Under-mine 5G communication beam alignment method based on millimeter wave array beam pattern

Abstract

The invention discloses an under-mine 5G communication beam alignment method based on a millimeter wave array beam pattern, and belongs to the field of physical layer signal processing of wireless communication. The objective of the invention is to solve the problem that the real-time performance and security of mine industrial communication are difficult to guarantee by using the existing beam alignment method. According to the method, a constant modulus super-large scale antenna array is equivalent to a discrete time complex exponential signal, the frequency spectrum gain is controlled by adjusting the time change rate of the signal frequency, and then an accurate triangular beam pattern for rapid beam alignment is generated based on the antenna array meeting the constant modulus constraint. A transmitting end transmits triangular beam patterns in different directions at different training frames and accumulates the triangular beam patterns in a staggered manner at a receiving end, and the receiving end can perform rapid channel angle estimation and beam alignment by analyzing gain differences caused by geometrical shapes and channel angles of the beam patterns of different frames, so that the real-time performance and safety of industrial communication under a mine are ensured. The method can be applied to communication beam alignment.

Description

Technical Field

[0001] The present invention belongs to the field of physical layer signal processing of wireless communications, and specifically relates to a method for 5G communication beam alignment in mines based on millimeter wave array beam patterns. Background Art

[0002] With the rapid development of wireless communication technology, 5G mobile communication technology has gradually matured and enabled digital industrial production. Among them, 5G wireless communication technology used in coal mining underground provides technical guarantees for low-latency, high-reliability, and fully automatic "digital smart mines". 5G wireless communication uses millimeter wave frequency bands (30GHz-300GHz) to greatly improve the information transmission rate and ensure the real-time transmission of industrial information underground. By increasing the carrier frequency to reduce the antenna size, ultra-large-scale antenna arrays can be integrated (the array can include hundreds or thousands of antennas), providing precise beam pointing, which is suitable for the tortuous and complex channel environment of mine tunnels. However, if the communication beam is not correctly pointed to the communication terminal, the communication performance will be reduced. Therefore, beam alignment technology is an important guarantee to support 5G wireless communication underground.

[0003] The existing beam alignment method assumes that the amplitude and phase of each antenna are independently adjustable, uses a rectangular beam pattern for beam scanning, and uses a binary tree method to search for angles so that the beam alignment error converges exponentially with the beam training time. However, millimeter-wave ultra-large-scale antenna arrays usually use an analog antenna array structure, and the amplitude of each antenna is equal, so only the phase is adjustable; secondly, the algorithm does not converge when the beam training time is short, that is, the alignment error is large when the beam training time is short, which is not conducive to fast beam alignment. In summary, the existing beam alignment method is difficult to ensure the real-time and security of underground industrial communications. Summary of the invention

[0004] The purpose of the present invention is to solve the problem that it is difficult to ensure the real-time and security of underground industrial communications using existing beam alignment methods, and to propose an underground 5G communication beam alignment method based on millimeter wave array beam patterns.

[0005] The technical solution adopted by the present invention to solve the above technical problems is:

[0006] A method for 5G communication beam alignment in mines based on millimeter wave array beam patterns, the method specifically comprising the following steps:

[0007] Step 1: The transmitter includes N T The simulated structure antenna array consists of N antennas. R A hybrid antenna array consisting of antennas and L RRF chains; and in the analog structure antenna array and the hybrid structure antenna array, the spacing between adjacent array elements is d;

[0008] The channel incident angle is θ∈[θ min ,θ max ), θ min is the minimum angle of incidence of the channel, θ max is the maximum angle of the channel incident, and the channel exit angle is φ∈[φ min ,φ max ), φ min is the minimum angle of the channel emission, φ max is the maximum angle of the channel emission. It is defined that there are M training frames in the communication beam training phase. The transmitter uses G T different beam patterns, and G T =M, the receiving end uses G R different beam patterns, and G R =L R ;

[0009] Step 2: Based on the beam pattern Solve the transmit antenna array simulated weighted beamforming vector

[0010] Step 3: Based on the beam pattern Solve the receiver antenna array simulated weighted beamforming vector

[0011] Step 4: According to Constructing the transmitter codebook dictionary according to Construct the receiving codebook dictionary The transmitter uses a weighted beamforming vector in the mth training frame The receiver always uses the weighted beam aggregation matrix during the M training frames training process Then the receiving end receives the mth frame received signal y according to the weighted beam aggregation matrix (m) , m=1,2,…,M;

[0012] Step 5: Obtain an estimated value of the strongest path incident angle based on the received signal of each frame at the receiving end

[0013] Step 6: Obtain an estimated value of the strongest path angle based on the received signal of each frame at the receiving end

[0014] Step 7: Use the estimate and Complete beam alignment.

