Method and system for AIS signal separation of four circular array antennas

Abstract

This invention discloses a method and system for separating AIS aliasing signals using a four-circular array antenna. The steps are as follows: S100: AIS signals are received through multiple antennas to obtain the spatial baseband signal matrix X, and the phase error Δφ is calculated using interval calculation. k S200: Calculate the phase compensation matrix p; S200: Compensate the baseband signal according to the phase compensation matrix to obtain the baseband signal X. comp S300: For baseband signal X comp By expanding the spatial and temporal dimensions, we obtain the spatiotemporal snapshot matrix X. st (t); S400: The covariance matrix R is obtained by estimating the spatiotemporal snapshot matrix. xx S500: For the covariance matrix R xx Eigenvalue decomposition is performed to obtain the signal subspace and noise subspace; S600: DOA estimation is performed using the MUSIC algorithm to construct the spatial spectrum P(θ); S700: The beam w at the optimal solution is obtained using the space-time constrained minimum variance algorithm; S800: The separated signal is output through the sparse separation method of protocol features. This method can significantly improve the availability of AIS self-organizing network and provide more reliable protection for ship navigation safety.

Description

Technical Field

[0001] This invention relates to the field of AIS aliasing signal separation, and particularly to a method and system for separating AIS aliasing signals using a four-circular array antenna. Background Technology

[0002] The Automatic Identification System (AIS) was developed in the 1990s and belongs to the second generation of maritime communication technology. As early as 2003, the International Maritime Organization had already mandated the promotion of AIS equipment. After more than ten years of development, AIS has become widely used.

[0003] The AIS network uses a decentralized self-organizing network (SOTDMA), and its physical layer modem technical specifications are as follows:

[0004] (1) Operating frequency:

[0005] (a) AIS1: Center frequency 161.975MHz, bandwidth 25kHz,

[0006] (b)AIS2: Center frequency 162.025MHz, bandwidth 25kHz;

[0007] (2) Frequency accuracy: better than ±500Hz;

[0008] (3) The dual-channel independent receiver can receive simultaneously on two separate channels; one transmitter serves as a backup for TDMA transmission on the two separate channels;

[0009] (4) Data code type: NRZI;

[0010] (5) Modulation method: GMSK-FM, modulation index 0.5, transmit BT product 0.4, receive BT product 0.5;

[0011] (6) Transmission rate: 9.6kbps±10ppm;

[0012] The expanded application of AIS technology has placed a huge load on AIS channels, resulting in high data link loads in some busy areas. When two or more AIS terminal devices transmit signals with the same center frequency in the same time slot, these AIS signals alias in time and frequency, and traditional AIS receiving equipment cannot decode and receive them properly, leading to AIS message loss. Severe AIS time slot conflicts may cause accidents such as ship collisions.

[0013] Traditional blind source separation algorithms can separate aliased signals in the case of a single antenna. However, for blind source separation algorithms to be successfully implemented, the aliased signals must meet the following characteristics:

[0014] (1) Independent Component Analysis (ICA): requires that the aliased signals are statistically contradictory and non-Gaussian distributed;

[0015] (2) Sparse Component Analysis (SCA): requires that the aliased signal has sparsity in a certain transform domain (such as the time-frequency domain);

[0016] (3) Non-negative matrix decomposition (NMF): requires that the aliased signal can be represented as a non-negative linear combination of basis vectors.

[0017] However, the AIS signal uses GMSK modulation, and this signal has the following characteristics:

[0018] (1) Independent Component Analysis (ICA):

[0019] (a) All AIS signals use the same GMSK modulation (constant envelope, continuous phase);

[0020] (b) The time-slot conflict signals completely overlap in the time / frequency domain;

[0021] (c) Signal independence is compromised (same modulation, same frequency, same power domain);

[0022] (d) Therefore, the ICA method cannot separate independent components from a single-channel aliased AIS signal (observation dimension < source dimension).

[0023] (2) Sparse component analysis (SCA):

[0024] (a) The GMSK signal energy is uniformly distributed in the time domain (without burst pulses);

[0025] (b) The frequency domain bandwidth is only 25kHz (adjacent signal spectra completely overlap);

[0026] (c) Lacks significant sparsity features;

[0027] (d) Therefore, the SCA method's separation performance for aliased AIS signals is close to random guessing.

