Positioning method, electronic device, program product, and storage medium

Abstract

The application discloses a positioning method, an electronic device, a program product and a storage medium. The positioning method comprises the following steps: acquiring a received signal received by an access point, wherein a sensing area of the access point is divided into a plurality of subareas; preprocessing the received signal by using a preset signal model to obtain a to-be-predicted signal; determining a target subarea corresponding to the to-be-predicted signal from the sensing area according to the to-be-predicted signal and a pre-trained area selection network; determining a target position prediction model corresponding to the target subarea, and inputting the to-be-predicted signal into the target position prediction model to obtain position data, wherein the plurality of subareas correspond to the plurality of position prediction models in a one-to-one manner. The application solves the technical problems of complex data collection and large calculation amount in the prior art positioning method.

Description

Technical Field

[0001] This invention relates to the field of signal processing technology, and more specifically, to a positioning method, electronic device, program product, and storage medium. Background Technology

[0002] Cellular Massive MIMO systems are widely used in 5G networks. Their superior performance lies in the fact that, under the same frequency resources and power conditions, they can provide much more channel capacity than traditional MIMO systems, thereby supporting more terminal access and higher service rates.

[0003] Massive MIMO systems primarily rely on multi-antenna array receivers to coherently process multipath signals, achieving advantages in frequency selectivity, spatial selectivity, and angular selectivity. These receivers then spatially multiplex and enable multiple access for multiple users. In sensor networks, target localization is typically achieved through ranging and angle measurement. Despite attempts to address the practical challenges of target localization, issues such as imperfect time synchronization, sensor position perturbations, and gain-phase mismatch in multi-channel receivers hinder further improvements in positioning accuracy. To overcome these challenges, deep learning, with its powerful nonlinear fitting capabilities, has identified a new paradigm for achieving high-precision localization. However, deep learning requires pre-training on large-scale data offline and its application across wide sensing areas leads to complex data acquisition and a significant computational burden.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This invention provides a positioning method, an electronic device, a program product, and a storage medium to at least solve the technical problems of complex data acquisition and large computational load in existing positioning methods.

[0006] According to one aspect of the present invention, a positioning method is provided, comprising: acquiring a received signal received by an access point, wherein the sensing area of ​​the access point is divided into multiple sub-regions; preprocessing the received signal using a preset signal model to obtain a signal to be predicted; determining a target sub-region corresponding to the signal to be predicted from the sensing area based on the signal to be predicted and a pre-trained region selection network; determining a target location prediction model corresponding to the target sub-region, and inputting the signal to be predicted into the target location prediction model to predict location data, wherein the multiple sub-regions correspond one-to-one with multiple location prediction models.

[0007] Optionally, the construction of the preset signal model includes: determining the communication signal expression based on the user equipment configuration information and the preset access point configuration information of the receiving access point, wherein the user equipment is included in the non-cellular MIMO network; determining the sensing signal expression based on the preset access point configuration information of the transmitting access point and the receiving access point, wherein the transmitting access point and the receiving access point are included in the non-cellular MIMO network; and determining the preset signal model based on the communication signal expression, the sensing signal expression, and additive noise.

[0008] Optionally, based on the user equipment configuration information and the preset access point configuration information of the receiving access point, the communication signal expression is determined, including: based on the communication channel representation and communication signal representation of the user equipment and the receiving access point, the communication channel representation and communication signal representation are included in the user equipment configuration information and the preset access point configuration information.

[0009] Optionally, based on the preset configuration information of the transmitting access point and the receiving access point, the sensing signal expression is determined, including: based on the sensing channel representation and sensing signal representation of the transmitting access point and the receiving access point, the sensing channel representation and sensing signal representation are included in the preset access point configuration information.

[0010] Optionally, the sensing channel representation includes: signal attenuation at the transmitting access point and the receiving access point, bistatic distance between the transmitting access point and the receiving access point, steering vector of the transmitting access point and the receiving access point, subcarrier spacing, speed of light, coherent processing interval, Doppler frequency shift, and carrier wavelength.

[0011] Optionally, the received signal is preprocessed using a preset signal model to obtain the signal to be predicted, including: determining multiple sensing signals from multiple received signals based on the preset signal model; determining a signal matrix based on the multiple sensing signals, wherein the signal matrix is ​​a complex matrix; and dividing the real and imaginary parts of the signal matrix to determine the signal to be predicted.

[0012] Optionally, based on the signal to be predicted and the pre-trained region selection network, the target sub-region corresponding to the signal to be predicted is determined from the perceived region, including: inputting the signal to be predicted into the feature extraction module in the region selection network for feature extraction to obtain the feature extraction result; inputting the feature extraction result into the regression module in the region selection network to determine the target sub-region.

[0013] Optionally, a target location prediction model corresponding to the target sub-region is determined, and the signal to be predicted is input into the target location prediction model to predict the location data. This includes: determining the target location prediction model from multiple location prediction models based on the target sub-region; inputting the signal to be predicted into the spatial attention module in the target location prediction model, using the spatial attention module to perform weighted processing on the signal to be predicted to obtain weighted features; and inputting the weighted features into the residual convolutional neural network in the target location prediction model for prediction to obtain the location data.

[0014] Optionally, the spatial attention module includes a pooling layer, a convolutional layer, and an activation function. The signal to be predicted is input into the spatial attention module in the target location prediction model. The spatial attention module performs weighted processing on the signal to be predicted to obtain weighted features, including: using a pooling layer to pool the signal to be predicted to obtain a pooling result; using a convolutional layer to perform a convolution operation on the pooling result to obtain a convolution result; using an activation function to impose weight restrictions on the convolution result based on a preset weight range to obtain spatial attention weights; and multiplying the spatial attention weights element-wise with the signal to be predicted to obtain weighted features.

[0015] Optionally, the target location prediction model includes a first convolutional module, a second convolutional module, and a regression module. Weighted features are input into the residual convolutional neural network in the target location prediction model for prediction to obtain location data. This includes: using the first convolutional module to perform a first processing on the weighted features to obtain a first processing result, wherein the first processing sequentially includes convolution processing, normalization processing, activation processing, and pooling processing; using the second convolutional module to perform a second processing on the first processing result to obtain a second processing result, wherein the second processing includes main branch operations and auxiliary branch operations; and using the regression module to predict the second processing result to obtain the location data.

[0016] Optionally, the second processing result is processed by the second convolution module to obtain the second processing result, including: performing a main branch operation on the first processing result using the second convolution module to obtain a main branch operation result, wherein the main branch operation sequentially includes: first convolution processing, normalization processing, activation processing, second convolution processing, and second normalization processing; performing an auxiliary branch operation on the first processing result using the second convolution module to obtain an auxiliary branch operation result, wherein the auxiliary branch operation includes: convolution processing based on a preset dimension; and determining the second processing result based on the main branch operation result and the auxiliary branch operation result.

