CN110401475A - Downlink beam training method, network equipment and terminal equipment - Google Patents

Abstract

The application discloses a downlink beam training method, network equipment and terminal equipment, which relate to the communication field and are used to reduce beam training overhead. A downlink beam training method includes: a terminal device receives a sounding beam from a network device, the sounding beam is modulated by an antenna weight vector, and the antenna weight vector includes a fixed codebook vector and a randomly generated random codebook vector; the terminal The device constructs an observation matrix according to the antenna weight vector; the terminal device performs signal reconstruction on a sparse vector of a beam search energy space through the observation matrix, and obtains beam pair information according to the sparse vector, wherein the beam pair The information includes sending end beam information; the terminal device sends the sending end beam information to the network device. The embodiment of the present application is applied to beam training.

Description

Downlink beam training method, network equipment and terminal equipment

technical field

The present application relates to the communication field, and in particular to a downlink beam training method, network equipment and terminal equipment.

Background technique

The 5th generation new radio (5G NR) technology supports millimeter-wave high-frequency communication. Millimeter-wave communication has outstanding advantages such as high transmission rate, high security, easy large-scale integration, and abundant license-free spectrum resources. However, high-frequency communication also brings high path loss. In order to compensate for the high path loss caused by high-frequency point communication, millimeter-wave communication uses antenna arrays and beamforming technology for directional communication to increase channel gain. Since the orientation of the transceiver device is unknown, beam training is required to complete directional beam alignment before communication. The purpose of beam training is to identify the optimal beam pair from all beam pairs and form a directional communication link under the given antenna weight vector (AWV).

In the initial access phase of 5G NR, beam management is required for multiple users at the same time. The beam training overhead is too large, the scanning time is long, and the initial access speed is slow. In the existing beam training technology, the training process between beam training users based on exhaustive search is independent of each other, and synchronous training can be implemented for multiple users, but the training overhead of a single user is too large; the search based on hierarchical feedback can reduce The training overhead of a single user, but the training between users is not independent. In a multi-user scenario, the training overhead will increase sharply with the increase in the number of users.

Contents of the invention

Embodiments of the present application provide a downlink beam training method, a network device, and a terminal device, so as to reduce beam training overhead.

In the first aspect, a downlink beam training method is provided, the method includes: a terminal device receives a sounding beam from a network device, the sounding beam is modulated by an antenna weight vector, and the antenna weight vector includes a fixed codebook vector and a randomly generated random codebook Vector; the terminal device constructs an observation matrix according to the antenna weight vector; the terminal device performs signal reconstruction on the sparse vector of the beam search energy space through the observation matrix, and calculates the beam pair information according to the sparse vector, wherein the beam pair information includes the beam information of the transmitting end; The terminal device sends the foregoing sending end beam information to the network device. In the downlink beam training method provided by the embodiment of the present application, the network device sends the sounding beam modulated by the antenna weight vector including the fixed codebook vector and the random codebook vector, and the terminal device uses the random antenna weight vector and the multipath channel response model The observation matrix is constructed, and the sparse reconstruction algorithm in the compressed sensing theory is used. Through the observation matrix, the problem of exhaustively enumerating a large number of beam search energy spaces is changed to the solution of the sparse vector of the energy space, so the beam can be reduced. Training overhead. And through the introduction of the fixed codebook vector, compared with the completely random codebook vector, the transmission gain is improved, the performance of the algorithm in the low SNR environment is improved, and the anti-noise ability is enhanced.

In a possible implementation manner, the observation matrix is: Among them, b _r,i ＝W _r (W _r W _r ^H ) ^{- 1} u _r,i , b _t,i ＝W _t (W _t W _t ^H ) ^-1 u _t,i , vec() means the matrix becomes a vector, W _r is the antenna weight vector of the beam at the receiving end during actual transmission, W _t is the antenna weight vector of the beam at the sending end during actual transmission, is the i-th receiving antenna weight vector, u _t,i is the i-th transmitting antenna weight vector, () ^H means conjugate transpose, () ^-1 means matrix inverse. This embodiment provides a solution to the observation matrix.

In a possible implementation manner, the beam pair information includes optimal beam pair information, and the terminal device performs signal reconstruction on the sparse vector of the beam search energy space through the observation matrix, and calculates the beam pair information according to the sparse vector, including: Any one of the above formulas can make an accurate estimate of the position χ of the maximum value in each component of the sparse vector q of the observation matrix θ: or, or, Among them, θ _χ represents the χth column of the observation matrix θ, q=(θ ^H θ) ^-1 θ ^H h, Indicates to take the corresponding χ when the calculation result of the following formula is the minimum value, Indicates that the χ corresponding to the calculation result of the following formula is the maximum value, || || ₂ indicates the 2-norm, || || indicates the modulus, || || ² indicates the modulus square, and h is after vectorization The multipath channel response element of ; according to the following formula, the optimal beam pair information (k _opt , l _opt ) is obtained as: l _opt =χ/K _r , k _opt =χ-K _r l _opt , where K _r is the receiving The number of end beams, l _opt is the optimal beam sequence number of the transmitting end, and k _opt is the optimal receiving end beam sequence number. This embodiment provides a solution to the optimal beam pair.