[0028] (3) Nonnegative matrix factorization (NMF):

[0029] (a) The radio frequency signal contains positive and negative oscillations (does not satisfy the non-negative constraint);

[0030] (b) No prior basis vector library (AIS signals have no fixed waveform template);

[0031] (c) Therefore, the decomposition of aliased AIS signals is meaningless.

[0032] In summary, traditional single-antenna blind source separation methods are unsuitable for aliased AIS signals. Therefore, multi-antenna separation techniques are the only viable solution. Summary of the Invention

[0033] The technical problem solved by this invention is to provide a method for separating AIS aliased signals using a four-circular array antenna, which can separate AIS signals that are mixed in time and frequency, thereby greatly improving the availability of AIS self-organizing networks.

[0034] The technical solution adopted by this invention to solve its technical problem is: a method for separating AIS aliasing signals using a four-circular array antenna, the steps of which are:

[0035] S100: Receives AIS signals via multiple antennas to obtain the space baseband signal matrix X. Uses the IMU (Inertial Measurement Unit) to acquire the current UAV's angular velocity ω, acceleration a, linear velocity ν, and position vector r. Calculates the phase error Δφ based on the interval used. k And calculate the phase compensation matrix p;

[0036] S200: The baseband signal is compensated and output according to the phase compensation matrix to obtain the compensated baseband signal X. comp ;

[0037] S300: By analyzing the baseband signal X comp By applying tap delays to expand the spatial and temporal dimensions, the spatiotemporal snapshot matrix X is obtained. st (t), where the i-th column of the space-time matrix is

[0038] S400: Obtain the covariance matrix R using spatiotemporal snapshot matrix estimation. xx ;

[0039] S500: For the covariance matrix R xx Perform feature decomposition to obtain the signal subspace and the noise subspace;

[0040] S600: The MUSIC algorithm is used to estimate DOA and construct the spatial spectrum P(θ);

[0041] S700: The beam w when the optimal solution is obtained using the space-time constrained minimum variance algorithm;

[0042] S800: Outputs the separated signal through a sparse separation method based on protocol features.

[0043] Furthermore, in step S100, the phase error Δφ k The calculation method is as follows:

[0044]

[0045] Where Δt is the sampling interval;

[0046] Next, the phase compensation matrix p;

[0047] P=diag[exp(-j·Δφ1), exp(-j·Δφ2), exp(-j·Δφ3), exp(-j·Δφ4)];

[0048] In step S200, the compensated baseband signal X is obtained. comp X comp =P⊙X.

[0049] Furthermore, in step S300, specifically:

[0050] For each baseband signal X comp The array element signals are spatiotemporally expanded to construct a P×1 vector:

[0051]

[0052] Where τ is the delay time;

[0053] Stack the P-dimensional vectors of the four array elements in order to form an N*P-dimensional column vector:

[0054]

[0055] For each time t, a 12-dimensional vector is constructed. Over the entire observation time Mτ, a 12×M matrix is ​​obtained, called the space-time snapshot matrix, denoted as X. st (t).

[0056]

[0057] Wherein, the i-th column of the space-time matrix is

[0058] Furthermore, in step S400, the covariance matrix R is obtained by estimating the spatiotemporal snapshot matrix. xx ;

[0059]

[0060] in, is the i-th column of the space-time matrix, and M is the number of samples.

[0061] Furthermore, in step S500, the covariance matrix R... xx Perform eigenvalue decomposition:

[0062] R xx =UΛU H ;

[0063] Where Λ is a diagonal matrix with eigenvalues ​​arranged in descending order on the main diagonal elements, and U is a unitary matrix composed of the eigenvectors corresponding to the eigenvalues.

[0064] Secondly, subspace extraction is performed, that is:

[0065] U s =[u1,u2,…,u K ];

[0066] K represents the number of signal sources;

[0067] Noise subspace extraction:

[0068] U n =[u K+1 ,u K+2 ,…,u N ].

[0069] Furthermore, in step S600, the spatial spectrum P(θ) is constructed;

[0070]

[0071] Where a(θ) is the extended guiding vector.

