A sonar distance-angle joint estimation method, device and equipment with fusion error correction

Abstract

The present invention belongs to the technical field of sonar signal processing, and in particular, relates to a sonar distance-angle joint estimation method, device and equipment with fusion error correction, the method comprising: preprocessing the received orthogonal frequency division multiplexing sonar signal to remove the cyclic prefix and tail samples of the received signal; performing fast Fourier transform on the preprocessed received signal to calculate the shifted modulation symbol sequence; according to the shifted modulation symbol sequence, using the multi-target interference suppression characteristics and beamforming technology, obtaining a rough estimate of the distance and angle of the target; constructing an optimization problem in matrix form, and gradually updating the sparse signal and phase error matrix through an iterative algorithm until a predetermined convergence standard is reached to obtain the distance and angle estimation results of the target. It has broad application prospects in system integration, underwater target detection, positioning and tracking, and is a promising research topic for the future development of sonar systems.

Description

Technical Field

[0001] The present invention belongs to the technical field of sonar signal processing, and in particular relates to a sonar distance-angle joint estimation method, device and equipment integrating error correction. Background Art

[0002] Orthogonal Frequency Division Multiplexing (OFDM) sonar offers interesting and promising applications in military or civilian target localization. Compared with traditional sonar systems, OFDM sonar is more sensitive to frequency offsets. Random phase errors caused by array uncertainty and acoustic wave propagation characteristics seriously affect the estimation accuracy.

[0003] In recent years, sonar detection using acoustic communication signals has received considerable attention due to its small installation space, low implementation cost, and high spectrum efficiency. With the great success of Orthogonal Frequency Division Multiplexing (OFDM) technology in communication, OFDM sonar system has been accepted as a promising scheme to improve the performance of traditional sonar systems using Linear Frequency Modulation (LFM) signals or Frequency Modulated Continuous Wave (FMCW) signals. Target localization is one of the most important tasks of sonar systems, whether for military or civilian purposes. Joint range-angle estimation in OFDM systems has become a hot topic.

[0004] Based on OFDM systems, researchers have proposed various range-angle estimation methods. The multiple signal classification (MUSIC) algorithm is applied in OFDM wireless positioning systems to estimate the angle of arrival and time of arrival (TOA) of each path. The matrix pencil (MP) method is proposed for angle and TOA estimation in OFDM communication systems. In recent years, MUSIC and Estimation of Signal Parameters through Rotational Invariant Technique (ESPRIT) have been introduced into range-angle estimation in OFDM radar systems. For OFDM signals, a combination of ESPRIT and MUSIC methods is used for range-angle estimation. Subspace high-resolution estimation methods for OFDM radar systems have good accuracy, but the computational cost is high due to eigenvalue decomposition and spectral peak search. Compressed sensing (CS) technology provides a new idea for joint range-angle estimation by solving the sparse-driven optimization problem, which enables accurate reconstruction of sparse signals from limited measurements. A low-complexity joint range-angle estimation method for OFDM radar based on CS has been proposed. Super-resolution range-angle estimation is recovered by constructing a redundant dictionary and solving the optimization problem.

[0005] The above-mentioned joint distance-angle estimation method is mainly used in wireless communication and radar systems in the air. However, the scheme discussed in this application focuses on active sonar systems. The sensor position or intrinsic parameters in the sonar system are not completely accurate, which will introduce unknown phase errors at each array element. The characteristics of sound wave propagation in the ocean, such as ocean turbulence and velocity changes, make the implementation of sonar systems face significant difficulties. In order to meet the accuracy requirements of wavelength-level angle estimation, array and propagation errors cannot be ignored. Efficient phase error compensation is crucial to sonar systems, so some phase error compensation methods are proposed. The maximum a posteriori (MAP) method is proposed to estimate the signal of interest in the presence of array errors, but the amount of memory and processing calculation required is too large. A low-complexity eigenstructure method is proposed for estimating angles and array errors. A CS-based method is also proposed to solve the measurement distortion problem caused by unknown errors by solving the recovery optimization of sparse signals and errors.

[0006] The received signal is a superposition of multiple target echoes. Target echoes from different range units can cause mutual interference, just like the inter-symbol interference (ISI) caused by multipath in communications. Referring to the OFDM technology of adding a cyclic prefix (CP) in communications, target interference can be suppressed in the OFDM sonar system. At the same time, OFDM is very sensitive to time and frequency offsets, especially in the presence of propagation media and array errors. Therefore, how to overcome this limitation is one of the main research directions of OFDM sonar systems. Summary of the invention

[0007] The purpose of the present invention is to overcome the defects of the prior art and propose a sonar distance-angle joint estimation method, device and equipment with fusion error correction.

