Positioning methods and related apparatuses

Abstract

The application provides a positioning method and related device, and relates to the technical field of mobile communication. The method comprises the following steps: obtaining an intermediate positioning result, wherein the intermediate positioning result comprises a first positioning result and a second positioning result, the first positioning result is obtained based on a first positioning mode, the second positioning result is obtained based on a second positioning mode, the first positioning mode is different from the second positioning mode, obtaining a positioning result of a terminal based on the intermediate positioning result and a dynamic environment parameter, the dynamic environment parameter is used for configuring the weight of the intermediate positioning result, and the dynamic environment parameter represents at least one of the number of base stations associated with the terminal, the number of satellites associated with the terminal and the state of a sensor associated with the terminal. The positioning result of the terminal not only combines the results of multiple positioning modes, but also considers the number of base stations, the number of satellites and the state of the sensor associated with the terminal, which is beneficial to further improving the accuracy of the positioning result of the terminal, and thus the positioning technology is improved.

Description

Technical Field

[0001] This application relates to the field of mobile communication technology, and in particular to a positioning method and related apparatus. Background Technology

[0002] With the continuous innovation and widespread adoption of communication technologies, positioning technologies based on mobile communication technologies have been further developed.

[0003] Currently, positioning technology based on mobile communication technology still has room for improvement, such as the need to improve positioning accuracy in certain scenarios. Summary of the Invention

[0004] This application provides a positioning method and related apparatus, aiming to improve existing positioning methods, such as increasing the accuracy of positioning results. The disclosed technical solution is as follows:

[0005] The first aspect of this application provides a positioning method. In some implementations, this method can be executed by a network device, or by a component (such as a circuit, chip, or chip system) configured in the network device, or by a logic module or software capable of implementing all or part of the functions of the network device. In other implementations, this method can be executed by a terminal, or by a component (such as a circuit, chip, or chip system) configured in the terminal, or by a logic module or software capable of implementing all or part of the terminal's functions. This application does not limit the scope of this method.

[0006] The method includes: obtaining intermediate positioning results, which include: a first positioning result and a second positioning result, wherein the first positioning result is obtained based on a first positioning method and the second positioning result is obtained based on a second positioning method, wherein the first positioning method and the second positioning method are different; obtaining the positioning result of the terminal based on the intermediate positioning results and dynamic environmental parameters, wherein the dynamic environmental parameters are used to configure the weight of the intermediate positioning results, and the dynamic environmental parameters represent at least one of the number of base stations associated with the terminal, the number of satellites associated with the terminal, and the state of the sensors associated with the terminal.

[0007] For example, the first positioning result is related to at least one dynamic environmental parameter, such as the parameter used to obtain the first positioning result, which is obtained using at least one of a base station, satellite, and sensor; the second positioning result is related to at least one dynamic environmental parameter, such as the parameter used to obtain the second positioning result, which is obtained using at least one of a base station, satellite, and sensor. Therefore, the dynamic environmental parameters can reflect the accuracy of the intermediate positioning results.

[0008] Because the first positioning method differs from the second positioning method, intermediate positioning results can be obtained from multiple dimensions. Since the terminal's positioning results are based on these intermediate results, the accuracy of the terminal's positioning results is improved. Furthermore, because dynamic environmental parameters are used to configure the weights of the intermediate positioning results, the terminal's positioning results also take into account the number of base stations, satellites, and sensor status associated with the terminal, which further improves the accuracy of the terminal's positioning results, thus improving the positioning technology.

[0009] In some implementations, the process of obtaining the second localization result based on the second localization method includes: calling a first deep learning model to obtain the location information output by the first deep learning model based on the parameter combination (also known as the comprehensive parameter combination), where the first deep learning model and the parameter combination have an adaptation relationship. Obtaining the second localization result based on the deep learning model is beneficial to improving the accuracy of the second localization result. Furthermore, because the first deep learning model and the parameter combination have an adaptation relationship, the utilization rate of the first deep learning model can be improved. For example, the computational power of the first deep learning model is well matched with the parameter combination, further improving the accuracy of the second localization result.

[0010] In some implementations, the adaptation relationship between the first deep learning model and the parameter combination includes: at least one of the features (such as data containing time series data or data with spatial relationships) and dimensions (such as the number of dimensions of the data) of the first deep learning model and the parameter combination having an adaptation relationship. The adaptation relationship is used to select the first deep learning model based on the parameter combination, which can both avoid wasting the resources of the first deep learning model and improve the accuracy of the output results of the first deep learning model.

[0011] In some implementations, the first deep learning model and the parameter combination have an adaptation relationship, including at least one of the following: a long short-term memory network-type deep learning model adapted to a parameter combination containing temporal data; a convolutional neural network-type deep learning model adapted to a parameter combination containing spatial relationships; in addition, the first target model in the first deep learning model is also adapted to the indoor or outdoor scene where the terminal is located, including: indoor, outdoor, from indoor to outdoor, or from outdoor to indoor. The different dimensions of adaptation between the first deep learning model and the parameter combination can lay the foundation for more flexible model selection and improve the universality of the localization method for different scenarios.

[0012] In some implementations, the parameter combination includes the following data: the number of transceiver antennas of the terminal, the frequency band used by the terminal, the number of historical data sets of Channel State Information (CSI) acquired by the terminal, and the number of historical data sets of Received Signal Strength Indication (RSSI). The method innovatively combines the above data for positioning, which is beneficial to improving the accuracy of the positioning results. Alternatively, the parameter combination may include at least one of the following data: satellite positioning data and sensor data. Satellite positioning data is more suitable for open outdoor areas and is easily affected by obstructions, while sensor data can compensate for the problems caused by obstructions, which is beneficial to improving the accuracy of the positioning results.

[0013] In some implementations, before obtaining intermediate localization results, the process includes: downloading a first deep learning model from a model management platform. This first deep learning model is selected from multiple models by the model management platform based on adaptation relationships. The model management platform facilitates the configuration and training of various models, thereby enabling efficient model processing.

[0014] In some implementations, obtaining the terminal's positioning result based on intermediate positioning results and dynamic environmental parameters includes: obtaining the positioning result output based on the intermediate positioning results and dynamic environmental parameters by calling a second deep learning model. This second deep learning model includes hidden layers used to assign weights to the intermediate positioning results. Using a deep learning model to implement weight assignment allows the model to fully learn the relationship between the intermediate positioning results and the weight configuration, which helps improve the accuracy of the terminal's positioning results.

[0015] In some implementations, the second deep learning model is adapted to the processing results of the intermediate localization results. The processing results include at least one of the following: obtaining information about the localization method of the intermediate localization results (such as the number and type), and information about the intermediate localization results obtained for each localization method (such as the number of groups and features), thereby improving the accuracy of the output results of the second deep learning model and making full use of the resources of the second deep learning model.

[0016] In some implementations, before obtaining the terminal's positioning result based on intermediate positioning results and dynamic environmental parameters, the process includes: downloading a second deep learning model from the model management platform, which selects the second deep learning model from multiple models based on the adaptation relationship between the second deep learning model and the processing results.

[0017] In some implementations, after obtaining the terminal's positioning result, the method further includes: comparing the terminal's positioning result with a first positioning result to obtain a first difference; and adjusting the configuration of the base station's transmitted signal based on the first difference and a first threshold, where the signal is the basis for the first positioning method to obtain the first positioning result. Adjusting the configuration of the base station's transmitted signal helps the terminal obtain a higher quality signal, thereby improving the accuracy of the first positioning result and further improving the accuracy of the terminal's positioning result.

[0018] In some implementations, after obtaining the positioning result of the terminal, the method further includes: comparing the positioning result of the terminal with a second positioning result to obtain a second difference; and based on the second difference and a second threshold, retraining at least one model of the first deep learning model for obtaining the second positioning result and the second deep learning model for obtaining the positioning result of the terminal, so as to improve the accuracy of the positioning result of the terminal when the first deep learning model and the second deep learning model are used for positioning again.

