decisions in complex environments with strong uncertainty, effectively identify and eliminate low-quality prediction results, reduce the possibility of misselecting base stations, improve the reliability of base stations selected according to TDOA positioning, and thus improve the positioning accuracy of base stations. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.
[0049] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.
[0050] Figure 1 A flowchart of the first embodiment of the TDOA positioning base station selection method of the present application is provided;
[0051] Figure 2 A schematic diagram of a brief flow chart of a TDOA positioning base station selection method provided in Example 1 of the present application;
[0052] Figure 3 A flow chart of the second embodiment of the TDOA positioning base station selection method provided in this application;
[0053] Figure 4A flowchart of the third embodiment of the TDOA positioning base station selection method of the present application is provided;
[0054] Figure 5 A schematic diagram of the AdaBoost model training process provided in Example 3 of the present application;
[0055] Figure 6 This is a schematic diagram of the module structure of the TDOA positioning base station selection device according to an embodiment of the present application;
[0056] Figure 7 This is a schematic diagram of the device structure of the hardware operating environment involved in the TDOA positioning base station selection method in the embodiment of the present application.
[0057] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0058] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.
[0059] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.
[0060] The main solution of the embodiment of the present application is: extracting the amplitude dispersion characteristics reflecting the signal power and the time dispersion characteristics reflecting the arrival time difference of the signal from the RRS data and TRS data collected from each candidate base station through the TDOA positioning device; inputting the amplitude dispersion characteristics and the time dispersion characteristics into the positioning base station prediction model to obtain the prediction results of each candidate base station as a positioning base station, and the positioning base station prediction model is obtained by inputting the time dispersion characteristics and amplitude dispersion characteristics extracted from the sample data into the AdaBoost model for training; performing confidence evaluation on the prediction results to obtain the confidence of the prediction results corresponding to each candidate base station; based on the confidence, selecting a base station that meets the preset confidence threshold requirements from each candidate base station as a positioning base station.
[0061] In this embodiment, for the convenience of description, the following description is made with the TDOA positioning device as the execution entity.
[0062] Most existing technologies assume that the signal propagation path is clear and unobstructed, that is, the ideal LOS condition. However, in actual applications, that is, in urban environments, buildings, vegetation, and even weather conditions can cause serious multipath effects, forming NLOS propagation, increasing the signal propagation time delay and uncertainty, resulting in a reduction in the reliability of the base station selected based on TDOA positioning, and thus reducing the positioning accuracy of the base station.
[0063] The present application provides a solution, which extracts amplitude dispersion features and time dispersion features from RRS data and TRS data to form a standardized feature vector, and then uses a positioning base station prediction model to select and predict the base station based on the feature vector. The positioning base station prediction model is obtained by inputting the time dispersion features and amplitude dispersion features extracted from sample data into an AdaBoost model for training. Since the AdaBoost model effectively improves the prediction accuracy in complex scenarios by integrating weak classifiers, and combines the amplitude dispersion and time dispersion features, as well as the base station working status parameters, the generated standardized feature vector more comprehensively reflects the characteristics of the TDOA base station. The positioning base station prediction model obtained by training with the AdaBoost model also has the characteristics of the AdaBoost model. The algorithm in the AdaBoost model is used for classification prediction. The strong learning ability and low complexity of the algorithm make the base station selection in complex environments more intelligent and accurate. Compared with traditional algorithms, the prediction effect is greatly improved and the calculation cost is greatly reduced. In order to ensure that the contribution of each candidate base station can be combined when finally selecting the base station, after the positioning base station prediction model outputs the prediction result, the prediction result is evaluated for confidence, so as to identify and give priority to those base stations with high positioning contribution and reliable prediction results, so as to make more reliable base station selection decisions in complex environments with strong uncertainty, effectively identify and eliminate low-quality prediction results, reduce the possibility of misselecting base stations, improve the reliability of base stations selected according to TDOA positioning, and thus improve the positioning accuracy of base stations.
[0064] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a TDOA positioning device, etc. The following takes the TDOA positioning device as an example to illustrate this embodiment and the following embodiments.
[0065] Based on this, the embodiment of the present application provides a TDOA positioning base station selection method, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the TDOA positioning base station selection method of the present application.
