Communication method and device

Abstract

The embodiment of the invention provides a communication method and device, and relates to the technical field of communication. The historical positioning data of the first node and the positioning error data at the moment corresponding to the historical positioning data are utilized to train an intelligent model (such as an AI / ML model) used for predicting the positioning error of the navigation system positioning adopted by the first node, so that the positioning information of the navigation system can be combined with the positioning error predicted by the intelligent model; and the positioning information output by the navigation system is compensated based on the positioning error, so that the positioning accuracy and flexibility of the first node are improved, and meanwhile, the model complexity is reduced. The method is especially suitable for a positioning scene of a high-speed mobile carrier, and effectively solves the problems of accuracy and sudden change of model positioning or navigation system positioning in a high-speed scene.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and in particular to a communication method and device. Background Art

[0002] In order to provide positioning-related services to UE (User Equipment), the network needs to locate the UE. The widespread deployment of mobile communication networks has spurred the demand for high-precision wireless positioning of UEs.

[0003] Because positioning methods based on geometric relationships (such as triangulation) are significantly affected by environmental factors, especially in current ultra-dense networking scenarios, they struggle to meet the requirements of high precision, low latency, real-time performance, and dynamic adaptability. Fifth-generation mobile communication technology (5G) introduces artificial intelligence (AI) and machine learning (ML), enabling the application of various AI / ML capabilities to mobile communication networks.

[0004] Therefore, how to use AI / ML models to more accurately locate UEs is an urgent problem that needs to be solved. Summary of the Invention

[0005] The embodiments of the present application provide a communication method and apparatus that can effectively improve the positioning accuracy of a UE.

[0006] In a first aspect, embodiments of the present application provide a communication method, applied to a first node, which may be a user equipment (UE), that receives first information from a second node, which may be a location management function (LMF) network element. The first information includes input data and expected output data for training an intelligent model. The input data and expected output data are determined based on a preset positioning scenario. The input data includes historical positioning data of the first node, and the expected output data includes positioning error data at corresponding moments in the historical positioning data. Based on the first information, an intelligent model is determined, and the intelligent model is used to train the positioning error based on the current positioning data.

[0007] The embodiments of the present application utilize the historical positioning data of the first node and the positioning error data at the corresponding moments of the historical positioning data to train an intelligent model (such as an AI / ML model) for predicting the positioning error of the first node using a navigation system. This model combines the positioning information of the navigation system with the positioning error predicted by the intelligent model and compensates the positioning information output by the navigation system based on the positioning error, thereby improving the accuracy and flexibility of the first node's positioning. This model is particularly suitable for positioning scenarios involving high-speed mobile carriers, effectively addressing the accuracy and variability issues associated with model-based positioning or navigation system-based positioning in high-speed scenarios.

[0008] In one optional implementation, the intelligent model can be a Gated Recurrent Unit (GRU) model. As a variant of recurrent neural networks, the GRU model has only two memory cell gates, resulting in fewer training parameters than other recurrent neural network variants, such as the Long Short-Term Memory (LSTM) model. This reduces computational complexity and significantly improves training efficiency.

[0009] In an optional implementation, the positioning scenario includes a first positioning scenario in which positioning is performed by a combination of multiple navigation systems, the multiple navigation systems including a primary navigation system and a secondary navigation system. The historical positioning data includes first historical positioning data of the primary navigation system, and the positioning error data includes positioning difference data between the multiple navigation systems at the time corresponding to the first historical positioning data. Determining the intelligent model based on the first information can be accomplished by using the first historical positioning data as input and the positioning difference data as output, and training a first mapping relationship between the input data and the output data in the intelligent model. The intelligent model is then determined based on the first mapping relationship.

[0010] Optionally, the main navigation system may include an inertial navigation system (INS), and the auxiliary navigation system may include a global navigation satellite system (GNSS).

[0011] In positioning scenarios where multiple navigation systems are combined, the UE's historical GNSS / INS navigation data is obtained from the LMF. The INS positioning data (i.e., the first historical positioning data) in the historical GNSS / INS navigation data is used as input, and the positioning difference data between the GNSS / INS positioning data and the INS positioning data (i.e., the difference in position navigation results referred to later) is used as the desired output to train the intelligent model. The AI ​​model can also be used to assist in error compensation for UE positioning, enabling correction of INS positioning information in the event of GNSS signal loss (unable to maintain satellite signal lock due to signal obstruction, etc.).

[0012] In an optional implementation, the positioning scenario includes a second positioning scenario performed by a navigation system, the historical positioning data includes second historical positioning data of the navigation system, and the positioning error data includes an error sequence corresponding to the second historical positioning data, where the error sequence is a sequence of positioning errors corresponding to positioning elements at each moment in the second historical positioning data. Determining the intelligent model based on the first information can be accomplished by using the second historical positioning data as input and the error sequence as output, and training a second mapping relationship between the input data and the output data in the intelligent model. The intelligent model is then determined based on the second mapping relationship.

[0013] Optionally, the first information may further include historical positioning true values ​​at each moment, and the error sequence may be determined based on the historical positioning true values.

[0014] For example, the UE can only use INS or other navigation systems for positioning. Considering that the UE may not have the ability of combined navigation in some scenarios, in order to effectively improve the positioning accuracy of the UE in such scenarios. For this positioning scenario, the data for training the intelligent model is the historical positioning data of the UE using the navigation system for positioning as input, and the error sequence corresponding to the historical positioning data is used as the expected output. The trained model can predict the positioning error of a single navigation system, thereby realizing error compensation for the single navigation system and improving positioning accuracy.

[0015] In an optional implementation, the above-mentioned determination of the intelligent model based on the first information can be carried out as follows: the intelligent model is determined based on a preset model training cycle and the first information; wherein the model training cycle is determined based on the time interval step of the historical positioning data and the amount of historical positioning data.

[0016] This embodiment of the present application takes into account the model training cycle and specifies the time range of the data used for model training (i.e., the historical positioning data and error data at corresponding moments in the first information). In combination with the various positioning-related historical data contained in the first information, the model learns this data during this specific training cycle and adjusts weight parameters. This allows the model to fully utilize the time-correlated positioning data and better learn the relationship between error and various input information, thereby improving the model's prediction accuracy and robustness and providing reliable support for subsequent positioning error correction.

[0017] In an optional implementation, the intelligent model may be updated according to a preset model update period and the first information, wherein the model update period is determined according to the motion state of the first node.

[0018] The embodiments of the present application take into account that when the UE is moving at high speed, the model may not be applicable to the current rapidly changing motion state characteristics, and the model prediction accuracy may also be affected as the motion state changes and time passes. By dynamically adjusting the model update cycle according to the error accumulation rate under different motion states, and updating the intelligent model in combination with the model update cycle, the model accuracy can be effectively improved.

[0019] In an optional implementation, before receiving the first information from the second node, a request message may be sent to the second node, requesting the first information from the second node. This allows the UE to promptly obtain the first information used to train the intelligent model when needed, thereby improving positioning accuracy, particularly in complex or high-speed movement scenarios.

[0020] In an optional implementation, before receiving the first information from the second node, second information may also be reported to the second node, where the second information is used to indicate the positioning capability information of the first node, and the second information includes at least one of the following: a model structure that supports model positioning, the data types of input and expected output required for training the model, a model training cycle, and a model update cycle.

[0021] In this way, by reporting the second information, the UE can better understand the specific positioning capabilities and supported technologies of the UE, thereby providing more optimized positioning services for the UE. For example, if the UE supports a specific model positioning technology, the LMF can provide corresponding auxiliary data (such as the first information) for the UE's model training, thereby enhancing the UE's positioning accuracy.

[0022] In a second aspect, embodiments of the present application provide a communication method, applied to a first node, for processing current positioning data according to a preset intelligent model to obtain a positioning error. The intelligent model is determined based on first information sent by a second node. Furthermore, a third message is sent to the second node, the third message including the positioning error, or the third message including positioning information determined based on the positioning error.

[0023] In an optional implementation, the process of processing the current positioning data according to the preset intelligent model may be as follows: when the moving speed of the first node reaches a preset threshold, the current positioning data is processed according to the preset intelligent model.

[0024] In an optional implementation, processing the current positioning data according to a preset intelligent model may be performed as follows: inputting the current positioning data into the intelligent model, processing the current positioning data based on the intelligent model, and obtaining the positioning error at the time corresponding to the current positioning data. The current positioning data includes at least one of the following: an angular velocity measurement value, a specific force measurement value, a calculated velocity navigation solution, and a calculated heading angle navigation solution.

[0025] In an optional implementation, positioning compensation information for current positioning data may be determined based on the positioning error, and positioning information may be determined based on the positioning compensation information and the current positioning data.

[0026] In one optional implementation, the current positioning data is first current positioning data obtained by positioning performed by the primary navigation system in a first positioning scenario, and the positioning compensation information is compensated positioning data that replaces positioning performed by the auxiliary navigation system in the first positioning scenario. Determining the positioning information based on the positioning compensation information and the current positioning data specifically involves fusing the positioning compensation information and the current positioning data using a Kalman filter to obtain the positioning information.

[0027] In one optional implementation, the current positioning data is second current positioning data obtained by positioning by the navigation system in a second positioning scenario, and the positioning compensation information is corrected positioning data that corrects the second current positioning data. The process of determining the positioning information based on the positioning compensation information and the current positioning data can be performed as follows: the current positioning data is corrected based on the positioning compensation information to obtain the positioning information.

[0028] In a third aspect, an embodiment of the present application provides a communication method, which is applied to a second node, by sending first information to a first node, where the first information includes input data and expected output data for training an intelligent model, and the input data and expected output data are determined based on a preset positioning scenario. The input data includes historical positioning data of the first node, and the expected output data includes positioning error data at a corresponding moment of the historical positioning data.

[0029] In an optional implementation, before sending the first information to the first node, the following step may be further included: receiving a request message sent by the first node, where the request message is used to request the first information from the second node.

[0030] In an optional implementation, before sending the first information to the first node, the following steps may also be included: receiving second information reported by the first node, the second information is used to indicate the positioning capability information of the first node, and the second information includes at least one of the following: a model structure that supports model positioning, the data types of input and expected output required for training the model, the model training cycle, and the model update cycle.

[0031] In a fourth aspect, an embodiment of the present application provides a communication device, which may be an electronic device or a chip or chip system within an electronic device. The communication device may include a first transceiver unit and a first processing unit. When the communication device is an electronic device, the first transceiver unit may be a Wi-Fi module, a cellular network module, or the like. The first transceiver unit is configured to perform transceiver steps to enable the electronic device to implement a communication method described in the first aspect or any possible implementation of the first aspect. When the communication device is an electronic device, the first processing unit may be a processor. The communication device may further include a storage unit, which may be a memory. The storage unit is configured to store instructions, and the first processing unit executes the instructions stored in the storage unit to enable the electronic device to implement a communication method described in the first aspect or any possible implementation of the first aspect. When the communication device is a chip or chip system within an electronic device, the first processing unit may be a processor. The first processing unit executes the instructions stored in the storage unit to enable the electronic device to implement a communication method described in the first aspect or any possible implementation of the first aspect. The storage unit may be a storage unit within the chip (eg, a register, a cache, etc.), or a storage unit within the electronic device that is located outside the chip (eg, a read-only memory, a random access memory, etc.).

