Ai model-based positioning method, device and storage medium

Abstract

An AI model-based positioning method, a device and a storage medium, relating to the technical field of communications. A terminal device uses a PRS and an environmental parameter combination to determine one or more matching AI model indexes and sends same to an LMF network element, so as to acquire a corresponding AI model. The terminal device uses the environmental parameter combination to determine the weighting of a positioning result output by each AI model, and uses the PRS, each AI model and the weighting corresponding to each AI model to determine a final positioning result. In addition, the LMF network element may also determine an AI model index, and the LMF network element may determine a final positioning result, and send the final positioning result to the terminal device. The present solution can perform positioning by means of using a plurality of AI models having relatively high matching degrees to current environmental parameters, and, on the basis of weightings, fuse positioning results so as to obtain an accurate final positioning result. The present solution is suitable for regions where the real environment is complex and changeable or where AI model training is insufficient.

Description

Positioning method, device and storage medium based on AI model

[0001] The present application claims priority to the Chinese patent application No. 202411250465.2, filed on September 6, 2024, and entitled "Positioning method, device and storage medium based on AI model", the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0002] The present application relates to the technical field of wireless communication, in particular to a positioning method, device and storage medium based on AI model. BACKGROUND

[0003] In a new radio (NR) system or a long term evolution (LTE) system, it is usually necessary to manage user equipment (UE) mobility based on a base station, for example, to position the UE. In order to achieve UE mobility management, the accuracy of positioning is particularly important. Artificial intelligence (AI) is introduced into a wireless communication network and can be applied to a UE positioning scenario.

[0004] In some related solutions, a plurality of AI models trained for different combinations of environmental parameters can be used to form a model library, and the model library can be deployed in a location management function (LMF) network element. When performing location determination, the actual combination of environmental parameters of the UE to be positioned is obtained, and an AI model whose combination of environmental parameters used in training set generation and model training is the same as the actual combination of environmental parameters is matched from the model library. The output of the AI model is taken as a positioning result.

[0005] However, in actual application, the number of AI models obtained by training is generally limited, and the real environment is very complex, and there may be no corresponding AI model for all combinations of environmental parameters, resulting in that the applicable AI model cannot be matched according to the actual combination of environmental parameters of the UE and the positioning cannot be completed. SUMMARY

[0006] In order to solve the above problems, the present application provides a positioning method, device and storage medium based on AI model, which can use a plurality of AI models with high matching degree of current environmental parameters to perform positioning and fuse the positioning results according to weights to obtain an accurate final positioning result.

[0007] In a first aspect, the present application provides an AI model-based positioning method, which can be applied to a terminal device. The method comprises: determining one or more AI model indexes matched by using a combination of a received positioning reference signal (PRS) and environment parameters, wherein the combination of environment parameters comprises state parameters of the terminal device and state parameters of a network device; sending first index information, wherein the first index information comprises the one or more AI model indexes; receiving an AI model corresponding to each AI model index in the one or more AI model indexes; determining a weight of a positioning result output by each AI model by using the combination of environment parameters; and determining a final positioning result by using the PRS, each AI model, and the weight corresponding to each AI model.

[0008] In this implementation, the terminal device can determine one or more AI model indexes corresponding to AI models used for positioning, and then send all the AI model indexes to a network device, i.e., to an LMF network element. The LMF network element matches corresponding AI models from an AI model library according to each AI model index and sends the AI models to the terminal device. The terminal device can determine a weight corresponding to each AI model, and determine a final positioning result by using each AI model and the corresponding weight. In the prior art, when an AI model that completely matches a current combination of environment parameters of a UE cannot be determined, it is directly considered that positioning cannot be completed. However, the present application can use multiple AI models that have a high degree of matching with a current combination of environment parameters of the terminal device to perform positioning, and fuse positioning results according to weights to obtain a more accurate final positioning result. This is particularly suitable for service areas with complex and changeable real environments, or areas where AI models used for positioning are insufficiently trained, and has high practicality.

[0009] Meanwhile, after the LMF network element sends the AI models to the terminal device, for a scenario in which the terminal device side needs to continuously perform positioning, the UE side can directly use the received AI models to perform positioning. If the combination of environment parameters does not change subsequently, the UE side no longer needs to frequently interact with the LMF network element, thereby reducing the time delay of positioning.

[0010] In a possible implementation, the PRS carries one or more of a cell identifier (Cell ID), a reference signal received power (RSRP), and an angle of arrival (AoA) relative to a base station.

[0011] In a possible implementation, the cell identifier Cell ID is carried in the PRS, and the state parameter of the terminal device and the state parameter of the network device are included in the environmental parameter combination; the state parameter of the terminal device includes one or more of a spatial region code, a time state parameter, and a scene state parameter, wherein the spatial region code indicates a grid region where a cell corresponding to each Cell ID is located, the time state parameter is used to indicate a positioning time, and the scene state parameter is an identifier of a scene where the terminal device is located; and the state parameter of the network device indicates a parameter configuration of the network device.

[0012] In a possible implementation, the scene state parameter is determined by the terminal device based on detection results of a plurality of sensors, and the sensors include one or more of an air pressure sensor, an accelerometer, and a gyroscope.

[0013] In a possible implementation, the cell identifier Cell ID is carried in the PRS, and the state parameter of the terminal device includes a spatial region code, a time state parameter, and a scene state parameter; the one or more AI model indexes that match are determined by using the received PRS and the environmental parameter combination, including: determining a first Cell ID list, and the first Cell ID list includes all Cell IDs carried in the PRS; respectively determining an intersection of a second Cell ID list corresponding to each spatial region code and the first Cell ID list, and the second Cell ID list includes Cell IDs of all cells in a grid region corresponding to the spatial region code; taking a spatial region code corresponding to an intersection in which a number of Cell IDs is greater than or equal to a first threshold value as a first screening condition; when a time interval between the positioning time and a reference value corresponding to a time interval is less than a preset interval, taking a time state parameter corresponding to the time interval as a second screening condition, and the reference value corresponding to the time interval is a middle value of the time interval; taking a scene state parameter as a third screening condition; determining a fourth screening condition according to the state parameter of the network device; and determining the one or more AI model indexes that match according to the first screening condition, the second screening condition, the third screening condition, and the fourth screening condition, and a mapping table of the environmental parameter combination and the AI model index.

[0014] In a possible implementation, the cell identifier Cell ID and the RSRP are carried in the PRS, the state parameters of the terminal device include the spatial region code, the time state parameter and the scene state parameter, and the one or more AI model indexes matched are determined by using the received positioning reference signal PRS and the environmental parameter combination, including: determining a first Cell ID list, the first Cell ID list including all the Cell IDs carried in the PRS; determining the intersection of each spatial region code corresponding second Cell ID list and the first Cell ID list respectively, the second Cell ID list including all the Cell IDs of the cells in the grid area corresponding to the spatial region code; taking the spatial region code corresponding to the intersection meeting the following condition as the first screening condition: the number of the Cell IDs with the RSRP greater than or equal to the preset power threshold and the AoA relative to the base station meeting the preset angle range is greater than the first threshold; when the time interval between the positioning time and the reference value corresponding to the time interval is less than the preset interval, taking the time state parameter corresponding to the time interval as the second screening condition, the reference value corresponding to the time interval being the middle value of the time interval; taking the scene state parameter as the third screening condition; determining the fourth screening condition according to the state parameters of the network device; and determining the one or more AI model indexes matched according to the first screening condition, the second screening condition, the third screening condition, the fourth screening condition, and the mapping table of the environmental parameter combination and the AI model index.

[0015] In a possible implementation, the cell identifier Cell ID, the RSRP and the AoA relative to the base station are carried in the PRS, and the one or more AI model indexes matched are determined by using the received positioning reference signal PRS and the environmental parameter combination, including: determining a first Cell ID list, the first Cell ID list including all the Cell IDs carried in the PRS; determining the intersection of each spatial region code corresponding second Cell ID list and the first Cell ID list respectively, the second Cell ID list including all the Cell IDs of the cells in the grid area corresponding to the spatial region code; taking the spatial region code corresponding to the intersection meeting the following condition as the first screening condition: the number of the Cell IDs with the RSRP greater than or equal to the preset power threshold and the AoA relative to the base station meeting the preset angle range is greater than the first threshold; when the time interval between the positioning time and the reference value corresponding to the time interval is less than the preset interval, taking the time state parameter corresponding to the time interval as the second screening condition, the reference value corresponding to the time interval being the middle value of the time interval; taking the scene state parameter as the third screening condition; determining the fourth screening condition according to the state parameters of the network device; and determining the one or more AI model indexes matched according to the first screening condition, the second screening condition, the third screening condition, the fourth screening condition, and the mapping table of the environmental parameter combination and the AI model index.

[0016] In a possible implementation, the state parameter of the network device indicates a parameter configuration of the network device, and the weight of the positioning result output by each AI model is determined by using the environment parameter combination, and specifically includes: determining a first weight coefficient of each AI model according to the number of Cell IDs in the intersection of the respective space region encodings as a first screening condition, the first weight coefficient being positively correlated with the number of Cell IDs in the intersection; determining a second weight coefficient of each AI model according to the time interval, the second weight coefficient being negatively correlated with the time interval; determining a third weight coefficient of each AI model according to the scene state parameter; determining a fourth weight coefficient of each AI model according to the state parameter of the network device; determining the weight coefficient of each AI model according to the respective first weight coefficients, the respective second weight coefficients, the respective third weight coefficients and the respective fourth weight coefficients; and performing normalization processing on the weight coefficient of each AI model to obtain the weight of each AI model, and the sum of the weights of the AI models being 1.

[0017] In a possible implementation, the state parameter of the network device indicates a parameter configuration of the network device in different time periods, and the weight coefficient of each AI model is determined according to the respective first weight coefficients, the respective second weight coefficients, the respective third weight coefficients and the respective fourth weight coefficients, including: for each AI model, performing the following steps: summing the product of a first influence factor and the first weight coefficient, the product of the first influence factor and the third weight coefficient, the product of a second influence factor and the second weight coefficient, and the product of the second influence factor and the fourth weight coefficient to determine the weight coefficient of each AI model, wherein the first influence factor is greater than the second influence factor.

[0018] In a possible implementation, the final positioning result is determined by using the PRS, the AI models and the weights corresponding to the AI models, including: taking the PRS as the input of each AI model to obtain the positioning result output by each AI model; multiplying the positioning result output by each AI model by the weight corresponding to the AI model to obtain a final positioning result component corresponding to each AI model; and superimposing the final positioning result components corresponding to the AI models to obtain the final positioning result.

[0019] In a second aspect, the present application also provides an AI model-based positioning method, which can be applied to a terminal device, and the method includes: sending an environment parameter combination and a positioning reference signal PRS, the environment parameter combination including a state parameter of the terminal device and a state parameter of a network device; receiving one or more AI models, each AI model in the one or more AI models corresponding to an AI model index, and each AI model index being determined by the environment parameter combination and the PRS; determining the weight of the positioning result output by each AI model by using the environment parameter combination; and determining a final positioning result by using the PRS, the AI models and the weights corresponding to the AI models.

[0020] In this implementation, after the terminal device sends the environment parameter combination and the PRS to the LMF network element, the LMF can determine one or more AI model indexes corresponding to the AI model for positioning, and then the LMF network element matches the corresponding AI model from the AI model library according to each AI model index and sends it to the terminal device. After the terminal device side determines the respective weights corresponding to each AI model, the terminal device determines the final positioning result by using each AI model and the corresponding weight. In the existing related scheme, when an AI model that completely matches the current environment parameter combination of the terminal device cannot be determined, it is directly considered that positioning cannot be completed. However, the scheme provided in the embodiments of the present application can use multiple AI models that have a high matching degree with the current environment parameters of the terminal device for positioning, and fuse the positioning results according to the weights to obtain a more accurate final positioning result. This scheme is especially suitable for service areas with complex and changeable real environments, or areas where the AI model for positioning is not fully trained, and has high practicability.

[0021] At the same time, after the LMF network element sends the AI model to the terminal device, for the scenario where the terminal device side needs to perform continuous positioning, the terminal device side can directly use the received AI model for positioning, and if the environment parameter combination does not change subsequently, it is no longer necessary to frequently interact with the LMF network element, thereby reducing the time delay of positioning.

[0022] In a possible implementation, the PRS carries one or more of a cell identifier Cell ID, a reference signal receiving power RSRP, and an angle of arrival AoA relative to the base station.

[0023] In a possible implementation, the PRS carries a cell identifier Cell ID, and the environment parameter combination includes a state parameter of the terminal device and a state parameter of the network device; the state parameter of the terminal device includes one or more of a spatial region code, a time state parameter, and a scene state parameter, wherein the spatial region code indicates a grid region where the cell corresponding to each Cell ID is located, the time state parameter is used to indicate a positioning time, and the scene state parameter is an identifier of a scene in which the terminal device is located; and the state parameter of the network device indicates a parameter configuration of the network device.

[0024] In a possible implementation, the cell identifier (Cell ID) is carried in the PRS, the state parameter of the terminal device includes a spatial region code, a time state parameter, and a scene state parameter, and a weight of a positioning result output by each AI model is determined by combining the environment parameters, specifically including: determining a first Cell ID list, the first Cell ID list including all Cell IDs carried in the PRS; determining an intersection of each spatial region code corresponding second Cell ID list and the first Cell ID list respectively, the second Cell ID list including Cell IDs of all cells in a grid region corresponding to the spatial region code; taking a spatial region code corresponding to an intersection whose number of Cell IDs is greater than or equal to a first threshold as a first screening condition; determining a first weight coefficient of each AI model according to the number of Cell IDs in the intersection corresponding to each spatial region code as the first screening condition, the first weight coefficient being positively correlated with the number of Cell IDs in the intersection; determining a second weight coefficient of each AI model according to a time interval, the second weight coefficient being negatively correlated with the time interval; determining a third weight coefficient of each AI model according to the scene state parameter; determining a fourth weight coefficient of each AI model according to the state parameter of the network device; determining a weight coefficient of each AI model according to the first weight coefficient, the second weight coefficient, the third weight coefficient, and the fourth weight coefficient; and performing normalization processing on the weight coefficient of each AI model to obtain a weight of each AI model, the sum of the weights of the AI models being 1.

[0025] In a possible implementation, the Cell IDs of all cells in the grid region corresponding to the spatial region code include Cell IDs of all cells in the spatial region corresponding to the code that can be measured by the positioning assistance data.

[0026] In a possible implementation, the Cell ID includes a cell global identifier (CGI) and / or an E-UTRAN cell global identifier (ECGI or ECI). The second Cell ID list can also include one or more of a physical cell identifier (PCI), an absolute radio-frequency channel number (ARFCN), and the like of all cells in the grid region corresponding to the spatial region code, to assist in identifying the cells.

[0027] In a possible implementation, the cell identification Cell ID and RSRP are carried in the PRS, the state parameters of the terminal device include a spatial region code, a time state parameter and a scene state parameter, a weight of a positioning result output by each AI model is determined by using an environmental parameter combination, and the weight specifically includes: determining a first Cell ID list, the first Cell ID list including all Cell IDs carried in the PRS; respectively determining intersections of each spatial region code corresponding second Cell ID list and the first Cell ID list, the second Cell ID list including Cell IDs of all cells in a grid region corresponding to the spatial region code; taking, as a first screening condition, a spatial region code corresponding to an intersection that meets the following condition: a number of Cell IDs with RSRP greater than or equal to a preset power threshold is greater than a first threshold; determining a first weight coefficient of each AI model according to a number of Cell IDs in the intersection corresponding to each spatial region code that is the first screening condition, the first weight coefficient being positively correlated with the number of Cell IDs in the intersection; determining a second weight coefficient of each AI model according to a time interval, the second weight coefficient being negatively correlated with the time interval; determining a third weight coefficient of each AI model according to the scene state parameter; determining a fourth weight coefficient of each AI model according to the state parameter of the network device; determining a weight coefficient of each AI model according to the first weight coefficient, the second weight coefficient, the third weight coefficient and the fourth weight coefficient; and performing normalization processing on the weight coefficient of each AI model to obtain a weight of each AI model, the sum of the weights of the AI models being 1.

[0028] In a possible implementation, the cell identifier Cell ID is carried in the PRS, the state parameter of the terminal device includes a spatial region code, a time state parameter and a scene state parameter, and a weight of a positioning result output by each AI model is determined by combining the environment parameters, and specifically includes: determining a first Cell ID list, the first Cell ID list including all Cell IDs carried in the PRS; determining an intersection of a second Cell ID list corresponding to each spatial region code and the first Cell ID list respectively, the second Cell ID list including Cell IDs of all cells in a grid region corresponding to the spatial region code; taking a spatial region code corresponding to an intersection meeting the following conditions as a first screening condition: the number of Cell IDs whose RSRP is greater than or equal to a preset power threshold and whose AoA relative to the base station meets a preset angle range is greater than a first threshold; the number of Cell IDs whose RSRP is greater than or equal to the preset power threshold is greater than the first threshold; determining a first weight coefficient of each AI model according to the number of Cell IDs in the intersection corresponding to each spatial region code that is the first screening condition, the first weight coefficient being positively correlated with the number of Cell IDs in the intersection; determining a second weight coefficient of each AI model according to a time interval, the second weight coefficient being negatively correlated with the time interval; determining a third weight coefficient of each AI model according to the scene state parameter; determining a fourth weight coefficient of each AI model according to the state parameter of the network device; determining the weight coefficient of each AI model according to the first weight coefficient, the second weight coefficient, the third weight coefficient and the fourth weight coefficient; performing normalization processing on the weight coefficient of each AI model to obtain the weight of each AI model, and the sum of the weights of the AI models being 1.

