5G-aware Doppler spread clutter suppression method based on joint dimensionality reduction processing in subcarrier domain and spatial domain

Abstract

This invention relates to a 5G sensing Doppler spread clutter suppression method based on joint subcarrier domain and spatial domain dimensionality reduction processing, belonging to the field of 5G sensing clutter suppression in urban environments. To address the Doppler spread clutter suppression problem under the sensing background of 5G integrated sensing systems, this invention includes: acquiring sensing echo signals using a 5G integrated sensing system to obtain three-dimensional data blocks; locating the Doppler spread clutter region; setting the size of the local processing region in the angle-distance joint domain; constructing a three-dimensional data block of the local processing region; rearranging the data of the three-dimensional data block of the local processing region to obtain a two-dimensional data matrix to be processed; constructing a training sample data matrix; constructing the target's subcarrier domain steering vector and spatial domain steering vector; then calculating the two-dimensional local steering vector after dimensionality reduction in the subcarrier and spatial domains to obtain the adaptive processing output result of the Doppler unit to be detected; and traversing all angle units to obtain the output result after joint dimensionality reduction and adaptive processing in the subcarrier and spatial domains.

Description

Technical Field

[0001] The present invention belongs to the field of 5G-aware clutter suppression in urban environments, and specifically relates to a 5G-aware Doppler diffusion clutter suppression method based on joint dimensionality reduction processing in subcarrier domain and spatial domain. Background Art

[0002] The 5G synaesthesia integrated system operates at 4.9GHz. By sharing hardware equipment and spectrum resources, the network can simultaneously realize communication and perception functions. The integrated system design not only reduces the size and power consumption of hardware equipment, but also can further improve spectrum efficiency. Its maximum perception range is between 1,400 meters and 2,000 meters. With the large-scale deployment of 5G antenna base stations, it can achieve effective coverage of urban areas. 5G base station perception also has the advantages of low cost, all-day, and all-weather. It is an effective means of monitoring low-altitude airspace in cities. It can realize the detection and tracking of low-altitude aircraft in cities, combat illegal flying and invading drones, and provide protection for the safety and privacy of sensitive areas. However, due to the presence of a large number of independent scatterers in the urban environment, such as roads, buildings, guardrails, trees, wind turbines, etc., strong static clutter is generated, and the influence of wind force causes static clutter at different distances and directions to diffuse in the Doppler dimension (dynamic clutter). This Doppler diffusion clutter almost occupies the entire Doppler bandwidth, causing targets with radial velocity to be submerged in the clutter, seriously affecting the detection of flying drones.

[0003] Space-time adaptive processing (STAP) technology is the most widely used and effective clutter suppression method in the field of airborne radar and high-frequency ground wave radar. Its development stems from the problem of airborne radar detecting slow-moving targets. The traditional airborne radar STAP method uses the joint processing of two-dimensional information in the spatial and temporal domains to adaptively form an oblique notch that matches the clutter ridge in the angle-Doppler domain, thereby filtering out ground clutter. This technology relies on the accurate estimation of the clutter noise covariance matrix (CCM) of the detection unit. According to the RMB criterion, at least 2 times the number of independent and identically distributed (IID) samples of the system degrees of freedom are required. In actual scenarios, due to the non-uniformity of clutter and the configuration of the antenna array, it is difficult for independent and identically distributed samples to meet the conditions. Therefore, using a small number of independent and identically distributed samples to accurately estimate the clutter covariance matrix is ​​a key problem facing the STAP technology. Since 5G sensing signals use OFDM modulation, which is different from traditional radars, how to apply STAP technology is the first problem to be faced. At the same time, the clutter characteristics are also different from the ground clutter in airborne radars. The full-dimensional STAP method lacks sufficient training samples, and the samples contain all non-uniform clutter information, which will cause the clutter suppression performance to be seriously degraded. Therefore, in the context of non-uniform clutter in 5G sensing, how to suppress Doppler spread clutter is a difficult problem. Summary of the invention

[0004] The problem to be solved by the present invention is the Doppler diffusion clutter suppression problem under the perception background of the 5G synaesthesia integrated system, and a 5G perception Doppler diffusion clutter suppression method based on joint dimensionality reduction processing in the subcarrier domain and the spatial domain is proposed.

[0005] To achieve the above object, the present invention is implemented through the following technical solutions:

[0006] A 5G-aware Doppler spread clutter suppression method based on subcarrier domain and spatial domain joint dimensionality reduction processing comprises the following steps:

[0007] S1. Use the 5G interawareness integrated system to collect and sense echo signals, and then perform distance processing, Doppler processing and azimuth processing to obtain a three-dimensional data block of distance-Doppler-angle;

[0008] S2. For the distance-Doppler-angle three-dimensional data block obtained in step S1, firstly select an angle unit to obtain the distance-Doppler spectrum RD spectrum, then calculate the signal power of each distance unit of the distance-Doppler spectrum RD spectrum, then estimate the noise power of each distance unit, perform threshold judgment, extract the distance unit where the Doppler diffusion clutter area of ​​the selected angle unit is located; traverse all angle units, extract the distance unit where the Doppler diffusion clutter area of ​​each angle unit is located, and complete the positioning of the Doppler diffusion clutter area;

[0009] S3. Select an angle unit and a distance unit where the Doppler diffusion clutter area of ​​the selected angle unit obtained in step S2 is located, set the size of the local processing area of ​​the angle-distance joint domain, construct a three-dimensional data block of the local processing area, rearrange the data of the three-dimensional data block of the local processing area, and obtain a two-dimensional data matrix to be processed;

[0010] S4. Select a Doppler unit from the Doppler units in the two-dimensional data matrix to be processed obtained in step S3 as the Doppler unit to be detected for processing, construct a training sample data matrix, and use the training sample data matrix to estimate the clutter covariance matrix of the Doppler unit to be detected;

[0011] S5. According to the angle unit and the corresponding distance unit selected in step S3, the subcarrier domain steering vector and the spatial domain steering vector of the target are constructed, and then the two-dimensional local steering vector after the subcarrier domain and the spatial domain dimensionality reduction is calculated;

[0012] S6. Using the clutter covariance matrix obtained in step S4 and the two-dimensional local steering vector after the subcarrier domain and spatial domain dimensionality reduction obtained in step S5, the adaptive weight vector of the Doppler unit to be detected set in step S4 is calculated to obtain the adaptive processing output result of the Doppler unit to be detected;

[0013] S7. Based on the method of step S3 to step S6, traverse all Doppler units of the selected angle unit to obtain the adaptive processing output result of the selected angle unit and the distance unit;

[0014] S8. Based on the method of step S3 to step S7, traverse all angle units and the distance units where the Doppler diffusion clutter area corresponding to each angle is located, and obtain the three-dimensional output result of distance-Doppler-angle after joint dimensionality reduction adaptive processing in subcarrier domain and spatial domain.

