A communication method, apparatus, medium, and product

Abstract

The embodiment of the application provides a communication method, device, medium and product, aiming at improving the positioning accuracy of a network device, the method comprising: sending first information, the first information comprising a time of arrival, an angle of arrival and a first identifier; sending second information, the second information indicating that a second network device generates a target execution action corresponding to the first identifier; receiving third information, the third information indicating the target execution action, the target execution action being determined by the second network device based on a confidence level of a predicted position and a corresponding relationship between the confidence level and the execution action; the confidence level of the predicted position being determined based on the time of arrival, the angle of arrival and a double-flow neural network, so that the first network device receives, from the second network device, the execution action determined by the second network device based on the corresponding relationship between the confidence level and the execution action, so that the first network device can execute the corresponding action in combination with the actual situation, realize dynamic adjustment of positioning resources, and improve the positioning accuracy.

Description

Technical Field

[0001] This application relates to the field of wireless communication technology, and in particular to a communication method, apparatus, medium and product. Background Technology

[0002] Integrated sensing and communication (ISAC) achieves target localization by fully utilizing the temporal and spatial resources of wireless communication and radar sensing. In one possible embodiment, to ensure positioning accuracy, the network device adjusts positioning resources at fixed intervals to avoid positioning errors caused by long-term operation. However, using fixed-period adjustments to positioning resources can result in adjustments being made too early or too late, and the adjustment method may not be suitable for the current network device, leading to low positioning accuracy. Therefore, improving positioning accuracy has become a key technical problem that needs to be solved. Summary of the Invention

[0003] This application provides a communication method, apparatus, medium, and product, with the aim of solving the problem of low positioning accuracy in the positioning function of network devices.

[0004] To achieve the above objectives, this application provides the following technical solution:

[0005] The first aspect of this application provides a communication method, which may be executed by a network device, or by a component (such as a circuit, core network unit, chip, or chip system) configured in the network device, or by a logic module or software capable of implementing all or part of the functions of the network device. This application does not limit this approach. The following description uses a first network device as an example, and the method includes:

[0006] Send first information, the first information including arrival time, arrival angle and first identifier; the arrival time and arrival angle are the arrival time and arrival angle of the signal sent from the predicted location received by the first network device; the predicted location is the position of the target terminal predicted by the first network device after a preset time; the first identifier is the identifier of the arrival time and the arrival angle.

[0007] Send a second message, which instructs a second network device to generate a target execution action corresponding to the first identifier;

[0008] The third information is received, which instructs the target to perform an action. The target's action is determined by the second network device based on the confidence level of the predicted location and the correspondence between the confidence level and the action. The confidence level of the predicted location is determined based on the arrival time, the arrival angle, and a two-stream neural network.

[0009] In the above scheme, by establishing a correspondence between confidence level and execution action, and enabling the second network device to determine the corresponding execution action based on the correspondence between confidence level and execution action, the execution action is sent to the first network device. This allows the first network device to execute the corresponding action according to the actual situation, thereby realizing dynamic adjustment of positioning resources and ensuring positioning accuracy.

[0010] In some possible implementations, the output of the two-stream neural network is a heteroscedastic covariance matrix, which is used to map the confidence level of the predicted location.

[0011] In the above scheme, the confidence level of the predicted position is mapped by the heteroscedasticity covariance matrix output by the dual-stream neural network to determine the accuracy of the current predicted position. This facilitates the subsequent determination of the corresponding target action based on the accuracy, so as to accurately perform dynamic adjustments to the first network device.

[0012] In some possible implementations, the two-stream neural network includes a first encoder and a second encoder;

[0013] The first encoder is used to take the arrival time as input and output a Gaussian parameter corresponding to the arrival time;

[0014] The second encoder is used to take the arrival angle as input and output Gaussian parameters corresponding to the arrival angle; the Gaussian parameters corresponding to the arrival time and the Gaussian parameters corresponding to the arrival angle are used to determine the heteroscedasticity covariance matrix.

[0015] In the above scheme, the arrival time and arrival angle are input into the neural network to construct a fusion network. By combining multiple positioning resources, the heteroscedasticity covariance matrix is ​​accurately determined, ensuring the accuracy of the confidence level.

[0016] In some possible implementations, if the confidence level is within a preset range and the duration exceeds a first duration, the target action is determined by the second network device based on the confidence level of the predicted location and the correspondence between the confidence level and the action.

[0017] In the above scheme, the target action is generated only after the confidence level is within a preset range for a certain duration. This avoids the need for repeated adjustments to the target action, ensures the accuracy of the target action, and reduces resource consumption.

[0018] In some possible implementations, the correspondence between the confidence level and the action to be performed includes:

[0019] If the confidence level is lower than the first threshold, the target action is to re-predict the confidence level using the arrival time or the arrival angle, or to delay the retest.

[0020] If the confidence level is higher than the first threshold and lower than the second threshold, the target performs the action of increasing the positioning reference signal density, deepening the beam, or increasing the beam gain.

[0021] If the confidence level is higher than the second threshold, the target action is to maintain the current state.

[0022] In the above scheme, corresponding execution actions are used as target execution actions for different confidence levels, so as to adjust the first network device according to the actual situation, realize the dynamic adjustment of the first network device, and ensure the accuracy of subsequent positioning.

[0023] In some possible implementations, the method further includes:

[0024] After generating the arrival time and the arrival angle, the Kalman filter is corrected based on the first reference signal.

[0025] In the above scheme, the Kalman filter is corrected after each generation of arrival time and angle of arrival to ensure the accuracy of subsequent Kalman filter predictions of arrival time and angle of arrival.

[0026] In some possible implementations, the correction of the Kalman filter based on the first reference signal includes:

[0027] The Kalman gain is calculated based on the state-measurement cross-covariance and the measurement covariance, where the measurement covariance is determined based on the first reference signal, and the state-measurement cross-covariance is determined based on the sampling points and the nonlinear motion model.

[0028] The predicted state is corrected based on the Kalman gain and the predicted state covariance to obtain the posterior state estimate; the predicted state covariance is determined based on the sampling points and the nonlinear motion model.

[0029] In the above scheme, the Kalman filter is corrected after each generation of arrival time and angle of arrival to ensure the accuracy of subsequent Kalman filter predictions of arrival time and angle of arrival.

[0030] In some possible implementations, the first identifier is the measurement window number of the arrival time and the arrival angle, as well as a timestamp.

[0031] In the above scheme, the measurement window number and timestamp used in the process of generating arrival time and arrival angle are used as unique identifiers corresponding to arrival time and arrival angle, so as to facilitate cross-message and cross-cell alignment.

[0032] A second aspect of this application provides a communication method, which may be executed by a network device, or by a component configured in the network device (such as a circuit, core network unit, chip, or chip system), or by a logic module or software capable of implementing all or part of the functions of the network device. This application does not limit the scope of the method. The following description uses a second network device as an example. The method includes:

[0033] The system receives first information, which includes arrival time, angle of arrival, and a first identifier. The arrival time and angle of arrival are the arrival time and angle of arrival of the signal transmitted at the predicted location received by the first network device. The predicted location is the position of the target terminal predicted by the first network device after a preset time. The first identifier is an identifier for the arrival time and the angle of arrival.

[0034] Receive second information, the second information instructing the second network device to generate a target execution action corresponding to the first identifier;

[0035] The target action is determined based on the confidence level of the predicted location and the correspondence between the confidence level and the action to be performed; the confidence level of the predicted location is determined based on the arrival time, the arrival angle, and a two-stream neural network.