In a possible implementation manner, the beam pair information further includes candidate beam pair information, and the method further includes: the terminal device sequentially makes an accurate estimate of the position χ of the maximum value among the remaining non-zero value components in q according to the formula, and obtains Candidate beam pair information (k _oth , l _oth ), where l _oth is the beam sequence number of the candidate sending end, and k _oth is the beam sequence number of the candidate receiving end. This embodiment provides a solution to the candidate beam pairs.

In a possible implementation manner, the beam information of the transmitting end includes a codebook vector corresponding to the beam of the transmitting end or a sequence number of the beam of the transmitting end. This implementation manner provides a possible implementation manner of beam information at the sending end.

In a possible implementation manner, the beam information of the sending end includes beam information of an optimal sending end and beam information of an alternative sending end. This implementation manner provides a possible implementation manner of beam information at the sending end.

In a possible implementation manner, the beam pair information further includes receiving end beam information, and the method further includes: the terminal device sends the receiving end beam information to the network device. This implementation manner provides a possible implementation manner of the beam pair information.

In the second aspect, a downlink beam training method is provided, including: a network device sends a sounding beam, the sounding beam is modulated by an antenna weight vector, the antenna weight vector includes a fixed codebook vector and a randomly generated random codebook vector, and the antenna weight vector uses The observation matrix is used to reconstruct the signal of the sparse vector of the beam search energy space, and the sparse vector is used to solve the beam pair information. The beam pair information includes the beam information of the sending end; the network device receives the beam information of the sending end from the terminal device . In the downlink beam training method provided by the embodiment of the present application, the network device sends the sounding beam modulated by the antenna weight vector including the fixed codebook vector and the random codebook vector, and the terminal device uses the random antenna weight vector and the multipath channel response model The observation matrix is constructed, and the sparse reconstruction algorithm in the compressed sensing theory is used. Through the observation matrix, the problem of exhaustively enumerating a large number of beam search energy spaces is changed to the solution of the sparse vector of the energy space, so the beam can be reduced. Training overhead. And through the introduction of the fixed codebook vector, compared with the completely random codebook vector, the transmission gain is improved, the performance of the algorithm in the low SNR environment is improved, and the anti-noise ability is enhanced.

In a possible implementation manner, the observation matrix is: Among them, b _r,i ＝W _r (W _r W _r ^H ) ^{- 1} u _r,i , b _t , _i ＝W _t (W _t W _t ^H ) ^-1 u _t,i , vec() means the matrix becomes a vector, W _r is the antenna weight vector of the beam at the receiving end during actual transmission, W _t is the antenna weight vector of the beam at the sending end during actual transmission, is the i-th receiving antenna weight vector, u _t,i is the i-th transmitting antenna weight vector, () ^H means conjugate transpose, () ^-1 means matrix inverse. This embodiment provides a solution to the observation matrix.

In a possible implementation manner, the beam pair information includes optimal beam pair information (k _opt , l _opt ): l _opt =χ/K _r , k _opt =χ-K _r l _opt , where K _r is the received The number of end beams, l _opt is the beam number of the optimal sending end, k _opt is the number of the optimal receiving end beam, χ is the position of the maximum value in each component of the sparse vector q of the observation matrix θ; χ is the An exact estimate of either term yields: or, or, Among them, θ _χ represents the χth column of the observation matrix θ, q=(θ ^H θ) ^-1 θ ^H h, Indicates to take the corresponding χ when the calculation result of the following formula is the minimum value, Indicates that the χ corresponding to the calculation result of the following formula is the maximum value, || || ₂ indicates the 2-norm, || || indicates the modulus, || || ² indicates the modulus square, and h is after vectorization The multipath channel response element of . This embodiment provides a solution to the optimal beam pair.

In a possible implementation manner, the beam pair information further includes candidate beam pair information (k _oth , _loth ), where l _oth is the beam sequence number of the candidate sending end, k _oth is the beam sequence number of the candidate receiving end, and the backup The selected beam pair information (k _oth , l _oth ) is obtained by accurately estimating the position χ of the maximum value among the remaining non-zero value components in q according to the above formula. This embodiment provides a solution to the candidate beam pairs.

