A hierarchical beam training method and apparatus based on channel coding

Abstract

The application provides a layered beam training method and device based on channel coding, which encodes an L-bit binary sequence of indexes (0 to N T -1) of N T angle directions into an M-bit binary sequence by using an encoder; constructs a beam set by using N T M-bit binary sequences; generates a wide-beam layered codebook corresponding to the beam set based on a narrow-beam codebook; and performs layered beam training on each user by using the wide-beam layered codebook and a decoder to select an optimal beamforming code word of each user in the narrow-beam codebook. The layered beam training problem is modeled as a channel coding and decoding problem, so that the reliability of the layered beam training is improved by using the coding gain of the channel coding to improve the data transmission reliability and reduce the bit error rate; and the low pilot overhead of the layered beam training is ensured by using the advantages of the layered beam training.

Description

Technical Field

[0001] This invention relates to the field of wireless mobile communication technology, and in particular to a hierarchical beam training method and apparatus based on channel coding. Background Technology

[0002] Terahertz communication technology, with its abundant spectrum resources, can exponentially increase transmission rates to meet the exponentially growing wireless data traffic demands of 6G communication. It is considered a potential key technology for 6G mobile communication and has broad application prospects. Meanwhile, massive MIMO systems, equipped with numerous antennas, can generate directional beams with extremely high array gain through different precoding designs, compensating for the significant communication path loss in the terahertz band and achieving orders-of-magnitude increases in data rates.

[0003] High-gain, ultra-large-scale MIMO transmission requires accurate channel state information. As antenna array size increases, accurate channel information estimation leads to significant pilot overhead. Therefore, codebook-based beamforming methods are widely used in practical systems to obtain channel state information. Existing codebook-based beamforming methods are mainly divided into narrow-beam codebook-based and wide-beam codebook-based methods. Narrow-beam codebook-based training methods typically employ beam scanning algorithms (see A. Alkhateeb, G. Leus, and RW Heath, “Limited feedback hybrid precoding for multi-user millimeter wave systems,” IEEE Trans. Wireless Commun., vol. 14, no. 11, pp. 6481–6494, 2015.). This method selects the optimal codeword based on the power of the received signal by sequentially testing codewords in the codebook. While beam scanning algorithms can achieve optimal performance, their training overhead increases linearly with the number of antennas, resulting in excessive overhead in large-scale MIMO systems. Therefore, hierarchical codebooks based on wide beams have been proposed to reduce pilot overhead caused by beam scanning. The performance and overhead of hierarchical beam training depend on the design of the hierarchical codebook. For hierarchical beam training (see C. Qi, K. Chen, OADobre, and GYLi, “Hierarchical codebook-based multiuser beamtraining for millimeter wave massive MIMO,” IEEE Trans. Wireless Commun., vol. 19, no. 12, pp. 8142–8152, 2020.), a predefined hierarchical codebook is typically used, where the angular region covered by codewords in the upper-layer codebook is divided into several smaller angular regions covered by codewords in the lower-layer codebook. However, hierarchical codebooks have low beam directional gain, are highly susceptible to noise at low signal-to-noise ratios, and exhibit poor beam training performance.

[0004] Currently, no training scheme has been proposed that can guarantee both high reliability and low pilot overhead for beam training in terahertz massive MIMO systems under low signal-to-noise ratio conditions. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides a hierarchical beam training method and apparatus based on channel coding. The hierarchical beam training problem is modeled as a channel coding and decoding problem, thereby leveraging the advantages of channel coding gain in improving data transmission reliability and reducing bit error rate to enhance the reliability of hierarchical beam training. Simultaneously, the advantages of hierarchical beam training are utilized to ensure low pilot overhead.

[0006] In a first aspect, the present invention provides a hierarchical beam training method based on channel coding, for large-scale MIMO terahertz systems, wherein the base station of the large-scale MIMO terahertz system has N T The method includes: [Number of antennas] and the user has a single antenna.

[0007] Divide the spatial angle [-1, 1] into N equal parts. T There are several angles and directions, numbered N in sequence. T Each angle direction is assigned from 0 to N. T -1 index;

[0008] Using a pre-built encoder, the L-bit binary sequence representing the index of each angular direction is encoded into an M-bit binary sequence; where L = log₂N T M≥L;

[0009] Using N T A beam set is constructed using M-bit binary sequences; wherein the beam set has M layers, each layer containing one beam, and the m-th position of the beam in the l-th layer is assigned the l-th bit of the M-bit binary sequence obtained by index m; m∈(0, 1, 2, ..., N) T -1), l∈(0,1,2,…,M);

[0010] Based on beamforming codewords with a number of N T The narrow beam codebook is used to generate a wide beam hierarchical codebook corresponding to the beam set;

[0011] Using the wide-beam hierarchical codebook and the pre-built decoder, hierarchical beam training is performed on each user in the large-scale MIMO terahertz system to select the optimal beamforming codeword for each user in the narrow-beam codebook.