[0066] In this embodiment, the TDOA positioning base station selection method includes steps S10 to S40:
[0067] Step S10, extracting amplitude dispersion features and time dispersion features from the collected RRS data and TRS data of each candidate base station;
[0068] It should be noted that the RRS (Received Reference Signal) data is the base station downlink signal received by the user equipment (UE, User Experience), including the characteristics of information such as timestamp, signal strength, channel quality indication, reference signal received power, and reference signal received quality. Among them, the collection method of RRS data can be UE reporting, using the built-in measurement report function of the UE to report the received RRS data to the network side regularly or by event triggering. A monitoring module can also be deployed in network equipment (such as eNB, gNB) to capture downlink RRS data in real time. The monitoring module must have the ability to process signals of different standards, frequency bands, and beams to ensure the comprehensiveness of the data. Among them, the reported content includes UE identification, base station identification, RRS measurement results, etc., which can be transmitted through standard protocols (such as LTE RRC measurement report, NR RLM measurement report). TRS (Transmitted Reference Signal) data is the base station uplink signal received by the base station, including timestamp, signal power, signal quality, channel state information, reference signal sequence and other information. Among them, a data recording module is configured inside the base station to automatically collect and store TRS data. The recording module should have the ability to uniformly manage and process TRS data of different UEs and different uplink resource scheduling modes, and use the synchronization mechanism between network devices, such as GPS synchronization, to ensure the time alignment of TRS data between different base stations. The network side synchronously obtains TRS data from each base station through control signaling or data interface. Amplitude dispersion can be a phenomenon in which the signal amplitude fluctuates and diffuses randomly due to the multipath propagation effect of the channel in the communication system. The amplitude dispersion characteristics reflect the signal power, which can be amplitude dispersion energy, amplitude dispersion maximum value, amplitude dispersion mean, amplitude dispersion variance, amplitude dispersion kurtosis, and amplitude dispersion skewness. Time dispersion can be a phenomenon in which the signal is diffused and overlapped in time at the receiving end due to the influence of multipath propagation in the communication system. The time dispersion characteristics reflect the arrival time difference of the signal, which can be the time dispersion mean, time dispersion variance, time dispersion kurtosis, and time dispersion skewness.
[0069] It can be understood that, due to the TDOA positioning technology, the RRS data and TRS data of the base station contain rich spatiotemporal information, and the positioning performance of these spatiotemporal information is crucial to determining the contribution of the base station to positioning. Therefore, step S10 is performed to extract the time dispersion characteristics and amplitude dispersion characteristics that can effectively characterize the positioning performance and contribution of the candidate base station from the RRS data and TRS data, so as to directly reflect the complexity of the wireless environment, multipath effects, signal propagation stability and other factors, thereby providing a basis for the subsequent selection of base stations through the model.
[0070] Step S20, inputting the amplitude dispersion feature and the time dispersion feature into a positioning base station prediction model to obtain a prediction result of each candidate base station as a positioning base station, wherein the positioning base station prediction model is obtained by inputting the time dispersion feature and the amplitude dispersion feature extracted from the sample data into an AdaBoost model for training;
[0071] It should be noted that the AdaBoost (Adaptive Boosting) algorithm is an adaptive boosting ensemble learning method that aims to combine multiple weak learners into a strong learner to improve prediction accuracy. The AdaBoost model is a machine learning model trained based on the AdaBoost algorithm. The prediction result can be the result of the positioning base station prediction model weighted summation of the prediction memory of each weak classifier based on the amplitude time dispersion and time dispersion features.
[0072] In addition, it should be noted that a weak classifier can be a classifier whose classification accuracy is only slightly higher than random guessing in ensemble learning. Weak classifiers usually have limited performance when used alone, but can significantly improve prediction accuracy after being combined through ensemble methods. The accuracy of weak classifiers is usually between 60% and 80%. A strong classifier is a machine learning model with high classification accuracy, and its classification accuracy is usually above 90%.
[0073] It can be understood that the amplitude dispersion features and time dispersion features are classified and predicted using the positioning base station prediction model, which is trained by the AdaBoost model. The AdaBoost model can effectively improve the prediction accuracy in complex scenarios by integrating weak classifiers.
[0074] In the specific implementation, the standardized amplitude dispersion characteristics and time dispersion characteristics of the candidate base stations are first input into the trained positioning base station prediction model to obtain the prediction results of each candidate base station. Then the confidence of the prediction results of each base station is calculated.
[0075] Step S30, performing confidence evaluation on the prediction result to obtain the confidence of the prediction result corresponding to each candidate base station;
[0076] It should be noted that the confidence level can be the degree to which the population parameter value falls around the sample statistical value measurement result.
[0077] It is understandable that the prediction results output by the positioning base station prediction model may contain low-quality predictions. Since low-quality predictions cannot reduce the possibility of misselecting base stations, in order to reduce the misselection of candidate base stations, the prediction results can be confidence evaluated to select candidate base stations with higher confidence from the prediction results output by the positioning base station prediction model, so as to provide more information for subsequent base station selection strategies in scenarios with changing signal environments.
[0078] In the specific implementation, each iteration of the AdaBoost algorithm trains a weak classifier and adjusts the sample weights according to the error rate of the previous round. Each weak classifier has a weight that indicates its importance in the final integrated classifier. The final prediction is the result of the weighted sum of the prediction results of all weak classifiers. Therefore, the prediction confidence of each candidate base station can be estimated based on this information.
[0079] The specific process is: Weak classifier weight: For the tth weak classifier, record its weight α t , which is determined during the training process based on the performance of the classifier on the training set; record the cumulative decision value of each candidate base station after passing through all weak classifiers. For each base station, its total decision value is the sum of all weak classifiers D t The sum of the prediction result of the base station and its weight product; finally, the absolute value mapping is used to convert the added decision value into a confidence score.