[0015] Furthermore, the spacing d=λ / 2, where λ is the carrier wavelength.

[0016] Furthermore, the beam pattern adopted by the transmitting end is:

[0017]

[0018] in, is the gth T beam patterns, Beam pattern Angle at the geometric center;

[0019] The beam pattern used by the receiving end is:

[0020]

[0021] in, is the gth R beam patterns, Beam pattern The angle at the geometric center.

[0022] Furthermore, the specific process of step 2 is as follows:

[0023] Step 21: Let ω = (2πd / λ)sin(φ), according to φ∈[φ min ,φ max ) to get ω∈[ω min ,ω max ),ω min is a lower bound of ω, ω max is the upper bound of ω, substituting ω into the beam pattern Get function

[0024]

[0025] in, The superscript H stands for conjugate transpose;

[0026] Step 22: Solve the function In the interval ω∈[ω min ,ω max ), we get the continuous time function:

[0027]

[0028] Where t is time;

[0029] The constant term C and They are:

[0030]

[0031]

[0032] Step 23: Find the inverse function of equation (4) ω = f(t), and then use Find the original function of f(t) Then the complex exponential time signal s(t) is defined as:

[0033]

[0034] Where e is the base of the natural logarithm, j is the imaginary unit, and rect(·) represents the rectangular function;

[0035] Step 24: According to t=n T T, T = 1, the complex exponential time signal s(t) is sampled to obtain the discrete time signal s[n T ],n T =0,...,N T -1, that is, the transmitting end beamforming vector for:

[0036]

[0037] in, express The nth T elements.

[0038] Furthermore, the specific process of step three is:

[0039] Step 31: Let ω = (2πd / λ)sin(θ), according to θ∈[θ min ,θ max ) to get ω∈[ω min ,ω max ), substituting ω into the beam pattern Get function

[0040]

[0041] in, The superscript H stands for conjugate transpose;

[0042] Step 32: Solve the function In the interval ω∈[ω min ,ω max ), we get the continuous time function:

[0043]

[0044] Where t is time;

[0045] The constant term C and They are:

[0046]

[0047]

[0048] Step 3. Find the inverse function of equation (10) ω = f(t), and then use Find the original function of f(t) Then the complex exponential time signal s(t) is defined as:

[0049]

[0050] Where e is the base of the natural logarithm, j is the imaginary unit, and rect(·) represents the rectangular function;

[0051] Step 3 and 4: According to t=n R T, T = 1, sample s(t) and obtain the discrete time signal s[n R ],n R =0,...,N R -1, that is, the receiving end beamforming vector is obtained for:

[0052]

[0053] in, express The nth R elements.

[0054] Furthermore, the receiving end receives the signal y in the mth frame (m) for:

[0055]

[0056] Where L represents the number of channel path clusters, α l is the complex gain of the lth path cluster, Represents the plural, is the channel steering vector at the receiving end, is the channel steering vector at the transmitting end, θ l is the incident angle of the lth path cluster, φ l is the exit angle of the lth path cluster, Receive signal y for the mth frame (m) Gaussian white noise in;

[0057]

[0058]

[0059] Furthermore, the incident angle is estimated according to each frame of the received signal at the receiving end. The specific process is:

[0060] Step 51: y (m) Simplify and rewrite it into the following form:

[0061]

[0062] in, g θ (θ l )=(W RF W BS ) H a R (θ l );

[0063] Step 52: From y (m) , m=1,2,…,M, select the frame m containing the most received signal energy * :

[0064]

[0065] Among them, ||·|| 2 is the 2-norm;

[0066] Step 53: The mth * Frame reception signal Simplified to:

[0067]

[0068] in, Indicates the interference of other multipath signals on the strongest path received signal. For m * Gaussian white noise in the frame received signal, is the complex gain of the strongest path, is the exit angle of the strongest path, is the incident angle of the strongest path;

[0069] Step 54: According to and Estimated angle of incidence

[0070] Furthermore, the specific process of step 54 is as follows:

[0071] Step 541: Ignore distractions and have right The first R Elements modulo:

[0072]

[0073] in, Yes The first R The result of taking the modulus of the elements;

[0074]

[0075] in, Indicates the number of elements corresponding to the maximum modulo result;

[0076] The incident angle is For the The angle at the geometric center of the beam pattern;

[0077] Step 542: Definition Yes No. The result of taking the modulus of the elements is Yes No. The result of taking the modulus of the elements is the angle of incidence. which is reduced to:

[0078]