[0029] In a possible implementation, the state parameter of the network device indicates parameter configurations of the network device in different time periods, and the weight coefficient of each AI model is determined according to the first weight coefficient, the second weight coefficient, the third weight coefficient and the fourth weight coefficient, including: for each AI model, performing the following steps: summing a product of a first influence factor and the first weight coefficient, a product of the first influence factor and the third weight coefficient, a product of a second influence factor and the second weight coefficient, and a product of the second influence factor and the fourth weight coefficient to determine the weight coefficient of each AI model, wherein the first influence factor is greater than the second influence factor.

[0030] In a possible implementation, the PRS, the AI models, and the weights corresponding to the AI models are used to determine the final positioning result, including: taking the PRS as input of the AI models to obtain positioning results output by the AI models; multiplying the positioning results output by the AI models by the weights corresponding to the AI models to obtain final positioning result components corresponding to the AI models; and superimposing the final positioning result components corresponding to the AI models to obtain the final positioning result.

[0031] In a third aspect, the embodiments of the present application also provide a positioning method of an AI model, which can be applied to a terminal device and includes: determining one or more AI model indexes matched by using a combination of a received positioning reference signal (PRS) and environment parameters, the combination of environment parameters including state parameters of the terminal device and state parameters of a network device; sending first index information, the PRS, and the combination of environment parameters, the first index information including the one or more AI model indexes; and receiving position information, the position information including a final positioning result, the final positioning result being determined by the PRS, one or more AI models, and a weight corresponding to each AI model in the one or more AI models, each AI model corresponding to one AI model index in the one or more AI model indexes, and the weight being a weight of a positioning result output by each AI model determined by using the combination of environment parameters.

[0032] In this implementation, the terminal device can determine one or more AI model indexes corresponding to AI models used for positioning, and then send all the AI model indexes to the LMF network element. The LMF network element matches corresponding AI models from the AI model library according to the AI model indexes, and determines weights corresponding to the AI models. Then, the LMF can determine a final positioning result by using the AI models and the corresponding weights, and send the final positioning result to the terminal device. In the existing scheme, when an AI model completely matched with the current combination of environment parameters of the terminal device cannot be determined, it is directly considered that positioning cannot be completed. However, the scheme of the present application can use multiple AI models that are highly matched with the current combination of environment parameters of the terminal device for positioning, and fuse the positioning results according to the weights to obtain a more accurate final positioning result. This scheme is particularly suitable for service areas with complex and changeable real environments, or areas where AI models used for positioning are insufficiently trained, and has high practicability.

[0033] Meanwhile, for scenarios in which continuous positioning is not required at the terminal device side, such as scenarios in which single positioning or positioning with a small number of times is performed, the terminal device side can directly obtain the positioning result sent by the LMF network element. In this way, the terminal device side does not need to receive the AI models sent by the LMF network element, the amount of data received can be reduced, and the time delay of single positioning is also reduced.

[0034] In a fourth aspect, the embodiments of the present application also provide a positioning method of an AI model, which can be applied to a terminal device and includes: sending an environment parameter combination and a positioning reference signal (PRS), wherein the environment parameter combination includes a state parameter of the terminal device and a state parameter of a network device; and receiving position information, wherein the position information includes a final positioning result, the final positioning result is determined by the PRS, one or more AI models, and a weight corresponding to each AI model in the one or more AI models, the weight is a weight of a positioning result output by each AI model using the environment parameter combination, each AI model corresponds to an AI model index, and each AI model index is determined using the environment parameter combination and the PRS.

[0035] In this implementation, the PRS and the environment parameter combination are sent by the terminal device to the LME network element, the LMF network element can determine the AI model index using the PRS and the environment parameter combination, the LMF network element matches the corresponding AI model from the AI model library according to each AI model index and determines the weight corresponding to each AI model. Then the LMF can determine the final positioning result using each AI model and the corresponding weight, and sends the final positioning result to the terminal device. In the existing related scheme, when an AI model completely matching the current environment parameter combination of the terminal device cannot be determined, it is directly considered that the positioning cannot be completed, while the scheme provided in the embodiments of the present application can use multiple AI models with high matching degree with the current environment parameter of the terminal device for positioning, and fuse the positioning results according to the weights, thereby obtaining a more accurate final positioning result. It is especially suitable for service areas with complex and changeable real environment, or areas where the AI model for positioning is insufficiently trained, and has high practicability.

[0036] Meanwhile, after the LMF network element sends the AI model to the terminal device, for the scenario that the terminal device side needs to continuously position, the terminal device side can directly use the received AI model for positioning, and if the environment parameter combination does not change subsequently, it is no longer necessary to frequently interact with the LMF network element, thereby reducing the time delay of positioning.

[0037] In a fifth aspect, the embodiments of the present application further provide a positioning method of an AI model, which can be applied to a network device, and the network device can be an LMF network element. The method comprises the following steps: receiving first index information, wherein the first index information comprises one or more AI model indexes, and the one or more AI model indexes are determined by combining received positioning reference signals (PRS) and environmental parameters, and the environmental parameters comprise state parameters of a terminal device and state parameters of the network device; and sending AI models corresponding to each of the one or more AI model indexes respectively, wherein each AI model is used to determine a final positioning result together with the PRS and a weight corresponding to the AI model, and the weight corresponding to each AI model is a weight of a positioning result output by each AI model determined by using the environmental parameters.

[0038] In this implementation, the LMF network element matches corresponding AI models from an AI model library according to each AI model index and determines weights corresponding to each AI model, and then sends one or more AI models matched and determined to a terminal device for fusion positioning. This method is suitable for a service area with complex and changeable real environment, or an area where AI models for positioning are insufficiently trained, and has high practicability.

[0039] In a sixth aspect, the embodiments of the present application further provide a positioning method of an AI model, which can be applied to a network device, and the network device can be an LMF network element. The method comprises the following steps: receiving environmental parameters and positioning reference signals (PRS), wherein the environmental parameters comprise state parameters of a terminal device and state parameters of the network device; determining one or more AI model indexes by using the environmental parameters and the PRS; and sending AI models corresponding to each of the one or more AI model indexes respectively, wherein each AI model is used to determine a final positioning result together with the PRS and a weight corresponding to the AI model, and the weight corresponding to each AI model is a weight of a positioning result output by each AI model determined by using the environmental parameters.

[0040] In this implementation, one or more AI model indexes are determined by the LMF network element, and corresponding AI models are matched from an AI model library according to each AI model index and weights corresponding to each AI model are determined, and then one or more AI models matched and determined are sent to a terminal device for fusion positioning. This method is suitable for a service area with complex and changeable real environment, or an area where AI models for positioning are insufficiently trained, and has high practicability.

[0041] In a possible implementation, the PRS carries one or more of a cell identifier (Cell ID), a reference signal received power (RSRP), and an angle of arrival (AoA) relative to a base station.

[0042] In a possible implementation, the cell identifier Cell ID is carried in the PRS, and the state parameter of the terminal device and the state parameter of the network device are included in the environment parameter combination; the state parameter of the terminal device includes one or more of a space area code, a time state parameter, and a scene state parameter, wherein the space area code indicates a grid area in which a cell corresponding to each Cell ID is located, the time state parameter is used to indicate a positioning time, and the scene state parameter is an identifier of a scene in which the terminal device is located; and the state parameter of the network device indicates a parameter configuration of the network device.

[0043] In a possible implementation, the cell identifier Cell ID is carried in the PRS, and the state parameter of the terminal device includes a space area code, a time state parameter, and a scene state parameter; and one or more AI model indexes that match are determined by using the environment parameter combination and the PRS, specifically including: determining a first Cell ID list, the first Cell ID list including all Cell IDs carried in the PRS; respectively determining an intersection of a second Cell ID list corresponding to each space area code and the first Cell ID list, the second Cell ID list including Cell IDs of all cells in a grid area corresponding to the space area code; taking, as a first screening condition, a space area code corresponding to an intersection in which a number of Cell IDs is greater than or equal to a first threshold value; taking, as a second screening condition, a time state parameter corresponding to a time interval when a time interval between a positioning time and a reference value corresponding to the time interval is less than a preset interval, the reference value corresponding to the time interval being a middle value of the time interval; taking, as a third screening condition, a scene state parameter in which the terminal device is located; determining a fourth screening condition according to the state parameter of the network device; and determining the one or more AI model indexes that match according to the first screening condition, the second screening condition, the third screening condition, and the fourth screening condition, and a mapping table of the environment parameter combination and the AI model index.

[0044] In a possible implementation, the cell identification Cell ID and the RSRP are carried in the PRS, the state parameters of the terminal device include the spatial region code, the time state parameter and the scene state parameter, and one or more AI model indexes that match are determined by using the environmental parameter combination and the PRS, specifically including: determining a first Cell ID list, the first Cell ID list including all the Cell IDs carried in the PRS; determining an intersection of a second Cell ID list corresponding to each spatial region code and the first Cell ID list respectively, the second Cell ID list including all the Cell IDs of the cells in the grid region corresponding to the spatial region code; taking the spatial region code corresponding to the intersection that meets the following condition as a first screening condition: the number of the Cell IDs whose RSRP is greater than or equal to a preset power threshold is greater than a first threshold; when a time interval between the positioning time and a reference value corresponding to the time interval is less than a preset interval, taking the time state parameter corresponding to the time interval as a second screening condition, the reference value corresponding to the time interval being a middle value of the time interval; taking the scene state parameter as a third screening condition; determining a fourth screening condition according to the state parameters of the network device; and determining the one or more AI model indexes that match according to the first screening condition, the second screening condition, the third screening condition and the fourth screening condition, and a mapping table of the environmental parameter combination and the AI model index.

[0045] In a possible implementation, the cell identification Cell ID, the RSRP and the AoA relative to the base station are carried in the PRS, and one or more AI model indexes that match are determined by using the environmental parameter combination and the PRS, specifically including: determining a first Cell ID list, the first Cell ID list including all the Cell IDs carried in the PRS; determining an intersection of a second Cell ID list corresponding to each spatial region code and the first Cell ID list respectively, the second Cell ID list including all the Cell IDs of the cells in the grid region corresponding to the spatial region code; taking the spatial region code corresponding to the intersection that meets the following condition as a first screening condition: the number of the Cell IDs whose RSRP is greater than or equal to a preset power threshold and whose AoA relative to the base station meets a preset angle range is greater than a first threshold; when a time interval between the positioning time and a reference value corresponding to the time interval is less than a preset interval, taking the time state parameter corresponding to the time interval as a second screening condition, the reference value corresponding to the time interval being a middle value of the time interval; taking the scene state parameter as a third screening condition; determining a fourth screening condition according to the state parameters of the network device; and determining the one or more AI model indexes that match according to the first screening condition, the second screening condition, the third screening condition and the fourth screening condition, and a mapping table of the environmental parameter combination and the AI model index.

[0046] In a seventh aspect, the embodiments of the present application further provide a positioning method of an AI model, which can be applied to a network device, and the network device can be an LMF network element. The method comprises the following steps: receiving first index information, a positioning reference signal (PRS), and an environment parameter combination. The first index information comprises one or more AI model indexes. The one or more AI model indexes are determined by using the PRS and the environment parameter combination. The environment parameter combination comprises state parameters of a terminal device and state parameters of the network device. The method further comprises the following steps: determining an AI model corresponding to each AI model index in the one or more AI model indexes; determining a weight of a positioning result output by each AI model by using the environment parameter combination; determining a final positioning result by using the PRS, the AI models, and the weights corresponding to the AI models; and sending position information, wherein the position information comprises the final positioning result.

[0047] In this implementation, the LMF network element can determine the AI model according to the AI model index, determine the final positioning result, and directly send the final positioning result to the terminal device. The specific implementation of how the LMF network element determines the weight of the positioning result output by each AI model and determines the final positioning result can be referred to the implementation provided in the first aspect.

[0048] In an eighth aspect, the embodiments of the present application further provide a positioning method of an AI model, which can be applied to a network device, and the network device can be an LMF network element. The method comprises the following steps: receiving an environment parameter combination and a positioning reference signal (PRS). The environment parameter combination comprises state parameters of a terminal device and state parameters of the network device. The method further comprises the following steps: determining one or more AI model indexes matched by using the environment parameter combination and the PRS; determining an AI model corresponding to each AI model index in the one or more AI model indexes; determining a weight of a positioning result output by each AI model by using the environment parameter combination; determining a final positioning result by using the PRS, the AI models, and the weights corresponding to the AI models; and sending position information, wherein the position information comprises the final positioning result.

[0049] In this implementation, the LMF network element can determine the AI model index, determine the AI model according to the AI model index, determine the final positioning result, and directly send the final positioning result to the terminal device. The specific implementation of how the LMF network element determines the AI model index, determines the weight of the positioning result output by each AI model, and determines the final positioning result can be referred to the implementation provided in the first aspect.

[0050] In a ninth aspect, the present application also provides a terminal device, comprising a processor and a memory; the processor is coupled with the memory; the memory is configured to store computer programs and / or instructions; and the processor is configured to execute the computer programs and / or instructions stored in the memory to implement the AI model-based positioning method provided in the first aspect and any one of the implementation manners of the first aspect, or the second aspect and any one of the implementation manners of the second aspect, or the third aspect, or the fourth aspect.

[0051] In a tenth aspect, the present application also provides a communication apparatus, comprising a processor, which is configured to execute computer programs or instructions stored in a memory to enable the apparatus to perform the AI model-based positioning method provided in the fifth aspect, or the sixth aspect and any one of the implementation manners of the sixth aspect, or the seventh aspect, or the eighth aspect.

[0052] In an eleventh aspect, the present application also provides a computer program product, which comprises computer programs or instructions for executing the AI model-based positioning method provided in the first aspect and any one of the implementation manners of the first aspect, or the second aspect and any one of the implementation manners of the second aspect, or the third aspect, or the fourth aspect; or computer programs or instructions for executing the AI model-based positioning method provided in the fifth aspect, or the sixth aspect and any one of the implementation manners of the sixth aspect, or the seventh aspect, or the eighth aspect.

[0053] In a twelfth aspect, the present application also provides a computer-readable storage medium, which stores computer programs or instructions, and when the computer programs or instructions are executed, the AI model-based positioning method provided in any one of the aspects or any one of the implementation manners of any one of the aspects is implemented. BRIEF DESCRIPTION OF DRAWINGS

[0054] FIG. 1 is a schematic diagram of a communication system provided by the present application;

[0055] FIG. 2 is a schematic diagram of an NR system in 5G provided by an embodiment of the present application;

[0056] FIG. 3 is a flowchart of an AI model-based positioning method provided by an embodiment of the present application;

[0057] FIG. 4 is a schematic diagram of a scene provided by an embodiment of the present application;

[0058] FIG. 5 is a schematic diagram of a framework for AI model-based positioning provided by an embodiment of the present application;

[0059] FIG. 6 is a flowchart of another AI model-based positioning method provided by an embodiment of the present application;

[0060] FIG. 7 is a schematic diagram of a framework for positioning based on an AI model according to an embodiment of the present application;

[0061] FIG. 8 is a flowchart of another AI model based positioning method according to an embodiment of the present application;

[0062] FIG. 9 is a schematic diagram of a framework for positioning based on an AI model according to an embodiment of the present application;

[0063] FIG. 10 is a flowchart of another AI model based positioning method according to an embodiment of the present application;

[0064] FIG. 11 is a schematic diagram of a framework for positioning based on an AI model according to an embodiment of the present application;

[0065] FIG. 12 is a schematic diagram of a terminal device according to an embodiment of the present application;

[0066] FIG. 13 is a schematic diagram of a communication apparatus according to an embodiment of the present application. DETAILED DESCRIPTION

[0067] In the present application, the word "exemplary" or "for example" is used to mean serving as an example, instance, or illustration. Any implementation described as "exemplary" or "for example" is not necessarily to be construed as preferred or advantageous over other implementations. The term "exemplary" or "for example" is intended to present concepts in a concrete manner.

[0068] It is to be understood that the terms "first", "second", etc. are used herein only to describe different instances and do not connote any relative importance of the associated objects. Thus, a feature defined with "first", "second" etc. can include one or more of the features implicitly or explicitly. It is to be understood that the terms used herein are merely for the purpose of describing particular embodiments and are not intended to be limiting. As used in the description of the various described examples, the singular forms "a", "an", and "the" are intended to include plural forms as well, unless the context clearly indicates otherwise.

[0069] The technical solutions of the embodiments of the present application can be applied to various communication systems, for example: a 5th generation (5G) or new radio (NR) system, a long term evolution (LTE) system, an LTE frequency division duplex (FDD) system, an LTE time division duplex (TDD) system, a wireless local area network (WLAN) system, a satellite communication system, a future communication system such as a 6th generation mobile communication system, or a converged system of multiple systems, and the like.

[0070] The technical solutions provided in the present application can also be applied to device to device (D2D) communication, vehicle-to-everything (V2X) communication, machine to machine (M2M) communication, machine type communication (MTC), and internet of things (IoT) communication system or other communication systems. Of course, future communication systems can also have other naming ways, which are still covered in the inclusive scope of the present application, and the present application does not make any limitation thereto.

[0071] A network element in a communication system can send a signal to another network element or receive a signal from another network element. The signal can include information, signaling, or data, etc. The network element can also be replaced by an entity, a network entity, a device, a communication device, a communication module, a node, a communication node, etc. The present disclosure describes the network element as an example. For example, the communication system can include at least one terminal device and at least one network device. The network device can send a downlink signal to the terminal device, and / or the terminal device can send an uplink signal to the network device. It can be understood that the terminal device in the present disclosure can be replaced by a first network element, and the network device can be replaced by a second network element, both of which perform the corresponding communication method in the present disclosure.

[0072] The terminal device in the embodiments of the present application includes various devices with wireless communication functions, which can be used to connect people, things, machines, etc. The terminal device can be widely used in various scenarios, such as: cellular communication, D2D, V2X, peer to peer (P2P), M2M, MTC, IoT, virtual reality (VR), augmented reality (AR), industrial control, autonomous driving, remote medical treatment, smart power grid, smart furniture, smart office, smart wear, smart traffic, smart city drone, robot, remote sensing, passive sensing, positioning, navigation and tracking, autonomous delivery, etc.