[0012] According to the channel-coding-based hierarchical beamforming method provided by the present invention, the step of encoding an L-bit binary sequence representing the index of each angular direction into an M-bit binary sequence using a pre-constructed encoder includes:

[0013] Let the index m of the angle direction be represented by an L-bit binary sequence denoted as b. m ;

[0014] The encoder's encoding algorithm will convert the b m Encoded as an M-bit binary sequence x m ;

[0015] The encoding algorithm of the encoder is as follows:

[0016] x = fen (b)

[0017] In the above formula, b is the input L-bit binary sequence, x is the output M-bit binary sequence, and f en (·) represents the encoding function of the encoder.

[0018] According to the channel coding-based hierarchical beam training method provided by the present invention, the narrow beam codebook The expression is as follows:

[0019]

[0020] in, The guiding vector can be represented as For spatial angles.

[0021] According to the channel coding-based hierarchical beamforming training method provided by the present invention, the number of beamforming codewords is N. T The narrow-beam codebook is used to generate a wide-beam hierarchical codebook corresponding to the beam set, including:

[0022] Let the l-th layer beam in the beam set be denoted as B(l), and let the beam obtained by bit flipping B(l) be denoted as B. * (l);

[0023] The wide-beam layered codebook is defined to have M layers, and the l-th layered codebook contains beamforming codewords corresponding to beam B(l). and beam Corresponding beamforming codeword

[0024] Using the narrow beam codebook and the beam B(l) / the beam Generate the This leads to the generation of the wide-beam hierarchical codebook.

[0025] According to the channel coding-based hierarchical beamforming training method provided by the present invention, the number of beamforming codewords is N. T The narrow-beam codebook is used to generate a wide-beam hierarchical codebook corresponding to the beam set, including:

[0026] Let the l-th layer beam in the beam set be denoted as B(l);

[0027] The wide-beam layered codebook is defined as having M+1 layers, and the l-th layer of the first M layers contains the beamforming codeword corresponding to beam B(l). The final layer of the codebook is the narrow beam codebook;

[0028] The narrow beam codebook and the beam B(l) are used to generate the... This leads to the generation of the wide-beam hierarchical codebook.

[0029] According to the channel coding-based hierarchical beam training method provided by the present invention, the generation of the beam using the narrow beam codebook and the beam B(l) is... The process includes:

[0030] Based on the beam B(l), determine Corresponding beam coverage area

[0031] According to the above Determine that the beam coverage range in the narrow beam codebook is within the range of the specified beam coverage range. Beamforming codewords in the text;

[0032] The beam coverage area in the narrow beam codebook is within the range of The weighted superposition result of the beamforming codewords in the above is used as the

[0033] According to the channel coding-based hierarchical beamforming training method provided by the present invention, the step of performing hierarchical beamforming training on each user in the large-scale MIMO terahertz system using the wide-beam hierarchical codebook and a pre-built decoder to select the optimal beamforming codeword for each user in the narrow-beam codebook includes:

[0034] The control base station sequentially sends the following to each user: and the Make each of the users compare the received The signal amplitude and the received The signal amplitude, when receiving the The signal amplitude is greater than that of the received signal. The signal amplitude is recorded as 0 when receiving the signal. The signal amplitude is less than that of the receiver. The signal amplitude is recorded as 1, and the recorded 0 or 1 is regarded as the beam training result of each user under the l-th layer beam in the beam set;

[0035] Control each user to provide the codewords composed of the beam training results of each layer of beams in the beam set;

[0036] The decoder is used to decode the codeword to obtain the index of the angle direction of each user.

[0037] Search for the index in the narrow beam codebook Corresponding beamforming codewords and the This serves as the optimal beamforming codeword for each of the aforementioned users.

[0038] According to the channel coding-based hierarchical beamforming training method provided by the present invention, the step of performing hierarchical beamforming training on each user in the large-scale MIMO terahertz system using the wide-beam hierarchical codebook and a pre-built decoder to select the optimal beamforming codeword for each user in the narrow-beam codebook includes:

[0039] The control base station sends to each of the users And enable each user to record and provide feedback. The power P(l) of the received signal;

[0040] Based on P(l) and the decoder, the index of the angular direction of each user is calculated.

[0041] The control base station tests the index in the narrow beam codebook for each user. and Beamforming code and Compare the data received by each user and stated The signal amplitude is determined, and the beamforming codeword with the larger signal amplitude is used as the optimal beamforming codeword for each user.

[0042] Secondly, the present invention provides a channel-coding-based hierarchical beam training device for massive MIMO terahertz systems, wherein the base station of the massive MIMO terahertz system has N T The device includes one antenna and a user has a single antenna.