[0036] A third message is sent, which instructs the target to perform an action.

[0037] In the above scheme, by establishing a correspondence between confidence level and execution action, and enabling the second network device to determine the corresponding execution action based on the correspondence between confidence level and execution action, the execution action is sent to the first network device. This allows the first network device to execute the corresponding action according to the actual situation, thereby realizing dynamic adjustment of positioning resources and ensuring positioning accuracy.

[0038] In some possible implementations, the correspondence between the confidence level and the action to be performed is pre-configured in the second network device.

[0039] In the above scheme, by pre-configuring the correspondence between confidence level and execution action in the second network device, the second network device does not need to obtain the correspondence between confidence level and execution action from other devices, thereby reducing the transmission resources required.

[0040] In some possible implementations, the correspondence between confidence level and action execution includes:

[0041] If the confidence level is lower than the first threshold, the target action is to re-predict the confidence level using the arrival time or the arrival angle, or to delay the retest.

[0042] If the confidence level is higher than the first threshold and lower than the second threshold, the target action is to increase the PRS density, deepen the beam, or increase the beam gain.

[0043] If the confidence level is higher than the second threshold, the target action is to maintain the current state.

[0044] In some possible implementations, the output of the two-stream neural network is a heteroscedastic covariance matrix, which is used to map the confidence level of the predicted location.

[0045] In some possible implementations, the two-stream neural network includes a first encoder and a second encoder;

[0046] The first encoder is used to take the arrival time as input and output a Gaussian parameter corresponding to the arrival time;

[0047] The second encoder is used to take the arrival angle as input and output Gaussian parameters corresponding to the arrival angle; the Gaussian parameters corresponding to the arrival time and the Gaussian parameters corresponding to the arrival angle are used to determine the heteroscedasticity covariance matrix.

[0048] It should be noted that the technical effects of the above implementation methods can be referred to the technical effects of the corresponding implementation methods in the first aspect.

[0049] A third aspect of this application provides a communication device, including a module for performing the method provided in the first aspect, or a module for performing the method provided in the second aspect.

[0050] A fourth aspect of this application provides a computer-readable storage medium storing a computer program or instructions that, when executed by a communication device, implement the method provided in the first aspect or the method provided in the second aspect.

[0051] The fifth aspect of this application provides a computer program product including instructions that, when executed, cause the method provided in the first aspect or the method provided in the second aspect.

[0052] A sixth aspect of this application provides a chip including a processor coupled to a memory for executing a computer program or instructions stored in the memory, such that the chip implements the method provided in the first aspect or the method provided in the second aspect.

[0053] The seventh aspect of this application provides a communication device, including a processor and an interface circuit. The interface circuit is used to receive signals from other communication devices and transmit them to the processor, or to send signals from the processor to other communication devices. The processor is used to implement the method provided in the first aspect or the method provided in the second aspect through logic circuits or executing code instructions.

[0054] An eighth aspect of this application provides a communication system, including the communication apparatus provided in the seventh aspect. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the system architecture of the communication system provided in the embodiments of this application;

[0056] Figure 2 A flowchart illustrating a communication method provided in an embodiment of this application;

[0057] Figure 3 A flowchart illustrating another communication method provided in an embodiment of this application;

[0058] Figure 4 A schematic diagram of the cumulative distribution function (CDF) curves under different schemes provided in the embodiments of this application;

[0059] Figure 5 A schematic diagram illustrating the average error under different schemes provided in the embodiments of this application;

[0060] Figure 6 A schematic diagram of the structure of a communication device provided in this application;

[0061] Figure 7 A schematic diagram of another communication device provided in this application;

[0062] Figure 8 A schematic diagram of the structure of an electronic device provided in this application;

[0063] Figure 9 A schematic diagram of the structure of another electronic device provided in this application. Detailed Implementation

[0064] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. The terminology used in the following embodiments is for the purpose of describing specific embodiments only and is not intended to be a limitation of this application. As used in the specification and appended claims of this application, the singular expressions "a," "an," "the," "the," "the," and "this" are intended to also include expressions such as "one or more," unless the context clearly indicates otherwise. It should also be understood that in the embodiments of this application, "one or more" refers to one, two, or more; "and / or" describes the relationship between related objects, indicating that three relationships may exist; for example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship.

[0065] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0066] The "multiple" mentioned in the embodiments of this application refers to two or more. It should be noted that in the description of the embodiments of this application, terms such as "first" and "second" are used only for the purpose of distinguishing descriptions and should not be construed as indicating or implying relative importance, nor should they be construed as indicating or implying order.

[0067] The embodiments of this application are applied to communication systems, which can be second-generation (2G) communication systems, third-generation (3G) communication systems, LTE systems, fifth-generation (5G) communication systems, LTE and 5G hybrid architectures, 5G new radio (5G NR) systems, and new communication systems that will emerge in the future development of communication.

[0068] A communication system includes a first device, a second device, and a third device. The first and second devices can be network-side devices used to provide network communication functions; in some cases, they are also called network equipment or network elements. Network equipment can typically be a base station (including functional units of a base station, or a combination of functional units of base stations) or a core network unit. The core network unit can be a functional unit within the core network, including but not limited to access and mobility management function (AMF) units or session management function (SMF) units. The third device can be a device accessing the network, typically a terminal. An example of a communication system is shown below. Figure 1 As shown, Figure 1 It includes base station 1, network element 2 and terminal equipment 3.

[0069] In the embodiments provided in this application, the base station can be any device with wireless transceiver capabilities, including but not limited to: evolved base stations (NodeB, eNB, or e-NodeB) in Long Term Evolution (LTE), base stations (gNodeB or gNB) or transmission receiving points / transmission reception points (TRPs) in New Radio (NR), and base stations in subsequent 3GPP evolutions. The base station can be: a macro base station, micro base station, pico base station, small cell, relay station, or balloon station, etc. The base station can include one or more co-located or non-co-located transmission reception points (TRPs). The base station can also be a radio controller, centralized unit (CU), and / or distributed unit (DU) in a cloud radio access network (CRAN) scenario. The base station can communicate with the terminal, or it can communicate with the terminal through a relay station. The terminal can communicate with multiple base stations using different technologies. For example, the terminal can communicate with base stations that support LTE networks, base stations that support 5G networks, and can also establish dual connections with both LTE and 5G base stations.

[0070] In the embodiments provided in this application, the terminal can take various forms, such as a mobile phone, tablet computer, computer with wireless transceiver capabilities, virtual reality (VR) terminal device, augmented reality (AR) terminal device, wireless terminal in industrial control, vehicle-mounted terminal device, wireless terminal in self-driving, wireless terminal in remote medical care, wireless terminal in smart grid, wireless terminal in transportation safety, wireless terminal in smart city, wireless terminal in smart home, wearable terminal device, etc. The terminal may also be referred to as terminal equipment, user equipment (UE), access terminal equipment, vehicle-mounted terminal, industrial control terminal, UE unit, UE station, mobile station, mobile station, remote station, remote terminal equipment, mobile device, UE terminal equipment, terminal equipment, wireless communication equipment, UE agent, or UE device, etc. The terminal can also be a fixed terminal or a mobile terminal.

[0071] In ISAC (Inter-Application Configuration Control), network devices utilize positioning resources to locate terminal devices and adjust these resources periodically to avoid positioning errors caused by prolonged operation. However, this fixed-period adjustment of positioning resources can result in adjustments being made too early or too late, and the adjustment method may not be suitable for the current network devices, leading to low positioning accuracy. Therefore, improving positioning accuracy has become a key technical problem that needs to be solved.

