WO2023116196A1 - Aoa and tof joint estimation method and apparatus for indoor positioning, and storage medium - Google Patents

Abstract

Disclosed in the present invention are an AoA and ToF joint estimation method and apparatus for indoor positioning, and a storage medium. The method comprises: pre-processing acquired CSI data to obtain processed data; and inputting the processed data into a pre-trained joint estimation model to obtain an AoA and a ToF which correspond to the processed data, wherein the joint estimation model is obtained by means of training by a deep convolutional neural network, and the deep convolutional neural network is formed by convolutional kernels of different sizes which are connected in parallel. In the present invention, training is performed on the basis of a deep convolutional neural network which is constructed by convolutional kernels of different sizes which are connected in parallel, such that a joint estimation model is obtained, thereby improving the precision, resolution and anti-noise capability of an estimation result.

Description

AOA and TOF joint estimation method, device and storage medium for indoor positioning technical field

The invention relates to an AOA and TOF joint estimation method, device and storage medium for indoor positioning, and belongs to the technical field of wireless signal processing.

Background technique

Wireless signals are widely used in various fields of daily life. In the fields of health perception, fire positioning and rescue, and augmented reality-based navigation, it is very important to accurately estimate the direction of arrival (AOA) and time of arrival (TOF) of each path of the wireless signal. However, during the multipath propagation of wireless signals, objects near indoor APs and mobile clients will reflect wireless signals, resulting in low estimation accuracy of direction of arrival (AOA) and time of arrival (TOF).

At present, most of the research is to obtain a corresponding relationship from the received signal to the AOA by constructing a mathematical model and combining the relevant algorithms of signal processing. However, a large number of approximations will be made in the process of model establishment and solution, and the approximation operation will lose part of the information. , there are certain obstacles to the accurate estimation of AOA.

Contents of the invention

The purpose of the present invention is to overcome the deficiencies in the prior art, and provide an AOA and TOF joint estimation method, device and storage medium for indoor positioning.

In a first aspect, the present invention provides an AOA and TOF joint estimation method for indoor positioning, the method comprising:

Preprocessing the acquired CSI data to obtain the processed data;

Input the processed data into the pre-trained joint estimation model to obtain the AOA and TOF corresponding to the processed data;

The joint estimation model is obtained by training a deep convolutional neural network, and the deep convolutional neural network uses convolution kernels of different sizes in parallel.

Further, the acquired CSI data is represented as a complex data matrix image, and the processed data is represented as a real number matrix image.

Further, the processed data is represented as a three-channel real number matrix image.

Further, the acquisition of the three-channel real matrix image includes:

Divide the complex data matrix of the obtained CSI data and its conjugate matrix into four sub-matrices according to the real part and the imaginary part and reorganize to obtain the recombination matrix;

Through the sliding windows with sliding steps of 2, 4 and 6, the recombination matrix is shifted and intercepted respectively to obtain several sub-matrices respectively, and then the several sub-matrices obtained are spliced to obtain three matrices. Matrix composition is the acquired three-channel real matrix image.

Further, the deep convolutional neural network includes an input layer, a 7*7 convolutional layer, a maximum pooling layer, a 1*1 convolutional layer, a 3*3 convolutional layer, a pooling layer, 2 Inception structures, Maximum pooling layer, 3 Inception structures, maximum pooling layer, 2 Inception structures, average pooling layer, fully connected layer, output layer.

Further, the Inception structure includes four layers: the first layer is an input layer, the second layer is a 1*1 convolutional layer, a 1*1 convolutional layer, a 3*3 maximum pooling layer, and a 1*1 convolutional layer , the third layer is 3*3 convolutional layer, 5*5 convolutional layer and 1*1 convolutional layer, and the fourth layer is the output layer.

Further, the training of the joint estimation model includes:

Obtain historical CSI data and the corresponding AOA and TOF of CSI data, and construct a data set;

Preprocessing the data set to construct a training set;

The training set is trained through a deep convolutional neural network to obtain a joint estimation model.