[0008] In order to achieve the above object, the present invention proposes a sonar distance-angle joint estimation method with fusion error correction, comprising:

[0009] Preprocessing the received OFDM sonar signal to remove the cyclic prefix and tail samples of the received signal;

[0010] Performing fast Fourier transform on the preprocessed received signal to calculate the shifted modulation symbol sequence;

[0011] According to the shifted modulation symbol sequence, the multi-target interference suppression characteristics and beamforming technology are used to obtain a rough estimate of the target's distance and angle;

[0012] An optimization problem in matrix form is constructed, and the sparse signal and phase error matrices are gradually updated through an iterative algorithm until the predetermined convergence criterion is reached, and the distance and angle estimation results of the target are obtained.

[0013] Preferably, the OFDM sonar comprises M arrays of equally spaced linear arrays, the spacing between adjacent arrays is d=λ / 2, where λ is the signal wavelength, and the transmitted OFDM signal s(t) is:

[0014]

[0015] Where N is the number of subcarriers, k is the subcarrier number, Δf is the subcarrier spacing, Δf = B / N, B is the bandwidth, w k is a modulation symbol sequence obtained by binary data and orthogonal amplitude modulation, j represents the imaginary part, and t represents the time domain.

[0016] Preferably, the received signal g of the mth array in the orthogonal frequency division multiplexing sonar m (t) is:

[0017]

[0018] Among them, a(R,θ) is the reflection function of the target, R is the discrete distance set, θ is the angle, exp[je m (θ)] represents the phase error and c is the speed of light.

[0019] Preferably, the shifted modulation symbol sequence w CP for:

[0020]

[0021] Among them, N CP is the cycle shift amount.

[0022] Preferably, a rough estimate of the distance and angle of the target for:

[0023]

[0024] Where A(n,q) is the element of matrix A, n is the number of n-th subcarriers, q is the q-th angle grid, ρ R is the distance resolution, is a rough estimate of the angle, is a rough estimate of the time delay, and then a rough estimate of the distance can be obtained, Q is the number of angle grids;

[0025] Preferably, the matrix form is: a matrix form of performing an N-point Fourier transform on the m-th array element, satisfying the following formula:

[0026] Y=WAVE+N

[0027] Among them, matrix Y is the measurement data, matrix A and matrix N correspond to the target coefficient and additive white Gaussian noise respectively, matrix V represents the steering vector, matrix E represents the phase error terms of different array elements, and matrix W represents the modulation symbols after displacement.

[0028] Preferably, the optimization problem in matrix form is:

[0029]

[0030] Among them, J(a,E) is the objective function, y=vec(Y), a=vec(A), Z E is the dictionary matrix, ρ represents the regularization parameter, are the estimated values ​​of the coefficient vector and phase error term, respectively.

[0031] Preferably, the method of gradually updating the sparse signal and the phase error matrix through an iterative algorithm until a predetermined convergence criterion is reached to obtain the distance and angle estimation results of the target comprises:

[0032] Step s1) l = 0, initialize the corresponding coefficient vector and the estimated value of the phase error term respectively and

[0033] Step s2) Calculate the Hessian matrix Update by solving the following

[0034]

[0035] Update the angle estimate by

[0036]

[0037] Among them, ∠[·] represents the complex angle, Ξ m is an M×M square matrix, except that the mth diagonal element is equal to e m All elements are zero; (.) H is the conjugate of the matrix; is the Kronecker product;

[0038] Step s3) Determination If the judgment result is true, execute l+1 and go to step s2); otherwise, go to step s4);

[0039] Step s4) Output the distance estimate of the target and angle estimation

[0040] On the other hand, the present invention provides a sonar distance-angle joint estimation device with fusion error correction, comprising:

[0041] A preprocessing module, used for preprocessing the received OFDM sonar signal to remove the cyclic prefix and tail samples of the received signal;

[0042] A calculation module, used for performing fast Fourier transform on the preprocessed received signal and calculating the shifted modulation symbol sequence;

[0043] A coarse estimation module, used to obtain a coarse estimate of the distance and angle of the target based on the shifted modulation symbol sequence, using multi-target interference suppression characteristics and beamforming technology; and

[0044] The optimization iteration module is used to construct the optimization problem in matrix form. Through the iterative algorithm, the sparse signal and phase error matrix are gradually updated until the predetermined convergence standard is reached to obtain the distance and angle estimation results of the target.