[0032] Exemplarily, a first transceiver unit is used to receive first information from a second node, where the first information includes input data and expected output data for training an intelligent model, where the input data and the expected output data are determined based on a preset positioning scenario, where the input data includes historical positioning data, and the expected output data includes positioning error data at a corresponding moment of the historical positioning data; and a first processing unit is used to determine the intelligent model based on the first information, where the intelligent model is used to train the positioning error based on the current positioning data.

[0033] In an optional implementation, the intelligent model may be a gated recurrent unit model.

[0034] In an optional implementation, the positioning scenario includes a first positioning scenario in which positioning is performed by a combination of multiple navigation systems, the multiple navigation systems include a main navigation system and a supplementary navigation system, the historical positioning data includes the first historical positioning data of the main navigation system, and the positioning error data includes the positioning difference data between the multiple navigation systems at the corresponding moment of the first historical positioning data; the above-mentioned first processing unit is specifically used to take the first historical positioning data as input and the positioning difference data as output, and train the first mapping relationship between the input data and the output data in the intelligent model; and determine the intelligent model based on the first mapping relationship.

[0035] In an optional implementation, the primary navigation system may include an inertial navigation system, and the secondary navigation system may include a global navigation satellite system.

[0036] In an optional implementation, the positioning scenario includes a second positioning scenario performed by a navigation system, the historical positioning data includes second historical positioning data of the navigation system, and the positioning error data includes an error sequence corresponding to the second historical positioning data, where the error sequence is a sequence consisting of positioning errors corresponding to positioning position elements at each moment in the second historical positioning data. The first processing unit is specifically configured to use the second historical positioning data as input and the error sequence as output, train a second mapping relationship between input data and output data in the intelligent model, and determine the intelligent model based on the second mapping relationship.

[0037] In an optional implementation, the first information may further include historical positioning true values ​​at each moment, and the error sequence is determined based on the historical positioning true values.

[0038] In an optional implementation, the first processing unit is specifically used to determine the intelligent model based on a preset model training cycle and the first information; wherein the model training cycle is determined based on the time interval step of the historical positioning data and the amount of historical positioning data.

[0039] In an optional implementation, the first processing unit is further configured to update the intelligent model according to a preset model update cycle and the first information; wherein the model update cycle is determined based on the motion state of the first node.

[0040] In an optional implementation, the first transceiver unit is further configured to send a request message to the second node after receiving the first information from the second node, where the request message is used to request the first information from the second node.

[0041] In an optional implementation, the first transceiver unit is further used to report second information to the second node before receiving the first information from the second node, where the second information is used to indicate the positioning capability information of the first node, and the second information includes at least one of the following: a model structure that supports model positioning, the data types of input and expected output required for training the model, a model training cycle, and a model update cycle.

[0042] In a fifth aspect, an embodiment of the present application provides another communication device, which may be an electronic device or a chip or chip system within an electronic device. The communication device may include a second transceiver unit and a second processing unit. When the communication device is an electronic device, the second transceiver unit may be a Wi-Fi module or a cellular network module, etc. The second transceiver unit is configured to perform transceiver steps to enable the electronic device to implement a communication method described in the second aspect or any possible implementation of the second aspect. When the communication device is an electronic device, the second processing unit may be a processor. The communication device may also include a storage unit, which may be a memory. The storage unit is configured to store instructions, and the second processing unit executes the instructions stored in the storage unit to enable the electronic device to implement a communication method described in the second aspect or any possible implementation of the second aspect. When the communication device is a chip or chip system within an electronic device, the second processing unit may be a processor. The second processing unit executes the instructions stored in the storage unit to enable the electronic device to implement a communication method described in the second aspect or any possible implementation of the second aspect. The storage unit may be a storage unit within the chip (eg, a register, a cache, etc.), or a storage unit within the electronic device that is located outside the chip (eg, a read-only memory, a random access memory, etc.).

[0043] Exemplarily, the second processing unit is used to process the current positioning data according to a preset intelligent model to obtain a positioning error; wherein the intelligent model is determined based on the first information sent by the second node; the second transceiver unit is used to send third information to the second node, the third information including the positioning error, or the third information including the positioning information determined based on the positioning error.

[0044] In an optional implementation, the second processing unit is specifically configured to process the current positioning data according to a preset intelligent model when the moving speed of the first node reaches a preset threshold.

[0045] In an optional implementation, the second processing unit is specifically configured to input current positioning data into an intelligent model, process the current positioning data based on the intelligent model, and obtain a positioning error at a corresponding moment in the current positioning data. The current positioning data includes at least one of the following: an angular velocity measurement value, a specific force measurement value, a calculated velocity navigation solution, and a calculated heading angle navigation solution.

[0046] In an optional implementation, the second processing unit is further configured to determine positioning compensation information for current positioning data based on the positioning error, and determine positioning information based on the positioning compensation information and the current positioning data.

[0047] In one optional implementation, the current positioning data is first current positioning data obtained by positioning performed by the primary navigation system in a first positioning scenario, and the positioning compensation information is compensated positioning data that replaces the positioning performed by the auxiliary navigation system in the first positioning scenario. The second processing unit is specifically configured to fuse the positioning compensation information and the current positioning data using a Kalman filter to obtain positioning information.

[0048] In one optional implementation, the current positioning data is second current positioning data obtained by positioning by the navigation system in a second positioning scenario, and the positioning compensation information is corrected positioning data used to correct the second current positioning data. The second processing unit is specifically configured to correct the current positioning data based on the positioning compensation information to obtain the positioning information.

[0049] In a sixth aspect, an embodiment of the present application provides another communication device, which can be an electronic device or a chip or chip system within an electronic device. The communication device may include a third transceiver unit. When the communication device is an electronic device, the third transceiver unit may be a Wi-Fi module or a cellular network module, etc. The third transceiver unit is used to perform the steps of transceiving so that the electronic device implements a communication method described in the third aspect or any possible implementation of the third aspect.

[0050] Exemplarily, the third transceiver unit is used to send first information to the first node, where the first information includes input data and expected output data for training an intelligent model, where the input data and the expected output data are determined based on a preset positioning scenario, where the input data includes historical positioning data of the first node, and where the expected output data includes positioning error data at a corresponding moment of the historical positioning data.

[0051] In an optional implementation, the third transceiver unit is further configured to receive a request message sent by the first node before sending the first information to the first node, where the request message is used to request the first information from the second node.

[0052] In an optional implementation, the third transceiver unit is further used to receive second information reported by the first node before sending the first information to the first node, where the second information is used to indicate the positioning capability information of the first node, and the second information includes at least one of the following: a model structure that supports model positioning, the data types of input and expected output required for training the model, a model training cycle, and a model update cycle.

[0053] In the seventh aspect, an embodiment of the present application provides an electronic device, comprising a processor and a memory, the memory being used to store computer-executable instructions, and the processor being used to run the computer-executable instructions stored in the memory to execute the method described in the first aspect or any possible implementation of the first aspect; or, to execute the method described in the second aspect or any possible implementation of the second aspect; or, to execute the method described in the third aspect or any possible implementation of the third aspect.

[0054] In an eighth aspect, an embodiment of the present application provides a computer-readable storage medium, in which a computer program or instruction is stored. When the computer program or instruction is run on a computer, the computer executes the method described in the first aspect or any possible implementation of the first aspect; or, executes the method described in the second aspect or any possible implementation of the second aspect; executes the method described in the third aspect or any possible implementation of the third aspect.

[0055] In the ninth aspect, an embodiment of the present application provides a computer program product including a computer program. When the computer program is run, it enables the computer to execute the method described in the first aspect or any possible implementation of the first aspect; or, execute the method described in the second aspect or any possible implementation of the second aspect; execute the method described in the third aspect or any possible implementation of the third aspect.

[0056] In a tenth aspect, the present application provides a chip system, the chip or chip system including at least one processor and a communication interface, the communication interface and the at least one processor being interconnected by a line, the at least one processor being used to run a computer program or instruction to execute the method described in the first aspect or any possible implementation of the first aspect, or to execute the method described in the second aspect or any possible implementation of the second aspect; to execute the method described in the third aspect or any possible implementation of the third aspect. The communication interface in the chip may be an input / output interface, a pin, or a circuit, etc.

[0057] In one possible implementation, the chip or chip system described above in this application further includes at least one memory, wherein instructions are stored in the at least one memory. The memory may be a storage unit within the chip, such as a register or cache, or a storage unit of the chip (such as a read-only memory or random access memory).

[0058] It should be understood that the second to tenth aspects of this application correspond to the technical solutions of the first aspect of this application, and the beneficial effects achieved by each aspect and the corresponding feasible implementation methods are similar, and the relevant descriptions will not be repeated. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A schematic diagram of the architecture of a possible communication system provided in an embodiment of the present application;

[0060] Figure 2a A schematic diagram of a possible application scenario provided by an embodiment of the present application;

[0061] Figure 2b A schematic diagram of another possible application scenario provided by an embodiment of the present application;

[0062] Figure 3 One of the flow charts of the communication method provided in the embodiment of the present application;

[0063] Figure 4 This is an example diagram of GNSS / INS integrated navigation in an embodiment of the present application;

[0064] Figure 5 This is a structural diagram of the GRU model in the embodiment of the present application;

[0065] Figure 6 This is one of the training process diagrams of the GRU model in the embodiment of this application;

[0066] Figure 7 This is one of the prediction process diagrams of the GRU model in the embodiment of this application;

[0067] Figure 8This is one of the training process diagrams of the GRU model in the embodiment of this application;

[0068] Figure 9 This is the second schematic diagram of the prediction process of the GRU model in the embodiment of the present application;

[0069] Figure 10 This is an example diagram of the training cycle and update cycle in the embodiment of the present application;

[0070] Figure 11 The second flowchart of the communication method provided in the embodiment of the present application;

[0071] Figure 12 Flowchart 3 of the communication method provided in the embodiment of the present application;

[0072] Figure 13 Flowchart 4 of the communication method provided in the embodiment of the present application;

[0073] Figure 14 This is one of the structural diagrams of the communication device provided in the embodiment of the present application;

[0074] Figure 15 The second structural diagram of the communication device provided in the embodiment of the present application;

[0075] Figure 16 The third structural diagram of the communication device provided in the embodiment of the present application;

[0076] Figure 17 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application;

[0077] Figure 18 A schematic diagram of the structure of the chip system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0078] In order to clearly describe the technical solutions of the embodiments of the present application, the following first briefly introduces the relevant technical background and related technologies of the embodiments of the present application:

[0079] First, let's briefly introduce navigation systems. Navigation involves guiding a moving vehicle (hereinafter referred to as a vehicle, such as a user equipment (UE), vehicle, drone, etc.) from its current location to a predetermined target location along a reasonable route. This process requires accurate real-time location information. Navigation and positioning technologies play a crucial role in traditional fields such as navigation, aviation, and surveying and mapping, as well as in emerging fields such as autonomous driving and drones. Navigation and positioning technologies primarily rely on two methods: direct positioning and dead reckoning to determine the vehicle's actual position. Direct positioning uses recognizable external information to directly determine the vehicle's position. The Global Navigation Satellite System (GNSS) is a typical direct positioning system. In addition to satellite positioning using radio signals, there are also laser ranging positioning based on infrared signals and sonar positioning based on acoustic waves. Dead reckoning uses vehicle motion information such as velocity, attitude, and acceleration, combined with motion models, to infer the vehicle's motion state. The Inertial Navigation System (INS) is a typical dead reckoning positioning system.