[0072] Furthermore, in step S600, the beam w at the optimal solution is obtained using the space-time constrained minimum variance algorithm, specifically as follows:

[0073] The space-time constrained minimum variance algorithm is adopted, with the following constraints:

[0074] ;

[0075] stC H w = f;

[0076] Where C is the constraint matrix, constraining the target direction + interference null. f is the response vector, with the value

[10] indicating that the main lobe gain is 1 and the interference gain is 0;

[0077] The closed-form solution can be obtained using the Lagrange multiplier method, and the beam w at the optimal solution is:

[0078]

[0079] Furthermore, in the S800, the separation signal is output through a sparse separation method based on protocol features, specifically as follows:

[0080] S801: Constructing an AIS signal time-frequency dictionary:

[0081]

[0082] S802: Solving sparse optimization problems:

[0083]

[0084] in, This is a function to indicate CRC check failure.

[0085] S802: Final output separation signal:

[0086]

[0087] The present invention also discloses a four-circular array antenna AIS aliasing signal separation system, including a motion sensing unit, a radio frequency unit and a signal separation unit;

[0088] The motion sensing unit is used to fuse IMU or BeiDou data and generate a bit compensation matrix.

[0089] The radio frequency unit is used for signal ADC sampling;

[0090] The signal separation unit is used to output the final separated signal.

[0091] The beneficial effects of this invention are:

[0092] 1. By introducing a spatial (4 elements) × temporal (3 taps) dimension through multiple antennas, this patent is able to handle 12-dimensional space.

[0093] 2. By introducing motion compensation and joint modeling of Doppler frequency shift, the method mentioned in this patent can tolerate angular velocities of 200° / s.

[0094] 3. By introducing CRC feedback from the AIS protocol and frame header matching operations, the algorithm mentioned in this patent improves the success rate of weak signal separation by 2.5 times.

[0095] 4. Through hardware acceleration via FPGA and algorithm optimization and simplification, real-time processing capability was achieved, significantly improving the efficiency of AIS signal separation. Attached Figure Description

[0096] Figure 1 This is a block diagram of the AIS separation system architecture for a four-circular array antenna hybrid system.

[0097] Figure 2 This is a flowchart of the space-time joint separation process for aliased AIS signals. Detailed Implementation

[0098] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0099] like Figure 1 and Figure 2As shown, the embodiments of this application disclose a method for separating AIS aliasing signals using a four-circular array antenna. S100: AIS signals are received through multiple antennas to obtain a spatial baseband signal matrix X. The angular velocity ω, acceleration a, linear velocity ν, and position vector r of the current UAV are obtained through an IMU (Inertial Measurement Unit). The phase error Δφ is calculated based on the interval. k And calculate the phase compensation matrix p;

[0100] S200: The baseband signal is compensated and output according to the phase compensation matrix to obtain the compensated baseband signal X. comp ;

[0101] S300: By analyzing the baseband signal X comp By applying tap delays to expand the spatial and temporal dimensions, the spatiotemporal snapshot matrix X is obtained. st (t), where the i-th column of the space-time matrix is

[0102] S400: Obtain the covariance matrix R using spatiotemporal snapshot matrix estimation. xx ;

[0103] S500: For the covariance matrix R xx Perform feature decomposition to obtain the signal subspace and the noise subspace;

[0104] S600: The MUSIC algorithm is used to estimate DOA and construct the spatial spectrum P(θ);

[0105] S700: The beam w when the optimal solution is obtained using the space-time constrained minimum variance algorithm;

[0106] S800: Outputs the separated signal through a sparse separation method based on protocol features.

[0107] The above method can efficiently separate aliased AIS signals received by a four-circular array antenna. The method first receives AIS signals through multiple antennas and combines this with the motion state of the UAV obtained by the IMU (Inertial Measurement Unit) to calculate and compensate for the phase error, obtaining the compensated baseband signal. Next, by applying tap delay processing to the baseband signal, spatial and temporal dimensions are expanded to construct a space-time snapshot matrix. Then, the covariance matrix is ​​estimated using this matrix, and eigenvalue decomposition is performed to obtain the signal subspace and noise subspace. Based on this, the MUSIC algorithm is used for DOA estimation to construct the spatial spectrum, and then the space-time constrained minimum variance algorithm is used to obtain the optimal beam. Finally, the separated signal is output through a sparse separation method of protocol features. The entire processing flow is compact and efficient, significantly improving the availability of AIS ad hoc networks and providing more reliable protection for ship navigation safety.