[0079] When Σ≥0, execute step 543, and when Σ<0, execute step 544;

[0080] Step 543: According to formula (2), and Further expressed as:

[0081]

[0082]

[0083] definition Then there is

[0084]

[0085]

[0086]

[0087] in, For the The angle at the geometric center of the beam pattern;

[0088] Calculate an estimate of u for:

[0089]

[0090] in, represents the "pseudo-inverse" of A;

[0091] The incident angle The estimated value is yes The first element in yes The second element in ;

[0092] Step 544: According to formula (2), and Further expressed as:

[0093]

[0094]

[0095] in, For the The angle at the geometric center of the beam pattern, For the The angle at the geometric center of the beam pattern;

[0096] definition Then there is

[0097]

[0098]

[0099]

[0100] The estimated value of u is calculated as:

[0101]

[0102] in, represents the “pseudo-inverse” of A′;

[0103] The incident angle The estimated value is yes The first element in yes The second element in .

[0104] Furthermore, the specific process of step six is ​​as follows:

[0105] Step 61: y (m) Simplify and rewrite it into the following form:

[0106]

[0107] in,

[0108] Step 6.2. From y (m) , m=1,2,…,M L R Select the element with the most received signal energy from the elements, that is:

[0109]

[0110] Step 6.3: Using the received signal of the M training frames Elements

[0111]

[0112]

[0113] in, is the complex gain of the strongest path, is the incident angle of the strongest path, Indicates that the M training frames at the transmitter are at angles Beamforming gain in direction, is the exit angle of the strongest path, Indicates the interference of other multipath signals on the strongest path received signal. is the Gaussian white noise in the received signal of the first frame The elements, is the Gaussian white noise in the received signal of the Mth frame The elements,

[0114] Step 64: According to and Get an estimate of the angle of incidence of the strongest path

[0115] Furthermore, the specific process of step seven is as follows:

[0116] The mean absolute errors of the incident angle estimation and the exit angle estimation are:

[0117]

[0118]

[0119] in, It represents the pursuit of expectations;

[0120] The transmitting and receiving ends use the following beam patterns respectively:

[0121]

[0122]

[0123] Substitute g(φ) into step 2 to obtain the weighted beamforming vector Ff at the transmitting end; substitute g(θ) into step 3 to obtain the weighted beamforming vector Ww at the receiving end;

[0124] The transmitting and receiving antenna arrays use Ff and Ww for beamforming and beam aggregation respectively to complete beam alignment.

[0125] The beneficial effects of the present invention are:

[0126] The present invention equates the constant modulus ultra-large-scale antenna array to a discrete-time complex exponential signal, controls the spectral gain by adjusting the time rate of change of the signal frequency, and then generates an accurate triangular beam pattern for fast beam alignment based on the antenna array that satisfies the constant modulus constraint. The transmitting end transmits triangular beam patterns in different directions in different training frames and accumulates them at the receiving end. The receiving end can perform fast channel angle estimation and beam alignment by analyzing the gain differences caused by the geometric shapes and channel angles of the beam patterns in different frames. Compared with the existing beam alignment methods, the method of the present invention can obtain more channel angle information through accurate beam patterns, and has a faster channel angle estimation error convergence speed during the beam training process, so it can achieve fast beam alignment and ensure the real-time and safety of underground industrial communications. BRIEF DESCRIPTION OF THE DRAWINGS

[0127] Figure 1 It is a flowchart of the transmitting end and the receiving end of the present invention;

[0128] Figure 2 is the beam pattern of the transmitting end;

[0129] In the figure: the dotted line is the theoretical value, and the solid line is the simulation value;

[0130] Figure 3a is the mean absolute error of the incident angle at different signal-to-noise ratios;

[0131] Figure 3b is the root mean square error of the incident angle at different signal-to-noise ratios;

[0132] Figure 3c is the average absolute error of the emission angle at different signal-to-noise ratios;

[0133] Figure 3d is the root mean square error of the exit angle at different signal-to-noise ratios. DETAILED DESCRIPTION

[0134] Specific implementation method 1: Combination Figure 1 This embodiment describes a method for aligning a 5G communication beam in a mine based on a millimeter wave array beam pattern, wherein the millimeter wave array refers to a carrier frequency of millimeter waves, and the method specifically includes the following steps:

[0135] Step 1: The transmitter includes N T The simulated structure antenna array consists of N antennas. R A hybrid antenna array consisting of antennas and L R RF chains; and in the analog structure antenna array and the hybrid structure antenna array, the spacing between adjacent array elements is d;

