Multi-target position and velocity joint estimation method based on terahertz near-field MIMO-ofdm

Abstract

The present invention provides a multi-target position and velocity joint estimation method based on terahertz near-field MIMO-OFDM, comprising: a terahertz near-field MIMO-OFDM model establishment step: establishing a near-field MIMO-OFDM signal model, and representing a received signal in the near-field MIMO-OFDM signal model as a third-order tensor; a tensor decomposition step: decomposing the received signal sensor, and on the basis of the uniqueness of the tensor decomposition, performing estimation by using a factor matrix obtained after the tensor decomposition to obtain estimated values of intermediate parameters; a parameter preliminary estimation step: using the estimated values of the intermediate parameters and a least squares (LS) algorithm for preliminary calculation so as to obtain iteration starting points, and establishing an optimization equation; and a precise positioning and velocity estimation step: performing Taylor expansion and second-order Taylor approximation on nonlinear terms of the optimization equation at the starting points, converting the optimization equation into a convex optimization equation, and using a cvx toolbox to iteratively solve the convex optimization equation to obtain estimated values of the position and the velocity of a target. The present invention improves the accuracy of target parameter estimation by jointly processing an inter-antenna phase difference and an inter-subcarrier phase difference by means of tensor decomposition.

Description

A terahertz near-field MIMO-OFDM multi-target position and velocity joint estimation method TECHNICAL FIELD

[0001] The present application relates to electromagnetic wave target estimation technology, in particular to a terahertz near-field MIMO-OFDM multi-target position and velocity joint estimation technology. BACKGROUND

[0002] In the face of the rapid growth of wireless service demand, wireless technology is developing towards higher frequencies to seek greater bandwidth, so terahertz wireless communication located at 0.1THz to 10THz is considered as a very key technology for future wireless communication. At the same time, the continuous increase of wireless communication frequency makes the frequency band of communication and radar gradually coincide. In the face of limited spectrum resources, integrating the past independent communication system and radar system to realize communication and sensing integration (ISAC) can improve the spectrum efficiency and hardware resources. In particular, terahertz can provide hundreds of GHz of spectrum resources, which can provide large communication capacity and high-precision sensing, so the research on terahertz communication and sensing integration is a very popular topic.

[0003] Due to the high frequency of terahertz, the traditional far-field model is no longer applicable, and a more accurate near-field spherical wave model needs to be used. High frequency means high attenuation, so multiple antennas are needed to improve gain. In order to ensure no positioning ambiguity, the antenna spacing is usually controlled within half a wavelength, which leads to too small antenna spacing and affects the sensing accuracy due to antenna coupling. In terms of integrated waveform, it is difficult for a simple radar waveform to embed information bits. If the OFDM waveform commonly used in wireless communication is used in the near-field spherical wave model, the phase difference between antennas is not only related to the azimuth angle but also related to the distance between the antenna array and the target, which leads to a significant increase in the complexity of parameter estimation in sensing. In summary, the following problems need to be solved:

[0004] 1. The high frequency of terahertz makes the far-field plane wave model no longer applicable. In the case of double-array MIMO active sensing, an accurate terahertz MIMO-OFDM near-field spherical wave model needs to be established.

[0005] 2. Terahertz attenuation is large, multiple antennas are needed to improve gain, and it is necessary to study how to reduce the number of antennas and increase the antenna spacing to avoid coupling without causing position ambiguity and reducing the aperture of the antenna array.

[0006] 3. Using OFDM waveform in the near-field model will significantly increase the complexity of parameter estimation in sensing. How to improve sensing accuracy while reducing algorithm complexity needs further research.

[0007] Many sensing models are based on far-field assumptions. Researchers have used second-order Taylor expansions of near-field spherical models to approximate near-field signal models and reduce model complexity. Research on sensing based on near-field spherical wave models and employing OFDM waveforms commonly used in communications has seen some scholars estimate target parameters by directly performing maximum likelihood gridded search on near-field OFDM multidimensional data. Podkurkov et al. (2018) utilized the advantages of tensors in processing multidimensional data, jointly estimating target parameters through tensor decomposition and least squares algorithms.