[0017] According to another aspect of the present invention, an electronic device is also provided, comprising: a memory storing an executable program; and a processor for running the program, wherein the program executes the positioning method described in any of the preceding embodiments.

[0018] According to another aspect of the present invention, a computer program product is also provided, including a computer program that, when executed by a processor, implements the positioning method described in any of the above embodiments.

[0019] According to another aspect of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is executed, it controls the device where the computer-readable storage medium is located to perform the positioning method described above.

[0020] In this embodiment of the invention, a received signal from an access point is acquired, wherein the access point's sensing area is divided into multiple sub-regions; the received signal is preprocessed using a preset signal model to obtain a signal to be predicted; based on the signal to be predicted and a pre-trained region selection network, a target sub-region corresponding to the signal to be predicted is determined from the sensing area; a target location prediction model corresponding to the target sub-region is determined, and the signal to be predicted is input into the target location prediction model to predict location data, wherein multiple sub-regions correspond one-to-one with multiple location prediction models. This invention, by dividing the sensing area into multiple regions using a region selection network and employing different positioning models for different regions, solves the problem of data acquisition and computational burden caused by applying deep learning to achieve target localization within a wide sensing area, thereby addressing the technical problems of complex data acquisition and high computational load in existing localization methods. Attached Figure Description

[0021] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0022] Figure 1 This is a flowchart of a positioning method according to one embodiment of the present invention;

[0023] Figure 2 This is a schematic diagram of a non-cellular massive MIMO system according to one embodiment of the present invention;

[0024] Figure 3 This is a schematic diagram of data preprocessing according to one embodiment of the present invention;

[0025] Figure 4 This is a schematic diagram of a region selection network according to one embodiment of the present invention;

[0026] Figure 5 This is a schematic diagram of a spatial attention mechanism according to one embodiment of the present invention;

[0027] Figure 6This is a schematic diagram of a residual convolutional neural network according to one embodiment of the present invention;

[0028] Figure 7 This is a flowchart illustrating a positioning method according to one embodiment of the present invention;

[0029] Figure 8 This is a schematic diagram comparing the performance of a region selection network at different signal-to-noise ratios according to one embodiment of the present invention;

[0030] Figure 9 This is a schematic diagram comparing the performance of a region selection network at different grid intervals according to one embodiment of the present invention;

[0031] Figure 10 This is a schematic diagram comparing the RMSE of a localization method according to one embodiment of the present invention with other methods at different signal-to-noise ratios;

[0032] Figure 11 This is a schematic diagram comparing the CDF of the positioning method according to one embodiment of the present invention with that of other methods.

[0033] Figure 12 This is a structural block diagram of a collaborative sensing system according to one embodiment of the present invention. Detailed Implementation

[0034] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0035] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0036] According to an embodiment of the present invention, an embodiment of a positioning method is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system containing at least one set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0037] This method embodiment can also be executed in an electronic device including a memory and a processor, a similar control device, or in the cloud. Taking an electronic device as an example, the electronic device may include one or more processors and a memory for storing data. Optionally, the electronic device may also include a communication device for communication functions and a display device. Those skilled in the art will understand that the above structural description is merely illustrative and does not limit the structure of the electronic device. For example, the electronic device may include more or fewer components than described above, or have a different configuration than described above.

[0038] A processor may include one or more processing units. For example, a processor may include a central processing unit (CPU), a graphics processing unit (GPU), a digital signal processing (DSP) chip, a microcontroller unit (MCU), a field-programmable gate array (FPGA), a neural network processing unit (NPU), a tensor processing unit (TPU), or an artificial intelligence (AI) processor. Different processing units may be independent components or integrated into one or more processors. In some instances, electronic devices may also include one or more processors.

[0039] The memory can be used to store computer programs, such as the computer program corresponding to the positioning method in this embodiment of the invention. The processor implements the positioning method by running the computer program stored in the memory. The memory may include high-speed random access memory and non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to electronic devices via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0040] The communication device is used to receive or send data via a network. Specific examples of the network mentioned above may include a wireless network provided by the mobile terminal's communication provider. In one example, the communication device includes a network interface controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the communication device may be a radio frequency (RF) module, used for wireless communication with the Internet. In some embodiments of this solution, the communication device is used to connect to mobile devices such as mobile phones and tablets, enabling the mobile device to send commands to the electronic device.

[0041] The display device can be a touchscreen liquid crystal display (LCD) or a touch display (also referred to as a "touchscreen" or "touch display screen"). The LCD allows a user to interact with the user interface of the electronic device. In some embodiments, the electronic device has a graphical user interface (GUI), which allows the user to interact with the GUI by touching a touch-sensitive surface with fingers and / or gestures. Executable instructions for performing these human-computer interaction functions are configured / stored in one or more processor-executable computer program products or readable storage media.

[0042] Figure 1 This is a flowchart of a positioning method according to one embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0043] Step S101: Obtain the received signal received by the access point, wherein the sensing area of ​​the access point is divided into multiple sub-areas.

[0044] Specifically, in wireless communication, an access point (AP) refers to a device that can connect wireless devices to a network, such as a Wi-Fi router. Received signals refer to the wireless signals that the access point can detect; these signals can come from mobile devices, sensors, etc. The sensing area refers to the geographical range within which the access point can detect wireless signals. To improve positioning accuracy, the sensing area is divided into smaller sub-regions, each with its own characteristics and boundaries.

[0045] Step S102: Using a preset signal model, the received signal is preprocessed to obtain the signal to be predicted.

[0046] Specifically, the preset signal model is a mathematical model pre-defined based on the physical environment and wireless propagation characteristics, used to describe the signal's propagation pattern in space. Preprocessing refers to performing necessary mathematical processing on the received signal to eliminate noise, enhance signal characteristics, etc., making it more suitable for subsequent location prediction. The preprocessed signal is then used for the next step of location prediction.

[0047] Step S103: Based on the signal to be predicted and the pre-trained region selection network, determine the target sub-region corresponding to the signal to be predicted from the perception region.

[0048] Specifically, a region selection network is a pre-trained machine learning model that can identify the sub-region from which the signal originates based on the signal characteristics of the input signal to be predicted. After analysis by the region selection network, the specific sub-region from which the signal to be predicted originates is determined and designated as the target sub-region.

[0049] Step S104: Determine the target location prediction model corresponding to the target sub-region, and input the signal to be predicted into the target location prediction model to predict the location data. Here, multiple sub-regions correspond one-to-one with multiple location prediction models.

[0050] Specifically, the target location prediction model is a location prediction model corresponding to the target sub-region. Each sub-region has its own corresponding location prediction model, which is trained based on the wireless signal propagation characteristics of different sub-regions. When the signal to be predicted is input into the location prediction model corresponding to the target sub-region, the model will output a location prediction result, i.e., location data. This result is obtained based on signal characteristics and sub-region characteristics.

[0051] Understandably, each sub-region has its own location prediction model, which allows for more accurate location prediction based on signal characteristics.