[0045] In a third aspect, the present invention further provides a computer device, comprising a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the above method.

[0046] Compared with the prior art, the advantages of the present invention are:

[0047] The present invention proposes a novel distance-angle estimation and phase error compensation method for OFDM sonar systems. Due to inaccurate sensor position, system intrinsic parameters, ocean turbulence and velocity changes, phase error needs to be additionally considered. Utilizing the structure of the OFDM sonar signal model and taking into account the phase error, the optimization problem recovers a joint distance-angle estimation with high accuracy and low sidelobe. In the presence of phase error and strong noise, the estimation accuracy and computational efficiency are improved. The new joint distance-angle estimation method for OFDM sonar systems has broad application prospects in system integration, underwater target detection, positioning and tracking, and is a promising research topic for the future development of sonar systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a flow chart of the sonar distance-angle joint estimation method with fusion error correction of the present invention;

[0049] Figure 2 are waveforms with different normalized range distributions;

[0050] FIG3 is a normalized distance estimation based on OFDM waveforms with different CP lengths, where FIG3(a) is an OFDM waveform with CP length = 100, and FIG3(b) is an OFDM waveform with CP length = 0;

[0051] FIG4 is a normalized distance and angle estimation profile using different methods and waveforms, wherein FIG4(a) is a normalized distance estimation profile, and FIG4(b) is a normalized angle estimation profile;

[0052] FIG5 is a diagram of the behavior of phase compensation, where FIG5(a) is a normalized angle estimation profile and FIG5(b) is a phase error estimation result;

[0053] Figure 6 It is a comparison of the accuracy of distance and angle estimation. DETAILED DESCRIPTION

[0054] The present invention proposes a distance-angle estimation method for a sonar system based on compressed sensing and fusion error correction. In the implementation of the method, a double iterative algorithm is used to achieve high-resolution distance-angle estimation and error correction, solve the challenges brought by phase error and target parameter estimation in the OFDM sonar system, and the method also uses properties such as Hadamard product to accelerate the calculation of matrix multiplication.

[0055] 1 Principle Description

[0056] The method processes Orthogonal Frequency Division Multiplexing (OFDM) sonar signals and achieves high-precision joint distance and angle estimation in the presence of phase errors.

[0057] This method first removes the CP and tail samples of the received signal. The second step is to perform discrete Fourier transform (DFT) on the received signal and obtain w CP The third step is to divide the received signal by w CP , and perform IDFT processing on it to estimate the distance of the target. The fourth step is to estimate in the angle domain through beam forming technology, thereby realizing joint distance-angle estimation. The fifth step is to gradually update the sparse signal and phase error matrix through an iterative algorithm until the predetermined convergence standard is reached. The method flow chart is shown in Figure 1 shown.

[0058] 2 Method Structure

[0059] 2.1 Signal Model

[0060] Sonar is a single-input multiple-output (SIMO) system in which the M array elements are equally spaced linear arrays. The spacing between adjacent elements is d = λ / 2, where λ is the signal wavelength. The transmitted OFDM signal with symbol period T, number of subcarriers N and bandwidth B is described as follows:

[0061]

[0062] Where Δf = B / N is the subcarrier spacing, wk is a modulation symbol sequence obtained by binary data such as phase shift keying (PSK) and quadrature amplitude modulation (QAM), and The sampling rate is T s =T / N=1 / B, the OFDM signal after sampling is:

[0063]

[0064] The sampled OFDM signal is the modulation symbol w = [w0, w1, ..., w N-1 ] T Each OFDM symbol is added with a length of N CP The total transmitted OFDM length is T+T CP , as shown below:

[0065]

[0066] Considering the carrier frequency f c , the OFDM signal transmitted after adding CP is:

[0067]

[0068] We consider a two-way propagation case and choose the first element as the reference point. For a far-field point target of the distance-angle pair (R,θ), the delay of the mth array element is expressed as:

[0069]

[0070] Where c is the speed of light and v is the target speed. After down-conversion to baseband, the received signal of the mth array element is expressed as:

[0071]

[0072] Where a(R,θ) is the reflectance function of the target. As we all know, the requirement of distance estimation accuracy matches the distance resolution, Δτ m (θ) can be neglected. However, the angle estimation accuracy requirement is matched to the wavelength. Sensor position inaccuracy, system intrinsic parameters, ocean turbulence, and velocity variations also affect the estimation performance and should be considered in the signal model.