In a possible implementation manner, the beam information of the transmitting end includes a codebook vector corresponding to the beam of the transmitting end or a sequence number of the beam of the transmitting end. This implementation manner provides a possible implementation manner of beam information at the sending end.

In a possible implementation manner, the beam information of the sending end includes beam information of an optimal sending end and beam information of an alternative sending end. This implementation manner provides a possible implementation manner of beam information at the sending end.

In a possible implementation manner, the beam pair information further includes receiving end beam information, and the method further includes: the network device receives the above receiving end beam information from the terminal device. This implementation manner provides a possible implementation manner of the beam pair information.

In a third aspect, a terminal device is provided, including: a receiving unit, configured to receive a sounding beam from a network device, the sounding beam is modulated by an antenna weight vector, and the antenna weight vector includes a fixed codebook vector and a randomly generated random codebook vector The construction unit is used to construct the observation matrix from the antenna weight vector received by the receiving unit; the solution unit is used to construct the observation matrix constructed by the unit to perform signal reconstruction on the sparse vector of the beam search energy space, and solve the beam pair information according to the sparse vector, Wherein, the beam pair information includes the beam information of the sending end; the sending unit is configured to send the beam information of the sending end solved by the solving unit to the network device. Based on the same inventive concept, since the principle and beneficial effect of the terminal device to solve the problem can refer to the above-mentioned first aspect and the beneficial effects brought by various possible implementations of the first aspect, the implementation of the terminal device can refer to the above-mentioned first aspect. Various possible implementation manners of the aspect and the first aspect, the overlapping parts will not be repeated.

In a fourth aspect, a network device is provided, including: the network device sends a sounding beam, the sounding beam is modulated by an antenna weight vector, the antenna weight vector includes a fixed codebook vector and a randomly generated random codebook vector, and the antenna weight vector is used to construct Observation matrix. The observation matrix is used to reconstruct the signal of the sparse vector in the beam search energy space. The sparse vector is used to solve the beam pair information. The beam pair information includes the beam information of the sending end; the network device receives the beam information of the sending end from the terminal device. Based on the same inventive concept, since the problem-solving principle and beneficial effects of the network device can refer to the above-mentioned second aspect and the beneficial effects brought by various possible implementations of the second aspect, the implementation of the network device can refer to the above-mentioned second aspect. Various possible implementation manners of the aspect and the second aspect, the overlapping parts will not be repeated.

A fifth aspect provides a communication system, including the terminal device described in the third aspect and the network device described in the fourth aspect.

In a sixth aspect, an embodiment of the present application provides a terminal device, including: a processor and a memory, the memory is used to store a program, and the processor invokes the program stored in the memory to execute the method described in any one of the above first aspects.

In a seventh aspect, an embodiment of the present application provides a storage medium on which a computer program is stored, and when the computer program is executed by a processor, the method described in any one of the above-mentioned first aspects is implemented.

In an eighth aspect, an embodiment of the present application provides a chip system, including: a processor, configured to support a terminal device in implementing the method described in any one of the foregoing first aspects.

In a ninth aspect, an embodiment of the present application provides a network device, including: a processor and a memory, the memory is used to store a program, and the processor invokes the program stored in the memory to execute the method described in any one of the above second aspects.

In a tenth aspect, an embodiment of the present application provides a storage medium on which a computer program is stored, and when the computer program is executed by a processor, the method described in any one of the above-mentioned second aspects is implemented.

In an eleventh aspect, an embodiment of the present application provides a chip system, including: a processor, configured to support a network device to implement the method described in any one of the foregoing second aspects.

For the technical effects of the fifth aspect to the eleventh aspect, reference may be made to the content described in the first aspect and the second aspect.

Description of drawings

FIG. 1 is a schematic structural diagram of a communication system provided by an embodiment of the present application;

FIG. 2 is a first structural schematic diagram of a terminal device provided in an embodiment of the present application;

FIG. 3 is a schematic structural diagram of a network device provided in an embodiment of the present application;

FIG. 4 is a schematic flowchart of a downlink beam training method provided in an embodiment of the present application;

FIG. 5 is a schematic diagram of a sounding beam corresponding to a random codebook vector generated by three Bernoulli random sequences provided in an embodiment of the present application;

FIG. 6 is a schematic diagram of the energy distribution of the search space E provided by the embodiment of the present application;

7 is a schematic diagram of a simulation of the ratio of the channel capacity to the ideal value under different schemes provided by the embodiment of the present application;

FIG. 8 is a schematic diagram of a simulation of a cell scenario provided in an embodiment of the present application;

FIG. 9 is a second structural schematic diagram of a terminal device provided in an embodiment of the present application;

FIG. 10 is a third structural schematic diagram of a terminal device provided by an embodiment of the present application;

FIG. 11 is a fourth schematic structural diagram of a terminal device provided in an embodiment of the present application;

FIG. 12 is a second schematic structural diagram of a network device provided by an embodiment of the present application;

FIG. 13 is a schematic structural diagram III of a network device provided in an embodiment of the present application;

FIG. 14 is a fourth structural schematic diagram of a network device provided by an embodiment of the present application.