In a second aspect, the present invention provides a device, including a processor and a storage medium;

The storage medium is used to store instructions;

The processor is configured to operate according to the instructions to perform the steps of the method of the first aspect.

In a third aspect, the present invention provides a storage medium on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method described in the first aspect are implemented.

Compared with the prior art, the beneficial effects of the present invention are: the present invention obtains a joint estimation model through deep convolutional neural network training based on parallel construction of convolution kernels of different sizes, and achieves a higher accuracy than that estimated by traditional methods. Accuracy, higher resolution and better noise immunity.

Description of drawings

Fig. 1 is the working flowchart of the AOA and TOF joint estimation method for indoor positioning according to the embodiment of the present invention;

FIG. 2 is a schematic diagram of an antenna array at a receiving end according to an embodiment of the present invention;

3 is a schematic diagram of a single-channel construction method of CSI data in a data set according to an embodiment of the present invention;

4 is a schematic diagram of a multi-channel construction method for CSI data in a data set according to an embodiment of the present invention;

5 is a schematic diagram of a multi-scale convolution kernel and CSI data convolution method according to an embodiment of the present invention;

Fig. 6 is a neural network structure diagram of parallel design of convolution kernels of different sizes according to an embodiment of the present invention;

Fig. 7 is the root mean square error (RMSE) contrast chart of joint estimated value and multiple method estimated results of the embodiment of the present invention;

Fig. 8 is a comparison diagram of the error distribution between the joint estimated value and the estimated results of various methods according to the embodiment of the present invention.

Detailed ways

Embodiments of the present invention are described in detail below, examples of which are shown in the drawings, wherein the same or similar reference numerals designate the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the figures are exemplary only for explaining the present invention and should not be construed as limiting the present invention.

In the description of the present invention, several means more than one, and multiple means more than two. Greater than, less than, exceeding, etc. are understood as not including the original number, and above, below, within, etc. are understood as including the original number. If the description of the first and second is only for the purpose of distinguishing the technical features, it cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features relation.

In the description of the present invention, unless otherwise clearly defined, words such as setting, installation, and connection should be understood in a broad sense, and those skilled in the art can reasonably determine the specific meanings of the above words in the present invention in combination with the specific content of the technical solution.

In the description of the present invention, reference to the terms "one embodiment," "some embodiments," "exemplary embodiments," "examples," "specific examples," or "some examples" is intended to mean that the embodiments are A specific feature, structure, material, or characteristic described by or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

The present invention provides an AOA and TOF joint estimation method, device and storage medium for indoor positioning. The present invention will be further described below in conjunction with the accompanying drawings and embodiments, wherein:

Example 1:

As shown in Figure 1, an embodiment of the present invention provides an AOA and TOF joint estimation method for indoor positioning, including:

The OFDM signal is used as the transmission signal, the transmission channel is set to Gaussian channel, the channel bandwidth is 40MHz, the center frequency is 5.32GHz, the interval between adjacent subcarriers is set to 312.5KHz, and 30 subcarriers are selected, as shown in Figure 2, the receiving end adopts A linear array of three antennas generates a data set by setting different channel parameters. The data set samples include input data and labels, and the received CSI is used as input data, and AOA and TOF are used as labels.

The CSI data in the data set is a matrix of complex numbers. However, the current deep convolutional neural network does not have a proper processing method for complex input. In order to obtain a CSI image that can be used as an input, the present invention uses the complex CSI data received from the antenna array The data matrix H 90×1 is divided into HR, H I, where HR is the real part matrix of H, and H I is the imaginary part matrix of H, and then convert it to the formula (1) to obtain the amplitude that can effectively retain the CSI and phase information;

Among them, csi m, k, m=1, 2, 3, k=1, 2, ..., 30 represent the kth subcarrier on the mth antenna, which is complex data, and Re(csi) represents the real value of csi part, Im(csi) means to take the imaginary part of csi.