[0072] To address the aforementioned problems, embodiments of this application provide a communication method, apparatus, medium, and product. To make the technical solutions of this application clearer and easier to understand, the communication method, apparatus, medium, and product provided in the embodiments of this application will be described below with reference to the accompanying drawings.

[0073] See Figure 2 The flowchart shown illustrates a communication method, which includes:

[0074] S201: The first network device sends first information, which includes arrival time, angle of arrival, and a first identifier; the arrival time and angle of arrival are the arrival time and angle of the signal received by the first network device from the predicted location; the predicted location is the predicted location of the target terminal after a preset time; and the first identifier is an identifier for the arrival time and angle of arrival. Correspondingly, the second network device receives the first information.

[0075] The first network device predicts the future location of the target terminal device to obtain the predicted location. Specifically, the predicted location is the location of the target terminal device after a preset time, so that the first network device can drive beamforming or pointing based on the predicted location of the target terminal to enhance the measurement signal-to-noise ratio and angular spectral resolution of the positioning reference signal (PRS) or sounding reference signal (SRS), and obtain accurate time of arrival (TOA) and / or angle of arrival (AOA).

[0076] The time of arrival is the time when the first network device receives the signal sent by the target terminal device when the target terminal device sends a signal to the first network device from the predicted location. The angle of arrival is the angle at which the first network device receives the signal sent by the target terminal device when the target terminal device sends a signal to the first network device from the predicted location.

[0077] After determining the arrival time and / or angle of arrival, the first network device sends the arrival time and / or angle of arrival, along with a first identifier corresponding to the arrival time and angle of arrival, to the second network device. This enables the second network device to store the arrival time and angle of arrival and to distinguish the arrival time and angle of arrival sent by the first network device this time from those sent by other network devices, as well as the arrival time and angle of arrival sent by the first network device at other times, so as to facilitate subsequent cross-message or cross-cell alignment.

[0078] S202: The first network device sends second information, which instructs the second network device to generate a target execution action corresponding to the first identifier. Correspondingly, the second network device receives the second information.

[0079] The second information is specifically a feature report. The first network device sends a feature report to the second network device. The feature report carries a first identifier, enabling the second network device to perform cross-message or cross-cell alignment based on the first identifier. It retrieves the arrival time and angle of arrival corresponding to the first identifier from its local storage for analysis, and determines the target execution action corresponding to the arrival time and angle of arrival. The target execution action is used to adjust the positioning resources of the first network device to ensure the positioning accuracy of the target terminal device.

[0080] It should be noted that when the first network device does not require adjustment, the target action can be empty to save resources on the first network device side.

[0081] S203: The first network device receives third information, which instructs the target to perform an action. The target's action is determined by the second network device based on the confidence level of the predicted location and the correspondence between the confidence level and the action. The confidence level of the predicted location is determined based on the time of arrival, the angle of arrival, and a two-stream neural network. Accordingly, the second network device sends the third information.

[0082] The second network device sends third information to the first network device. This third information includes the target action corresponding to the confidence level. The confidence level is determined by the second network device after inputting the angle of arrival and time of arrival into a two-stream neural network, based on the confidence level for the predicted location.

[0083] Taking the first network device as the base station and the second network device as an LMF network element as an example, the flowchart of a communication method in this embodiment is as follows: Figure 3 As shown, the method includes:

[0084] S301: LMF pre-obtains the correspondence between confidence level and execution action.

[0085] Set a confidence threshold to divide the confidence level into multiple ranges, with each confidence level range corresponding to a different execution action.

[0086] The correspondence between confidence levels and actions stored in the LMF can be pre-set in the LMF or sent to the LMF by the base station in advance. This embodiment will be described based on the premise that the correspondence between confidence levels and actions is pre-set in the LMF, that is, the correspondence between confidence levels and actions is pre-configured in the second network device to reduce the overhead of the LMF and the base station.

[0087] It should be noted that, in addition to pre-storing the correspondence between confidence and execution actions, LMF may also include the maximum value to which location resources can be adjusted, as well as hold window parameters, so as to facilitate subsequent determination of whether to generate the target execution action, and to select the currently executable execution action as the target execution action from multiple execution actions corresponding to the confidence level.

[0088] S302: The base station measures the arrival time and / or angle of arrival of the predicted location.

[0089] The base station performs routine measurements for PRS or SRS, and at the same time determines the predicted location of the target terminal device based on an unscented Kalman filter (UKF), and drives beamforming or pointing to enhance the measurement signal-to-noise ratio and angular spectral resolution of PRS or SRS, and determine a more accurate time of arrival and angle of arrival.

[0090] In one alternative embodiment, the arrival time and angle of arrival are determined based on an unscented Kalman filter and a first reference signal, which is either a positioning reference signal or a detection reference signal.

[0091] Specifically, the arrival time and angle of arrival are determined based on the unscented Kalman filter. The parameters of the unscented Kalman filter are then corrected based on the first reference signal so that the corrected unscented Kalman filter can continue to determine the next arrival time and angle of arrival. This process of cyclic correction of the unscented Kalman filter ensures the accuracy of the arrival time and angle of arrival determined by the unscented Kalman filter.

[0092] Furthermore, when determining the arrival time and angle of arrival based on an unscented Kalman filter, the arrival time is specifically determined based on the Euclidean distance between the predicted location's position vector and the first network device's position vector, the measurement noise of the arrival time, and the ranging bias term of the arrival time; the predicted location's position vector is determined based on the unscented Kalman filter and the first reference signal. The angle of arrival is specifically determined based on the position of the first network device, the predicted location, the measurement noise of the angle of arrival, and the systematic bias in the angle of arrival; the predicted location is determined based on the unscented Kalman filter and the first reference signal.

[0093] The unscented Kalman filter includes a state and process model, as well as a measurement model. The state and process model includes: under discrete-time conditions, defining the system state vector at time k as... ,in This represents the two-dimensional spatial coordinates of the target terminal device at time k, which is also the predicted position. The x-axis is... The vertical axis is , This represents the velocity components of the target terminal device on the coordinate axis. The velocity component of the target terminal device on the horizontal axis. This represents the velocity component of the target terminal device on the vertical axis.

[0094] The process model is adopted as a uniform motion model or an equivalent model incorporating acceleration noise. The process model is as follows:

[0095] ;

[0096] in, The function represents the time interval between adjacent discrete moments. Used to describe the motion relationship of the target terminal device between adjacent moments. Process noise. Used to characterize uncertainties introduced by factors such as random acceleration and motion modeling errors. ,Right now It follows a zero-mean Gaussian distribution. Let be the process noise covariance matrix.

[0097] The coordinates of the predicted location are substituted into the measurement model to determine the arrival time and angle. The measurement model is for the j-th base station at time k, and consists of two types of observations: distance measurement and azimuth measurement.

[0098] The distance measurement model is as follows:

[0099] ;

[0100] in, Represents the predicted location vector. Indicates the first The location vectors of several base stations, where the j-th base station is the first network device. This is for Euclidean distance calculation. Measurement noise, representing the time of arrival, is used to describe ranging error. The distance offset term, representing the time of arrival, is used to characterize the systematic positive deviation generated under non-line-of-sight (NLOS) propagation conditions.