[0043] Index marker units are used to divide the spatial angle [-1, 1] into N equal parts. T There are several angles and directions, numbered N in sequence. T Each angle direction is assigned from 0 to N. T -1 index;

[0044] An encoding unit is used to encode an L-bit binary sequence, represented by the index of each angular direction, into an M-bit binary sequence using a pre-built encoder; where L = log₂N. T M≥L;

[0045] Beamforming unit, used to utilize N TA beam set is constructed using M-bit binary sequences; wherein the beam set has M layers, each layer containing one beam, and the m-th position of the beam in the l-th layer is assigned the l-th bit of the M-bit binary sequence obtained by index m; m∈(0, 1, 2, ..., N) T -1), l∈(0,1,2,…,M);

[0046] Wide-beam hierarchical codebook generation unit, used for beamforming codewords with N characters. T The narrow beam codebook is used to generate a wide beam hierarchical codebook corresponding to the beam set;

[0047] A hierarchical beam training unit is used to perform hierarchical beam training on each user in the massive MIMO terahertz system using the wide-beam hierarchical codebook and a pre-built decoder, so as to select the optimal beamforming codeword for each user in the narrow-beam codebook.

[0048] Thirdly, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the channel coding-based hierarchical beam training method as described in the first aspect.

[0049] This invention provides a hierarchical beam training method based on channel coding, for large-scale MIMO terahertz systems, wherein the base station of the large-scale MIMO terahertz system has N T Each user has a single antenna, including: dividing the spatial angle [-1, 1] into N equal parts. T There are several angles and directions, numbered N in sequence. T Each angle direction is assigned from 0 to N. T The index is -1; using a pre-built encoder, the L-bit binary sequence represented by the index of each angle direction is encoded into an M-bit binary sequence; where L = log2 N T M≥L; using N T A beam set is constructed using M-bit binary sequences; wherein the beam set has M layers, each layer containing one beam, and the m-th position of the beam in the l-th layer is assigned the l-th bit of the M-bit binary sequence obtained by index m; m∈(0, 1, 2, ..., N) T -1), l∈(0, 1, 2, ..., M); the number of beamforming codewords is N. TThe invention generates a wide-beam hierarchical codebook corresponding to the beam set from a narrow-beam codebook. Using the wide-beam hierarchical codebook and a pre-built decoder, hierarchical beam training is performed on each user in the large-scale MIMO terahertz system to select the optimal beamforming codeword for each user from the narrow-beam codebook. This invention introduces channel coding into beam training based on a wide-beam hierarchical codebook, leveraging the coding gain of channel coding to improve data transmission reliability and reduce bit error rate, thereby enhancing the reliability of hierarchical beam training. Simultaneously, it utilizes the advantages of hierarchical beam training to ensure low pilot overhead, ultimately achieving more reliable beam training with lower pilot overhead and lower SNR. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0051] Figure 1 This is one of the flowcharts of the hierarchical beam training method based on channel coding provided by the present invention;

[0052] Figure 2 This is a schematic diagram comparing the reachability and rate of various beam training algorithms provided by this invention;

[0053] Figure 3 This is a schematic diagram of the structure of the channel coding-based hierarchical beam training device provided by the present invention;

[0054] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention;

[0055] Figure label:

[0056] 410: Processor; 420: Communication interface; 430: Memory; 440: Communication bus. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0058] The following is combined with Figures 1-4The present invention describes a channel-coded hierarchical beam training method and apparatus.

[0059] In a large-scale MIMO terahertz system, N T An antenna base station simultaneously serves K users with single antennas, and the spacing between the base station antennas is... λ is the carrier wavelength. f is the carrier frequency, and c is the speed of light.

[0060] Consider any user k, assuming its channel is... The Saleh-Valenzuela channel model, widely used in high-frequency bands such as terahertz, can generally be considered as a superposition of channel components from different paths. k It can be represented as:

[0061]

[0062] In the above formula, L k This represents the number of multipath channels (including line-of-sight and non-line-of-sight channels) between user k and the base station. and These are the complex gain and transmission angle of the l-th path between user k and the base station, respectively. The guiding vector is expressed as follows:

[0063]

[0064] Among them, spatial angle It can be defined as This refers to the user's physical orientation angle.

[0065] At terahertz frequencies, scattering can cause signal attenuation exceeding 20 dB, resulting in a small number of non-line-of-sight components. Line-of-sight components dominate data transmission in terahertz communication systems. Therefore, this invention primarily considers the channel for line-of-sight components.

[0066] Considering the downlink channel transmission from the base station to the user, assume that the signal transmitted by user k is a unity-power signal s. k Then the signal y received by user k k It can be represented as:

[0067] y k =h k w k s k +n k

[0068] In the above formula, w k Let n be the precoding matrix between user k and the base station. k Gaussian white noise

[0069] Without loss of generality, assume s k =1, then the beam training problem can be expressed as:

[0070]

[0071]

[0072] in, The number of beamforming codewords is N. T The narrow beam codebook is expressed as follows:

[0073]

[0074] Existing beam training methods can be broadly divided into two types. The first type is a beam training method based on narrow beam codebooks, which involves sequentially testing the codebook. N T One approach uses beamforming codewords to select the optimal beamforming codeword relative to the user, but its training overhead increases linearly with the number of antennas, leading to excessive overhead in large-scale MIMO systems. The second approach is beam training based on wide-beam codebooks, typically employing predefined hierarchical codebooks. The angular regions covered by codewords in the upper-layer codebook are divided into smaller angular regions covered by codewords in the lower-layer codebooks. However, the performance and overhead of hierarchical beam training depend on the design of the hierarchical codebooks. Existing hierarchical codebooks have low beam directional gain, are highly susceptible to noise at low signal-to-noise ratios, and exhibit poor beam training performance.