[0136] The channel incident angle is θ∈[θ min ,θ max ), θ min is the minimum angle of incidence of the channel, θ max is the maximum angle of the channel incident, and the channel exit angle is φ∈[φ min ,φ max ), φ min is the minimum angle of the channel emission, φ max is the maximum angle of the channel emission. It is defined that there are M training frames in the communication beam training phase. The transmitter uses G T different beam patterns, and G T =M, the beam pattern of the transmitting end is as follows Figure 2 As shown, the beam pattern of the receiving end is the same as that of the transmitting end. The receiving end uses G R different beam patterns, and G R =L R ;

[0137] Step 2: Based on the beam pattern Solve the transmit antenna array simulated weighted beamforming vector

[0138] Step 3: Based on the beam pattern Solve the receiver antenna array simulated weighted beamforming vector

[0139] Step 4: According to Constructing the transmitter codebook dictionary according to Construct the receiving codebook dictionary The transmitter uses a weighted beamforming vector in the mth training frame The receiver always uses the weighted beam aggregation matrix during the M training frames training process Then the receiving end receives the mth frame received signal y according to the weighted beam aggregation matrix (m) , m=1,2,…,M;

[0140] Step 5: Obtain an estimated value of the strongest path incident angle based on the received signal of each frame at the receiving end

[0141] Step 6: Obtain an estimated value of the strongest path angle based on the received signal of each frame at the receiving end

[0142] Step 7: Use the estimate and Complete beam alignment.

[0143] Specific implementation method 2: This implementation method is different from specific implementation method 1 in that the spacing d=λ / 2, where λ is the carrier wavelength.

[0144] The other steps and parameters are the same as those in the first embodiment.

[0145] Specific implementation method three: This implementation method is different from specific implementation method two in that the beam pattern used by the transmitting end is:

[0146]

[0147] in, is the gth T beam patterns, Beam pattern Angle at the geometric center;

[0148] The beam pattern used by the receiving end is:

[0149]

[0150] in, is the gth R beam patterns, Beam pattern The angle at the geometric center.

[0151] The other steps and parameters are the same as those in the second specific implementation manner.

[0152] Specific implementation method 4: This implementation method is different from specific implementation method 3 in that the specific process of step 2 is as follows:

[0153] Step 21: Let ω = (2πd / λ)sin(φ), according to φ∈[φ min ,φ max ) to get ω∈[ω min ,ω max ),ω min is a lower bound of ω, ω max is the upper bound of ω, substituting ω into the beam pattern Get function

[0154]

[0155] in, The superscript H stands for conjugate transpose;

[0156] Step 22: Solve the function In the interval ω∈[ω min ,ω max ), we get the continuous time function:

[0157]

[0158] Where t is time;

[0159] It is easy to know that the continuous time function t = f -1 (ω) is an increasing function, and we define t∈[0,N T ), then f -1 (ω min )=0,f -1 (ω max )=N T , thus obtaining the constant term C and They are:

[0160]

[0161]

[0162] Step 23: Find the inverse function of equation (4) ω = f(t), and then use Find the original function of f(t) Then the complex exponential time signal s(t) is defined as:

[0163]

[0164] Among them, e is the base of natural logarithm, j is the imaginary unit, rect(·) represents the rectangular function, and it is easy to know that t∈[0,N T );

[0165] Step 24: According to t=n T T, T = 1, the complex exponential time signal s(t) is sampled to obtain the discrete time signal s[n T ],n T =0,...,N T -1, that is, the transmitting end beamforming vector for:

[0166]

[0167] in, express The nth T elements.

[0168] The other steps and parameters are the same as those in the third specific implementation manner.

[0169] Specific implementation method five: This implementation method is different from specific implementation method three in that the specific process of step three is as follows:

[0170] Step 31: Let ω = (2πd / λ)sin(θ), according to θ∈[θ min ,θ max ) to get ω∈[ω min ,ω max ), substituting ω into the beam pattern Get function

[0171]

[0172] in, The superscript H stands for conjugate transpose;

[0173] Step 32: Solve the function In the interval ω∈[ω min ,ω max ), we get the continuous time function:

[0174]

[0175] Where t is time;

[0176] It is easy to know that the continuous time function t = f -1 (ω) is an increasing function, and we define t∈[0,N R ), then f -1 (ω min )=0,f -1 (ω max )=N R , thus obtaining the constant term C and They are:

[0177]

[0178]

[0179] Step 3. Find the inverse function of equation (10) ω = f(t), and then use Find the original function of f(t) Then the complex exponential time signal s(t) is defined as:

[0180]

[0181] Among them, e is the base of natural logarithm, j is the imaginary unit, rect(·) represents the rectangular function, and it is easy to know that t∈[0,N R );

[0182] Step 3 and 4: According to t=n R T, T = 1, sample s(t) and obtain the discrete time signal s[n R ],n R =0,...,N R -1, that is, the receiving end beamforming vector is obtained for:

[0183]

[0184] in, express The nth R elements.