Detailed ways

The network architecture and business scenarios described in the embodiments of the present application are for more clearly illustrating the technical solutions of the embodiments of the present application, and do not constitute limitations on the technical solutions provided by the embodiments of the present application. For the evolution of architecture and the emergence of new business scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

The embodiment of the present application may be applied to a scenario of time division duplexing (time division duplexing, TDD), and may also be applicable to a scenario of frequency division duplexing (frequency division duplexing, FDD). The technical solutions provided in this application can be applied to 5G NR systems.

It should be noted that although the embodiments of this application are described based on the scenario of a 5G NR network in a wireless communication network, it should be pointed out that the solutions in this embodiment of this application can also be applied to other wireless communication networks, and the corresponding names can also be Replace it with the name of the corresponding function in other wireless communication networks.

An embodiment of the present application provides a communication system, as shown in FIG. 1 , including at least one terminal (terminal) device 11 and a network device 12 .

Optionally, the terminal device 11 involved in the embodiment of the present application may include various handheld devices with wireless communication functions, vehicle-mounted devices, wearable devices, computing devices or other processing devices connected to wireless modems; it may also include Subscriber unit (subscriber unit), cellular phone (cellular phone), smart phone (smartphone), wireless data card, personal digital assistant (personal digital assistant, PDA) computer, tablet computer, wireless modem (modem), handheld device (handheld ), laptop computer (laptop computer), cordless phone (cordless phone) or wireless local loop (wireless local loop, WLL) station, machine type communication (machine type communication, MTC) terminal, user equipment (user equipment, UE ), a mobile station (mobile station, MS), a terminal device (terminal device) or a relay user equipment, etc. Wherein, the relay user equipment may be, for example, a 5G residential gateway (residential gateway, RG). For convenience of description, in this application, the above-mentioned devices are collectively referred to as terminal devices.

Taking the terminal device 11 as a mobile phone as an example, the general hardware architecture of the mobile phone will be described. As shown in Figure 2, the mobile phone may include: a radio frequency (radio frequency, RF) circuit 110, a memory 120, other input devices 130, a display screen 140, a sensor 150, an audio circuit 160, an I/O subsystem 170, a processor 180, And parts such as power supply 190. Those skilled in the art can understand that the structure of the mobile phone shown in the figure does not constitute a limitation to the mobile phone, and may include more or less components than shown in the figure, or combine certain components, or split certain components, or Different component arrangements. Those skilled in the art can understand that the display screen 140 belongs to a user interface (user interface, UI), and the display screen 140 may include a display panel 141 and a touch panel 142 . Although not shown, the mobile phone may also include functional modules or devices such as a camera and a Bluetooth module, which will not be repeated here.

Further, the processor 180 is respectively connected to the RF circuit 110 , the memory 120 , the audio circuit 160 , the I/O subsystem 170 , and the power supply 190 . The I/O subsystem 170 is connected to other input devices 130 , the display screen 140 and the sensor 150 respectively. Among them, the RF circuit 110 can be used for receiving and sending signals during sending and receiving information or talking, especially, after receiving downlink information from the base station, send it to the processor 180 for processing. The memory 120 can be used to store software programs as well as modules. The processor 180 executes various functional applications and data processing of the mobile phone by running the software programs and modules stored in the memory 120, such as executing the methods and functions of the terminal device in the embodiments of the present application. Other input devices 130 can be used to receive input numbers or character information, and generate keyboard signal input related to user settings and function control of the mobile phone. The display screen 140 can be used to display information input by or provided to the user and various menus of the mobile phone, and can also accept user input. Sensor 150 may be a light sensor, motion sensor, or other sensor. Audio circuitry 160 may provide an audio interface between the user and the handset. The I/O subsystem 170 is used to control input and output external devices, and the external devices may include other device input controllers, sensor controllers, and display controllers. The processor 180 is the control center of the mobile phone 200, using various interfaces and lines to connect various parts of the entire mobile phone, by running or executing software programs and/or modules stored in the memory 120, and calling data stored in the memory 120, Execute various functions of the mobile phone 200 and process data, thereby monitoring the mobile phone as a whole. The power supply 190 (such as a battery) is used to supply power to the above-mentioned components. Preferably, the power supply can be logically connected to the processor 180 through the power management system, so that functions such as charging, discharging, and power consumption can be managed through the power management system.