[0075] In view of this, the present invention provides a hierarchical beam training method based on channel coding, such as... Figure 1 As shown, this method is aimed at massive MIMO terahertz systems, where the base station of the massive MIMO terahertz system has N T Each user has a single antenna, primarily including:

[0076] S11: Divide the spatial angle [-1, 1] into N equal parts. T There are several angles and directions, numbered N in sequence. T Each angle direction is assigned from 0 to N. T -1 index;

[0077] Wherein, the angular coverage range of the angular direction with index m is D. m = [-1+m / 2] L-1 -1+(m+1) / 2 L-1 ], m∈(0, 1, 2, ..., N T -1), L=log2 N T .

[0078] S12: Using a pre-built encoder, encode the L-bit binary sequence represented by the index of each angle direction into an M-bit binary sequence; where M≥L;

[0079] For example: when L equals 4, the L-bit binary sequence at index 7 is represented as {0, 1, 1, 1}, and the L-bit binary sequence at index 13 is represented as {1, 1, 0, 1}.

[0080] It should be noted that the encoder can employ channel coding algorithms such as Hamming codes, polar codes, and convolutional codes.

[0081] S13: Utilizing N T A beam set is constructed using M-bit binary sequences; wherein the beam set has M layers, each layer containing one beam, and the m-th position of the beam in the l-th layer is assigned the l-th bit of the M-bit binary sequence obtained by index m; m∈(0, 1, 2, ..., N) T -1), l∈(0,1,2,…,M);

[0082] S14: The number of beamforming codewords is N. T The narrow beam codebook is used to generate a wide beam hierarchical codebook corresponding to the beam set;

[0083] S15: Using the wide-beam hierarchical codebook and the pre-built decoder, hierarchical beam training is performed on each user in the large-scale MIMO terahertz system to select the optimal beamforming codeword for each user in the narrow-beam codebook.

[0084] This invention provides a hierarchical beam training method based on channel coding, which models the hierarchical beam training problem as a channel coding and decoding problem. This allows the coding gain of channel coding to improve data transmission reliability and reduce bit error rate, thereby improving the reliability of hierarchical beam training. At the same time, the advantages of hierarchical beam training are used to ensure low pilot overhead, ultimately achieving more reliable beam training with lower pilot overhead and lower SNR.

[0085] Specifically, S12 includes:

[0086] S12.1: Let b be an L-bit binary sequence representing the angular direction index m. m Among them, b m =(b 1m b 2m , ..., b Lm ), b Lm For b m The Lth bit;

[0087] S12.2: The encoder's encoding algorithm, which converts the b...m Encoded as an M-bit binary sequence x m ;

[0088] The encoding algorithm of the encoder is as follows:

[0089] x = f en (b)

[0090] In the above formula, b is the input L-bit binary sequence, x is the output M-bit binary sequence, and f en (·) represents the encoding function of the encoder.

[0091] Polar coding is the coding method used in the control channel of the 5G NR standard, and it is the only coding method theoretically proven to reach the Shannon limit. Polar coding recursively polarizes the channel, resulting in different reliability levels for each sub-channel. By transmitting information on the "good" channels and freezing the "poor" channels, transmission reliability is improved. Taking the encoder as an example, which is a polar code encoder, the encoding algorithm of the encoder can be specifically written as follows:

[0092]

[0093] In the above formula, b is the input L-bit binary sequence, and x is the output M-bit binary sequence. It is a matrix composed of rows corresponding to the high-reliability channel indices selected from the generator matrix G using the Gaussian approximation algorithm. It is an M-order Kronecker factorial. One can conceive that... For N T Line 2 M A matrix of columns.

[0094] Specifically, this invention models the layered beam training problem as a channel coding and decoding problem. Since channel decoding has two methods, soft decoding and hard decoding, the wide-beam layered codebook generation and layered beam training of this invention differ for soft decoding and hard decoding.

[0095] In hardware decoding mode, S13 includes:

[0096] S13.1: Let the l-th layer beam in the beam set be B(l), and let the beam obtained by bit flipping B(l) be B. * (l);

[0097] S13.2: Define the number of layers in the wide-beam layered codebook as M, and the l-th layered codebook contains the beamforming codeword corresponding to beam B(l). and beam Corresponding beamforming codeword

[0098] S13.3: Utilizing the narrow beam codebook and the beam B(l) / the beam Generate the This leads to the generation of the wide-beam hierarchical codebook.