[0008] However, applying a second-order Taylor approximation to the near-field spherical wave model introduces systematic errors, affecting the accuracy of parameter estimation. Directly performing maximum likelihood estimation on multidimensional data leads to excessive computational complexity. Podkurkov et al. only used near-field characteristics to estimate target position parameters; as the target moves away from the array, the near-field characteristics weaken, causing a decline in target localization performance. Technical issues

[0009] The technical problem to be solved by this invention is to provide a method for joint estimation of position and velocity of multiple targets in terahertz near-field MIMO-OFDM with lower complexity and higher accuracy. Technical solutions

[0010] The technical solution adopted by this invention to solve the above-mentioned technical problems is a joint estimation method for the position and velocity of multiple targets in terahertz near-field MIMO-OFDM, comprising the following steps:

[0011] Steps for establishing a terahertz near-field MIMO-OFDM model: Establish a near-field MIMO-OFDM signal model, and represent the received signal in the near-field MIMO-OFDM signal model as a third-order tensor;

[0012] The near-field MIMO-OFDM signal model is as follows:

[0013] ;

[0014] in, Indicates from the first root transmitting antenna to the first The received signal of the nth subcarrier under the pth FO-OFDM symbol index in the propagation direction of the root receiving antenna. For the first Group of emitter subarrays to the first The propagation direction of the group receiving subarray The speed of the target Indicates the first The first transmitting antenna to the The propagation direction of the receiving antenna subcarriers in the additive white Gaussian noise under the th FO-OFDM symbol index; transmit antenna to the th target in the propagation direction from the transmit array center to the receive array center, th target, th target; subcarrier spacing, th target in the propagation direction from the transmit array center to the receive array center, single FO-OFDM signal duration, carrier frequency, speed of light;

[0015]

[0016] wherein, reflection coefficient of the th target, modulation symbol of the th subcarrier of the th transmit antenna under the th FO-OFDM symbol index, preset value for making signals of each transmit antenna not intersect in the frequency domain, th target in the propagation distance difference between the th transmit antenna and the transmit array center, th target in the propagation distance difference between the th receive antenna and the receive array center, th target in the propagation distance difference between the th receive antenna and the receive array center, signal propagation delay of the th target;

[0017] tensor decomposition step: decomposing the received signal tensor, based on the uniqueness of the tensor decomposition, using the factor matrix after the tensor decomposition to estimate the estimated value of the intermediate parameters, the intermediate parameters including , and ;

[0018] parameter preliminary estimation step: using the estimated value of the intermediate parameters and the least square (LS) algorithm to obtain the initial value of the coordinate value of the th target relative to the origin on the x-axis , and the​​​ The distance of the signal propagation path from the target to the center antenna of the receiving array initial value ,according to and Solving the geometric relationship between them The azimuth angle of each target relative to the center antenna of the receiving array The estimated value , using parameters express and Iteration initial point and And establish optimization equations;

[0019] Precise positioning and velocity estimation steps: Perform a Taylor expansion of the nonlinear terms of the optimization equation at the initial point, apply a second-order Taylor approximation, transform the optimization equation into a convex optimization equation, and use the CVX toolbox to iteratively solve the convex optimization equation to obtain the desired result. and The estimated value and Then, based on the geometric relationship of the target position, the first... The y-coordinates of the targets relative to the origin The estimated value Finally, the target position parameters and the parameters obtained from tensor decomposition are used. The estimated number is 1. The velocity values ​​of each target on the x-axis and y-axis respectively. The estimated value ;No. The joint estimation of the position and velocity of the target has been completed. Beneficial effects

[0020] The beneficial effects of this invention are that by establishing an accurate terahertz MIMO-OFDM near-field spherical wave model, and utilizing OFDM signals with orthogonal frequencies at the transmitting end, the unambiguous condition is transformed from the antenna spacing to the frequency spacing between the transmitting antennas. By using tensor decomposition to jointly process multi-dimensional information such as the phase difference between antennas and the phase difference between subcarriers, the accuracy of target parameter estimation is improved, and the antenna array is grouped to reduce the complexity of the tensor decomposition algorithm. Attached Figure Description

[0021] Figure 1 shows a model of a terahertz near-field MIMO uniform linear antenna array.

[0022] Figure 2 is a flowchart of the joint estimation of the position and velocity of multiple targets in terahertz near-field MIMO-OFDM based on tensor decomposition.