[0015] Furthermore, the specific implementation method of step S1 includes the following steps:

[0016] S1.1. Collect and sense echo signals using the 5G interawareness integrated system;

[0017] S1.2. Perform distance processing, Doppler processing and azimuth processing on the sensing echo signal collected in step S1.1:

[0018] The distance processing is performed by inverse fast Fourier transform IFFT, which transforms the subcarrier domain data into the distance domain, and sets the distance dimension of the processed data to N;

[0019] Doppler processing is implemented by fast Fourier transform FFT, which transforms OFDM symbol domain data into Doppler domain, and sets the Doppler dimension of the processed data to L;

[0020] Azimuth processing is achieved through digital beamforming, which transforms the array element domain data into the angle domain, and sets the angle dimension of the processed data to M;

[0021] S1.3. Set the distance-Doppler-angle three-dimensional data block obtained in step S1.2 to {data}, and the dimension of the three-dimensional data block is N×L×M;

[0022] Sets a data element x for a range-Doppler-angle 3D data block nlm ∈{data}, where the distance unit n∈{1,2,…,N-1,N}, the Doppler unit l∈{1,2,…,L-1,L}, the angle unit m∈{1,2,…,M-1,M}, and the angle corresponding to the angle system of the mth angle unit is θ m ,θ m ∈{θ1,θ2,…,θ M-1 ,θ M}.

[0023] Furthermore, the specific implementation method of step S2 includes the following steps:

[0024] S2.1. For the distance-Doppler-angle three-dimensional data block obtained in step S1, first select an angle unit to obtain the distance-Doppler spectrum RD spectrum;

[0025] Specifically, the angle unit m is set to 1, and the RD spectrum is obtained. The modulus value is squared to obtain |x nl1 | 2 ;

[0026] S2.2. Assume that the static clutter and the nearby Doppler units with higher energy account for at most η% of all Doppler units. Use this priori value to locate the Doppler area of ​​static clutter and side lobes, set the static clutter and the Doppler area with higher energy to zero, and reduce the influence of large energy units on the average noise floor estimation. The expression is:

[0027]

[0028] S2.3. Calculate the signal power A of distance unit n rn , the calculation formula is:

[0029]

[0030] Then calculate the set A of signal powers of all distance cells r ={A r1 ,A r2 ,…,A rN};

[0031] S2.4. Estimate the noise power of each distance unit and take A r The signal power of the smallest β1% to β2% distance unit is calculated to get the average noise floor A. Noise , where β1 is the lower limit of the noise floor estimation proportional coefficient, and β2 is the upper limit of the noise floor estimation proportional coefficient;

[0032] Set the threshold coefficient α, count the distance units where the signal power is greater than the threshold, and get the distance unit n where the Doppler diffusion clutter area is located. Clutter , a total of n C There is Doppler diffusion clutter in each range unit, and the expression is:

[0033] n Clutter ={n|A rn >αA Noise}={n1,L,n C};

[0034] S2.5. Traverse all angle units, obtain the distance unit where the corresponding Doppler diffusion clutter area is located, and complete the positioning of the Doppler diffusion clutter area.

[0035] Furthermore, the specific implementation method of step S3 includes the following steps:

[0036] S3.1. Select the angle unit m to be processed and a distance unit n in the distance unit where the Doppler diffusion clutter area corresponding to m is located;

[0037] S3.2. Set the local processing area size of the angle-distance joint domain to: the number of angle units m a and the number of distance units n r ; Then the angle unit range of the local processing area corresponding to the specified angle unit m is:

[0038] {m-(m a -1) / 2,…,m-1,m,m+1,…,m+(m a -1) / 2};

[0039] The local processing area distance unit range corresponding to the specified distance unit n is:

[0040] {n-(n r -1) / 2,…,n-1,n,n+1,…,n+(n r -1) / 2};

[0041] Then all Doppler unit data of the local processing area are obtained to form a three-dimensional data block {dataRDA} of the local processing area, whose dimension is n r ×L×m a ;

[0042] S3.3. Rearrange the local processing area three-dimensional data block {dataRDA} into an angle-range-Doppler data format with a dimension of m a ×n r ×L, and then vectorize the angle-distance two-dimensional data to get m a n r ×L two-dimensional data matrix {dataARD} to be processed.

[0043] Furthermore, the specific implementation method of step S4 includes the following steps:

[0044] S4.1. Select a Doppler unit l from the two-dimensional data matrix {dataARD} to be processed obtained in step S3 as the Doppler unit to be detected, and form a local processing area data vector X of the Doppler unit to be detected LPR , whose dimension is m a n r ×1;

[0045] S4.2. Process the Doppler unit to be detected and set the protection unit, the number is 2L g , the protection Doppler unit range is {lL g ,…,l-1,l+1,…,l+L g}, protect the Doppler unit from participating in the estimation of the clutter covariance matrix;

[0046] S4.3. Construct the training sample data matrix. According to the RMB criterion, the minimum number of training samples is L. r =2×m a ×n r , then select L from the two-dimensional data matrix {dataARD} r The data of Doppler units are used as the training sample data matrix. For the Doppler unit to be detected, the Doppler unit l of the training sample TS for:

[0047] l TS ∈{lL g -L r / 2,…,lL g -1,l+L g +1,l+L g +L r / 2}

[0048] If the endpoint of the Doppler unit interval of the training sample exceeds the range of all Doppler units, then the starting point of the Doppler interval of the training sample is cycled to the end point of all Doppler ranges, or the end point of the Doppler interval of the training sample is cycled to the starting point of all Doppler ranges, thereby forming a training sample data matrix {dataTS} with a dimension of m a n r ×L r ;

[0049] S4.4. Estimate the clutter covariance matrix of the Doppler unit to be detected using the data in the training sample data matrix The expression is:

[0050]

[0051] in,· H represents the conjugate transpose.