[0099] Furthermore, in S13.3, the generation of the narrow beam codebook and the beam B(l) is described. The process includes:

[0100] Based on the beam B(l), determine Corresponding beam coverage area

[0101] According to the above Determine that the beam coverage range in the narrow beam codebook is within the range of the specified beam coverage range. Beamforming codewords in the text;

[0102] The beam coverage area in the narrow beam codebook is within the range of The weighted superposition result of the beamforming codewords in the above is used as the

[0103] In detail, let's denote the codewords of the l-th layer (l∈{1, 2, ..., M}) layer codebook. The coverage area corresponding to the wide beam is According to the encoding scheme, It can be represented as:

[0104]

[0105] Where B(l, m) is the value assigned to the m-th position in beam B(l).

[0106] The coverage area of ​​a wide beam can be viewed as the union of the coverage areas of the underlying narrow beam; therefore, the wide beam codeword can be expressed as the underlying narrow beam codebook. Weighted summation of codewords w(n):

[0107]

[0108] Among them, ψ is introduced n To avoid low-gain ripples within the beam coverage area with additional degrees of freedom, ψ n It can be set to also, Let w(n) be the beam coverage area.

[0109] In practice, to facilitate comparison with other beam training schemes, it is usually necessary to... Normalization, i.e.

[0110] It is conceivable that the narrow beam codebook and the beam could be used... Generate the The process involves using the narrow beam codebook and the beam B(l) to generate the... The process is essentially the same, and will not be repeated here.

[0111] S14 includes:

[0112] S14.1: The control base station sequentially sends the following to each user: and the Make each of the users compare the received The signal amplitude and the received The signal amplitude, when receiving the The signal amplitude is greater than that of the received signal. The signal amplitude is recorded as 0 when receiving the signal. The signal amplitude is less than that of the receiver. The signal amplitude is recorded as 1, and the recorded 0 or 1 is regarded as the beam training result of each user under the l-th layer beam in the beam set;

[0113] S14.2: Control each user to feed back the codeword composed of the beam training results of each layer of beams in the beam set;

[0114] S14.3: Decode the codeword using the decoder to obtain the index of the angle direction of each user.

[0115] S14.4: Search for the index in the narrow beam codebook. Corresponding beamforming codewords and the This serves as the optimal beamforming codeword for each of the aforementioned users.

[0116] Taking the decoder as a polar code decoder as an example, the decoder can use the serial cancellation algorithm as the decoding algorithm.

[0117] In software decoding mode, S13 includes:

[0118] S13.A: The l-th layer beam in the beam set is denoted as B(l);

[0119] S13.B: Define the number of layers in the wide beam layered codebook as M+1, where the l-th layer of the first M layers contains the beamforming codeword corresponding to beam B(l). The final layer of the codebook is the narrow beam codebook;

[0120] S13.C: Generate the following using the narrow beam codebook and the beam B(l) This leads to the generation of the wide-beam hierarchical codebook.

[0121] That is, at this time, the wide-beam layered codebook It contains M+1 layers, and the number of layers in the hierarchical codebook is M (L = log2 N). T The code consists of an information bit layer, an ML redundancy check layer, and a bottom-level codebook. The bottom layer contains N codewords. T Narrow beam codebook The performance of beam training can be improved by utilizing the high gain of narrow beams. The other M layers are codebooks designed according to the coding algorithm, with each layer containing only one codeword.

[0122] Of course, in S13.C, the narrow beam codebook and the beam B(l) are used to generate the... The process in S13.3 involves using the narrow beam codebook and the beam B(l) to generate the... The process is exactly the same.

[0123] S14 includes:

[0124] S14.A: Control the base station to send to each of the users And enable each user to record and provide feedback. The power P(l) of the received signal;

[0125] S14.B: Based on P(l) and the decoder, calculate the index of the angular direction of each user.

[0126] S14.C: Control the base station to test the index in the narrow beam codebook for each user. and Beamforming code and Compare the data received by each user and stated The signal amplitude is determined, and the beamforming codeword with the larger signal amplitude is used as the optimal beamforming codeword for each user.

[0127] The goal of beam decoding is to achieve narrow beam codebooks The optimal beamforming codeword is selected. Based on the received signal power, it can be determined whether the user is using the codeword. beam coverage This allows the selection of codewords based on the decoding algorithm.

[0128] Similarly, taking the decoder as a polar code decoder as an example, the decoder can use the serial cancellation algorithm as the decoding algorithm to achieve decoding. The serial cancellation algorithm is designed based on the construction principle of polar codes. It eliminates the uncertainty of the information bits to be decoded one by one through an inorder traversal operation similar to a binary tree, thereby completing the decoding. The basis of the serial cancellation algorithm is to obtain an accurate log-likelihood ratio (LLR).

[0129] Therefore, S14.B can be broken down into two steps, as follows:

[0130] S14.B-1: Derive the above from P(l) log-likelihood ratio

[0131] S14.B-2: Utilization Together with the polar code decoder, the index of the angular direction of each user is calculated.