[0052] In this embodiment of the invention, a received signal from an access point is acquired, wherein the access point's sensing area is divided into multiple sub-regions; the received signal is preprocessed using a preset signal model to obtain a signal to be predicted; based on the signal to be predicted and a pre-trained region selection network, a target sub-region corresponding to the signal to be predicted is determined from the sensing area; a target location prediction model corresponding to the target sub-region is determined, and the signal to be predicted is input into the target location prediction model to predict location data, wherein multiple sub-regions correspond one-to-one with multiple location prediction models. This invention, by dividing the sensing area into multiple regions using a region selection network and employing different positioning models for different regions, solves the problem of data acquisition and computational burden caused by applying deep learning to achieve target localization within a wide sensing area, thereby addressing the technical problems of complex data acquisition and high computational load in existing localization methods.

[0053] Optionally, the construction of the preset signal model includes: determining the communication signal expression based on the user equipment configuration information and the preset access point configuration information of the receiving access point, wherein the user equipment is included in the non-cellular MIMO network; determining the sensing signal expression based on the preset access point configuration information of the transmitting access point and the receiving access point, wherein the transmitting access point and the receiving access point are included in the non-cellular MIMO network; and determining the preset signal model based on the communication signal expression, the sensing signal expression, and additive noise.

[0054] Specifically, such as Figure 2 As shown, consider a cellular-free massive MIMO system consisting of U single-antenna user equipment (UEs) and N randomly distributed access points (APs), where each AP is equipped with an M antenna per line. x Antenna and each column of M y In a uniform planar array (UPA) of antennas, the total number of antennas per AP is M = M x M y Assuming T APs transmit sensing signals and R APs simultaneously receive sensing signals and communication signals from UEs, then the received signal of the r-th receiving AP on the k-th subcarrier and the l-th symbol... for:

[0055]

[0056] in, This represents the communication channel of the u-th UE. For the sensing channel from the t-th transmitting AP to the r-th receiving AP, s u(k,l) and Let these represent the communication signal from the u-th UE and the sensing signal from the t-th transmitting AP, respectively. It is additive noise that follows a complex Gaussian distribution.

[0057] It should be noted that the above formula (1) is the preset signal model, which includes the communication signal expression, the sensing signal expression and additive noise.

[0058] Optionally, based on the user equipment configuration information and the preset access point configuration information of the receiving access point, the communication signal expression is determined, including: based on the communication channel representation and communication signal representation of the user equipment and the receiving access point, the communication channel representation and communication signal representation are included in the user equipment configuration information and the preset access point configuration information.

[0059] Specifically, the communication signal expression is as follows:

[0060] Optionally, based on the preset configuration information of the transmitting access point and the receiving access point, the sensing signal expression is determined, including: based on the sensing channel representation and sensing signal representation of the transmitting access point and the receiving access point, the sensing channel representation and sensing signal representation are included in the preset access point configuration information.

[0061] Specifically, the expression for the sensing signal is:

[0062] Optionally, by separating the sensing signal and the communication signal using a time-division multiplexing method and applying orthogonal pilot signals, the receiving AP can distinguish sensing signals from different transmitting APs. Simultaneously, to improve the directivity of the received signal, a receiving beamformer is employed on the receiving AP. Therefore, the sensing signal from the t-th transmitting AP to the r-th receiving AP is:

[0063]

[0064] Among them, w H This indicates the conjugate transpose operation. It is additive white Gaussian noise.

[0065] Optionally, the sensing channel representation includes: signal attenuation at the transmitting access point and the receiving access point, bistatic distance between the transmitting access point and the receiving access point, steering vector of the transmitting access point and the receiving access point, subcarrier spacing, speed of light, coherent processing interval, Doppler frequency shift, and carrier wavelength.

[0066] Specifically, the sensing channel is represented as:

[0067]

[0068] Where, α r,t The attenuation is represented from the t-th transmitting AP to the r-th receiving AP, Δf represents the subcarrier spacing, c represents the speed of light, and T represents the speed of light. p f represents the coherent processing interval. d Let λ represent the Doppler frequency shift, and λ represent the carrier wavelength. The numerical calculation formula is as follows: Where f c It is the subcarrier frequency, d r,t The bistatic distance between the t-th transmitting AP and the r-th receiving AP is represented by the formula d. r +d t , where d r and d t Let d represent the distances between the target and the r-th receiving AP, and the distances between the target and the t-th transmitting AP, respectively. Specifically, d r and d t The calculation formula is as follows:

[0069]

[0070] Among them, (x q ,y q ), (x r ,y r ) and (x t ,y t ) represent the coordinates of the target, the r-th receiving AP, and the t-th transmitting AP, respectively. and Let represent the steering vectors of the r-th receiving AP and the t-th transmitting AP, respectively, as follows:

[0071]

[0072] Accompanying,

[0073]

[0074] Among them, d, θ r , and θ t Let represent the antenna spacing of the UPA, the azimuth angle of the r-th receiving AP, the elevation angle of the r-th receiving AP, the azimuth angle of the t-th transmitting AP, and the elevation angle of the t-th transmitting AP, respectively. T and These represent the transpose operation and the Kronecker product, respectively.

[0075] Optionally, the received signal is preprocessed using a preset signal model to obtain the signal to be predicted, including: determining multiple sensing signals from multiple received signals based on the preset signal model; determining a signal matrix based on the multiple sensing signals, wherein the signal matrix is ​​a complex matrix; and dividing the real and imaginary parts of the signal matrix to determine the signal to be predicted.

[0076] For example, the perceptual region is divided along the XY plane as A = {A1, A2, ..., A...} S}, where A represents the set of subregions, and each element represents a subregion that has been divided. For example... Figure 3 As shown, assume the sensing signal determined from the received signals of the t-th transmitting AP and the r-th receiving AP is... Stack all signals to form a signal matrix Where KL and TR are obtained by multiplying the subcarriers and symbols, and by multiplying the number of transmitting APs and receiving APs, respectively. Signal Matrix The signal matrix is ​​a complex matrix, but deep learning algorithms require real numbers as input. Therefore, the signal matrix is... The real and imaginary parts are divided into two channels for the input feature, represented as... This is the signal to be predicted.

[0077] Optionally, based on the signal to be predicted and the pre-trained region selection network, the target sub-region corresponding to the signal to be predicted is determined from the perceived region, including: inputting the signal to be predicted into the feature extraction module in the region selection network for feature extraction to obtain the feature extraction result; inputting the feature extraction result into the regression module in the region selection network to determine the target sub-region.

[0078] Specifically, such as Figure 4 As shown, the region selection network aims to identify the sub-region to which a target belongs, and includes a feature extraction module and a regression module. The feature extraction module consists of three convolutional layers, each including a convolution operation, batch normalization, an activation function, and max pooling. The regression module includes global average pooling and fully connected layers, and the output label is... The elements corresponding to the sub-region to which the target belongs are set to 1, and the rest are set to 0.