[0019] A second aspect of this application provides a positioning device, including modules for implementing the method provided in the first aspect of this application.

[0020] A third aspect of this application provides a positioning device, comprising: one or more processors and a memory; the memory is used to store program code; the processor is used to run the program code, causing the positioning device to implement the method provided in the first aspect of this application.

[0021] A fourth aspect of this application provides a computer-readable storage medium having instructions stored thereon that, when executed on an electronic device, cause the electronic device to perform the method provided in the first aspect of this application.

[0022] The fifth aspect of this application provides a computer program product having instructions stored thereon, which, when run on an electronic device, cause the electronic device to implement the method provided in the first aspect of this application.

[0023] A sixth aspect of this application provides a chip system comprising: at least one processor and an interface for receiving code instructions and transmitting them to the at least one processor; the at least one processor executes the code instructions to implement the method provided in the first aspect of this application. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is an example diagram of a mobile communication system;

[0026] Figure 2 This is an example diagram of a communication system that implements a positioning method according to an embodiment of this application;

[0027] Figure 3 This is an example diagram illustrating the acquisition of positioning results based on various models in a model management platform, as provided in the embodiments of this application.

[0028] Figure 4 This is an example diagram of a meta-learner provided in an embodiment of this application;

[0029] Figure 5 This is a flowchart of a positioning method provided in an embodiment of this application;

[0030] Figure 6 This is a flowchart of yet another positioning method provided in the embodiments of this application;

[0031] Figure 7 This is an example diagram illustrating the composition of a positioning device provided in an embodiment of this application;

[0032] Figure 8 This is an example diagram illustrating the composition of another positioning device provided in an embodiment of this application;

[0033] Figure 9 This is an example diagram illustrating the composition of another positioning device provided in the embodiments of this application. Detailed Implementation

[0034] The terms "first," "second," and "third," etc., used in this application specification, claims, and drawings are used to distinguish different objects, not to limit a specific order.

[0035] In the embodiments of this application, the words "in some implementations" or "for example" are used to indicate examples, illustrations or descriptions, and should not be construed as being more preferred or more advantageous than other embodiments or designs.

[0036] Figure 1 This is an example of a mobile communication system, which includes terminals, base stations, and core network equipment.

[0037] Mobile communication systems can be second-generation (2G) communication systems, third-generation (3G) communication systems, long-term evolution (LTE) systems, fifth-generation (5G) communication systems, LTE and 5G hybrid architectures, 5G new radio (5G NR) systems, and other new communication systems that will emerge in the future development of communication.

[0038] Terminal devices can take various forms, such as mobile phones, tablets, computers with wireless transceiver capabilities, virtual reality (VR) terminal devices, augmented reality (AR) terminal devices, wireless terminal devices in industrial control, vehicle-mounted terminal devices, wireless terminal devices in self-driving technology, wireless terminal devices in remote medical care, wireless terminal devices in smart grids, wireless terminal devices in transportation safety, wireless terminal devices in smart cities, wireless terminal devices in smart homes, wearable terminal devices, and so on. Terminal devices are sometimes also referred to as user equipment (UE), access terminal devices, vehicle-mounted terminal devices, industrial control terminal devices, UE units, UE stations, mobile stations, mobile terminals, remote stations, remote terminal devices, mobile devices, UE terminal devices, wireless communication devices, UE agents, or UE devices. Terminal devices can also be fixed terminal devices or mobile terminal devices.

[0039] Access network equipment can be terrestrial base stations or non-terrestrial network (NTN) equipment. NTN equipment can also be called base stations and / or satellite access nodes (SAN).

[0040] A base station is any device located on the network side with wireless transceiver capabilities, including but not limited to: evolved Node Bs (NodeBs, eNBs, or e-NodeBs) in LTE, base stations (gNodeBs or gNBs) or transmission receiving points / transmission reception points (TRPs) in new radio (NR), base stations evolved later in 3GPP, access nodes, wireless relay nodes, and wireless backhaul nodes in Wi-Fi systems. Base stations can be macro base stations, micro base stations, pico base stations, small cells, relay stations, or balloon stations, etc. A base station can contain one or more co-located or non-co-located TRPs. A base station can also be a radio controller, centralized unit (CU), and / or distributed unit (DU) in a cloud radio access network (CRAN) scenario. Base stations can communicate with terminal devices, or they can communicate with terminal devices through relay stations.

[0041] Core network equipment and base stations can be independent and different physical devices, or the functions of core network equipment and the logical functions of base stations can be integrated into the same physical device, or a physical device can integrate some of the functions of core network equipment and some of the functions of base stations. Figure 1 In this example, the core network equipment includes the location management function (LMF).

[0042] based on Figure 1 The example flow of the system implementing the positioning function shown includes: the base station sends the positioning parameters obtained from the terminal device to the LMF, and the LMF obtains the positioning result based on the received data.

[0043] Based on the different types of parameters used for positioning, positioning methods are classified into several types, as described below:

[0044] 1. Traditional cellular positioning: uses cellular network signals such as Positioning Reference Signal (PRS) and positioning algorithms to obtain positioning results.

[0045] Traditional cellular positioning is constrained by base station layout and signal propagation characteristics. In remote and complex urban environments, the positioning accuracy is as low as tens or even hundreds of meters. It has poor scalability and is prone to communication bandwidth congestion when the positioning is poor.

[0046] 2. Location based on Channel State Information (CSI) and / or Received Signal Strength Indicator (RSSI).

[0047] CSI-based positioning has stringent hardware requirements and poor environmental adaptability, while RSSI-based positioning is affected by multipath effects, hotspot distribution, and signal stability, making it difficult to meet the accuracy requirements of high-precision scenarios.

[0048] 3. Satellite-based positioning: Location information is obtained based on satellite positioning data acquired by the terminal, such as Global Positioning System (GPS) data.

[0049] Satellite-based positioning suffers from insufficient scalability and low accuracy in complex environments such as urban high-rise areas, mountainous areas, and inclement weather. This is because satellite signals are interfered with by factors such as buildings, terrain, or weather.

[0050] 4. Sensor data-based positioning: The terminal acquires sensor data collected by the sensors mounted on the terminal and sends the sensor data to the LMF through the base station. The LMF obtains the terminal's location information based on the sensor data.

[0051] Location based on sensor data has the following problems:

[0052] The accuracy of the sensors on the terminal is limited. For example, the error of inertial sensors is easy to accumulate. Furthermore, because the fusion algorithms of various sensor data are complex and the sensors are greatly affected by the environment, the accuracy of sensor data is low, which reduces the accuracy of the positioning results. In addition, there are problems such as high resource consumption of sensor equipment and difficulty in balancing performance and resources.

[0053] In summary, current positioning technologies based on wireless communication systems are unable to meet the high-precision positioning requirements in complex environments, and it is also difficult to achieve a balance between performance improvement and equipment resource consumption. This poses a significant challenge to improving the accuracy, stability, and universality of positioning.

[0054] To address the aforementioned problems, embodiments of this application provide a positioning method. The positioning method provided by the embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0055] by Figure 2 For example, the positioning method provided in the embodiments of this application is by Figure 2 The system implementation shown includes at least one base station ( Figure 2(Taking multiple base stations as an example), at least one terminal and core network equipment. Figure 2 and Figure 1 The difference is at least in that, Figure 2 In addition to LMF, the core network equipment also includes a model management platform. Figure 2 Taking the LMF and model management platform as independent devices as an example, the model management platform can also be integrated into the LMF.

[0056] The model management platform is used to store and manage various types of artificial intelligence (AI) models, and can provide downloads of AI models to terminals or network devices. Besides being set up in the core network as a server, the model management platform can also be a cloud server, a personal computer (PC), or a laptop.