[0073] The terminal device can be a terminal in any of the above scenarios, such as an MTC terminal, an IoT terminal, etc. The terminal device can be a user equipment (UE) of the 3rd generation partnership project (3GPP) standard, a terminal, a fixed device, a mobile station device or a mobile device, a subscriber unit

[0074] (such as a cellular phone, a smart phone, a session initialization protocol (SIP) phone, a wireless data card, a personal digital assistant (PDA), a computer, a tablet computer, a notebook computer, a wireless modem, a handset, a laptop computer, a computer with wireless transceiver function, a smart book, a vehicle, a satellite, a global positioning system (GPS) device, a target tracking device, an aircraft (such as a drone, a helicopter, a multi-helicopter, a quad-helicopter, or an airplane, etc.), a ship, a remote control device, a smart home device, an industrial device, or a device built in the above device (such as a communication module, a modem or a chip in the above device, etc.), or other processing devices connected to the wireless modem. For the convenience of description, the terminal device will be described as a terminal or a UE hereinafter.

[0075] It should be understood that in some scenarios, the UE can also be used to act as a base station. For example, the UE can act as a scheduling entity which provides sidelink signals between UEs in scenarios such as V2X, D2D or P2P, etc.

[0076] In the embodiments of the present application, the apparatus for implementing the function of the terminal device can be a terminal device, or can be an apparatus capable of supporting the terminal device to implement the function, for example, a chip system or a chip, which can be installed in the terminal device. In the embodiments of the present application, the chip system can be composed of a chip, or can include a chip and other discrete devices.

[0077] The network device in the embodiments of the present application can be a device for communicating with the terminal device, and can also be referred to as an access network device or a radio access network device, for example, the network device can be a base station. The network device in the embodiments of the present application can refer to a radio access network (RAN) node (or device) for accessing the terminal device to a wireless network. The base station can broadly cover various names in the following or be replaced by the following names, such as: Node B (NodeB), evolved Node B (eNB), next generation Node B (gNB), relay station, access point, transmission point (TRP), transmitting point (TP), primary station, secondary station, motor slide retainer (MSR) node, home base station, network controller, access node, wireless node, access point (AP), transmission node, transceiver node, baseband unit (BBU), remote radio unit (RRU), active antenna unit (AAU), remote radio head (RRH), central unit (CU), distributed unit (DU), positioning node, etc.

[0078] The base station can be a macro base station, a micro base station, a relay node, a donor node or the like, or a combination thereof. The base station can also refer to a communication module, modem or chip for being arranged in the foregoing devices or apparatuses. The base station can also be a mobile switching center and a device assuming a base station function in D2D, V2X, M2M communication, a network side device in a 6G network, a device assuming a base station function in a future communication system, etc. The base station can support networks of the same or different access technologies. The embodiments of the present application do not limit the specific technology and specific device form of the network device.

[0079] The base station can be fixed or mobile. For example, a helicopter or a drone can be configured to act as a mobile base station, and one or more cells can move according to the location of the mobile base station. In other examples, a helicopter or a drone can be configured to act as a device that communicates with another base station.

[0080] In the embodiments of the present application, the device for implementing the function of the network device can be the network device, or a device capable of supporting the network device to implement the function, such as a chip system or a chip, which can be installed in the network device. In the embodiments of the present application, the chip system can be composed of a chip, or can include a chip and other discrete devices.

[0081] The network device and the terminal device can be deployed on land, including indoors or outdoors, handheld or vehicle-mounted; can also be deployed on the water surface; and can also be deployed on aircraft, balloons and satellites in the air. The scenarios in which the network device and the terminal device are located are not limited in the embodiments of the present application. In addition, the terminal device and the network device can be hardware devices, or software functions running on special hardware, software functions running on general hardware, such as virtualized functions instantiated on a platform (for example, a cloud platform), or entities including special or general hardware devices and software functions. The specific forms of the terminal device and the network device are not limited in the present application.

[0082] In order for those skilled in the art to more clearly understand the solutions of the present application, the application scenarios involved in the technical solutions of the present application will be described first as follows.

[0083] Referring to FIG. 1, this figure is a schematic diagram of a communication system provided by the present application.

[0084] The wireless communication system includes an access network 400 and a core network 200. Optionally, the communication system can also include an Internet 300.

[0085] The wireless access network 100 can be a next-generation (for example, 6G or higher version) wireless access network, or a traditional (for example, 5G, 4G, 3G or 2G) wireless access network.

[0086] One or more terminal devices 420 (only one of which is shown in the figure) can be connected to each other or connected to one or more network devices in the access network 400. In FIG. 1, it is taken as an example that the access network 400 includes a network device 410a and a network device 410b.

[0087] In FIG. 1, only for illustrative purposes, other devices can also be included in the wireless communication system, such as core network devices, wireless relay devices and / or wireless backhaul devices, etc.

[0088] In practical applications, the wireless communication system can comprise multiple network devices (also referred to as access network devices) at the same time, and can comprise multiple terminal devices at the same time, without limitation. One network device can serve one or more terminal devices at the same time. One terminal device can access one or more network devices at the same time.

[0089] Currently, artificial intelligence (AI) is introduced into wireless communication networks and has been widely applied to many application scenarios of air interface technology. A common application scenario is AI-based positioning of terminal devices.

[0090] The AI node for AI-based positioning of terminal devices can be deployed in one or more of the following: a network device, a terminal device, a core network, or a positioning device; or the AI node can also be deployed separately, such as in a location other than any of the above devices.

[0091] Optionally, the AI node is configured to perform AI-related operations. As an example, the AI-related operations can include one or more of model failure testing, model performance testing, model training testing, or data collection.

[0092] For example, the network device can forward data related to the AI model reported by the terminal device to the AI node, and the AI node performs AI-related operations. For another example, the network device or the terminal device can forward data related to the AI model to the AI node, and the AI node performs AI-related operations.

[0093] For another example, the AI node can send one or more of the following outputs of AI-related operations to the network device and / or the terminal device: a trained neural network model, model evaluation, or test results. Optionally, the AI node can directly send the outputs of AI-related operations to the network device and the terminal device. For another example, the AI node can send the outputs of AI-related operations to the terminal device through the network device. For another example, the AI node can send the outputs of AI-related operations to the network device through the terminal device.

[0094] It can be understood that the number of AI nodes is not limited in the present application. For example, when there are multiple AI nodes, the multiple AI nodes can be divided based on functions, such as different AI nodes being responsible for different functions.

[0095] In a related solution, an AI node (including multiple AI models for positioning) can be deployed in a location management function (LMF) network element of the core network 200. The LMF network element can be used to collect positioning-related measurement results, determine the location information of the terminal device according to the collected measurement results, and the like. The LMF network element can also be referred to as a positioning server, a location management device, and the like.

[0096] It can be understood that, in the present application, the apparatus for implementing the function of the location management device can be the location management device, or can be an apparatus capable of supporting the location management device to implement the function, for example, a chip system, a hardware circuit, a software module, or a hardware circuit plus a software module, which can be installed in the location management device or can be used in matching with the location management device.

[0097] For ease of description, the apparatus for implementing the function of the location management device is taken as the location management device, and the location management device is taken as the LMF for example, to describe the technical solutions of the present application.

[0098] Referring to FIG. 2, it is a schematic diagram of an NR system in 5G provided by an embodiment of the present application.

[0099] The base station in the next generation radio access network (NG-RAN) can include a next generation base station (gNB) and a next generation-evolved base station (ng-eNB). Among them, the gNB is a base station in 5G, and the ng-eNB is a 4G base station accessing the 5G core network.

[0100] The core network includes an access and mobility management function (AMF) network element and a location management function (LMF) network element, and the like.

[0101] The user equipment (UE) and the gNB communicate through the NR-Uu interface, and the UE and the ng-eNB communicate through the LTE-Uu interface. Both the NR-Uu interface and the LTE-Uu interface can be used to transmit positioning-related signaling. The capital letter U represents the user to network interface (UNI), and the lowercase letter u represents universal.

[0102] Among them, LTE-Uu and NR-Uu use non-access layer (non-access stratum, NAS) or radio resource control (Radio Resource Control, RRC) protocol for transmission. The gNB and AMF, ng-eNB and AMF communicate through the next generation control plane (next generation core network, NG-C) interface, which can be used to transmit positioning-related signaling.

[0103] The NG-C uses the NG application protocol (NG application protocol, NGAP). The AMF network element and the LMF network element communicate through the NL1 interface, which can be used to transmit positioning-related signaling. N in NL1 represents the AMF network element, and L represents the location service device, that is, the LMF network element.

[0104] All or part of the functions implemented by one or more of the UE, base station, or core network can be virtualized, that is, implemented by a special processor or a general processor and a corresponding software module. The UE and the base station involve air interface transmission interfaces, and the transceiver function of the interface can be implemented by hardware. The core network, such as the aforementioned AMF network element or LMF network element, can be virtualized. Optionally, the virtualized UE, base station or core network function can be implemented by a cloud device, such as a cloud device in an over the top (over the top, OTT) system.

[0105] Among them, the UE and the LMF can communicate based on the LTE positioning protocol (LTE positioning protocol, LPP). The LPP protocol specifies the process of interacting information between the UE and the LMF. In actual deployment, the UE and the LMF are not directly connected, but are connected through the UE-base station-AMF-LMF. The LPP message can be transparently transmitted across the base station and the AMF, and the interaction between the UE and the LMF is realized. Alternatively, the UE and the LMF can interact information through the forwarding of the base station and the AMF, without limitation.

[0106] Among them, the interaction between the base station and the LMF can be based on the NR positioning protocol A (NR positioning protocol A, NRPPa). The NRPPa protocol specifies the process of interacting information between the base station and the LMF. In actual deployment, the base station and the LMF are connected through the AMF. The NRPPa protocol is transparent to the AMF, and the NRPPa data unit is transparently transmitted across the AMF, realizing the interaction between the LMF and the base station.

[0107] Currently, in order to adapt to more general scenarios, an AI model is generally not deployed on the LMF network element side for positioning, but a plurality of different AI models need to be generated to form an AI model library to adapt to different environments, and each AI model corresponds to a different combination of environmental parameters.

[0108] The UE can determine the current combination of environmental parameters in which it is located, and select the AI model that best matches the current actual combination of environmental parameters from the AI model library of the LMF network element. The UE takes the positioning reference signal (PRS) sent by one or more base stations as the input of the AI model, and the output of the AI model is the position information of the UE.

[0109] In related solutions, the matching standard for selecting the best matching AI model from the AI model library of the LMF is that the actual combination of environmental parameters during position prediction and the combination of environmental parameters corresponding to the AI model are completely consistent. The reason for this is that when the training set of the AI model is generated and the model is trained, it is for a specific combination of environmental parameters, so when the AI model is used for position prediction, it should also be applied to the same combination of environmental parameters, so that the position information of the UE output by the AI model is accurate and the position error is reduced.

[0110] In theory, if the model training is sufficient, the number of AI models in the AI model library is sufficient, and it can be considered that the optimal AI model can be matched for all combinations of environmental parameters. However, in actual application, the number of AI models obtained by training is generally limited, and the real environment is very complex, and there may not be a corresponding AI model for all combinations of environmental parameters, resulting in that the applicable AI model cannot be matched according to the actual combination of environmental parameters of the UE and positioning cannot be completed.

[0111] To solve the above technical problems, the present application provides a positioning method based on AI model, device and storage medium, which can use a plurality of AI models with high matching degree with the current environmental parameters of the UE for positioning, and fuse the positioning results according to the weight, and then obtain a more accurate final positioning result. It is especially suitable for service areas with complex and variable real environments, or areas where the AI model for positioning is not fully trained, and has high practicality.

[0112] The technical solutions provided by the present application will be described in detail below with reference to the accompanying drawings.

[0113] Referring to FIG. 3, it is a flowchart of a positioning method based on AI model provided by an embodiment of the present application.

[0114] In the following description, the terminal device is taken as UE, and the AI model is deployed in the LMF network element as an example, and the method includes the following steps:

[0115] S21: The UE receives a positioning reference signal.

[0116] The positioning reference signal (PRS) received by the UE can be sent by multiple base stations. For the convenience of illustration, the base station is taken as an example of gNB in the embodiments of the present application, but this does not constitute a limitation on the type of base station.

[0117] The PRS sent by each base station can carry one or more of the following: cell ID (Cell ID), reference signal received power (RSRP), and angle of arrival (AoA) relative to the base station.

[0118] The Cell ID is a unique identifier assigned to each cell. In one possible implementation, the PRS sent by multiple base stations can carry the Cell ID of the serving cell and the neighboring cells.

[0119] In some embodiments, the UE side can not receive the Cell ID of the neighboring cells, in which case the physical cell identifier (PCI) of the neighboring cells can be used to distinguish the neighboring cells.

[0120] After receiving the PRS sent by each base station, the UE can aggregate the Cell ID of all cells, the RSRP corresponding to each cell, and the AoA relative to each base station.

[0121] The following is an example.

[0122] Referring to FIG. 4, which is a schematic diagram of a scenario provided by the embodiments of the present application.

[0123] In FIG. 4, a schematic diagram of the division of the service area into regions and the encoding of the spatial regions is shown. The service area can be divided according to the region information and the characteristics of the region, for example, a rectangular region of a certain area can be used as a division reference to divide the service area into multiple grid regions, such as grid region L1, grid region L2, grid region L3, and grid region L4 in the figure. Each grid region has a corresponding spatial region code.

[0124] Each grid region generally includes multiple cells, and the number of cells included in each grid region can be the same or different. In FIG. 4, the number of cells included in each grid region is taken as an example for illustration. The box labeled 50 in FIG. 4 represents a UE, i.e., UE 50. Each black and white circle in FIG. 4 represents a cell, i.e., the circle labeled 40 represents a cell, and each cell corresponds to a Cell ID.

[0125] After the UE 50 receives the PRS sent by each base station, the Cell IDs of all cells are obtained, and each cell is one of the nine cells in the dashed box in the figure. As can be seen, because the UE is at the boundary position of different grid areas, the ranges of the nine cells are also distributed in different grid areas.

[0126] S22: The UE obtains the environmental parameter combination of the current environment.

[0127] In order to improve the positioning accuracy of the AI model, it is necessary to make the environment of the UE consistent when the model is trained and applied to positioning, that is, to make the PRS used in model training and the PRS obtained in positioning obtained in the same environment, so as to ensure the accuracy of positioning.

[0128] The attribute characteristics of the environment in which the UE is located can be represented by the environmental parameter combination.

[0129] The environmental parameter combination includes a state parameter of a terminal device and a state parameter of a network device.

[0130] The state parameter of the terminal device can be used to represent the state parameter of the environment in which the UE is currently located, which can include but is not limited to a space state parameter, a time state parameter, and a scene state parameter.

[0131] The space state parameter can be space region coding, which can refer to dividing continuous space into discrete grid regions for discretization, and coding the discretized regions. As shown in FIG. 4, the grid regions L1 to L4 can have corresponding space region codes.

[0132] In actual application, different regions have different space region codes, and different AI models are generally used for positioning. That is, the AI model uses a training data set including data in a specific region during training, so the AI model generally specifically processes the positioning task in the region.

[0133] In one possible implementation, the service area can be divided into discrete grid regions by using Geohash geographic coding or Google S2 coding, and the space region codes corresponding to the grid regions are determined.

[0134] The spatial state coding of the grid area where the UE is located can reflect the approximate physical location range where the UE is currently located, and the spatial state coding can be determined by the Cell ID carried in the PRS received by the UE. For example, in FIG. 4, the Cell IDs carried in the respective PRSs received by the UE correspond to nine cells, and it is determined that the spatial area coding given corresponds to the grid area L1-L4, which indicates that the UE is currently located in the area covered by the grid area L1-L4.

[0135] In a possible implementation, the UE can embody the correspondence between the Cell ID list and the spatial area coding. In the correspondence, the Cell IDs of all cells included in the grid area corresponding to each spatial area coding are included. For example, refer to the example shown in Table 1 below.

[0136] Table 1: Correspondence table between Cell ID list and spatial area coding

[0137] The above correspondence can be stored in the form of a data table on the UE, or can be obtained in real time through the network when the UE performs positioning. It can be understood that the above spatial area coding and the data structure of the Cell ID are only for convenient explanation, and do not constitute a limitation on the technical solutions of the present application. In actual application, other data structures conforming to the requirements of related standards can be used. In actual application, the Cell IDs in the grid area corresponding to different spatial area codings can overlap, that is, the same cell can exist in different spatial areas, because at the boundary position of different spatial areas, all cells that can be measured by the positioning assistance data of a certain spatial area can include the cells of the adjacent spatial area. In Table 1, for the convenience of explanation, only the cells of different Cell ID lists are directly obtained according to the cell distribution shown in FIG. 4. If the cell overlap scenario is considered, the number of Cell IDs included in each spatial area coding can increase. For example, when the cell overlap scenario is considered, the Cell ID list corresponding to the spatial area coding L1 can include Cell1-Cell18; and the cells Cell19, Cell125, and Cell131 in L2 that are relatively adjacent to L1, that is, the three cells in the leftmost column of L2; and the cells Cell37-Cell42 in L3 that are relatively adjacent to L1, that is, the six cells in the topmost row of L3; and the cell Cel55 in L4 that is relatively adjacent to L1, that is, the cell in the top-left corner of L4.

[0138] For the convenience of explanation, the Cell ID list in Table 1 is still taken as an example in the following description. It can be understood that when the cell overlap scenario is considered, the cells included in the Cell ID list and the specific number of cells are changed. However, the principle of determining the first screening condition according to the Cell ID list is similar.

[0139] The time state parameter is parameter information obtained by discretizing a continuous time variable, and represents time information when the UE is positioned.

