Subspace time difference direct positioning method based on propagation operator
By using a subspace time difference direct positioning method based on propagation operators, integrating base station signals for Fourier transform and constructing a noise subspace, the high computational complexity and insufficient accuracy of existing technologies are solved, achieving high-precision and low-complexity positioning results.
Patent Information
- Application Number
- CN202410984574.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-22
- Publication Date
- 2024-11-22
AI Technical Summary
Existing passive positioning technologies suffer from high computational complexity, high hardware costs, and insufficient positioning accuracy. In particular, the DPD algorithm based on TDOA has a large matrix operation dimension and high computational complexity, making it difficult to achieve high-precision positioning.
A subspace time difference direct positioning method based on propagation operators is adopted. By integrating the time domain signals of the base station and performing Fourier transform, the covariance matrix and propagation operators are calculated to construct the noise subspace. The loss function is constructed using the Lagrange multiplier method, which reduces computational complexity and improves positioning accuracy.
It achieves high-precision positioning results while reducing computational complexity, alleviating hardware and computational burden, meeting real-time requirements, and exhibiting good positioning performance.
Smart Images

Figure CN119001597A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of wireless positioning technology, and in particular to a subspace time difference direct positioning method based on a propagation operator. Background Art
[0002] Positioning technology has played a huge role in civil and military fields such as radar, wireless communication, vehicle-mounted technology, and indoor navigation. Passive positioning, as an important part of this technology, has attracted more and more attention due to its advantages such as high accuracy and strong concealment. Early passive positioning was generally implemented using a two-step positioning method. The first step is to estimate the intermediate parameters using the signals received by the observation station. The second step is to establish the corresponding equation based on the above parameters and the distribution of the observation station. Then a specific method is used to solve the equation to obtain the position of the radiation source. Although the two-step method has been widely used, it still has some shortcomings. Direct Position Determination (DPD) estimates the target position directly from the measurement data without the need for parameter estimation. Compared with the two-step positioning method, the direct positioning technology estimates the radiation source position in a single step based on the received data, which significantly improves the positioning accuracy and robustness. The more classic direct positioning algorithms include the direct positioning algorithm based on the maximum likelihood (ML) criterion, the direct positioning algorithm based on subspace data fusion (SDF), and the direct positioning algorithm based on minimum variance distortionless response (MVDR).
[0003] The above-mentioned DPD algorithms are all based on the use scenarios of array antennas. With the continuous development of materials and process technologies, most materials tend to be thinner and lighter, and the hardware cost of array antennas is high, which may not be realized in practical applications. On the contrary, the positioning method based on Time Difference Of Arrival (TDOA) is implemented through distributed single sensor observation stations, but there are few DPD algorithms based on TDOA. The more classic one is the direct positioning algorithm using the maximum likelihood estimator. After that, the determinant-based algorithm constructed using subspace orthogonality came into being, but the algorithm has a large matrix operation dimension and high computational complexity. It has great requirements on computing hardware and is difficult to implement. On top of this, the Multiple-frequency Function Fusion (MFF) algorithm based on eigenvalue decomposition at multiple useful frequencies of the fused broadband signal reduces the algorithm complexity while retaining a high positioning accuracy. However, in actual use, matrix decomposition also brings about a large computational complexity. Summary of the invention
[0004] The purpose of the present invention is to propose a subspace time difference direct positioning method based on a propagation operator, which reduces the complexity of constructing the subspace and realizes high-precision direct positioning.
[0005] In order to achieve the above technical objectives, the technical solution adopted by the present invention is:
[0006] A subspace time difference direct positioning method based on a propagation operator, the subspace time difference direct positioning method comprising the following steps:
[0007] Step 1: Integrate all base station time domain received signals and perform discrete Fourier transform in segments to obtain the frequency domain matrix form of the received signal;
[0008] Step 2: Calculate the covariance matrix of different frequencies and use the covariance matrix to calculate the propagation operator and construct the noise subspace;
[0009] Step 3: Use the noise subspace to construct the loss function;
[0010] Step 4: Divide the area to be searched into grids and calculate the loss function value of each grid point;
[0011] Step 5: Search for the extreme value of the loss function to obtain the radiation source position.