[0023] Figure 3 shows the grouping of terahertz near-field antennas and the coarse estimation model.

[0024] Figure 4 shows the relationship between the root mean square error of the localization algorithm of the present invention and the least squares algorithm and the target distance.

[0025] Figure 5 shows the relationship between the root mean square error and signal-to-noise ratio of the localization algorithm of the present invention and the least squares algorithm.

[0026] Figure 6 shows the relationship between the root mean square error of positioning and the target distance when the signal-to-noise ratio is 15dB.

[0027] Figure 7 shows the relationship between the root mean square error of target distance and velocity and the target spacing when the signal-to-noise ratio is 15dB.

[0028] Figure 8 shows the relationship between signal bandwidth, root mean square error of positioning, and target distance.

[0029] Figure 9 shows the relationship between signal bandwidth, root mean square error of velocity, and target distance.

[0030] Figure 10 shows the relationship between antenna grouping, root mean square error of positioning, and signal-to-noise ratio. Embodiments of the present invention

[0031] For ease of description, the symbols that appear are explained: subscript or superscript Indicates launch. Indicates receipt; This variable represents the subarray index of the transmitting array. , This represents the subarray index variable of the receiving array. The total number of transmission arrays is The total number of receiver arrays is ; Indicates the number of transmitting array antennas. Indicates the number of receiving array antennas. This represents the number of antennas in a transmit array. This indicates the number of antennas in a receiver array; , ; Indicates the first The first in the group of emitter subarrays Root transmitting antenna, For the first The first group of receiving subarrays Root receiving antenna; For the total target number, For the target ordinal number variable, ; For the first The first of the root transmitting antennas The subcarrier at the _ ... Under the index of FO-OFDM symbols; Indicates the first The first transmitting antenna to the The propagation direction of the receiving antenna The subcarrier at the _ ... Under the index of FO-OFDM symbols; For the first The first goal is to start from the first root transmitting antenna to The direction of propagation of the root receiving antenna; From the center of the launch array to the One goal, Receive array center to the One goal; For the first The first goal is to start from the first The subarray emitted to the reflection was the first Group subarray reception; the total number of signal subcarriers is The subcarrier sequence number variable is The total number of FO-OFDM symbol indices is The FO-OFDM symbol index number variable is The number of transmit / receive antenna pairs formed by a set of transmit subarrays and a set of receive subarrays is: The antenna pair number variable is .

[0032] The distance along the signal propagation path. It is additive white Gaussian noise. For signal propagation delay, In order to be in Time-transmitted signal expression, Indicates modulation symbols, In order to be in Time-received signal / received signal expression This represents the coordinate value on the x-axis. This represents the coordinate value on the y-axis. Due to the difference in propagation distance, This is the speed value. To represent the reflection coefficient, Indicates the azimuth angle.

[0033] As shown in Figure 1, the number of transmitting array antennas is The number of receiving array antennas is The transmitting array and the receiving array are divided into and Each subarray has an adjacent antenna spacing of [number]. Assuming the antenna array is located on the x-axis, the coordinates of the transmitting antenna are: , The coordinates of the receiving antenna are , ,in and They represent the first The first subarray root transmitting antenna and the first The first subarray Root receiving antenna. Number of moving targets: , No. The coordinates of the targets relative to the origin are: The speed is , The first The x- and y-coordinates of each target relative to the origin. For the first The velocity values ​​of each target on the x-axis and y-axis respectively.

[0034] Consider transmitting frequency orthogonal OFDM signals, i.e., FO-OFDM signals. Subcarrier spacing, subcarrier index . The duration of a single FO-OFDM signal. For the duration of the protection interval, For the duration of the effective symbol, If it is the carrier frequency, then in time... The The expression for the root transmitting antenna signal is:

[0035] ;

[0036] in, For FO-OFDM symbol index, For the first The modulation symbol of the nth subcarrier of the root transmit antenna under the pth FO-OFDM symbol index. To ensure that the signals from each transmitting antenna do not intersect in the frequency domain, a preset value is greater than [value missing]. Integers.