[0140] In a positioning scenario, considering the generality of the AI model, a corresponding time state parameter is generally set for each time period in a day, for example, a day is discretized by hours into 24 specific time periods, namely [0:00, 1:00), [1:00, 2:00), …, [23:00, 24:00). The signal characteristics received by the UE in each time period can be different, which can affect the positioning accuracy, and therefore a corresponding AI model exists for each hour time period.

[0141] Different time periods correspond to different time state parameters, for example, when the time when the UE is positioned is in [0:00, 1:00), the corresponding time state parameter is T1, or is directly represented by the time interval [0:00, 1:00); when the time when the UE is positioned is in [23:00, 24:00), the corresponding time state parameter is T24, or is directly represented by the time interval [23:00, 24:00).

[0142] In a possible implementation, the positioning time when the UE is positioned can be determined by the network, and then the time period in which the positioning time is located is determined, and then the time state parameter is obtained.

[0143] It can be understood that the above division manner is only an example and does not constitute a limitation on the technical solutions of the present application, for example, a day can also be divided into 12 specific time periods, each including two hours; or a day can be divided into 48 specific time periods, each including 0.5 hours, etc.

[0144] In another possible implementation, the plurality of time periods divided can also have an intersection, for example, the time periods in a day can be [0:00, 2:00), [1:00, 3:00), [2:00, 4:00), …, [23:00, 24:00), etc.

[0145] The scene state parameter represents the scene in which the UE is located when positioned, different scenes have different scene characteristics, which can affect the accuracy of the positioning result of the AI model, and therefore generally an AI model suitable for different scenes needs to be trained for different scenes. The scene can include but is not limited to a high-speed rail scene, a subway scene, an indoor scene, an underground garage scene, a general outdoor scene, etc.

[0146] For example, for a high-speed rail scenario, when the UE is located on a high-speed rail, the UE has the characteristics of fast moving speed, large signal penetration loss, and large Doppler shift of the signal. For an underground garage scenario, when the UE is located in an underground garage, the signal strength is low, and signal interruption may occur, causing data update delay. Therefore, the AI model used for positioning in the high-speed rail scenario and the model parameters and model types of the AI model suitable for positioning in the underground garage scenario may be different.

[0147] Each scenario is pre-configured with a corresponding identifier, for example, the high-speed rail scenario corresponds to identifier L1, and the underground garage scenario corresponds to identifier L2. The identifier can be used as a scenario state parameter.

[0148] The UE can detect the scenario through a sensor equipped by the UE.

[0149] For example, the accelerometer equipped by the UE can detect the acceleration of the UE in different directions, and the gyroscope can be used to determine the motion posture of the UE. The UE can determine, according to the detection results of the accelerometer and the gyroscope, that the UE has accelerated and decelerated motion multiple times in a short time interval, for example, multiple times in tens of minutes or a few minutes. When the altitude of the UE is determined to be lower than the altitude of the grid area corresponding to the space area code according to the air pressure detected by the air pressure sensor, it can be determined that the current scenario is a subway scenario.

[0150] For another example, when the UE determines, according to the air pressure detected by the air pressure sensor, that the altitude of the UE is lower than the altitude of the grid area corresponding to the space area code, and the UE determines, according to the detection results of the accelerometer and the gyroscope, that the speed variation of the UE is small, it can be determined that the current scenario is an underground garage scenario (similar to a basement scenario).

[0151] For another example, when the UE determines, according to the detection results of the accelerometer and the gyroscope, that the acceleration and speed of the UE meet the characteristics of the high-speed rail scenario, it can be determined that the current scenario is a high-speed rail scenario.

[0152] For another example, the UE can determine whether it is in an indoor scenario or an outdoor scenario according to whether it can currently initiate global navigation satellite system (GNSS) positioning. If it can initiate GNSS positioning, it is considered to be in an outdoor scenario, otherwise it is considered to be in an indoor scenario.

[0153] In addition, the UE can also determine the scenario according to the attributes of the currently connected base station device. For example, when the currently connected base station of the UE is identified as a high-speed rail dedicated base station, it can be determined that the current scenario is a high-speed rail scenario. For another example, when the currently connected base station of the UE is identified as a special base station of a large shopping mall, it can be determined that the current scenario is an indoor scenario.

[0154] The network device state parameter is generally related to the configuration parameter of the network device, and is mainly related to the network optimization measure of the operator of the network device. For the same network device, the operator can set different configuration parameters on different dates or in different time periods of the same day. The configuration parameter is specifically related to the load condition of the network device at different times, the maintenance ability of the operator, the business behavior of the operator, and the like.

[0155] For example, for the same network device, the configuration parameters of different times such as weekdays (for example, Monday to Friday), ordinary holidays (for example, Saturday and Sunday), and other statutory holidays are different, resulting in differences in information processing delay, network quality, and the like of the network device, which will affect the positioning result.

[0156] Therefore, different AI models are generally trained for different network device state parameters, and when positioning, an AI model matched with the current network device state parameter is selected for use, thereby improving the accuracy of the positioning result.

[0157] The parameters of the network device can be identified by anonymization and discretization.

[0158] For example, for time-related state parameters, continuous time can be discretized into different time periods, such as weekday time periods, ordinary holiday time periods, and other statutory holiday time periods. When the positioning time of the UE is in the discretized weekday time period, it can be determined that the current time-related state parameter is the state parameter corresponding to the weekday time period.

[0159] S23: The UE determines one or more AI model indexes matched by using the combination of the received positioning reference signal PRS and the environmental parameter.

[0160] Compared with the existing related scheme, in the scheme of the present application, multiple available AI models can be determined.

[0161] It can be understood that the plurality of AI models determined by the present application, the combination of environmental parameters of the training data used in training, and the screening conditions when determining the AI model index are not completely consistent. For example, both the time screening condition and the space screening condition can obtain a plurality of matched AI models. This is because the present application scheme takes into account that due to the complexity of the environment in actual application, the model training is not sufficient, but in some special cases, the UE cannot match a unique optimal model according to the current combination of environmental parameters, but can only match a few relatively optimal AI models. The combination of environmental parameters corresponding to the training data of these AI models is more or less inconsistent with the real combination of environmental parameters in positioning. At this time, instead of directly determining that the positioning cannot be completed, the present application scheme considers that even if the most reliable positioning result cannot be obtained at this time, the positioning results of these nearly optimal AI models can be fused to obtain a relatively accurate final positioning result.

[0162] Referring to FIG. 5, this figure is a schematic diagram of a framework for positioning based on AI models provided by an embodiment of the present application.

[0163] The UE aggregates the information of the received PRS, and determines one or more AI model indexes corresponding to the AI models currently applicable in combination with the acquired combination of environmental parameters.

[0164] In the embodiment of the present application, the determination of part of the environmental parameters in the combination of environmental parameters by using sensor information is taken as an example. In actual application, the environmental parameters in the combination of environmental parameters can also be obtained by user active configuration, or through network equipment, Internet of Things, Internet, etc. The embodiment of the present application does not make specific limitation.

[0165] In selecting the AI model currently applicable, the main basis is to match the AI model according to the combination of environmental parameters.

[0166] Taking the combination of environmental parameters including the state parameters of the terminal device and the state parameters of the network equipment, and the state parameters of the terminal device including the spatial region code (i.e. spatial state parameter), time state parameter and scene state parameter as an example.

[0167] The AI model index is a data structure used for fast query and AI model, which can be understood as the identification of the AI model.

[0168] Each AI model has a unique corresponding AI model index. The specific data structure of the AI model index is not limited in the embodiment of the present application. The following description is only for example.

[0169] Referring to Table 2 below, this table is a correspondence table of each combination of environmental parameters and AI model index.

[0170] Table 2: Correspondence table of each combination of environmental parameters and AI model index

[0171] The correspondence in Table 2 can also be referred to as a mapping table. It can be understood that the data structure of each data in Table 2 above is only for the convenience of description, and does not constitute a limitation on the technical solutions of the present application. In actual application, other data structures conforming to the requirements of relevant standards can be used.

[0172] The UE can first aggregate all the PRSs received, and then obtain all the Cell IDs, see Table 1 and FIG. 4 for details. At this time, the determined Cell IDs are Cell11, Cell12, Cell25, Cell17, Cell18, Cell31, Cell41, Cell42 and Cell55 in turn.

[0173] All the Cell IDs indicated in the PRSs can form a first Cell ID list. That is, the first Cell ID list is the union set of the Cell IDs indicated in all the PRSs.

[0174] The Cell ID list corresponding to the spatial region code can be referred to as a second Cell ID list.

[0175] The intersection of the first Cell ID list and the second Cell ID list corresponding to each spatial region code is determined.

[0176] The intersection of the two lists means that there are Cell IDs that appear in both lists. For example, continuing to refer to Table 1, the intersection of the first Cell ID list and the second Cell ID list corresponding to the spatial region code L1 is Cell11, Cell12, Cell17 and Cell18; the intersection of the first Cell ID list and the second Cell ID list corresponding to the spatial region code L2 is Cell25 and Cell31; the intersection of the first Cell ID list and the second Cell ID list corresponding to the spatial region code L3 is Cell41 and Cell42; and the intersection of the first Cell ID list and the second Cell ID list corresponding to the spatial region code L4 is Cell55.

[0177] It should be noted that in the above intersection process, if the scenario of cell overlap is considered, the number of Cell IDs included in the second Cell ID list corresponding to each spatial region code in Table 1 can increase. However, the principle of determining the intersection of the first Cell ID list and the second Cell ID list corresponding to each spatial region code is similar, and will not be described here.

[0178] In a possible implementation, the UE can reduce the range of the grid area or the number of grid areas participating in matching according to the service area information (for example, city information) to reduce the amount of calculation, for example, the UE can determine the location according to the current base station identifier and the Cell ID, and reduce the number of grid areas participating in matching in combination with the area information, as shown in FIG. 4, only the grid areas corresponding to the space area codes L1-L4 are matched, and other grid areas such as L5 do not need to be matched. In addition, for the grid areas participating in matching, the range can be reduced, for example, for the grid area L1, when the first Cell ID list of the UE includes Cell 12, because of the limited communication capability of the UE and the base station, the left 9 cells corresponding to the grid area of the space area code L1 can not participate in the calculation, because the distance between the 9 cells and Cell 12 is too large, and there is no intersection with the first Cell ID list.

[0179] When the number of Cell IDs included in the intersection is greater than or equal to the first threshold, the UE considers that the corresponding grid area satisfies the screening condition of the space state parameter, that is, the first screening condition.

[0180] The purpose of determining the first screening condition is to determine in which grid area the UE is located. The traditional non-AI Time Difference of Arrival (TDOA) positioning scheme needs to know the position coordinates of the base station or antenna, but in the AI-based positioning scheme, it is not necessary, and the AI model can realize positioning based on the signal fingerprint information received by the UE.

[0181] Specifically, when the number of Cell IDs included in the intersection is greater than or equal to the first threshold, the UE takes the space area code of the second Cell ID list corresponding to the intersection as the space screening condition of the AI model index.

[0182] The present embodiment does not specifically limit the first threshold, and the following description takes 3 as an example.

[0183] Continuing to refer to the above example, among the grid areas corresponding to the space area codes L1-L4, only the number of Cell IDs in the intersection corresponding to L1 is greater than 3, so the space area code L1 is taken as the space screening condition of the AI model index. At this time, P1-P8 in Table 2 satisfy the space screening condition in each AI model index.

[0184] Further, the screening process of the space area code can also refer to the number of Cell ID intersections and other information in the PRS.

[0185] For example, RSRP is further added for joint judgment. For each Cell ID in the intersection, it is further determined whether the RSRP corresponding to each Cell ID is greater than or equal to a preset power threshold, and whether the number of Cell IDs whose RSRP is greater than or equal to the preset power threshold is greater than or equal to a first threshold. If yes, the spatial region code of the second Cell ID list corresponding to the intersection is taken as the spatial screening condition of the AI model index. Similarly, AoA can also be further added for joint judgment.

[0186] For example, RSRP and AoA are further added for joint judgment. For each Cell ID in the intersection, it is further determined whether the RSRP corresponding to each Cell ID is greater than or equal to a preset power threshold, and whether the AoA corresponding to each Cell ID satisfies a preset angle range, and then it is determined whether the number of Cell IDs that satisfy the above conditions at the same time is greater than or equal to a first threshold. If yes, the spatial region code of the second Cell ID list corresponding to the intersection is taken as the spatial screening condition of the AI model index.

[0187] For the time state parameter, since it has been discretized into time intervals, the UE can compare the positioning time with each time interval to determine the time interval in which the positioning time is located, or the time interval closest to the positioning time. The time interval determined by the time parameter can be one or more. The following will be further described with reference to Table 2.

[0188] In one example, the time interval corresponding to the time parameter T1 in Table 2 is [7:00, 8:00], and the time interval corresponding to the time parameter T2 is [9:00, 10:00]. When the positioning time of the UE is 7:50, since it is in [7:00, 8:00], the time parameter T1 is taken as the time screening condition. At this time, among P1-P8 which satisfy the spatial screening condition, the AI model indexes that also satisfy the time screening condition are P1-P4.

[0189] In another example, the time interval corresponding to the time parameter T1 in Table 2 is [7:00, 8:00], and the time interval corresponding to the time parameter T2 is [9:00, 10:00]. When the positioning time of the UE is 6:50, since it is not in any of the above time periods, the time parameter of the time period closest to the positioning time is determined as the time screening condition, that is, the time parameter T1 is taken as the time screening condition.

[0190] In another example, the time interval corresponding to the time parameter T1 in Table 2 is [7:00, 8:00], and the time interval corresponding to the time parameter T2 is [7:30, 8:30]. When the positioning time of the UE is 7:50, the UE is simultaneously in the above two time intervals, and therefore the time parameters T1 and T2 are both used as the time screening condition, and P1-P8 simultaneously satisfy the time screening condition and the space screening condition.

[0191] In another example, the basis for determining the time screening condition is whether the time interval between the positioning time and the reference value corresponding to the time interval is greater than a preset interval. If it is less than, the time state parameter corresponding to the time interval is used as the time screening condition, and the time screening condition is the second screening condition. The determined time interval can be one or more.

[0192] The reference value is the middle value of the time interval, and the preset interval can be set according to the actual AI model corresponding to the geographical area, and the embodiments of the present application are not limited, and examples are given below.

[0193] For example, the time interval corresponding to the time parameter T1 in Table 2 is [8:00, 9:00], and the reference value corresponding to [8:00, 9:00] is 8:30. The time interval corresponding to the time parameter T2 is [9:00, 10:00], and the reference value corresponding to [9:00, 10:00] is 9:30. The preset interval can be set to 1 hour. When the positioning time of the UE is 9:20, the time interval between the positioning time and 8:30 is 50 minutes, which is less than the preset interval; the time interval between the positioning time and 9:30 is 10 minutes, which is less than the preset interval, and therefore the time parameter T1 and the time parameter T2 can be used as the time screening condition.

[0194] The scene state parameter can identify the scene in which the UE is located when positioning. After the UE determines the scene in which it is located, the scene state parameter corresponding to the scene can be used as the scene screening condition, and the scene screening condition is the third screening condition, which is illustrated by examples below.

[0195] Referring to Table 2, the scene corresponding to the scene state parameter S1 is a high-speed rail scene, and the scene corresponding to the scene state parameter S1 is a subway scene. When the UE determines that the scene in which it is located is a high-speed rail scene, the scene state parameter S1 is determined as the time screening condition. At this time, among P1-P8 that satisfy the space screening condition, the AI model indexes that also satisfy the scene screening condition are P1, P2, P5, and P6.

[0196] For the network device state parameter, the network device state parameter can be quantified into discrete intervals for matching. Taking the time-related network device state parameter as an example, the network device screening condition can be determined in a similar manner to the time state parameter, which is illustrated by examples below.

[0197] Continuing to refer to Table 2 above, the network device state parameter is Q1, indicating that the current configuration parameter of the network device is the working day configuration parameter, and the network device state parameter is Q2, indicating that the current configuration parameter of the network device is the holiday configuration parameter. The UE can determine that the current time is in a working day according to the positioning time, and the network device state parameter Q1 is taken as the network device screening condition. At this time, among P1, P2, P5, P6 that simultaneously satisfy the spatial screening condition (L1), the time screening condition (T1, T2), and the scene screening condition (S1), the AI model indexes that also satisfy the network device screening condition are P1 and P5.

[0198] The network device state parameter related to the space can determine the network device screening condition in a similar manner to the space area coding. The network device screening condition is the fourth screening condition, which will not be described again here.

[0199] In the above manner, one or more AI model indexes that match are determined.

[0200] S24: The UE sends first index information to the LMF network element.

[0201] The first index information includes one or more AI model indexes.

[0202] In a possible implementation, the UE and the LMF network element can communicate based on the LTE positioning protocol (LPP). The first index information can be carried in LPP-related signaling.

[0203] S25: The LMF network element determines an AI model corresponding to each AI model index in the one or more AI model indexes.

[0204] Continuing to refer to FIG. 5, the LMF network element side is deployed with an AI model library for UE positioning. After the LMF network element obtains one or more AI model indexes, the AI model corresponding to each AI model index is matched from the AI model library.

[0205] S26: The LMF network element sends all determined AI models to the UE.

[0206] In a possible implementation, the UE and the LMF can communicate based on the LPP, and the related data of the AI model can be carried in LPP-related signaling.

[0207] S27: The UE determines the weight of the positioning result output by each AI model by using the environmental parameter combination.

[0208] The application scheme needs to determine the weight of each AI model when fusing the positioning result, and the weight is the weight of the positioning result output by the AI model. After determining the weight corresponding to each AI model, the positioning result of each AI model is multiplied by the corresponding weight, and then the obtained calculation result is accumulated to determine the final positioning result.

[0209] It can be understood that if only one AI model index is determined in S23, the weight of the AI model can be determined as 1 when the LMF network element sends the corresponding AI model to the UE, that is, the positioning result of the AI model can be directly used as the final positioning result. The following describes the case when multiple AI model indexes are determined in S23.