[0079] Optionally, a target location prediction model corresponding to the target sub-region is determined, and the signal to be predicted is input into the target location prediction model to predict the location data. This includes: determining the target location prediction model from multiple location prediction models based on the target sub-region; inputting the signal to be predicted into the spatial attention module in the target location prediction model, using the spatial attention module to perform weighted processing on the signal to be predicted to obtain weighted features; and inputting the weighted features into the residual convolutional neural network in the target location prediction model for prediction to obtain the location data.

[0080] Specifically, such as Figure 5 As shown, the application of spatial attention mechanisms aims to assign appropriate target localization contribution weights to different transceiver stations. (Connection) The outputs of max pooling and average pooling are used as input to the convolutional layer, and then a sigmoid activation function is applied to constrain the attention weights between 0 and 1. Perform element-wise multiplication to obtain the weighted characteristics.

[0081] Finally, the weighted features output by the spatial attention mechanism are used as input to the residual convolutional neural network, aiming to output the target's location coordinates. For example... Figure 6 As shown, the residual convolutional neural network includes a basic convolutional module, a refined convolutional module, and a regression module. The basic convolutional module consists of a single convolutional layer, including convolution operations, batch normalization, activation functions, and max pooling. The refined convolutional module contains three residual blocks. The input features of each residual block undergo a sequential process: first, a convolution operation; second, batch normalization and a ReLU activation function; and third, another convolution operation and batch normalization are applied. This is called the main branch operation. The auxiliary branch operation has two methods: if the stride is 1, the input and output features of the residual block have the same dimension, allowing them to be directly connected; if the stride is 2, the input features of the residual block are processed by a convolution operation with a kernel size of [1,1], resulting in the same dimension as the output features. The output of the main branch and the output of the auxiliary branch are added element-wise to represent the result of residual learning. Finally, the result of residual learning is passed through a ReLU activation function.

[0082] Optionally, the spatial attention module includes a pooling layer, a convolutional layer, and an activation function. The signal to be predicted is input into the spatial attention module in the target location prediction model. The spatial attention module performs weighted processing on the signal to be predicted to obtain weighted features, including: using a pooling layer to pool the signal to be predicted to obtain a pooling result; using a convolutional layer to perform a convolution operation on the pooling result to obtain a convolution result; using an activation function to impose weight restrictions on the convolution result based on a preset weight range to obtain spatial attention weights; and multiplying the spatial attention weights element-wise with the signal to be predicted to obtain weighted features.

[0083] Specifically, such as Figure 5 As shown, the signal to be predicted The outputs of max pooling and average pooling layers are concatenated and used as input to a convolutional layer. A sigmoid activation function is then applied to constrain the attention weights of the convolutional layer output to between 0 and 1. These attention weights are then compared with... Perform element-wise multiplication to obtain the weighted characteristics.

[0084] Optionally, the target location prediction model includes a first convolutional module, a second convolutional module, and a regression module. Weighted features are input into the residual convolutional neural network in the target location prediction model for prediction to obtain location data. This includes: using the first convolutional module to perform a first processing on the weighted features to obtain a first processing result, wherein the first processing sequentially includes convolution processing, normalization processing, activation processing, and pooling processing; using the second convolutional module to perform a second processing on the first processing result to obtain a second processing result, wherein the second processing includes main branch operations and auxiliary branch operations; and using the regression module to predict the second processing result to obtain the location data.

[0085] Specifically, such as Figure 6 As shown, the weighted features are first processed using a basic convolutional module (i.e., the first convolutional module). Then, the output of the basic convolutional module is processed using a refined convolutional module (i.e., the second convolutional module). Finally, the output of the refined convolutional module is predicted using a regression module to obtain the target's location coordinates, i.e., location data.

[0086] Optionally, the second processing result is processed by the second convolution module to obtain the second processing result, including: performing a main branch operation on the first processing result using the second convolution module to obtain a main branch operation result, wherein the main branch operation sequentially includes: first convolution processing, normalization processing, activation processing, second convolution processing, and second normalization processing; performing an auxiliary branch operation on the first processing result using the second convolution module to obtain an auxiliary branch operation result, wherein the auxiliary branch operation includes: convolution processing based on a preset dimension; and determining the second processing result based on the main branch operation result and the auxiliary branch operation result.

[0087] Specifically, such as Figure 6 As shown, the refined convolutional module (i.e., the second convolutional module) contains three residual blocks. The processing of the input features of each residual block is as follows: First, a convolution operation (i.e., the first convolution processing) is performed; second, batch normalization and ReLU activation function processing (i.e., normalization and activation processing) are performed; then, another convolution operation and batch normalization are used (i.e., the second convolution processing and the second normalization processing). This process is called the main branch operation. The auxiliary branch operation has two methods: if the stride is 1, the input and output features of the residual block have the same dimension, allowing them to be directly connected; if the stride is 2, the input features of the residual block are processed by a convolution operation with a kernel size of [1,1], resulting in the same dimension as the output features. The output of the main branch and the output of the auxiliary branch are added element-wise to represent the result of residual learning. Finally, the result of residual learning is passed through a ReLU activation function to obtain the second processing result.

[0088] For example, such as Figure 7As shown, the positioning method provided by this invention is as follows:

[0089] When target localization is required, the received signal is first preprocessed to obtain input features. The target's sub-region is determined based on a pre-trained region selection network. Then, within the determined sub-region, the target's position coordinates are output based on a pre-trained spatial attention mechanism and a residual convolutional neural network, as follows:

[0090] The received signal is preprocessed to determine the sensing signal; the sensing area is divided into S sub-regions along the XY plane; the sensing signal is processed to obtain the signal matrix. signal matrix Convert to That is, for the signal matrix The real and imaginary parts are divided into two channels for the input feature, represented as... Acquire training data and divide the training and test sets in an 8:2 ratio; set loss functions l1 and l2, batch sizes B1 and B2, iteration counts E1 and E2, and optimizers O1 and O2; train a region selection network based on l1, B1, E1, and O1; train a spatial attention mechanism and a residual convolutional neural network based on l2, B2, E2, and O2; utilize the trained network to achieve target localization, specifically, through the trained region selection network... Determine the sub-region of the target, and then, based on a trained spatial attention mechanism and a residual convolutional neural network,... Achieve precise target positioning; output target coordinates.