[0108] In this embodiment, in step S100, the phase error Δφ k The calculation method is as follows:

[0109]

[0110] Where Δt is the sampling interval;

[0111] Next, the phase compensation matrix p;

[0112] P=diag[exp(-j·Δφ1), exp(-j·Δφ2), exp(-j·Δφ3), exp(-j·Δφ4)];

[0113] In step S200, the compensated baseband signal X is obtained. comp X comp =P⊙X.

[0114] The settings of steps S100 and S200 above ensure that the AIS signal received by the UAV in high-speed motion can be accurately compensated, thereby effectively reducing the impact of phase errors caused by motion on signal separation performance. By performing phase compensation on the baseband signal, the accuracy and reliability of subsequent signal processing can be significantly improved.

[0115] In this embodiment, step S300 specifically includes:

[0116] Suppose there are N array elements (e.g., 4), and the baseband signal received by each element, after motion compensation, is in complex form, denoted as:

[0117] (Dimension: 4×P, where P is the number of time sampling points);

[0118] For the signal of each array element, not only the current sample is used, but also samples from multiple past times (time delay taps) are used. Let the number of time delay taps be P (e.g., P = 3), then for each array element, a P-dimensional time vector is constructed.

[0119] For the signal of the nth element (n = 1, 2, 3, 4) Construct a P×1 vector:

[0120]

[0121] Where τ is the delay time (usually taken as one symbol period or sampling period; in AIS, the symbol period is approximately 1 / 9600 seconds ≈ 104 μs, but in practice it may be adjusted according to the Doppler spread).

[0122] Stack the P-dimensional vectors of the four array elements sequentially to form an N*P-dimensional column vector, where N=4 and P=3:

[0123]

[0124]

[0125] For each time t, a 12-dimensional vector is constructed. Over the entire observation time Mτ, a 12×M matrix is ​​obtained, called the space-time snapshot matrix, denoted as X. st (t).

[0126]

[0127] Wherein, the i-th column of the space-time matrix is

[0128] Specifically, the settings in step S300 above can increase the system's degrees of freedom to counteract the Doppler frequency shift introduced by the relative motion of the transceiver system. The spatial dimension (array elements) provides the ability to distinguish signals from different directions, while the temporal dimension (time delay taps) provides the ability to distinguish signals with different Doppler frequency shifts. The extended space-time signal can be viewed as a virtual array (its number of array elements is the original number of array elements multiplied by the number of time delay taps). Therefore, the system's degrees of freedom increase from N to N*P, thereby enabling the simultaneous suppression of more interference and the separation of more signals.

[0129] In this embodiment, in step S400, the covariance matrix R is obtained by estimating the spatiotemporal snapshot matrix. xx ;

[0130]

[0131] in, is the i-th column of the space-time matrix, and M is the number of samples.

[0132] Specifically, in S500, for the covariance matrix R xx Perform eigenvalue decomposition:

[0133] R xx =UΛU H ;

[0134] Where Λ is a diagonal matrix with eigenvalues ​​arranged in descending order on the main diagonal elements, and U is a unitary matrix composed of the eigenvectors corresponding to the eigenvalues.

[0135] Secondly, subspace extraction is performed, that is:

[0136] U s =[u1,u2,…,u K ];

[0137] K represents the number of signal sources;

[0138] Noise subspace extraction:

[0139] U n =[u K+1 ,u K+2 ,…,u N ].

[0140] It should be explained that the covariance matrix is ​​derived from the signal subspace U. s and noise subspace U n It consists of two subspaces that are orthogonal to each other.

[0141] In the steps described above, the signal subspace and noise subspace obtained through eigenvalue decomposition provide the foundation for subsequent signal separation. The signal subspace contains eigenvectors related to the signal source, reflecting the spatial characteristics of the signal. The noise subspace, on the other hand, contains eigenvectors related to noise; these vectors are orthogonal to the signal subspace, providing an effective basis for signal separation.

[0142] In this embodiment, in step S600, the spatial spectrum P(θ) is constructed;

[0143]

[0144] Here, a(θ) is the extended steering vector, which is obtained by finding the maximum value of P(θ). The steering vector can be obtained by the spectral peak search algorithm.