[0185] The other steps and parameters are the same as those in the third specific implementation manner.

[0186] Specific implementation method 6: This implementation method is different from specific implementation methods 4 or 5 in that the receiving end receives the signal y in the mth frame (m) for:

[0187]

[0188] Where L represents the number of channel path clusters, α l is the complex gain of the lth path cluster, Represents the plural, is the channel steering vector at the receiving end, is the channel steering vector at the transmitting end, θ l is the incident angle of the lth path cluster, φ l is the exit angle of the lth path cluster, Receive signal y for the mth frame (m) Gaussian white noise in;

[0189]

[0190]

[0191] The other steps and parameters are the same as those in the fourth or fifth embodiment.

[0192] Specific implementation method 7: This implementation method is different from specific implementation method 6 in that the strongest path of the multipath channel is defined as l * , the incident angle is estimated according to the received signal of each frame at the receiving end The specific process is:

[0193] Step 51: y (m) Simplify and rewrite it into the following form:

[0194]

[0195] in, Right now

[0196] Step 52: From y (m) , m=1,2,…,M, select the frame m containing the most received signal energy * :

[0197]

[0198] Among them, ||·|| 2 is the 2-norm;

[0199] Step 53: The mth * Frame reception signal Simplified to:

[0200]

[0201] in, Indicates the interference of other multipath signals on the strongest path received signal. For m * Gaussian white noise in the frame received signal, is the complex gain of the strongest path, is the exit angle of the strongest path, is the incident angle of the strongest path,

[0202] Step 54: According to and Estimated angle of incidence

[0203] The other steps and parameters are the same as those in the sixth embodiment.

[0204] Specific implementation eight: This implementation is different from specific implementation seven in that the specific process of step five-four is as follows:

[0205] Step 541: Ignore distractions and have right The first R Elements modulo:

[0206]

[0207] in, Yes The first R The result of taking the modulus of the elements;

[0208]

[0209] in, Indicates the number of elements corresponding to the maximum modulo result;

[0210] but At the maximum modulus Within the angle range of For the The angle at the geometric center of the beam pattern;

[0211] Step 542: By comparing beam patterns and In Angle The gain size on the The value range of Yes No. The result of taking the modulus of the elements is Yes No. The result of taking the modulus of the elements is the angle of incidence. which is reduced to:

[0212]

[0213] When Σ≥0, execute step 543, and when Σ<0, execute step 544;

[0214] Step 543: According to formula (2), and Further expressed as:

[0215]

[0216]

[0217] definition Then there is

[0218]

[0219]

[0220]

[0221] in, For the The angle at the geometric center of the beam pattern;

[0222] Calculate the estimated value of u using the LS algorithm for:

[0223]

[0224] in, represents the "pseudo-inverse" of A, that is The incident angle The estimated value is yes The first element in yes The second element in ;

[0225] Step 544: According to formula (2), and Further expressed as:

[0226]

[0227]

[0228] in, For the The angle at the geometric center of the beam pattern, For the The angle at the geometric center of the beam pattern;

[0229] definition Then there is

[0230]

[0231]

[0232]

[0233] The estimated value of u calculated by the LS algorithm is:

[0234]

[0235] in, represents the "pseudo-inverse" of A', that is The incident angle The estimated value is yes The first element in yes The second element in .

[0236] The other steps and parameters are the same as those in the seventh embodiment.

[0237] Specific implementation method 9: This implementation method is different from specific implementation method 6 in that the specific process of step 6 is as follows:

[0238] Step 61: y (m) Simplify and rewrite it into the following form:

[0239]

[0240] in,

[0241] Step 6.2. From y (m) , m=1,2,...,M L R Select the element with the most received signal energy from the elements, that is:

[0242]

[0243] Step 6.3: Using the received signal of the M training frames Elements

[0244]

[0245]

[0246] in, is the complex gain of the strongest path, is the incident angle of the strongest path, Indicates that the M training frames at the transmitter are at angles Beamforming gain in direction, is the exit angle of the strongest path, Indicates the interference of other multipath signals on the strongest path received signal. is the Gaussian white noise in the received signal of the first frame The elements, is the Gaussian white noise in the received signal of the Mth frame The elements, g φ (φ l )=[g 1 (φ l ),...,g M (φ l )] T ,

[0247] Step 64: According to and Get an estimate of the angle of incidence of the strongest path The estimation method is the same as step 54.