[0057] Figure 2 The document also shows examples of components for acquiring parameters used for positioning, such as GPS, gyroscopes, accelerometers, and barometric pressure sensors. It is understood that these components can be independent of the terminal or can be integrated into the terminal. When the components are independent of the terminal, the components can send the acquired parameters to the terminal.

[0058] In positioning scenarios, LMF is used to efficiently store and manage various positioning-related data, such as base station locations and signal propagation model parameters. Simultaneously, it accurately manages and maintains the terminal's context information, which includes positioning technology support and mobility status, and can update its status changes in real time, providing crucial support for the core network's positioning services and other key functions.

[0059] The following section will first explain the various models in the model management platform and their training strategies.

[0060] Figure 3 This illustrates the technical logic for obtaining location results based on various models within the model management platform:

[0061] The model management platform is configured with base learners and meta-learners.

[0062] A base learner can be understood as a model that obtains localization results based on at least one type of localization data. For example, a base learner includes:

[0063] Cellular localization base learner: also known as the first-class base learner, it is used to implement traditional cellular localization functions. For example, the cellular localization base learner obtains the localization result based on the cellular localization method without the participation of a model.

[0064] Deep learning models based on CSI and RSSI: referred to as the second type of base learner, used to implement localization functions based on CSI data and / or RSSI data.

[0065] GPS and sensor-based deep learning models: also known as third-class base learners, are used to achieve positioning functions based on GPS data and / or sensor data.

[0066] It is understandable that deep learning models based on CSI and RSSI, and deep learning models based on GPS and sensors, are the base learners that need to be called to obtain positioning results.

[0067] A meta-learner can be understood as a model that outputs localization results based on the output of the base learner.

[0068] return Figure 3 For example, in the prediction stage, the positioning results (X1, Y1) are output by the cellular positioning base learner, the positioning results (X2, Y2) are output by the deep learning model based on CSI and RSSI, and the positioning results (X3, Y3) are output by the deep learning model based on GPS and sensors.

[0069] The meta-learner is based on (X1, Y1), (X2, Y2), (X3, Y3) and dynamic environmental parameters (number of base stations, number of satellites, sensor status) and outputs the final positioning result (X, Y).

[0070] The localization result output by the meta-learner can be denoted as: (X, Y) = MLP([X1, Y1, X2, Y2, X3, Y3, number of base stations, number of satellites, sensor status]).

[0071] Among them, the number of base stations refers to the number of base stations participating in positioning, the number of satellites refers to the number of satellites participating in positioning, and the sensor status indicates whether the sensor is normal, such as normal or faulty. The sensor status can also be the sensor's accuracy, etc.

[0072] For example, the meta-learner uses a multilayer perceptron (MLP), with the specific structure as follows: Figure 4 As shown, it includes an input layer, a hidden layer, and an output layer.

[0073] For example, the input layer contains 9 neurons, corresponding to 9-dimensional feature input. In the hidden layer, the first hidden layer has 128 neurons and uses a Rectified Linear Unit (ReLU) as the activation function. It learns the correlation between dynamic environment parameters and the output of the base learners. Through feature crossing, it realizes the dynamic weight allocation of the localization results output by each base learner based on the dynamic environment parameters. That is, the first hidden layer can dynamically allocate weights to the localization results output by each base learner based on the dynamic environment parameters.

[0074] The second hidden layer has 64 neurons and also uses the ReLU activation function to further extract higher-order features. After the two hidden layers, a dropout layer is set. Figure 4 (Not shown in the image) The Dropout layer randomly discards 20% of the neurons to prevent overfitting and enhance the model's generalization ability. The output layer consists of two neurons, corresponding to the final output latitude and longitude (X, Y).

[0075] The training process for the base learner includes:

[0076] For a deep learning model based on CSI and RSSI (i.e., the second type of base learner), input sample data containing CSI and / or RSSI related parameters (referred to as CSI / RSSI sample data) is used to obtain the localization result output by the second type of base learner, denoted as (X2, Y2). The model parameters of the second type of base learner are adjusted using (X2, Y2) and the accurate location labels corresponding to the CSI / RSSI sample data to obtain the trained second type of base learner.

[0077] For a GPS and sensor-based deep learning model (i.e., a third-type base learner), input sample data, including GPS data and / or sensor data (referred to as GPS / sensor sample data), yields the localization result output by the third-type base learner, denoted as (X3, Y3). The model parameters of the third-type base learner are adjusted using (X3, Y3) and the accurate location labels corresponding to the GPS / sensor sample data.

[0078] For example, during the training of the base learner, LMF may also select an appropriate base learner to be trained based on the features of the sample data, which will be described in conjunction with subsequent embodiments. The specific content of the sample data used for training will also be described in subsequent embodiments.

[0079] After training the second and third base learners, a meta-learner is trained based on these learned base learners, as follows:

[0080] Input CSI / RSSI sample data (different from the sample data used to train the second type of base learner) into the second type of base learner to obtain the localization result output by the second type of base learner, denoted as (X2, Y2). Input GPS / sensor sample data (different from the sample data used to train the third type of base learner) into the second type of base learner to obtain the localization result output by the third type of base learner, denoted as (X3, Y3).

[0081] LMF calls the first type of base learner and uses cellular positioning to obtain the positioning result, denoted as (X1, Y1).

[0082] Input (X1, Y1), (X2, Y2), (X3, Y3) along with base station number samples, satellite number samples, and sensor state samples into the meta-learner to obtain the predicted coordinates (X, Y) output by the meta-learner. Adjust the model parameters of the meta-learner based on the sample labels (i.e., the real location labels) and the predicted coordinates (X, Y).

[0083] For example, in adjusting the model parameters of the meta-learner based on sample labels and predicted coordinates (X, Y), the mean squared error (MSE) is chosen as the loss function. The difference between the model's prediction and the true location label is measured by calculating the squared Euclidean distance between the predicted coordinates and the true location label. The optimizer chosen is the Adam optimizer, which features an adaptive learning rate. The initial learning rate is set to 0.001, and this adaptive learning rate adjustment helps the model converge to the optimal solution more quickly. During training, the gradient is calculated using the backpropagation algorithm and the chain rule, thereby updating the weights and biases of each neuron in the MLP, allowing the model to continuously optimize.

[0084] The trained base learners and meta-learners are stored on the model management platform for easy access for subsequent localization.

[0085] The following will combine Figure 2 The various devices shown will be used to provide a more detailed description of the positioning method provided in the embodiments of this application.

[0086] Figure 5 This application discloses a positioning method flowchart, which includes the following steps:

[0087] S101. The base station sends a Positioning Reference Signal (PRS) to the terminal, and the terminal receives the PRS accordingly.

[0088] For example, a base station sends a PRS to terminals within a target area, and the target area can be determined based on relevant protocols.

[0089] PRS includes base station identifier, timestamp, and signal configuration information.

[0090] S102. The terminal performs positioning measurements based on PRS and obtains measurement data.

[0091] For example, the measurement data includes at least one of signal strength, time of arrival, and angle of arrival.

[0092] S103. The terminal sends measurement data to the base station, and the base station receives the measurement data accordingly.

[0093] S104. The base station processes the measurement data to obtain the processing results.

[0094] For example, the processing methods include, but are not limited to, time calibration, angle calibration, filtering, and removal of outliers.

[0095] S105. The base station sends the processing result to the LMF, and the LMF receives the processing result accordingly.

[0096] S106 and LMF obtain the positioning results based on the processing results.

[0097] To distinguish it from subsequent location results, this is referred to as the first type of location result.

[0098] For example, in this step, LMF achieves the purpose of using traditional cellular positioning by calling the first type of base learner.