Optionally, the network device 12 involved in this embodiment of the present application may be a base station, and the general hardware architecture of the base station will be described. As shown in FIG. 3 , the base station 12 may include an indoor baseband processing unit (building baseband unit, BBU) 1201 and a remote radio frequency module (remote radio unit, RRU) 1202, and the RRU 1202 is connected to an antenna feeder system (ie, an antenna) 1203, and the BBU 1201 and RRU 1202 can be disassembled and used as required. The antenna 1203 can be an antenna array, including multiple sub-antennas, that is, array elements. The base station can convert the signal processing in the antenna domain into a beam domain signal through the codebook, and control the direction angles of these array elements. In this application, the base station can control some The array element is modulated by a fixed codebook, and the other part is modulated by a random codebook. The base station may include various forms of base stations, for example: macro base stations, micro base stations (also called small stations), relay stations, access points, and the like. The base station can execute the method and function of the network device in the embodiments of the present application.

The embodiments of the present application can be applied to beam training during initial fast access of a 5G NR network and fast recovery after beam failure. During the initial access, the fast beam training of the embodiment of the present application realizes the rapid access of the device; when the link is interrupted and the beam fails due to reasons such as antenna position, angle movement, and obstacle occlusion, the embodiment of the present application can also be applied Fast beam training enables fast recovery of beams. At the same time, in the embodiment of the present application, multiple sets of beam pairs including the optimal beam pair can be simultaneously obtained, and the group with the largest SNR can be selected as the transmission beam pair according to the base station strategy, and the other beam pairs can be used as candidate beam pairs.

The embodiment of the present application provides a beam training method, which is applied to the above system, as shown in FIG. 4, the method includes:

S101. The network device sends a detection beam.

The sounding beam is modulated by the antenna weight vector, which includes a fixed codebook vector and a random codebook vector.

For example, suppose one antenna weight vector for the sounding beam is Where s1 is a fixed codebook vector, r1 is a random codebook vector, if the antenna domain signal to send the sounding beam is X=[x1 x2], then the beam domain signal that the network device finally sends the sounding beam to the air interface is: In other words, Y is the transmitted signal after beamforming (antenna modulation).

The radio frequency signal modulated by the codebook is a directional shaped beam. The number of codebook vectors is the same as the number of array elements, for example, if there are 4 fixed codebook vectors, 4 array elements are required to send fixed codebook vectors, and if there are 8 random codebook vectors, 8 array elements are required to send random codes This vector.

Regarding the ratio of the number of fixed codebook vectors to random codebook vectors, it is found in the simulation that if the number of fixed codebooks is too small, the antenna weight vector will not show directionality; if the number of fixed codebooks is too large, the codebook will be affected randomness, resulting in poor final signal reconstruction effect. Therefore, the number of fixed codebook vectors should be adjusted according to the specific communication environment. For example, if the number of antennas at the transceiver end is 16, the ratio of the number of codebooks in the random codebook vector to the fixed codebook vector is 1:1, and the first 8 antennas send the sounding beams pointing to the fixed codebook vector in the 90° direction, and the fixed The codebook vector is [1,-1,1,-1,1,-1,1,-1], and the last 8 antennas send probing beams of random codebook vectors, which are generated by Bernoulli random sequences .

The fixed codebook vector and the random codebook vector are described in detail below:

For fixed codebook vectors:

The main lobe direction of the fixed codebook vector selects the center direction of the coverage area of the network equipment (such as a base station) as much as possible; if it is not enough to cover the entire coverage area of the entire network equipment, multiple fixed codebook vectors with different coverage directions can be designed, The beam training is carried out in a time-divided and direction-divided manner by scanning.

Optionally, consecutive N array elements can be selected for transmitting a fixed codebook, and its position in the entire antenna can be the beginning or end of the continuous N array elements, or the continuous N array elements at any position in the middle Yuan.

The fixed codebook vector may adopt fixed codebook design methods such as discrete Fourier transform (DFT), Floor, and Beam, or other design methods. Take the DFT fixed codebook design method as an example to illustrate:

Wherein, M is the number of array elements using a fixed codebook, u _m-1 is the fixed codebook vector of the mth column of U, m=0,1,...,M-1,

For random codebook vectors:

In the antenna array, except that some array elements can use fixed codebook vectors, the remaining array elements can use random codebook vectors. There are many ways to generate random codebook vectors, such as Bernoulli random codebook, [1,-1,0] multi-valued random codebook, etc. Figure 5 shows the detection beams corresponding to the random codebook vectors generated by three Bernoulli random sequences. The random codebook vector corresponding to (a) in Figure 5 is [1,1,1,-1,-1,1,-1,1], and the random codebook vector corresponding to (b) in Figure 5 is [- 1,1,-1,1,-1,1,-1,-1], the random codebook vector corresponding to (c) in Figure 5 is [1,-1,1,1,-1,-1, -1,1].