[0012] Step 1 further comprises:
[0013] There are L observation base stations in the two-dimensional plane, and the lth observation base station is located at S l (X l ,Y l ), each observation station is equipped with only one sensor to capture the signal and each observation station is time synchronized; suppose there are M unknown far-field radiation sources, and they emit uncorrelated broadband signals; let the mth unknown transmitter be located at T m (x m ,y m ), at any time t within the observation time interval, the signal obtained by the lth base station is expressed as:
[0014]
[0015] Where T s is the observation time interval, α l,m is the unknown attenuation coefficient on the propagation path of the signal from the mth radiation source to the lth observation station, s m (t) is the signal emitted by the mth radiation source at any time t within the observation time interval, w l (t) is white zero-mean complex additive Gaussian noise, τ l,m is the time delay of the mth radiation source reaching the lth observation station, expressed as:
[0016]
[0017] Where c is the propagation speed of electromagnetic waves, integrating the received signals of all base stations to obtain:
[0018] Z(t)=[z 1 (t)z 2 (t)…z L (t)] T ,0≤t≤T s ;
[0019] Divide the received signals of each base station into J segments evenly, that is:
[0020] Z=[Y 1 (t′)Y 2 (t′)…Y J (t′)];
[0021] Then, for each segment of the received signal Y j (t′) performs N-point DFT and obtains:
[0022]
[0023] in, and Y j (t′), z l (t j ),s m (t j ) and w l (t j ) DFT results, f n The frequency corresponding to the nth frequency point; let q be the radiation source position vector, and
[0024]
[0025] A(f n ,q)=[a(f n ,q 1 ) a(f n ,q 2 ) … a(f n ,q M )]
[0026]
[0027] Then we have:
[0028]
[0029] Among them, S j (f n) represents the nth narrowband frequency domain component of the transmitted signal when the frequency domain snapshot is j, W j (f n ) represents the nth narrowband frequency domain component of the transmitted signal when the frequency domain snapshot is j, A(f n ,q) represents the signal manifold matrix, a(f n ,q m ) is a vector containing attenuation coefficient, delay and frequency information in the n-th narrowband frequency domain;
[0030] Integrate all frequency domain snapshot data to get the frequency domain matrix form of the received signal:
[0031]
[0032] Step 2 further includes:
[0033] Frequency f n The covariance matrix of the corresponding components is:
[0034]
[0035] Among them, R S (f n ) is the signal spectral density matrix, R W (f n ) is the noise spectral density matrix; the signal manifold matrix is decomposed into:
[0036]
[0037] Among them, A 1 (f n ,q) is A(f n ,q) first M rows, A 2 (f n ,q) is A(f n ,q) of the posterior LM; then there exists a unique propagation operator So that the following is true:
[0038]
[0039] Among them, I L-M Represents the identity matrix of L=M dimensions; definition:
[0040]
[0041] Then we have:
[0042]
[0043] Q nN and A(f n ,q) are orthogonal, for Q nNThe column vector of the noise subspace U is orthogonalized nN :
[0044]
[0045] The covariance matrix is divided into blocks:
[0046]
[0047] Among them, G n Depend on The first M columns of H n Depend on The propagation operator of the nth frequency point is obtained as follows:
[0048]
[0049] The corresponding noise subspace is constructed based on the propagation operator.
[0050] Step 3 further includes:
[0051] Let the search position be p = [x, y] T , when the search position is the radiation source position, according to the orthogonality of the noise subspace and the signal manifold, we get:
[0052]
[0053] Where diag(·) represents diagonalization, a(f n ,p) represents the vector of the time delay and frequency information from the target location to each base station, represents the attenuation vector from the target location to each base station, where α 1 ,α 2 ,··,α L Indicates the attenuation coefficient to each base station.