[0037] The steps for joint estimation of multi-target position and velocity in terahertz near-field MIMO-OFDM based on tensor decomposition are shown in Figure 2:

[0038] S1, a near-field MIMO-OFDM signal model is established, and grouping is performed according to different antenna subarrays, which specifically includes:

[0039] Step S110, a near-field MIMO-OFDM signal model is established.

[0040] Ignoring the Doppler shift caused by the subcarrier, the received signal from the first transmit antenna to the first receive antenna can be represented as:

[0041] ;

[0042] wherein, is the reflection coefficient of the first target, is the Doppler shift of the first target caused by the carrier, is the speed of light, is the speed of the first target in the propagation direction from the first transmit antenna to the first receive antenna, is the signal propagation delay of the first target in the propagation direction from to , represents the additive white Gaussian noise in the propagation direction from to . The signal propagation delay can be specifically represented as:

[0043] ;

[0044] According to the geometric relationship, the signal propagation path distance of the first target to the center antenna of the transmit array and the signal propagation distance of the first target to the center of the receive array are:

[0045] ;

[0046] wherein, is the coordinate value of the center of the transmit array relative to the origin on the x-axis, is the coordinate value of the center of the receive array relative to the origin on the x-axis.

[0047] The first target is in the first root transmit antenna and the first The difference in signal propagation distance between the root receiving antenna and its respective array center antenna , for:

[0048] ;

[0049] No. The goal is from arrive Signal propagation delay in the propagation direction and distance difference , The relationship can be represented as:

[0050] ;

[0051] For the first The signal propagation delay of a target in the propagation direction from the center of the transmitting array to the center of the receiving array.

[0052] Step S120: Simplify the signal model.

[0053] Assuming the antenna aperture is small enough, there is And relative to the duration of the symbol Propagation delay Very small, with Furthermore, the targets have approximately equal velocities relative to the same subarray antenna. The received signal model is simplified, and a Discrete Fourier Transform (DFT) is performed. The final near-field MIMO-OFDM signal model is as follows:

[0054] ;

[0055] in, Indicates from the first Transmitting antenna to the The received signal of the nth subcarrier under the pth FO-OFDM symbol index in the propagation direction of the receiving antenna. For the first Group of emitter subarrays to the first The propagation direction of the group receiving subarray The speed of the target Indicates the first The first transmitting antenna to the The propagation direction of the receiving antenna The subcarrier at the _ ... Additive white Gaussian noise under the FO-OFDM symbol index; Indicates the first transmit antenna to the receive antenna of the intermediate quantity of the target, the intermediate quantity of the

[0056] .

[0057] Step S130 signal tensor representation:

[0058] For the group of subarray transmit signals reflected by the target and received by the group of subarrays, the non-noise item can be expressed in the form of a third-order tensor satisfying CP decomposition :

[0059] ;

[0060] wherein, represents the tensor outer product, and the vector is expressed as:

[0061] ;

[0062] wherein, represents the sequence number of the transceiving antenna pair from the group of transmit subarrays to the group of receive subarrays.

[0063] In fact, the received signal of the th subcarrier in the propagation direction from the th transmit antenna to the th receive antenna under the th FO-OFDM symbol index is , similar to the construction of the non-noise item third-order tensor , the received signal containing noise can be easily constructed as a third-order tensor .

[0064] S2, decompose the grouped received signal tensor, and estimate the intermediate parameters using the factor matrices after tensor decomposition, and the specific steps include:

[0065] Step S210 CP tensor decomposition

[0066] the third-order tensor of the group of subarray transmit signals reflected by the target and received by the group of subarrays, the rank of which is equal to the number of targets . The third-order tensor with a rank of ​ After CP decomposition, it can be expressed as the sum of three-order rank 1 tensors, and the CP decomposition of the tensor has uniqueness under the Kruskal condition:

[0067] ;

[0068] Generally, the number of OFDM symbols , the number of subcarriers , and the number of antenna pairs will be greater than the target symbol number , which can satisfy the Kruskal condition, because is composed of known or constant, so the tensor After CP decomposition, the vectors of the factor matrix can be obtained as follows:

[0069] .