[0073] exp[je m (θ)] corresponds to the time delay difference Δτ between array elements m(θ) = (m-1)dcos(θ) / c-2vt / c, the phase error caused by array uncertainty and acoustic wave propagation error at the m array element. In the above approximation exp(-j4πvt / λ) is also neglected. We discretize the sonar monitoring area into a P×Q grid in the range and angle domain. The discrete range set is denoted by R = [R0,...,R p ,...R p-1 The discrete angle set is represented by Θ = [θ0, ..., θ q ,...,θ Q-1 ]. The distance resolution is ρ R =c / 2B=cT s / 2. In the pth distance unit, in formula (6) The approximate calculation of is expressed as:

[0074] 2R p / c≈2pρ R / c=pT s . (7)

[0075] Assume that the distance and angle of the target belong to (R, Θ) and at t = lT s (l=0,1,...,N+N CP +P-2), the received signal of the mth array element is expressed as:

[0076]

[0077] The elements of the matrix A are and

[0078]

[0079] 2.2 Phase Error Compensation and Range-Angle Estimation in OFDM Sonar Systems

[0080] Due to its excellent characteristics, OFDM technology has been introduced into the development of sonar / radar systems and has become a hot topic of research. Researchers in the field of sonar detection have noticed this topic and proposed a CS-based OFDM sonar parameter estimation method.

[0081] Similar to the application of OFDM technology in communication, in the sonar parameter estimation application, we assume that:

[0082] To ensure the orthogonality of multi-target echoes, the length of CP satisfies N CP ≥P-1.

[0083] To get a non-interfering build, remove the N at the head CP sample.

[0084] To reflect the impact of all targets, the P-1 samples at the tail need to be removed.

[0085] Remove N from the head CP After removing P-1 samples from the tail, the received signal of the mth array element is expressed as:

[0086]

[0087] Based on the basic principles and methods of using DFT to calculate linear convolution, we perform N-point DFT on the m-th array element and obtain:

[0088] y m =x m w CP , (12)

[0089] where y m and x m They are g m and h m DFT of w CP is the vector s(l+N CP )(l=0,1,...,N-1), which is the cyclic shift N of vector (2) CP , can be calculated by the following formula:

[0090]

[0091] Receive signal x m The DFT calculation formula is:

[0092]

[0093] x m After performing an N-point IDFT, the target coefficient of the m-th array element on the distance set R can be estimated:

[0094]

[0095] Then, by using the multi-target interference suppression feature, the high-resolution distance can be estimated. After applying the beamforming technique, the joint range-angle estimate is:

[0096]

[0097] The phase error exp[je m (θ)] caused by Δθ m It directly affects the accuracy of parameter estimation.

[0098] Taking into account the inevitable noise, the signal model in formula (12) is constructed in matrix form:

[0099] Y=WAVE+N, (17)

[0100] The matrix is the measured data, and y m =DET[g m (N CP ),g m (N CP +1),g m (N CP +N-1)] T .matrix and matrix Corresponding to the target coefficient and additive white Gaussian noise respectively. The matrix represents the steering vector, where:

[0101] matrix represents the phase error terms of different array elements. The matrix It is expressed as [w CP,0 ,w CP,1 ,...,w CP,N-1 ] T ,in As shown in formula (13).

[0102] Using vectorization and Kronecker product, the two-dimensional model is converted to a one-dimensional model as follows:

[0103]

[0104] Where y=vec(Y), a=vec(A).

[0105] Since the number of array elements M is less than the number of angular grids Q, reconstructing a from the measured data y is an ill-posed problem. CS technology is an effective method to solve the ill-posed inverse problem by using sparse prior information of signal a and the restricted isometry property (RIP) of the dictionary matrix ZE. CS requires that a is sparse and the coefficient vector a has high sparsity because the number of targets is significantly less than the number of grids in the range-angle plane. RIP is related to the non-correlation property of the dictionary matrix. In the range-angle estimation of OFDM sonar, it is proved that the dictionary matrix can meet the RIP requirement with high probability. Ignoring the phase error term, the l1 norm regularization method is used to solve the range-angle estimation problem in the OFDM sonar system as follows:

[0106]

[0107] where ρ represents the regularization parameter. Many sparse recovery methods have been proposed to solve the minimum l1 norm reconstruction problem, such as orthogonal matching pursuit (OMP), adaptive moment estimation (ADaM), and Broyden-Fletcher-Goldfarb-Shanno (BFGS).