S102. The terminal device receives the sounding beam from the network device, and constructs an observation matrix θ according to the antenna weight vector.

Assuming that φ _tl is the transmission angle of the physical transmission channel, φ _rl is the arrival angle of the physical transmission channel, M is the number of antennas at the transmitting end, N is the number of antennas at the receiving end, L is the number of multipath channels, λ _l is the channel coefficient, ( ) ^H is the conjugate transpose symbol, g _l is the receiving azimuth vector, p _l is the sending azimuth vector. Then the multipath channel propagation model of millimeter wave is:

in, is a simplification of H, g _l and p _l are:

In the process of training the sounding beam, the transmitting end of the network device sends the antenna domain signal as a training pilot to assist training. The training pilot is composed of an energy-normalized sequence x=[x ₁ ,x ₂ ,...,x _M ] ^T , where M is the number of antennas at the transmitting end. According to the multipath channel response model, the received symbol y _i of the i-th pilot sequence _xi can be expressed as:

Among them, u _t,i is the transmitting antenna weight vector of the i-th pilot sequence x _i , u _r,i is the receiving antenna weight vector of the i-th pilot sequence x _i , M is the number of transmitting antennas, γ is the transmission Signal SNR, n _i is additive Gaussian white noise. It should be noted that if the terminal device uses an omnidirectional antenna to receive, the power consumption will be too large and the effect will be poor. Therefore, the receiving antenna of the terminal device also has directionality, and its receiving direction is determined by the receiving antenna weight vector u _r,i .

In the multipath channel response model, the multipath channel response element h _i (l) of the l-th path of the i-th pilot (that is, the coefficient before x _i in formula (4)) can be described as:

If the i-th pilot sequence x _i is transmitted, the receiving antenna weight vector of the k-th receiving beam is u _r,i , the transmitting antenna weight vector of the l-th transmitting beam is u _t,i , and the number of antennas at the receiving end is N , then the SNR of the receiving end of the terminal device can be expressed as formula (6):

The goal of beam training is to find the beam pair information (k _opt , l _opt ) with the highest channel SNR satisfying formula (6) from all possible beam pair information, where k _opt is the beam search space E (for example, in Figure 6 As shown in ), the optimal receiving end beam sequence number of the terminal device side, l _opt is the optimal transmitting end beam sequence number of the network device side in the beam search space E. which is:

(k _opt ,l _opt )=argmaxSNR(k,l) (7)

It can be seen from the formula (6) that the SNR is the square mean value of h _i (l), and the larger the SNR is, the larger h is actually.

If the optimal beam pair is searched exhaustively for a beam search space with M antennas at the transmitting end and N antennas at the receiving end, the cost is ξ=MN. According to the core idea of compressive sensing theory, the observation matrix can be constructed through the above-mentioned random antenna weight vector, and the exhaustive search problem in the beam search space E can be transformed into the solution problem of the sparse vector q through the observation matrix. Specifically, the observation matrix θ can be constructed by formulas (8)-(10).

First, in order to find the beam search space E conveniently, define the vectors e _r and e _t :

Among them, W _r is the antenna weight vector of the beam at the receiving end during actual transmission, W _t is the antenna weight vector of the beam at the sending end during actual transmission, is the channel gain coefficient, γ is the transmission signal SNR, N is the number of antennas at the receiving end, and λ _l is the channel coefficient of the lth transmit beam. W _r and W _t have been stored in network devices and terminal devices before beam training. Then according to the least square estimation (least square, LS), we can get:

() ^-1 means matrix inverse. Then put formula (9) into formula (5), the multipath channel response elements of all the paths (regardless of l) of the ith pilot in the multipath channel response model are:

Formula (10) can be expressed by vectorization:

Then θ is the constructed observation matrix its i-th behavior Among them, g is the receiving azimuth vector, p is the sending azimuth vector, is the i-th receive antenna weight vector, u _ti is the i-th transmit antenna weight vector, b _r,i =W _r (W _r W _r ^H ) ^-1 u _r,i , b _t,i =W _t (W _t W _t ^H ) ^-1 u _t,i , E is the exhaustive beam search energy space, vec() means turning the matrix into a vector, q=vec(E).

In the expression, q is a vector formed by the millimeter-wave beam search energy space E and has significant sparsity. Each non-zero element of q represents a set of available beam pairs. According to formula (11) q and h are linearly related, and according to formula (6), the larger the SNR is, the larger h is, so the value of q represents the beam pair. The SNR channel gain strength of q, the position with the largest value in q corresponds to the optimal beam pair information (k _opt , l _opt ), and the positions of other non-zero values in q correspond to other available candidate beam pair information (k _oth , _loth ).