[0091] For example, the positioning method provided by the present invention is simulated as follows:

[0092] First, the simulation experiment parameters are constructed, as shown in Table 1 below:

[0093] Table 1

[0094] parameter value carrier frequency 30GHz Subcarrier spacing 60kHz Number of subcarriers 1024 Symbolic Number 2 Antenna spacing half wavelength Sending AP coordinates (4.8,1.2)，(5.2,9.2) Coordinates of the receiving AP (2.4,0)，(7.2,0)，(2.8,10)，(7.6,10)

[0095] The simulation constructs a scenario with a coverage area of ​​10m × 10m, in which two transmitting APs and four receiving APs are deployed in specific locations, and each AP is equipped with an M... x =8 and M yThe UPA antenna array with a resolution of 4 is shown in Table 1, along with the system and waveform parameters used in the simulation. In the data preprocessing stage, the large sensing area is first divided into four equal-sized sub-regions. Then, the entire area is uniformly divided at 0.1m intervals, and the received signals at these points are labeled with indices corresponding to the sub-regions, thus constructing a dataset of size 10,000 for the region selection network. Furthermore, to achieve high-precision target localization, points are uniformly selected at 0.025m intervals within each sub-region, and the received signals at these points are labeled with corresponding coordinates, thus constructing a dataset of size 4×40,000 for the spatial attention mechanism and the residual convolutional neural network. In the region selection network, the kernel size, stride, and padding of the three convolutional operations are [3,3], 1, and 2, respectively, with output channels of 32, 64, and 128, respectively. In the spatial attention mechanism, the kernel size and padding of the convolutional operations are [1,7] and [0,3], respectively. In the residual convolutional neural network, the kernel size for all convolutional operations is [3,3]. A stride of 2 is chosen to connect the residual blocks, and the stride for the auxiliary channel convolutional operations is 4. The network parameters are updated using the Adam optimizer, and cross-entropy loss and minimum mean square error loss are used to handle the region selection and object localization tasks, respectively. The learning rate and mini-batch size are set to 0.001 and 40, respectively. Furthermore, the learning rate is reduced by a factor of 0.15 every 25 epochs.

[0096] Figure 8 , Figure 9 The effects of different signal-to-noise ratios and grid spacings on the training accuracy of region selection networks are shown. Figure 8 It can be seen that training accuracy is positively correlated with signal-to-noise ratio (SNR), especially in the 10dB–15dB range, where it increases by approximately 5.63%. Furthermore, the convergence speed is unaffected by SNR. Meanwhile, at an SNR of 15dB, the region selection network achieves a training accuracy of 96.65% and a test accuracy of 96.52%, demonstrating excellent classification accuracy and strong generalization ability. In addition, Figure 9 The effect of grid spacing on training accuracy is shown when the signal-to-noise ratio is fixed at 15 dB. Clearly, performance improves as the grid spacing decreases. However, a smaller grid spacing means more data is required, leading to higher deployment costs. An optimal trade-off between classification accuracy and cost should be sought based on actual requirements.

[0097] To further compare the performance of target localization, the proposed method is compared with BackProjection, Transformer, and FedSLoc. Figures 10 to 11 This paper presents a performance comparison of different methods from multiple perspectives. Notably, the proposed method is named RSAResNet. Figure 10It can be seen that the RMSE of all four methods decreases with the increase of signal-to-noise ratio. Specifically, Back Projection performs the worst, while FedSLoc performs slightly better than Transformer. The proposed method is significantly better than the other three methods, improving the positioning accuracy by 68.2%, 49.6%, and 47.1%, respectively. This indicates that the proposed method performs well in high-precision target positioning. Figure 11 The cumulative distribution function (CDF) of the estimation errors of different methods is described for sub-region 1 and a signal-to-noise ratio of 15 dB. It can be seen that the other three methods exhibit similar positioning accuracy, but they all lag significantly behind the proposed method, further strengthening the validation of the effectiveness of the proposed method.

[0098] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0099] This invention also provides a collaborative sensing system for implementing the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0100] Figure 12 This is a structural block diagram of a collaborative sensing system 200 according to one embodiment of the present invention, such as... Figure 12As shown, the system includes: an acquisition module 201, used to acquire the received signal received by the access point, wherein the sensing area of ​​the access point is divided into multiple sub-regions; a preprocessing module 202, used to preprocess the received signal using a preset signal model to obtain the signal to be predicted; a determination module 203, used to determine the target sub-region corresponding to the signal to be predicted from the sensing area based on the signal to be predicted and a pre-trained region selection network; and a prediction module 204, used to determine the target position prediction model corresponding to the target sub-region, and input the signal to be predicted into the target position prediction model to predict the position data, wherein multiple sub-regions correspond one-to-one with multiple position prediction models.

[0101] Optionally, the preprocessing module 202 is further configured to: determine a communication signal expression based on user equipment configuration information and preset access point configuration information of the receiving access point, wherein the user equipment is included in a non-cellular MIMO network; determine a sensing signal expression based on preset access point configuration information of the transmitting and receiving access points, wherein the transmitting and receiving access points are included in a non-cellular MIMO network; and determine a preset signal model based on the communication signal expression, the sensing signal expression, and additive noise.

[0102] Optionally, the preprocessing module 202 is further configured to: determine a communication signal expression based on the communication channel representation and communication signal representation of the user equipment and the receiving access point, wherein the communication channel representation and communication signal representation are included in the user equipment configuration information and the preset access point configuration information.

[0103] Optionally, the preprocessing module 202 is further configured to: determine a sensing signal expression based on the sensing channel representation and sensing signal representation of the transmitting access point and the receiving access point, wherein the sensing channel representation and sensing signal representation are included in preset access point configuration information.

[0104] Optionally, the sensing channel representation determined by the preprocessing module 202 includes: signal attenuation of the transmitting access point and the receiving access point, bistatic distance between the transmitting access point and the receiving access point, steering vector of the transmitting access point and the receiving access point, subcarrier spacing, speed of light, coherent processing interval, Doppler frequency shift, and carrier wavelength.

[0105] Optionally, the preprocessing module 202 is further configured to: determine multiple sensing signals from multiple received signals based on a preset signal model; determine a signal matrix based on the multiple sensing signals, wherein the signal matrix is ​​a complex matrix; and divide the real and imaginary parts of the signal matrix to determine the signal to be predicted.

[0106] Optionally, the determining module 203 is further configured to: extract features from the feature extraction module in the input region of the signal to be predicted to obtain the feature extraction result; and input the feature extraction result into the regression module in the region selection network to determine the target sub-region.

[0107] Optionally, the prediction module 204 is further configured to: determine a target location prediction model from multiple location prediction models based on the target sub-region; input the signal to be predicted into the spatial attention module in the target location prediction model, use the spatial attention module to perform weighted processing on the signal to be predicted to obtain weighted features; input the weighted features into the residual convolutional neural network in the target location prediction model for prediction to obtain location data.

[0108] Optionally, the spatial attention module includes a pooling layer, a convolutional layer, and an activation function. The prediction module 204 is further configured to: use the pooling layer to perform pooling processing on the signal to be predicted to obtain the pooling processing result; use the convolutional layer to perform a convolution operation on the pooling processing result to obtain the convolution processing result; use the activation function to impose weight restrictions on the convolution processing result based on a preset weight range to obtain spatial attention weights; and multiply the spatial attention weights element-wise with the signal to be predicted to obtain weighted features.