[0054] Cut G effective frequency bands and integrate G effective frequency bands to get:
[0055]
[0056] Define K(p) as:
[0057]
[0058] So:
[0059]
[0060] Assumptions The first element of is 1, and the conversion is:
[0061]
[0062] Among them, e 1 =[1,0,…,0] T , according to the Lagrange multiplier method, the loss function F(p) is:
[0063] F(p)=e 1 H K(p) -1 e 1 .
[0064] Compared with the prior art, the present invention has the following beneficial effects:
[0065] First, the subspace time difference direct positioning method based on the propagation operator of the present invention has an estimation accuracy that is better than the traditional two-step positioning method, a lower complexity than the MFF method and a similar positioning performance.
[0066] Second, the subspace time difference direct positioning method based on the propagation operator of the present invention can reduce the complexity of calculating the subspace by constructing the noise subspace through data segmentation and multi-band segmentation and using the propagation operator instead of the traditional matrix decomposition, thereby alleviating the hardware and computing burden of the base station to meet the real-time requirements of the time difference positioning system.
[0067] Third, the subspace time difference direct positioning method based on the propagation operator of the present invention uses multi-frequency data, extracts the effective part from the received signal to construct the loss function and suppresses the unknown attenuation through the Lagrange multiplier method, which effectively improves the positioning accuracy of the algorithm. In computer simulation, the method proposed by the present invention shows better positioning accuracy and lower complexity than other time difference positioning methods. BRIEF DESCRIPTION OF THE DRAWINGS
[0068] Figure 1 It is a flow chart of the subspace time difference direct positioning method based on the propagation operator of the present invention;
[0069] Figure 2 This is a schematic diagram of distributed single sensor base station positioning;
[0070] Figure 3 It is a two-dimensional spatial spectrum of the method of the present invention;
[0071] Figure 4 A scatter plot of the results of locating multiple radiation sources by the method of the present invention;
[0072] Figure 5 Schematic diagram of the root mean square error performance of the present invention and other positioning algorithms under different signal-to-noise ratios;
[0073] Figure 6 Schematic diagram of the complexity of the subspace construction of the present invention and the MFF algorithm. DETAILED DESCRIPTION
[0074] The embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings.
[0075] Symbolic representation: In this paper, bold uppercase letters, bold lowercase letters, and italic letters, such as A, a, and a, represent matrices, vectors, and scalars, respectively. p and [A] p,q denote the pth element of vector a and the element of pth row and qth column of matrix A respectively, (·) T , (·) H and(·) -1 denote the transpose, conjugate transpose, and inverse operations of a matrix, respectively; ||·|| and |·| denote the Euclidean norm of a vector and the absolute value of a scalar, respectively; ε{·} denotes the mathematical expectation; letters with triangular superscripts represents the estimated value of R.
[0076] The present invention provides a subspace time difference direct positioning method based on a propagation operator, such as Figure 1 As shown, the specific steps include:
[0077] Step 1: Integrate all base station time domain received signals and perform discrete Fourier transform on a segmented basis to obtain the frequency domain matrix form of the received signal.