S103. The terminal device performs signal reconstruction on the sparse vector q of the beam search energy space E through the observation matrix, and obtains beam pair information according to the sparse vector q.

After the terminal device obtains the observation matrix, it can use the sparse reconstruction algorithm of compressed sensing principle, such as matching pursuit (MP), orthogonal matching pursuit (OMP), basis pursuit (basis pursuit, BP), noise reduction basis pursuit (basis pursuit de-noising, BPD), etc., to solve the sparsest solution of q, for example, using the matching pursuit algorithm to solve the optimal beam pair information method is as follows:

According to the least square estimation of formula (11), the sparse vector q of the observation matrix θ can be obtained:

q＝(θ ^H θ) ^-1 θ ^H h (12)

Among them, h is the multipath channel response element.

Since the vector q is a sparse vector, and the matrix θ has significant randomness, according to the theory of compressed sensing, we can use the minimum square error to calculate the position χ of the maximum value in each component of the sparse vector q according to any of the following formulas Precise estimate:

or,

or,

Among them, θ _χ represents the χth column of the observation matrix θ, q=(θ ^H θ) ^-1 θ ^H h, Indicates to take the corresponding χ when the calculation result of the following formula is the minimum value, Indicates that the χ corresponding to the calculation result of the following formula is the maximum value, || || ₂ indicates the 2-norm, || || indicates the modulus, || || ² indicates the modulus square, and h is after vectorization The multipath channel response element of .

Therefore, the optimal beam pair information (k _opt , l _opt ) is obtained:

Among them, K _r is the number of beams at the receiving end, l _opt is the beam number of the optimal sending end, and k _opt is the beam number of the optimal receiving end.

After obtaining the optimal beam pair information (k _opt , l _opt ), the terminal device can make an accurate estimate of the maximum position χ of the remaining non-zero value components in q according to formula (13) to formula (16), The candidate beam pair information (k _oth , l _oth ) is obtained, where l _oth is the beam sequence number of the candidate transmitting end, and k _oth is the beam sequence number of the candidate receiving end.

S104. The terminal device sends the above beam information of the sending end to the network device.

As the sending end, the network device can only care about which sending end beam the local end uses to communicate with the terminal device, so the terminal device can only feed back the sending end beam information to the network device. Optionally, the beam information at the transmitting end may include a codebook vector corresponding to the beam at the transmitting end or a sequence number of the beam at the transmitting end. The beam information at the sending end may further include link channel amplitude information of the beam at the sending end.

Optionally, the beam information of the transmitter may include the beam information of the optimal transmitter. Therefore, the beam information of the optimal transmitter may include a codebook vector corresponding to the beam of the optimal transmitter or a serial number l _opt of the optimal beam of the transmitter. The beam information of the transmitting end may also include the beam information of the alternative transmitting end. Therefore, the beam information of the transmitting end may include the codebook vector corresponding to the beam of the transmitting end or the sequence number _loth of the beam of the transmitting end.

Optionally, the terminal device may also send the receiving end beam information to the network device, and the receiving end beam information may include a codebook vector corresponding to the receiving end beam or a receiving end beam sequence number. The receiving end beam information and the transmitting end beam information may be collectively referred to as beam pair information.

Optionally, the receiving end beam information may include optimal receiving end beam information, so the optimal receiving end beam information may include a codebook vector corresponding to the optimal receiving end beam or an optimal receiving end beam sequence number k _opt . The receiving end beam information may also include candidate receiving end beam information, so the candidate receiving end beam information may include a codebook vector corresponding to the candidate receiving end beam or a sequence number _loth of the candidate receiving end beam.

That is to say, the terminal device may also send the optimal beam pair information to the network device. The optimal beam pair information may include a codebook vector corresponding to the optimal beam pair or a sequence number (k _opt , l _opt ) of the optimal beam pair. The terminal device may also send candidate beam pair information to the network device, where the candidate beam pair information may include a codebook vector corresponding to the candidate beam pair or a sequence number (k _oth , _loth ) of the candidate beam pair. This enables the terminal device and the network device to switch to an alternative beam pair when the signal deteriorates due to factors such as occlusion during the communication process through the optimal beam pair.

S105. The network device and the terminal device communicate by using the beam pair.

According to the strategy, the two parties can select the optimal beam pair (k _opt , l _opt ) as the actual transmission beam pair, or select an alternate beam pair as the actual transmission beam pair.

In the downlink beam training method provided by the embodiment of the present application, the network device sends the sounding beam modulated by the antenna weight vector including the fixed codebook vector and the random codebook vector, and the terminal device uses the random antenna weight vector and the multipath channel response model The observation matrix is constructed, and the sparse reconstruction algorithm in the compressed sensing theory is used. Through the observation matrix, the problem of exhaustively enumerating a large number of beam search energy spaces is changed to the solution of the sparse vector of the energy space, so the beam can be reduced. Training overhead. And through the introduction of the fixed codebook vector, compared with the completely random codebook vector, the transmission gain is improved, the performance of the algorithm in the low SNR environment is improved, and the anti-noise ability is enhanced.