[0099] S107, LMF sends the first type of positioning result to the terminal, and the terminal receives the first type of positioning result accordingly.

[0100] It is understandable that S101-S107 is a process for positioning using traditional cellular positioning methods. For more specific implementation details of each step, please refer to the relevant protocols, which will not be elaborated here.

[0101] S108. The terminal acquires and preprocesses CSI data and RSSI data.

[0102] For example, CSI data includes channel gain data for CSI, which may be a channel gain factor, a channel gain vector, or a channel gain matrix. CSI data may be currently acquired data or historical CSI data, i.e., previously acquired CSI data that has been stored.

[0103] For example, RSSI data includes RSSI intensity and RSSI variation sequences, and the RSSI data may be currently acquired RSSI data or historical RSSI data.

[0104] For example, preprocessing methods for CSI data include, but are not limited to: using a min-max normalization method to map the channel gain data to [0, 1] to eliminate significant differences in the channel gain range under different scenarios; and using mean filtering for noise reduction to effectively reduce the adverse effects of noise interference in the wireless channel on the accuracy of CSI.

[0105] Preprocessing methods for historical CSI data include, but are not limited to: accurately sorting historical CSI data from different times and measurement periods according to time sequence using timestamps to achieve data alignment; if there are inconsistencies in time precision, using time interpolation to interpolate historical CSI data to ensure that the data is evenly and continuously distributed on the time axis.

[0106] For example, preprocessing methods for RSSI data include, but are not limited to, using Kalman filtering to smooth signal fluctuations and remove anomalies to optimize RSSI data quality. Historical RSSI data can be processed using similar preprocessing methods to historical CSI data, which will not be elaborated upon here.

[0107] S109. The terminal extracts the first comprehensive parameter combination based on CSI data and RSSI data.

[0108] To distinguish it from the following content, the combined parameter combination in this step will be referred to as the first combined parameter combination.

[0109] For example, the first combination of integrated parameters includes the following parameters: number of transceiver antennas, frequency band used, number of CSI historical data sets, and number of RSSI historical data sets.

[0110] An example of how to extract each parameter from the first comprehensive parameter combination is as follows:

[0111] By analyzing the channel gain data of CSI, such as the dimensions of the channel gain matrix, the number of transceiver antennas can be determined by combining the relationship between multiple matrix dimensions and communication protocols in a complex MIMO system.

[0112] In current communication systems such as NR systems, frequency band identifiers are extracted by parsing signaling such as the Master Indication Block (MIB) and System Information Block 1 (SIB1) to determine the frequency band being used.

[0113] The preprocessed CSI historical data is grouped and statistically analyzed according to a set time interval. When data loss or anomalies occur, the time window is corrected or redefined to obtain the number of CSI historical data groups, i.e., how many groups of CSI historical data there are.

[0114] First, synchronize the RSSI data with the CSI data. Then, using a statistical method similar to CSI, count the number of RSSI historical information groups according to the set time interval to obtain the number of RSSI historical data groups.

[0115] S110, The terminal sends the first integrated parameter combination to the model management platform, and the model management platform receives the first integrated parameter combination accordingly.

[0116] S111. The model management platform selects a suitable model based on the first comprehensive parameter combination. In this step, it is assumed that the model selected by the model management platform to be suitable for the first comprehensive parameter combination is a second type of base learner.

[0117] For example, the model management platform selects a suitable AI model from a predefined model library based on at least one of the features and dimensions of the first comprehensive parameter combination. For instance, if the first comprehensive parameter combination contains time-series data (such as the number of historical CSI and / or RSSI data sets), a base learner of the Long Short-Term Memory (LSTM) or Transformer type is selected. If spatial relationships need to be processed (i.e., the data has spatial relationships) (such as the first comprehensive parameter combination containing Multiple-Input Multiple-Output (MIMO) data), a base learner of the Convolutional Neural Network (CNN) type is selected. In this step, because the first comprehensive parameter combination contains the number of historical CSI and RSSI data sets, a base learner based on LSTM or Transformer is selected, referred to as the second type of base learner.

[0118] For example, the above selection rules can be pre-configured on the model management platform, and the adaptation relationship (i.e. correspondence) of each base learner and parameter combination features and / or dimensions can be pre-configured on the model management platform to lay the foundation for selecting base learners based on comprehensive parameter combinations.

[0119] Choosing a base learner that is compatible with the first combination of integrated parameters is beneficial to improving the accuracy of the localization results output by the base learner based on the first combination of integrated parameters.

[0120] S112. The terminal downloads the second type of base learner from the model management platform.

[0121] For example, the model management platform sends information such as the identifier of the second type of base learner to the terminal, and the terminal downloads the second type of base learner from the model management platform based on this information.

[0122] S113. The terminal uses the first comprehensive parameter combination as the input of the second type of base learner, calls the second type of base learner, and obtains the second type of localization result output by the second type of base learner.

[0123] It is understandable that the second type of positioning result is the positioning result obtained based on the first comprehensive parameter combination, namely CSI data and RSSI data.

[0124] S114. The terminal acquires and preprocesses GPS data and sensor data.

[0125] For example, sensor data includes at least one of the following: gyroscope data, accelerometer data, and barometric pressure sensor data.

[0126] For example, preprocessing methods include, but are not limited to: cleaning the data, removing obviously erroneous or abnormal data points by setting reasonable data range thresholds, normalizing the data from different sensors, and aligning the data using timestamps to ensure the temporal synchronization of data from different sensors.

[0127] For example, preprocessing can also be performed on certain types of sensor data. For gyroscope data, a low-pass filter can be used to filter out abnormal high-frequency fluctuations caused by measurement noise or high-frequency interference. For accelerometer data, high-frequency noise caused by vibration or other reasons can be removed to make the data smoother and better reflect the true motion state of the object.

[0128] S115. The terminal extracts a second comprehensive parameter combination based on GPS data and sensor data.

[0129] For example, the second combination of integrated parameters includes at least one of the following: the average value of rotational angular velocity, the variance of rotational angular velocity, the cumulative change of rotational angle, the peak value of acceleration, the trough value of acceleration, the frequency of acceleration change, the rate of change of air pressure, the position accuracy factor, longitude, latitude, and altitude.

[0130] An example of how to extract each parameter from the above second comprehensive parameter combination is as follows:

[0131] The gyroscope data features are extracted as the average and variance of the rotational angular velocity over a period of time, as well as the cumulative change in rotational angle, to reflect the rotational state and stability of the device. Accelerometer features are extracted as the peak and trough values ​​of acceleration, and the frequency of acceleration changes. Barometric pressure sensor features are extracted as the rate of change of barometric pressure. The location information (longitude, latitude, and altitude) contained in the GPS data is fused with data from other sensors to obtain a second comprehensive parameter combination.

[0132] S116. The terminal sends the second integrated parameter combination to the model management platform, and the model management platform receives the second integrated parameter combination accordingly.

[0133] S117. The model management platform selects a third type of base learner that is suitable for the second comprehensive parameter set based on the second comprehensive parameter set.

[0134] For example, in this step, a third type of base learner is selected based on the dimensions of the data contained in the second combination of comprehensive parameters.

[0135] For example, a third type of base learner can also be selected based on the type of data contained in the second combination of synthesis parameters:

[0136] Based on the type of sensor data and the corresponding accuracy data of the sensor data contained in the second comprehensive parameter combination, the environment in which the terminal is located is analyzed. For example, the current environment type (such as urban canyon, open area, indoor, etc.) can be determined based on altitude.

[0137] Based on the types of sensor data included in the second comprehensive parameter combination, the motion state of the terminal is identified, and the user's current motion state (walking, driving, stationary, etc.) is identified through rotation and acceleration features;

[0138] Assess the current reliability of each sensor based on the variance and rate of change of the sensor data.