[0078] Consider positioning scenarios such as Figure 2 There are L observation base stations in the two-dimensional plane, and the lth observation base station is located at S l (X l ,Y l ), each observation station is equipped with only one sensor to capture the signal and each observation station is time synchronized. Assume that there are M unknown far-field radiation sources, and they emit uncorrelated broadband signals. Assume that the mth unknown emitter is located at T m (x m ,y m ). At any time t within the observation time interval, the signal obtained by the lth base station can be expressed as:
[0079]
[0080] Where 0≤t≤T s is the observation time interval, α l,m is the unknown attenuation coefficient on the propagation path of the signal from the mth radiation source to the lth observation station, s m (t) is the signal emitted by the mth radiation source at any time t within the observation time interval, w l (t) is white zero-mean complex additive Gaussian noise, τ l,m is the time delay of the mth radiation source reaching the lth observation station, expressed as;
[0081]
[0082] Where c is the propagation speed of electromagnetic waves. By integrating the received signals of all base stations, we can get:
[0083] Z(t)=[z 1 (t) z 2 (t) … z L (t)] T ,0≤t≤T s ;
[0084] Divide the received signals of each base station into J segments evenly, that is:
[0085] Z=[Y 1 (t′) Y 2 (t′) … Y J (t′)];
[0086] Then, DFT is performed on each segment of the received signal, that is, Y j (t′) performs N-point DFT, and we can get:
[0087]
[0088]
[0089] in, and Y j (t′), z l (t j ),s m (t j ) and w l (t j ) of the DFT results, where f n The frequency corresponding to the nth frequency point. Let q be the radiation source position vector, and
[0090]
[0091] A(f n ,q)=[a(f n ,q 1 ) a(f n ,q 2 ) … a(f n ,q M )]
[0092]
[0093] Then we have:
[0094]
[0095] Among them, S j (f n ) represents the nth narrowband frequency domain component of the transmitted signal when the frequency domain snapshot is j, W j (f n ) represents the nth narrowband frequency domain component of the transmitted signal when the frequency domain snapshot is j, A(f n ,q) represents the signal manifold matrix, which is a vector containing attenuation coefficient, delay and frequency information.
[0096] Integrating all frequency domain snapshot data gives:
[0097]
[0098] Step 2: Calculate the covariance matrix of different frequencies and use the covariance matrix to calculate the propagation operator and construct the noise subspace:
[0099] Frequency f n The covariance matrix of the corresponding components is:
[0100]
[0101] Among them, R S (f n ) is the signal spectral density matrix, R W (f n ) is the noise spectral density matrix. The propagation operator is defined below. The definition of the propagation operator depends on the block division of the signal manifold. Since the signal manifold is full rank, it can be decomposed into:
[0102]
[0103] Among them, A 1 (f n ,q) is A(f n ,q) first M rows, A 2 (f n ,q) is A(f n ,q) of the post-LM. Then there exists a unique propagation operator So that the following is true:
[0104]
[0105] Among them, I L-M Represents the identity matrix of LM dimension. From the above formula, we can define:
[0106]
[0107] Then we have:
[0108]
[0109] That is Q nN and A(f n ,q) are orthogonal, so Q nN It is the same space as the noise subspace. Using this property, we can nN The noise subspace U can be obtained by orthogonalizing the column vector of nN :
[0110]
[0111] Next, we estimate the propagation operator and divide the covariance matrix into blocks to obtain:
[0112]
[0113] Among them, G n Depend on The first M columns of H n Depend on The propagation operator of the nth frequency point can be obtained as follows:
[0114]
[0115] After obtaining the propagation operator, the corresponding noise subspace can be constructed.
[0116] Step 3: Construct the loss function using the noise subspace:
[0117] Let the search position be p = [x, y] T , when the position is the radiation source position, according to the orthogonality of the noise subspace and the signal manifold, we can get:
[0118]
[0119]
[0120] Where diag(·) represents diagonalization, a(f n ,p) contains the vector of the time delay and frequency information from the target location to each base station, The attenuation coefficient from the target location to each base station. is unknown, so the above formula cannot be used directly as the loss function.
[0121] The narrowband data in different frequency bands of the same broadband transmission signal correspond to the same attenuation coefficient, that is, the same However, in reality, not all frequency bands of the signal are valid and uniform, so G valid frequency bands are usually intercepted for use during processing. By integrating these valid frequency bands, we can get:
[0122]
[0123] Define K(p) as:
[0124]
[0125] So:
[0126]
[0127] At this point, assuming If the first element of is 1, then we can convert it to:
[0128]
[0129] Among them, e 1 =[1,0,…,0] T , according to the Lagrange multiplier method, we can get the loss function as:
[0130] F(p)=e 1 H K(p) -1 e 1 .