[0210] For each AI model, the weight coefficient K of the corresponding AI model can be determined by the following formula:

[0211] Wherein, n is the number of each environmental parameter in the environmental parameter combination, n is a positive integer; K i is the weight coefficient corresponding to the ith environmental parameter; p i is the influence factor of the ith environmental parameter, and the influence factor is related to the attribute of the environmental parameter, for example, the first influence factor related to space is greater than the second influence factor related to time, and the influence factor is used to calibrate and adjust the weight coefficient corresponding to each environmental parameter. In formula (1), each environmental parameter is weighted and summed with the corresponding influence factor to obtain the weight coefficient K of each AI model.

[0212] The greater the weight coefficient, the higher the credibility of the AI model selected by the environmental parameter corresponding to the weight coefficient, and the closer the positioning result of the AI model to the real position.

[0213] In the embodiment of the application, taking the environmental parameters including space state parameters, time state parameters, scene state parameters and network device state parameters as an example, n is 4.

[0214] For the space state parameter, that is, the space region code, when selecting the model, the more the number of Cell IDs in the intersection, the higher the weight coefficient of the corresponding space state parameter, and the weight of the space state parameter is the first weight coefficient. The first weight coefficient is positively correlated with the number of Cell IDs in the intersection.

[0215] In one possible implementation, the weight coefficient is the ratio of the number of Cell IDs in the intersection corresponding to the space region code to the number of Cell IDs included in the first Cell ID list, which is illustrated by the following example.

[0216] For example, all the Cell IDs indicated in the PRS can form a first Cell ID list including 10 Cell IDs. The intersection of the second Cell ID list corresponding to the spatial area code L1 and the first Cell ID list includes 5 Cell IDs, and the weight coefficient corresponding to the spatial area code L1 is 0.5; the intersection of the second Cell ID list corresponding to the spatial area code L2 and the first Cell ID list includes 4 Cell IDs, and the weight coefficient corresponding to the spatial area code L1 is 0.4.

[0217] In another possible implementation, the weight coefficient is the ratio between the number of Cell IDs in the intersection corresponding to the spatial area code and the number of Cell IDs in the intersection corresponding to all the AI model indexes matched in S23. The following is an example.

[0218] For example, the AI model indexes matched in S23 are P1 and P9 in Table 2, wherein the spatial area code of P1 is L1, and the intersection of the second Cell ID list corresponding to L1 and the first Cell ID list includes 5 Cell IDs. The spatial area code of P9 is L2, and the intersection of the second Cell ID list corresponding to L2 and the first Cell ID list includes 3 Cell IDs. Then the weight coefficient corresponding to L1 is 5 / (5+3) = 0.625, and the weight coefficient corresponding to L2 is 3 / (5+3) = 0.375.

[0219] The size of the second weight coefficient of the time state parameter is negatively related to the time interval between the reference value of the time interval and the positioning time. The larger the time interval, the smaller the second weight coefficient. The following is an example.

[0220] For example, the time interval corresponding to the time parameter T1 in Table 2 is [8:00, 9:00], and the reference value corresponding to [8:00, 9:00] is 8:30. The time interval corresponding to the time parameter T2 is [9:00, 10:00], and the reference value corresponding to [9:00, 10:00] is 9:30. When the positioning time of the UE is 9:20, the time interval between the positioning time and 8:30 is 50 minutes, and the corresponding second weight coefficient is K 21 ; the time interval between the positioning time and 9:30 is 10 minutes, and the corresponding second weight coefficient is K 22 . Then K 22 is greater than K 21 , indicating that the prediction result of the AI model obtained by taking T2 as the time screening condition has a higher credibility.

[0221] In a possible implementation, the weight coefficient K2 of the time state parameter can be determined by the following formula: K2 = 1-Ts / T0 (2)

[0222] wherein Ts is a time interval between the reference value of the time interval and the positioning time, and T0 is a preset interval.

[0223] For example, taking the time interval corresponding to the time parameter T1 in Table 2 as [8:00, 9:00], the reference value is 8:30, the positioning time is 9:20, and the preset interval is one hour, the weight coefficient corresponding to T1 is K2 = 0.167.

[0224] For example, taking the time interval corresponding to the time parameter T2 in Table 2 as [9:00, 10:00], the reference value is 9:30, the positioning time is 9:20, and the preset interval is one hour, the weight coefficient corresponding to T1 is K2 = 0.833.

[0225] The above is only an example, and the weight of the time state parameter can be determined in other ways, which will not be described here.

[0226] For the scene state parameter, the third weight coefficient corresponding to each scene can be calibrated by means of big data and AI model analysis. For example, the third weight coefficient corresponding to the high-speed rail scene is 0.2, and the weight coefficient corresponding to the indoor scene is 0.5, and the application embodiment is not limited, the correspondence between the scene and the weight coefficient can be determined in advance and stored locally in the UE, or obtained by the UE through the network when positioning.

[0227] For the network device state parameter, the weight coefficient corresponding to each network device state parameter can also be determined in a similar manner. For example, for the time-related network device state parameter, a similar manner to the time state parameter described above can be adopted, and for the space-related network device state parameter, a similar manner to the space state parameter described above can be adopted, which will not be described here.

[0228] After determining the weight coefficient K of each model by formula (1), the weight coefficients of each AI model are normalized to obtain the weight of the AI model distributed in [0, 1]. After normalization, the weight coefficient K of each AI model is converted into the weight k, and the weight coefficients of all AI models add up to 1.

[0229] The following is an example. Continue to refer to Table 2, taking AI model indexes P1 and P15 as examples.

[0230] For P1, the weight coefficient K1 corresponding to the spatial region encoding is 0.5, the weight coefficient K2 corresponding to the time state parameter is 0.4, the weight coefficient K3 corresponding to the scene state parameter is 0.8, and the weight coefficient K4 corresponding to the network device state parameter is 0.5. Among these, K1 and K3 are spatially related environmental parameters with relatively large influence factors; for example, the first influence factor is set to 1.1. K2 and K4 are time-related environmental parameters with relatively small influence factors; for example, the second influence factor is set to 0.8. Therefore, the weight coefficient K corresponding to the AI ​​model index P1 is determined according to equation (1). P1 =1.1*0.5+0.8*0.4+1.1*0.8+0.8*0.5=2.15.

[0231] For P15, the weight coefficient K1 corresponding to the spatial region encoding is 0.4, the weight coefficient K2 corresponding to the temporal state parameter is 0.8, the weight coefficient K3 corresponding to the scene state parameter is 0.6, and the weight coefficient K4 corresponding to the network device state parameter is 0.5. The weight coefficient K corresponding to the AI ​​model index P15 is determined according to equation (1). P15 =1.1*0.4+0.8*0.4+1.1*0.6+0.8*0.5=1.82.

[0232] For K P1 and K P15 Normalization is performed to obtain the weight k corresponding to P1. P1 =(K P1 ) / (K P1 +K P15 The weight k corresponding to P2 is approximately 0.542. P15 =(K P15 ) / (K P1 +K P15 ), approximately 0.458.

[0233] S28: The UE uses the PRS, each AI model, and the weights corresponding to each AI model to determine the final localization result.

[0234] The UE uses the PRS as input to each AI model to obtain the positioning results output by each AI model. The positioning results output by the AI ​​model can be the absolute position or the relative position of the UE, and this application embodiment does not specifically limit this.

[0235] Understandably, the localization results output by each AI model are generally of the same type, such as both absolute position and both relative position.

[0236] The absolute position can be a latitude and longitude coordinate. The relative position can be a position coordinate of the UE relative to a serving base station and / or a base station corresponding to a neighbor cell. The relative position can also be represented in other manners, and embodiments of the present application do not limit the manner of representing the relative position.

[0237] The final positioning result position can be determined by the following formula:

[0238] where j is the number of AI models, and j is a positive integer. k j is a weight of the jth AI model. Pos j is a positioning result output by the jth AI model. That is, the final positioning result component corresponding to each AI model is obtained by multiplying the positioning result output by each AI model by the weight corresponding to each AI model, and the final positioning result is obtained by superimposing the final positioning result components corresponding to each AI model.

[0239] The following describes an example in which the positioning result is an absolute position and the number of AI models is 2.

[0240] In one example, the positioning result of the AI model P1 is a latitude and longitude coordinate (x1, y1), and the weight k1 of P1 is 0.542. The positioning result of the AI model P15 is a latitude and longitude coordinate (x2, y2), and the weight k1 of P15 is 0.458. The final positioning result can be a latitude and longitude coordinate (0.542x1+0.458x2, 0.542y1+0.458y2).

[0241] In embodiments of the present application, the UE can take the weight of the AI model with the highest weight as the confidence of the final positioning result this time. The higher the weight, the higher the confidence. For example, when only one AI model that best matches is determined, the weight of the AI model is 1, and the corresponding confidence is 1, indicating that the positioning result this time is predicted by an AI model that completely matches the current environmental parameter combination, and the result is accurate and reliable.

[0242] It can be understood that the above steps are only divided for convenience of description, and do not constitute a limitation on the technical solutions of the present application. In actual applications, the order of the above steps can be changed, for example, S27 can be moved forward to before S24, or S27 can be performed in parallel with S24.

[0243] In summary, the positioning method provided in the embodiments of the present application can determine one or more AI model indexes corresponding to the AI models for positioning by the UE, and then send all the AI model indexes to the LMF network element. The LMF network element matches the corresponding AI models from the AI model library according to the AI model indexes and sends them to the UE. The UE can determine the weights corresponding to the AI models, and determine the final positioning result by using the AI models and the corresponding weights. In the prior art, when an AI model that completely matches the current environmental parameter combination of the UE cannot be determined, it is directly considered that the positioning cannot be completed. However, the scheme provided in the embodiments of the present application can use multiple AI models that have a high matching degree with the current environmental parameters of the UE for positioning, and fuse the positioning results according to the weights to obtain a more accurate final positioning result. This scheme is especially suitable for service areas with complex and changeable real environments, or areas where the AI models for positioning are not fully trained, and has high practicability. At the same time, after the LMF network element sends the AI models to the UE, for the scenario where the UE needs to continuously position, the UE can directly use the received AI models for positioning, and if the environmental parameter combination does not change, the UE no longer needs to frequently interact with the LMF network element, thereby reducing the time delay of positioning.

[0244] In the above embodiments, the LMF network element sends the determined AI models to the UE, and the UE performs fusion positioning. The following describes an implementation manner in which the LMF network element directly performs fusion positioning according to the determined AI models and sends the final positioning result to the UE.

[0245] Referring to FIG. 6, this figure is a flowchart of another AI model-based positioning method provided in the embodiments of the present application.

[0246] In the following description, the terminal device is taken as the UE, and the AI model is deployed in the LMF network element as an example. The method includes the following steps.

[0247] S41: The UE receives PRS.

[0248] The PRS received by the UE can be sent by multiple base stations. In order to facilitate the description, the base station is taken as gNB in the embodiments of the present application, but this does not constitute a limitation on the type of base station.

[0249] The PRS sent by each base station can carry one or more of the Cell ID, RSRP, and AoA relative to the base station. In a possible implementation manner, the PRS sent by multiple base stations can carry the Cell IDs of the serving cell and the neighboring cells. In some embodiments, the UE side can not receive the Cell IDs of the neighboring cells, and in this case, the PCIs of the neighboring cells can be used to distinguish the neighboring cells.

[0250] S42: The UE acquires an environment parameter combination of the current environment.

[0251] The environment parameter combination includes a state parameter of the terminal device and a state parameter of the network device.

[0252] The state parameter of the terminal device can be used to represent the state parameter of the environment in which the terminal device, i.e., the UE, is currently located, which can include but is not limited to a spatial state parameter, a time state parameter, and a scene state parameter.

[0253] For specific descriptions of the environment parameter combination, refer to the above S22, which will not be repeated here.

[0254] S43: The UE determines one or more AI model indexes that match by using the received positioning reference signal PRS and the environment parameter combination.

[0255] The AI model index is a data structure used for fast query and AI model, which can be understood as an identifier of the AI model. Each AI model has a unique corresponding AI model index, and the specific data structure of the AI model index is not limited in the embodiments of the present application. The UE can pre-store the corresponding relationship between each environment parameter combination and the AI model index in the form of a mapping table, which can be referred to Table 2 above.

[0256] The UE can first summarize all the PRS received, and then obtain all the Cell IDs. All the Cell IDs indicated in the PRS can form a first Cell ID list.

[0257] The Cell ID list corresponding to the spatial region code can be referred to as a second Cell ID list.

[0258] The intersection of the first Cell ID list and the second Cell ID list corresponding to each spatial region code is determined.

[0259] When the number of Cell IDs included in the intersection is greater than or equal to a first threshold, the UE considers that the corresponding grid region satisfies the screening condition of the spatial state parameter, and takes the spatial region code of the second Cell ID list corresponding to the intersection as the spatial screening condition of the AI model index.

[0260] Further, the screening process of the spatial region code can also refer to the number of Cell ID intersection and other information in the PRS at the same time. For example, further adding RSRP and / or AoA for joint judgment, that is, after determining the number of Cell ID in the intersection, further determining whether the RSRP of the base station corresponding to each Cell ID in the intersection satisfies the preset power threshold, and / or whether the AoA of the base station corresponding to each Cell ID in the intersection satisfies the preset angle range. If yes, the spatial region code of the second Cell ID list corresponding to the intersection is taken as the spatial screening condition of the AI model index.

[0261] For the time state parameter, since it has been discretized into time intervals, the UE can compare the positioning time with each time interval, determine the time interval in which the positioning time is located, or the time interval closest to the positioning time. The time interval determined by the time parameter can be one or more.

[0262] In a possible implementation, the basis for determining the time screening condition is whether the time interval between the positioning time and the reference value corresponding to the time interval is greater than the preset interval. If less, the time state parameter corresponding to the time interval is taken as the time screening condition. Wherein, the reference value is the middle value of the time interval, and the preset interval can be set according to the actual AI model corresponding to the geographical area, and the embodiments of the present application are not limited specifically.

[0263] The scene state parameter can identify the scene in which the UE is located when positioning. After the UE determines the scene in which it is located, the scene state parameter corresponding to the scene can be taken as the scene screening condition.

[0264] For the network device state parameter, the network device state parameter can be quantified into discrete intervals for matching. Taking the time-related network device state parameter as an example, the network device screening condition can be determined in a similar manner to the time state parameter; the space-related network device state parameter can be determined in a similar manner to the spatial region code.

[0265] In the above manner, one or more AI model indexes that match are determined.

[0266] S44: The UE sends the first index information, the PRS, and the environment parameter combination to the LMF network element.

[0267] Referring to FIG. 7, which is a schematic diagram of a second framework for positioning based on an AI model according to an embodiment of the present application.

[0268] Figure 7 corresponds to the scheme of Figure 5, the difference between them is that in Figure 7, the UE needs to send PRS and environmental parameter combination to the network device. Among them, the PRS is used as the input of each AI model, and the environmental parameter combination is used to determine the weight of each AI model corresponding to the LMF side.

[0269] In a possible implementation, the communication between the UE and the LMF can be based on LPP, and the first index information, PRS and environmental parameter combination can be carried in the LPP related signaling.

[0270] It can be understood that in another possible implementation, the UE can also directly send the weight of each AI model to the LMF network element after determining the weight of each AI model locally, that is, the first index information, PRS and the weight of each AI model corresponding to the LPP related signaling. In this implementation, the LMF network element can directly use the weight of each AI model corresponding to the UE to perform fusion positioning.

[0271] S45: The LMF network element determines an AI model corresponding to each AI model index in one or more AI model indexes.

[0272] The LMF network element side is deployed with an AI model library for UE positioning. The AI model library includes a plurality of trained AI models, each AI model corresponding to an AI model index.

[0273] After the LMF network element obtains one or more AI model indexes, it matches the AI model corresponding to each AI model index from the AI model library.

[0274] S46: The LMF network element determines the weight of the positioning result output by each AI model using the environmental parameter combination.

[0275] The scheme of the embodiment of the application determines the weight of the positioning result output by each AI model by the LMF network element, not on the UE, which can utilize the computing power of the LMF network element to quickly obtain the weight result and improve the data processing speed.

[0276] It can be understood that if only one AI model index is determined in S45, the weight of the AI model can be directly determined as 1, that is, the positioning result of the AI model can be directly used as the final positioning result. The following introduces the case when multiple AI model indexes are determined in S45.

[0277] For each AI model, the weight coefficient K of the corresponding AI model can be determined by the above formula (1).

[0278] The greater the weight coefficient is, the higher the credibility of the AI model filtered by the environmental parameter corresponding to the weight coefficient is, and the closer the positioning result of the AI model is to the real position.

[0279] In the embodiments of the present application, the environmental parameters include space state parameters, time state parameters, scene state parameters, and network device state parameters, for example, and n is 4.

[0280] For the space state parameters, that is, the space region code, the greater the number of Cell IDs in the intersection is, the higher the weight coefficient of the corresponding space state parameter is when filtering the model.

[0281] In a possible implementation, the weight coefficient is the ratio of the number of Cell IDs in the intersection corresponding to the space region code to the number of Cell IDs included in the first Cell ID list.

[0282] In another possible implementation, the weight coefficient is the ratio between the number of Cell IDs in the intersection corresponding to the space region code and the number of Cell IDs in the intersection corresponding to all AI model indexes matched in S45.

[0283] For the time state parameters, the size of the weight coefficient is negatively related to the time interval between the reference value of the time interval and the positioning time, and the greater the time interval is, the smaller the weight coefficient is.

[0284] In a possible implementation, the weight coefficient K2 of the time state parameter can be determined by the above formula (2).

[0285] For the scene state parameters, the weight coefficients corresponding to various scenes can be calibrated by big data and AI model analysis, and the correspondence between the scene and the weight is saved in the LMF network element.