[0139] Based on the environment, motion state, and current reliability of each sensor, a third type of base learner is selected. For example, when the position accuracy factor (PDOP) is high, the acceleration changes frequently but with moderate amplitude, the air pressure changes slowly, and the GPS signal is intermittent, a third type of base learner adapted to urban driving scenarios with dense buildings is selected.

[0140] Correspondingly, data from various scenarios can be collected as sample data, and the sample data can be used to train a third type of base learner that matches each scenario. For example, a large amount of GPS and sensor data can be collected in different environmental scenarios (city, suburbs, indoors, etc.), the data in each scenario can be labeled, the real location coordinates can be recorded, and sensor patterns in different motion states (stationary, walking, driving, etc.) can be collected.

[0141] During the model training phase: specialized localization models are trained for different environment types and motion states. Deep learning architectures (such as CNN, LSTM, etc.) are used to process spatiotemporal sequence data, and transfer learning techniques are employed to enable the model to adapt to new environments.

[0142] Model classification and storage: The trained models are classified and stored according to the applicable environment and motion features. A model index database is established to record the best applicable conditions for each model (i.e., the correspondence between the third type of base learning and the adapted scenario).

[0143] S118. The terminal downloads the third type of base learner from the model management platform.

[0144] S119. The terminal uses the second comprehensive parameter combination as the input of the third type base learner, calls the third type base learner, and obtains the third type localization result output by the third type base learner.

[0145] It is understandable that the third type of positioning result is a positioning result obtained based on the second comprehensive parameter combination, namely GPS data and sensor data.

[0146] S120: The terminal processes the first type of positioning result, the second type of positioning result, and the third type of positioning result to obtain the third comprehensive parameter combination.

[0147] The third comprehensive parameter combination includes the number of positioning modes and the number of sets of positioning results for each positioning mode.

[0148] The positioning mode refers to the positioning method. In this embodiment, we take the cellular positioning mode implemented by the first type of base learner, the positioning mode based on CSI and RSSI data implemented by the second type of base learner, and the positioning mode based on GPS and sensor data implemented by the third type of base learner as examples. In this case, based on the aforementioned steps, there are three positioning modes in this embodiment. One positioning mode can obtain multiple sets of positioning results. For example, if the terminal obtains GPS and sensor data at 10 times, then the data at each of these 10 times can obtain a set of latitude and longitude based on the third type of base learner. Therefore, a total of 10 sets of second-type positioning results can be obtained at these 10 times. The array of positioning results for each positioning mode refers to the number of positioning results obtained by each positioning mode.

[0149] S121. The terminal sends the third integrated parameter combination to the model management platform, and the model management platform receives the third integrated parameter combination accordingly.

[0150] S122. The model management platform selects a meta-learner that is suitable for the third comprehensive parameter group.

[0151] For example, an appropriate meta-learner is selected based on at least one of the information of the positioning mode (also known as the positioning method) included in the third comprehensive parameter combination and / or the information of the positioning results of each positioning mode (also known as the intermediate positioning results obtained by each positioning method).

[0152] The third comprehensive parameter combination includes, but is not limited to, information about the positioning mode, including but not limited to the type and number of positioning modes. Examples of types include, but are not limited to, positioning mode types such as cellular positioning, CSI and RSSI-based positioning, sensor-based positioning, and GPS-based positioning. The number indicates how many positioning modes are available.

[0153] The information for the positioning results of each positioning mode includes: the number of sets of positioning results for each positioning mode and the characteristics of multiple sets of positioning results for the same positioning mode.

[0154] Information on the positioning mode and / or at least one of the number of sets of positioning results for each positioning mode can reveal the richness of the positioning data (how many sets of valid data each technology provides) and assess environmental complexity.

[0155] Assume there are three positioning modes: cellular positioning, CSI- and RSSI-based positioning, and sensor-based positioning. The number of positioning result sets for each mode and the characteristics of multiple sets of positioning results for each mode are as follows:

[0156] • Cellular positioning: 3 groups (with relatively large dispersion);

[0157] • Location based on CSI and RSSI: 2 groups (high consistency);

[0158] • Sensor-based positioning: 5 groups (with a few outliers).

[0159] As can be seen from the above positioning mode information, GPS-based positioning is not available, and based on the characteristics of multiple positioning results for each positioning mode, a meta-learner optimized for "indoor-outdoor transition areas" can be matched.

[0160] It is understandable that various meta-learners with suitable third comprehensive parameter sets can be pre-trained. These meta-learners are suitable for various scenarios, such as urban canyons (densely populated areas with high-rise buildings), open suburbs, indoor environments, and transportation hubs.

[0161] S123. The terminal downloads the meta-learner from the model management platform.

[0162] For example, the model management platform sends information such as an identifier of the meta-learner to the terminal, and the terminal downloads the meta-learner from the model management platform based on this information.

[0163] S124. The terminal inputs the first type of positioning result, the second type of positioning result, the third type of positioning result, and the dynamic environment parameters into the meta-learner, and calls the meta-learner to output the positioning result.

[0164] For example, the number of base stations in the dynamic environmental parameters can be obtained based on terminal measurements, the number of satellites in the dynamic environmental parameters can also be obtained by terminal measurements or by the terminal from satellites communicating with the terminal, and the sensor status in the dynamic environmental parameters can be obtained by the terminal from each sensor, such as the sensor status carried in the sensor data.

[0165] As mentioned earlier, because the meta-learner learns the correlation between dynamic environment parameters and base learner outputs during training, it can dynamically assign weights to the localization results output by each base learner based on the dynamic environment parameters, thus further improving the accuracy of the meta-learner output results.

[0166] Figure 5 The process shown has the following beneficial effects:

[0167] 1. Multiple base learners obtain positioning results from different dimensions. The positioning results from different dimensions are fused to obtain the final positioning result. Compared with a single positioning method, the final positioning result has higher accuracy. Moreover, when the accuracy of some positioning methods in different dimensions is limited, such as when the accuracy of cellular positioning and satellite positioning is reduced in a basement, the fusion of positioning results from different dimensions can still ensure the accuracy of the final positioning result. Therefore, it has strong tolerance and universality to the environment.

[0168] 2. For both base learners and meta-learners in deep learning models, selection is based on information from the output data (i.e., the combination of comprehensive parameters), such as at least one of type, feature, and dimension. This ensures a better fit between the model and the input data, further improving the accuracy of the localization results. Furthermore, selecting the model based on parameters (such as combinations of comprehensive parameters) also helps improve resource utilization. In other words, it avoids wasting the model's computational resources while fully utilizing them to obtain localization results with the required accuracy.

[0169] 3. The meta-learner can output the final localization result based on dynamic environmental parameters, which can further improve the accuracy of the localization result.

[0170] Figure 6 This is yet another positioning method provided in the embodiments of this application, and... Figure 5 The difference between the methods shown is that after obtaining the first type of positioning result, the LMF does not send the first type of positioning result to the terminal. After obtaining the second type of positioning result, the terminal sends the second type of positioning result to the LMF. After obtaining the third type of positioning result, the terminal sends the third type of positioning result to the LMF.

[0171] Unlike S123, as in S224, the LMF downloads the meta-learner from the model management platform. Also unlike S124, as in S225, the LMF inputs the first-class localization result, the second-class localization result, the third-class localization result, and dynamic environment parameters into the meta-learner to obtain the localization result output by the meta-learner. After obtaining the localization result, it sends the result to the terminal, and the terminal receives the result. Aside from the differences mentioned above, Figure 6 For other steps, please refer to Figure 5 This will not be elaborated upon here.

[0172] The positioning method provided in the above embodiments will be illustrated below with examples of two specific scenarios.

[0173] High-precision positioning in dynamic urban canyon environments:

[0174] In this example, the devices involved in localization include: multiple densely distributed base stations in the city, terminals equipped with various sensors, and an LMF (Model Management Module), which contains a model management platform. The first type of base learner, the second type of base learner, the third type of base learner, and the MLP meta-learner mentioned above have all been trained and stored on the model management platform.