[0131] Step 4: Divide the area to be searched into grids and calculate the loss function value of each grid point;
[0132] like Figure 3 The figure shows the two-dimensional spatial spectrum obtained by calculating the loss function in the possible area by the method of the present invention, where the radiation source is located at T 1 (310m,380m),T 2 (810m, 520m), we can see that the two-dimensional spatial spectrum has two maximum peaks at the positions corresponding to the two radiation sources, which means that the radiation source can be located by finding the positions of the M largest peaks in the two-dimensional spatial spectrum.
[0133] Step 5: Search for the extreme value of the loss function.
[0134] The effect of the algorithm of the present invention is further described below in conjunction with a simulation example.
[0135] We simulated a single-channel multi-base station positioning scenario and considered two far-field broadband radiation sources with a sampling frequency of 125 MHz. Their actual positions are located at T 1 (410m,480m),T 2 (810m,120m). Use 4 1(200m,0m),S 2 (900m,0m),S 3 (200m,600m),S 4 The observation station (900m, 600m) is used to locate it. N = 1024, G = 25, and SNR = 10dB. The search area is 1000m × 1000m, the grid division accuracy is 1m, and 50 simulation experiments are carried out independently. Figure 4 It is a scatter plot of the positioning results of the method of the invention.
[0136] The performance estimation standard of the present invention in the simulation is the relative root mean square error (RRMSE) defined as:
[0137]
[0138] Among them, Mc is the number of Monte Carlo experiments, represents the position estimation result of the pth radiation source in the mcth experiment, q p represents the true value of the position of the pth radiation source, where P is the number of radiation sources. In the following simulation, the number of Monte Carlo experiments is 500. The complexity performance is measured by the number of complex multiplications performed in the algorithm.
[0139] Figure 5 The performance curve of the root mean square error of the positioning results of the method of the present invention, the two-step algorithm, and the MFF algorithm as the signal-to-noise ratio changes. Figure 4 Same. From Figure 5 It can be seen that compared with the two-step algorithm, the present invention achieves a very high positioning accuracy and is similar to the positioning accuracy of the MFF algorithm. However, the present algorithm does not require eigenvalue decomposition, and the complexity of the corresponding steps is reduced.
[0140] Figure 6 This is a bar chart comparing the complexity of the constructed noise subspace of the method of the present invention and the MFF algorithm. Figure 4 Same. From Figure 6 It can be seen that the present invention achieves lower complexity in constructing the noise subspace.
[0141] In summary, from the analysis of the simulation effect diagram and the complexity bar chart, it can be seen that the subspace time difference direct positioning method based on the propagation operator proposed in the present invention can greatly reduce the computational complexity of constructing the noise subspace, and can meet the requirements of single-channel multi-base station direct positioning for resolution and real-time performance. At the same time, the algorithm can significantly improve the estimation accuracy and have better performance under harsh environmental conditions and when the signal-to-noise ratio is low.
[0142] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The scheme in the embodiments of the present application can be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.
[0143] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0144] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0145] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0146] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.
[0147] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.
Claims
1. A subspace time difference direct positioning method based on a propagation operator, characterized in that: The subspace time difference direct positioning method comprises the following steps: Step 1: Integrate all base station time domain received signals and perform discrete Fourier transform in segments to obtain the frequency domain matrix form of the received signal; Step 2: Calculate the covariance matrix of different frequencies and use the covariance matrix to calculate the propagation operator and construct the noise subspace; Step 3: Use the noise subspace to construct the loss function; Step 4: Divide the area to be searched into grids and calculate the loss function value of each grid point; Step 5: Search for the extreme value of the loss function to obtain the radiation source position.