[0132] Among them, in S14.B-1

[0133]

[0134]

[0135] In the above formula, g l (Ω) is the beam gain of B(l), σ 2 Let p(UE∈B) be the noise power of the channel. l |P(l)=x) is when each of the users in Corresponding beam coverage area During this time, each of the users receives the The probability density distribution of the received signal. For when each of the aforementioned users is not Corresponding beam coverage area During this time, each of the users receives the The probability density distribution of the received signal is given by I0, where I0 is the modified 0th-order Bessel function.

[0136] Here, g represents the normalized codeword energy. l (Ω) can be represented as:

[0137]

[0138] Where Ω∈[-1, 1], This represents the beam coverage area.

[0139] In detail, this invention first obtains an accurate log-likelihood ratio (LLR) when the user is within the beam coverage area. Inside, the received signal is |g l (Ω)+n| 2 n is Gaussian white noise with a mean of 0 and a variance of σ. 2 Because the beam coverage of the codebook designed in this invention... Therefore, beam gain When the user is not within beam coverage area Inside, the received signal is |n| 2 It is evident that the received power does not follow the Gaussian channel model commonly used in polar codes, but rather a chi-square distribution. Therefore, the original polar code decoding algorithm cannot be directly applied, and the LLR expression needs to be re-derived for the beam training problem to make the decoding algorithm suitable for the beam training problem.

[0140] When the user is in / out of coverage area Within the range, the probability density distribution of the received signal is:

[0141]

[0142]

[0143] Where I0 is the modified 0th-order Bessel function. The LLR distribution derived from this is:

[0144]

[0145] After obtaining the LLR, beam decoding can be performed using a serial cancellation algorithm to calculate the angle index.

[0146] Finally, the base station tested the codebook for each user in turn. The index is and Beamforming code and The final selected beamforming codeword is obtained, thereby improving the beamforming codeword selection accuracy.

[0147] In soft decoding, this invention adaptively improves the LLR calculation method in the polar code decoding process to adapt to beam training.

[0148] Figure 2 This is a diagram comparing the reachability and rate of various beam training algorithms, such as... Figure 2 As shown, the beam training method proposed in this invention can effectively improve the reliability of beam training under low signal-to-noise ratio with the same pilot overhead.

[0149] Secondly, the channel-coding-based hierarchical beam training apparatus provided by the present invention will be described. The channel-coding-based hierarchical beam training apparatus described below can be referred to in correspondence with the channel-coding-based hierarchical beam training method described above. Figure 3 As shown, the device is designed for massive MIMO terahertz systems, where the base station of the massive MIMO terahertz system has N T The device includes one antenna and a user has a single antenna.

[0150] Index marker unit 21 is used to divide the spatial angle [-1, 1] into N equal parts. T There are several angles and directions, numbered N in sequence. T Each angle direction is assigned from 0 to N. T -1 index;

[0151] Encoding unit 22 is used to encode an L-bit binary sequence, represented by the index of each angular direction, into an M-bit binary sequence using a pre-built encoder; where L = log₂N T M≥L;

[0152] Beamforming unit 23, used to utilize N T A beam set is constructed using M-bit binary sequences; wherein the beam set has M layers, each layer containing one beam, and the m-th position of the beam in the l-th layer is assigned the l-th bit of the M-bit binary sequence obtained by index m; m∈(0, 1, 2, ..., N) T -1), l∈(0,1,2,…,M);

[0153] Wide-beam hierarchical codebook generation unit 24 is used for generating codebooks with N beamforming codewords. T The narrow beam codebook is used to generate a wide beam hierarchical codebook corresponding to the beam set;

[0154] The hierarchical beam training unit 25 is used to perform hierarchical beam training on each user in the massive MIMO terahertz system using the wide-beam hierarchical codebook and the pre-built decoder, so as to select the optimal beamforming codeword for each user in the narrow-beam codebook.

[0155] This invention provides a hierarchical beam training device based on channel coding.

[0156] By modeling the hierarchical beamforming training problem as a channel coding and decoding problem, the coding gain of channel coding can be used to improve the reliability of data transmission and reduce the bit error rate, thereby improving the reliability of hierarchical beamforming training. At the same time, the advantages of hierarchical beamforming training can be used to ensure low pilot overhead, ultimately achieving more reliable beamforming training with lower pilot overhead and lower SNR.