[0175] To improve the response speed of localization and reduce latency, commonly used and lightweight base learners are downloaded and cached by the terminal from the model management platform.

[0176] The terminal, base station, and LMF work together to execute the positioning process provided in the above embodiments to obtain positioning results.

[0177] In this example, after obtaining the location results, the algorithms and / or models used in the location process can also be optimized and adjusted based on the location results:

[0178] One example of optimization adjustment is to compare the final positioning result with the positioning result output by the first type of base learner. If the difference is greater than a first threshold, the corresponding network device is triggered to reconfigure the base station, such as adjusting the transmission power and time-frequency resource allocation of the base station's PRS signal, or optimizing the signal coverage in indoor and outdoor edge areas, in order to increase the strength of the PRS signal.

[0179] Another example of optimization adjustment is to monitor the difference between the first type of localization result, the second type of localization result, the third type of localization result and the final localization result in real time. If the difference is greater than a second threshold (such as 5 meters), the base learner and / or meta learner are retrained to ensure the long-term stability of the model in complex scenarios.

[0180] Understandably, the terminal or LMF can upload the final positioning result and the data used to obtain that result (such as PRS, CSI, RSSI, GPS, and sensor data) to the model management platform. This allows the platform to store the data as new training samples for subsequent model retraining. Training samples obtained in the dynamic environment of urban canyons can enhance the model's ability to resist multipath interference from urban canyons.

[0181] Seamless positioning enhancement in indoor / outdoor switching scenarios:

[0182] In this example, the devices involved in positioning include: base stations in indoor shopping malls and outdoor plazas, terminals equipped with multi-band antennas and various high-precision sensors, and the aforementioned LMF and model management platform.

[0183] The difference from the above embodiments is that, in addition to adapting to the features and / or dimensions of various types of data, the models in the model management platform are also adapted for indoor and outdoor scenarios. For example, indoor and outdoor scenarios are divided into being indoors, outdoors, from indoors to outdoors, and from outdoors to indoors. Each scenario is adapted to at least one model. For example, the main model adapted for indoor scenarios is the aforementioned second type of base learner, and the auxiliary model is a model based on inertial sensor data for positioning. As another example, the model adapted for outdoor scenarios is the aforementioned first type of base learner and a model based on GPS positioning.

[0184] Similarly, meta-learners can also be adapted to indoor and outdoor scenes. For example, a first-order meta-learner trained on outdoor data can be adapted to outdoor scenes.

[0185] The model management platform also deploys a lightweight scene classification model, which is used to analyze and identify the current scene of the terminal, i.e., whether the terminal is currently indoors, outdoors, moving from indoors to outdoors, or moving from outdoors to indoors.

[0186] The terminal can download lightweight scene classification models and some base learners from the model management platform.

[0187] In this example, the difference from the aforementioned embodiments lies in the deployment of a backup positioning mechanism in the LMF. This backup positioning mechanism can be one of the aforementioned positioning methods, such as positioning based on GPS data and sensor data, or it can be the positioning method provided in the embodiments of this application. The purpose of deploying backup is to obtain a positioning result using another positioning process when one positioning process cannot provide a result. For example, the LMF deploys a first positioning network element and a second positioning network element. The first positioning network element uses the positioning method provided in this embodiment for positioning, while the second positioning network element performs positioning based on GPS data and sensor data. By default, the first positioning network element is used for positioning. If the first positioning network element cannot perform positioning (e.g., due to equipment failure), the second positioning network element is activated to ensure the stability of the positioning service.

[0188] The process of the positioning method based on the above configuration includes:

[0189] The terminal invokes a lightweight scene classification model, combining it with acquired data such as current CSI data, to identify the current scene. If the current location is indoors, it invokes a second-type base learner and a model based on inertial sensor data for localization to obtain the localization result. For example, the localization results output by the second-type base learner and the model based on inertial sensor data are input into a meta-learner adapted to the indoor scene to obtain the final localization result output by the meta-learner. Inertial sensor data can compensate for errors caused by CSI signal occlusion, thus improving the accuracy of the final localization result.

[0190] In indoor scenarios, if the inertial sensor is disturbed by vibration, causing abnormal inertial sensor data, the meta-learner can reduce the weight of the positioning result output by the model based on the inertial sensor data based on the dynamic environmental parameter of sensor state, and give priority to the positioning result output by the second type of base learner (i.e., give priority to CSI data), thus improving the accuracy of the final positioning result.

[0191] If the current location is outdoors, the first type of base learner adapted to the outdoor scene and the GPS-based positioning model are called to obtain the positioning results respectively. Then, the final positioning result is obtained based on the meta learner adapted to the outdoor scene. The meta learner can suppress the noise caused by GPS signal jumps, which helps to improve the accuracy of the final positioning result.

[0192] It is understandable that, because the model being called is adapted to the outdoor scene, it is necessary to accurately obtain the model's input data. Therefore, if the terminal determines that the GPS signal is lost, it will activate a backup model (base learner), such as a sensor-based positioning model.

[0193] Combined with a redundancy backup mechanism, the above process can be implemented by the first positioning network element. If the first positioning network element cannot locate the device (e.g., due to equipment failure), the second positioning network element will be activated for positioning.

[0194] In this example, the optimized measurement described in the previous example can also be performed to further improve the positioning accuracy.

[0195] In this example, for indoor-outdoor switching scenarios, the positioning performance of different scenarios can be evaluated periodically. For example, the positioning results can be compared with the actual measurement results to obtain the performance evaluation results. For areas with poor performance, such as shopping mall entrances and exits, which are areas with high frequency of indoor-outdoor switching, an MLP meta-learner specifically for that area can be trained to further reduce switching latency and positioning jitter.

[0196] In summary, the positioning method provided by the above embodiments has the following beneficial effects:

[0197] 1. Multi-technology integration improves positioning accuracy:

[0198] Existing technologies (such as traditional cellular networks, GPS, and sensor positioning) rely on a single or few technologies. Static weight fusion cannot dynamically perceive changes in the environment, resulting in insufficient accuracy in complex scenarios.

[0199] The embodiments of this application, through an integrated learning framework, combine the preliminary results of traditional positioning, deep learning-based CSI / RSSI positioning, and GPS / sensor positioning with dynamic environmental parameters (number of base stations, number of satellites, sensor status), and input them into an MLP meta-learner. The dynamic environmental parameters are used to adjust the weights of each technology in real time (e.g., reducing the weight of traditional positioning when there are insufficient base stations). By suppressing the deviation caused by signal interference and environmental changes through feature cross-referencing, multi-dimensional data complementarity and scene adaptive fusion are achieved, significantly improving the positioning accuracy in complex environments (such as urban canyons, indoor-outdoor switching).

[0200] 2. Integrated learning enhances environmental adaptability:

[0201] Existing solutions only support switching between traditional and AI positioning, lack deep integration of multi-technology data, have limited environmental adaptability, and are difficult to cope with complex scenarios such as multipath interference and non-line-of-sight propagation.

[0202] The embodiments of this application train an MLP meta-learner based on a large amount of real-world scenario data. Through the high-order feature interaction of the MLP hidden layer, the potential relationship between different positioning technologies is mined (such as using sensors to compensate for GPS signal obstruction). By combining the feature interaction of multi-source positioning technologies with scene recognition parameters, the system can dynamically adapt to different environments (such as indoor, outdoor, and mountainous areas) to achieve highly robust positioning across scenarios.

[0203] 3. Synergistic optimization of resource efficiency and reliability:

[0204] Existing technologies do not achieve dynamic optimization of resources and accuracy, rely on fixed model configurations, resulting in resource waste or insufficient accuracy, and lack fault tolerance mechanisms, leading to a high risk of common-mode failure.