[0080] It should be noted that the calculation formula for the total decision value is:
[0081]
[0082] The confidence calculation formula is:
[0083]
[0084] Among them, T means there are T weak classifiers, and the confidence formula maps any real number to (0,1). The closer the value is to 1, the more accurate the prediction is.
[0085] It can be understood that using the output of the positioning base station prediction model to calculate the confidence of each base station prediction result is conducive to formulating efficient base station classification and selection rules.
[0086] Step S40: Based on the confidence level, a base station that meets a preset confidence threshold requirement is selected from the candidate base stations as a positioning base station.
[0087] It should be noted that the preset confidence threshold data may be updated regularly or on demand based on information such as actual positioning results, network status changes, and user feedback.
[0088] In the specific implementation, refer to Figure 2, Step 1: Get the reference signal sample RRS data and TRS data, and clean and time-align the RRS data and TRS data; Step 2: Since the time dispersion characteristics and amplitude dispersion characteristics of the base station contain features that are strongly related to the NLOS propagation conditions, the key features of the sample data are extracted from the RRS data and TRS data; Step 3: Train a reliable machine learning model; Step 4: Set a threshold, and classify the base stations whose prediction confidence exceeds the threshold as "optional base stations", and the rest as "unfavorable base stations". Select the base station with the highest prediction confidence from the "optional base stations" to obtain the TDOA base station selection result.
[0089] It is understandable that by combining the collaborative work of UE and network side, not only the data source is broadened, but also the high quality and consistency of data samples in complex environments are ensured through data cleaning and time alignment technology. This method enhances the resistance to the influence of multipath effects and provides a solid foundation for subsequent processing.
[0090] This embodiment provides a TDOA positioning base station selection method, which extracts amplitude dispersion features and time dispersion features from RRS data and TRS data to form a standardized feature vector, and then uses a positioning base station prediction model to select and predict the base station based on the feature vector. The positioning base station prediction model is obtained by inputting the time dispersion features and amplitude dispersion features extracted from sample data into an AdaBoost model for training. Since the AdaBoost model effectively improves the prediction accuracy in complex scenarios by integrating weak classifiers, and combines the amplitude dispersion and time dispersion features, as well as the base station working status parameters, the generated standardized feature vector more comprehensively reflects the characteristics of the TDOA base station. The positioning base station prediction model obtained by training with the AdaBoost model also has the characteristics of the AdaBoost model. The algorithm in the AdaBoost model is used for classification prediction. The strong learning ability and low complexity of the algorithm make the base station selection in complex environments more intelligent and accurate. Compared with traditional algorithms, the prediction effect is greatly improved and the calculation cost is greatly reduced. In order to ensure that the contribution of each candidate base station can be combined when finally selecting the base station, after the positioning base station prediction model outputs the prediction result, the prediction result is evaluated for confidence, so as to identify and give priority to those base stations with high positioning contribution and reliable prediction results, so as to make more reliable base station selection decisions in complex environments with strong uncertainty, effectively identify and eliminate low-quality prediction results, reduce the possibility of misselecting base stations, improve the reliability of base stations selected according to TDOA positioning, and thus improve the positioning accuracy of base stations.
[0091] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. Figure 3 , step S10 also includes steps S11 to S14:
[0092] Step S11, pre-processing the collected RRS data and TRS data of each candidate base station;
[0093] It should be noted that preprocessing may include data cleaning and checking data integrity.
[0094] In the specific implementation, the collected RRS and TRS data are preliminarily cleaned to remove invalid, erroneous, and duplicate data records, such as measurement values beyond the physical range, reference signals that are not successfully decoded, etc. At the same time, the integrity of the reported data is checked to ensure that key fields are not omitted.
[0095] Step S12, aligning the timestamps of the pre-processed RRS data and TRS data;
[0096] It is understandable that since RRS data and TRS data are received or sent at the same time, and there cannot be a time difference in the modules that collect data, there will be a time difference in the data collection received or sent at the same time. In order to eliminate the impact of this time difference on the prediction, it is necessary to align the timestamps of RRS data and TRS data.
[0097] In the specific implementation, the RRS and TRS data are accurately aligned according to their respective timestamps to ensure that the RRS and TRS data at the same time can be associated. For data with time offset, interpolation, smoothing and other methods are used to correct it.
[0098] Step S13, determining a cross-correlation value between the RRS data and the TRS data based on the acquisition time points of the RRS data and the TRS data after the timestamps are aligned;
[0099] It should be noted that the cross-correlation value may be a numerical representation for measuring the similarity between the RRS data and the TRS data.
[0100] In the specific implementation, let N s is the total number of samples in the signal waveform, r j [n] and s j [n] represent the RRS and TRS sampling data exchanged between the positioning target and the jth base station, then r j [n] and s j The positive cross-correlation coefficient of [n] is:
[0101]
[0102] Where n = 1, 2, ..., N s -1, |·| means taking the absolute value, and k means the current base sample point.
[0103] Step S14: determining an amplitude dispersion feature reflecting signal power and a time dispersion feature reflecting a time difference of arrival of the signal based on the cross-correlation value.