[0248] The other steps and parameters are the same as those in the sixth embodiment.

[0249] The specific process of step 64 is:

[0250] Step 641: Ignore distractions and have right Taking the mth element modulo we get:

[0251]

[0252]

[0253] Among them, m max Indicates the frame corresponding to the maximum modulo result;

[0254] but At the maximum modulus Within the angle range, the strongest angle is obtained. The mth max The angle at the geometric center of the beam pattern;

[0255] Step 642: By comparing beam patterns and In Angle The gain size on the output angle can be further reduced. The value range of Yes The mth max -1 element modulo result, Yes The mthmax +1 element modulo result, then the exit angle which is reduced to:

[0256]

[0257] When Σ≥0, execute step 643, and when Σ<0, execute step 644;

[0258] Step 643: According to formula (1), and Further expressed as:

[0259]

[0260]

[0261] definition Then there is

[0262]

[0263]

[0264]

[0265] in, The mth max -1 Angle at the geometric center of the beam pattern;

[0266] The estimated value of u is calculated by LS algorithm for:

[0267]

[0268] in, represents the "pseudo-inverse" of A, then the exit angle The estimated value is yes The first element in yes The second element in ;

[0269] Step 644: According to formula (1), and Further expressed as:

[0270]

[0271]

[0272] in, The mth max+1 Angle at the geometric center of the beam pattern;

[0273] definition Then there is

[0274]

[0275]

[0276]

[0277] The estimated value of u calculated by the LS algorithm is:

[0278]

[0279] The exit angle The estimated value is yes The first element in yes The second element in .

[0280] Specific implementation method 10: This implementation method is different from specific implementation methods 8 or 9 in that the specific process of step 7 is as follows:

[0281] The mean absolute error (MAE) of the incident angle estimation and the mean absolute error (MAE) of the outgoing angle estimation are:

[0282]

[0283]

[0284] in, It represents the pursuit of expectations;

[0285] The transmitting and receiving ends use the following beam patterns respectively:

[0286]

[0287]

[0288] Substitute g(φ) into step 2 to obtain the weighted beamforming vector Ff at the transmitting end; substitute g(θ) into step 3 to obtain the weighted beamforming vector Ww at the receiving end;

[0289] The transmitting and receiving antenna arrays use Ff and Ww for beamforming and beam aggregation respectively to complete beam alignment.

[0290] The other steps and parameters are the same as those in the eighth or ninth embodiment.

[0291] Simulation verification:

[0292] like Figure 2 As shown in FIG. 1 , the feasibility of the beam pattern design of the method proposed in the present invention is verified by simulating and analyzing the beam pattern. Taking the beam pattern of the transmitting end as an example (the beam pattern of the receiving end is the same), assume that N T =1024,G T =10, the method proposed by the present invention is based on a given beam pattern Solve to get the corresponding like Figure 2 As shown, The beam pattern simulation value (represented by the solid line) fits the theoretical value (represented by the dotted line) well, verifying the feasibility of the beam pattern design method of the present invention.

[0293] like Figure 3a , Figure 3b , Figure 3c and Figure 3d As shown in FIG. 1 , the effectiveness of the method proposed in the present invention is verified by simulating and analyzing the channel angle error of beam alignment. T =512,N R =1024,L R =4, the number of beam training frames M = 4. The number of channel multipath L ~ min{max{1,Poisson(μ)},L max}, where Poisson(μ) represents the Poisson distribution with mean μ, L max represents the maximum value of the possible number of path clusters, μ = 1.6 in the case of millimeter wave carrier frequency, L max =4. Channel gain in U l ~U[0,1], In the case of millimeter wave carrier frequency, τ =2.8, ζ = 4. Channel angle φ l ,θ l ∈U[-π / 3,π / 3].

[0294] like Figure 3a and Figure 3c As shown, the present invention adopts the mean absolute error of angle The method proposed in the present invention uses a triangular beam pattern with a beam width of 45° in the m=1, 2, 3, and 4 training frames respectively. By comparing the gain differences of different beam patterns, the MAE can be reduced to infinitesimal (theoretically), such as Figure 3cAs shown in the figure, the MAE of AoD (angle of departure) can be reduced to less than 1° in the case of a single-path channel. Because the triangular beam pattern can obtain more channel angle information than the traditional rectangular beam pattern, the method proposed in the present invention greatly reduces the beam training time and can achieve fast beam alignment. Figure 3b and Figure 3d As shown, the present invention adopts the root mean square error To analyze the extreme outlier value of the angle alignment error, the extreme outlier value of the method proposed in the present invention is significantly less in the case of a single-path channel. It should be noted that since the receiving end can use L R The RF chains are trained for beamforming. Therefore, the error convergence rate of AoA (angle of incidence) is faster than that of AoD (angle of departure). After M=4 beamforming frames, the angle alignment error of AoA is obviously smaller than that of AoD.