[0286] For the network device state parameters, the weight coefficients corresponding to various network device state parameters can also be determined in a similar manner. For example, for the time-related network device state parameters, a similar manner to the time state parameters described above can be adopted, and for the space-related network device state parameters, a similar manner to the space state parameters described above can be adopted. The embodiments of the present application will not be repeated here.

[0287] When the weight coefficients K of the models are determined by formula (1), the weight coefficients of the AI models are normalized to obtain the weights of the AI models distributed between [0, 1]. After normalization, the weight coefficients K of the AI models are converted into weights k, and the weight coefficients of all AI models add up to 1.

[0288] S47: The LMF network element determines the final positioning result by using the PRS, the AI models, and the weights corresponding to the AI models.

[0289] The LMF network element takes the PRS sent by the UE as the input of each AI model to obtain the positioning result output by each AI model. The positioning result output by a single AI model can be the absolute position or the relative position of the UE, and the embodiments of the present application are not limited in this regard.

[0290] It can be understood that the types of the positioning results output by each AI model are generally the same, for example, the absolute position or the relative position at the same time.

[0291] The absolute position can be the latitude and longitude coordinates. The relative position can be the position coordinates of the UE relative to the serving base station and / or the base station corresponding to the neighboring cell, and the relative position can also be represented in other ways, and the embodiments of the present application are not limited in this regard.

[0292] The final positioning result position can be determined by the above formula (3), that is, the final positioning result component corresponding to each AI model is obtained by multiplying the positioning result output by each AI model and the weight corresponding to each AI model, and the final positioning result corresponding to each AI model is obtained by superimposing the final positioning result components corresponding to each AI model.

[0293] S48: The LMF network element sends the location information to the UE.

[0294] The location information includes the final positioning result.

[0295] In a possible implementation, the LMF network element can carry the location information in the related signaling of the LTE positioning protocol.

[0296] It can be understood that the division of the above steps is only for the convenience of description, and does not constitute a limitation on the technical solutions of the present application. In actual application, the order of the above steps can be changed, for example, S47 can be moved forward to before S45, or S47 can be performed in parallel with S45.

[0297] In summary, the positioning method provided in the embodiments of the present application can determine one or more AI model indexes corresponding to the AI models for positioning by the UE, and then send all the AI model indexes to the LMF network element. The LMF network element can match the corresponding AI models from the AI model library according to the AI model indexes, and determine the weights corresponding to the AI models. Then, the LMF can determine the final positioning result by using the AI models and the corresponding weights, and send the final positioning result to the UE. In the existing scheme, when an AI model that completely matches the current environmental parameter combination of the UE cannot be determined, it is directly considered that the positioning cannot be completed. However, the scheme of the present application can use multiple AI models that have a high matching degree with the current environmental parameters of the UE to perform positioning, and fuse the positioning results according to the weights to obtain a more accurate final positioning result. The scheme is especially suitable for service areas with complex and changeable real environments, or areas where the AI models for positioning are not fully trained, and has high practicability.

[0298] Meanwhile, for a scenario in which the UE side does not need to perform continuous positioning, such as a scenario in which single positioning or positioning with a small number of times is performed, the UE side can directly obtain the positioning result sent by the LMF network element. In this way, the UE side does not need to receive the AI model sent by the LMF network element, which can significantly reduce the amount of data received and also reduce the time delay of single positioning.

[0299] In the above description, the UE side stores the mapping table and determines the AI model index as an example. The following describes an implementation manner in which the LMF network element determines the AI model index.

[0300] Referring to FIG. 8, this figure is a flowchart of another AI model-based positioning method provided in the embodiments of the present application.

[0301] In the following description, the terminal device is taken as the UE, and the AI model is deployed in the LMF network element as an example. The method includes the following steps.

[0302] S61: The UE receives PRS.

[0303] The PRS received by the UE can be sent by multiple base stations.

[0304] The PRS sent by each base station can carry one or more of the Cell ID, the RSRP, and the AoA relative to the base station. In a possible implementation manner, the PRS sent by the multiple base stations can carry the Cell IDs of the serving cells and the neighboring cells. In some embodiments, the UE side can not receive the Cell IDs of the neighboring cells. In this case, the PCIs of the neighboring cells can be used to distinguish the neighboring cells.

[0305] S62: The UE obtains the environmental parameter combination of the current environment.

[0306] The environmental parameter combination includes a state parameter of the terminal device and a state parameter of a network device.

[0307] The state parameter of the terminal device can be used to represent a state parameter of an environment in which the terminal device, i.e., the UE, is currently located, which can include, but is not limited to, a spatial state parameter, a time state parameter, and a scene state parameter.

[0308] Specific descriptions about the environmental parameter combination can be referred to the above S22, which will not be repeated here.

[0309] S63: The UE sends the environmental parameter combination and the PRS to the LMF network element.

[0310] Referring to FIG. 9, which is a schematic diagram of a framework for positioning based on an AI model according to an embodiment of the present application.

[0311] The difference between the scheme corresponding to FIG. 9 and FIGS. 5 and 7 is that the mapping table is not stored at the UE side but stored at the LMF network element side, so the LMF network element needs to determine the AI model index according to the mapping table, and therefore the UE needs to send the environmental parameter combination and the PRS to the LMF network element.

[0312] In a possible implementation, the UE and the LMF can communicate based on LPP, and the PRS and the environmental parameter combination can be carried in LPP-related signaling.

[0313] S64: The LMF network element determines one or more AI model indexes that match by using the environmental parameter combination and the PRS.

[0314] The UE can send the PRS sent by each base station to the LMF in sequence, or aggregate and send them to the LMF at a time, which is not limited in the embodiments of the present application.

[0315] The mapping table, i.e., the correspondence table of each environmental parameter combination and the AI model index, can be referred to Table 2 above.

[0316] After receiving the PRS, the LMF network element aggregates information of all the PRS, and further obtains all the Cell IDs. All the Cell IDs indicated in the PRS can form a first Cell ID list.

[0317] The LMF network element determines an intersection of the first Cell ID list and a second Cell ID list corresponding to each spatial region code.

[0318] When the number of Cell IDs included in the intersection is greater than or equal to a first threshold, the LMF network element considers that the corresponding grid region satisfies the screening condition of the spatial state parameter, and takes the spatial region code of the second Cell ID list corresponding to the intersection as the spatial screening condition of the AI model index.

[0319] Further, the screening process of the spatial region code can also refer to the number of Cell ID intersection and other information in the PRS at the same time. For example, further adding RSRP and / or AoA for joint judgment, that is, after determining the number of Cell ID in the intersection, further determining whether the RSRP of the base station corresponding to each Cell ID in the intersection satisfies the preset power threshold, and / or whether the AoA of the base station corresponding to each Cell ID in the intersection satisfies the preset angle range. If yes, the spatial region code of the second Cell ID list corresponding to the intersection is taken as the spatial screening condition of the AI model index.

[0320] For the time state parameter, since it has been discretized into time intervals, the LMF network element can compare the positioning time with each time interval to determine the time interval in which the positioning time is located, or the time interval closest to the positioning time. The time interval determined by the time parameter can have one or more.

[0321] In one possible implementation, the basis for determining the time screening condition is whether the time interval between the positioning time and the reference value corresponding to the time interval is greater than the preset interval. If less, the time state parameter corresponding to the time interval is taken as the time screening condition. Wherein, the reference value is the middle value of the time interval, and the preset interval can be set according to the actual AI model corresponding to the geographical area, and the embodiments of the present application are not limited specifically.

[0322] The scene state parameter can identify the scene in which the UE is located when positioning. After the UE determines the scene in which it is located, the scene state parameter corresponding to the scene can be taken as the scene screening condition.

[0323] For the network device state parameter, the network device state parameter can be quantified into discrete intervals for matching. Taking the time-related network device state parameter as an example, the network device screening condition can be determined in a similar manner to the time state parameter; the space-related network device state parameter can be determined in a similar manner to the spatial region code.

[0324] In the above manner, the LMF network element can determine one or more AI model indexes matched by the mapping table.

[0325] S65: The LMF network element determines the AI model corresponding to each AI model index in one or more AI model indexes respectively.

[0326] The LMF network element side is deployed with an AI model library for UE positioning. The AI model library includes a plurality of trained AI models, each of which corresponds to an AI model index.

[0327] After the LMF network element obtains one or more AI model indexes, it matches the AI models corresponding to the AI model indexes from the AI model library.

[0328] S66: The LMF sends all the determined AI models to the UE.

[0329] In a possible implementation, the UE and the LMF can communicate based on LPP, and the related data of the AI model can be carried in the LPP-related signaling.

[0330] S67: The UE determines the weight of the positioning result output by each AI model using the environmental parameter combination.

[0331] In the fusion of the positioning result, the weight of each AI model needs to be determined, which is the weight of the positioning result output by the AI model. After the weight corresponding to each AI model is determined, the positioning result of each AI model is multiplied by the corresponding weight, and then the obtained calculation result is accumulated to determine the final positioning result.

[0332] It can be understood that if only one AI model index is determined, the weight of the AI model can be directly determined as 1, that is, the positioning result of the AI model can be directly used as the final positioning result. The following describes the case when multiple AI model indexes are determined.

[0333] For each AI model, the weight coefficient K of the corresponding AI model can be determined by the above formula (1).

[0334] The greater the weight coefficient, the higher the credibility of the AI model filtered by the environmental parameter corresponding to the weight coefficient, and the closer the positioning result of the AI model to the true position.

[0335] For the spatial state parameter, that is, the spatial region code, the more the number of Cell IDs in the intersection when filtering the model, the higher the weight coefficient of the corresponding spatial state parameter.

[0336] In a possible implementation, the weight coefficient is the ratio of the number of Cell IDs in the intersection corresponding to the spatial region code to the number of Cell IDs included in the first Cell ID list.

[0337] In another possible implementation, the weight coefficient is the ratio between the number of Cell IDs in the intersection corresponding to the spatial region code and the number of Cell IDs in the intersection corresponding to all AI model indexes.

[0338] For the time state parameter, the size of the weight coefficient is negatively related to the size of the time interval between the reference value of the time interval and the positioning time. The larger the time interval, the smaller the weight coefficient.

[0339] In a possible implementation, the weight coefficient K2 of the time state parameter can be determined by the above formula (2).

[0340] For the scene state parameter, the weight coefficient corresponding to each scene can be calibrated by means of big data and AI model analysis, and the correspondence between the scene and the weight is saved in the UE.

[0341] After determining the weight coefficient K of each model by formula (1), the weight coefficients of each AI model are normalized to obtain the weight of the AI model distributed in [0, 1]. After normalization, the weight coefficient K of each AI model is converted into the weight k, and the weight coefficients of all AI models add up to 1.

[0342] S68: The UE determines the final positioning result by using the PRS, the AI models, and the weight corresponding to each AI model.

[0343] The UE takes the acquired PRS as the input of each AI model to obtain the positioning result output by each AI model. The positioning result output by a single AI model can be the absolute position or the relative position of the UE, which is not limited in the embodiments of the present application. It can be understood that the types of the positioning results output by each AI model are generally the same, for example, both are absolute positions or both are relative positions.

[0344] The absolute position can be the latitude and longitude coordinates. The relative position can be the position coordinates of the UE relative to the serving base station and / or the base station corresponding to the neighboring cell, and the relative position can also be represented in other ways, which are not limited in the embodiments of the present application.

[0345] The final positioning result position can be determined by the above formula (3), that is, the positioning result output by each AI model is multiplied by the weight corresponding to each AI model to obtain the final positioning result component corresponding to each AI model, and the final positioning result components corresponding to each AI model are superimposed to obtain the final positioning result corresponding to each AI model.

[0346] It can be understood that the above steps are only divided for convenience of explanation, and do not constitute a limitation on the technical solutions of the present application. In actual application, the order of the above steps can be changed, for example, S67 can be moved forward to before S63, or S67 can be performed in parallel with S63.

[0347] In summary, the positioning method provided in the embodiments of the present application, after the UE sends the environment parameter combination and the PRS to the LMF network element, the LMF can determine one or more AI model indexes corresponding to the AI model for positioning, and then the LMF network element matches the corresponding AI model from the AI model library according to each AI model index and sends it to the UE. After the UE side determines the weight corresponding to each AI model, the final positioning result is determined by using each AI model and the corresponding weight. In the existing related scheme, when the AI model that completely matches the current environment parameter combination of the UE cannot be determined, it is directly considered that the positioning cannot be completed, while the scheme provided in the embodiments of the present application can use multiple AI models that have a high matching degree with the current environment parameters of the UE for positioning, and fuse the positioning results according to the weights, and then obtain a more accurate final positioning result. It is especially suitable for service areas with complex and changeable real environment, or areas where the AI model for positioning is not fully trained, and has high practicability.

[0348] At the same time, after the LMF network element sends the AI model to the UE, for the scenario that the UE side needs to perform continuous positioning, the UE side can directly use the received AI model for positioning, and if the environment parameter combination does not change, it is no longer necessary to frequently interact with the LMF network element, thereby reducing the time delay of positioning.

[0349] The following describes an implementation manner in which the AI model index is determined by the LMF network element, and the fusion positioning is performed by the LMF network element.

[0350] Referring to FIG. 10, which is a flowchart of another AI model-based positioning method provided in the embodiments of the present application.

[0351] In the following description, the terminal device is taken as the UE, and the AI model is deployed in the LMF network element as an example. The method includes the following steps.

[0352] S81: The UE receives the PRS.

[0353] The PRS received by the UE can be sent by multiple base stations.

[0354] The PRS sent by each base station can carry one or more of the Cell ID, the RSRP, and the AoA relative to the base station. In a possible implementation manner, the PRS sent by the multiple base stations can carry the Cell IDs of the serving cell and the neighboring cells. In some embodiments, the UE side can not receive the Cell ID of the neighboring cell, and in this case, the PCI of the neighboring cell can be used to distinguish the neighboring cell.

[0355] S82: The UE obtains the environment parameter combination of the current environment.

[0356] The environmental parameter combination includes a state parameter of the terminal device and a state parameter of a network device.

[0357] The state parameter of the terminal device can be used to represent a state parameter of an environment in which the terminal device, i.e., the UE, is currently located, which can include, but is not limited to, a spatial state parameter, a time state parameter, and a scene state parameter.

[0358] For specific descriptions of the environmental parameter combination, refer to the descriptions in S22 above, which are not repeated here.

[0359] S83: The UE sends the environmental parameter combination and the PRS to the LMF network element.

[0360] Referring to FIG. 11, which is a schematic diagram of a framework for positioning based on an AI model according to an embodiment of the present application.

[0361] In the scheme shown in FIG. 11, the UE side does not store a mapping table, and the mapping table is stored at the LMF network element side. Therefore, the LMF network element needs to determine the AI model index according to the mapping table, and thus the UE needs to send the environmental parameter combination and the PRS to the LMF network element.

[0362] In a possible implementation, the UE and the LMF can communicate based on LPP, and the PRS and the environmental parameter combination can be carried in LPP-related signaling.

[0363] S84: The LMF network element determines one or more AI model indexes that match the environmental parameter combination and the PRS.

[0364] The UE can send the PRS sent by each base station to the LMF one by one, or aggregate and send them to the LMF at a time, which is not limited in the embodiments of the present application.

[0365] The mapping table, i.e., the correspondence table of the environmental parameter combination and the AI model index, can refer to Table 2 above.

[0366] After receiving the PRS, the LMF network element aggregates the information of all the PRS, and further obtains all the Cell IDs. All the Cell IDs indicated in the PRS can form a first Cell ID list.

[0367] The LMF network element determines the intersection of the first Cell ID list and a second Cell ID list corresponding to each spatial region code. When the number of Cell IDs included in the intersection is greater than or equal to a first threshold, the LMF network element considers that the corresponding grid region satisfies the screening condition of the spatial state parameter, and takes the spatial region code of the second Cell ID list corresponding to the intersection as the spatial screening condition of the AI model index.

[0368] Further, the screening process of the spatial region code can also refer to the number of Cell ID intersection and other information in the PRS at the same time. For example, further adding RSRP and / or AoA for joint judgment, that is, after determining the number of Cell ID in the intersection, further determining whether the RSRP of the base station corresponding to each Cell ID in the intersection satisfies the preset power threshold, and / or whether the AoA of the base station corresponding to each Cell ID in the intersection satisfies the preset angle range. If yes, the spatial region code of the second Cell ID list corresponding to the intersection is taken as the spatial screening condition of the AI model index.

[0369] For the time state parameter, since it has been discretized into time intervals, the LMF network element can compare the positioning time with each time interval to determine the time interval in which the positioning time is located, or the time interval closest to the positioning time. The time interval determined by the time parameter can have one or more.

[0370] In one possible implementation, the basis for determining the time screening condition is whether the time interval between the positioning time and the reference value corresponding to the time interval is greater than the preset interval. If less, the time state parameter corresponding to the time interval is taken as the time screening condition. Wherein, the reference value is the middle value of the time interval, and the preset interval can be set according to the actual AI model corresponding to the geographical area, and the embodiments of the present application are not limited specifically.

[0371] The scene state parameter can identify the scene in which the UE is located when positioning. After the UE determines the scene in which it is located, the scene state parameter corresponding to the scene can be taken as the scene screening condition.

[0372] For the network device state parameter, the network device state parameter can be quantified into discrete intervals for matching. Taking the time-related network device state parameter as an example, the network device screening condition can be determined in a similar manner to the time state parameter; the space-related network device state parameter can be determined in a similar manner to the spatial region code.

[0373] In the above manner, the LMF network element can determine one or more AI model indexes matched by the mapping table.

[0374] S85: The LMF network element determines the AI model corresponding to each AI model index in one or more AI model indexes respectively.

[0375] The LMF network element side is deployed with an AI model library for UE positioning. The AI model library includes a plurality of trained AI models, each of which corresponds to an AI model index. After the LMF network element obtains one or more AI model indexes, it matches the AI models corresponding to the AI model indexes from the AI model library.

[0376] S86: The LMF network element determines the weight of the positioning result output by each AI model by combining the environmental parameters.