[0080] Each navigation system has unique advantages. For example, INS has strong autonomy, can provide more navigation parameter data, and has a higher output frequency. However, the shortcomings of INS are also more obvious. Its errors gradually accumulate over time, resulting in low long-term reliability. The advantage of GNSS is its accurate positioning accuracy, and its errors do not accumulate over time. However, its disadvantages are that the navigation parameter information is not comprehensive, and the signal is easily interrupted by interference. In many complex environments, due to the carrier's strict requirements for positioning precision and accuracy, relying solely on a single navigation technology is difficult to achieve the desired effect. Therefore, in some scenarios, the advantages of combined navigation technology that integrates data from multiple navigation systems are becoming increasingly prominent. GNSS / INS combined navigation technology is currently the most widely used combined navigation solution for vehicles and drones.

[0081] The Next Generation Radio Access Network (NG-RAN) can use a variety of positioning methods to determine the UE's location, including: (1) signal measurement. (2) Position estimation based on measurement and optional velocity calculation. These processes can be performed by the UE or the serving base station, and then the positioning information is reported to the Location Management Function (LMF) through different data transmission methods. The positioning methods supported by NG-RAN include GNSS navigation, inertial navigation positioning, WLAN positioning, and Bluetooth positioning. Different positioning methods are applicable to mobile terminals in different scenarios.

[0082] With the widespread deployment of mobile communication networks, the demand for high-precision wireless positioning of carriers such as user equipment (UE) continues to increase. With the application of AI / ML models in UE positioning, the possibility of UE positioning has been proposed. When sending a function message to the LMF, the UE can choose the following positioning service modes: (1) traditional positioning methods based on geometric relationships (such as triangulation); (2) AI / ML direct positioning; (3) AI / ML assisted positioning.

[0083] However, traditional positioning methods based on geometric relationships (such as triangulation) are significantly affected by environmental factors in non-line-of-sight environments, complex urban areas, and densely populated networks, making them difficult to meet the requirements of high precision, low latency, real-time performance, and dynamic adaptability. Current AI / ML positioning algorithms primarily rely on historical positioning data to train models to predict UE location information for positioning. The positioning accuracy of these models is highly dependent on historical data, resulting in high computational complexity and limited flexibility. Furthermore, when the UE moves slowly, the training and prediction errors of the AI ​​model are small. However, in high-speed movement scenarios, the UE's position changes, causing a sharp decline in AI prediction performance.

[0084] Therefore, current UE positioning technology faces the following problems:

[0085] (1) Traditional positioning methods based on geometric relationships have limited accuracy in non-line-of-sight environments, complex urban areas, and ultra-dense networking environments.

[0086] (2) Current AI / ML models rely on a large amount of positioning data for training, and the cost of obtaining high-quality data is high. Furthermore, the model computational complexity is high, which limits its application on resource-constrained terminals and results in poor flexibility.

[0087] (3) Current AI / ML models are difficult to adapt to high-speed mobile scenarios such as UE.

[0088] Therefore, there is an urgent need to study AI / ML-based assisted positioning or direct positioning algorithms to improve the accuracy and robustness of UE positioning technology.

[0089] In light of this, embodiments of the present application propose a communication method and apparatus in which a first node (a mobile carrier, such as a UE) receives first information from a second node (a node carrying the mobile carrier's historical positioning data, such as an LMF network element). The first information includes input data and expected output data for training an intelligent model. The input data and expected output data are determined based on a preset positioning scenario. The input data includes the first node's historical positioning data, and the expected output data includes positioning error data at the time corresponding to the historical positioning data. Based on the first information, an intelligent model is determined, which is used to train the positioning error based on the current positioning data. In this process, the first node's historical positioning data and the positioning error data at the time corresponding to the historical positioning data are used to train an intelligent model (such as an AI / ML model) for predicting the positioning error of the first node using a navigation system. This method combines the positioning information of the navigation system with the positioning error predicted by the intelligent model, and compensates the positioning information output by the navigation system based on the positioning error, thereby improving the accuracy and flexibility of the first node's positioning. This method is particularly suitable for positioning high-speed mobile carriers, effectively addressing the accuracy and variability issues associated with model-based positioning or navigation system-based positioning in high-speed scenarios. In addition, compared with the related art of training models to predict location information, the embodiments of the present application focus more on error characteristics, have lower model complexity, and the data used to train the model is also more streamlined, reducing excessive reliance on large amounts of historical positioning data.

[0090] Below, the embodiments of this application will use a variety of currently commonly used navigation and positioning technologies as solution embodiments to explain in detail the specific implementation details of the AI / ML assisted positioning mode.

[0091] First, it should be noted that in the embodiments of this application, terms such as "first" and "second" are used to distinguish between identical or similar items with substantially the same functions and effects. For example, the terms "first chip" and "second chip" are used solely to distinguish between different chips and do not limit their order. Those skilled in the art will understand that terms such as "first" and "second" do not limit the number or execution order, and that terms such as "first" and "second" do not necessarily define differences.

[0092] It should be noted that in the embodiments of this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described in this application as "exemplary" or "for example" should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0093] In the embodiments of the present application, "at least one" refers to one or more, and "more" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: the existence of A alone, the existence of A and B at the same time, and the existence of B alone, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, a--c, bc, or abc, where a, b, c can be single or plural.

[0094] In this application, "sending information / data" refers solely to the direction of information / data transmission, including direct transmission via a device's communication interface (e.g., an air interface). "Sending" can also be understood as "output" from a module interface. "Sending" can also include indirect transmission by a processing unit via a communication interface, i.e., after the processing unit outputs information / data via a module interface, it is then transmitted to the device's communication interface and then transmitted. "Receiving information / data" refers solely to the direction of information / data transmission, including direct reception via a communication interface. "Receiving" can also be understood as "input" from a module interface. "Receiving information / data" can also include indirect reception by a processing unit via a communication interface, i.e., after the communication interface receives information / data, it is then transmitted to the processing unit's module interface and then input into the processing unit. "Sending information / data to... (e.g., a terminal)" can be understood as the destination of the information being the terminal. This can include direct or indirect transmission of information / data to the terminal. "Receiving information / data from... (e.g., a terminal)" can be understood as the source of the information being the terminal. This can include direct or indirect reception of information / data from the terminal. Information / data may be processed between the source and destination of the information / data, such as format changes, but the destination can understand the valid information / data from the source. Similar expressions in this application can be understood similarly and will not be repeated here.

[0095] The technical solutions of the embodiments of this application can be applied to various communication systems, such as long-term evolution (LTE) systems, fifth-generation (5G) communication systems, satellite communication systems, and wireless fidelity (WiFi) systems. The solutions provided in this application can also be applied to future communication systems or other communication systems. This application does not limit this.

[0096] Figure 1It is a schematic diagram of the architecture of a communication system suitable for the communication method provided in this application. Figure 1 A schematic diagram of a possible, non-limiting system architecture is shown. Figure 1 As shown, the communication system includes a radio access network (RAN) 10 and a core network (CN) 20. Optionally, the communication system also includes the Internet 30. The RAN 10 includes at least one RAN node (e.g. Figure 1 110a and 110b, collectively referred to as 110) and at least one terminal (such as Figure 1 120a-120j in the figure are collectively referred to as 120). The RAN 10 may also include other RAN nodes, such as wireless relay equipment and / or wireless backhaul equipment ( Figure 1 Terminal 120 is wirelessly connected to RAN node 110. RAN node 110 is wirelessly or wiredly connected to core network 20. The core network equipment in core network 20 and RAN node 110 in RAN 10 can be different physical devices, or they can be the same physical device that integrates core network logical functions and radio access network logical functions.

[0097] The RAN 10 may be a cellular system related to the Third Generation Partnership Project (3GPP), such as a 4G or 5G mobile communication system, or a future-oriented evolutionary system. The RAN 10 may also be an open access network (O-RAN or ORAN), a cloud radio access network (CRAN), or a wireless fidelity (Wi-Fi) system. The RAN 10 may also be a communication system that integrates two or more of the above systems.

[0098] RAN node 110, sometimes also called access network equipment, RAN entity or access node, is part of the communication system and is used to help terminals achieve wireless access. Multiple RAN nodes 110 in the communication system can be nodes of the same type or different types. In some scenarios, the roles of RAN node 110 and terminal 120 are relative, for example, Figure 1 The network element 120i can be a helicopter or a drone, which can be configured as a mobile base station. For the terminals 120j that access the RAN 10 through the network element 120i, the network element 120i is a base station; but for the base station 110a, the network element 120i is a terminal. The RAN node 110 and the terminal 120 are sometimes referred to as communication devices, for example Figure 1The network elements 110a and 110b may be understood as communication devices having base station functions, and the network elements 120a-120j may be understood as communication devices having terminal functions.

[0099] In one possible scenario, a RAN node may be a base station, an evolved NodeB (eNodeB), an access point (AP), a transmission reception point (TRP), a next generation NodeB (gNB), a next generation base station in a 6G mobile communication system, a base station in a future mobile communication system, or an access node in a WiFi system. A RAN node may be a macro base station (e.g. Figure 1 110a in), micro base stations or indoor stations (such as Figure 1 RAN nodes can be 110b in the example above), relay nodes or donor nodes, or wireless controllers in CRAN scenarios. Alternatively, RAN nodes can be servers, wearable devices, vehicles, or onboard devices. For example, the access network device in vehicle-to-everything (V2X) technology can be a roadside unit (RSU).

[0100] In another possible scenario, multiple RAN nodes collaborate to assist terminals in achieving wireless access, with different RAN nodes implementing portions of the base station's functionality. For example, a RAN node can be a centralized unit (CU), a distributed unit (DU), a CU-control plane (CP), a CU-user plane (UP), or a radio unit (RU). The CU and DU can be separate or included in the same network element, such as a baseband unit (BBU). The RU can be included in a radio frequency device or radio unit, such as a remote radio unit (RRU), an active antenna unit (AAU), or a remote radio head (RRH).

[0101] In different systems, CU (or CU-CP and CU-UP), DU or RU may have different names, but those skilled in the art will understand their meanings. For example, in the ORAN system, CU may also be called O-CU (Open Central Unit, open CU, responsible for processing control plane protocols, including managing packet data convergence protocol (PDCP), service data adaptation protocol (SDAP) and radio resource control (RRC) protocol entities), DU may also be called (Open Distributed Unit, O-DU, with baseband processing functions, complete protocol layer functions, mainly responsible for high-level protocol functions such as data encryption and integrity protection. It also has physical layer high-level processing functions), CU-CP may also be called O-CU-CP, CU-UP may also be called O-CU-UP, and RU may also be called (Open Radio Unit, O-RU, with physical layer underlying signal processing functions, mainly responsible for sending and receiving radio frequency signals). For convenience of description, this application uses CU, CU-CP, CU-UP, DU and RU as examples for description. Any of the CU (or CU-CP, CU-UP), DU and RU in this application may be implemented by a software module, a hardware module, or a combination of a software module and a hardware module.