[0101] The specific orientation measurement model is as follows:

[0102] ;

[0103] in, Indicates the first The two-dimensional coordinates of each base station Let j be the x-coordinate of the j-th base station. Let j be the ordinate of the j-th base station. It is a four-quadrant arctangent function used to calculate the incident angle or azimuth angle of a target relative to a base station. The measurement noise representing the angle of arrival. To represent the systematic bias in the angle, used to indicate the systematic bias that may exist in the angle measurement.

[0104] The distance measurement model and the azimuth measurement model can be uniformly represented as: ,in, For the first The nonlinear measurement function corresponding to each base station To observe the noise vector, This is the bias vector. The unified representation of the ranging and azimuth measurement models facilitates subsequent multimodal joint processing within the unscented Kalman filtering framework.

[0105] S303: The base station sends first information, which includes arrival time, arrival angle and first identifier, and the corresponding LMF receives the first information.

[0106] The first identifier is the measurement window number of arrival time and arrival angle, and the timestamp. The measurement window number is the number of the measurement window that generates the arrival time and arrival angle, and the timestamp is used to indicate the time when the arrival time and arrival angle are generated. Thus, by combining the measurement window number of arrival time and arrival angle, and the time when the arrival time and arrival angle are generated, the first identifier can uniquely represent an arrival time and arrival angle.

[0107] S304: The base station sends second information, which instructs the second network device to generate a target execution action corresponding to the first identifier, and the corresponding LMF receives the second information.

[0108] After receiving the second information, LMF determines the first information that needs to be generated for the target execution action based on the timestamp and side window number in the second information. The second information is specifically a feature report, the contents of which are shown in Table 1.

[0109]

[0110] Table 1

[0111] The feature report must include at least a timestamp field and a measurement window field. The timestamp field is of type unsigned integer, with a value range limited to 0 to 2. 32-1 The measurement window field is of type unsigned integer, with a value range limited to 0 to 65535, and a size of 2 bytes. The calibration fingerprint field is of type octet string and occupies 4 bytes. The version field is of type enumerated, with no byte limit; its size can be set based on actual needs.

[0112] S305: LMF determines the confidence level corresponding to the predicted location based on arrival time, arrival angle, and a two-stream neural network.

[0113] Arrival time and arrival angle are input into a two-stream neural network (2NN) to determine the confidence level corresponding to the predicted location based on the network's output. Specifically, the 2NN outputs a heteroscedasticity covariance matrix (HCC), which maps the confidence level of the predicted location. The 2NN includes a first encoder and a second encoder. The first encoder takes the arrival time as input and outputs Gaussian parameters corresponding to that time. The second encoder takes the arrival angle as input and outputs Gaussian parameters corresponding to both the arrival angle and arrival time. These parameters are used to determine the HCC. Further, in this embodiment, the HCC can be replaced with a diagonal approximation of the HCC, and the first and second encoders are lightweight encoders.

[0114] It should be noted that the output of the two-stream neural network includes not only the heteroscedasticity covariance matrix, but also the coordinates of the predicted position, in order to establish the relationship between the heteroscedasticity covariance matrix and the predicted position, so as to determine whether the output of the two-stream neural network is the heteroscedasticity covariance matrix corresponding to the predicted position.

[0115] Specifically, the data stream corresponding to the arrival time is input to the first encoder. The content input to the first encoder can be a one-dimensional time feature sequence or a vector of the arrival time, such as the peak value, sidelobe ratio, peak width, and delay spread statistics of the arrival time. The data stream corresponding to the arrival angle is input to the second encoder. The content input to the second encoder can be the angular spectral peak, main lobe angle, interpeak ratio, angular spread, array consistency index, etc., so that the first encoder (TOA encoder) and the second encoder (AOA encoder) each output Gaussian parameters corresponding to the input. The Gaussian parameters can be parameters such as mean and variance. For example, the TOA encoder outputs (μ T , Σ T ), μ T Σ represents the predicted position output by the first encoder. T The AOA encoder outputs (μ) the covariance or uncertainty for the predicted position, which is the output of the first encoder. A , Σ A ), μ A Σ represents the predicted position output by the second encoder. A This refers to the covariance or uncertainty of the predicted position output by the second encoder. To predict the location or mean, it is a two-dimensional vector. , Covariance, or uncertainty, is a 2×2 matrix. Specifically, it can be a diagonal matrix.

[0116] Perform a product of experts (PoE) on the Gaussian parameters output by the first and second encoders to obtain the mean and covariance of the fused sample. , .in, The covariance after fusion This is the average value after fusion. In the case where either the first or second encoder is unreliable, due to the unreliable encoder's... The size is relatively small, so it has little impact on the accuracy of the fused result, thus ensuring the accuracy of the fused result.

[0117] The training loss of a two-stream neural network includes regression loss, scene classification auxiliary loss, distance / angle auxiliary loss, etc., to improve the robustness of non-line-of-sight (NLOS) or multipath scenes.

[0118] Specifically, for the true location (G) of the target terminal device, Gaussian negative log-likelihood (NLL) or weighted multi-task loss is used as the main regression term to obtain the loss function of the two-stream neural network:

[0119] ;

[0120] If diagonal covariance is used, then in the loss function... , σ x σ is the x-coordinate of the predicted location. y To predict the ordinate of the location, the loss function can be decomposed into a scalar form, which facilitates stable training. To avoid numerical problems, the network can output the loss function at the implementation layer. and will Cut off to To ensure training stability, among which This is the logarithmic parameterization of the diagonal covariance.

[0121] Other auxiliary losses can be weighted together to obtain the final loss function. :

[0122] ;

[0123] in Cross-entropy for scene classification To address the auxiliary regression error for TOA, To address the auxiliary regression error of AOA, To and The corresponding weights To and The corresponding weights To and The corresponding weights To and The corresponding weights.

[0124] Furthermore, after constructing the loss function, the confidence level corresponding to the predicted location is determined based on the output of the two-stream neural network, including:

[0125] Extracting precision (prec) from the fused covariance:

[0126] .

[0127] Will Mapped to finite integer confidence scalar A parameterizable monotonic mapping is used:

[0128]

[0129] in For calibration parameters, a calibration set is used to map common accuracy ranges to the desired confidence interval. To round up or down, The output is in the range of 0 to 1023. This mapping facilitates the efficient transmission of confidence in signaling in small integer form, serving as the basis for decision-making regarding subsequent target actions.

[0130] S306: LMF determines the target action based on the confidence level corresponding to the predicted location and the correspondence between the confidence level and the action to be performed.

[0131] In an optional embodiment, if the confidence level is within a preset range and the duration exceeds a first duration, the target action is determined by the second network device based on the confidence level of the predicted location and the correspondence between the confidence level and the action.

[0132] To avoid frequently generating multiple target execution actions, this embodiment introduces a first duration. When the confidence level remains within a specific range and the duration reaches the first duration, the LMF generates the target execution action corresponding to that confidence level. If the duration of the confidence level remaining within any range is less than the first duration, then the first duration is not generated, or the duration of the confidence level within each range is determined, and the target execution action is determined based on the range with the longest cumulative time. The first duration can specifically be a hold window length, or other information that can indicate the duration for which the confidence level needs to be maintained. The hold window length is the duration of the hold window, used to require the confidence level to remain within a single preset range for at least H time slots before triggering the instruction to generate the target execution action, where H is an integer greater than 0.