[0052] Furthermore, the specific implementation method of step S5 includes the following steps:

[0053] S5.1. Construct the subcarrier domain steering vector b(f) of the target according to the angle unit and the corresponding distance unit selected in step S3. subc,n ) and the spatial steering vector a(θ m ), the expression is:

[0054]

[0055] Among them, b(f subc,n ) is an N×1-dimensional column vector, f subc,nis the baseband subcarrier frequency corresponding to the selected distance unit, f s =B is the OFDM signal bandwidth, f subc,n / f s is the normalized subcarrier frequency;

[0056] a(θ m ) is a K×1 dimensional column vector, K is the number of array elements, d is the uniform linear array element spacing, λ n is the wavelength of the subcarrier corresponding to the selected distance unit, dsinθ m / λ n is the normalized spatial angular frequency;

[0057] S5.2. Calculate the target spatial-subcarrier domain two-dimensional steering vector s0, expressed as:

[0058]

[0059] In the formula represents the Kronecker direct product of two vectors, and the dimension of the target spatial-subcarrier domain two-dimensional steering vector is KN×1;

[0060] Construct a dimensionality reduction transformation matrix T for joint dimensionality reduction in the subcarrier domain and the spatial domain, whose dimension is KN×m a n r , the calculation formula is:

[0061]

[0062] Calculate the two-dimensional local steering vector s after dimensionality reduction in the subcarrier domain and spatial domain LPR The expression is:

[0063] s LPR =T Η s0.

[0064] Furthermore, the specific implementation method of step S6 includes the following steps:

[0065] S6.1. Using the clutter covariance matrix obtained in step S4 and the two-dimensional local steering vector after the subcarrier domain and spatial domain dimensionality reduction obtained in step S5, calculate the adaptive weight vector w of the Doppler unit to be detected set in step S4 opt , the expression is:

[0066]

[0067] Where μ is a constant, w opt The dimension is m a n r ×1;

[0068] S6.2. Calculate the adaptive processing output result y of the Doppler unit to be detected using the adaptive weight vector out (n, l, m), the output expression of the Doppler unit to be detected of the distance unit n, the Doppler unit l, and the angle unit m after adaptive filtering is:

[0069]

[0070] Among them, X LPR is the data vector of the local processing area of ​​the Doppler unit to be detected.

[0071] Furthermore, the specific implementation method of step S7 is to set l = 1, ..., L, and calculate the output y corresponding to each Doppler unit out (n,l,m), get the adaptive processing output result y of the specified angle unit m and distance unit n out (n,m).

[0072] Furthermore, the specific implementation method of step S8 includes the following steps:

[0073] S8.1. The distance unit n where Doppler spread clutter exists in the RD spectrum of the specified angle unit m obtained in step S2 C lutter={n1,L,n C}, let the distance unit n = n1,…,n C , calculate the adaptive processing output y of all Doppler units in each distance unit where the clutter is located out (m);

[0074] S8.2. Let angle unit m = 1, ..., M, calculate all clutter distance units n corresponding to each angle unit Clutter All Doppler units output y out .

[0075] Beneficial effects of the present invention:

[0076] The 5G sensing Doppler diffusion clutter suppression method based on joint dimensionality reduction processing in the subcarrier domain and the spatial domain described in the present invention can suppress the Doppler diffusion clutter in the 5G sensing signal echo, improve the signal-to-noise ratio of the target, and is beneficial to the detection and tracking of targets such as drones in urban low-altitude airspace. It has the characteristics of low computational burden, flexible processing, and adaptive change of weights. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 It is a flow chart of a 5G-aware Doppler spread clutter suppression method based on joint dimensionality reduction processing in subcarrier domain and spatial domain according to the present invention;

[0078] Figure 2 This is a schematic diagram of the result of step S1 of the present invention;

[0079] Figure 3 This is a schematic diagram of the result of step S8 of the present invention. DETAILED DESCRIPTION

[0080] In order to make the purpose, technical solution and advantages of the present invention more clear, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention, that is, the specific embodiments described are only part of the embodiments of the present invention, rather than all of the specific embodiments. The components of the specific embodiments of the present invention described and shown in the drawings herein can be arranged and designed in various different configurations, and the present invention can also have other embodiments.

[0081] Therefore, the following detailed description of the specific embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents the selected specific embodiments of the present invention. Based on the specific embodiments of the present invention, all other specific embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present invention.

[0082] In order to further understand the content, features and effects of the present invention, the following specific implementation methods are given as examples, and the attached Figure 1 -Attached Figure 3 The detailed instructions are as follows:

[0083] Embodiment 1:

[0084] A 5G-aware Doppler spread clutter suppression method based on subcarrier domain and spatial domain joint dimensionality reduction processing comprises the following steps:

[0085] S1. Use the 5G interawareness integrated system to collect and sense echo signals, and then perform distance processing, Doppler processing and azimuth processing to obtain a three-dimensional data block of distance-Doppler-angle;

[0086] Furthermore, the specific implementation method of step S1 includes the following steps:

[0087] S1.1. Collect and sense echo signals using the 5G interawareness integrated system;

[0088] S1.2. Perform distance processing, Doppler processing and azimuth processing on the sensing echo signal collected in step S1.1:

[0089] The distance processing is performed by inverse fast Fourier transform IFFT, which transforms the subcarrier domain data into the distance domain, and sets the distance dimension of the processed data to N;

[0090] Doppler processing is implemented by fast Fourier transform FFT, which transforms OFDM symbol domain data into Doppler domain, and sets the Doppler dimension of the processed data to L;

[0091] Azimuth processing is achieved through digital beamforming, which transforms the array element domain data into the angle domain, and sets the angle dimension of the processed data to M;

[0092] S1.3. Set the distance-Doppler-angle three-dimensional data block obtained in step S1.2 to {data}, and the dimension of the three-dimensional data block is N×L×M;

[0093] Sets a data element x for a range-Doppler-angle 3D data block nlm ∈{data}, where the distance unit n∈{1,2,…,N-1,N}, the Doppler unit l∈{1,2,…,L-1,L}, the angle unit m∈{1,2,…,M-1,M}, and the angle corresponding to the angle system of the mth angle unit is θ m ,θ m ∈{θ1,θ2,…,θ M-1 ,θ M};