2. The subspace time difference direct positioning method based on the propagation operator according to claim 1 is characterized in that: Step 1 further comprises: There are L observation base stations in the two-dimensional plane, and the lth observation base station is located at S l (X l ,Y l ), each observation station is equipped with only one sensor to capture the signal and each observation station is time synchronized; suppose there are M unknown far-field radiation sources, and they emit uncorrelated broadband signals; let the mth unknown transmitter be located at T m (x m ,y m ), at any time t within the observation time interval, the signal obtained by the lth base station is expressed as: Where T s is the observation time interval, α l,m is the unknown attenuation coefficient on the propagation path of the signal from the mth radiation source to the lth observation station, s m (t) is the signal emitted by the mth radiation source at any time t within the observation time interval, w l (t) is white zero-mean complex additive Gaussian noise, τ l,m is the time delay of the mth radiation source reaching the lth observation station, expressed as: Where c is the propagation speed of electromagnetic waves, integrating the received signals of all base stations to obtain: Z(t)=[z1(t)z2(t)…z L (t)] T ,0≤t≤T s ; Divide the received signals of each base station into J segments evenly, that is: Z=[Y1(t′)Y2(t′)…Y J (t′)]; Then, for each segment of the received signal Y j (t′) performs N-point DFT and obtains: in, and Y j (t′), z l (t j ),s m (t j ) and w l (t j ) DFT results, f n The frequency corresponding to the nth frequency point; let q be the radiation source position vector, and A(f n ,q)=[a(f n ,q1) a(f n ,q2) … a(f n ,q M )] Then we have: Among them, S j (f n ) represents the nth narrowband frequency domain component of the transmitted signal when the frequency domain snapshot is j, W j (f n ) represents the nth narrowband frequency domain component of the transmitted signal when the frequency domain snapshot is j, A(f n ,q) represents the signal manifold matrix, a(f n ,q m ) is a vector containing attenuation coefficient, delay and frequency information in the n-th narrowband frequency domain; Integrate all frequency domain snapshot data to get the frequency domain matrix form of the received signal:
3. The subspace time difference direct positioning method based on the propagation operator according to claim 1 is characterized in that: Step 2 further includes: Frequency f n The covariance matrix of the corresponding components is: Among them, R S (f n ) is the signal spectral density matrix, R W (f n ) is the noise spectral density matrix; the signal manifold matrix is decomposed into: Among them, A1(f n ,q) is A(f n ,q) first M rows, A2(f n ,q) is A(f n ,q) of the posterior LM; then there exists a unique propagation operator So that the following is true: Among them, I L-M Represents the identity matrix of LM dimension; definition: Then we have: Q nN and A(f n ,q) are orthogonal, for Q nN The column vector of the noise subspace U is orthogonalized nN : The covariance matrix is divided into blocks: Among them, G n Depend on The first M columns of H n Depend on The propagation operator of the nth frequency point is obtained as follows: The corresponding noise subspace is constructed based on the propagation operator.
4. The subspace time difference direct positioning method based on the propagation operator according to claim 1 is characterized in that: Step 3 further includes: Let the search position be p = [x, y] T , when the search position is the radiation source position, according to the orthogonality of the noise subspace and the signal manifold, we get: Where diag(·) represents diagonalization, a(f n ,p) represents the vector of the time delay and frequency information from the target location to each base station, represents the attenuation vector from the target location to each base station, where α1,α2,··,α L Indicates the attenuation coefficient to each base station. Cut G effective frequency bands and integrate G effective frequency bands to get: Define K(p) as: So: Assumptions The first element of is 1, and the conversion is: Where, e1=[1,0,…,0] T , according to the Lagrange multiplier method, the loss function F(p) is: F(p)=e1 H K(p) -1 e1。
Citation Information
Patent Citations
Compression spatial spectrum-based single base MIMO (Multiple Input Multiple Output) radar target DOA (Direction of Arrival) estimation method
CN104251989A
Broadband signal direct positioning method based on multi-frequency cost function fusion under unknown attenuation coefficient
CN115052246A
Direct positioning method of distributed uniform linear array based on unknown mutual coupling
CN115685061A
Method for carrying out multi-radiation-source positioning by using unmanned aerial vehicle group
CN116184316A
Vehicle positioning method and device based on spatial spectrum estimation method
CN116540176A