[0070] Step S220 estimates the intermediate parameters

[0071] The vector is represented by the time delay , so the time delay can be estimated by averaging the vector :

[0072] ;

[0073] wherein represents a function of taking the phase angle of a complex number;

[0074] The vector is represented by the velocity , so the velocity can be estimated by averaging the vector :

[0075] ;

[0076] Similarly, the vector is represented by the distance difference , so the distance difference can be estimated by averaging the vector :

[0077] ;

[0078] wherein .

[0079] S3, estimate the parameters , and establish an optimization equation, specifically including:

[0080] Step S310 Preliminary parameter estimation :

[0081] Distance difference The expression contains parameters The information is used to estimate intermediate parameters after tensor decomposition. Then you can Make a preliminary estimate, for The parameters can be estimated by expanding the expression and using the minimization algorithm. initial value for:

[0082] ;

[0083] in, It contains only the known antenna coordinates and the estimated distance difference. , and All are intermediate values.

[0084] Step S320: Estimate parameters :

[0085] As shown in Figure 3, the parameters It is the azimuth angle of the target relative to the center antenna of the receiving array. From geometric relationships, we know that:

[0086] ;

[0087] Literature indicates that the estimated value of the azimuth angle... The estimation has high accuracy, and its accuracy remains almost constant as the distance from the target to the antenna array increases.

[0088] Step S330: Establish the optimization equation:

[0089] Target position parameters Not only with distance difference Related to time delay It also includes target location information and joint parameters. and It can improve positioning accuracy. However, and These are all estimated intermediate parameters, which contain estimation errors. Substituting them into the equation will not strictly satisfy the equation. Therefore, the following optimization equation can be constructed:

[0090] ;

[0091] in, denote the target position parameter, for F-norm, for the estimate of for the estimate of In order to reduce the iteration process, a good initial point needs to be selected, and the preliminary estimate of will have a large error when the target distance is far away, while still has high accuracy, so according to the geometric relationship, the parameter is used to represent the initial point :

[0092] .

[0093] S4, convert the equation into a convex optimization problem, solve the optimization equation to further estimate the parameter , and estimate the target position and velocity, which specifically includes:

[0094] Step S410 converts the equation into a convex optimization problem:

[0095] It can be seen that the optimization equation has a nonlinear term, which makes the optimization non-convex, and it is very difficult to solve directly. The Taylor expansion is performed on the nonlinear term at the initial point:

[0096] ;

[0097] The nonlinear term is approximated by the first order linear, so the optimization equation becomes the following convex optimization equation:

[0098] ;

[0099] where, is the L2 norm, the intermediate quantity , and the intermediate quantity .

[0100] Step S430 estimates the velocity :

[0101] As shown in FIG. 3, the target velocity is composed of two parts: , where, and + are the radial velocities relative to the transmitting and receiving sub-arrays, respectively. The projections of the radial velocities of the transmitting and receiving arrays on the

[0102] coordinate axis are superimposed to be , respectively. The target velocity can be estimated using the target position parameter obtained and the parameter obtained by tensor decomposition:

[0103] ;

[0104] in,

[0105] , .

[0106] Step S440: Estimate the value As the first The estimated location of the nth target will be used as the basis for determining the nth target's position. The speed of the target Finally, the positions and velocities of L targets are output to complete the joint estimation.

[0107] Complexity analysis:

[0108] The computational complexity of key steps in this invention includes CP decomposition and parameter... The estimation and solution of the convex optimization equations. The computational complexity of tensor decomposition for each set of received data is:

[0109] ;

[0110] Each group The estimated computational complexity is Meanwhile, the computational complexity of locating the target location by solving an optimization problem is... , where T is the iteration number. Due to the existence of There are groups, therefore, the total computational complexity is: At the same time, it can be seen that when and In comparison, it can be ignored. Finally, we can obtain:

[0111] ;

[0112] As can be seen, the computational complexity varies with the number of groups. When Greater than and In this case, appropriately increasing the number of groups can reduce the complexity caused by the number of antennas, thereby reducing the overall complexity.

[0113] Simulation test

[0114] The simulation results further demonstrate the performance of the proposed implementation. In Figures 4 to 10, the system parameters were set as follows: the system carrier frequency was... GHz, subcarrier spacing kHz, OFDM symbol duration us. The spacing between adjacent antennas is... cm, total number of transmit antennas , total number of receive antennas . Target region is set to , . Without any special instructions, the transmit and receive arrays are divided into , and the signal-to-noise ratio SNR is defined as:

[0115] ;

[0116] The root mean square error (RMSE) of positioning is defined as:

[0117] ;

[0118] The root mean square error (RMSE) of velocity is defined as:

[0119] .