[0094] S2. For the distance-Doppler-angle three-dimensional data block obtained in step S1, firstly select an angle unit to obtain the distance-Doppler spectrum RD spectrum, then calculate the signal power of each distance unit of the distance-Doppler spectrum RD spectrum, then estimate the noise power of each distance unit, perform threshold judgment, extract the distance unit where the Doppler diffusion clutter area of ​​the selected angle unit is located; traverse all angle units, extract the distance unit where the Doppler diffusion clutter area of ​​each angle unit is located, and complete the positioning of the Doppler diffusion clutter area;

[0095] Furthermore, the specific implementation method of step S2 includes the following steps:

[0096] S2.1. For the distance-Doppler-angle three-dimensional data block obtained in step S1, first select an angle unit to obtain the distance-Doppler spectrum RD spectrum;

[0097] Specifically, the angle unit m is set to 1, and the RD spectrum is obtained. The modulus value is squared to obtain |x nl1 | 2 ;

[0098] S2.2. Assume that the static clutter and the nearby Doppler units with higher energy account for at most η% of all Doppler units. Use this priori value to locate the Doppler area of ​​static clutter and side lobes, set the static clutter and the Doppler area with higher energy to zero, and reduce the influence of large energy units on the average noise floor estimation. The expression is:

[0099]

[0100] S2.3. Calculate the signal power A of distance unit n rn , the calculation formula is:

[0101]

[0102] Then calculate the set A of signal powers of all distance cells r ={A r1 ,A r2 ,…,A rN};

[0103] S2.4. Estimate the noise power of each distance unit and take A r The signal power of the smallest β1% to β2% distance unit is calculated to get the average noise floor A. Noise , where β1 is the lower limit of the noise floor estimation proportional coefficient, and β2 is the upper limit of the noise floor estimation proportional coefficient;

[0104] Set the threshold coefficient α, count the distance units where the signal power is greater than the threshold, and get the distance unit n where the Doppler diffusion clutter area is located. Clutter , a total of n C There is Doppler diffusion clutter in each range unit, and the expression is:

[0105] n Clutter ={n|A rn >αA Noise}={n1,L,n C};

[0106] S2.5. Traverse all angle units, obtain the distance unit where the corresponding Doppler spread clutter area is located, and complete the positioning of the Doppler spread clutter area;

[0107] S3. Select an angle unit and a distance unit where the Doppler diffusion clutter area of ​​the selected angle unit obtained in step S2 is located, set the size of the local processing area of ​​the angle-distance joint domain, construct a three-dimensional data block of the local processing area, rearrange the data of the three-dimensional data block of the local processing area, and obtain a two-dimensional data matrix to be processed;

[0108] Furthermore, the specific implementation method of step S3 includes the following steps:

[0109] S3.1. Select the angle unit m to be processed and a distance unit n in the distance unit where the Doppler diffusion clutter area corresponding to m is located;

[0110] S3.2. Set the local processing area size of the angle-distance joint domain to: the number of angle units m a and the number of distance units n r; Then the angle unit range of the local processing area corresponding to the specified angle unit m is:

[0111] {m-(m a -1) / 2,…,m-1,m,m+1,…,m+(m a -1) / 2};

[0112] The local processing area distance unit range corresponding to the specified distance unit n is:

[0113] {n-(n r -1) / 2,…,n-1,n,n+1,…,n+(n r -1) / 2};

[0114] Then all Doppler unit data of the local processing area are obtained to form a three-dimensional data block {dataRDA} of the local processing area, whose dimension is n r ×L×m a ;

[0115] S3.3. Rearrange the local processing area three-dimensional data block {dataRDA} into an angle-range-Doppler data format with a dimension of m a ×n r ×L, and then vectorize the angle-distance two-dimensional data to get m a n r ×L two-dimensional data matrix {dataARD} to be processed;

[0116] S4. Select a Doppler unit from the Doppler units in the two-dimensional data matrix to be processed obtained in step S3 as the Doppler unit to be detected for processing, construct a training sample data matrix, and use the training sample data matrix to estimate the clutter covariance matrix of the Doppler unit to be detected;

[0117] Furthermore, the specific implementation method of step S4 includes the following steps:

[0118] S4.1. Select a Doppler unit l from the two-dimensional data matrix {dataARD} to be processed obtained in step S3 as the Doppler unit to be detected, and form a local processing area data vector X of the Doppler unit to be detected LPR , whose dimension is m a n r ×1;

[0119] S4.2. Process the Doppler unit to be detected and set the protection unit, the number is 2L g , the protection Doppler unit range is {lL g ,…,l-1,l+1,…,l+L g}, protect the Doppler unit from participating in the estimation of the clutter covariance matrix;

[0120] S4.3. Construct the training sample data matrix. According to the RMB criterion, the minimum number of training samples is L. r =2×m a ×n r , then select L from the two-dimensional data matrix {dataARD} r The data of Doppler units are used as the training sample data matrix. For the Doppler unit to be detected, the Doppler unit l of the training sample TS for:

[0121] l TS ∈{lL g -L r / 2,…,lL g -1,l+L g +1,l+L g +L r / 2}

[0122] If the endpoint of the Doppler unit interval of the training sample exceeds the range of all Doppler units, then the starting point of the Doppler interval of the training sample is cycled to the end point of all Doppler ranges, or the end point of the Doppler interval of the training sample is cycled to the starting point of all Doppler ranges, thereby forming a training sample data matrix {dataTS} with a dimension of m a n r ×L r ;

[0123] S4.4. Estimate the clutter covariance matrix of the Doppler unit to be detected using the data in the training sample data matrix The expression is:

[0124]

[0125] Here, ·H represents the conjugate transpose.

[0126] S5. According to the angle unit and the corresponding distance unit selected in step S3, the subcarrier domain steering vector and the spatial domain steering vector of the target are constructed, and then the two-dimensional local steering vector after the subcarrier domain and the spatial domain dimensionality reduction is calculated;

[0127] Furthermore, the specific implementation method of step S5 includes the following steps:

[0128] S5.1. Construct the subcarrier domain steering vector b(f) of the target according to the angle unit and the corresponding distance unit selected in step S3. subc,n ) and the spatial steering vector a(θ m ), the expression is:

[0129]

[0130] Among them, b(f subc,n ) is an N×1-dimensional column vector, f subc,n is the baseband subcarrier frequency corresponding to the selected distance unit, f s =B is the OFDM signal bandwidth, f subc,n / f s is the normalized subcarrier frequency;

[0131] a(θ m ) is a K×1 dimensional column vector, K is the number of array elements, d is the uniform linear array element spacing, λ n is the wavelength of the subcarrier corresponding to the selected distance unit, dsinθ m / λ n is the normalized spatial angular frequency;

[0132] S5.2. Calculate the target spatial-subcarrier domain two-dimensional steering vector s0, expressed as:

[0133]

[0134] In the formula represents the Kronecker direct product of two vectors, and the dimension of the target spatial-subcarrier domain two-dimensional steering vector is KN×1;

[0135] Construct a dimensionality reduction transformation matrix T for joint dimensionality reduction in the subcarrier domain and the spatial domain, whose dimension is KN×m a n r , the calculation formula is:

[0136]

[0137] Calculate the two-dimensional local steering vector s after dimensionality reduction in the subcarrier domain and spatial domain LPR The expression is:

[0138] s LPR =T Η s0.