[0377] In the fusion of the positioning result, the weight of each AI model needs to be determined, which is the weight of the positioning result output by the AI model. After the weight corresponding to each AI model is determined, the positioning result of each AI model is multiplied by the corresponding weight, and then the obtained calculation result is accumulated to determine the final positioning result.

[0378] It can be understood that if only one AI model index is determined, the weight of the AI model can be directly determined as 1, that is, the positioning result of the AI model can be directly used as the final positioning result. The following describes the case when multiple AI model indexes are determined.

[0379] For each AI model, the weight coefficient K of the corresponding AI model can be determined by the above formula (1).

[0380] The greater the weight coefficient, the higher the credibility of the AI model screened by the environmental parameter corresponding to the weight coefficient, and the closer the positioning result of the AI model to the true position.

[0381] For the spatial state parameter, that is, the spatial region code, the more the number of Cell IDs in the intersection when screening the model, the higher the weight coefficient of the corresponding spatial state parameter.

[0382] In one possible implementation, the weight coefficient is the ratio of the number of Cell IDs in the intersection corresponding to the spatial region code to the number of Cell IDs included in the first Cell ID list.

[0383] In another possible implementation, the weight coefficient is the ratio between the number of Cell IDs in the intersection corresponding to the spatial region code and the number of Cell IDs in the intersection corresponding to all AI model indexes.

[0384] For the time state parameter, the size of the weight coefficient is negatively related to the size of the time interval between the reference value of the time interval and the positioning time. The larger the time interval, the smaller the weight coefficient.

[0385] In one possible implementation, the weight coefficient K2 of the time state parameter can be determined by the above formula (2).

[0386] For the scene state parameters, the weight coefficients corresponding to each scene can be calibrated through big data and AI model analysis, and the corresponding relationship between the scene and the weight is saved in the UE.

[0387] After the LMF determines the weight coefficients K corresponding to each model through formula (1), the weight coefficients of each AI model are normalized to obtain the weight of the AI model distributed in [0, 1]. After normalization, the weight coefficient K of each AI model is converted into the weight k, and the weight coefficients of all AI models are added to equal 1.

[0388] S87: The LMF network element determines the final positioning result by using the PRS, each AI model, and the weight corresponding to each AI model.

[0389] The LMF network element takes the obtained PRS as the input of each AI model to obtain the positioning result output by each AI model. The positioning result output by a single AI model can be the absolute position or the relative position of the UE, and the embodiments of the present application are not limited specifically. It can be understood that the types of the positioning results output by each AI model are generally the same, for example, both are absolute positions or both are relative positions.

[0390] The final positioning result position can be determined through the above formula (3), that is, the positioning result output by each AI model is multiplied by the weight corresponding to each AI model to obtain the final positioning result component corresponding to each AI model, and the final positioning result components corresponding to each AI model are superimposed to obtain the final positioning result corresponding to each AI model.

[0391] S88: The LMF network element sends the position information to the UE.

[0392] The position information includes the final positioning result. In one possible implementation, the LMF network element can carry the position information in the related signaling of the LTE positioning protocol.

[0393] It can be understood that the division of the above steps is only for convenience of explanation, and does not constitute a limitation on the technical solutions of the present application. In actual application, the order of the above steps can be changed, for example, S87 can be moved forward to before S84, or S47 can be performed in parallel with S45.

[0394] In summary, the positioning method provided in the embodiments of the present application sends the PRS and the environment parameter combination to the LME network element from the UE side, the LMF network element can determine the AI model index by using the PRS and the environment parameter combination, and the LMF network element matches the corresponding AI model from the AI model library according to each AI model index and determines the weight corresponding to each AI model. Then the LMF can determine the final positioning result by using each AI model and the corresponding weight, and sends the final positioning result to the UE. In the existing related scheme, when the AI model completely matched with the current environment parameter combination of the UE cannot be determined, it is directly considered that the positioning cannot be completed, while the scheme provided in the embodiments of the present application can use multiple AI models with high matching degree with the current environment parameter of the UE for positioning, and fuse the positioning results according to the weight, and then obtain a more accurate final positioning result. It is especially suitable for service areas with complex and changeable real environment, or areas with insufficient AI model training for positioning, and has high practicability.

[0395] Meanwhile, after the LMF network element sends the AI model to the UE in this way, for the scenario that the UE side needs to continuously position, the UE side can directly use the received AI model for positioning, and if the environment parameter combination does not change subsequently, it is no longer necessary to frequently interact with the LMF network element, thereby reducing the time delay of positioning.

[0396] Based on the method provided in the above embodiments, the embodiments of the present application further provide a terminal device, which will be specifically described below with reference to the accompanying drawings.

[0397] Referring to FIG. 12, it is a schematic diagram of a terminal device provided in the embodiments of the present application.

[0398] The terminal device 100 can include a processor 110, an external memory interface 120, an internal memory 121, a universal serial bus (USB) interface 130, a charge management module 140, a power management module 141, a battery 142, an antenna 1, an antenna 2, a mobile communication module 150, a wireless communication module 160, a sensor module 180, a key 190, a motor 191, an indicator 192, a camera 193, a display screen 194, etc.

[0399] The sensor module 180 can include a gyroscope sensor 180A, an air pressure sensor 180B, an acceleration sensor 180C, etc.

[0400] It can be understood that the structure shown in the embodiments of the present application does not constitute a specific limitation on the terminal device 100. In other embodiments of the present application, the terminal device 100 can include more or fewer components than shown, or combine certain components, or split certain components, or different component arrangements. The components shown can be implemented in hardware, software, or a combination of software and hardware.

[0401] The processor 110 can include one or more processing units, for example: the processor 110 can include an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Among them, different processing units can be independent devices, or can be integrated in one or more processors. For example, the controller can generate operation control signals according to instruction operation codes and timing signals, complete the control of fetching instructions and executing instructions.

[0402] The memory in the processor 110 can also be configured to store instructions and data. In some embodiments, the memory in the processor 110 is a cache memory. The memory can save instructions or data that the processor 110 has just used or repeatedly uses. If the processor 110 needs to use the instructions or data again, it can directly call from the memory. Avoiding repeated access, reducing the waiting time of the processor 110, thus improving the efficiency of the system.

[0403] The NPU is a neural-network (NN) computing processor. When the terminal device receives the AI model sent by the LMF network element, the NPU can be used to run the AI model to obtain the final positioning result.

[0404] The barometric pressure sensor 180B is used to measure the barometric pressure. In some embodiments, the terminal device 100 calculates the altitude by the barometric pressure value measured by the barometric pressure sensor 180B, to assist in determining the current scene of the terminal device.

[0405] The gyroscope sensor 180A can be used to determine the motion posture of the terminal device 100. The acceleration sensor 180C can detect the magnitude of acceleration of the terminal device 100 in each direction (generally three axes). The terminal device can use the detection results output by the gyroscope sensor 180A and the acceleration sensor 180C to assist in determining the current scene in which the terminal device is located.

[0406] The wireless communication function of the terminal device 100 can be implemented by the antenna 1, the antenna 2, the mobile communication module 150, the wireless communication module 160, the modem processor, and the baseband processor, etc.

[0407] The antenna 1 and the antenna 2 are used to transmit and receive electromagnetic wave signals. Each antenna in the terminal device 100 can be used to cover a single or multiple communication frequency bands. Different antennas can also be multiplexed to improve the utilization rate of the antennas. For example, the antenna 1 can be multiplexed as a diversity antenna of a wireless local area network. In some other embodiments, the antennas can be used in combination with a tuning switch.

[0408] The mobile communication module 150 can provide a solution for wireless communication including 2G / 3G / 4G / 5G, etc. applied to the terminal device 100. The wireless communication module 160 can provide a solution for wireless communication including a wireless local area network (WLAN) (such as a Wi-Fi network), Bluetooth (BT), a global navigation satellite system (GNSS), frequency modulation (FM), near field communication (NFC), infrared technology (IR), etc. applied to the terminal device 100.

[0409] The processor 110 of the terminal device can be used to execute the computer programs and / or instructions stored in the memory to implement the AI model-based positioning method described in the above embodiments.

[0410] The embodiments of the present application also provide a communication apparatus which can be used as a positioning device. The communication apparatus includes but is not limited to a base station, a core network element, and other network devices. Taking the core network device as an example for description.

[0411] Referring to FIG. 13, which is a schematic diagram of a communication apparatus provided by an embodiment of the present application.

[0412] The illustrated communication apparatus 1100 includes a processor 1110, a memory 1120, and a transceiver 1130.

[0413] The processor 1110 is mainly used for baseband processing, controlling the communication device 1100, etc. The processor 1110 is usually the control center of the communication device 1100, and is used to control the communication device 1100 to perform the AI model-based positioning method in the above method embodiments.

[0414] The memory 1120 is mainly used for storing computer program codes and data.

[0415] For example, for the schemes shown in FIGS. 5 and 7, the AI model library can be stored on the memory, and the AI model library can include a plurality of AI models.

[0416] For example, for the schemes shown in FIGS. 9 and 11, the mapping table and the AI model library can be stored on the memory, and the AI model library can include a plurality of AI models.

[0417] The transceiver 1130 is mainly used for transceiving radio frequency signals and converting radio frequency signals and baseband signals. The transceiver 1130 can also be referred to as a transceiver, a transceiving circuit, etc.

[0418] The transceiving module of the transceiver 1130 can include an antenna 1133 and a radio frequency circuit (not shown in the figure), wherein the radio frequency circuit is mainly used for radio frequency processing.

[0419] Optionally, the devices in the transceiver 1130 for implementing the receiving function can be regarded as a receiver 1032, and the devices for implementing the sending function can be regarded as a transmitter 1031. The receiver 1032 can also be referred to as a receiving module, a receiver, or a receiving circuit, etc. The transmitter 1031 can be referred to as a transmitting module, a transmitter, or a transmitting circuit, etc.

[0420] The processor 1110 and the memory 1120 can include one or more single boards, and each single board can include one or more processors and one or more memories.

[0421] The processor 1110 is used to read and execute the program in the memory 1120 to realize the control of the communication device. If there are multiple single boards, the single boards can be interconnected to enhance the processing capability. As an optional implementation, multiple single boards can share one or more processors, or multiple single boards can share one or more memories, or multiple single boards can share one or more processors at the same time.

[0422] For example, in an implementation, the transceiving module of the transceiver 1130 is used to perform the process of transceiving the related information performed by the LMF network element in the above method embodiments. The processor 1110 is used to perform the positioning method performed by the LMF network element in the above method embodiments.

[0423] It should be understood that FIG. 13 is merely an example and not limiting, and the communication device including the processor, the memory, and the transceiver described above can not depend on the structure shown in FIG. 13.

[0424] The embodiments of the present application also provide a storage medium. The computer readable storage medium can be any available medium or data storage device that can be accessed by a computing device. The available medium can be a magnetic medium (e.g., a floppy diskette, a hard disk drive, a magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid state hard drive) or the like. The computer readable storage medium includes instructions that instruct a terminal device or a communication device to perform the AI model based positioning method described above. The embodiments of the present application also provide another computer readable storage medium. The computer readable storage medium includes instructions that instruct a terminal device or a communication device to perform the AI model based positioning method described above.

[0425] The embodiments of the present application also provide a computer program product including instructions. The computer program product can be software or a program product including instructions that can be run on a terminal device or a communication device, or stored in any available medium. When the computer program product is run on a terminal device or a communication device, the terminal device or the communication device is caused to perform the AI model based positioning method described above. The embodiments of the present application also provide a computer program product including instructions. When the computer program product is run on a terminal device or a communication device, the terminal device or the communication device is caused to perform the AI model based positioning method described above.

[0426] In the embodiments of the present application, "at least one" means one or more, and "multiple" means two or more. "At least one of the following" or the like means any combination of the items, including any combination of single item or multiple items. For example, at least one of a, b, or c can represent a, b, c, a-b, a-c, b-c, or a-b-c, where a, b, and c can be single or multiple.

[0427] The above description and the above embodiments are merely used to illustrate the technical solutions of the present application, but not to limit the present application; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An AI model-based positioning method, characterized by, The method comprises: determining one or more AI model indexes matched by using a combination of a received positioning reference signal (PRS) and environment parameters, wherein the combination of environment parameters comprises state parameters of a terminal device and state parameters of a network device; sending first index information, wherein the first index information comprises the one or more AI model indexes; receiving an AI model corresponding to each of the one or more AI model indexes respectively; determining a weight of a positioning result output by each of the AI models by using the combination of environment parameters; determining a final positioning result by using the PRS, the AI models, and the weights corresponding to the AI models.

2. The method of claim 1, wherein, The PRS carries one or more of a cell identifier (Cell ID), a reference signal received power (RSRP), and an angle of arrival (AoA) relative to a base station.

3. The method of claim 2, wherein, The PRS carries a cell identifier (Cell ID), and the combination of environment parameters comprises state parameters of a terminal device and state parameters of a network device. The state parameters of the terminal device comprise one or more of a spatial region code, a time state parameter, and a scene state parameter, wherein the spatial region code indicates a grid region in which a cell corresponding to each of the Cell IDs is located, the time state parameter is used to indicate a positioning time, and the scene state parameter is an identifier of a scene in which the terminal device is located. The state parameters of the network device indicate a parameter configuration of the network device.

4. The method of claim 3, wherein, The network device is a base station, and the configuration parameter comprises a beam configuration parameter of a downlink PRS signal of the base station.

5. The method of claim 3, wherein, The scene state parameter is determined by the terminal device by using detection results of a plurality of sensors, wherein the sensors comprise one or more of an air pressure sensor, an accelerometer, and a gyroscope.

6. The method of claim 3, wherein, The PRS carries a cell identifier (Cell ID), and the state parameters of the terminal device comprise a spatial region code, a time state parameter, and a scene state parameter. The method comprises: determining a first Cell ID list comprising all Cell IDs carried in the PRS; determining an intersection of a second Cell ID list corresponding to each spatial region code and the first Cell ID list respectively, wherein the second Cell ID list comprises all Cell IDs of cells in a grid region corresponding to the spatial region code; taking a spatial region code corresponding to an intersection whose number of Cell IDs is greater than or equal to a first threshold value as a first screening condition; when a time interval between a positioning time and a reference value corresponding to a time interval is less than a preset interval, taking a time state parameter corresponding to the time interval as a second screening condition, wherein the reference value corresponding to the time interval is a middle value of the time interval; taking a scene state parameter as a third screening condition; determining a fourth screening condition according to the state parameters of the network device; and According to the first screening condition, the second screening condition, the third screening condition, the fourth screening condition, and a mapping table of environmental parameter combinations and AI model indexes, one or more AI model indexes that match are determined.

7. The method of claim 3, wherein, The PRS carries a cell identifier Cell ID and the RSRP, the state parameters of the terminal device include spatial region encoding, time state parameters, and scene state parameters, and the determination of one or more AI model indexes that match by using the received positioning reference signal PRS and the environmental parameter combination includes: A first Cell ID list is determined, and the first Cell ID list includes all Cell IDs carried in the PRS; An intersection of each spatial region encoding corresponding second Cell ID list and the first Cell ID list is determined respectively, and the second Cell ID list includes all Cell IDs of cells in a grid region corresponding to the spatial region encoding; A spatial region encoding corresponding to an intersection that meets the following conditions is taken as a first screening condition: The number of Cell IDs whose RSRP is greater than or equal to a preset power threshold is greater than a first threshold value; When a time interval between a positioning time and a reference value corresponding to a time interval is less than a preset interval, a time state parameter corresponding to the time interval is taken as a second screening condition, and the reference value corresponding to the time interval is a middle value of the time interval; A scene state parameter in which the terminal device is located is taken as a third screening condition; A fourth screening condition is determined according to the state parameters of the network device; According to the first screening condition, the second screening condition, the third screening condition, the fourth screening condition, and a mapping table of environmental parameter combinations and AI model indexes, one or more AI model indexes that match are determined.

8. The method of claim 3, wherein, The PRS carries a cell identifier Cell ID, the RSRP, and the AoA of the relative base station, and the determination of one or more AI model indexes that match by using the received positioning reference signal PRS and the environmental parameter combination includes: A first Cell ID list is determined, and the first Cell ID list includes all Cell IDs carried in the PRS; An intersection of each spatial region encoding corresponding second Cell ID list and the first Cell ID list is determined respectively, and the second Cell ID list includes all Cell IDs of cells in a grid region corresponding to the spatial region encoding; A spatial region encoding corresponding to an intersection that meets the following conditions is taken as a first screening condition: The number of Cell IDs whose RSRP is greater than or equal to a preset power threshold and whose AoA of the relative base station meets a preset angle range is greater than a first threshold value; When a time interval between a positioning time and a reference value corresponding to a time interval is less than a preset interval, a time state parameter corresponding to the time interval is taken as a second screening condition, and the reference value corresponding to the time interval is a middle value of the time interval; A scene state parameter in which the terminal device is located is taken as a third screening condition; A fourth screening condition is determined according to the state parameters of the network device; According to the first screening condition, the second screening condition, the third screening condition, the fourth screening condition, and a mapping table of environmental parameter combinations and AI model indexes, one or more AI model indexes that match are determined.

9. The method according to any one of claims 6-8, characterized in that, The state parameter of the network device indicates a parameter configuration of the network device, and the weight of the positioning result output by each AI model is determined according to the environmental parameter combination. According to the number of Cell IDs in the corresponding intersection of each spatial region code as the first screening condition, a first weight coefficient of each AI model is determined, and the first weight coefficient is positively correlated with the number of Cell IDs in the intersection. According to the time interval, a second weight coefficient of each AI model is determined, and the second weight coefficient is negatively correlated with the time interval. According to the scene state parameter, a third weight coefficient of each AI model is determined. According to the state parameter of the network device, a fourth weight coefficient of each AI model is determined. According to the first weight coefficient, the second weight coefficient, the third weight coefficient, and the fourth weight coefficient, a weight coefficient of each AI model is determined. The weight coefficients of each AI model are normalized to obtain the weight of each AI model, and the sum of the weights of each AI model is 1.