[0295] Depend on Figure 3a , Figure 3b , Figure 3c and Figure 3d It can be seen that under the condition that the signal-to-noise ratio at the receiving end of the communication system is not less than 10dB, the method proposed in the present invention only needs M=4 beam training frames to reduce the single-path channel beam alignment error to within 2° and the multi-path channel beam alignment error to within 5°. Considering that there will be channel angle expansion in the actual channel (the millimeter wave channel angle expansion is about 10°), the beam alignment accuracy of the beam alignment method proposed in the present invention can meet the needs of the 5G wireless communication system in the mine and realize fast beam alignment.

[0296] The above calculation examples of the present invention are only used to explain the calculation model and calculation process of the present invention in detail, and are not intended to limit the implementation methods of the present invention. For ordinary technicians in the relevant field, other different forms of changes or modifications can be made based on the above description. It is impossible to list all the implementation methods here. All obvious changes or modifications derived from the technical solution of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for 5G communication beam alignment in mines based on millimeter wave array beam patterns, characterized in that: The method specifically comprises the following steps: Step 1: The transmitter includes N T The simulated structure antenna array consists of N antennas. R A hybrid antenna array consisting of antennas and L R RF chains; and in the analog structure antenna array and the hybrid structure antenna array, the spacing between adjacent array elements is d; The channel incident angle is θ∈[θ min ,θ max ), θ min is the minimum angle of incidence of the channel, θ max is the maximum angle of the channel incident, and the channel exit angle is φ∈[φ min ,φ max ), φ min is the minimum angle of the channel emission, φ max is the maximum angle of the channel emission. It is defined that there are M training frames in the communication beam training phase. The transmitter uses G T different beam patterns, and G T =M, the receiving end uses G R different beam patterns, and G R =L R ; Step 2: Based on the beam pattern Solve the transmit antenna array simulated weighted beamforming vector Step 3: Based on the beam pattern Solve the receiver antenna array simulated weighted beamforming vector Step 4: According to Constructing the transmitter codebook dictionary according to Construct the receiving codebook dictionary The transmitter uses a weighted beamforming vector in the mth training frame The receiver always uses the weighted beam aggregation matrix during the M training frames training process Then the receiving end receives the mth frame received signal y according to the weighted beam aggregation matrix (m) , m=1,2,…,M; Step 5: Obtain an estimated value of the strongest path incident angle based on the received signal of each frame at the receiving end Step 6: Obtain an estimated value of the strongest path angle based on the received signal of each frame at the receiving end Step 7: Use the estimate and Complete beam alignment.

2. The method for 5G communication beam alignment in mines based on millimeter wave array beam pattern according to claim 1 is characterized in that: The spacing d=λ / 2, where λ is the carrier wavelength.

3. The method for 5G communication beam alignment in mines based on millimeter wave array beam pattern according to claim 2 is characterized in that: The beam pattern used by the transmitting end is: in, is the gth T beam patterns, Beam pattern Angle at the geometric center; The beam pattern used by the receiving end is: in, is the gth R beam patterns, Beam pattern The angle at the geometric center.

4. The method for 5G communication beam alignment in mines based on millimeter wave array beam pattern according to claim 3 is characterized in that: The specific process of step 2 is as follows: Step 21: Let ω = (2πd / λ)sin(φ), according to φ∈[φ min ,φ max ) to get ω∈[ω min ,ω max ),ω min is a lower bound of ω, ω max is the upper bound of ω, substituting ω into the beam pattern Get function in, The superscript H stands for conjugate transpose; Step 22: Solve the function In the interval ω∈[ω min ,ω max ), we get the continuous time function: Where t is time; The constant term C and They are: Step 23: Find the inverse function of equation (4) ω = f(t), and then use Find the original function of f(t) Then the complex exponential time signal s(t) is defined as: Where e is the base of the natural logarithm, j is the imaginary unit, and rect(·) represents the rectangular function; Step 24: According to t=n T T, T = 1, the complex exponential time signal s(t) is sampled to obtain the discrete time signal s[n T ],n T =0,...,N T -1, that is, the transmitting end beamforming vector for: in, express The nth T elements.