[0139] S6. Using the clutter covariance matrix obtained in step S4 and the two-dimensional local steering vector after the subcarrier domain and spatial domain dimensionality reduction obtained in step S5, the adaptive weight vector of the Doppler unit to be detected set in step S4 is calculated to obtain the adaptive processing output result of the Doppler unit to be detected;

[0140] Furthermore, the specific implementation method of step S6 includes the following steps:

[0141] S6.1. Using the clutter covariance matrix obtained in step S4 and the two-dimensional local steering vector after the subcarrier domain and spatial domain dimensionality reduction obtained in step S5, calculate the adaptive weight vector w of the Doppler unit to be detected set in step S4 opt, the expression is:

[0142]

[0143] Where μ is a constant, w opt The dimension is m a n r ×1;

[0144] S6.2. Calculate the adaptive processing output result y of the Doppler unit to be detected using the adaptive weight vector out (n,l,m), the output expression of the Doppler unit to be detected of the distance unit n, Doppler unit l, and angle unit m after adaptive filtering is:

[0145]

[0146] Among them, X LPR is the data vector of the local processing area of ​​the Doppler unit to be detected.

[0147] S7. Based on the method of step S3 to step S6, traverse all Doppler units of the selected angle unit to obtain the adaptive processing output result of the selected angle unit and the distance unit;

[0148] Furthermore, the specific implementation method of step S7 is to set l = 1, ..., L, and calculate the output y corresponding to each Doppler unit out (n,l,m), get the adaptive processing output result y of the specified angle unit m and distance unit n out (n,m).

[0149] S8. Based on the method of step S3 to step S7, traverse all angle units and the distance unit where the Doppler diffusion clutter area corresponding to each angle is located, and obtain the distance-Doppler-angle three-dimensional output result after the subcarrier domain and the spatial domain joint dimensionality reduction adaptive processing;

[0150] Furthermore, the specific implementation method of step S8 includes the following steps:

[0151] S8.1. The distance unit n where Doppler spread clutter exists in the RD spectrum of the specified angle unit m obtained in step S2 C lutter={n1,L,n C}, let the distance unit n = n1,…,n C , calculate the adaptive processing output y of all Doppler units in each distance unit where the clutter is located out (m);

[0152] S8.2. Let angle unit m = 1, ..., M, calculate all clutter distance units n corresponding to each angle unitClutter All Doppler units output y out .

[0153] Based on a 5G-aware Doppler spread clutter suppression method based on joint dimensionality reduction processing in the subcarrier domain and the spatial domain in this embodiment, the technical solution for conducting an experiment is as follows:

[0154] Step 1: Process the distance, Doppler and azimuth of the perception echo signal of the 5G synaesthesia integrated system to obtain a three-dimensional data block of distance-Doppler-angle. The specific steps are as follows:

[0155] The distance processing of the perception echo signal of the 5G synaesthesia integrated system is realized by IFFT, which transforms the subcarrier domain data into the distance domain, and the distance dimension of the processed data is 300; Doppler processing is realized by FFT, which transforms the OFDM symbol domain data into the Doppler domain, and the Doppler dimension of the processed data is 256; azimuth processing is realized by digital beamforming, which transforms the array element domain data into the angle domain, and the angle dimension of the processed data is 61. The obtained distance-Doppler-angle (RDA) three-dimensional data block is {data}, and its dimension is 300×256×61. Then for a data element x in the three-dimensional data block nlm ∈{data}, where the distance unit n∈{1,2,…,300}, the Doppler unit l∈{1,2,…,256}, the angle unit m∈{1,2,…,61}, and the angle corresponding to the angle system of the mth angle unit is θ m ,θ m ∈{-90°,-87°,…,0°,…,87°,90°}. Draw the distance-Doppler diagram (RD diagram) of the 35th angle unit, such as Figure 2 shown.

[0156] Step 2: Locate the Doppler diffusion clutter area. First, fix a certain angle unit, calculate the signal power of each distance unit of the RD spectrum, estimate the noise power level of each distance unit, make a threshold judgment, obtain the distance unit that needs to be suppressed, and then traverse all angle units to obtain the distance unit that needs to be suppressed. The specific steps are as follows:

[0157] (1) Input the range-Doppler-angle three-dimensional data block {data}, where the range unit n = 1, ..., 300, the Doppler unit l = 1, ..., 256, the angle unit m = 1, ..., 61, x nlm ∈{data};

[0158] (2) Fix the angle unit m = 1, obtain the RD spectrum, and square the modulus value to obtain |x nl1 | 2 ;

[0159] (3) Assuming that the static clutter and the nearby Doppler units with higher energy account for at most 10% of all Doppler units, use this priori value to locate the Doppler area of ​​the static clutter and side lobes, set the static clutter and the Doppler area with higher energy to zero, and reduce the impact of the large energy units on the average noise floor estimation:

[0160] x nl1 =0(l=116,117,…,141,142)

[0161] (4) Calculate the signal power of each distance unit:

[0162]

[0163] (5) Get the set A of signal powers of all distance units r ={A r1 ,A r2 ,…,A r300}, the noise floor power is calculated as follows: Take A r The average noise floor A is calculated by calculating the signal power of the 10% to 20% distance units with the smallest values. Noise ;

[0164] (6) Set the threshold coefficient to 4, count the distance cells whose signal power is greater than the threshold, and obtain the distance cells where the Doppler diffusion clutter area is located. A total of 73 distance cells have Doppler diffusion clutter, that is, n Clutter ={n|A rn >4A Noise}={n1,L,n 73};

[0165] (7) Traverse each angle unit and obtain the distance unit n where the corresponding Doppler diffusion clutter area is located Clutter .