[0102] Terminals can also be referred to as terminal devices, user equipment (UE), mobile stations, or mobile terminals. They are widely used in various scenarios, such as device-to-device (D2D), vehicle-to-everything (V2X) communication, machine-type communication (MTC), the Internet of Things (IoT), virtual reality, augmented reality, industrial control, autonomous driving, telemedicine, smart grids, smart furniture, smart offices, smart wearables, smart transportation, and smart cities. Terminals can be mobile phones, tablets, computers with wireless transceiver capabilities, wearable devices, vehicles, drones, helicopters, airplanes, ships, robots, robotic arms, and smart home appliances.

[0103] In the embodiments of the present application, the terminal and the network device may be hardware devices, or software functions running on dedicated hardware, or software functions running on general-purpose hardware, such as virtualization functions instantiated on a platform (e.g., a cloud platform), or entities including dedicated or general-purpose hardware devices and software functions. The present application does not limit the specific forms of the terminal and the network device.

[0104] The following is a further introduction to possible application scenarios of the embodiments of the present application, taking UE positioning as an example. Figure 2a and Figure 2b shown. Figure 2a It shows that the UE has a combined navigation capability. For example, the UE has a GNSS navigation system and an INS navigation system and can use CNSS / INS combined navigation positioning. The UE can receive input data and expected output data sent by the LMF for training the intelligent model. The input data and expected output data can be the historical positioning data of the UE using CNSS / INS combined navigation positioning (such as INS historical positioning data obtained by differential GNSS / INS) and corresponding positioning error data (such as the positioning difference data between the GNSS / INS positioning result and the INS positioning result) to train an intelligent model that can be used to predict the positioning error. In particular, when the UE is in a high-speed motion state, the GNSS positioning accuracy is no longer accurate due to the Doppler effect, the signal being blocked by buildings, bridges, etc. in an urban environment, or the multipath effect (signal reflection). The UE can use the trained intelligent model to combine the INS navigation data to predict the positioning error data corresponding to the GNSS, and use the positioning error data to determine the positioning compensation information, thereby achieving positioning compensation for the INS navigation data and solving the problem of poor positioning accuracy when the UE is in high-speed motion. Figure 2b The UE uses a navigation system, such as an INS navigation system, for navigation and positioning. Figure 2a The difference between the illustrated scenarios is that the input data and expected output data used by the UE to train the intelligent model can be the historical positioning data of the INS, as well as the positioning error sequence corresponding to the historical positioning data. This effectively improves the positioning accuracy of the UE in a navigation system scenario and improves positioning accuracy in high-speed mobile scenarios. It should be understood that in some application scenarios, the UE can also obtain or store a trained intelligent model from other devices or equipment without training the intelligent model, and can directly apply the intelligent model to predict the positioning error.

[0105] Combined with the above possible application scenarios, Figure 3 This is one of the flow charts of the communication method proposed in the embodiment of the present application. The first node can be any terminal or a device configured in the terminal (such as a chip or module, etc.). For example, the first node can be Figure 2a or Figure 2b Shows the UE in the application scenario or the device configured in the UE (such as a chip or module, etc.), which can also be Figure 1 Any mobile device or device (such as a chip or module, etc.) configured in the mobile device in the terminal 120 shown. The second node can be Figure 2a or Figure 2b The LMF shown can also be Figure 1 Any core network device shown or a device configured in the core network device (such as a chip or module, etc.). The following takes the first node as a UE and the second node as an LMF as an example to illustrate the embodiment of the present application. Figure 3 As shown, the method may include step S301 and step S302.

[0106] Step S301: The UE receives first information from the LMF. The first information includes input data and expected output data for training the intelligent model. The input data and expected output data are determined according to a preset positioning scenario. The input data includes historical positioning data of the first node, and the expected output data includes positioning error data at a corresponding moment of the historical positioning data.

[0107] The Location Management Function (LMF) is a key component in the communications network architecture, responsible for managing and providing services related to the location of user equipment (UE). As the centralized location management entity in the network, the LMF collects and integrates positioning data from multiple sources, including base station signal measurements and time synchronization information. This centralized data management ensures data integrity and reliability, providing UEs with comprehensive and accurate historical positioning data. This data is network-verified and highly reliable, helping to improve model training results. Furthermore, by obtaining data from the LMF, the UE can align its positioning model with the network's overall positioning strategy, improving the quality and consistency of positioning services.

[0108] Optionally, the LMF may send the first information to the UE via an LTE Positioning Protocol (LPP) Provide Assistance Information message. The first information may be provided by the LMF to the UE upon a request from the UE, or may be provided proactively by the LMF to the UE, which is not particularly limited in this embodiment.

[0109] In this embodiment, the positioning scenario may include scenarios corresponding to multiple navigation systems, or may also include scenarios corresponding to one navigation system. The following are two example methods provided in this embodiment.

[0110] Example 1: The positioning scenario may include a first positioning scenario in which positioning is performed by a combination of multiple navigation systems, where the multiple navigation systems include a main navigation system and a supplementary navigation system. The historical positioning data includes the first historical positioning data of the main navigation system, and the positioning error data includes the positioning difference data between the multiple navigation systems at the corresponding moment of the first historical positioning data.

[0111] In this example, the primary navigation system may be an inertial navigation system (INS), and the auxiliary navigation system may be a global navigation satellite system (GNSS). In some examples, the primary navigation system may also be a GNSS, and the auxiliary navigation system may be an INS. Alternatively, the primary navigation system and the auxiliary navigation system may each be another navigation system, which is not specifically limited in this embodiment.

[0112] Combined with the above, INS can obtain the position, velocity and attitude information of the carrier (UE in this embodiment) by integrating the measurement data of the accelerometer and gyroscope. However, data drift, measurement error and error accumulation over time may cause the position and velocity information obtained during navigation to diverge rapidly. Therefore, GNSS auxiliary positioning information can be used to compensate for the positioning error of INS. The compensation principle is as follows: Figure 4 As shown in the figure, the Inertial Measurement Unit (IMU), consisting of accelerometers and gyroscopes, measures the vehicle's acceleration and angular velocity, among other measurements. The INS then interprets the IMU input, integrating the accelerometer and gyroscope measurements to obtain the vehicle's position, velocity, and attitude. GNSS determines position and other information by receiving satellite signals, providing highly accurate auxiliary positioning. A Kalman Filter (KF) receives the position, velocity, and attitude information derived from the INS data, along with the auxiliary positioning information provided by the GNSS. By fusing these two sets of information, the INS performs error correction and outputs corrected positioning information. This corrected positioning significantly improves accuracy compared to information provided by the INS alone, effectively addressing the problem of positioning divergence caused by INS error accumulation. The KF filter is an optimal estimation algorithm based on a state-space model that accurately estimates and corrects system states in the presence of noise and uncertainty.

[0113] However, GNSS has improved the accuracy of INS positioning to a certain extent. However, the GNSS system is easily affected by the observation environment, and the satellite signal is easily affected by obstructions such as high buildings and tree shades, which causes interference to the satellite signal received by the GNSS. Even in tunnels and other places with severe obstruction, the GNSS signal is completely interrupted. At this time, the combined navigation system degenerates into a single INS for positioning, and the positioning accuracy and precision will also decrease. Therefore, in this positioning scenario, the present application obtains the GNSS / INS historical navigation data of the UE from the LMF, and uses the INS positioning data (i.e., the first historical positioning data) in the GNSS / INS historical navigation data as input, and the positioning difference data between the GNSS / INS positioning data and the INS positioning data (i.e., the difference in the position navigation result in the following text) is used. ) as the desired output for subsequent intelligent model training (such as AI / ML models). The AI ​​model can also be used to assist in UE positioning error compensation, enabling correction of INS positioning information in the GNSS signal loss state (due to signal obstruction, etc., unable to maintain satellite signal lock).

[0114] It should be noted that in various other navigation system scenarios, as the accuracy of one of the navigation systems decreases, the overall positioning accuracy will also decrease. This embodiment only uses GNSS / INS as an example to illustrate the technical solution of this application, and does not limit the technical solution of this application.

[0115] Example 2: The positioning scenario includes a second positioning scenario positioned by a navigation system, the historical positioning data includes second historical positioning data of the navigation system, and the positioning error data includes an error sequence corresponding to the second historical positioning data, which is a sequence composed of positioning errors corresponding to the positioning position elements at each moment in the second historical positioning data.

[0116] In this example, the UE can use only INS or other navigation systems for positioning. Compared with the combination positioning of multiple navigation systems, the positioning accuracy of a single navigation system will be relatively lower. Considering that the UE may not have the ability of combined navigation in some scenarios, in order to effectively improve the positioning accuracy of the UE in such scenarios. This embodiment can also improve the positioning accuracy of the UE for the positioning scenario of a navigation system. Similar to the combined navigation scenario, the data used to train the intelligent model is the historical positioning data of the UE using the navigation system for positioning as input, and the error sequence corresponding to the historical positioning data is used as the expected output to train the intelligent model to predict the positioning error of a single navigation system, thereby achieving error compensation and improving positioning accuracy.

[0117] Step S302: The UE determines an intelligent model based on the first information, where the intelligent model is used to train the positioning error based on the current positioning data.

[0118] Optionally, the intelligent model may adopt a gated recurrent unit (GRU) model.

[0119] Considering that the error of the navigation system is usually correlated with time, especially for INS navigation, since the error of IMU is correlated with time, the neural network model can be used to train and predict the error change of INS and provide error compensation for the INS positioning solution.

[0120] Recurrent Neural Networks (RNNs) are a type of neural network model that excels at processing sequential data. Unlike traditional feedforward neural networks, RNNs incorporate feedback loops, enabling them to capture temporal dependencies in sequential data. However, RNNs are prone to the vanishing gradient problem when processing long-term data, which in turn affects model prediction accuracy. Consequently, two variants of the RNN model have emerged to address this issue. Both the Long Short-Term Memory (LSTM) and the Gated Recurrent Unit (GRU) incorporate gating mechanisms to address the vanishing gradient problem in RNN models. The difference between the two is that the LSTM incorporates three gating mechanisms: an input gate, a forget gate, and an output gate, which control the input, forget, and output of information, respectively. It also incorporates memory cells to store key information. However, since the three gates of the LSTM contribute differently to predictive accuracy, omitting gates with smaller contributions and their corresponding weights can improve computational speed and reduce computational space when computing power and storage resources are limited. Therefore, the GRU model, based on the LSTM model, optimizes these three gates into two: a reset gate and an update gate. The update gate implements the input gate in the LSTM model, which is used to receive data input from the outside and process the data. The update gate implements the forget gate and output gate in the LSTM model. The internal structure of the GRU model is as follows: Figure 5 As shown, the mathematical formula is described as follows:

[0121] (1)

[0122] (2)

[0123] (3)

[0124] (4)

[0125] Where Z t , r t are update gate and reset gate respectively, X t represents the input vector at the current time t, h t-1 is the memory variable (hidden state) at the previous moment t-1; W and U are the corresponding weight parameters. Different subscripts indicate the weights used in different calculation processes, such as W z 、U z Calculation for updating gate, W r 、U r Calculations used to reset the gate, etc. t is the candidate set, h tThe hidden state at the current time t is the final output state, integrating information from the previous state and the candidate hidden state. σ represents the sigmoid activation function. ∘ represents element-wise multiplication (Hadamard product), which multiplies corresponding elements of two vectors of the same dimension. The above formula works as follows: First, the input data and the memory variable from the previous time are permuted and combined according to the formula. The sigmoid function transforms the input data into a value between (0, 1) and inputs it to the update gate and reset gate. The reset gate, input data, and past state variable are then linearly transformed and combined. Finally, the information from these two is added together to produce the output at the current time.