[0108] Due to the presence of phase error, high-precision range-angle estimation can be obtained by sparse recovery methods, which cannot be directly applied to recover a from y in (19). Phase error compensation for sonar detection should be introduced. Existing works on sparse phase error compensation are mainly based on traditional (LFM / FMCW) radar / sonar systems, while OFDM sonar is rarely discussed.

[0109] This paper proposes a distance-angle estimation method for OFDM sonar system in the presence of phase error based on compressed sensing. The phase error term E is random and has nothing to do with V and W, ensuring that the dictionary matrix Z E The non-correlation characteristics and RIP of . The sparse optimization problem becomes:

[0110]

[0111] In this section, an improved quasi-Newton algorithm is developed based on dual iteration and Hessian update scheme to achieve sparse signal reconstruction in the presence of phase error. The two unknown quantities a and E are obtained by utilizing their derivatives. Let us recall that we initialized The steps of the distance-angle estimation method based on OFDM sonar are as follows: first, remove the CP and tail of the received signal; then, perform DFT on the signal received in the time domain and calculate w CP ; In addition, each array element of the signal obtained in the second step is divided by w CP , and apply IDFT in the time domain to estimate the distance of the target; finally, the angle is estimated by applying beamforming technology in the angle domain. Without any help of phase error information, the matrix The initialization is I, ie(e m ) 0 =0,(0≤m≤M-1).

[0112] To overcome the non-differentiable nature of ||a||1 near the origin, the following smooth approximation is used:

[0113]

[0114] where δ is a small non-negative number, approximately 0.1% of the noise variance.

[0115] First, the derivative of the objective function J(a,E) with respect to a is solved by the following formula:

[0116]

[0117] The approximate expression of the Hessian matrix is:

[0118] H(a,E)=(Z E ) H Z E +ρ·Λ(a), (24)

[0119] in The Hessian matrix H(a,E) is also a function of a. Solve H(a,E)a=(Z E ) H y cannot directly return the estimated result. Use the quasi-Newton method to solve the problem:

[0120]

[0121] in and They represent the l-th estimation of a and E respectively, is the dictionary matrix updated in the lth iteration. γ is the step size. Substituting (24) into (25), assuming γ = 1, rewrite the lth update iteration as:

[0122]

[0123] Second, to estimate E in the quasi-Newton iteration, we compute the objective function with respect to e m The derivative of , which is the mth element on the diagonal of E. We rearrange the objective function as follows:

[0124]

[0125] For each element e, the following formula is used m Derivative of the objective function:

[0126]

[0127] Among them m is an M×M square matrix, except that the mth diagonal element is equal to e m All elements are zero. b m and β m Signal The amplitude and phase of the objective function are about e i The second-order derivative of is calculated as follows:

[0128]

[0129] By combining the zero points of the first-order derivative and the positive values ​​of the second-order derivative, the objective function can be minimized and the angle estimate is obtained at the lth iteration as follows:

[0130]

[0131] where ∠[·] represents a complex angle.

[0132] and The iterations continue until the algorithm converges to a predefined threshold η.

[0133]

[0134] According to the above analysis and derivation, the phase error compensation and distance-angle estimation process methods of the OFDM sonar system are shown in Table 1.

[0135] Table 1 Phase error compensation and distance-angle estimation process methods for OFDM sonar system

[0136]

[0137] The main computational cost of the proposed method in each iteration is the cost of matrix inversion and matrix multiplication. Fast matrix inversion and matrix multiplication are the key to the method. In order to improve the efficiency of the method, the conjugate gradient algorithm (CGA) is used to solve the inverse matrix of H(a,E). We can obtain This requires a lot of repeated calculations as follows:

[0138]

[0139] in is the updated value of the CGA algorithm in each iteration. We adopt the extended formula in formula (19) and simplify the first term of formula (32) as follows:

[0140]

[0141] in is a vector The matrix form of W. H W=I N , where I N represents an N×N identity matrix. We can further simplify (33) as follows:

[0142]

[0143] It is a diagonal matrix, and the Hadamard product is used to speed up the second term of matrix multiplication:

[0144]

[0145] in Depend on The diagonal elements of .

[0146] The main computational cost of solving for a in each iteration is as follows:

[0147] According to I N Matrix properties, calculations Only addition operations are required.

[0148] ·calculate It is equivalent to three matrix multiplication, and the computational complexity is O(2NQM).

[0149] Using the Hadamard product property, calculate The computational cost is O(NQ).