[0205] The embodiments of this application flexibly adjust the technology combination according to environmental conditions (such as selecting an appropriate model based on indoor and outdoor scenarios), combine the selection of models based on comprehensive parameter combinations, and redundant design to achieve a dynamic balance between resource consumption and positioning accuracy, which is applicable to safety-critical fields such as rail transit.

[0206] Figure 7 This is an example of the composition of a positioning device provided in an embodiment of this application. The positioning device can be a terminal, including but not limited to mobile phones, smart wearable devices (such as smartwatches), and other electronic devices. Taking a mobile phone as an example below, the positioning device may include a processor 110, an internal memory 120, a display screen 130, an antenna 1, an antenna 2, a mobile communication module 140, and a wireless communication module 150, etc.

[0207] It is understood that the structure illustrated in this embodiment does not constitute a specific limitation on the positioning device. In other embodiments, the positioning device may include more or fewer components than illustrated, or combine some components, or split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0208] The processor 110 may include one or more processing units, such as an application processor (AP), a modem processor, a digital signal processor (DSP), and / or a baseband processor.

[0209] Internal memory 120 can be used to store executable program code, which includes instructions. Processor 110 performs various functions of the electronic device by executing the instructions stored in internal memory 120.

[0210] The wireless communication function of electronic devices can be realized through antenna 1, antenna 2, mobile communication module 140, wireless communication module 150, modem processor, and baseband processor.

[0211] Antenna 1 and antenna 2 are used to transmit and receive electromagnetic wave signals.

[0212] The mobile communication module 140 can provide solutions for wireless communication applications, including 2G / 3G / 4G / 5G, in electronic devices.

[0213] In some embodiments, the mobile communication module 140 includes a communication interface coupled to the processor 110. This communication interface may be a transceiver or an input / output interface. In some embodiments, when the positioning device is a chip configured in the terminal, the communication interface may be an input / output interface.

[0214] The wireless communication module 150 can provide solutions for wireless communication applications in electronic devices, including wireless local area networks (WLAN) (such as wireless fidelity (Wi-Fi) networks), Bluetooth (BT), global navigation satellite system (GNSS), frequency modulation (FM), near field communication (NFC), and infrared (IR) technologies.

[0215] In addition, an operating system runs on top of the aforementioned components. Examples include iOS, Android, and Windows. Applications can be installed and run on this operating system.

[0216] Figure 8 Another example of the composition of a positioning device provided in an embodiment of this application. This positioning device may be a network device, such as an LMF. Figure 8 A simplified schematic diagram of a network device is shown. The network device includes at least one processor 210, at least one memory 220, at least one transceiver 230, at least one network interface 240, and one or more antennas 250. The processor 210, memory 220, transceiver 230, and network interface 240 are connected, for example, via a bus. In this embodiment, the connection may include various interfaces, transmission lines, or buses, etc., and this embodiment is not limited in this respect. The antenna 250 is connected to the transceiver 230. The network interface 240 is used to enable the network element to connect to other communication devices through a communication link. For example, the network interface 240 may include a network interface between the network element and network elements in the core network, such as an S1 interface, or a network interface between the network element and other network elements, such as an X2 or Xn interface.

[0217] Figure 8The processor 210 shown can specifically perform the functions of each step executed by the LMF in the above positioning method, the memory 220 can perform the storage action in the above positioning method, the transceiver 230 and the antenna 250 can perform the transmission and reception action in the above positioning method, and the network interface 240 can perform the interaction action between the LMF and the terminal and / or the base station in the above positioning method.

[0218] Processor 210 may include, but is not limited to, at least one of the following: a central processing unit (CPU), a microprocessor, a digital signal processor (DSP), a microcontroller unit (MCU), or an artificial intelligence processor, etc., which are various computing devices that run software. Each computing device may include one or more cores for executing software instructions to perform calculations or processing. The processor may be a standalone semiconductor chip or integrated with other circuits into a single semiconductor chip. For example, it may form a System-on-a-Chip (SoC) with other circuits (such as encoding / decoding circuits, hardware acceleration circuits, or various bus and interface circuits), or it may be integrated as a built-in processor within an ASIC. The ASIC with the integrated processor may be packaged separately or together with other circuits. In addition to the cores for executing software instructions to perform calculations or processing, the processor may further include necessary hardware accelerators, such as field-programmable gate arrays (FPGAs), programmable logic devices (PLDs), or logic circuits that implement dedicated logic operations.

[0219] The memory 220 may include at least one of the following types, but is not limited to: read-only memory (ROM) or other types of static storage devices that can store static information and instructions, random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or electrically erasable programmable-only memory (EEPROM).

[0220] Transceiver 230 can be used to support the reception or transmission of radio frequency (RF) signals between network elements and other devices. Transceiver 230 can be connected to antenna 250. Transceiver 230 includes a transmitter Tx and a receiver Rx. Specifically, one or more antennas 250 can receive RF signals. The receiver Rx of transceiver 230 is used to receive RF signals from the antennas, convert the RF signals into digital baseband signals or digital intermediate frequency (IF) signals, and provide the digital baseband signals or IF signals to processor 210 so that processor 210 can perform further processing on the digital baseband signals or IF signals, such as demodulation and decoding. In addition, the transmitter Tx in transceiver 230 is also used to receive modulated digital baseband signals or IF signals from processor 210, convert the modulated digital baseband signals or IF signals into RF signals, and transmit the RF signals through one or more antennas 250. Specifically, the receiver Rx can selectively perform one or more stages of downmixing and analog-to-digital conversion on the radio frequency signal to obtain a digital baseband signal or a digital intermediate frequency (IF) signal. The order of the downmixing and IF processing is adjustable. The transmitter Tx can selectively perform one or more stages of upmixing and digital-to-analog conversion on the modulated digital baseband or digital IF signal to obtain a radio frequency signal. The order of the upmixing and IF processing is also adjustable. Digital baseband signals and digital IF signals can be collectively referred to as digital signals.

[0221] The transceiver 230 can also be referred to as an input / output interface or a communication interface, etc. In some embodiments, when the positioning device described above is a chip configured in a satellite, the transceiver 230 can be an input / output interface.

[0222] It should be understood that Figure 8 This is for illustrative purposes only and not as a limitation. The network devices mentioned above, including processors, memory, and transceivers, may be independent of... Figure 8 The structure shown.

[0223] This application also provides a positioning device.

[0224] like Figure 9 As shown, the positioning device 300 includes a processing module 301 and a transceiver module 302. In some embodiments, the positioning device may further include a storage module 303, which can be used to store instructions (code or program) and / or data. The processing module 301 and the transceiver module 302 can be coupled to the storage module 303. For example, the processing module 301 can read instructions (code or program) and / or data from the storage module to implement the corresponding method. The above modules can be set independently, or partially or completely integrated.

[0225] The processing module 301, transceiver module 302, and storage module 303 in this application embodiment are used to enable the positioning device 300 to perform the functions of the network device (such as core network device or base station) in the above method embodiment, or to enable the positioning device 300 to perform the functions of the terminal in the above method embodiment.

[0226] The following describes each module in the positioning device 300, which is used to implement the functions of the terminal in the above method embodiment.

[0227] In some embodiments, the positioning device 300 can correspondingly implement the functions or steps implemented by the terminal in the above-described method embodiments. Specifically, the processing module 301 is used to obtain intermediate positioning results, which include: a first positioning result and a second positioning result. The first positioning result is obtained based on a first positioning method, and the second positioning result is obtained based on a second positioning method. The first positioning method and the second positioning method are different. Based on the intermediate positioning results and dynamic environmental parameters, the terminal's positioning result is obtained. The dynamic environmental parameters are used to configure the weight of the intermediate positioning results. The dynamic environmental parameters represent at least one of the number of base stations associated with the terminal, the number of satellites associated with the terminal, and the sensor status associated with the terminal.