[0157] Thirdly, Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4 As shown, the electronic device may include: a processor 410, a communication interface 420, a memory 430, and a communication bus 440. The processor 410, communication interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logic instructions from the memory 430 to execute a channel-coding-based hierarchical beamforming training method. This method is designed for massive MIMO terahertz systems, where the base station of the massive MIMO terahertz system has N... T The method involves dividing the spatial angle [-1, 1] into N equal parts, where each user has a single antenna and the user has a single antenna. T There are several angles and directions, numbered N in sequence. T Each angle direction is assigned from 0 to N. T The index is -1; using a pre-built encoder, the L-bit binary sequence represented by the index of each angle direction is encoded into an M-bit binary sequence; where L = log2 N T M≥L; using N T A beam set is constructed using M-bit binary sequences; wherein the beam set has M layers, each layer containing one beam, and the m-th position of the beam in the l-th layer is assigned the l-th bit of the M-bit binary sequence obtained by index m; m∈(0, 1, 2, ..., N) T -1), l∈(0, 1, 2, ..., M); the number of beamforming codewords is N. TThe narrow-beam codebook is used to generate a wide-beam hierarchical codebook corresponding to the beam set. Using the wide-beam hierarchical codebook and a pre-built decoder, hierarchical beam training is performed on each user in the large-scale MIMO terahertz system to select the optimal beamforming codeword for each user from the narrow-beam codebook. Furthermore, the logic instructions in the memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0158] Fourthly, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the channel coding-based hierarchical beam training method provided by the above methods. This method is aimed at large-scale MIMO terahertz systems, where the base station of the large-scale MIMO terahertz system has N... T The method involves dividing the spatial angle [-1, 1] into N equal parts, where each user has a single antenna and the user has a single antenna. T There are several angles and directions, numbered N in sequence. T Each angle direction is assigned from 0 to N. T The index is -1; using a pre-built encoder, the L-bit binary sequence represented by the index of each angle direction is encoded into an M-bit binary sequence; where L = log2 N T M≥L; using N T A beam set is constructed using M-bit binary sequences; wherein the beam set has M layers, each layer containing one beam, and the m-th position of the beam in the l-th layer is assigned the l-th bit of the M-bit binary sequence obtained by index m; m∈(0, 1, 2, ..., N) T -1), l∈(0, 1, 2, ..., M); the number of beamforming codewords is N. TThe method generates a wide-beam hierarchical codebook corresponding to the beam set from a narrow-beam codebook. Using the wide-beam hierarchical codebook and a pre-built decoder, hierarchical beam training is performed on each user in the massive MIMO terahertz system to select the optimal beamforming codeword for each user from the narrow-beam codebook. Fifthly, the present invention also provides a non-transitory computer-readable storage medium storing a computer program that, when executed by a processor, implements the channel-coding-based hierarchical beam training method provided by the above methods. This method is for massive MIMO terahertz systems where the base station has N... T The method involves dividing the spatial angle [-1, 1] into N equal parts, where each user has a single antenna and the user has a single antenna. T There are several angles and directions, numbered N in sequence. T Each angle direction is assigned from 0 to N. T The index is -1; using a pre-built encoder, the L-bit binary sequence represented by the index of each angle direction is encoded into an M-bit binary sequence; where L = log2 N T M≥L; using N T A beam set is constructed using M-bit binary sequences; wherein the beam set has M layers, each layer containing one beam, and the m-th position of the beam in the l-th layer is assigned the l-th bit of the M-bit binary sequence obtained by index m; m∈(0, 1, 2, ..., N) T -1), l∈(0, 1, 2, ..., M); the number of beamforming codewords is N. T The narrow beam codebook is used to generate a wide beam hierarchical codebook corresponding to the beam set; using the wide beam hierarchical codebook and the pre-built decoder, hierarchical beam training is performed on each user in the large-scale MIMO terahertz system to select the optimal beamforming codeword for each user in the narrow beam codebook.

[0159] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0160] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0161] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A hierarchical beamforming training method based on channel coding, for massive MIMO terahertz systems, wherein the base station of the massive MIMO terahertz system has N T The feature is that there are one antenna and the user has a single antenna. The method includes: Divide the spatial angle [-1, 1] into N T angle directions, and give 0 to N T -1 index to the N T angle directions in turn; Using a pre-constructed encoder, each of the index of the angle direction is represented as an L-bit binary sequence to an M-bit binary sequence; wherein L=log2N T , M≥L; Utilizing N T bit binary sequences to construct a beam set; wherein the beam set has M layers, each layer contains one beam, the assignment of the mth position of the beam contained in the lth layer is the lth bit of the M-bit binary sequence obtained by the index m; m ∈ (0, 1, 2, …, N T -1), l ∈ (0, 1, 2, …, M). Based on beamforming codewords with a number of N T The narrow beam codebook is used to generate a wide beam hierarchical codebook corresponding to the beam set; Using the wide-beam hierarchical codebook and the pre-built decoder, hierarchical beam training is performed on each user in the large-scale MIMO terahertz system to select the optimal beamforming codeword for each user in the narrow-beam codebook.

2. The hierarchical beamforming training method based on channel coding according to claim 1, characterized in that, The step of encoding an L-bit binary sequence, represented by the index of each angular direction, into an M-bit binary sequence using a pre-built encoder includes: Let the index m of the angle direction be represented by an L-bit binary sequence denoted as b. m ; According to the encoder's encoding algorithm, the b m Encoded as an M-bit binary sequence x m ; The encoding algorithm of the encoder is as follows: x=f en (b) In the above formula, b is the input L-bit binary sequence, x is the output M-bit binary sequence, and f en (·) represents the encoding function of the encoder.