[0133] The correspondence between confidence level and action is specifically the correspondence between the preset range of confidence level and the action to be performed. The preset range is the range defined by LMF based on the confidence level threshold. For example, when the confidence level threshold includes threshold 1 and threshold 2, and threshold 1 is less than threshold 2, the preset range can be divided into three ranges: confidence level less than threshold 1, confidence level greater than or equal to threshold 1 and less than threshold 2, and confidence level greater than or equal to threshold 2.

[0134] To facilitate understanding, the following examples will be provided:

[0135] With confidence thresholds of 0.7 and 0.8, and a window length of 10 time slots, the preset ranges include preset range a, preset range b, and preset range c. Preset range a has a confidence level less than 0.7, preset range b has a confidence level greater than or equal to 0.7 and less than 0.8, and preset range c has a confidence level greater than or equal to 0.8. If the confidence level is 0.75 for 10 consecutive time slots, then the preset range falls within preset range b for 10 consecutive time slots, triggering the generation of the target execution action instruction, which is generated based on a confidence level of 0.75. If the confidence level is 0.75 for 2 time slots, 0.71 for 3 time slots, 0.69 for 2 time slots, and 0.8 for 3 time slots, then none of the preset ranges reaches the window length required to generate the target execution action, so no target execution action is generated. Alternatively, since the time slots within the preset range b are 5 time slots, they account for the largest proportion among the 3 preset ranges and are the range with the longest cumulative time where the confidence threshold is located, the target execution action is generated based on the preset range b.

[0136] In one optional embodiment, the correspondence between confidence level and action execution includes:

[0137] If the confidence level is lower than the first threshold, the target action is to re-predict the confidence level using the arrival time or arrival angle, or to delay the retest.

[0138] If the confidence level is higher than the first threshold and lower than the second threshold, the target will perform the action of increasing the PRS density, deepening the beam, or increasing the beam gain.

[0139] If the confidence level is higher than the second threshold, the target will perform the action of maintaining the current state.

[0140] The target actions include: Increase PRS density to improve the proportion of PRS resource elements (REs) on time-frequency resources. Deepen the beam or increase beam gain to improve directivity and reduce sidelobes. Delayed retesting, pending UKF's next prediction. : Regress to a single modality, for example, predict confidence based solely on AOA or TOA, to avoid interference when the data is affected by the single modality regression, making the generated confidence scores more accurate. Maintain the current state; no additional action is required.

[0141] Taking time t as an example, LMF is based on the confidence level at time t. Determine the target and execute the action. :

[0142] ;

[0143] in, The confidence threshold. To avoid frequent flips, a hysteresis threshold is introduced, which includes an entry threshold. and exit threshold ,satisfy .

[0144] It should be noted that the correspondence between confidence level and execution action, as well as the confidence threshold, can be pre-configured in the LMF or sent by the base station to the LMF; this embodiment does not limit this. When the correspondence between confidence level and execution action, and the confidence threshold, are sent by the base station to the LMF, the content sent by the base station to the LMF is shown in Table 2:

[0145]

[0146] Table 2

[0147] The base station sends at least the Action Table field and the Confidence Threshold Setting field to the LMF. The Confidence Threshold Setting field is a dynamic array (sequence of integers), consisting of integers ranging from 0 to 1023, with a size of 3 × 10 bits. The Action Table field is also a dynamic array sequence containing an enumeration type to identify the action to be performed, and a sequence of sequences to encapsulate the threshold. The size of the dynamic array is dynamically adjusted based on the correspondence between thresholds and actions. The Resource Capability field is a sequence, including capability parameters such as maximum power, maximum beam depth, and maximum hold duration. The size of the Resource Capability field is dynamically adjusted based on the number of capabilities that need to be reported to the LMF. The Hysteresis or Hold field is a sequence, and its size can be set according to actual needs. The Version field is an enumerated type with no byte limit, and its size can be set according to actual needs.

[0148] S307: The LMF sends third information, which instructs the target to perform an action. Correspondingly, the base station receives the third information.

[0149] After determining the target action, LMF sends the target action to the base station through third information. The third information carries at least the target action, and the specific content of the third information is shown in Table 3.

[0150]

[0151] Table 3

[0152] The `decided action` field is an enumeration type used to enumerate the target actions, and its size is less than 3 bits. The `action parameter` field is a sequence type, and its size can be adjusted based on the number of action parameters, such as PRS period / density increment. The `valid window` field is an unsigned integer type, ranging from 0 to 65535, and its size is 2 bytes. The `reason code` field is an enumeration type, less than 3 bits in size, used to indicate the triggering reason for the anomaly, such as low confidence or data anomaly.

[0153] S308: The base station executes the target action.

[0154] After receiving the target's action, the base station executes the target's action to achieve dynamic adjustment of the base station and ensure the positioning accuracy of the base station.

[0155] In an optional embodiment, the above method further includes:

[0156] After generating the arrival time and angle of arrival, the Kalman filter is corrected based on the first reference signal.

[0157] After each time of arrival and angle of arrival is generated, the Kalman filter is corrected so that the base station can predict subsequent times of arrival and angles of arrival based on the corrected Kalman filter. This enables dynamic adjustment of the Kalman filter and ensures the accuracy of the times of arrival and angles of arrival.

[0158] Specifically, the Kalman filter is corrected based on the first reference signal, including:

[0159] The Kalman gain is calculated based on the state-measurement cross-covariance and the measurement covariance. The measurement covariance is determined based on the first reference signal, and the state-measurement cross-covariance is determined based on the sampling points and the nonlinear motion model.

[0160] The predicted state is corrected based on Kalman gain and predicted state covariance to obtain the posterior state estimate; the predicted state covariance is determined based on sampling points and nonlinear motion model.

[0161] Furthermore, the state-measurement cross-covariance is determined based on sampling points and a nonlinear motion model, including:

[0162] generate σ points, Each σ point is used to characterize the state distribution at time k-1;

[0163] The state distribution at time k-1 is propagated through a nonlinear motion model to obtain the predicted distribution of σ at time k.

[0164] The predicted distributions are weighted and summed according to their mean values ​​to obtain the mean of the predicted states;

[0165] The covariance of the predicted state is determined based on the degree of dispersion of the predicted distribution of σ at time k relative to the mean of the predicted state.

[0166] Substituting each σ point into the nonlinear measurement function yields the predicted measurement σ point;

[0167] The mean of the predicted quantities is obtained by weighting all predicted quantity σ points according to their mean weights.

[0168] The predicted measurement covariance is determined based on the predicted measurement points, the predicted measurement mean, the measurement noise covariance matrix, and the covariance weights.

[0169] Based on the predicted measurement points, the predicted measurement mean, the covariance weight, the predicted state mean, and the predicted distribution of the state at the current time, the state-measurement cross-covariance is determined.

[0170] Specifically, let the state dimension be... The Kalman filter is specifically the unscented Kalman filter (UKF). The UKF's scaling parameters include... The scaling factor is defined as:

[0171] ;

[0172] in, Used to control the expansion scale of point σ in the state space, where point σ is a sampling point. Used to introduce prior distribution information; for Gaussian distributions, it is usually taken as... , To adjust the parameters.

[0173] The mean weights and covariance weights corresponding to the σ point are as follows:

[0174] ;

[0175] ;

[0176] Mean weight Covariance weights This is used to ensure that the mean and covariance obtained through the σ-point approximation are consistent with the original state distribution in both first and second order senses. The mean of the state estimate is based on the current moment. and its covariance matrix ,generate σ points:

[0177] ;

[0178] in, Represents the square root operation of a matrix, with subscripts. This indicates retrieving the corresponding column. For at any time The σ point generated in the above manner statistically equivalently characterizes time. The state distribution is used to provide a basis for subsequent nonlinear propagation.