5. The method for 5G communication beam alignment in mines based on millimeter wave array beam pattern according to claim 3 is characterized in that: The specific process of step three is: Step 31: Let ω = (2πd / λ)sin(θ), according to θ∈[θ min ,θ max ) to get ω∈[ω min ,ω max ), substituting ω into the beam pattern Get function in, The superscript H stands for conjugate transpose; Step 32: Solve the function In the interval ω∈[ω min ,ω max ), we get the continuous time function: Where t is time; The constant term C and They are: Step 3. Find the inverse function of equation (10) ω = f(t), and then use Find the original function of f(t) Then the complex exponential time signal s(t) is defined as: Where e is the base of the natural logarithm, j is the imaginary unit, and rect(·) represents the rectangular function; Step 3 and 4: According to t=n R T, T = 1, sample s(t) and obtain the discrete time signal s[n R ],n R =0,...,N R -1, that is, the receiving end beamforming vector is obtained for: in, express The nth R elements.

6. The method for 5G communication beam alignment in mines based on millimeter wave array beam pattern according to claim 4 or 5, characterized in that: The receiving end receives the signal y in the mth frame (m) for: Where L represents the number of channel path clusters, α l is the complex gain of the lth path cluster, Represents the plural, is the channel steering vector at the receiving end, is the channel steering vector at the transmitting end, θ l is the incident angle of the lth path cluster, φ l is the exit angle of the lth path cluster, Receive signal y for the mth frame (m) Gaussian white noise in; 7. The method for 5G communication beam alignment in mines based on millimeter wave array beam pattern according to claim 6 is characterized in that: The incident angle is estimated according to each frame received signal of the receiving end The specific process is: Step 51: y (m) Simplify and rewrite it into the following form: in, Step 52: From y (m) , m=1,2,…,M, select the frame m containing the most received signal energy * : Among them, ||·||2 is the 2-norm; Step 53: The mth * Frame reception signal Simplified to: in, Indicates the interference of other multipath signals on the strongest path received signal. For m * Gaussian white noise in the frame received signal, is the complex gain of the strongest path, is the exit angle of the strongest path, is the incident angle of the strongest path; Step 54: According to and Estimated angle of incidence 8. The method for 5G communication beam alignment in mines based on millimeter wave array beam pattern according to claim 7, characterized in that: The specific process of step 54 is as follows: Step 541: Ignore distractions and have right The first R Elements modulo: in, Yes The first R The result of taking the modulus of the elements; in, Indicates the number of elements corresponding to the maximum modulo result; The incident angle is For the The angle at the geometric center of the beam pattern; Step 542: Definition Yes No. The result of taking the modulus of the elements is Yes No. The result of taking the modulus of the elements is the angle of incidence. which is reduced to: When Σ≥0, execute step 543, and when Σ<0, execute step 544; Step 543: According to formula (2), and Further expressed as: definition Then there is in, For the The angle at the geometric center of the beam pattern; Calculate an estimate of u for: in, represents the "pseudo-inverse" of A; The incident angle The estimated value is yes The first element in yes The second element in ; Step 544: According to formula (2), and Further expressed as: in, For the The angle at the geometric center of the beam pattern, For the The angle at the geometric center of the beam pattern; definition Then there is The estimated value of u is calculated as: in, represents the "pseudo-inverse" of A'; The incident angle The estimated value is yes The first element in yes The second element in .

9. The method for 5G communication beam alignment in mines based on millimeter wave array beam pattern according to claim 6, characterized in that: The specific process of step six is ​​as follows: Step 61: y (m) Simplify and rewrite it into the following form: in, Step 6.

2. From y (m) , m=1,2,...,M L R Select the element with the most received signal energy from the elements, that is: Step 6.3: Using the received signal of the M training frames Elements in, is the complex gain of the strongest path, is the incident angle of the strongest path, Indicates that the M training frames at the transmitter are at angles Beamforming gain in direction, is the exit angle of the strongest path, Indicates the interference of other multipath signals on the strongest path received signal. is the Gaussian white noise in the received signal of the first frame The elements, is the Gaussian white noise in the received signal of the Mth frame The elements, g φ (φ l )=[g1(φ l ),...,g M (φ l )] T , Step 64: According to and Get an estimate of the angle of incidence of the strongest path 10. The method for 5G communication beam alignment in mines based on millimeter wave array beam pattern according to claim 8 or 9, characterized in that: The specific process of step seven is as follows: The mean absolute errors of the incident angle estimation and the exit angle estimation are: in, It represents the pursuit of expectations; The transmitting and receiving ends use the following beam patterns respectively: Substitute g(φ) into step 2 to obtain the weighted beamforming vector Ff at the transmitting end; substitute g(θ) into step 3 to obtain the weighted beamforming vector Ww at the receiving end; The transmitting and receiving antenna arrays use Ff and Ww for beamforming and beam aggregation respectively to complete beam alignment.

Images