3. The hierarchical beamforming method based on channel coding according to claim 1, characterized in that, The narrow beam codebook The expression is as follows: in, The guiding vector can be represented as For spatial angles.

4. The hierarchical beamforming method based on channel coding according to claim 3, characterized in that, The number of beamforming codewords is N. T The narrow-beam codebook is used to generate a wide-beam hierarchical codebook corresponding to the beam set, including: Let the l-th layer beam in the beam set be denoted as B(l), and let the beam obtained by bit flipping B(l) be denoted as B. * (l); The wide-beam layered codebook is defined to have M layers, and the l-th layered codebook contains beamforming codewords corresponding to beam B(l). and beam Corresponding beamforming codeword Using the narrow beam codebook and the beam B(l) / the beam Generate the This leads to the generation of the wide-beam hierarchical codebook.

5. The hierarchical beamforming method based on channel coding according to claim 3, characterized in that, The number of beamforming codewords is N. T The narrow-beam codebook is used to generate a wide-beam hierarchical codebook corresponding to the beam set, including: Let the l-th layer beam in the beam set be denoted as B(l); The wide-beam layered codebook is defined as having M+1 layers, and the l-th layer of the first M layers contains the beamforming codeword corresponding to beam B(l). The final layer of the codebook is the narrow beam codebook; The narrow beam codebook and the beam B(l) are used to generate the... This leads to the generation of the wide-beam hierarchical codebook.

6. The hierarchical beamforming method based on channel coding according to claim 4 or 5, characterized in that, The method of generating the narrow beam codebook and the beam B(l) The process includes: Based on the beam B(l), determine Corresponding beam coverage area According to the above Determine that the beam coverage range in the narrow beam codebook is within the range of the specified beam coverage range. Beamforming codewords in the text; The beam coverage area in the narrow beam codebook is within the range of The weighted superposition result of the beamforming codewords in the above is used as the 7. The hierarchical beamforming method based on channel coding according to claim 4, characterized in that, The step of using the wide-beam hierarchical codebook and a pre-built decoder to perform hierarchical beamforming training on each user in the massive MIMO terahertz system, in order to select the optimal beamforming codeword for each user from the narrow-beam codebook, includes: The control base station sequentially sends the following to each user: and the Make each of the users compare the received The signal amplitude and the received The signal amplitude, when receiving the The signal amplitude is greater than that of the received signal. The signal amplitude is recorded as 0 when receiving the signal. The signal amplitude is less than that of the receiver. The signal amplitude is recorded as 1, and the recorded 0 or 1 is regarded as the beam training result of each user under the l-th layer beam in the beam set; Control each user to provide the codewords composed of the beam training results of each layer of beams in the beam set; The decoder is used to decode the codeword to obtain the index of the angle direction of each user. Search for the index in the narrow beam codebook Corresponding beamforming codewords and the This serves as the optimal beamforming codeword for each of the aforementioned users.

8. The hierarchical beamforming training method based on channel coding according to claim 5, characterized in that, The step of using the wide-beam hierarchical codebook and a pre-built decoder to perform hierarchical beamforming training on each user in the massive MIMO terahertz system, in order to select the optimal beamforming codeword for each user from the narrow-beam codebook, includes: The control base station sends to each of the users And enable each user to record and provide feedback. The power P(l) of the received signal; Based on P(l) and the decoder, the index of the angular direction of each user is calculated. The control base station tests the index in the narrow beam codebook for each user. and Beamforming code and Compare the data received by each user and stated The signal amplitude is determined, and the beamforming codeword with the larger signal amplitude is used as the optimal beamforming codeword for each user.

9. A channel-coding-based hierarchical beam training device for massive MIMO terahertz systems, wherein the base station of the massive MIMO terahertz system has N T The feature is that there are one antenna and the user has a single antenna. The device includes: Index marker units are used to divide the spatial angle [-1, 1] into N equal parts. T There are several angles and directions, numbered N in sequence. T Each angle direction is assigned from 0 to N. T -1 index; An encoding unit is used to encode an L-bit binary sequence, represented by the index of each angular direction, into an M-bit binary sequence using a pre-built encoder; where L = log₂N. T M≥L; Beamforming unit, used to utilize N T A beam set is constructed using M-bit binary sequences; wherein the beam set has M layers, each layer containing one beam, and the m-th position of the beam in the l-th layer is assigned the l-th bit of the M-bit binary sequence obtained by index m; m∈(0,1,2,…,N) T -1), l∈(0,1,2,…,M); Wide-beam hierarchical codebook generation unit, used for beamforming codewords with N characters. T The narrow beam codebook is used to generate a wide beam hierarchical codebook corresponding to the beam set; A hierarchical beam training unit is used to perform hierarchical beam training on each user in the massive MIMO terahertz system using the wide-beam hierarchical codebook and a pre-built decoder, so as to select the optimal beamforming codeword for each user in the narrow-beam codebook.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the channel-coded hierarchical beam training method as described in any one of claims 1 to 8.

Images