[0179] For each generated σ point, propagation is performed using the system's nonlinear process model:

[0180] ;

[0181] The step of predicting the distribution of point σ at time k is equivalent to propagating the state distribution from the previous time step through a nonlinear motion model, thereby obtaining the state at time k. The predicted distribution is represented by .

[0182] After obtaining the set of predicted σ points, the set of σ points is weighted and summed according to the mean weight to obtain the mean of the predicted state:

[0183] ;

[0184] Therefore, the predicted state mean It is not obtained from a single point of propagation, but is jointly determined by all predicted σ points in a statistical sense, thus more accurately reflecting the prior state of the nonlinear system.

[0185] Predicted state covariance Further calculations are performed based on the dispersion of the predicted σ point relative to the predicted mean:

[0186] ;

[0187] in, This is the process noise covariance matrix, used to compensate for uncertainties not explicitly modeled in the motion model.

[0188] After obtaining the predicted distribution of the state, it is necessary to map the predicted state to the observation space for measurement updates. To do this, each predicted σ point... Substitute into the nonlinear measurement function The corresponding predicted measurement σ point is obtained:

[0189] ;

[0190] This process is equivalent to propagating the predicted state distribution through a nonlinear measurement model, thereby constructing a set of σ points in the observation space that correspond one-to-one with the state space.

[0191] Subsequently, a weighted sum is performed on all predicted measurement σ points according to their mean weights to obtain the mean of the predicted measurements:

[0192] ;

[0193] It should be noted that the predicted measurement mean It is not simply a matter of calculating the mean of the predicted states. Instead of directly substituting the measurement function, it is obtained through the statistical propagation of the σ point, thus more accurately reflecting the impact of the nonlinear measurement model on state uncertainty.

[0194] The predicted measurement covariance and the state-measurement cross-covariance are calculated as follows:

[0195] ;

[0196] ;

[0197] in, For state-measurement cross-covariance, To measure covariance, This is the measurement noise covariance matrix, used to characterize the noise characteristics of each measurement stream in TOA and AOA.

[0198] Kalman gain calculated based on state-measurement cross-covariance and measurement covariance. Kalman gain is used to weigh the predicted state against the actual observation. The magnitude of Kalman gain reflects the system’s relative confidence in the predicted information versus the measured information.

[0199] The predicted state is corrected based on Kalman gain and predicted state covariance to obtain the posterior state estimate:

[0200]

[0201] ;

[0202] The above update process completes a full prediction-measurement-correction cycle. Among these, the measurement covariance... The system can be dynamically adjusted based on the noise characteristics of each stream in the TOA / AOA, the antenna array size, beamwidth, and PRS density; simultaneously, the predicted state estimate... It can be used to drive beam pointing and gain control in the next time slot, thus forming a closed-loop optimization mechanism. Let k be the state estimation error covariance matrix at the current time k. To predict state covariance, This is the interaction term between Kalman gain and measurement uncertainty, reflecting the reduction in estimation uncertainty after fusing measurements.

[0203] For ease of understanding, the following comparison is provided:

[0204] like Figure 4 and Figure 5 As shown, in a 100m × 100m scene, four base stations are placed at the four corners, randomly distributed within the target area. Parameters are set as follows: TOA ranging noise standard deviation 0.5m, NLOS path probability 50% with a positive bias of 1.0 ± 0.4m, and AOA angle error approximately 1.5°. A total of 3000 randomly distributed target point samples are generated, divided into training and test sets in an 8:2 ratio. Five schemes are used to test the positioning accuracy, from left to right: positioning using the dual-stream neural network in this embodiment, positioning using the TOA neural network, positioning using the AOA neural network, triangulation positioning using only TOA, and triangulation positioning using only AOA.

[0205] Based on the above comparison results, the two-stream neural network significantly outperforms all the comparison methods with an average localization error of 0.89m. It improves localization accuracy by approximately 36% compared to traditional TOA triangulation and by approximately 15% compared to single-stream TOA neural networks. The CDF curve also shows that the two-stream network has the highest localization success rate at all error thresholds, verifying the effectiveness of multi-source information fusion and attention mechanisms in suppressing NLOS interference.

[0206] This application establishes a correspondence between confidence level and execution action, and enables the second network device to send the execution action to the first network device after determining the corresponding execution action based on the correspondence between confidence level and execution action. This allows the first network device to execute the corresponding action according to the actual situation, thereby realizing dynamic adjustment of positioning resources and ensuring positioning accuracy.

[0207] This application also provides a communication device, including a module for performing a communication method.

[0208] This application also provides a computer-readable storage medium storing a computer program or instructions, which, when executed by a communication device, implements a communication method.

[0209] This application also provides a computer program product, including instructions that, when executed, enable a communication method to be implemented.

[0210] This application also provides a chip, including a processor coupled to a memory, for executing computer programs or instructions stored in the memory, thereby enabling the chip to implement a communication method.

[0211] This application also provides a communication device, including a processor and an interface circuit. The interface circuit is used to receive signals from other communication devices and transmit them to the processor, or to send signals from the processor to other communication devices. The processor implements the communication method through logic circuits or executing code instructions.

[0212] Figure 6 This is a schematic block diagram of a communication device provided in an embodiment of this application. Figure 6 As shown, the communication device 600 may include a communication module 610. The communication module 610 can implement corresponding communication functions, which can be internal communication functions of the communication device 600 or communication functions between the communication device 600 and other devices. Optionally, the communication module 610 may also be referred to as a communication interface or transceiver module. Optionally, the communication device 600 also includes a processing module 620. The processing module 620 can implement corresponding processing functions.

[0213] Optionally, the communication device 600 further includes a storage module, which can be used to store instructions and / or data; the processing module 620 can read the instructions and / or data in the storage module so that the communication device 600 can implement the aforementioned method embodiments.

[0214] In one possible design, the communication device 600 may correspond to the first network device in the above method embodiments, or to a component (such as a circuit, chip, or chip system) configured in the first network device. The communication device 600 can be used to perform the steps or processes performed by the network device in any of the above method embodiments.

[0215] For example, the communication module 610 is used to send first information, which includes arrival time, arrival angle, and first identifier; the arrival time and arrival angle are the arrival time and arrival angle of the signal sent from the predicted location received by the first network device; the predicted location is the position of the target terminal predicted by the first network device after a preset time; the first identifier is an identifier of the arrival time and arrival angle.

[0216] The communication module 610 is also used to instruct the second network device to generate a target execution action corresponding to the first identifier;

[0217] The communication module 610 is also used to receive third information, which instructs the target to perform an action. The target's action is determined by the second network device based on the confidence level of the predicted position and the correspondence between the confidence level and the action. The confidence level of the predicted position is determined based on the time of arrival, the angle of arrival, and a two-stream neural network.

[0218] The processing module 620 is used to execute the target execution action.

[0219] The above are merely examples; for detailed steps or procedures, please refer to the descriptions in the foregoing embodiments.

[0220] In one possible design, the communication device 600 may correspond to the second network device in the above method embodiments, or to a component (such as a circuit, chip, or chip system) configured in the second network device. The communication device 600 can be used to perform the steps or processes performed by the network device in any of the above method embodiments.