[0166] Step 3: Select an angle unit and a corresponding distance unit obtained in step 2 that needs to be clutter suppressed, set the size of the local processing area in the angle-distance joint domain, obtain the data of all Doppler units, form a three-dimensional data block in the local processing area, and then rearrange the data to form a data matrix to be processed. The specific steps are as follows:

[0167] (1) Select the angle unit to be processed, such as the 35th angle unit (corresponding to angle θ 35 =12°), and then select the n obtained in step 2 corresponding to the angle unit Clutter A distance unit in , such as the 60th distance unit;

[0168] (2) Set the local processing area size of the angle-distance joint domain to: the number of angle units m aand the number of distance units n r The angle unit range of the local processing area corresponding to the 35th angle unit is {31, 32, 33, 34, 35, 36, 37, 38, 39}, and the distance unit range of the local processing area corresponding to the 60th distance unit is {59, 60, 61}. All Doppler unit data of the local processing area are obtained to form a three-dimensional data block {dataRDA} of the local processing area, whose dimension is 3×256×9;

[0169] (3) The three-dimensional data block {dataRDA} of the local processing area is rearranged into an angle-range-Doppler data format with a dimension of 9×3×256, and then the angle-range two-dimensional data is vectorized to form a 27×256 two-dimensional data matrix {dataARD}.

[0170] Step 4: Select a Doppler unit from all Doppler units for processing. This Doppler unit is called the Doppler unit to be detected. Construct a training sample data matrix and use the training sample data matrix to estimate the clutter covariance matrix of the Doppler unit to be detected. The specific steps are as follows:

[0171] (1) Select the 24th Doppler unit in the two-dimensional data matrix {dataARD} as the Doppler unit to be detected, and construct the local processing area data vector X of the Doppler unit to be detected. LPR , whose dimension is 27×1;

[0172] (2) A protection unit is set for the Doppler unit to be detected. The number is 4, so the range of the protection Doppler unit is {22, 23, 25, 26}. The protection Doppler unit does not participate in the estimation of the clutter covariance matrix;

[0173] (3) Construct the training sample data matrix. According to the RMB criterion, the minimum number of training samples is L r =2×9×3=54, then the data of 54 Doppler units are selected from the two-dimensional data matrix {dataARD} as the training sample data matrix. Then for the Doppler unit to be detected, the Doppler unit l of its training sample TS ∈{251,252,253,254,255,256,1,2,…,21,27,28,…,53}, thus forming the training sample data matrix {dataTS}, whose dimension is 27×54;

[0174] (4) Use the training samples to estimate the clutter covariance matrix of the Doppler unit l to be detected:

[0175]

[0176] In the formula H represents the conjugate transpose.

[0177] Step 5: Based on the distance unit and angle unit selected in step 3, construct the subcarrier domain steering vector, spatial domain steering vector and two-dimensional local steering vector after subcarrier domain and spatial domain dimensionality reduction of the target. The specific steps are as follows:

[0178] (1) The subcarrier domain steering vector and spatial domain steering vector of the target are:

[0179]

[0180] Where b(f subc,60 ) is a 324×1-dimensional column vector, f subc,60 is the baseband subcarrier frequency corresponding to the selected distance unit, f s =B is the OFDM signal bandwidth, f subc,60 / f s is the normalized subcarrier frequency; a(θ 35 ) is a 16×1 dimensional column vector, 16 is the number of array elements, d is the uniform linear array element spacing, λ n is the wavelength of the subcarrier corresponding to the selected distance unit, dsinθ 35 / λ n is the normalized spatial angular frequency;

[0181] (2) Calculate the target's spatial-subcarrier domain two-dimensional steering vector:

[0182]

[0183] In the formula Represents the Kronecker direct product of two vectors, and the dimension of the target spatial-subcarrier domain two-dimensional steering vector is 5184×1.

[0184] (3) Construct a dimensionality reduction transformation matrix for joint dimensionality reduction in the subcarrier domain and the spatial domain. Its dimension is 5184 × 27 and the calculation formula is:

[0185]

[0186] (4) Then the two-dimensional local steering vector after dimensionality reduction in the subcarrier domain and the spatial domain can be calculated by the following formula:

[0187] s LPR =T Η s0

[0188] Step 6: Calculate the adaptive weight vector of the Doppler unit to be detected based on the clutter covariance matrix obtained in step 4 and the two-dimensional local steering vector obtained in step 5 to obtain the adaptive processing output result of the Doppler unit to be detected. The specific steps are as follows:

[0189] (1) Using the linear constrained minimum variance criterion, the optimal weight vector is calculated:

[0190]

[0191] in is a constant, w opt The dimension is 27×1.

[0192] (2) The output of the adaptive filtering of the distance unit 60, the 24th Doppler unit, and the 35th angle unit is calculated using the optimal weight vector:

[0193]

[0194] Step 7: Traverse all Doppler units and obtain the output results of the specified angle unit and distance unit. The specific steps are as follows:

[0195] Let l = 1, ..., 256, and calculate the output y corresponding to each Doppler unit out (60, l, 35), and get the adaptive processing output result y of the 35th angle unit and the 60th distance unit out (60,35).

[0196] Step 8: Traverse all angle units and the distance units where the Doppler spread clutter corresponding to each angle is located, and obtain the range-Doppler-angle three-dimensional output result after the subcarrier domain and spatial domain joint dimensionality reduction adaptive processing. The specific steps are as follows:

[0197] (1) The distance unit n where Doppler diffusion clutter exists in the RD spectrum of the 35th angle unit obtained in step 2 C lutter={n1,L,n C}, let the distance unit n = n1,…,n C , calculate the adaptive processing output y of all Doppler units in each distance unit where the clutter is located out (m);

[0198] (2) Let the angle unit m = 1, ..., M, and calculate all the clutter distance units n corresponding to each angle unit Clutter All Doppler units output y out .

[0199] Attached Figure 2 This is the range-Doppler diagram (RD diagram) of the 35th angle unit (12° direction) before clutter suppression. The range unit range is 1 to 150, and the Doppler unit range is 1 to 256. It can be seen that there is strong Doppler diffusion clutter in the 59th to 68th and 118th to 133rd range units. This clutter occupies all Doppler bandwidths, and moving targets will be submerged in the Doppler diffusion clutter, resulting in a decrease in target detection performance. In the RD spectrum, the Doppler diffusion clutter shows serious non-uniformity.