[0104] It is understandable that the time dispersion and amplitude dispersion characteristics of wireless signals, such as mean, variance, kurtosis and skewness, directly reflect factors such as the complexity of the wireless environment, multipath effects, and signal propagation stability. These factors have a significant impact on the TDOA positioning accuracy. Since the relationship between the transmitted signal and the received signal is a necessary condition for the ranging algorithm to estimate TDOA, the relationship (cross-correlation) between the transmitted signal and the received signal is usually used in the positioning algorithm to extract key features reflecting the amplitude dispersion and time dispersion characteristics of the wireless signal from the RRS and TRS data, providing high-quality feature input for subsequent model training or base station selection.
[0105] Furthermore, step S41 further includes:
[0106] Determining the amplitude dispersion energy of the signal power distribution in the RRS data and the TRS data based on the sum of the squares of the cross-correlation values at each of the acquisition time points;
[0107] It is understandable that by determining the amplitude dispersion energy of the signal power distribution or amplitude distribution in the RRS data and the TRS data, the stability of the signal during propagation can be reflected. Therefore, the amplitude dispersion energy can be calculated based on the square sum of the cross-correlation values at each acquisition time point.
[0108] Specifically, the calculation formula of amplitude dispersion energy is:
[0109]
[0110] Here, R[m] is the cross-correlation value when the sampling point is represented by the time point.
[0111] Filtering the maximum amplitude dispersion value corresponding to the target time point with the maximum cross-correlation value from the RRS data and the TRS data;
[0112] It can be understood that by determining the maximum deviation value of the signal power distribution or amplitude distribution, the degree of noise interference on the signal can be reflected. Therefore, the maximum amplitude dispersion value corresponding to the target time point with the maximum cross-correlation value can be screened from the RRS data and the TRS data.
[0113] Specifically, the calculation formula for the maximum amplitude dispersion is:
[0114] ν a =max m R[m].
[0115] Based on the cross-correlation values at each of the acquisition time points, averaging the signal power to determine an amplitude dispersion mean of the signal power;
[0116] It can be understood that the average value of the signal power or amplitude can reflect the average receiving strength of the signal. Therefore, the signal power can be averaged to determine the amplitude dispersion mean.
[0117] Specifically, the calculation formula for the amplitude dispersion mean is:
[0118]
[0119] Based on the amplitude dispersion mean and the corresponding cross-correlation value, performing a fluctuation degree analysis on the signal power to determine the amplitude dispersion variance of the signal power;
[0120] It can be understood that the variance of the signal power or amplitude can reflect the degree of fluctuation of the signal reception strength. Therefore, the degree of fluctuation of the signal power can be analyzed to determine the amplitude dispersion variance of the signal power.
[0121] Specifically, the calculation formula of amplitude dispersion variance is:
[0122]
[0123] Based on the amplitude dispersion mean, the corresponding amplitude dispersion variance and the cross-correlation value, performing a kurtosis analysis on the signal power to determine the amplitude dispersion kurtosis of the signal power;
[0124] It can be understood that the kurtosis of the signal power or amplitude can reflect the sharpness of the signal reception strength distribution. Therefore, the signal power can be subjected to a sharpness analysis to determine the amplitude dispersion kurtosis of the signal power.
[0125] Specifically, the calculation formula of amplitude dispersion kurtosis is:
[0126]
[0127] Based on the amplitude dispersion mean, the corresponding amplitude dispersion variance and the cross-correlation value, the signal power is analyzed for asymmetry to determine the amplitude dispersion deviation of the signal power.
[0128] It can be understood that the skewness of the signal power or amplitude can reflect the asymmetry of the signal reception strength distribution. Therefore, the asymmetry degree of the signal power can be analyzed to determine the amplitude dispersion skewness of the signal power.
[0129] Specifically, the calculation formula of amplitude dispersion skewness is:
[0130]
[0131] Furthermore, step S41 further includes:
[0132] Normalizing the cross-correlation values at each acquisition time point to obtain a normalized cross-correlation at each acquisition time point;
[0133] In the specific implementation, before obtaining the time diffusion feature, it is necessary to normalize , that is:
[0134]
[0135] Based on the normalized cross-correlation, averaging the signal arrival time differences in the RRS data and the TRS data to obtain a time dispersion mean of the arrival time differences;
[0136] It can be understood that the average value of the signal arrival time difference in RRS and TRS data can reflect the average delay of signal propagation. Therefore, the time dispersion mean of the arrival time difference can be obtained by averaging the signal arrival time difference in RRS data and TRS data.
[0137] Specifically, the calculation formula for the time diffusion mean is:
[0138]
[0139] Based on the time dispersion mean and each of the normalized cross-correlations, performing a fluctuation degree analysis on the arrival time difference to determine the time dispersion variance of the arrival time difference;
[0140] It can be understood that the variance of the arrival time difference can reflect the fluctuation degree of the signal propagation delay. Therefore, the time dispersion variance of the arrival time difference can be determined by analyzing the fluctuation degree of the arrival time difference.