[0126] Since the GRU model has only two memory unit gates, the training parameters of the GRU model are reduced compared to the LSTM model, the computational complexity is lower, and the training efficiency can be greatly improved.

[0127] Next, the training process of the intelligent model is further introduced for different positioning scenarios.

[0128] In the first positioning scenario corresponding to the combined positioning of multiple navigation systems, the above-mentioned determination of the intelligent model based on the first information can be carried out in the following manner: taking the first historical positioning data as input and the positioning difference data as output, training the first mapping relationship between the input data and the output data in the intelligent model; and determining the intelligent model based on the first mapping relationship.

[0129] For example, taking the GRU model as an example, combined with Figure 6 As shown, when the GNSS signal is not interrupted, the differential GNSS navigation information (denoted as P GNSS ) and INS mechanical arrangement (i.e. the output of the inertial measurement unit IMU, denoted as P INS ) The information obtained is sent to the Kalman filter to solve the positioning information, and then the INS is corrected (to obtain the corrected positioning information, including the position P GNSS / INS , can also include speed V GNSS / INS and acceleration A GNSS / INS ). In addition, the specific force f output by the IMU b , angular velocity ω b (Optionally, it can also include the speed information V calculated by INS INS And heading angle information (also known as attitude information) A INS) as the input matrix of the GRU model. Specifically, since the integrated navigation system degenerates into a single INS positioning system after the GNSS signal is lost, the model input is the INS output result. INS is a navigation system that uses an inertial measurement unit (such as an accelerometer and a gyroscope) to measure and solve the motion state. Therefore, the input of the GRU network can be the fusion result of the IMU measurement output and the INS navigation positioning result output (in some embodiments, only the position data solved by the INS can be used as input, which is not particularly limited in this embodiment). INS position solution generally includes carrier attitude solution, velocity ratio force solution and position solution. Optionally, in this embodiment, the GRU model input feature vector X t It can be:

[0130] (5)

[0131] Where, ω bT represents the three-axis angular velocity measurement value (i.e. angular velocity) from the gyroscope, f bT represents the triaxial specific force measurement (i.e., specific force) from the accelerometer, v nT represents the 3D velocity navigation solution (i.e., velocity information) derived from the INS mechanical arrangement, sinφ and cosφ represent the heading navigation solution (i.e., heading angle information), respectively. It should be understood that T stands for transpose.

[0132] Accordingly, the difference δP between the GNSS and INS position navigation results at the corresponding moment (optionally, it can also include the difference in velocity information δv and the difference in heading angle information δA) can be used as the desired output of GRU model training, establishing a mapping relationship between input and target output. It is understood that the data used for model training can be historical positioning data obtained from the LMF when the GNSS signal is not interrupted.

[0133] Combine Figure 7 As shown in the figure, when the GNSS signal is interrupted, the GNSS will be unable to provide corresponding navigation information. At this time, the UE can use the previously trained GRU model to predict the error difference at the corresponding time. This error difference is used to calculate the pseudo-GNSS position information (also known as positioning compensation information). This pseudo-GNSS position information and the INS solution information are then fed into the Kalman filter to suppress the error divergence problem of the INS and thus improve the UE's positioning accuracy.

[0134] As mentioned above, INS and other similar navigation systems exhibit time-dependent error dispersion. Intelligent models, such as the GRU model, a variant of recurrent neural networks, excel at processing data with time-series characteristics. Accordingly, the GRU model can also be used to estimate the error of a single navigation system, such as the INS, to correct for inherent flaws.

[0135] Accordingly, in a second positioning scenario corresponding to positioning using a navigation system, determining the intelligent model based on the first information can be performed as follows: using the second historical positioning data as input and the error sequence as output, training a second mapping relationship between the input data and the output data in the intelligent model, and determining the intelligent model based on the second mapping relationship.

[0136] Optionally, the first information may also include historical positioning true values ​​at each moment, and the error sequence is determined based on the historical positioning true values. For example, by comparing the positioning position element information at each moment with the corresponding historical positioning true value, an error sequence corresponding to a time series over a predefined time length (persons skilled in the art may adapt this predefined time length based on actual applications or experience, for example, it may be the time length corresponding to the training cycle or update cycle mentioned later).

[0137] The INS output result can be the IMU measurement result and the INS solution result (or only the INS solution result). The input feature matrix of the GRU model can be similar to the above-mentioned multiple navigation combinations, and can be the fusion result of the IMU measurement output and the INS navigation positioning result output, as shown in formula (5).

[0138] Compared to the positioning scenario described above involving multiple navigation systems, in this example, the intelligent model aims to correct errors for a single navigation system, such as an INS. Therefore, the GRU model in this embodiment expects the output to be an INS error sequence. This error sequence can be the difference between the corresponding position elements of the INS output and the actual positioning result, used to derive the INS inverse solution process and correct the compensation values ​​used during the INS solution. The GRU expected output matrix can be simply defined as follows.

[0139] (6)

[0140] Where Y t represents the expected output matrix, Corresponding to ω bT 、f bT 、v nT , sinφ and cosφ at the corresponding time, that is, the compensation value in the INS solution process. Optionally, for the UE positioning scenario using a navigation system, the GRU model training and prediction process can be Figure 8 and Figure 9 As shown. Among them, ΔP(t), ΔV(t), ΔA(t) are the position error, velocity error and heading angle error corresponding to time t. GRU / INS ,V GRU / INS , AGRU / INS are the positioning information after the positioning error is compensated by the model. In some embodiments, only ΔP(t), P GRU / INS This application does not impose any special limitation on this.

[0141] It should be understood that in the model, the mapping relationship refers to the relationship between input data and target outputs learned by the GRU model through training. This relationship can be represented in the model as a set of optimized parameters and weights that define how to convert input data into output predictions. During the training process, the model can continuously adjust its internal parameters and weights through the backpropagation algorithm to learn the mapping relationship between input data and output errors.

[0142] Through the above technical solution, the positioning accuracy compensation when using a single navigation system for positioning can be effectively improved, and the model complexity can be reduced.

[0143] Optionally, considering that the model needs to be trained to obtain optimal weight parameter information to achieve higher prediction accuracy, and that the navigation system's error is correlated with time, especially for INS navigation systems, where the error gradually accumulates over time, this embodiment sets a training cycle during the intelligent model training process to improve the model's robustness. Specifically, the above step S302 determines the intelligent model based on the first information, and the following methods can be used:

[0144] An intelligent model is determined according to a preset model training cycle and the first information; wherein the model training cycle is determined based on the time interval step of the historical positioning data and the amount of the historical positioning data.

[0145] For example, the training period T train The time interval step length τ of the historical positioning data and the number N of historical positioning data are determined by the calculation formula: . Among them, the time interval step τ represents the time interval between two adjacent historical positioning data; the number of historical positioning data N refers to the number of historical data used for model training. That is, the training cycle can be determined by the number of historical data used for training and the time interval of these data. For example, if there are 100 historical positioning data and the time interval step of adjacent data is 0.1 seconds, then the training cycle can be T train =100×0.1=10 seconds.

[0146] When training the intelligent model to determine the model, the aforementioned model training cycle is taken into account, specifying the time range of the data used for model training (i.e., the historical positioning data and the error data at the corresponding time in the first information). The model learns from the various historical positioning-related data contained in the first information during this specific training cycle and adjusts weight parameters. This allows the model to fully utilize the time-correlated positioning data and better learn the relationship between error and various input information, thereby improving the model's prediction accuracy and robustness and providing reliable support for subsequent positioning error correction.

[0147] Furthermore, considering that the model is not applicable to the current motion state characteristics when the UE is moving at high speed, and that the model prediction accuracy may be affected as the motion state changes and over time, this embodiment sets a model update period to further improve model accuracy. Specifically, the solution may also include the following steps: updating the intelligent model based on a preset model update period and the first information. The model update period is determined based on the motion state of the UE.

[0148] Exemplarily, the motion state of the UE can be identified and classified into different categories, such as stationary, low-speed movement, medium-speed movement, and high-speed movement (for example, different motion states can be identified by analyzing sensor data, such as accelerometer, gyroscope, GPS speed information, etc.). By analyzing the error accumulation characteristics under each motion state, generally, the error accumulation is faster in the high-speed movement state, while the error accumulation is slower in the stationary or low-speed state. The model update cycle is dynamically adjusted according to the error accumulation rate under different motion states. For example, in the high-speed movement state, a shorter update cycle is set to update the model more timely to adapt to the rapidly changing motion state. In the stationary or low-speed movement state, since the error accumulation is slower, the prediction accuracy of the model is less affected, and the update cycle can be extended. It can be understood that the first information in this embodiment can be real-time, for example, under the corresponding update cycle or training cycle conditions, the first information is obtained from the LMF in real time to update or train the model.

[0149] Taking the GNSS / INS combined positioning scenario as an example, the training cycle and update cycle of the intelligent model can be as follows: Figure 10 As shown, on the T time axis, it contains the training cycle T train and update period T update , where the training cycle Ttrain is a continuous time interval on the time axis T, which represents the time range covered by the data used for GRU model training, and the update cycle T update It means the time interval for retraining the trained GRU model, for example, every certain time (such as T updateThe model is retrained every 1 hour. At the moment of GNSS lock loss, i.e., time t, the trained or updated intelligent model can be used to predict positioning errors, thereby improving the UE's positioning accuracy.

[0150] In related art, a GNSS / INS integrated navigation solution utilizes online learning and compensation using a neural network model. During GNSS / INS integrated navigation, when GNSS is normal, a first neural network and a second neural network are alternately trained within a set period. This cycle ensures that one of the first and second neural networks is always in a training state and the other is in a trained standby state. The trained neural network is used to predict pseudo-GNSS signals based on input INS data. When GNSS is lost during navigation, the standby neural network is connected to the integrated navigation architecture, and INS data is input to the connected neural network. The pseudo-GNSS signals predicted by the neural network are then compensated for the original INS data using a Kalman filter to output a positioning result. This technical solution requires alternating training of two neural networks, resulting in a complex and inflexible model. Furthermore, the model predicts the actual positioning position. Furthermore, due to the limited computing and storage resources of the UE itself, training two neural networks can lead to insufficient UE computing resources. This embodiment, however, utilizes a training positioning error approach. When GNSS is lost, traditional positioning methods and intelligent model-assisted positioning can be used to optimize positioning accuracy and precision. This allows for predictions tailored to the user's current environment, offering enhanced real-time and flexibility. Furthermore, this implementation takes into account the limited resources of UEs, requiring only the training or updating of a single intelligent model, saving UE resources. Focusing on positioning error training, the model is simpler and less data-dependent. Furthermore, by setting model training and update cycles, the effective updating and dynamic adaptability of the AI / ML model ensures the continuous and efficient operation of positioning services and enables rapid response to environmental changes, enhancing the effectiveness of the model's positioning error predictions and further improving UE positioning accuracy.

[0151] Figure 11 This is the second flow chart of the communication method provided in the embodiment of the present application, including steps S1100-S1104. Figure 3 Corresponding to the embodiment, this embodiment further includes steps S1100 to S1101 before receiving the first information from the second node.

[0152] Step S1102: The UE sends a request message to the LMF, where the request message is used to request the first information from the second node.