[0150] ·Let the number of iterations of each iteration to solve (32)CGA be L CGA , the calculation is about O[L CGA (2NQM+NQ)].

[0151] The angle estimate in the lth iteration can be simply calculated as:

[0152]

[0153] in Y(:,m) and Π l (:,m) are the original return signal and Π l The mth column of .

[0154] The main computational cost of solving E in each iteration is as follows:

[0155] ·calculate It is equivalent to three matrix multiplication, and the computational complexity is O(NQM+QN 2 ).

[0156] Calculate Y(:,m) H Π l The computation cost of (:,m) is O(N).

[0157] · The computational cost is approximately O[M(NQM+QN 2 +N+1)].

[0158] In summary, the computational cost of the proposed method is approximately:

[0159] O[LL CGA (2NQM+NQ)]+O[LM(NQM+QN2 +N + 1)] is not a serious problem under the current computing power conditions.

[0160] The fast method of the OFDM sonar system is shown in Table 2.

[0161] Table 2 Fast Method of OFDM Sonar System

[0162]

[0163] The OFDM system is very sensitive to phase errors and frequency synchronization. When two or more targets with different range cells are placed together in the spectrum, they will interfere with each other and confound their estimates.

[0164] First, as Figure 1 shown, the proposed method constructs a signal model by selecting appropriate parameters and derives g m (l) = h m (p,q)*s(l + N CP ) The three conditions in (11) are as follows:

[0165] · If N > P, there are some zeros in the estimation vector A. [A(P,:),..., A(N - 1,:)] represents the non - existent reflection coefficients.

[0166] · If N = P, the estimation vector A(:,q) is exactly the reflection coefficient of the angle θ q .

[0167] · If N < P, according to the periodicity of the DFT, s contains multiple periods. g m (l) The received signal is the sum of the reflection coefficients of different range cells in a sub - carrier, that is, the signal in one range cell will interfere with other OFDM symbols, resulting in target interference.

[0168] In summary, the number of sub - carriers of the OFDM signal is at least equal to the number of range cells, that is, N ≥ P. In addition, at the receiving end, we need to remove the CP and the tail from the received signal. To reduce the redundancy between signal pulses, we specify a rule that the number of sub - carriers is equal to the number of range cells.

[0169] Second, in the OFDM sonar estimation application, the range resolution is a function of the transmit signal bandwidth. The performance of the OFDM system highly depends on the interference level in the system. To ensure the orthogonality between sub - carriers, the calculation of the p - th range cell R p in formula (6) is approximately an integer multiple of the range resolution, that is, R p ≈ pρ y , where p is an integer. Ignoring the residual error. To suppress the interference caused by the residual term, measure the norm of the estimation error The estimation accuracy is improved while minimizing the interference energy. Based on the OFDM sonar model, the proposed double iterative method is used to obtain a y =c / (2B) with high resolution distance estimation and the advantage of target interference suppression.

[0170] Third, in the phase error compensation process, the proposed method does not require a specific form of phase error. In other words, the method is applicable to different types of phase errors, which is conducive to improving the adaptability of the method in complex and variable marine environments and system errors. Unlike existing detection methods, the proposed method maintains the advantages of CP-based OFDM while compensating for different types of phase errors.

[0171] The technical solution of the present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0172] Example 1

[0173] Embodiment 1 of the present invention proposes a sonar distance-angle joint estimation method with fusion error correction. Using simulation experiments, we use computer simulation to evaluate the performance of the proposed phase error compensation and distance-angle estimation method. Detailed sonar parameters are shown in Table 3. The array element located at the origin of the coordinate transmits an OFDM signal. The receiving array is a uniform linear array of 25 elements along the x-axis. Assume that the sonar monitoring area is divided into a distance-angle plane of 256×90 pixels. The modulation symbols are generated by PSK from binary data. The parameters of the optimization problem are adjusted to δ=0.01.