[0141] Specifically, the calculation formula of time dispersion variance is:
[0142]
[0143] Based on the time dispersion mean, the time variance and each of the normalized cross-correlations, performing a kurtosis analysis on the arrival time difference to determine the time dispersion kurtosis of the arrival time difference;
[0144] It can be understood that the kurtosis of the arrival time difference can reflect the sharpness of the signal propagation delay distribution. Therefore, the time dispersion kurtosis of the arrival time difference can be determined by performing a sharpness analysis on the arrival time difference.
[0145] Specifically, the calculation formula of time dispersion kurtosis is:
[0146]
[0147] Based on the time dispersion mean, the time variance and each of the normalized cross-correlations, an asymmetry degree analysis is performed on the arrival time difference to determine the time dispersion skewness of the arrival time difference.
[0148] It can be understood that the skewness of the arrival time difference can reflect the asymmetry of the signal propagation delay distribution. Therefore, the time dispersion skewness of the arrival time difference can be determined by analyzing the asymmetry of the arrival time difference.
[0149] Specifically, the calculation formula of time dispersion skewness is:
[0150]
[0151] It is understandable that in order to improve the stability and robustness of the features, feature smoothing technology and feature normalization technology can be used to process the time diffusion feature and amplitude diffusion feature. Specifically, the extracted time diffusion and amplitude diffusion features are smoothed, such as moving average, exponential smoothing, Kalman filtering, etc., to eliminate short-term noise and improve the stability of the features, and all extracted features are normalized, such as the minimum-maximum normalization method, z-score standardization, etc., to ensure that different features have the same scale, which is conducive to subsequent model training.
[0152] Based on the first and second embodiments of the present application, in the third embodiment of the present application, the same or similar contents as those in the above embodiments can be referred to the above introduction, and will not be described in detail later. Figure 4 Before step S10, the TDOA positioning base station selection method further includes steps S1 to S4:
[0153] Step S1, collecting RRS data and TRS data of each sample base station;
[0154] Step S2, extracting time dispersion features and amplitude dispersion features from the RRS data and TRS data of each sample base station;
[0155] Step S3, iteratively training the pre-selected weak classifiers based on the time diffusion feature and the amplitude diffusion feature to obtain the weights of each of the weak classifiers;
[0156] Step S4: Based on the weights, the weak classifiers are combined into a strong classifier to obtain a positioning base station prediction model.
[0157] Understandably, since most of the existing data-driven machine learning solutions rely on high computing power, which is limited in pico base station devices, the AdaBoost algorithm is used to achieve efficient selection of TDOA base stations. Compared with other machine learning techniques, the main advantage of the AdaBoost algorithm is that it can build a simple and low-complexity model with an inherent probabilistic interpretation.
[0158] In the specific implementation, the time dispersion and amplitude dispersion features extracted from the RRS data and TRS data collected from each sample base station can be used to train a classification model that can effectively distinguish the positioning performance and contribution of base stations. For details, refer to Figure 5 First, according to the characteristic characteristics of RRS data and TRS data and actual needs, select a suitable weak classifier as the basic model of the AdaBoost model, and then set the number of iterations of the AdaBoost model, that is, the number of weak classifiers, which can be dynamically adjusted according to the number of samples, and set the learning rate of the AdaBoost model to control the update amplitude of the sample weight after each iteration to initialize the model and configure the initialization parameters, and then input the training data set containing time dispersion features and amplitude dispersion features, where this data set consists of N samples, and each sample is given the same weight.
[0159] That is:
[0160] For each weak classifier, train a weak classifier f t (x), and calculate its error rate on the training sample set:
[0161]
[0162] According to the error rate e t Calculate the weak classifier weights:
[0163]
[0164] Then update the sample weights
[0165]
[0166] where z t is the normalization factor, which is the sum of all sample weights and is calculated as follows
[0167]
[0168] The weights of the samples and the weights of the weak classifiers are iteratively updated, and then the weak classifiers trained in each iteration are combined to obtain a strong classifier (positioning base station prediction model).
[0169] That is,
[0170] It is understandable that in order to ensure the performance of the positioning base station prediction model, cross-validation methods, such as k-fold cross-validation, can also be used to evaluate the performance of the positioning base station prediction model on different training set partitions, and then adjust the hyperparameters of the positioning base station prediction model based on the cross-validation results, such as the number of iterations, learning rate, etc., to optimize the overall performance of the model, thereby improving the classification ability and robustness of the positioning base station prediction model.
[0171] Furthermore, step S30 may also include:
[0172] Calling the weights corresponding to the weak classifiers in the positioning base station prediction model and the reference results of the weak classifiers predicting each candidate base station;
[0173] Based on the reference results and the corresponding weights, the confidence level of the prediction results is evaluated to obtain a final decision value of the prediction results corresponding to each candidate base station;
[0174] The final decision value is mapped and converted to obtain the confidence level of the prediction result corresponding to each candidate base station.
[0175] It can be understood that at the output of the model, a quantitative assessment of the positioning contribution of each base station is provided through the prediction confidence calculation method, which enables the identification and priority of candidate base stations with high positioning contribution and reliable prediction results, thereby making more informed base station selection decisions in complex environments with strong uncertainty.