[0120] Figures 4 and 5 show the relationship between the RMSE of the algorithm, the distance from the target to the origin, and the signal-to-noise ratio, respectively. At the same time, the algorithm proposed in this application is compared with the near-field least squares algorithm proposed by Podkurkov et al. in 2018. As can be seen from Figure 4, the estimation performance of the proposed algorithm is significantly better than that of the least squares algorithm. In general, the accuracy of the algorithm decreases with the increase of the distance. But due to the use of a more refined signal model based on spherical wavefront, the performance of the algorithm proposed in this application decreases slower than the near-field least squares algorithm. In Figure 5, the target distance is set to 5m. It can be seen that with the improvement of the signal-to-noise ratio, the estimation performance of the two algorithms has improved. At the same time, compared with the existing LS algorithm, the estimation performance of the algorithm proposed in this application is better, and the calculation speed is faster.

[0121] In Figure 6, two targets are set, the distance from the first target to the origin (r1) is set to 10m, 20m, 30m and 200m. The x-coordinate represents the distance d between the two targets, r2 is the distance from the second target to the origin. It can be seen that the estimation accuracy increases with the decrease of r1. In addition, when the two targets are too close, it is difficult to distinguish the two targets, so the estimation accuracy decreases with the decrease of the distance between the two targets. In addition, when d is greater than 0.21m, the estimation performance increases slightly with the increase of d, and when d is greater than 0.068m, the estimation accuracy in the near-field scene can reach the centimeter level. In Figure 7, the velocities of the two targets are set to 0.5m / s and 0.7m / s, respectively. Since the estimation accuracy of the velocity is affected by the estimation accuracy of the position, it can be seen that the change trend of its estimation accuracy is the same as that in Figure 6.

[0122] In FIG. 8 and FIG. 9, r1 = 10m is set. The bandwidth is from 15MHz to 200MHz. It can be seen that the estimation accuracy gradually increases with the increase of bandwidth. In addition, when the distance between the two targets is greater than 0.030m, 0.044m and 0.15m respectively in the case of 120MHz, 30MHz and 15MHz, the estimation accuracy of the algorithm proposed in the present application reaches centimeter level. And it can be seen that with the increase of bandwidth, the resolution of the two targets also increases.

[0123] FIG. 10 illustrates the achievable estimation accuracy with respect to the number of subarray groups in the receiving array. The transmitting antenna array is generally considered as a group = 1). The number of receiving antennas is = 60. It can be seen from FIG. 10 that with the increase of the number of subarrays, the positioning estimation accuracy decreases, but when the signal-to-noise ratio is greater than -4dB, this deterioration disappears, which is usually the working condition of the system. On the other hand, as analyzed before, dividing the receiving antennas into more groups can reduce the computational complexity of the algorithm. Therefore, a reasonable choice can be made between the complexity of the algorithm and the estimation accuracy. And it can be seen from FIG. 10 that with the increase of the signal-to-noise ratio, the number of groups has little effect on the estimation performance. Therefore, when the signal-to-noise ratio is high, dividing the antennas into more groups can better reduce the computational complexity.