The convolutional neural network extracts the features of the input image through the convolution kernel. Considering that the estimated values of TOF and AOA are related to multiple subcarriers, as shown in Figure 3, the present invention proposes a method for constructing a single-channel input matrix, so that hi, j represents the value of the i (i=1,2,...,180) row and j (j=1,2) column of the matrix, setting the moving step size to 2 so that the convolution kernel can cover multiple subcarriers when training the network, When the window moves, the data from the front end is used to fill in the end, so that the matrix obtained by moving the window i times, and finally these 90 matrices with a size of 180×2 are merged into a new matrix:

In order to extract more features of each subcarrier signal combination, as shown in Figure 4, the present invention proposes a method for constructing a multi-channel input matrix, changing the moving step size to 4 and 6 respectively, and then performing the same filling operation on the window to obtain two The obtained three matrices are composed of three-dimensional matrix H 180×180×3 used as the input of the neural network according to the graph.

According to the obtained three-channel real matrix image and the AOA and TOF corresponding to the three-channel real matrix image, a training set is constructed.

The deep convolutional neural network is designed based on the parallel connection of convolution kernels of different sizes. The convolution kernels of different scales can be convoluted with the received signals of subcarriers of different frequencies according to the method shown in Figure 5, and then the data related to each subcarrier can be extracted. Features, as shown in Figure 6, the deep convolutional neural network includes an input layer, a 7*7 convolutional layer, a maximum pooling layer, a 1*1 convolutional layer, a 3*3 convolutional layer, a pooling layer, and 2 Inception structure, maximum pooling layer, 3 Inception structures, maximum pooling layer, 2 Inception structures, average pooling layer, fully connected layer, output layer; the Inception structure includes four layers: the first layer is the input layer, The second layer is 1*1 convolutional layer, 1*1 convolutional layer, 3*3 maximum pooling layer and 1*1 convolutional layer, the third layer is 3*3 convolutional layer, 5*5 convolutional layer And 1*1 convolutional layer, the fourth layer is the output layer.

The training set is trained through a deep convolutional neural network to obtain a joint estimation model.

The estimated CSI data is preprocessed to obtain a three-channel real matrix image, and the three-channel indicator matrix is used as the input of the joint estimation model, and the AOA and TOF corresponding to the estimated CSI data are output.

As shown in Figure 7, it is a comparison diagram of the root mean square error (RMSE) of the estimated results of the method of the present invention and other algorithms. It can be seen that compared with the proposed method of the present invention, the other two methods (SpoFi and Join- 2D) The estimated RMSE of AOA and TOF decreases with the increase of SNR; since the training samples include data under multiple signal-to-noise ratios, the trained network has good generalization ability, so in different signal-to-noise ratios There are more stable estimation results than the following; from the results, it can be seen that the method proposed in this paper performs better than the other two methods at any signal-to-noise ratio.

As shown in Figure 8, taking the AOA estimation error distribution in the left figure as an example, from the perspective of error distribution, the estimation errors of the AOA of the present invention are all distributed within 10°, while the estimation errors of the other two methods are within 10° The amount of data only accounts for 80% and 70% of the total data, and nearly 20% of the data has an estimation error of more than 30°, that is, it cannot be accurately identified.

Another advantage of the method proposed in the present invention is that even if the angle of arrival and time of arrival of two paths are similar, they can be identified accurately, while the identification of the other two paths in this case is not ideal. This is because the estimation algorithms in SpoFi and Join-2D are both improved algorithms based on the MUSIC algorithm, in which the MUSIC algorithm utilizes the orthogonality of the signal subspace and the noise subspace to construct a spatial spectral function, and searches through spectral peaks To estimate AOA, if the spectral peaks are similar, it is difficult to distinguish them, so the traditional algorithm has certain resolution limitations. The neural network-based method proposed in the present invention estimates AOA and TOF by extracting relevant features of the signal. The estimated values between the paths do not affect each other, so there is no problem of insufficient resolution. In the case of Table 1, SpoFi and Join-2D can only identify one path due to the low resolution, and the method proposed in the present invention can be used in Both paths are accurately identified within a certain error range.