[0176] In the specific implementation, the standardized candidate base station feature vector is first input into the trained positioning base station prediction model to obtain the prediction result of each base station. Then the confidence of the prediction result of each base station is calculated to provide more information for the subsequent base station selection strategy.
[0177] Each iteration of the AdaBoost algorithm trains a weak classifier and adjusts the sample weights based on the error rate of the previous round. Each weak classifier has a weight that indicates its importance in the final integrated classifier. The final prediction is the result of the weighted sum of the prediction results of all weak classifiers. Therefore, the prediction confidence of each candidate base station can be estimated based on this information, and then a threshold is set to classify the base stations with prediction confidence exceeding the threshold as "optional base stations" and the remaining base stations as "unfavorable base stations". From the "optional base stations", the base stations with the highest prediction confidence can be selected. Among them, based on the actual positioning effect, network status changes, user feedback and other information, the base station feature vector is updated regularly or on demand, the positioning base station prediction model is retrained, and the base station selection strategy is adjusted.
[0178] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the TDOA positioning base station selection method of the present application. More simple transformations based on this technical concept are all within the protection scope of the present application.
[0179] This application also provides a TDOA positioning base station selection device, please refer to Figure 6 , the TDOA positioning base station selection device comprises:
[0180] An extraction module 10 is used to extract amplitude dispersion features and time dispersion features from the collected RRS data and TRS data of each candidate base station;
[0181] A prediction module 20 is used to input the amplitude dispersion feature and the time dispersion feature into a positioning base station prediction model to obtain a prediction result of each candidate base station as a positioning base station, wherein the positioning base station prediction model is obtained by inputting the time dispersion feature and the amplitude dispersion feature extracted from the sample data into an AdaBoost model for training;
[0182] An evaluation module 30, configured to perform confidence evaluation on the prediction result to obtain the confidence of the prediction result corresponding to each candidate base station;
[0183] The selection module 40 is configured to select a base station that meets a preset confidence threshold requirement from among the candidate base stations based on the confidence as a positioning base station.
[0184] Optionally, the extraction module 10 is also used to preprocess the collected RRS data and TRS data of each candidate base station; perform timestamp alignment on the preprocessed RRS data and TRS data; determine the cross-correlation value between the RRS data and the TRS data based on the collection time point of the RRS data and the TRS data after the timestamp alignment; and determine the amplitude dispersion feature and the time dispersion feature based on the cross-correlation value.
[0185] Optionally, the amplitude dispersion characteristics include amplitude dispersion energy, amplitude dispersion maximum value, amplitude dispersion mean value, amplitude dispersion variance, amplitude dispersion kurtosis and amplitude dispersion skewness;
[0186] The extraction module 10 is also used to determine the amplitude dispersion energy of the signal power distribution in the RRS data and the TRS data based on the square sum of the mutual correlation values at each of the acquisition time points; to select the maximum amplitude dispersion value at the target time point corresponding to the maximum mutual correlation value from the RRS data and the TRS data; to average the signal power based on the mutual correlation values at each of the acquisition time points to determine the amplitude dispersion mean of the signal power; to analyze the degree of fluctuation of the signal power based on the amplitude dispersion mean and the corresponding mutual correlation value to determine the amplitude dispersion variance of the signal power; to analyze the degree of sharpness of the signal power based on the amplitude dispersion mean, the corresponding amplitude dispersion variance and the mutual correlation value to determine the amplitude dispersion kurtosis of the signal power; to analyze the degree of asymmetry of the signal power based on the amplitude dispersion mean, the corresponding amplitude dispersion variance and the mutual correlation value to determine the amplitude dispersion kurtosis of the signal power.
[0187] Optionally, the time dispersion characteristics include time dispersion mean, time dispersion variance, time dispersion kurtosis and time dispersion skewness;
[0188] The extraction module 10 is also used to normalize the cross-correlation values at each of the acquisition time points to obtain the normalized cross-correlation at each of the acquisition time points; based on the normalized cross-correlation, average the signal arrival time differences in the RRS data and the TRS data to obtain the time dispersion mean of the arrival time difference; based on the time dispersion mean and the normalized cross-correlations, perform a fluctuation degree analysis on the arrival time difference to determine the time dispersion variance of the arrival time difference; based on the time dispersion mean, the time variance and the normalized cross-correlations, perform a sharpness degree analysis on the arrival time difference to determine the time dispersion kurtosis of the arrival time difference; based on the time dispersion mean, the time variance and the normalized cross-correlations, perform an asymmetry degree analysis on the arrival time difference to determine the time dispersion skewness of the arrival time difference.
[0189] Optionally, the prediction module 20 is also used to obtain sample data from the receiving and transmitting data of each sample base station; extract time dispersion features and amplitude dispersion features from the RRS data and TRS data of each candidate base station; iteratively train the pre-selected weak classifiers based on the time dispersion features and the amplitude dispersion features to obtain the weights of each weak classifier; based on the weights, combine each weak classifier into a strong classifier to obtain a positioning base station prediction model.