[0200] Attached Figure 3 The RD diagram of the 35th angle unit after clutter suppression processing shows that the Doppler diffusion clutter of the 59th to 68th and 118th to 133rd distance units are suppressed, and the targets at the 60th distance unit and the 24th Doppler unit are retained, indicating that the method of the present invention can cope with the clutter suppression problem of Doppler diffusion and improve the target detection performance.

[0201] It should be noted that relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0202] Although the present application has been described above with reference to specific embodiments, various modifications may be made thereto and parts thereof may be replaced with equivalents without departing from the scope of the present application. In particular, as long as there is no structural conflict, the various features in the specific embodiments disclosed in the present application may be used in combination with each other in any manner, and the fact that these combinations are not exhaustively described in this specification is only for the sake of omitting space and saving resources. Therefore, the present application is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A 5G-aware Doppler spread clutter suppression method based on joint dimensionality reduction processing in subcarrier domain and spatial domain, characterized in that: The steps include: S1. Use the 5G interawareness integrated system to collect and sense echo signals, and then perform distance processing, Doppler processing and azimuth processing to obtain a three-dimensional data block of distance-Doppler-angle; S2. For the distance-Doppler-angle three-dimensional data block obtained in step S1, firstly select an angle unit to obtain the distance-Doppler spectrum RD spectrum, then calculate the signal power of each distance unit of the distance-Doppler spectrum RD spectrum, then estimate the noise power of each distance unit, perform threshold judgment, extract the distance unit where the Doppler diffusion clutter area of ​​the selected angle unit is located; traverse all angle units, extract the distance unit where the Doppler diffusion clutter area of ​​each angle unit is located, and complete the positioning of the Doppler diffusion clutter area; S3. Select an angle unit and a distance unit where the Doppler diffusion clutter area of ​​the selected angle unit obtained in step S2 is located, set the size of the local processing area of ​​the angle-distance joint domain, construct a three-dimensional data block of the local processing area, rearrange the data of the three-dimensional data block of the local processing area, and obtain a two-dimensional data matrix to be processed; S4. Select a Doppler unit from the Doppler units in the two-dimensional data matrix to be processed obtained in step S3 as the Doppler unit to be detected for processing, construct a training sample data matrix, and use the training sample data matrix to estimate the clutter covariance matrix of the Doppler unit to be detected; S5. According to the angle unit and the corresponding distance unit selected in step S3, the subcarrier domain steering vector and the spatial domain steering vector of the target are constructed, and then the two-dimensional local steering vector after the subcarrier domain and the spatial domain dimensionality reduction is calculated; S6. Using the clutter covariance matrix obtained in step S4 and the two-dimensional local steering vector after the subcarrier domain and spatial domain dimensionality reduction obtained in step S5, the adaptive weight vector of the Doppler unit to be detected set in step S4 is calculated to obtain the adaptive processing output result of the Doppler unit to be detected; S7. Based on the method of step S3 to step S6, traverse all Doppler units of the selected angle unit to obtain the adaptive processing output result of the selected angle unit and the distance unit; S8. Based on the method of step S3 to step S7, traverse all angle units and the distance units where the Doppler diffusion clutter area corresponding to each angle is located, and obtain the three-dimensional output result of distance-Doppler-angle after joint dimensionality reduction adaptive processing in subcarrier domain and spatial domain.

2. According to claim 1, a 5G-aware Doppler spread clutter suppression method based on subcarrier domain and spatial domain joint dimensionality reduction processing is characterized in that: The specific implementation method of step S1 includes the following steps: S1.

1. Collect and sense echo signals using the 5G interawareness integrated system; S1.

2. Perform distance processing, Doppler processing and azimuth processing on the sensing echo signal collected in step S1.1: The distance processing is performed by inverse fast Fourier transform IFFT, which transforms the subcarrier domain data into the distance domain, and sets the distance dimension of the processed data to N; Doppler processing is implemented by fast Fourier transform FFT, which transforms OFDM symbol domain data into Doppler domain, and sets the Doppler dimension of the processed data to L; Azimuth processing is achieved through digital beamforming, which transforms the array element domain data into the angle domain, and sets the angle dimension of the processed data to M; S1.

3. Set the distance-Doppler-angle three-dimensional data block obtained in step S1.2 to {data}, and the dimension of the three-dimensional data block is N×L×M; Sets a data element x for a range-Doppler-angle 3D data block nlm ∈{data}, where the distance unit n∈{1,2,…,N-1,N}, the Doppler unit l∈{1,2,…,L-1,L}, the angle unit m∈{1,2,…,M-1,M}, and the angle corresponding to the angle system of the mth angle unit is θ m ,θ m ∈{θ1,θ2,…,θ M-1 ,θ M }.

3. According to claim 2, a 5G-aware Doppler spread clutter suppression method based on subcarrier domain and spatial domain joint dimensionality reduction processing is characterized in that: The specific implementation method of step S2 includes the following steps: S2.

1. For the distance-Doppler-angle three-dimensional data block obtained in step S1, first select an angle unit to obtain the distance-Doppler spectrum RD spectrum; Specifically, the angle unit m is set to 1, and the RD spectrum is obtained. The modulus value is squared to obtain |x nl1 | 2 ; S2.

2. Assume that the static clutter and the nearby Doppler units with higher energy account for at most η% of all Doppler units. Use the prior value η% to locate the Doppler area of ​​static clutter and side lobes, set the static clutter and the Doppler area with higher energy to zero, and reduce the influence of large energy units on the average noise floor estimation. The expression is: S2.

3. Calculate the signal power A of distance unit n rn , the calculation formula is: Then calculate the set A of signal powers of all distance cells r ={A r1 ,A r2 ,…,A rN }; S2.

4. Estimate the noise power of each distance unit and take A r The signal power of the smallest β1% to β2% distance unit is calculated to get the average noise floor A. Noise , where β1 is the lower limit of the noise floor estimation proportional coefficient, and β2 is the upper limit of the noise floor estimation proportional coefficient; Set the threshold coefficient α, count the distance units where the signal power is greater than the threshold, and get the distance unit n where the Doppler diffusion clutter area is located. Clutter , a total of n C There is Doppler diffusion clutter in each range unit, and the expression is: n Clutter ={n|A rn >αA Noise }={n1,…,n C }; S2.