[0109] Optionally, the target location prediction model includes a first convolutional module, a second convolutional module, and a regression module. The prediction module 204 is further configured to: use the first convolutional module to perform a first processing on the weighted features to obtain a first processing result, wherein the first processing sequentially includes convolution processing, normalization processing, activation processing, and pooling processing; use the second convolutional module to perform a second processing on the first processing result to obtain a second processing result, wherein the second processing includes main branch operation and auxiliary branch operation; and use the regression module to predict the second processing result to obtain location data.

[0110] Optionally, the prediction module 204 is further configured to: perform main branch operations on the first processing result using the second convolution module to obtain the main branch operation result, wherein the main branch operation sequentially includes: first convolution processing, normalization processing, activation processing, second convolution processing, and second normalization processing; perform auxiliary branch operations on the first processing result using the second convolution module to obtain the auxiliary branch operation result, wherein the auxiliary branch operation includes: convolution processing based on a preset dimension; and determine the second processing result based on the main branch operation result and the auxiliary branch operation result.

[0111] Embodiments of the present invention also provide an electronic device, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the positioning method described in any of the above embodiments during runtime.

[0112] Optionally, in this embodiment, the processor in the above-described electronic device can be configured to run an executable program to perform the following steps:

[0113] Step S101: Obtain the received signal received by the access point, wherein the sensing area of ​​the access point is divided into multiple sub-areas.

[0114] Step S102: Using a preset signal model, the received signal is preprocessed to obtain the signal to be predicted.

[0115] Step S103: Based on the signal to be predicted and the pre-trained region selection network, determine the target sub-region corresponding to the signal to be predicted from the perception region.

[0116] Step S104: Determine the target location prediction model corresponding to the target sub-region, and input the signal to be predicted into the target location prediction model to predict the location data. Here, multiple sub-regions correspond one-to-one with multiple location prediction models.

[0117] Optionally, the processor in the aforementioned electronic device may be configured to run an executable program to perform the following steps: determining a communication signal expression based on user equipment configuration information and preset access point configuration information of the receiving access point, wherein the user equipment is included in a non-cellular MIMO network; determining a sensing signal expression based on preset access point configuration information of the transmitting and receiving access points, wherein the transmitting and receiving access points are included in a non-cellular MIMO network; and determining a preset signal model based on the communication signal expression, the sensing signal expression, and additive noise.

[0118] Optionally, the processor in the aforementioned electronic device may be configured to run an executable program to perform the following steps: determining a communication signal expression based on the communication channel representation and communication signal representation of the user equipment and the receiving access point, wherein the communication channel representation and communication signal representation are included in the user equipment configuration information and the preset access point configuration information.

[0119] Optionally, the processor in the aforementioned electronic device may be configured to run an executable program to perform the following steps: determining a sensing signal expression based on the sensing channel representation and sensing signal representation of the transmitting access point and the receiving access point, wherein the sensing channel representation and sensing signal representation are contained in preset access point configuration information.

[0120] Optionally, the processor in the aforementioned electronic device executes an executable program to determine the sensing channel representation, which includes: signal attenuation of the transmitting access point and the receiving access point, bistatic distance between the transmitting access point and the receiving access point, steering vector of the transmitting access point and the receiving access point, subcarrier spacing, speed of light, coherent processing interval, Doppler shift, and carrier wavelength.

[0121] Optionally, the processor in the aforementioned electronic device may be configured to run an executable program to perform the following steps: determining multiple sensing signals from multiple received signals based on a preset signal model; determining a signal matrix based on the multiple sensing signals, wherein the signal matrix is ​​a complex matrix; and dividing the real and imaginary parts of the signal matrix to determine the signal to be predicted.

[0122] Optionally, the processor in the aforementioned electronic device can be configured to run an executable program to perform the following steps: extracting features from the feature extraction module in the region selection network to obtain the feature extraction result; and inputting the feature extraction result into the regression module in the region selection network to determine the target sub-region.

[0123] Optionally, the processor in the aforementioned electronic device may be configured to run an executable program to perform the following steps: determining a target location prediction model from multiple location prediction models based on the target sub-region; inputting the signal to be predicted into the spatial attention module in the target location prediction model, using the spatial attention module to perform weighted processing on the signal to be predicted to obtain weighted features; and inputting the weighted features into the residual convolutional neural network in the target location prediction model for prediction to obtain location data.

[0124] Optionally, the spatial attention module includes a pooling layer, a convolutional layer, and an activation function. The processor in the aforementioned electronic device can be configured to run an executable program to perform the following steps: using the pooling layer to perform pooling processing on the signal to be predicted, obtaining a pooling processing result; using the convolutional layer to perform a convolution operation on the pooling processing result, obtaining a convolution processing result; using the activation function to impose weight restrictions on the convolution processing result based on a preset weight range, obtaining spatial attention weights; and multiplying the spatial attention weights element-wise with the signal to be predicted, obtaining weighted features.

[0125] Optionally, the target location prediction model includes a first convolutional module, a second convolutional module, and a regression module. The processor in the aforementioned electronic device can be configured to run an executable program to perform the following steps: using the first convolutional module to perform a first processing on the weighted features to obtain a first processing result, wherein the first processing sequentially includes convolution processing, normalization processing, activation processing, and pooling processing; using the second convolutional module to perform a second processing on the first processing result to obtain a second processing result, wherein the second processing includes main branch operations and auxiliary branch operations; and using the regression module to predict the second processing result to obtain location data.

[0126] Optionally, the processor in the aforementioned electronic device can be configured to run an executable program to perform the following steps: performing a main branch operation on the first processing result using a second convolution module to obtain a main branch operation result, wherein the main branch operation sequentially includes: first convolution processing, normalization processing, activation processing, second convolution processing, and second normalization processing; performing an auxiliary branch operation on the first processing result using the second convolution module to obtain an auxiliary branch operation result, wherein the auxiliary branch operation includes: convolution processing based on a preset dimension; and determining a second processing result based on the main branch operation result and the auxiliary branch operation result.

[0127] Embodiments of the present invention also provide a computer program product, including a computer program that, when executed by a processor, implements the positioning method described in any of the above embodiments.

[0128] Optionally, in this embodiment, the computer program, when executed by the processor, performs the following steps:

[0129] Step S101: Obtain the received signal received by the access point, wherein the sensing area of ​​the access point is divided into multiple sub-areas.

[0130] Step S102: Using a preset signal model, the received signal is preprocessed to obtain the signal to be predicted.

[0131] Step S103: Based on the signal to be predicted and the pre-trained region selection network, determine the target sub-region corresponding to the signal to be predicted from the perception region.

[0132] Step S104: Determine the target location prediction model corresponding to the target sub-region, and input the signal to be predicted into the target location prediction model to predict the location data. Here, multiple sub-regions correspond one-to-one with multiple location prediction models.