[0190] Optionally, the evaluation module 30 is also used to call the weights corresponding to each weak classifier in the positioning base station prediction model and the reference results of each weak classifier predicting each candidate base station; based on the reference results and the corresponding weights, the prediction results are evaluated for confidence to obtain the final decision value of the prediction results corresponding to each candidate base station; the final decision value is mapped and converted to obtain the confidence of the prediction results corresponding to each candidate base station.
[0191] The TDOA positioning base station selection device provided by the present application adopts the TDOA positioning base station selection method in the above embodiment, which can solve the technical problem that the signal propagation time delay increases and has uncertainty, resulting in a decrease in the reliability of the base station selected according to TDOA positioning, thereby reducing the positioning accuracy of the base station. Compared with the prior art, the beneficial effects of the TDOA positioning base station selection device provided by the present application are the same as the beneficial effects of the TDOA positioning base station selection method provided by the above embodiment, and the other technical features in the TDOA positioning base station selection device are the same as the features disclosed in the above embodiment method, which will not be repeated here.
[0192] The present application provides a TDOA positioning base station selection device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the TDOA positioning base station selection method in the above-mentioned embodiment one.
[0193] Reference below Figure 7 , which shows a schematic diagram of the structure of a TDOA positioning base station selection device suitable for implementing the embodiment of the present application. The TDOA positioning base station selection device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 7 The TDOA positioning base station selection device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.
[0194] like Figure 7As shown, the TDOA positioning base station selection device may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. Various programs and data required for the operation of the TDOA positioning base station selection device are also stored in the RAM 1004. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems may be connected to the I / O interface 1006: input devices 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 1003 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 1009. The communication device 1009 may allow the TDOA positioning base station selection device to communicate wirelessly or wired with other devices to exchange data. Although the TDOA positioning base station selection device with various systems is shown in the figure, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems may be implemented or have alternatively.
[0195] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.
[0196] The TDOA positioning base station selection device provided by the present application adopts the TDOA positioning base station selection method in the above embodiment, which can solve the technical problem that the signal propagation time delay increases and has uncertainty, resulting in a decrease in the reliability of the base station selected according to TDOA positioning, thereby reducing the positioning accuracy of the base station. Compared with the prior art, the beneficial effects of the TDOA positioning base station selection device provided by the present application are the same as the beneficial effects of the TDOA positioning base station selection method provided by the above embodiment, and the other technical features in the TDOA positioning base station selection device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.
[0197] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.
[0198] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
[0199] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the TDOA positioning base station selection method in the above-mentioned embodiment.
[0200] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.
[0201] The above-mentioned computer-readable storage medium may be included in the TDOA positioning base station selection device; or may exist independently without being assembled into the TDOA positioning base station selection device.
[0202] The above-mentioned computer-readable storage medium carries one or more programs. When the above-mentioned one or more programs are executed by the TDOA positioning base station selection device, the TDOA positioning base station selection device: extracts the amplitude dispersion characteristics reflecting the signal power and the time dispersion characteristics reflecting the arrival time difference of the signal from the collected RRS data and TRS data of each candidate base station; inputs the amplitude dispersion characteristics and the time dispersion characteristics into the positioning base station prediction model to obtain the prediction results of each candidate base station as a positioning base station, and the positioning base station prediction model is obtained by inputting the time dispersion characteristics and amplitude dispersion characteristics extracted from the sample data into the AdaBoost model for training; performs confidence evaluation on the prediction results to obtain the confidence of the prediction results corresponding to each candidate base station; and selects a base station that meets the preset confidence threshold requirement from each candidate base station as a positioning base station based on the confidence.
[0203] Computer program code for performing the operations of the present application may be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0204] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
[0205] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.
[0206] The readable storage medium provided by the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned TDOA positioning base station selection method, and can solve the technical problem that the signal propagation time delay increases and has uncertainty, resulting in a decrease in the reliability of the base station selected according to TDOA positioning, thereby reducing the positioning accuracy of the base station. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present application are the same as the beneficial effects of the TDOA positioning base station selection method provided by the above-mentioned embodiment, and will not be repeated here.
[0207] The present application also provides a computer program product, including a computer program, which implements the steps of the TDOA positioning base station selection method as described above when executed by a processor.
[0208] The computer program product provided by the present application can solve the technical problem that the signal propagation time delay increases and has uncertainty, resulting in a decrease in the reliability of the base station selected according to TDOA positioning, thereby reducing the positioning accuracy of the base station. Compared with the prior art, the beneficial effects of the computer program product provided by the present application are the same as the beneficial effects of the TDOA positioning base station selection method provided in the above embodiment, and will not be repeated here.
[0209] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.
Claims
1. A TDOA positioning base station selection method, characterized in that: The method includes: Extract amplitude dispersion features and time dispersion features from the collected RRS data and TRS data of each candidate base station; Inputting the amplitude dispersion feature and the time dispersion feature into a positioning base station prediction model to obtain a prediction result of each candidate base station as a positioning base station, wherein the positioning base station prediction model is obtained by inputting the time dispersion feature and the amplitude dispersion feature extracted from the sample data into an AdaBoost model for training; Performing confidence evaluation on the prediction result to obtain the confidence of the prediction result corresponding to each candidate base station; Based on the confidence, a base station that meets a preset confidence threshold requirement is selected from the candidate base stations as a positioning base station.