5. Traverse all angle units, obtain the distance unit where the corresponding Doppler diffusion clutter area is located, and complete the positioning of the Doppler diffusion clutter area.

4. According to claim 3, a 5G-aware Doppler spread clutter suppression method based on subcarrier domain and spatial domain joint dimensionality reduction processing is characterized in that: The specific implementation method of step S3 includes the following steps: S3.

1. Select the angle unit m to be processed and a distance unit n in the distance unit where the Doppler diffusion clutter area corresponding to m is located; S3.

2. Set the local processing area size of the angle-distance joint domain to: the number of angle units m a and the number of distance units n r ; Then the angle unit range of the local processing area corresponding to the specified angle unit m is: {m-(m a -1) / 2,…,m-1,m,m+1,…,m+(m a -1) / 2}; The local processing area distance unit range corresponding to the specified distance unit n is: {n-(n r -1) / 2,…,n-1,n,n+1,…,n+(n r -1) / 2}; Then all Doppler unit data of the local processing area are obtained to form a three-dimensional data block {dataRDA} of the local processing area, whose dimension is n r ×L×m a ; S3.

3. Rearrange the local processing area three-dimensional data block {dataRDA} into an angle-range-Doppler data format with a dimension of m a ×n r ×L, and then vectorize the angle-distance two-dimensional data to get m a n r ×L two-dimensional data matrix {dataARD} to be processed.

5. According to claim 4, a 5G-aware Doppler spread clutter suppression method based on subcarrier domain and spatial domain joint dimensionality reduction processing is characterized in that: The specific implementation method of step S4 includes the following steps: S4.

1. Select a Doppler unit l from the two-dimensional data matrix {dataARD} to be processed obtained in step S3 as the Doppler unit to be detected, and form a local processing area data vector X of the Doppler unit to be detected LPR , whose dimension is m a n r ×1; S4.

2. Process the Doppler unit to be detected and set the protection unit, the number is 2L g , the protection Doppler unit range is {lL g ,…,l-1,l+1,…,l+L g }, protect the Doppler unit from participating in the estimation of the clutter covariance matrix; S4.

3. Construct the training sample data matrix. According to the RMB criterion, the minimum number of training samples is L. r =2×m a ×n r , then select L from the two-dimensional data matrix {dataARD} r The data of Doppler units are used as the training sample data matrix. For the Doppler unit to be detected, the Doppler unit l of the training sample TS for: L TS ∈{lL g -L r / 2,…,lL g -1,l+L g +1,l+L g +L r / 2} If the endpoint of the Doppler unit interval of the training sample exceeds the range of all Doppler units, then the starting point of the Doppler interval of the training sample is cycled to the end point of all Doppler ranges, or the end point of the Doppler interval of the training sample is cycled to the starting point of all Doppler ranges, thereby forming a training sample data matrix {dataTS} with a dimension of m a n r ×L r ; S4.

4. Estimate the clutter covariance matrix of the Doppler unit to be detected using the data in the training sample data matrix The expression is: in,· H represents the conjugate transpose.

6. According to claim 5, a 5G-aware Doppler spread clutter suppression method based on subcarrier domain and spatial domain joint dimensionality reduction processing is characterized in that: The specific implementation method of step S5 includes the following steps: S5.

1. Construct the subcarrier domain steering vector b(f) of the target according to the angle unit and the corresponding distance unit selected in step S3. subc,n ) and the spatial steering vector a(θ m ), the expression is: Among them, b(f subc,n ) is an N×1-dimensional column vector, f subc,n is the baseband subcarrier frequency corresponding to the selected distance unit, f s =B is the OFDM signal bandwidth, f subc,n / f s is the normalized subcarrier frequency; a(θ m ) is a K×1 dimensional column vector, K is the number of array elements, d is the uniform linear array element spacing, λ n is the wavelength of the subcarrier corresponding to the selected distance unit, dsinθ m / λ n is the normalized spatial angular frequency; S5.

2. Calculate the target spatial-subcarrier domain two-dimensional steering vector s0, expressed as: In the formula represents the Kronecker direct product of two vectors, and the dimension of the target spatial-subcarrier domain two-dimensional steering vector is KN×1; Construct a dimensionality reduction transformation matrix T for joint dimensionality reduction in the subcarrier domain and the spatial domain, whose dimension is KN×m a n r , the calculation formula is: Calculate the two-dimensional local steering vector s after dimensionality reduction in the subcarrier domain and spatial domain LPR The expression is: s LPR =T H s0。 7. According to claim 6, a 5G-aware Doppler spread clutter suppression method based on subcarrier domain and spatial domain joint dimensionality reduction processing is characterized in that: The specific implementation method of step S6 includes the following steps: S6.

1. Using the clutter covariance matrix obtained in step S4 and the two-dimensional local steering vector after the subcarrier domain and spatial domain dimensionality reduction obtained in step S5, calculate the adaptive weight vector w of the Doppler unit to be detected set in step S4 opt , the expression is: Where μ is a constant, w opt The dimension is m a n r ×1; S6.

2. Calculate the adaptive processing output result y of the Doppler unit to be detected using the adaptive weight vector out (n,l,m), the output expression of the Doppler unit to be detected of the distance unit n, Doppler unit l, and angle unit m after adaptive filtering is: Among them, X LPR is the data vector of the local processing area of ​​the Doppler unit to be detected.

8. According to claim 7, a 5G-aware Doppler spread clutter suppression method based on subcarrier domain and spatial domain joint dimensionality reduction processing is characterized in that: The specific implementation method of step S7 is to set l = 1, ..., L, and calculate the output y corresponding to each Doppler unit out (n,l,m), get the adaptive processing output result y of the specified angle unit m and distance unit n out (n,m).

9. The 5G-aware Doppler spread clutter suppression method based on subcarrier domain and spatial domain joint dimensionality reduction processing according to claim 8 is characterized in that: The specific implementation method of step S8 includes the following steps: S8.

1. The distance unit n where Doppler spread clutter exists in the RD spectrum of the specified angle unit m obtained in step S2 Clutter ={n1,…,n C }, let the distance unit n = n1,…,n C , calculate the adaptive processing output y of all Doppler units in each distance unit where the clutter is located out (m); S8.

2. Let angle unit m = 1, ..., M, calculate all clutter distance units n corresponding to each angle unit Clutter All Doppler units output y out .

Images