[0228] During the execution of the above process, the information or data interaction function required by the processing module 301 is executed by the transceiver module 302. For details of the required information or data interaction, please refer to the above method embodiment.

[0229] It should be noted that the information interaction and execution process between the modules of the above-mentioned device are based on the same concept as the method embodiment of this application, and the resulting technical effects are the same as those of the method embodiment of this application. For details, please refer to the description in the method embodiment shown above in this application, and it will not be repeated here.

[0230] In other embodiments, the positioning device 300 can correspondingly implement the functions or steps implemented by LMF in the above-described method embodiments. Specifically, the processing module 301 is used to obtain intermediate positioning results, which include: a first positioning result and a second positioning result. The first positioning result is obtained based on a first positioning method, and the second positioning result is obtained based on a second positioning method. The first positioning method is different from the second positioning method. Based on the intermediate positioning results and dynamic environmental parameters, the positioning result of the terminal is obtained. The dynamic environmental parameters are used to configure the weight of the intermediate positioning results. The dynamic environmental parameters represent at least one of the number of base stations associated with the terminal, the number of satellites associated with the terminal, and the sensor status associated with the terminal.

[0231] During the execution of the above process, the information or data interaction function required by the processing module 301 is executed by the transceiver module 302. For details of the required information or data interaction, please refer to the above method embodiment.

[0232] It should be noted that the information interaction and execution process between the modules of the above-mentioned device are based on the same concept as the method embodiment of this application, and the resulting technical effects are the same as those of the method embodiment of this application. For details, please refer to the description in the method embodiment shown above in this application, and it will not be repeated here.

[0233] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the explanations and beneficial effects of the relevant contents in any of the positioning devices provided above can be referred to the corresponding method embodiments provided above, and will not be repeated here.

[0234] This application also provides a processor, including: an input circuit, an output circuit, and a processing circuit. The processing circuit receives signals through the input circuit and transmits signals through the output circuit, causing the processor to execute the positioning method described in the above embodiments.

[0235] In specific implementation, the processor can be one or more chips, the input circuit can be input pins, the output circuit can be output pins, and the processing circuit can be transistors, gate circuits, flip-flops, and various logic circuits. The input signal received by the input circuit can be received and input by, for example, but not limited to, a receiver, and the signal output by the output circuit can be output to, for example, but not limited to, a transmitter and transmitted by the transmitter. Furthermore, the input circuit and the output circuit can be the same circuit, which is used as the input circuit and the output circuit at different times. This application does not limit the specific implementation of the processor and various circuits.

[0236] This application also provides a chip system including one or more processors for calling and executing instructions stored in memory, causing the positioning method described in the above embodiments to be executed. The chip system may be composed of a chip or may include a chip and other discrete devices. The chip system may include input circuitry or an interface for transmitting information or data, and output circuitry or an interface for receiving information or data.

[0237] This application also provides a computer-readable storage medium storing instructions that, when executed on one or more computing devices, cause the one or more computing devices to perform the positioning method described in the above embodiments.

[0238] Computer-readable storage media can be non-transitory computer-readable storage media, such as read-only memory (ROM), random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage devices.

[0239] This application also provides a computer program product. When executed by one or more computing devices, the computer program product allows the computing devices to perform any of the aforementioned location methods. The computer program product can be a software installation package. When any of the aforementioned location methods is required, the computer program product can be downloaded and executed on a computer.

[0240] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. 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 of the technical features. 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.

Claims

1. A positioning method, characterized in that, include: The intermediate positioning results are obtained, including a first positioning result and a second positioning result. The first positioning result is obtained based on a second type of base learner and a first comprehensive parameter combination. The second positioning result is obtained based on a third type of base learner and a second comprehensive parameter combination. The first comprehensive parameter combination is extracted based on Channel State Information (CSI) data and Received Signal Strength Indication (RSSI) data obtained by the terminal. If the first comprehensive parameter combination contains time-series data, the second type of base learner includes a Long Short-Term Memory (LSTM) network-type deep learning model. If the first comprehensive parameter combination contains multi-antenna input-output (MIO) data, the second type of base learner includes a convolutional neural network-type deep learning model. The scenario reflected by the second comprehensive parameter combination includes: the environment in which the terminal is located, the motion state, and the current reliability of each sensor of the terminal. The third type of base learner is adapted to the scenario. The intermediate positioning results are processed to obtain a third comprehensive parameter combination. The third comprehensive parameter combination includes the number of positioning modes, the number of groups of positioning results for each positioning mode, and the characteristics of multiple groups of positioning results for the same positioning mode. The characteristics include the magnitude of dispersion, the level of consistency, and the number of outliers. The positioning modes include positioning methods used to obtain the first positioning result and the second positioning result. Based on the intermediate positioning results, dynamic environmental parameters, and a meta-learner that is appropriately matched with the third comprehensive parameter set, the positioning result of the terminal is obtained. The dynamic environmental parameters are used to configure the weights of the intermediate positioning results. The dynamic environmental parameters represent at least one of the number of base stations associated with the terminal, the number of satellites associated with the terminal, and the sensor status associated with the terminal.

2. The method according to claim 1, characterized in that, The third type of base learner is adapted to the indoor and outdoor scene where the terminal is located. The indoor and outdoor scene includes: indoor, outdoor, from indoor to outdoor, or from outdoor to indoor.

3. The method according to any one of claims 1-2, characterized in that, The first combination of comprehensive parameters includes the following data: The number of transceiver antennas of the terminal, the frequency band used by the terminal, the number of historical data sets of Channel State Information (CSI) acquired by the terminal, and the number of historical data sets of Received Signal Strength Indication (RSSI). The second combination of integrated parameters includes at least one of the following data: satellite positioning data and sensor data.

4. The method according to any one of claims 1-2, characterized in that, Before obtaining the intermediate positioning results, the following is also included: The second type of base learner and the third type of base learner are downloaded from the model management platform. The second type of base learner is selected from multiple models by the model management platform based on the first comprehensive parameter combination, and the third type of base learner is selected from multiple models by the model management platform based on the second comprehensive parameter combination.

5. The method according to claim 1, characterized in that, The meta-learner includes a hidden layer for assigning the weights to the intermediate localization results.

6. The method according to claim 5, characterized in that, Before obtaining the terminal's location result, the following steps are also included: The meta-learner is downloaded from the model management platform, and the meta-learner is selected from multiple models by the model management platform based on the third comprehensive parameter combination.

7. The method according to claim 1 or 5, characterized in that, After obtaining the location result of the terminal, the method further includes: The positioning result of the terminal is compared with the positioning result output by the first type of base learner to obtain the first difference, and the first type of base learner implements the cellular positioning mode. Based on the first difference and the first threshold, the configuration of the base station's transmitted signal is adjusted.

8. The method according to claim 1 or 5, characterized in that, After obtaining the location result of the terminal, the method further includes: The positioning result of the terminal is compared with the second positioning result to obtain the second difference; Based on the second difference and the second threshold, at least one model is retrained to obtain the second localization result from a third type of base learner and the localization result from the terminal.

9. A positioning device, characterized in that, include: A module for implementing the positioning method according to any one of claims 1-8.

10. A positioning device, characterized in that, include: One or more processors, and a memory; the memory is used to store program code; The processor is used to run the program code, so that the positioning device implements the positioning method as described in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, It stores instructions that, when executed on an electronic device, cause the electronic device to perform the positioning method as described in any one of claims 1 to 8.

12. A chip system, characterized in that, include: At least one processor and an interface, the interface being used to receive code instructions and transmit them to the at least one processor; The at least one processor executes the code instructions to implement the positioning method according to any one of claims 1 to 8.

Images