[0145] Next, the beam w at the optimal solution is obtained using the space-time constrained minimum variance algorithm, specifically:

[0146] The space-time constrained minimum variance algorithm is adopted, with the following constraints:

[0147] ;

[0148] stC H w = f;

[0149] Where C is the constraint matrix, which constrains the target direction plus the interference null, that is, through linear constraints, the beam is forced to have a predetermined response (e.g., gain = 1) in a specific direction / frequency, and the response is 0 (null) in the interference direction.

[0150] f is the response vector, and the value

[10] indicates that the main lobe gain is 1 and the interference gain is 0;

[0151] The closed-form solution can be obtained using the Lagrange multiplier method, and the beam w at the optimal solution is:

[0152]

[0153] In the above steps, the optimal beam w was obtained through the space-time constrained minimum variance algorithm. This beam has the maximum gain in the target direction and forms nulls in the interference direction, thereby effectively suppressing the interference signal.

[0154] In this embodiment, in S800, the separation signal is output through a sparse separation method based on protocol features, specifically as follows:

[0155] S801: Constructing an AIS signal time-frequency dictionary:

[0156] D = [d1…d N ],

[0157] Where, d N GMSK represents the prior information of the Nth known sample in the AIS waveform, and the baseband waveform generated by modulation is GMSK. X represents the complex signal matrix received by multiple antennas, where each column is an observation sample.

[0158] S802: Solving sparse optimization problems:

[0159]

[0160] in, This is the CRC check failure indicator function, where α is a sparse matrix representing the sparse representation of each observation sample in the dictionary D.

[0161] The goal of this optimization step is to identify which dictionary atoms represent linear combinations of the signals received by each antenna.

[0162] S802: Final output separation signal:

[0163]

[0164] Among them, D k It is a selected atom dictionary from the dictionary, α k That is the corresponding sparsity coefficient. For the estimated Kth source signal, D k and α k All are obtained through step S802.

[0165] The above design steps ensure that the sparse representation method can effectively separate individual signals from aliased AIS signals. The constructed time-frequency dictionary D contains various possible waveforms of the AIS signal; these waveforms serve as dictionaries to represent the received signals. By solving the sparse optimization problem, a sparse matrix α is obtained, representing the sparse representation of each observation sample in dictionary D—that is, which linear combinations of dictionary atoms can represent each sample. This step is crucial for signal separation; it leverages the sparsity of the signal, reducing computational complexity while improving the accuracy of signal separation. Finally, the separated signals are obtained by multiplying the selected dictionary atoms by their corresponding sparse coefficients. This method is not only applicable to AIS signals received by a four-circular array antenna but can also provide a useful reference for signal separation problems in other similar scenarios.

[0166] The present invention also discloses a four-circular array antenna AIS aliasing signal separation system, including a motion sensing unit, a radio frequency unit and a signal separation unit;

[0167] The motion sensing unit is used to fuse IMU or BeiDou data and generate a bit compensation matrix.

[0168] The radio frequency unit is used for signal ADC sampling;

[0169] The signal separation unit is used to output the final separated signal.

[0170] Specifically, the motion sensing unit includes GPS, IMU, motors, etc., and can be used to output predicted attitude angles (φ, θ, Ψ). Simultaneously, it calculates the phase compensation matrix p;

[0171] The radio frequency unit consists of four independent radio frequency front-ends and a 4-channel ADC. It can perform independent ADC sampling on four channels and then output a baseband signal matrix X.

[0172]

[0173] Finally, signal separation is performed using a signal separation unit.

[0174] Through the aforementioned hardware design, efficient separation of AIS aliased signals from the four circular array antennas was achieved. The motion sensing unit integrates high-precision positioning data from IMU or BeiDou, enabling real-time acquisition of the UAV's motion status and accurate calculation of the phase compensation matrix, effectively reducing phase errors caused by UAV movement. The radio frequency unit employs four independent radio frequency front-ends and a four-channel ADC sampling system, ensuring the independence and accuracy of the four-channel signals and providing a high-quality data foundation for subsequent signal processing. The signal separation unit, based on the aforementioned non-negative matrix factorization method, outputs the separated AIS signal through a series of complex signal processing steps. The entire system is compact, functionally complete, and significantly improves the availability of AIS self-organizing networks and the safety of ship navigation.