[0133] Optionally, when the above computer program is executed by the processor, it performs the following steps: determining a communication signal expression based on user equipment configuration information and preset access point configuration information of the receiving access point, wherein the user equipment is included in a non-cellular MIMO network; determining a sensing signal expression based on preset access point configuration information of the transmitting and receiving access points, wherein the transmitting and receiving access points are included in a non-cellular MIMO network; and determining a preset signal model based on the communication signal expression, the sensing signal expression, and additive noise.

[0134] Optionally, when the above computer program is executed by the processor, it performs the following steps: determining a communication signal expression based on the communication channel representation and communication signal representation of the user equipment and the receiving access point, wherein the communication channel representation and communication signal representation are included in the user equipment configuration information and the preset access point configuration information.

[0135] Optionally, when the above computer program is executed by the processor, it performs the following steps: determining a sensing signal expression based on the sensing channel representation and sensing signal representation of the transmitting access point and the receiving access point, wherein the sensing channel representation and sensing signal representation are included in preset access point configuration information.

[0136] Optionally, the sensing channel representation determined by the computer program when executed by the processor includes: signal attenuation of the transmitting access point and the receiving access point, bistatic distance between the transmitting access point and the receiving access point, steering vector of the transmitting access point and the receiving access point, subcarrier spacing, speed of light, coherent processing interval, Doppler shift, and carrier wavelength.

[0137] Optionally, when the above computer program is executed by the processor, it performs the following steps: based on a preset signal model, it determines multiple sensing signals from multiple received signals; based on the multiple sensing signals, it determines a signal matrix, wherein the signal matrix is ​​a complex matrix; it divides the real part and imaginary part of the signal matrix to determine the signal to be predicted.

[0138] Optionally, when the above computer program is executed by the processor, it performs the following steps: extracting features from the feature extraction module in the region selection network of the signal to be predicted to obtain the feature extraction result; and inputting the feature extraction result into the regression module in the region selection network to determine the target sub-region.

[0139] Optionally, when the above computer program is executed by the processor, it performs the following steps: determining the target location prediction model from multiple location prediction models based on the target sub-region; inputting the signal to be predicted into the spatial attention module in the target location prediction model, using the spatial attention module to perform weighted processing on the signal to be predicted to obtain weighted features; inputting the weighted features into the residual convolutional neural network in the target location prediction model for prediction to obtain location data.

[0140] Optionally, the spatial attention module includes a pooling layer, a convolutional layer, and an activation function. When the above computer program is executed by the processor, it performs the following steps: using the pooling layer, the signal to be predicted is pooled to obtain the pooling result; using the convolutional layer, the pooling result is convolved to obtain the convolution result; using the activation function, the convolution result is weighted according to a preset weight range to obtain the spatial attention weights; and the spatial attention weights are multiplied element-wise with the signal to be predicted to obtain the weighted features.

[0141] Optionally, the target location prediction model includes a first convolutional module, a second convolutional module, and a regression module. When the computer program is executed by the processor, it performs the following steps: using the first convolutional module, the weighted features are subjected to a first processing to obtain a first processing result, wherein the first processing includes convolution processing, normalization processing, activation processing, and pooling processing in sequence; using the second convolutional module, the first processing result is subjected to a second processing to obtain a second processing result, wherein the second processing includes main branch operation and auxiliary branch operation; using the regression module, the second processing result is predicted to obtain location data.

[0142] Optionally, when the above computer program is executed by the processor, it performs the following steps: using the second convolution module to perform main branch operations on the first processing result to obtain the main branch operation result, wherein the main branch operation includes, in sequence: first convolution processing, normalization processing, activation processing, second convolution processing, and second normalization processing; using the second convolution module to perform auxiliary branch operations on the first processing result to obtain the auxiliary branch operation result, wherein the auxiliary branch operation includes: convolution processing based on a preset dimension; and determining the second processing result based on the main branch operation result and the auxiliary branch operation result.

[0143] Embodiments of the present invention also provide a computer-readable storage medium, which includes a stored executable program, wherein, when the executable program is executed, it controls the device where the computer-readable storage medium is located to perform the positioning method described in any of the above embodiments.

[0144] Optionally, in this embodiment, the executable program can be configured to store an executable program for performing the following steps:

[0145] Step S101: Obtain the received signal received by the access point, wherein the sensing area of ​​the access point is divided into multiple sub-areas.

[0146] Step S102: Using a preset signal model, the received signal is preprocessed to obtain the signal to be predicted.

[0147] Step S103: Based on the signal to be predicted and the pre-trained region selection network, determine the target sub-region corresponding to the signal to be predicted from the perception region.

[0148] Step S104: Determine the target location prediction model corresponding to the target sub-region, and input the signal to be predicted into the target location prediction model to predict the location data. Here, multiple sub-regions correspond one-to-one with multiple location prediction models.

[0149] Optionally, the executable program can be configured to store an executable program for performing the following steps: determining a communication signal expression based on user equipment configuration information and preset access point configuration information of the receiving access point, wherein the user equipment is included in a non-cellular MIMO network; determining a sensing signal expression based on preset access point configuration information of the transmitting and receiving access points, wherein the transmitting and receiving access points are included in a non-cellular MIMO network; and determining a preset signal model based on the communication signal expression, the sensing signal expression, and additive noise.

[0150] Optionally, the executable program described above can be configured to store an executable program for performing the following steps: determining a communication signal expression based on the communication channel representation and communication signal representation of the user equipment and the receiving access point, wherein the communication channel representation and communication signal representation are included in the user equipment configuration information and the preset access point configuration information.

[0151] Optionally, the executable program described above can be configured to store an executable program for performing the following steps: determining a sensing signal expression based on the sensing channel representation and sensing signal representation of the transmitting access point and the receiving access point, wherein the sensing channel representation and sensing signal representation are contained in preset access point configuration information.

[0152] Optionally, the sensing channel representation determined when the above executable program is executed includes: signal attenuation of the transmitting access point and the receiving access point, bistatic distance between the transmitting access point and the receiving access point, steering vector of the transmitting access point and the receiving access point, subcarrier spacing, speed of light, coherent processing interval, Doppler frequency shift, and carrier wavelength.

[0153] Optionally, the executable program described above can be configured to store an executable program for performing the following steps: determining multiple sensing signals from multiple received signals based on a preset signal model; determining a signal matrix based on the multiple sensing signals, wherein the signal matrix is ​​a complex matrix; and dividing the real and imaginary parts of the signal matrix to determine the signal to be predicted.

[0154] Optionally, the executable program described above can be configured to store an executable program for performing the following steps: extracting features from the feature extraction module in the region selection network to obtain the feature extraction result; and inputting the feature extraction result into the regression module in the region selection network to determine the target sub-region.

[0155] Optionally, the executable program described above can be configured to store an executable program for performing the following steps: determining a target location prediction model from multiple location prediction models based on the target sub-region; inputting the signal to be predicted into the spatial attention module in the target location prediction model, using the spatial attention module to perform weighted processing on the signal to be predicted to obtain weighted features; inputting the weighted features into the residual convolutional neural network in the target location prediction model for prediction to obtain location data.