[0174] Table 3 OFDM sonar parameters

[0175] parameter value Carrier frequency 50kHz bandwidth 1.5kHz Speed ​​of sound 1500m / s Number of subcarriers 256 Sampling speed 0.667ms Distance resolution 0.5m Number of array elements 25 Distance between adjacent components 0.015m

[0176] First, in order to verify the sidelobe suppression performance of the proposed method, the normalized range profile and its partial magnified view are shown, e.g. Figure 2As shown. The proposed method using OFDM waveform is compared with the traditional pulse compression method using LFM waveform. At the same angle, there are two targets located at R = 50m and R = 52m respectively. The length of CP is set to 255, which is sufficient here. For the proposed method using OFDM waveform, the sidelobe of the range profile is lower than -250dB, which is much lower than the sidelobe of the traditional method using LFM waveform. To further demonstrate the sidelobe suppression performance, Figure 3 shows the normalized range profile using different lengths of CP. OFDM signals are added with different lengths of CP, 100 and 0 respectively. As shown in Figure 3(a) and Figure 3(b). The rest of the sonar parameters are the same as the above simulation. When the CP length is 100, the estimated range of the target matches the true range, and the sidelobe is lower than -32dB, while the sidelobe of the method using LFM waveform is -10dB. When the CP length becomes zero, the sidelobe level increases significantly and becomes -10dB due to interference between targets. However, the sidelobe near the target is still lower than that of the method using LFM waveform. Therefore, if a CP of sufficient length is inserted, the proposed method has the characteristics of high accuracy and low sidelobe. As shown in Figure 4, the normalized distance and angle estimation profiles using different methods and waveforms, where Figure 4 (a) is the normalized distance estimation profile, and Figure 4 (b) is the normalized angle estimation profile; Figure 5 is the behavior of phase compensation, where Figure 5 (a) is the normalized angle estimation profile, and Figure 5 (b) is the phase error estimation result; Figure 6 The figure shows the comparison of the accuracy of distance and angle estimation. The blue line shows the relationship between the root mean square error (RMSE) and signal-to-noise ratio (SNR) of distance estimation. The red line shows the relationship between the root mean square error (RMSE) and signal-to-noise ratio (SNR) of angle estimation.

[0177] Example 2

[0178] Embodiment 2 of the present invention provides a sonar distance-angle joint estimation device with fusion error correction, which is implemented based on the method of embodiment 1 and includes:

[0179] A preprocessing module, used for preprocessing the received OFDM sonar signal to remove the cyclic prefix and tail samples of the received signal;

[0180] A calculation module, used for performing fast Fourier transform on the preprocessed received signal and calculating the shifted modulation symbol sequence;

[0181] A coarse estimation module is used to obtain a coarse estimate of the distance and angle of the target based on the shifted modulation symbol sequence, using the multi-target interference suppression characteristics and beamforming technology;

[0182] The optimization iteration module is used to construct the optimization problem in matrix form. Through the iterative algorithm, the sparse signal and phase error matrix are gradually updated until the predetermined convergence standard is reached to obtain the distance and angle estimation results of the target.

[0183] Example 3

[0184] Embodiment 3 of the present invention provides a computer device, comprising: at least one processor, a memory, at least one network interface and a user interface. The various components in the device are coupled together through a bus system. It can be understood that the bus system is used to realize the connection and communication between these components. In addition to the data bus, the bus system also includes a power bus, a control bus and a status signal bus.

[0185] The user interface may include a display, a keyboard, or a pointing device (eg, a mouse, a trackball, a touch pad, or a touch screen).

[0186] It is understood that the memory in the disclosed embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus random access memory (DRRAM). The memories described herein are intended to include, but are not limited to, these and any other suitable types of memories.

[0187] In some embodiments, the memory stores the following elements, executable modules or data structures, or a subset thereof, or an extended set thereof: an operating system and applications.

[0188] The operating system includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., which are used to implement various basic services and process hardware-based tasks. The application includes various application programs, such as a media player (Media Player), a browser (Browser), etc., which are used to implement various application services. The program for implementing the method of the embodiment of the present disclosure can be included in the application.

[0189] In the above embodiment, the processor may be used to execute the steps of the method of embodiment 1 by calling a program or instruction stored in the memory, specifically, a program or instruction stored in an application.

[0190] The method of embodiment 1 can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor may be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field programmable gate array (Field Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in embodiment 1 can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with embodiment 1 can be directly embodied as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in a decoding processor. The software module may be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in a memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0191] It is understood that the embodiments described in the present invention can be implemented by hardware, software, firmware, middleware, microcode or a combination thereof. For hardware implementation, the processing unit can be implemented in one or more application specific integrated circuits (ASIC), digital signal processors (DSP), digital signal processing devices (DSPD), programmable logic devices (PLD), field programmable gate arrays (FPGA), general processors, controllers, microcontrollers, microprocessors, other electronic units for performing the functions described in the present application or a combination thereof.