[0175] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for separating AIS aliasing signals using a four-circular array antenna, characterized in that: S100: Receives AIS signals via multiple antennas to obtain the space baseband signal matrix X, and acquires the current angular velocity of the UAV via the IMU (Inertial Measurement Unit). acceleration linear velocity and position vector The phase error is obtained by calculating the interval. And calculate the phase compensation matrix p; S2 00: The baseband signal is compensated and output according to the phase compensation matrix to obtain the compensated baseband signal. ; S300: By analyzing baseband signals By applying tap delays, the spatial and temporal dimensions are expanded to obtain the spacetime snapshot matrix. The i-th column of the space-time matrix is ; S400: Obtain the covariance matrix using spatiotemporal snapshot matrix estimation. ; S500: For the covariance matrix Perform feature decomposition to obtain the signal subspace and the noise subspace; S600: DOA estimation is performed using the MUSIC algorithm to construct a spatial spectrum. ; S700: Beams that obtain the optimal solution using the space-time constrained minimum variance algorithm. ; S800: Outputs a separated signal through a sparse separation method based on protocol features; In step S100, the phase error The calculation method is as follows: ； in, The sampling interval; Next, the phase compensation matrix p; ； In step S200, the compensated baseband signal is obtained. , .

2. The method for separating AIS aliasing signals using a four-circular array antenna as described in claim 1, characterized in that: In step S300, specifically: For each baseband signal The array element signals are spatiotemporally expanded to construct a P×1 vector: ； in, It is a delay time; Stack the P-dimensional vectors of the four array elements in order to form an N*P-dimensional column vector: ； ； ； ； For each time t, a 12-dimensional vector is constructed; for the entire observation time... We can obtain a 12×M matrix, called the space-time snapshot matrix, denoted as . ; ； Wherein, the i-th column of the space-time matrix is .

3. The method for separating AIS aliasing signals using a four-circular array antenna as described in claim 2, characterized in that: In step S400, the covariance matrix is ​​obtained by estimating the spatiotemporal snapshot matrix. ; ； in, is the i-th column of the space-time matrix, and M is the number of samples.

4. The AIS signal separation method for a four-circular array antenna as described in claim 3, characterized in that: In step S500, the covariance matrix is... Perform eigenvalue decomposition: ； in, This is a diagonal matrix whose eigenvalues ​​are arranged in descending order on the main diagonal. The unitary matrix is ​​composed of the eigenvectors corresponding to the features; Secondly, subspace extraction is performed, that is: ； K represents the number of signal sources; Noise subspace extraction: 。 5. The method for separating AIS aliasing signals of a four-circular array antenna as described in claim 4, characterized in that: In step S600, the spatial spectrum is constructed. ; ； in, To extend the guide vector.

6. The method for separating AIS aliasing signals using a four-circular array antenna as described in claim 5, characterized in that: In step S700, the beam at the optimal solution is obtained using the space-time constrained minimum variance algorithm. Specifically: The space-time constrained minimum variance algorithm is adopted, with the following constraints: ； ； in, For the constraint matrix, The response vector takes the value [1 0], which indicates that the main lobe gain is 1 and the interference gain is 0. The closed-form solution can be obtained using the Lagrange multiplier method, and the beam at the optimal solution is obtained. for: 。 7. The method for separating AIS aliasing signals using a four-circular array antenna as described in claim 6, characterized in that: In the S800, the separation signal is output through a sparse separation method based on protocol features, specifically as follows: S801: Constructing an AIS signal time-frequency dictionary: ； S802: Solving sparse optimization problems: ； in, This is a function to indicate CRC check failure. S802: Final output separation signal: 。 8. A four-circular array antenna AIS aliasing signal separation system, used to execute the four-circular array antenna AIS aliasing signal separation method according to any one of claims 1 to 7, characterized in that: It includes a motion sensing unit, a radio frequency unit, and a signal separation unit; The motion sensing unit is used to fuse IMU or BeiDou data and generate a bit compensation matrix. The radio frequency unit is used for signal ADC sampling; The signal separation unit is used to output the final separated signal.

Images