[0156] Optionally, the spatial attention module includes a pooling layer, a convolutional layer, and an activation function. The executable program can be configured to store an executable program for performing the following steps: using the pooling layer to perform pooling processing on the signal to be predicted, obtaining the pooling processing result; using the convolutional layer to perform a convolution operation on the pooling processing result, obtaining the convolution processing result; using the activation function to impose weight restrictions on the convolution processing result based on a preset weight range, obtaining spatial attention weights; and multiplying the spatial attention weights element-wise with the signal to be predicted, obtaining weighted features.

[0157] Optionally, the target location prediction model includes a first convolutional module, a second convolutional module, and a regression module. The executable program can be configured to store an executable program for performing the following steps: using the first convolutional module to perform a first processing on the weighted features to obtain a first processing result, wherein the first processing sequentially includes convolution processing, normalization processing, activation processing, and pooling processing; using the second convolutional module to perform a second processing on the first processing result to obtain a second processing result, wherein the second processing includes main branch operations and auxiliary branch operations; and using the regression module to predict the second processing result to obtain location data.

[0158] Optionally, the executable program can be configured to store an executable program for performing the following steps: performing a main branch operation on the first processing result using the second convolution module to obtain a main branch operation result, wherein the main branch operation sequentially includes: first convolution processing, normalization processing, activation processing, second convolution processing, and second normalization processing; performing an auxiliary branch operation on the first processing result using the second convolution module to obtain an auxiliary branch operation result, wherein the auxiliary branch operation includes: convolution processing based on a preset dimension; and determining a second processing result based on the main branch operation result and the auxiliary branch operation result.

[0159] Optionally, specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated here.

[0160] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0161] In the embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be 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 displayed or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between units or modules may be electrical or other forms.

[0162] The units described as separate components may or may not be physically separate. 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 units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0163] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0164] If the integrated unit 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 the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0165] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A positioning method, characterized in that, include: The received signal received by the access point is acquired, wherein the sensing area of ​​the access point is divided into multiple sub-regions; The received signal is preprocessed using a preset signal model to obtain the signal to be predicted; The signal to be predicted is input into the feature extraction module of the pre-trained region selection network for feature extraction to obtain the feature extraction result. The feature extraction result is input into the regression module in the region selection network to determine the target sub-region corresponding to the signal to be predicted from the perception region; Based on the target sub-region, determine the target location prediction model corresponding to the target sub-region from multiple location prediction models; The signal to be predicted is input into the spatial attention module in the target location prediction model, and the spatial attention module is used to perform weighted processing on the signal to be predicted to obtain weighted features. The weighted features are input into the residual convolutional neural network in the target location prediction model for prediction to obtain location data, wherein the multiple sub-regions correspond one-to-one with the multiple location prediction models.

2. The positioning method according to claim 1, characterized in that, The construction of the preset signal model includes: Based on the user equipment configuration information and the preset access point configuration information of the receiving access point, a communication signal expression is determined, wherein the user equipment is included in a non-cellular multiple-input multiple-output network; Based on the preset access point configuration information of the transmitting access point and the receiving access point, a sensing signal expression is determined, wherein the transmitting access point and the receiving access point are included in a non-cellular multiple-input multiple-output network; The preset signal model is determined based on the communication signal expression, the sensing signal expression, and the additive noise.

3. The positioning method according to claim 2, characterized in that, The step of determining the communication signal expression based on user equipment configuration information and preset access point configuration information of the receiving access point includes: Based on the communication channel representation and communication signal representation of the user equipment and the receiving access point, the communication signal expression is determined, wherein the communication channel representation and the communication signal representation are included in the user equipment configuration information and the preset access point configuration information.

4. The positioning method according to claim 2, characterized in that, The process of determining the sensing signal expression based on the preset configuration information of the transmitting access point and the receiving access point includes: Based on the sensing channel representation and sensing signal representation of the transmitting access point and the receiving access point, the sensing signal expression is determined, wherein the sensing channel representation and the sensing signal representation are included in the preset access point configuration information.

5. The positioning method according to claim 4, characterized in that, The sensing channel representation includes: signal attenuation of the transmitting access point and the receiving access point, bistatic distance between the transmitting access point and the receiving access point, steering vector of the transmitting access point and the receiving access point, subcarrier spacing, speed of light, coherent processing interval, Doppler shift, and carrier wavelength.

6. The positioning method according to claim 1, characterized in that, The step of preprocessing the received signal using a preset signal model to obtain the signal to be predicted includes: Based on the preset signal model, multiple sensing signals are determined from the multiple received signals; Based on the plurality of sensing signals, a signal matrix is ​​determined, wherein the signal matrix is ​​a complex matrix; The real and imaginary parts of the signal matrix are divided to determine the signal to be predicted.

7. The positioning method according to claim 1, characterized in that, The spatial attention module includes a pooling layer, a convolutional layer, and an activation function. The signal to be predicted is input into the spatial attention module of the target location prediction model. The spatial attention module performs weighted processing on the signal to be predicted to obtain weighted features, including: The pooling layer is used to perform pooling processing on the signal to be predicted to obtain the pooling processing result; The convolutional layer is used to perform a convolution operation on the pooling result to obtain the convolutional processing result; Using the activation function, the spatial attention weights are obtained by weighting the convolution result based on a preset weight range. The spatial attention weights are multiplied element-wise with the signal to be predicted to obtain the weighted features.

8. The positioning method according to claim 7, characterized in that, The target location prediction model includes a first convolutional module, a second convolutional module, and a regression module. The step of inputting the weighted features into the residual convolutional neural network of the target location prediction model for prediction to obtain the location data includes: The weighted features are processed using the first convolution module to obtain a first processing result, wherein the first processing includes convolution processing, normalization processing, activation processing and pooling processing in sequence. The second convolution module is used to perform a second processing on the first processing result to obtain a second processing result, wherein the second processing includes a main branch operation and an auxiliary branch operation; The regression module is used to predict the second processing result to obtain the location data.

9. The positioning method according to claim 8, characterized in that, The second convolution module is used to perform a second processing on the first processing result to obtain a second processing result, including: The second convolution module is used to perform main branch operations on the first processing result to obtain the main branch operation result, wherein the main branch operation includes, in sequence: first convolution processing, normalization processing, activation processing, second convolution processing, and second normalization processing; The second convolution module is used to perform auxiliary branch operations on the first processing result to obtain the auxiliary branch operation result, wherein the auxiliary branch operation includes: convolution processing based on a preset dimension; The second processing result is determined based on the main branch operation result and the auxiliary branch operation result.

10. An electronic device, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the method according to any one of claims 1 to 9.

11. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the storage medium is located to perform the method according to any one of claims 1 to 9.

Images