Claims

1. A method for joint estimation of multi-target position and velocity in terahertz near-field MIMO-OFDM, characterized in that, The method comprises the steps of: The step of establishing a near-field MIMO-OFDM model: the received signal in the near-field MIMO-OFDM signal model is expressed as a third-order tensor; The near-field MIMO-OFDM signal model is: ； wherein Indicates from the first Root transmit antenna to first a received signal of an nth subcarrier in a propagation direction of a reception antenna at a pth FO-OFDM symbol index, For the first group transmit subarray to the the first sub-array of the group of reception sub-arrays in the direction of propagation the speed of the target, represents the 1 Root transmit antenna to the first the direction of propagation of the receiving antenna one subcarrier in the first an additive white Gaussian noise under a FO-OFDM symbol index; represents the 1 Transmit antenna to first the first intermediate quantities of the target, For the first intermediate quantities of the target; for subcarrier spacing, For the first a signal propagation delay of the target in the propagation direction from the center of the transmitting array to the center of the receiving array, for a single FO-OFDM signal duration, for the carrier frequency, The speed of light is c. ； wherein, For the first a reflection coefficient of the target, For the first first one subcarrier in the first modulation symbols under a FO-OFDM symbol index, a preset value for each transmitting antenna signal not to intersect in the frequency domain, For the first The object is in the first The propagation distance difference between the root transmit antenna and the center of the transmit array, For the first The object is in the first The difference in propagation distance between the receiving antenna and the center of the receiving array is d. Tensor decomposition step: decomposing the received signal tensor, based on the uniqueness of the tensor decomposition, estimating the intermediate parameters including 、 and ； Parameter preliminary estimation step: the preliminary solution of the first equation is obtained by using the estimated value of the intermediate parameter and the least square (LS) algorithm coordinate value of the target relative to the origin on the x-axis initial value of the counter and the first a target to a receive array center antenna signal propagation path distance initial value of the counter , according to and solving the geometric relationship between the first azimuth of a target relative to a center antenna of a receiving array estimated value of the number of the users , using parameters denotes and iteration initial point and And an optimization equation is established. Precise positioning and speed estimation step: Taylor expansion of the nonlinear term of the optimization equation at the initial point is performed to obtain a second-order Taylor approximation, the optimization equation is converted into a convex optimization equation, and the convex optimization equation is solved iteratively using the cvx toolbox to obtain and estimated value of the number of the users and Further, the first target position is obtained according to the target position geometry relationship coordinate value of the target relative to the origin on the y-axis estimated value of the number of the users ; finally using the target position parameters obtained and the parameters obtained from the tensor decomposition estimated to be the first velocity values of the respective targets on the x- and y-axes estimated value of the number of the users ; 1st The joint estimation of the position and speed of the target is completed.

2. The method of claim 1, wherein, The received signal in the near-field MIMO-OFDM signal model is expressed as a third-order tensor, and the specific expression is as follows: ； wherein For the first The group sub-array transmit signal is reflected by the target and received by the first The third-order tensor of the group sub-array receiving signals, whose rank is equal to the target number ； denotes the tensor outer product; tensor After CP decomposition, the vectors that constitute the factor matrix are obtained as 。 3. The method of claim 2, wherein, the vectors of the factor matrices resulting from the tensor decomposition the average estimate the estimate of the average the average estimate the estimate of the average the estimate of the average .

4. The method of claim 3, wherein, solving Specifically: ； wherein the intermediate quantity ， ， denotes the transpose, For the first The root receiving antenna has a coordinate value of the origin on the x-axis; denotes the total number of receive array antennas, the first The root receiving antenna is the center of the receiving array; For the first The root receiving antenna has a coordinate value of the origin on the x axis, For the first a target first Root receiving antenna and signal propagation distance difference from center antenna of receiving array 。 5. The method of claim 4, wherein, According to and the geometric relationship between the first and the second target relative to the center antenna of the receiving array The estimated value of the azimuth angle is specifically: 。 6. The method of claim 5, wherein, Usage parameters Indicates And Iteration initial point And Specifically: ； wherein the intermediate quantity ， The coordinate value of the center of the transmitting array on the y-axis relative to the origin is y.

7. The method of claim 6, wherein, The convex optimization equation is as follows: ； wherein the intermediate quantity , intermediate amount ， For an estimated value of the number of times of the user's use of the application, For estimated value of the target position parameter For the first The signal propagation path distance from the target to the center of the receiving array is d.

8. The method of claim 1, wherein, According to the target position geometry relationship, the estimated value of the coordinate value of the target relative to the origin on the y-axis is obtained Specifically:​​ ； The coordinate value of the center of the receiving array on the x-axis relative to the origin is x.

9. The method of claim 8, wherein, Finally, the target position parameters obtained and the parameters obtained by tensor decomposition are used to estimate the velocity values of the target in the x-axis and y-axis respectively of the target in the x-axis and y-axis respectively Specifically: ； wherein denotes the transpose, for the intermediate quantity, the total number of transmit arrays is , the total number of receiving arrays is ， for by one a vector of components, ， ； ， 。

Images