10. The method of claim 9, wherein, The state parameter of the network device indicates a parameter configuration of the network device in different time periods, and the weight coefficient of each AI model is determined according to the first weight coefficient, the second weight coefficient, the third weight coefficient, and the fourth weight coefficient, including: For each AI model, the following steps are performed: The product of the first influence factor and the first weight coefficient, the product of the first influence factor and the third weight coefficient, the product of the second influence factor and the second weight coefficient, and the product of the second influence factor and the fourth weight coefficient are summed to determine the weight coefficient of each AI model, wherein the first influence factor is greater than the second influence factor.

11. The method of claim 9, wherein, The final positioning result is determined by using the PRS, the AI models, and the weights corresponding to the AI models, including: The PRS is used as the input of each AI model to obtain the positioning result output by each AI model. The positioning result output by each AI model is multiplied by the weight corresponding to the AI model to obtain a final positioning result component corresponding to each AI model. The final positioning result components corresponding to each AI model are superimposed to obtain the final positioning result.

12. The method according to claim 6 or 7 or 8, characterized in that, The Cell IDs of all cells in the grid region corresponding to the spatial region code include the Cell IDs of all cells in the spatial region code for which the positioning assistance data can be measured.

13. The method of claim 6 or 7 or 8, wherein, The Cell IDs include a cell global identifier (CGI) and / or an E-UTRAN cell global identifier (ECI).

14. The method of claim 6 or 7 or 8, wherein, The second Cell ID list further includes physical cell identifiers (PCIs) and / or absolute radio frequency channel numbers (ARFCNs) of all cells in a grid area corresponding to the spatial area code.

15. An AI model-based positioning method, characterized by, The method comprises: sending an environment parameter combination and a positioning reference signal (PRS), wherein the environment parameter combination includes state parameters of a terminal device and state parameters of a network device; receiving one or more AI models, each of the one or more AI models corresponding to an AI model index, and each AI model index being determined based on the environment parameter combination and the PRS; determining a weight of a positioning result output by each AI model based on the environment parameter combination; determining a final positioning result based on the PRS, each AI model, and the weight corresponding to each AI model.

16. The method of claim 15, wherein, The PRS carries one or more of a cell identifier (Cell ID), a reference signal received power (RSRP), and an angle of arrival (AoA) relative to a base station.

17. The method of claim 16, wherein, The PRS carries a cell identifier (Cell ID), and the environment parameter combination includes state parameters of a terminal device and state parameters of a network device. The state parameters of the terminal device include one or more of a spatial area code, a time state parameter, and a scene state parameter, wherein the spatial area code indicates a grid area in which a cell corresponding to each Cell ID is located, the time state parameter is used to indicate a positioning time, and the scene state parameter is an identifier of a scene in which the terminal device is located. The state parameters of the network device indicate a parameter configuration of the network device.

18. The method of claim 17, wherein, The network device is a base station, and the configuration parameter includes a beam configuration parameter of a downlink PRS signal of the base station.

19. The method of claim 17, wherein, The PRS carries a cell identifier (Cell ID), and the state parameters of the terminal device include a spatial area code, a time state parameter, and a scene state parameter. The method comprises: determining a first Cell ID list, wherein the first Cell ID list includes all Cell IDs carried in the PRS; determining an intersection of each spatial area code corresponding second Cell ID list and the first Cell ID list, wherein the second Cell ID list includes Cell IDs of all cells in a grid area corresponding to the spatial area code; taking a spatial area code corresponding to an intersection whose number of Cell IDs is greater than or equal to a first threshold value as a first screening condition; determining a first weight coefficient of each AI model based on a number of Cell IDs in the intersection corresponding to each spatial area code serving as the first screening condition, wherein the first weight coefficient is positively correlated with the number of Cell IDs in the intersection; determining a second weight coefficient of each AI model based on a time interval between a positioning time and a reference value corresponding to a time interval, wherein the second weight coefficient is negatively correlated with the time interval; determining a third weight coefficient of each AI model based on a scene state parameter; and determining a final positioning result based on the PRS, each AI model, and the weight corresponding to each AI model. determine a fourth weight coefficient of each AI model according to the state parameter of the network device; determine a weight coefficient of each AI model according to the first weight coefficient, the second weight coefficient, the third weight coefficient and the fourth weight coefficient; normalize the weight coefficient of each AI model to obtain the weight of each AI model, and the sum of the weights of the AI models is 1.

20. The method of claim 17, wherein, The PRS carries a cell identifier Cell ID and the RSRP, the state parameter of the terminal device includes a spatial region code, a time state parameter and a scene state parameter, and the weight of the positioning result output by each AI model is determined according to the environmental parameter combination, specifically including: determining a first Cell ID list, the first Cell ID list including all Cell IDs carried in the PRS; determining the intersection of each spatial region code corresponding second Cell ID list and the first Cell ID list respectively, the second Cell ID list including all cell IDs of cells in the grid area corresponding to the spatial region code; the spatial region code corresponding to the intersection meeting the following conditions is taken as the first screening condition: the number of Cell IDs whose RSRP is greater than or equal to a preset power threshold is greater than a first threshold; determining a first weight coefficient of each AI model according to the number of Cell IDs in the intersection corresponding to each spatial region code as the first screening condition, the first weight coefficient being positively correlated with the number of Cell IDs in the intersection; determining a second weight coefficient of each AI model according to the time interval between the positioning time and the reference value corresponding to the time interval, the second weight coefficient being negatively correlated with the time interval; determining a third weight coefficient of each AI model according to the scene state parameter; determining a fourth weight coefficient of each AI model according to the state parameter of the network device; determining a weight coefficient of each AI model according to the first weight coefficient, the second weight coefficient, the third weight coefficient and the fourth weight coefficient; normalizing the weight coefficient of each AI model to obtain the weight of each AI model, and the sum of the weights of the AI models is 1.

21. The method of claim 17, wherein, The PRS carries a cell identifier Cell ID, the state parameter of the terminal device includes a spatial region code, a time state parameter and a scene state parameter, and the weight of the positioning result output by each AI model is determined according to the environmental parameter combination, specifically including: determining a first Cell ID list, the first Cell ID list including all Cell IDs carried in the PRS; determining the intersection of each spatial region code corresponding second Cell ID list and the first Cell ID list respectively, the second Cell ID list including all cell IDs of cells in the grid area corresponding to the spatial region code; A space region corresponding to an intersection of the following conditions is encoded as a first screening condition: The RSRP is greater than or equal to a preset power threshold, and the number of Cell IDs whose AoA relative to the base station satisfies a preset angle range is greater than a first threshold; The number of Cell IDs whose RSRP is greater than or equal to a preset power threshold is greater than a first threshold; A first weight coefficient of each AI model is determined according to the number of Cell IDs in the intersection corresponding to each space region code as the first screening condition, and the first weight coefficient is positively correlated with the number of Cell IDs in the intersection; A second weight coefficient of each AI model is determined according to a time interval between a positioning time and a reference value corresponding to a time interval, and the second weight coefficient is negatively correlated with the time interval; A third weight coefficient of each AI model is determined according to a scene state parameter; A fourth weight coefficient of each AI model is determined according to a state parameter of the network device; A weight coefficient of each AI model is determined according to the first weight coefficient, the second weight coefficient, the third weight coefficient, and the fourth weight coefficient; The weight coefficients of each AI model are normalized to obtain the weight of each AI model, and the sum of the weights of each AI model is 1.

22. The method of any one of claims 19-21, wherein, The state parameter of the network device indicates the parameter configuration of the network device at different time periods, and the determination of the weight coefficient of each AI model according to the first weight coefficient, the second weight coefficient, the third weight coefficient, and the fourth weight coefficient comprises: For each AI model, the following steps are performed: The products of a first influence factor and the first weight coefficient, the first influence factor and the third weight coefficient, the product of a second influence factor and the second weight coefficient, and the product of a second influence factor and the fourth weight coefficient are summed to determine the weight coefficient of each AI model, wherein the first influence factor is greater than the second influence factor.

23. The method of any one of claims 19-21, wherein, The determination of the final positioning result using the PRS, each AI model, and the weight corresponding to each AI model comprises: The PRS is used as the input of each AI model to obtain the positioning result output by each AI model; The positioning result output by each AI model is multiplied by the weight corresponding to each AI model to obtain a final positioning result component corresponding to each AI model; The final positioning result components corresponding to each AI model are superimposed to obtain the final positioning result.

24. The method of claim 19 or 20 or 21, wherein, The Cell IDs of all cells within the grid region corresponding to the space region code include the Cell IDs of all cells within the space region corresponding to the code for which the positioning assistance data can be measured.

25. The method of claim 19 or 20 or 21, wherein, The Cell IDs include a cell global identifier (CGI) and / or an E-UTRAN cell global identifier (ECI).

26. The method of claim 19 or 20 or 21, wherein, The second Cell ID list also includes the physical cell identifier (PCI) and / or the absolute radio frequency channel number (ARFCN) of all cells within the grid region corresponding to the space region code. 27.A positioning method based on an AI model, characterized in that, The method comprises: determining one or more AI model indexes matched by using a combination of a received positioning reference signal PRS and an environment parameter, the environment parameter combination comprising a state parameter of a terminal device and a state parameter of a network device; sending first index information, the PRS and the environment parameter combination, the first index information comprising the one or more AI model indexes; receiving position information, the position information comprising a final positioning result determined by the PRS, one or more AI models, a weight corresponding to each AI model in the one or more AI models, each AI model corresponding to one AI model index in the one or more AI model indexes, and the weight being a weight of a positioning result output by each AI model determined by using the environment parameter combination.

28. An AI model based positioning method, comprising: The method comprises: sending an environment parameter combination and a positioning reference signal PRS, the environment parameter combination comprising a state parameter of a terminal device and a state parameter of a network device; receiving position information, the position information comprising a final positioning result determined by the PRS, one or more AI models, a weight corresponding to each AI model in the one or more AI models, each AI model corresponding to one AI model index, and the weight being a weight of a positioning result output by each AI model determined by using the environment parameter combination. 29.A positioning method based on an AI model, characterized in that, The method comprises: receiving first index information, the first index information comprising one or more AI model indexes, the one or more AI model indexes being determined by using a combination of a received positioning reference signal PRS and an environment parameter, the environment parameter combination comprising a state parameter of a terminal device and a state parameter of a network device; sending an AI model corresponding to each AI model index in the one or more AI model indexes, each AI model being used to determine a final positioning result together with the PRS and a weight corresponding to each AI model, and the weight being a weight of a positioning result output by each AI model determined by using the environment parameter combination.

30. An AI model based positioning method, comprising: The method comprises: receiving an environment parameter combination and a positioning reference signal PRS, the environment parameter combination comprising a state parameter of a terminal device and a state parameter of a network device; determining one or more AI model indexes matched by using the environment parameter combination and the PRS; sending an AI model corresponding to each AI model index in the one or more AI model indexes, each AI model being used to determine a final positioning result together with the PRS and a weight corresponding to each AI model, and the weight being a weight of a positioning result output by each AI model determined by using the environment parameter combination.

31. The method of claim 30, wherein, The PRS carries one or more of a cell identifier Cell ID, a reference signal received power RSRP and an angle of arrival AoA relative to a base station.

32. The method of claim 31, wherein, The PRS carries a cell identifier Cell ID, and the environmental parameter combination includes a state parameter of the terminal device and a state parameter of a network device. The state parameter of the terminal device includes one or more of a spatial region code, a time state parameter, and a scene state parameter, wherein the spatial region code indicates a grid region in which a cell corresponding to each Cell ID is located, the time state parameter is used to indicate a positioning time, and the scene state parameter is an identifier of a scene in which the terminal device is located. The state parameter of the network device indicates a parameter configuration of the network device.

33. The method of claim 32, wherein, The network device is a base station, and the configuration parameter includes a beam configuration parameter of a downlink PRS signal of the base station.

34. The method of claim 31, wherein, The PRS carries a cell identifier Cell ID, and the state parameter of the terminal device includes a spatial region code, a time state parameter, and a scene state parameter. The method specifically includes: determining a first Cell ID list, wherein the first Cell ID list includes all Cell IDs carried in the PRS; determining an intersection of a second Cell ID list corresponding to each spatial region code and the first Cell ID list, respectively, wherein the second Cell ID list includes all Cell IDs of cells in a grid region corresponding to the spatial region code; taking, as a first screening condition, a spatial region code corresponding to an intersection whose number of Cell IDs is greater than or equal to a first threshold value; when a time interval between a positioning time and a reference value corresponding to a time interval is less than a preset interval, taking, as a second screening condition, a time state parameter corresponding to the time interval, wherein the reference value corresponding to the time interval is a middle value of the time interval; taking, as a third screening condition, a scene state parameter; determining a fourth screening condition according to the state parameter of the network device; 35. The method of claim 31, wherein, and determining one or more AI model indexes that match according to the first screening condition, the second screening condition, the third screening condition, and the fourth screening condition, and a mapping table of environmental parameter combinations and AI model indexes. The PRS carries a cell identifier Cell ID and the RSRP, and the state parameter of the terminal device includes a spatial region code, a time state parameter, and a scene state parameter. The method specifically includes: determining a first Cell ID list, wherein the first Cell ID list includes all Cell IDs carried in the PRS; determining an intersection of a second Cell ID list corresponding to each spatial region code and the first Cell ID list, respectively, wherein the second Cell ID list includes all Cell IDs of cells in a grid region corresponding to the spatial region code; taking, as a first screening condition, a spatial region code corresponding to an intersection that satisfies the following conditions: The number of Cell IDs whose RSRP is greater than or equal to a preset power threshold is greater than a first threshold value; When a time interval between a positioning time and a reference value corresponding to a time interval is less than a preset interval, a time state parameter corresponding to the time interval is taken as a second screening condition, and the reference value corresponding to the time interval is a middle value of the time interval; A scene state parameter in which the network device is located is taken as a third screening condition; A fourth screening condition is determined according to a state parameter of the network device; The one or more AI model indexes that match are determined according to the first screening condition, the second screening condition, the third screening condition, the fourth screening condition, and a mapping table of an environmental parameter combination and an AI model index.

36. The method of claim 31, wherein, The PRS carries a cell identifier Cell ID, the RSRP, and the AoA of the relative base station, and the one or more AI model indexes that match are determined according to the environmental parameter combination and the PRS, specifically including: A first Cell ID list is determined, and the first Cell ID list includes all Cell IDs carried in the PRS; An intersection of a second Cell ID list corresponding to each space region code and the first Cell ID list is respectively determined, and the second Cell ID list includes Cell IDs of all cells in a grid region corresponding to the space region code; A space region code corresponding to an intersection that meets the following condition is taken as a first screening condition: The number of Cell IDs whose RSRP is greater than or equal to a preset power threshold and whose AoA of the relative base station meets a preset angle range is greater than a first threshold value; When a time interval between a positioning time and a reference value corresponding to a time interval is less than a preset interval, a time state parameter corresponding to the time interval is taken as a second screening condition, and the reference value corresponding to the time interval is a middle value of the time interval; A scene state parameter in which the network device is located is taken as a third screening condition; A fourth screening condition is determined according to a state parameter of the network device; The one or more AI model indexes that match are determined according to the first screening condition, the second screening condition, the third screening condition, the fourth screening condition, and a mapping table of an environmental parameter combination and an AI model index.

37. The method of claim 34 or 35 or 36, wherein, The Cell IDs of all cells in the grid region corresponding to the space region code include Cell IDs of all cells in the space region corresponding to the code that can be measured by the positioning assistance data.

38. The method of claim 34 or 35 or 36, wherein, The Cell IDs include a cell global identifier CGI and / or an E-UTRAN cell global identifier code ECI.

39. The method of claim 34 or 35 or 36, wherein, The second Cell ID list also includes physical cell identifiers PCIs and / or absolute radio frequency channel numbers ARFCNs of all cells in the grid region corresponding to the space region code.

40. An AI model based positioning method, comprising: The method includes: receiving first index information, a positioning reference signal (PRS), and an environment parameter combination, the first index information including one or more AI model indexes determined using the PRS and the environment parameter combination, the environment parameter combination including a state parameter of a terminal device and a state parameter of a network device; determining an AI model corresponding to each of the one or more AI model indexes; determining a weight of a positioning result output by each of the AI models using the environment parameter combination; determining a final positioning result using the PRS, the AI models, and the weights corresponding to the AI models; sending position information including the final positioning result.

41. An AI model based positioning method, comprising: The method includes: receiving an environment parameter combination and a positioning reference signal (PRS), the environment parameter combination including a state parameter of a terminal device and a state parameter of a network device; determining one or more AI model indexes matched using the environment parameter combination and the PRS; determining an AI model corresponding to each of the one or more AI model indexes; determining a weight of a positioning result output by each of the AI models using the environment parameter combination; determining a final positioning result using the PRS, the AI models, and the weights corresponding to the AI models; sending position information including the final positioning result.

42. A terminal device, comprising: including a processor and a memory; the processor is coupled to the memory; the memory is configured to store computer programs and / or instructions; the processor is configured to execute the computer programs and / or instructions stored in the memory to implement the AI model-based positioning method according to any one of claims 1 to 28.

43. A communications device, characterized by The apparatus includes a processor configured to execute computer programs or instructions stored in a memory to cause the apparatus to perform the AI model-based positioning method according to any one of claims 29 to 41.

44. A computer program product, characterised in that, The computer program product includes computer programs or instructions for executing the AI model-based positioning method according to any one of claims 1 to 28, or includes computer programs or instructions for executing the AI model-based positioning method according to any one of claims 29 to 41.

45. A computer-readable storage medium, comprising: The computer readable storage medium stores computer programs or instructions that, when executed, perform the AI model-based positioning method according to any one of claims 1 to 41.

Images