[0221] For example, the communication module 610 is used to receive first information, which includes arrival time, angle of arrival, and first identifier; the arrival time and angle of arrival are the arrival time and angle of arrival of the signal transmitted at the predicted location received by the first network device; the predicted location is the position of the target terminal predicted by the first network device after a preset time; the first identifier is an identifier of the arrival time and angle of arrival.

[0222] The communication module 610 is also used to receive second information, which instructs the second network device to generate a target execution action corresponding to the first identifier;

[0223] The processing module 620 is used to determine the target execution action based on the confidence level of the predicted position and the correspondence between the confidence level and the execution action; the confidence level of the predicted position is determined based on the arrival time, the angle of arrival, and a two-stream neural network;

[0224] The communication module 610 is also used to send third information, which instructs the target to perform an action.

[0225] The above are merely examples; for detailed steps or procedures, please refer to the descriptions in the foregoing embodiments.

[0226] Figure 7 This is another schematic block diagram of the communication device 700 provided in the embodiments of this application. The communication device 700 may be a chip, chip system, or processor, etc., in a terminal device or network device that implements the above-described methods. The communication device 700 can be used to implement the methods described in the above-described method embodiments; for details, please refer to the descriptions in the above-described method embodiments.

[0227] like Figure 7 As shown, the communication device 700 may include one or more processors 710, which may also be referred to as processing units or processing modules, and can implement certain control functions. The processor 710 may be a general-purpose processor or a dedicated processor, such as a baseband processor or a central processing unit. The baseband processor can be used to process communication protocols and communication data, while the central processing unit can be used to control the communication device 700 (e.g., a base station, baseband chip, user, user chip), execute software programs, and process data from the software programs.

[0228] In an alternative design, the processor 710 may also store instructions and / or data, which can be executed by the processor 710 to cause the communication device 700 to perform the methods described in the above method embodiments.

[0229] In another alternative design, the communication device 700 may include a communication interface 720 for implementing receiving and transmitting functions. For example, the communication interface 720 may be a transceiver circuit, interface, interface circuit, or transceiver. The transceiver circuit, interface, interface circuit, or transceiver for implementing receiving and transmitting functions may be separate or integrated. The aforementioned transceiver circuit, interface, interface circuit, or transceiver may be used for reading and writing code / data, or it may be used for transmitting or relaying signals.

[0230] Optionally, the communication device 700 may include one or more memories 730, which may store instructions that can be executed on the processor 710, causing the communication device 700 to perform the methods described in the above method embodiments. Optionally, the memories 730 may also store data. Optionally, the processor 710 may also store instructions and / or data. The processor 710 and the memories 730 may be provided separately or integrated together.

[0231] It should be understood that, in one possible design, the steps in the method embodiments provided in this application can be implemented by integrated logic circuits in the processor's hardware or by instructions in software form. The steps of the methods disclosed in the embodiments of this application can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules in the processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method. To avoid repetition, detailed descriptions are not provided here.

[0232] In one implementation, the communication device 700 may correspond to the terminal device in the above method embodiments and may be used to execute the various steps and / or processes executed by the terminal device in the above method embodiments. The processor 710 may be used to execute instructions stored in the memory 730, and when the processor 710 executes the instructions stored in the memory, the processor 710 is used to execute the various steps and / or processes of the above method embodiments corresponding to the terminal device.

[0233] In another implementation, the communication device 700 may correspond to the network device in the above method embodiments and may be used to execute the various steps and / or processes executed by the network device in the above method embodiments. The processor 710 may be used to execute instructions stored in the memory 730, and when the processor 710 executes the instructions stored in the memory, the processor 710 is used to execute the various steps and / or processes of the above method embodiments corresponding to the network device.

[0234] It should be understood that the aforementioned processing device can be one or more chips. For example, the processing device can be a field-programmable gate array (FPGA), an application-specific integrated circuit (ASIC), a system-on-chip (SoC), a central processor unit (CPU), a network processor (NP), a digital signal processor (DSP), a microcontroller unit (MCU), a programmable logic device (PLD), or other integrated chips.

[0235] It is understood that the memory in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous linked dynamic random access memory (SLDRAM), and direct rambus RAM (DR RAM). It should be noted that the memory used in the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0236] Figure 8 This application provides an example of the composition of an electronic device. The electronic device can be a terminal device, including but not limited to mobile phones, smart wearable devices (such as smartwatches), and other electronic devices. Taking a mobile phone as an example, the electronic device may include a processor 810, an external memory interface 820, an internal memory 821, a display screen 830, a camera 840, antenna 1, antenna 2, a mobile communication module 850, and a wireless communication module 860, etc.

[0237] It is understood that the structure illustrated in this embodiment does not constitute a specific limitation on the electronic device. In other embodiments, the electronic device may include more or fewer components than illustrated, or combine some components, or split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0238] It is understood that the interface connection relationships between the modules illustrated in this embodiment are merely illustrative and do not constitute a limitation on the structure of the electronic device. In other embodiments of this application, the electronic device may also employ different interface connection methods or combinations of multiple interface connection methods as described in the above embodiments.

[0239] The external memory interface 820 can be used to connect external memory cards, such as Micro SD cards, to expand the storage capacity of electronic devices.

[0240] Internal memory 821 can be used to store executable program code, which includes instructions. Processor 810 executes various functional applications and data processing of electronic devices by running the instructions stored in internal memory 821.

[0241] The wireless communication function of electronic devices can be implemented through antenna 1, antenna 2, mobile communication module 850, wireless communication module 860, modem processor, and baseband processor.

[0242] The mobile communication module 850 can provide solutions for wireless communication applications, including 2G / 3G / 4G / 5G, in electronic devices. The mobile communication module 850 may include at least one filter, switch, power amplifier, low noise amplifier (LNA), etc.

[0243] Furthermore, an operating system runs on top of the aforementioned components. Examples include iOS, Android, and Windows operating systems. Applications can be installed and run on this operating system. Those skilled in the art will understand that, for the sake of convenience and brevity, explanations and beneficial effects of the relevant content in any of the above-described electronic devices can be found in the corresponding method embodiments provided above, and will not be repeated here.

[0244] Figure 9 This application provides another example of the composition of an electronic device. The electronic device may be a first device, including but not limited to a base station and a core network unit. Figure 9A simplified schematic diagram of a base station structure is shown. Base station 900 includes parts 910, 920, and 930. Part 910 is mainly used for baseband processing and base station control; part 910 is typically the control center of the base station, often referred to as a processor, used to control the base station to perform the processing operations on the first device side in the above method embodiments. Part 920 is mainly used for storing computer program code and data. Part 930 is mainly used for transmitting and receiving radio frequency signals and converting radio frequency signals to baseband signals; part 930 is often referred to as a transceiver module, transceiver, transceiver circuit, or transceiver unit. The transceiver module of part 930, also referred to as a transceiver or transceiver unit, includes an antenna 933 and radio frequency circuitry (…). Figure 9 (Not shown in the diagram), where the radio frequency circuitry is primarily used for radio frequency processing. Optionally, the device in section 930 used to implement the receiving function can be considered a receiver, and the device used to implement the transmitting function can be considered a transmitter; that is, section 930 includes receiver 932 and transmitter 931. A receiver can also be called a receiving module, receiver circuit, or receiving circuit, etc., and a transmitter can be called a transmitting module, transmitter, or transmitting circuit, etc.