[0153] For example, when the UE has a positioning requirement, for example, when the UE needs to enter a high-speed mobile scenario and needs a higher-precision positioning service, the UE may send a request message to the LMF to request the LMF to send the first information. For example, the request message may be a message from the UE to the LMF requesting positioning assistance information.

[0154] In this way, the UE can obtain the first information for training the intelligent model in a timely manner when needed, thereby improving positioning accuracy, especially in complex or high-speed moving scenarios.

[0155] In some examples, the LMF may also actively send the first information to the UE, which is not particularly limited in this embodiment.

[0156] Step S1103: The UE receives first information from the LMF, where the first information includes input data and expected output data for training the intelligent model. The input data and expected output data are determined based on a preset positioning scenario. The input data includes historical positioning data of the first node, and the expected output data includes positioning error data at a corresponding moment of the historical positioning data.

[0157] Step S1104: The UE determines an intelligent model based on the first information, where the intelligent model is used to train the positioning error based on the current positioning data.

[0158] It should be noted that step S1103 and step S1104 correspond to Figure 3 Step S301 and step S302 have similar principles. For relevant descriptions, please refer to the above text and will not be elaborated here.

[0159] Furthermore, before receiving the first information from the second node, the following step S1101 may be included:

[0160] Step S1101: The UE reports the second information to the LMF, where the second information is used to indicate the positioning capability information of the UE. The second information includes at least one of the following: a model structure supporting model positioning, data types of input and expected output required for training the model, a model training cycle, and a model update cycle.

[0161] Exemplarily, the second information may be LPP Provide Capabilities information.

[0162] In this embodiment, by reporting the second information, the LMF can better understand the specific positioning capabilities and supported technologies of the UE, thereby providing more optimized positioning services for the UE. For example, if the UE supports a specific model positioning technology, the LMF can provide corresponding auxiliary data (such as the first information) for the UE's model training, thereby enhancing the UE's positioning accuracy.

[0163] Optionally, before the UE reports the second information to the LMF, the LMF may request the second information from the UE. That is, the method may further include step S1100, where the LMF requests the second information from the UE, and the request may be sent via positioning request capabilities (LPP RequestCapabilities) information.

[0164] For example, in the intelligent model (AI / ML model) assisted positioning mode provided in this embodiment, the LMF requests information about the UE's currently supported positioning capabilities from the UE. When the UE reports this positioning capability information to the LMF, if the UE currently supports AI / ML direct positioning or assisted positioning, the UE may include in this message the AI / ML model used, the model input and output content (i.e., the data types required for model training and expected output), the model training cycle, the model update cycle, and the model structure. Based on the positioning capability information reported by the UE, the LMF provides positioning assistance information to the UE. This positioning assistance information can be requested by the UE or sent unsolicited by the LMF. If the UE supports AI / ML positioning and initiates an assistance information request, the assistance information provided by the LMF may include, but is not limited to, AI / ML model training input and expected output data, historical positioning accuracy values ​​(i.e., historical positioning truth values), and other information. After sending the assistance information, the LMF sends a positioning request message to the UE. After the UE uses AI / ML assisted positioning, it reports the positioning results to the LMF.

[0165] The above technical solution adopts the auxiliary positioning method of intelligent model, which can be applied to scenarios with high UE movement speed such as drones or vehicles. It can compensate for the positioning error problem of traditional positioning methods (such as INS positioning), effectively improve the positioning accuracy, and at the same time reduce the model complexity and increase the flexibility.

[0166] Figure 12 This is the third flow chart of the communication method provided in the embodiment of the present application, and continues to illustrate by taking the first node as UE and the second node as LMF as an example.

[0167] Step S1201: The UE processes current positioning data according to a preset intelligent model to obtain a positioning error, wherein the intelligent model is determined based on first information sent by the second node.

[0168] Optionally, the intelligent model can be trained by the UE, and its training process can be found in the above embodiment. In some embodiments, the intelligent model can also be obtained through real-time interaction with other devices or servers, or pre-stored in the UE, and is not necessarily trained by the UE. Accordingly, the model training process in the above embodiment can also be trained for other UEs or devices, and transmitted to other UEs or devices through interaction to perform model positioning. This embodiment does not specifically limit whether the model training subject and the application subject are the same subject.

[0169] It should be understood that when the model training subject and the model application subject are not the same subject, the relevant data used to train the model, such as historical positioning data, is relevant data for the model application subject.

[0170] Step S1202: The UE sends third information to the LMF, where the third information includes a positioning error, or the third information includes positioning information determined based on the positioning error.

[0171] Optionally, the UE may report the third information to the LMF after the LMF sends the positioning request.

[0172] In this embodiment, the UE uses an intelligent (AI / ML) model for assisted positioning. Specifically, in AI / ML-assisted positioning mode, the positioning message reported by the UE to the LMF may include two protocol messages. In one case, the UE reports the output of the AI / ML model and traditional positioning results (such as INS positioning results) separately for processing by the LMF. In this case, the third message may include both the positioning error and the INS positioning result. When the UE selects this reporting type, it may include the output of the AI / ML model (such as error compensation information or pseudo-observation information) in the report to facilitate data processing by the LMF. In another case, the UE autonomously combines the output of the AI / ML-assisted positioning with the traditional positioning results and reports the final positioning result to the LMF. In this case, the third message includes positioning information determined using the positioning error.

[0173] This approach, combined with the positioning error prediction of the AI / ML model, allows the UE to more accurately process positioning data, thereby improving overall positioning accuracy. Secondly, the UE has the flexibility to report the output of the AI / ML model and traditional positioning results separately, or to combine the two and report the final positioning result. By flexibly configuring the application scenarios of the intelligent model and the reporting format of location information, the advantages of the intelligent model in different positioning scenarios are fully utilized.

[0174] Optionally, considering that the accuracy of traditional positioning methods is more affected in high-speed UE movement scenarios, while the accuracy impact is smaller in non-high-speed movement scenarios, AI / ML models can be used for auxiliary positioning in high-speed movement scenarios to save UE computing resources. Specifically, the above process of processing the current positioning data according to the preset intelligent model can be carried out as follows: when the movement speed of the first node reaches a preset threshold, the current positioning data is processed according to the preset intelligent model.

[0175] It should be noted that those skilled in the art may adaptively set the preset threshold value based on actual applications or experience, and this embodiment does not specifically limit this.

[0176] In this way, by using AI / ML models for assisted positioning in high-speed mobile scenarios, positioning accuracy can be effectively improved while optimizing the use of computing resources.

[0177] Optionally, the processing of the current positioning data according to the preset intelligent model can be performed as follows: the current positioning data is input into the intelligent model, and the current positioning data is processed based on the intelligent model to obtain the positioning error at the time corresponding to the current positioning data. The current positioning data includes at least one of the following: an angular velocity measurement value, a specific force measurement value, a calculated velocity navigation solution, and a calculated heading angle navigation solution.

[0178] It should be noted that the model processing process is similar to the model training process, and can be combined with the above embodiment Figure 8 Please understand the corresponding content and no further explanation will be given here.

[0179] Optionally, when the UE reports the fused positioning information, the UE may also determine positioning compensation information for current positioning data based on the positioning error, and determine positioning information based on the positioning compensation information and the current positioning data.

[0180] For example, using GNSS / INS combined positioning as an example, the positioning error output by the model corresponds to the error caused by GNSS signal interruption. The AI / ML model is used to predict the error difference at the corresponding moment, and this error difference is used to calculate the pseudo-GNSS position information (for example, the predicted error difference is used to correct the current position calculated by the INS. Specifically, the error difference is added to the current position calculation result of the INS to generate the pseudo-GNSS position information. This process is equivalent to simulating the position information provided by the GNSS system when the GNSS signal is not interrupted, that is, the positioning compensation information). Then, by feeding the pseudo-GNSS position information and the INS solution information into the Kalman filter, the error divergence problem of the INS is suppressed.

[0181] In an optional implementation of this example, the current positioning data is first current positioning data obtained by positioning performed by the primary navigation system in a first positioning scenario, and the positioning compensation information is compensated positioning data that replaces positioning performed by the auxiliary navigation system in the first positioning scenario. Determining the positioning information based on the positioning compensation information and the current positioning data specifically involves fusing the positioning compensation information and the current positioning data using a Kalman filter to obtain the positioning information.

[0182] It should be noted that the model processing process is similar to the model training process, and can be understood in conjunction with the corresponding content in the above embodiments. The relevant explanation will not be repeated here.

[0183] In another optional implementation of this example, the current positioning data is second current positioning data obtained by positioning by the navigation system in a second positioning scenario, and the positioning compensation information is corrected positioning data that corrects the second current positioning data. The above process of determining positioning information based on the positioning compensation information and the current positioning data can be carried out as follows: the current positioning data is corrected based on the positioning compensation information to obtain positioning information.

[0184] Exemplarily, the positioning compensation information may be superimposed on the current positioning data to obtain positioning information.

[0185] It should be noted that the model processing process is similar to the model training process, and can be understood in conjunction with the corresponding content in the above embodiments. The relevant explanation will not be repeated here.

[0186] Accordingly, the embodiment of the present application also provides a communication method, Figure 3 In an example, the LMF sends a first message to the UE, where the first message includes input data and expected output data for training the intelligent model. The input data and the expected output data are determined according to a preset positioning scenario. The input data includes the historical positioning data of the first node, and the expected output data includes the positioning error data at the corresponding moment of the historical positioning data.

[0187] In an optional implementation, before sending the first information to the first node, the LMF may also receive a request message sent by the first node, where the request message is used to request the first information from the second node.

[0188] In an optional implementation, before sending the first information to the first node, the LMF may also receive second information reported by the first node, where the second information is used to indicate the positioning capability information of the first node, and the second information includes at least one of the following: a model structure that supports model positioning, the data types of input and expected output required for training the model, a model training cycle, and a model update cycle.

[0189] It should be noted that the above embodiment is a counterpart embodiment of the UE. For related descriptions and beneficial effects, please refer to the contents of the embodiment on the UE side, which will not be elaborated here.

[0190] Figure 13 This is the fourth flow chart of the communication method provided by the embodiment of the present application. In this embodiment, the UE can be used as the subject of model training and model application, and realize the UE positioning and uploading of positioning information through interaction with the LMF. Figure 13 As shown, the method may include the following process:

[0191] Step S1301: The LMF sends an LPP request to the UE to request second information of the UE, that is, positioning capability related information.

[0192] Step S1302: After receiving the request from the LMF, the UE provides the second information to the LMF, that is, the UE's own positioning capability information, such as the model structure supported by the UE for model positioning, the data types of input and expected output required for training the model, the model training cycle, and the model update cycle, etc.

[0193] Step S1303: LMF may choose to send the first information, i.e., positioning assistance information, to the UE proactively or based on the UE's request. The information may be sent without the UE sending a request message to assist the UE in model training and positioning.

[0194] Step S1304: The LMF sends a positioning request to the UE to obtain the UE's location information.

[0195] Step S1305: The UE trains an intelligent model using the first information fed back by the LMF, and uses the intelligent model, such as a time series model such as GRU, to train the error change of the navigation system (such as INS) to predict the positioning error used to compensate for the navigation system.

[0196] Step S1306: The UE reports the third information to the LMF. The reporting content corresponding to the third information may include the positioning measurement information of the navigation system and the positioning error predicted by the intelligent model (here is the GRU model). Alternatively, the reporting content may be the UE autonomously integrating the intelligent model-assisted positioning output result with the traditional positioning result, and reporting the final positioning result to the LMF so that the LMF can more accurately determine the UE's position.