[0192] For software implementation, the technology of the present invention can be implemented by executing the functional modules (such as procedures, functions, etc.) of the present invention. The software code can be stored in a memory and executed by a processor. The memory can be implemented in the processor or outside the processor.

[0193] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention is described in detail with reference to the embodiments, it should be understood by those skilled in the art that any modification or equivalent replacement of the technical solutions of the present invention does not depart from the spirit and scope of the technical solutions of the present invention and should be included in the scope of the claims of the present invention.

Claims

1. A sonar range-angle joint estimation method with fusion error correction, comprising: Preprocessing the received OFDM sonar signal to remove the cyclic prefix and tail samples of the received signal; Performing fast Fourier transform on the preprocessed received signal to calculate the shifted modulation symbol sequence; According to the shifted modulation symbol sequence, the multi-target interference suppression characteristics and beamforming technology are used to obtain a rough estimate of the target's distance and angle; Construct an optimization problem in matrix form, and gradually update the sparse signal and phase error matrix through an iterative algorithm until the predetermined convergence standard is reached to obtain the target distance and angle estimation results; A rough estimate of the distance and angle of the target for: Where A(n,q) is the element of matrix A, n is the number of n-th subcarriers, q is the q-th angle grid, ρ R is the distance resolution, is a rough estimate of the angle, is a rough estimate of the time delay, and then a rough estimate of the distance can be obtained, Q is the number of angle grids; The matrix form is: the matrix form of performing N-point Fourier transform on the m-th array element satisfies the following formula: Y=WAVE+N Among them, matrix Y is the measurement data, matrix A and matrix N correspond to the target coefficient and additive white Gaussian noise respectively, matrix V represents the steering vector, matrix E represents the phase error terms of different array elements, and matrix W represents the modulation symbols after displacement; The optimization problem in matrix form is: Among them, J(a,E) is the objective function, y=vec(Y), a=vec(A), Z E is the dictionary matrix, ρ represents the regularization parameter, are the estimated values ​​of the coefficient vector and phase error term, respectively; The method of gradually updating the sparse signal and the phase error matrix through an iterative algorithm until a predetermined convergence standard is reached to obtain the distance and angle estimation results of the target includes: Step s1) l = 0, initialize the corresponding coefficient vector and the estimated value of the phase error term respectively and Step s2) Calculate the Hessian matrix Update by solving the following Update the angle estimate by Among them, ∠{·} represents the complex angle, Ξ m is an M×M square matrix, except that the mth diagonal element is equal to e m All elements are zero; (.) H is the conjugate of the matrix; is the Kronecker product; Step s3) Determination If the judgment result is true, execute l+1 and go to step s2); otherwise, go to step s4); Step s4) Output the distance estimate of the target and angle estimation 2. The sonar distance-angle joint estimation method with fusion error correction according to claim 1 is characterized in that: The OFDM sonar includes M arrays of equally spaced linear arrays, and the spacing between adjacent arrays is d=λ / 2, where λ is the signal wavelength. The transmitted OFDM signal s(t) is: Where N is the number of subcarriers, k is the subcarrier number, Δf is the subcarrier spacing, Δf = B / N, B is the bandwidth, w k is a modulation symbol sequence obtained by binary data and orthogonal amplitude modulation, j represents the imaginary part, and t represents the time domain.

3. The sonar distance-angle joint estimation method with fusion error correction according to claim 2 is characterized in that: The received signal g of the mth array in the OFDM sonar m (t) is: Among them, a(R,θ) is the reflection function of the target, R is the discrete distance set, θ is the angle, exp[je m (θ)] represents the phase error and c is the speed of light.

4. The sonar distance-angle joint estimation method with fusion error correction according to claim 2 is characterized in that: The shifted modulation symbol sequence w CP for: Among them, N CP is the circular shift amount.

5. A device for the sonar distance-angle joint estimation method based on the fusion error correction of claim 1, characterized in that: include: A preprocessing module, used for preprocessing the received OFDM sonar signal to remove the cyclic prefix and tail samples of the received signal; A calculation module, used for performing fast Fourier transform on the preprocessed received signal and calculating the shifted modulation symbol sequence; A coarse estimation module is used to obtain a coarse estimate of the distance and angle of the target based on the shifted modulation symbol sequence, using the multi-target interference suppression characteristics and beamforming technology; and The optimization iteration module is used to construct the optimization problem in matrix form. Through the iterative algorithm, the sparse signal and phase error matrix are gradually updated until the predetermined convergence standard is reached to obtain the distance and angle estimation results of the target.

6. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method of claim 1.

Images