2. The method according to claim 1, characterized in that The step of extracting amplitude dispersion features and time dispersion features from the collected RRS data and TRS data of each candidate base station includes: Preprocessing the collected RRS data and TRS data of each candidate base station; Align the timestamps of the preprocessed RRS data with the TRS data; Determine a cross-correlation value between the RRS data and the TRS data based on the acquisition time points of the RRS data and the TRS data after the timestamps are aligned; Based on the cross-correlation value, the amplitude dispersion feature and the time dispersion feature are determined.
3. The method according to claim 2, characterized in that The amplitude dispersion characteristics include amplitude dispersion energy, amplitude dispersion maximum value, amplitude dispersion mean value, amplitude dispersion variance, amplitude dispersion kurtosis and amplitude dispersion skewness. The step of determining the amplitude dispersion characteristics based on the cross-correlation value includes: Determining the amplitude dispersion energy of the signal power distribution in the RRS data and the TRS data based on the sum of the squares of the cross-correlation values at each of the acquisition time points; Filtering the maximum amplitude dispersion value corresponding to the target time point with the maximum cross-correlation value from the RRS data and the TRS data; Based on the cross-correlation values at each of the acquisition time points, averaging the signal power to determine an amplitude dispersion mean of the signal power; Based on the amplitude dispersion mean and the corresponding cross-correlation value, performing a fluctuation degree analysis on the signal power to determine the amplitude dispersion variance of the signal power; Based on the amplitude dispersion mean, the corresponding amplitude dispersion variance and the cross-correlation value, performing a kurtosis analysis on the signal power to determine the amplitude dispersion kurtosis of the signal power; Based on the amplitude dispersion mean, the corresponding amplitude dispersion variance and the cross-correlation value, the signal power is analyzed for asymmetry to determine the amplitude dispersion deviation of the signal power.
4. The method according to claim 2, characterized in that The time dispersion features include a time dispersion mean, a time dispersion variance, a time dispersion kurtosis, and a time dispersion skewness. The step of determining the time dispersion features based on the cross-correlation value includes: Normalizing the cross-correlation values at each acquisition time point to obtain a normalized cross-correlation at each acquisition time point; Based on the normalized cross-correlation, averaging the signal arrival time differences in the RRS data and the TRS data to obtain a time dispersion mean of the arrival time differences; Based on the time dispersion mean and each of the normalized cross-correlations, performing a fluctuation degree analysis on the arrival time difference to determine the time dispersion variance of the arrival time difference; Based on the time dispersion mean, the time variance and each of the normalized cross-correlations, performing a kurtosis analysis on the arrival time difference to determine the time dispersion kurtosis of the arrival time difference; Based on the time dispersion mean, the time variance and each of the normalized cross-correlations, an asymmetry degree analysis is performed on the arrival time difference to determine the time dispersion skewness of the arrival time difference.
5. The method according to claim 1, characterized in that The positioning base station prediction model is obtained based on the following steps: Collect RRS data and TRS data of each sample base station; Extracting time dispersion features and amplitude dispersion features from the RRS data and TRS data of each sample base station; Iteratively training the preselected weak classifiers based on the time diffusion feature and the amplitude diffusion feature to obtain the weights of each of the weak classifiers; Based on the weights, the weak classifiers are combined into a strong classifier to obtain the positioning base station prediction model.
6. The method according to claim 1, characterized in that The step of performing confidence evaluation on the prediction result to obtain the confidence of the prediction result corresponding to each candidate base station comprises: Calling the weights corresponding to the weak classifiers in the positioning base station prediction model and the reference results of the weak classifiers predicting each candidate base station; Based on the reference results and the corresponding weights, the confidence level of the prediction results is evaluated to obtain a final decision value of the prediction results corresponding to each candidate base station; The final decision value is mapped and converted to obtain the confidence level of the prediction result corresponding to each candidate base station.
7. A TDOA positioning base station selection device, characterized in that: The device comprises: An extraction module is used to extract amplitude dispersion features and time dispersion features from the collected RRS data and TRS data of each candidate base station; A prediction module, used for inputting the amplitude dispersion feature and the time dispersion feature into a positioning base station prediction model to obtain a prediction result of each candidate base station as a positioning base station, wherein the positioning base station prediction model is obtained by inputting the time dispersion feature and the amplitude dispersion feature extracted from the sample data into an AdaBoost model for training; An evaluation module, used to perform confidence evaluation on the prediction result to obtain the confidence of the prediction result corresponding to each candidate base station; The selection module is used to select a base station that meets a preset confidence threshold requirement from each of the candidate base stations based on the confidence as a positioning base station.
8. A TDOA positioning base station selection device, characterized in that: The device comprises: a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the TDOA positioning base station selection method as described in any one of claims 1 to 6.
9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the TDOA positioning base station selection method according to any one of claims 1 to 6 are implemented.
10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the TDOA positioning base station selection method as described in any one of claims 1 to 6 are implemented.