[0197] Figure 14 This is a schematic diagram of the structure of a communication device provided in an embodiment of the present application. The communication device may be an electronic device or a chip or chip system in an electronic device. Figure 14As shown, the communication device 1400 may include a first transceiver unit 1401 and a first processing unit 1402. When the communication device is an electronic device, the first transceiver unit 1401 may be a Wi-Fi module or a cellular network module, etc. The first transceiver unit 1401 is used to perform the transceiver steps so that the electronic device implements the communication method described in the above method embodiment. When the communication device is an electronic device, the first processing unit 1402 may be a processor. The communication device may also include a storage unit, which may be a memory. The storage unit is used to store instructions, and the first processing unit 1402 executes the instructions stored in the storage unit so that the electronic device implements the communication method described in the above method embodiment. When the communication device is a chip or chip system in an electronic device, the first processing unit 1402 may be a processor. The first processing unit 1402 executes the instructions stored in the storage unit so that the electronic device implements a communication method described in the first aspect or any possible implementation of the first aspect. The storage unit may be a storage unit within the chip (eg, a register, a cache, etc.), or a storage unit within the electronic device that is located outside the chip (eg, a read-only memory, a random access memory, etc.).

[0198] Exemplarily, the first transceiver unit 1401 is used to receive first information from the second node, where the first information includes input data and expected output data for training an intelligent model, where the input data and the expected output data are determined based on a preset positioning scenario, where the input data includes historical positioning data, and the expected output data includes positioning error data at a corresponding moment of the historical positioning data; the first processing unit 1402 is used to determine the intelligent model based on the first information, where the intelligent model is used to train the positioning error based on the current positioning data.

[0199] In an optional implementation, the intelligent model may be a gated recurrent unit model.

[0200] In an optional implementation, the positioning scenario includes a first positioning scenario in which positioning is performed by a combination of multiple navigation systems, the multiple navigation systems include a main navigation system and a supplementary navigation system, the historical positioning data includes the first historical positioning data of the main navigation system, and the positioning error data includes the positioning difference data between the multiple navigation systems at the corresponding moment of the first historical positioning data; the above-mentioned first processing unit 1402 is specifically used to take the first historical positioning data as input and the positioning difference data as output, and train the first mapping relationship between the input data and the output data in the intelligent model; and determine the intelligent model based on the first mapping relationship.

[0201] In an optional implementation, the primary navigation system may include an inertial navigation system, and the secondary navigation system may include a global navigation satellite system.

[0202] In an optional implementation, the positioning scenario includes a second positioning scenario in which positioning is performed by a navigation system, the historical positioning data includes second historical positioning data of the navigation system, and the positioning error data includes an error sequence corresponding to the second historical positioning data, where the error sequence is a sequence consisting of positioning errors corresponding to positioning position elements at each moment in the second historical positioning data. The first processing unit 1402 is specifically configured to use the second historical positioning data as input and the error sequence as output, train a second mapping relationship between input data and output data in the intelligent model, and determine the intelligent model based on the second mapping relationship.

[0203] In an optional implementation, the first information may further include historical positioning true values ​​at each moment, and the error sequence is determined based on the historical positioning true values.

[0204] In an optional implementation, the first processing unit 1402 is specifically used to determine the intelligent model based on a preset model training cycle and the first information; wherein the model training cycle is determined based on the time interval step of the historical positioning data and the amount of historical positioning data.

[0205] In an optional implementation, the first processing unit 1402 is further configured to update the intelligent model according to a preset model update period and the first information; wherein the model update period is determined according to the motion state of the first node.

[0206] In an optional implementation, the first transceiver unit 1401 is further configured to send a request message to the second node after receiving the first information from the second node, where the request message is used to request the first information from the second node.

[0207] In an optional implementation, the first transceiver unit 1401 is also used to report second information to the second node before receiving the first information from the second node, where the second information is used to indicate the positioning capability information of the first node, and the second information includes at least one of the following: a model structure that supports model positioning, the data types of input and expected output required for training the model, a model training cycle, and a model update cycle.

[0208] It should be noted that the device in the above embodiment can execute the steps corresponding to the UE side (model training) method embodiment. For relevant instructions and beneficial effects, please refer to the content of the above method embodiment, and no further details will be given here.

[0209] Figure 15 This is a structural diagram of another communication device provided by an embodiment of the present application. The communication device may be an electronic device or a chip or chip system in an electronic device. Figure 15As shown, the communication device 1500 may include a second transceiver unit 1501 and a second processing unit 1502. When the communication device is an electronic device, the second transceiver unit 1501 may be a Wi-Fi module, a cellular network module, or the like. The second transceiver unit 1501 is configured to perform transceiver steps to enable the electronic device to implement a communication method described in the second aspect or any possible implementation of the second aspect. When the communication device is an electronic device, the second processing unit 1502 may be a processor. The communication device may also include a storage unit, which may be a memory. The storage unit is configured to store instructions, and the second processing unit 1502 executes the instructions stored in the storage unit to enable the electronic device to implement the communication method described in the method embodiment. When the communication device is a chip or chip system within an electronic device, the second processing unit 1502 may be a processor. The second processing unit 1502 executes the instructions stored in the storage unit to enable the electronic device to implement a communication method described in the second aspect or any possible implementation of the second aspect. The storage unit may be a storage unit within the chip (eg, a register, a cache, etc.), or a storage unit within the electronic device that is located outside the chip (eg, a read-only memory, a random access memory, etc.).

[0210] Exemplarily, the second processing unit 1502 is used to process the current positioning data according to a preset intelligent model to obtain a positioning error; wherein the intelligent model is determined based on the first information sent by the second node; the second transceiver unit 1501 is used to send third information to the second node, and the third information includes the positioning error, or the third information includes positioning information determined based on the positioning error.

[0211] In an optional implementation, the second processing unit 1502 is specifically configured to process the current positioning data according to a preset intelligent model when the moving speed of the first node reaches a preset threshold.

[0212] In an optional implementation, the second processing unit 1502 is specifically configured to input the current positioning data into the intelligent model, process the current positioning data based on the intelligent model, and obtain the positioning error at the time corresponding to the current positioning data. The current positioning data includes at least one of the following: an angular velocity measurement value, a specific force measurement value, a calculated velocity navigation solution, and a calculated heading angle navigation solution.

[0213] In an optional implementation, the second processing unit 1502 is further configured to determine positioning compensation information for current positioning data based on the positioning error, and determine positioning information based on the positioning compensation information and the current positioning data.

[0214] In one optional implementation, the current positioning data is first current positioning data obtained by positioning performed by the primary navigation system in a first positioning scenario, and the positioning compensation information is compensated positioning data that replaces positioning performed by the auxiliary navigation system in the first positioning scenario. Second processing unit 1502 is specifically configured to fuse the positioning compensation information and the current positioning data using a Kalman filter to obtain positioning information.

[0215] In an optional implementation, the current positioning data is second current positioning data obtained by positioning by the navigation system in a second positioning scenario, and the positioning compensation information is corrected positioning data used to correct the second current positioning data. The second processing unit 1502 is specifically configured to correct the current positioning data based on the positioning compensation information to obtain positioning information.

[0216] It should be noted that the device in the above embodiment can execute the steps corresponding to the UE side (model application) method embodiment. For relevant instructions and beneficial effects, please refer to the content of the above method embodiment, and no further details will be given here.

[0217] Figure 16 This is a structural diagram of another communication device provided in an embodiment of the present application. The communication device can be an electronic device or a chip or chip system in an electronic device. Figure 16 As shown, the communication device 1600 may include a third transceiver unit 1601. When the communication device is an electronic device, the third transceiver unit 1601 may be a Wi-Fi module or a cellular network module, etc. The third transceiver unit 1601 is used to perform the steps of transceiving, so that the electronic device implements the communication method provided by the method embodiment.

[0218] Exemplarily, the third transceiver unit 1601 is used to send first information to the first node, where the first information includes input data and expected output data for training an intelligent model, where the input data and the expected output data are determined based on a preset positioning scenario, where the input data includes historical positioning data of the first node, and the expected output data includes positioning error data at a corresponding moment of the historical positioning data.

[0219] In an optional implementation, the third transceiver unit 1601 is further configured to receive a request message sent by the first node before sending the first information to the first node, where the request message is used to request the first information from the second node.

[0220] In an optional implementation, the third transceiver unit 1601 is also used to receive second information reported by the first node before sending the first information to the first node, where the second information is used to indicate the positioning capability information of the first node, and the second information includes at least one of the following: a model structure that supports model positioning, the data types of input and expected output required for training the model, a model training cycle, and a model update cycle.

[0221] It should be noted that the device in the above embodiment can execute the steps corresponding to the LMF side method embodiment. Relevant descriptions and beneficial effects can be found in the content of the above method embodiment, and will not be elaborated here.

[0222] Figure 17 This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. The above method embodiment can be applied to electronic devices with communication functions. Figure 17 As shown, the electronic device 1700 can be applied to Figure 1 In the system shown, the functions of the terminal or network device in the above method embodiment are performed. Figure 17 As shown, the electronic device 1700 includes a processor 1701 and a transceiver 1702. Optionally, the electronic device 1700 also includes a memory 1703. The processor 1701, the transceiver 1702, and the memory 1703 can communicate with each other via an internal connection path to transmit control and / or data signals. The memory 1703 is used to store computer programs, and the processor 1701 is used to call and execute the computer programs from the memory 1703 to control the transceiver 1702 to transmit and receive signals. Optionally, the electronic device 1700 may also include an antenna 1704 for transmitting uplink data or uplink control signaling output by the transceiver 1702 via wireless signals.

[0223] The processor 1701 and the memory 1703 may be combined into a processing device, and the processor 1701 is configured to execute program code stored in the memory 1703 to implement the aforementioned functions. In a specific implementation, the memory 1703 may also be integrated into the processor 1701 or independent of the processor 1701. The processor 1701 may correspond to the processing unit or processor in the aforementioned embodiments.

[0224] The transceiver 1702 may correspond to the transceiver unit in the above embodiment. The transceiver 1702 may include a receiver (or receiver, receiving circuit) and a transmitter (or transmitter, transmitting circuit). The receiver is used to receive signals, and the transmitter is used to transmit signals.

[0225] It should be understood that Figure 17The electronic device 1700 shown is capable of implementing the various processes related to the terminal or network device in the above-described method embodiments. The operations and / or functions of the various modules in the electronic device 1700 are respectively for implementing the corresponding processes in the above-described method embodiments. For details, please refer to the description of the above-described method embodiments. To avoid repetition, detailed descriptions are omitted here.

[0226] The processor 1701 can be used to execute the actions described in the previous method embodiments, which are implemented within the terminal or network device, and the transceiver 1702 can be used to execute the actions described in the previous method embodiments, such as the network device sending to the terminal or the terminal receiving from the network device. For details, please refer to the description in the previous method embodiments, which will not be repeated here.

[0227] Optionally, the electronic device 1700 may further include a power supply 1705 for providing power to various devices or circuits in the terminal.

[0228] In addition, in order to make the functions of the terminal more complete, the electronic device 1700, when being a terminal, may also include one or more of an input unit 1706, a display unit 1707, an audio circuit 1708, a camera 1709 and a sensor 1710, and the audio circuit may also include a speaker 1708a, a microphone 1708b, etc.