[0245] Sections 910 and 920 may include one or more circuit boards, each of which may include one or more processors and one or more memories. The processors are used to read and execute programs from the memories to implement baseband processing functions and control the base station. If multiple circuit boards exist, they can be interconnected to enhance processing capabilities. As an alternative implementation, multiple circuit boards may share one or more processors, multiple circuit boards may share one or more memories, or multiple circuit boards may simultaneously share one or more processors.

[0246] For example, in one implementation, the transceiver module in section 930 is used to execute the transceiver-related processes performed by the base station (first device) in the aforementioned method embodiments. The processor in section 910 is used to execute the processing-related processes performed by the base station in the aforementioned method embodiments.

[0247] It should be understood that Figure 9 This is for illustrative purposes only and not as a limitation. The network devices mentioned above, including processors, memory, and transceivers, may be independent of... Figure 9 The structure shown.

[0248] This application also provides a chip system including a processor for supporting terminal devices or network devices in implementing the functions involved in the above aspects, such as transmitting or processing data and / or information involved in the above methods. In one possible design, the chip system also includes a memory for storing necessary program instructions and data for the terminal device or network device. The chip system may be composed of chips or may include chips and other discrete devices.

[0249] In the embodiments of this application, the terms and English abbreviations are exemplary examples given for ease of description and should not be construed as limiting the application in any way. This application does not preclude the possibility of defining other terms that can achieve the same or similar functions in existing or future agreements.

[0250] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated.

[0251] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0252] It should be understood that in the various embodiments of this application, the sequence number of each process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0253] In summary, the above are merely preferred embodiments of the technical solutions of this application and are not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A communication method characterized by comprising: Applied to a first network device, the method includes: Send first information, the first information including arrival time, arrival angle and first identifier; the arrival time and arrival angle are the arrival time and arrival angle of the signal sent from the predicted location received by the first network device; the predicted location is the position of the target terminal predicted by the first network device after a preset time; the first identifier is the identifier of the arrival time and the arrival angle. Send a second message, which instructs a second network device to generate a target execution action corresponding to the first identifier; The third information is received, which instructs the target to perform an action. The target's action is determined by the second network device based on the confidence level of the predicted location and the correspondence between the confidence level and the action. The confidence level of the predicted location is determined based on the arrival time, the arrival angle, and a two-stream neural network.

2. The method according to claim 1, characterized in that, The output of the two-stream neural network is a heteroscedasticity covariance matrix, which is used to map the confidence level of the predicted location.

3. The method according to claim 2, characterized in that, The dual-stream neural network includes a first encoder and a second encoder; The first encoder is used to take the arrival time as input and output a Gaussian parameter corresponding to the arrival time; The second encoder is used to take the arrival angle as input and output Gaussian parameters corresponding to the arrival angle; the Gaussian parameters corresponding to the arrival time and the Gaussian parameters corresponding to the arrival angle are used to determine the heteroscedasticity covariance matrix.

4. The method according to claim 1, characterized in that, If the confidence level is within a preset range and the duration exceeds a first duration, the target action is determined by the second network device based on the confidence level of the predicted location and the correspondence between the confidence level and the action.

5. The method according to claim 1 or claim 4, characterized in that, The correspondence between the confidence level and the executed action includes: If the confidence level is lower than the first threshold, the target action is to re-predict the confidence level using the arrival time or the arrival angle, or to delay the retest. If the confidence level is higher than the first threshold and lower than the second threshold, the target performs the action of increasing the positioning reference signal density, deepening the beam, or increasing the beam gain. If the confidence level is higher than the second threshold, the target action is to maintain the current state.

6. The method according to any one of claims 1 to 4, characterized in that, The arrival time and the arrival angle are determined based on an unscented Kalman filter and a first reference signal, which is a positioning reference signal or a detection reference signal.

7. The method according to claim 6, characterized in that, The arrival time is determined based on the Euclidean distance between the position vector of the predicted location and the position vector of the first network device, the measurement noise of the arrival time, and the ranging bias term of the arrival time; the position vector of the predicted location is determined based on the unscented Kalman filter and the first reference signal.

8. The method according to claim 6, characterized in that, The angle of arrival is determined based on the location of the first network device, the predicted location, the measurement noise of the angle of arrival, and the systematic bias in the angle of arrival; the predicted location is determined based on the unscented Kalman filter and the first reference signal.

9. The method according to claim 6, characterized in that, The method further includes: After generating the arrival time and the arrival angle, the Kalman filter is corrected based on the first reference signal.

10. The method according to claim 9, characterized in that, The correction of the Kalman filter based on the first reference signal includes: The Kalman gain is calculated based on the state-measurement cross-covariance and the measurement covariance, wherein the measurement covariance is determined based on the first reference signal, and the state-measurement cross-covariance is determined based on the sampling points and the nonlinear motion model. The predicted state is corrected based on the Kalman gain and the predicted state covariance to obtain the posterior state estimate; the predicted state covariance is determined based on the sampling points and the nonlinear motion model.

11. The method according to claim 1, characterized in that, The first identifier is the measurement window number for the arrival time and the arrival angle, as well as a timestamp.

12. A communication method, characterized in that, Applied to a second network device, the method includes: The system receives first information, which includes arrival time, angle of arrival, and a first identifier. The arrival time and angle of arrival are the arrival time and angle of arrival of the signal transmitted at the predicted location received by the first network device. The predicted location is the position of the target terminal predicted by the first network device after a preset time. The first identifier is an identifier for the arrival time and the angle of arrival. Receive second information, the second information instructing the second network device to generate a target execution action corresponding to the first identifier; The target action is determined based on the confidence level of the predicted location and the correspondence between the confidence level and the action to be performed; the confidence level of the predicted location is determined based on the arrival time, the arrival angle, and a two-stream neural network. A third message is sent, which instructs the target to perform an action.

13. The method according to claim 12, characterized in that, The correspondence between the confidence level and the action to be performed is pre-configured in the second network device.

14. The method according to claim 12 or 13, characterized in that, The correspondence between the confidence level and the executed action includes: If the confidence level is lower than the first threshold, the target action is to re-predict the confidence level using the arrival time or the arrival angle, or to delay the retest. If the confidence level is higher than the first threshold and lower than the second threshold, the target performs the action of increasing the positioning reference signal density, deepening the beam, or increasing the beam gain. If the confidence level is higher than the second threshold, the target action is to maintain the current state.

15. The method according to claim 12, characterized in that, The output of the two-stream neural network is a heteroscedasticity covariance matrix, which is used to map the confidence level of the predicted location.

16. The method according to claim 15, characterized in that, The dual-stream neural network includes a first encoder and a second encoder; The first encoder is used to take the arrival time as input and output a Gaussian parameter corresponding to the arrival time; The second encoder is used to take the arrival angle as input and output Gaussian parameters corresponding to the arrival angle; the Gaussian parameters corresponding to the arrival time and the Gaussian parameters corresponding to the arrival angle are used to determine the heteroscedasticity covariance matrix.

17. A communication device, characterized in that, The device includes a processor and an interface circuit, the interface circuit being used to receive signals from other communication devices and transmit them to the processor or to send signals from the processor to other communication devices, the processor being used through logic circuits or executing code instructions to implement the method as claimed in any one of claims 1 to 11, or any one of claims 12 to 16.

18. A computer-readable storage medium, characterized in that, The storage medium stores a computer program or instructions that, when executed by a communication device, implement the method as claimed in any one of claims 1 to 11, or any one of claims 12 to 16.

19. A computer program product, characterized in that, Includes instructions that, when executed, cause the method as claimed in any one of claims 1 to 11, or any one of claims 12 to 16, to be implemented.

Images