Table 1 Estimation results under similar AOA paths (null means recognition failure)

True value SpoFi Join-2D method path 1 of the present invention: (123°, 96ns) (125°, 107ns) (124°, 99ns) (123.6°, 85ns) path 2: (132°, 102ns) null null (133.7° ,94ns)

Example 2:

This embodiment provides a device, including a processor and a storage medium;

The storage medium is used to store instructions;

The processor is configured to operate according to the instructions to execute the steps of the method described in Embodiment 1.

Example 3:

This embodiment provides a storage medium on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method described in Embodiment 1 are implemented.

Those skilled in the art should understand that the embodiments of the present application may be provided as methods, systems, or computer program products. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present application is described with reference to flowcharts and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions The device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow chart or blocks of the flowchart and/or the block or blocks of the block diagrams.

The above is only a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the technical principle of the present invention, some improvements and modifications can also be made. It should also be regarded as the protection scope of the present invention.

Claims (9)

1. AOA and TOF joint estimation method for indoor positioning, characterized in that the method comprises:

   Preprocessing the acquired CSI data to obtain the processed data;

   Input the processed data into the pre-trained joint estimation model to obtain the AOA and TOF corresponding to the processed data;

   The joint estimation model is obtained by training a deep convolutional neural network, and the deep convolutional neural network uses convolution kernels of different sizes to be connected in parallel.
2. The AOA and TOF joint estimation method for indoor positioning according to claim 1, wherein the acquired CSI data is represented as a complex data matrix image, and the processed data is represented as a real number matrix image.
3. The AOA and TOF joint estimation method for indoor positioning according to claim 2, wherein the processed data is represented as a three-channel real number matrix image.
4. The AOA and TOF joint estimation method for indoor positioning according to claim 3, wherein the acquisition of the three-channel real number matrix image comprises:

   Divide the complex data matrix of the obtained CSI data and its conjugate matrix into four sub-matrices according to the real part and the imaginary part and reorganize to obtain the recombination matrix;

   Through the sliding windows with sliding steps of 2, 4 and 6, the recombination matrix is shifted and intercepted respectively to obtain several sub-matrices respectively, and then the several sub-matrices obtained are spliced to obtain three matrices. Matrix composition is the acquired three-channel real matrix image.
5. The AOA and TOF joint estimation method for indoor positioning according to claim 1, wherein the deep convolutional neural network comprises an input layer, a 7*7 convolutional layer, a maximum pooling layer, and a 1*1 Convolutional layer, 3*3 convolutional layer, pooling layer, 2 Inception structures, maximum pooling layer, 3 Inception structures, maximum pooling layer, 2 Inception structures, average pooling layer, fully connected layer, output layer.
6. The AOA and TOF joint estimation method for indoor positioning according to claim 5, wherein the Inception structure includes four layers: the first layer is an input layer, the second layer is a 1*1 convolutional layer, 1 *1 convolutional layer, 3*3 maximum pooling layer and 1*1 convolutional layer, the third layer is 3*3 convolutional layer, 5*5 convolutional layer and 1*1 convolutional layer, the fourth layer is output layer.
7. The AOA and TOF joint estimation method for indoor positioning according to claim 1, wherein the training of the joint estimation model comprises:

   Obtain historical CSI data and the corresponding AOA and TOF of CSI data, and construct a data set;

   Preprocessing the data set to construct a training set;

   The training set is trained through a deep convolutional neural network to obtain a joint estimation model.
8. A device, characterized in that it includes a processor and a storage medium;

   The storage medium is used to store instructions;

   The processor is configured to operate according to the instructions to perform the steps of the method according to any one of claims 1-7.
9. A storage medium, on which a computer program is stored, wherein when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are realized.

Images