[0229] The above method embodiments can be applied to or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method embodiment can be completed by hardware integrated logic circuits in the processor or by software instructions.

[0230] The processors described above may be general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or any combination thereof. A general-purpose processor may be a microprocessor or any conventional processor.

[0231] The steps of the method disclosed in the embodiments of this application can be directly implemented and executed by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium well-known in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in the memory, and the processor reads the information in the memory and, in conjunction with its hardware, completes the steps of the above method.

[0232] The memory in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. The non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM RAM (DR RAM). It should be noted that the memory of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0233] Figure 18 This is a chip system provided by an embodiment of the present application, such as Figure 18 As shown, the chip system 1800 includes at least one processor 1801 and a communication interface 1802. The communication interface 1802 and the at least one processor 1801 are interconnected via a line. The at least one processor 1801 is used to run computer programs or instructions to support the implementation of the functions of the terminal or network device involved in any of the above method embodiments, such as sending, receiving, or processing the information involved in the above method. The communication interface in the chip system can be an input / output interface, a pin, or a circuit.

[0234] In one possible design, the chip system further includes a memory 1803, which is used to store computer program instructions and data. The memory 1803 may be a storage unit within the chip, such as a register or cache, or a storage unit of the chip (e.g., a read-only memory or a random access memory). In some embodiments, the memory 1803 may also be an external memory.

[0235] The chip system can be composed of chips, or can include chips and other discrete devices.

[0236] In a possible implementation, the chip or chip system described above in this application further includes at least one memory, in which instructions are stored.

[0237] An embodiment of the present application also provides a computer program product, which includes: a computer program (also referred to as code, or instructions). When the computer program is run, the method executed by the terminal in the above method embodiment is executed, or the method executed by the network device is executed.

[0238] The present application also provides a computer-readable storage medium that stores a computer program (also referred to as code or instructions). When the computer program is executed, the method executed by the terminal in the above-described method embodiment is executed, or the method executed by the network device is executed.

[0239] The present application also provides a communication system, which includes the aforementioned terminal and network equipment.

[0240] The methods provided in the above embodiments can be implemented in whole or in part through software, hardware, firmware, or any combination thereof. When implemented using software, they can be implemented in whole or in part in the form of a computer program product. The computer program product may include one or more computer instructions. When loaded and executed on a computer, the computer program instructions fully or partially generate the processes or functions described in the embodiments of the present application. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium accessible by a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium may be magnetic media (e.g., floppy disk, hard disk, magnetic disk), optical media (e.g., DVD), or semiconductor media (e.g., solid-state drive (SSD)).

[0241] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0242] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0243] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the unit is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0244] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of the solution of this embodiment according to actual needs.

[0245] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0246] If this function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application can essentially or partially be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method of each embodiment of this application. The aforementioned storage medium includes: USB flash drives, mobile hard drives, read-only memories, random access memories, magnetic disks, optical disks, and other media that can store program code.

[0247] In the various embodiments of the present application, unless otherwise specified or there is a logical conflict, the terms and / or descriptions between different embodiments are consistent and can be referenced by each other. The technical features in different embodiments can be combined to form new embodiments according to their inherent logical relationships.

[0248] In one possible implementation, computer-readable media may include RAM, ROM, compact disc read-only memory (CD-ROM) or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium designed to carry or store the desired program code in the form of instructions or data structures and accessible by a computer. Furthermore, any connection is appropriately termed a computer-readable medium. For example, if software is transmitted from a website, server, or other remote source using a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. Disk and disc, as used herein, include optical disc, laser disc, optical disc, digital versatile disc (DVD), floppy disk, and Blu-ray disc, where disks typically reproduce data magnetically, while discs reproduce data optically using lasers. Combinations of the above are also intended to be included within the scope of computer-readable media.

[0249] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processing unit of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable device to produce a machine, so that the instructions executed by the processing unit of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0250] The above specific implementation methods further explain in detail the purpose, technical solutions and beneficial effects of the embodiments of the present application. It should be understood that the above are only specific implementation methods of the embodiments of the present application and are not intended to limit the scope of protection of the embodiments of the present application. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solutions of the embodiments of the present application should be included in the scope of protection of the embodiments of the present application.

Claims

1. A communication method, characterized in that: Applied to the first node, including: receiving first information from a second node, the first information including input data and expected output data for training an intelligent model, the input data and the expected output data being determined based on a preset positioning scenario, the input data including historical positioning data of the first node, and the expected output data including positioning error data at a corresponding moment of the historical positioning data; The intelligent model is determined according to the first information, and the intelligent model is used to train the positioning error according to the current positioning data.

2. The method according to claim 1, characterized in that The positioning scenario includes a first positioning scenario in which positioning is performed by a combination of multiple navigation systems, the multiple navigation systems include a main navigation system and a guided navigation system, the historical positioning data includes first historical positioning data of the main navigation system, and the positioning error data includes positioning difference data between the multiple navigation systems at a time corresponding to the first historical positioning data; The step of determining the intelligent model according to the first information includes: Taking the first historical positioning data as input and the positioning difference data as output, training a first mapping relationship between the input data and the output data in the intelligent model; The intelligent model is determined according to the first mapping relationship.

3. The method according to claim 2, characterized in that The main navigation system includes an inertial navigation system, and the auxiliary navigation system includes a global navigation satellite system.

4. The method according to claim 1, wherein The positioning scenario includes a second positioning scenario performed by a navigation system, the historical positioning data includes second historical positioning data of the navigation system, and the positioning error data includes an error sequence corresponding to the second historical positioning data, wherein the error sequence is a sequence composed of positioning errors corresponding to positioning position elements at each moment in the second historical positioning data; The step of determining the intelligent model according to the first information includes: Using the second historical positioning data as input and the error sequence as output, training a second mapping relationship between input data and output data in the intelligent model; The intelligent model is determined according to the second mapping relationship.

5. The method according to claim 4, characterized in that The first information also includes historical positioning true values ​​at each moment, and the error sequence is determined based on the historical positioning true values.

6. The method according to any one of claims 1 to 5, characterized in that The step of determining the intelligent model according to the first information includes: The intelligent model is determined according to a preset model training cycle and the first information; wherein the model training cycle is determined based on the time interval step of the historical positioning data and the amount of the historical positioning data.

7. The method according to any one of claims 1 to 5, characterized in that Also includes: The intelligent model is updated according to a preset model update cycle and the first information; wherein the model update cycle is determined according to the motion state of the first node.

8. The method according to any one of claims 1 to 5, characterized in that Before receiving the first information from the second node, the method further includes: A request message is sent to the second node, where the request message is used to request the first information from the second node.

9. The method according to any one of claims 1 to 5, characterized in that Before receiving the first information from the second node, the method further includes: Report second information to the second node, where the second information is used to indicate the positioning capability information of the first node, and the second information includes at least one of the following: a model structure that supports model positioning, the data types of input and expected output required for training the model, the model training cycle, and the model update cycle.

10. The method according to any one of claims 1 to 5, characterized in that The intelligent model includes a gated recurrent unit model.

11. A communication method, characterized in that: Applied to the first node, the method includes: Processing the current positioning data according to a preset intelligent model to obtain a positioning error; wherein the intelligent model is determined based on the first information sent by the second node; Sending third information to the second node, where the third information includes the positioning error, or the third information includes positioning information determined based on the positioning error.

12. The method according to claim 11, characterized in that The processing of the current positioning data according to the preset intelligent model includes: When the moving speed of the first node reaches a preset threshold, the current positioning data is processed according to a preset intelligent model.

13. The method according to claim 11 or 12, characterized in that The processing of the current positioning data according to the preset intelligent model includes: Inputting the current positioning data into the intelligent model, processing the current positioning data based on the intelligent model, and obtaining the positioning error at the corresponding moment of the current positioning data; The current positioning data includes at least one of the following: an angular velocity measurement value, a specific force measurement value, a velocity navigation solution obtained through calculation, and a heading angle navigation solution.

14. The method according to claim 11 or 12, characterized in that Also includes: determining positioning compensation information for the current positioning data based on the positioning error; The positioning information is determined according to the positioning compensation information and the current positioning data.

15. The method according to claim 14, characterized in that The current positioning data is first current positioning data obtained by positioning by the main navigation system in a first positioning scenario, and the positioning compensation information is compensation positioning data used to replace the positioning by the auxiliary navigation system in the first positioning scenario; The determining the positioning information according to the positioning compensation information and the current positioning data includes: The positioning compensation information and the current positioning data are fused through a Kalman filter to obtain the positioning information.

16. The method according to claim 14, characterized in that The current positioning data is second current positioning data obtained by positioning by the navigation system in a second positioning scenario, and the positioning compensation information is corrected positioning data correcting the second current positioning data; The determining the positioning information according to the positioning compensation information and the current positioning data includes: According to the positioning compensation information, the current positioning data is corrected to obtain the positioning information.

17. A communication method, characterized in that: Applied to the second node, including: First information is sent to a first node, where the first information includes input data and expected output data for training an intelligent model, where the input data and the expected output data are determined according to a preset positioning scenario, the input data includes historical positioning data of the first node, and the expected output data includes positioning error data at a corresponding moment of the historical positioning data.

18. The method according to claim 17, characterized in that Before sending the first information to the first node, the method further includes: A request message sent by the first node is received, where the request message is used to request the first information from the second node.

19. The method according to claim 17 or 18, characterized in that Before sending the first information to the first node, the method further includes: Receive second information reported by the first node, where the second information is used to indicate positioning capability information of the first node, and the second information includes at least one of the following: a model structure supporting model positioning, data types of input and expected output required for training the model, a model training cycle, and a model update cycle.

20. A communication device, characterized in that: include: a first transceiver unit, configured to receive first information from a second node, the first information including input data and expected output data for training an intelligent model, the input data and the expected output data being determined based on a preset positioning scenario, the input data including historical positioning data, and the expected output data including positioning error data at a corresponding moment of the historical positioning data; The first processing unit is used to determine the intelligent model according to the first information, and the intelligent model is used to train the positioning error according to the current positioning data.

21. A communication device, characterized in that: include: a second processing unit, configured to process the current positioning data according to a preset intelligent model to obtain a positioning error; wherein the intelligent model is determined based on the first information sent by the second node; The second transceiver unit is configured to send third information to the second node, where the third information includes the positioning error, or the third information includes positioning information determined based on the positioning error.

22. A communication device, characterized in that: include: The third transceiver unit is used to send first information to the first node, where the first information includes input data and expected output data for training the intelligent model, where the input data and the expected output data are determined according to a preset positioning scenario, where the input data includes historical positioning data of the first node, and the expected output data includes positioning error data at a corresponding moment of the historical positioning data.

23. An electronic device, characterized in that: include: processor and memory; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory, so that the electronic device performs the communication method according to any one of claims 1 to 19.

24. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the communication method according to any one of claims 1 to 19 is implemented.

25. A chip system, characterized in that: The system comprises at least one processor and a communication interface, wherein the communication interface and the at least one processor are interconnected via a line, and the at least one processor is used to run a computer program or instruction to execute the communication method according to any one of claims 1 to 19.

26. A computer program product, characterized in that The invention comprises a computer program, which, when being executed, causes a computer to execute the communication method according to any one of claims 1 to 19.

Images