Simulation results

As shown in Figure 6, if a regular linear array (uniform linear array, ULA) antenna is used, the number of antenna elements at the transceiver end is 16, the number of beams at the transceiver end is 32, and when the fixed codebook is a DFT codebook, each beam pair The distribution diagram of SNR at the receiving end is equivalent to the energy distribution diagram of the search space E in formula (11). It can be seen that the energy of the search space is relatively concentrated and has strong sparsity. The goal of beam training is to find out the optimal SNR in the channel formed by each beam pair. The optimal beam pair in this scenario is (23,3).

If a ULA antenna array is used, the distance between the array elements is λ/2, the radio frequency carrier is 28 GHz, the antennas at the receiving end are N=8, and the antennas at the transmitting end are M=64. The fixed codebook for directional transmission at the transceiver end is a DFT codebook with a 90° main lobe direction, such as [1,-1,1,-1,…], and the random codebook is generated by a Bernoulli random sequence. The multipath channel includes 1 direct path and 4 reflection paths. The sparse reconstruction algorithm uses orthogonal matching pursuit. In the simulation, the number of fixed codebooks is selected as 1/2 and 1/4 of the total number of antennas for simulation. The position of the fixed codebook is selected from the first 1/2 or 1/4 of the antennas. The channel is noise-free and exhaustive search is used. The communication channel capacity of the optimal beam pair is the theoretical upper limit. The simulation results are shown in Figure 7: it can be seen that the compressed sensing scheme of this application can basically approach the upper limit of performance in the case of high SNR, with only about 2-3% performance loss. Compared with the completely random compressed sensing scheme, the performance of the scheme of the present application is greatly improved under low transmission signal SNR, and the anti-noise ability of beam training is enhanced.

The simulation of the cell scene is shown in Figure 8, where the transmission SNR at a distance of 100m from the base station is set to be 10dB, and the abscissa is the furthest distance from the cell user to the base station. It can be seen that when the coverage range is about 800m, the performance of the compressed sensing training scheme reaches 90% of the exhaustive search.

The training cost of exhaustive search is: ξ=MN, the cost of this application scheme is: ξ=α(log ₂ (MN)), α is an empirical coefficient, and the simulation results show that it is generally 2.5-3. The comparison of the two overheads is shown in Table 1. In this table, the empirical coefficient α=3. It can be seen from this that the application scheme can greatly reduce the overhead of beam training, and with the increase in the number of antennas at the transceiver end, the overhead advantage is more obvious.

Table 1

MN 64*8 128*8 128*16 128*32 256*32 This program 27 30 33 36 39 exhaustive search 512 1024 2048 4096 8192 ratio 5.2% 2.9% 1.6% 0.88% 0.11%

The present application provides a terminal device configured to execute the above method. The embodiment of the present application may divide the terminal device into functional modules according to the above method example, for example, each functional module may be divided corresponding to each function, or two or more functions may be integrated into one processing module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software function modules. It should be noted that the division of modules in this application is schematic, and is only a logical function division, and there may be other division methods in actual implementation.

In the case of dividing each functional module corresponding to each function, FIG. 9 shows a possible structural diagram of the terminal device involved in the above embodiment. The terminal device 11 includes: a receiving unit 1111, a construction unit 1112, and a solving unit 1113. The sending unit 1114. Each of the above units is used to support the terminal device to execute the related method in any one of the drawings in FIG. 4 . The terminal device provided in this application is used to execute the corresponding method provided above. Therefore, its corresponding features and beneficial effects can be referred to the beneficial effects of the corresponding method provided above, and will not be repeated here repeat.

Exemplarily, the receiving unit 1111 is used to support the terminal device 11 to execute the process S102 in FIG. 4; the construction unit 1112 is used to support the terminal device 11 to execute the process S102 in FIG. Process S103 in FIG. 4 ; the sending unit 1114 is used to support the terminal device 11 in executing the process S104 in FIG. 4 . Wherein, all relevant content of each step involved in the above-mentioned method embodiment can be referred to the function description of the corresponding function module, and will not be repeated here.

In the case of using an integrated unit, FIG. 10 shows a possible structural diagram of the terminal device involved in the above embodiment. The terminal device 11 includes: a storage module 1121 , a processing module 1122 , and a communication module 1123 . The above modules are used to support the terminal device to execute the related method in any one of the drawings in FIG. 4 . The terminal device provided in this application is used to execute the corresponding method provided above. Therefore, its corresponding features and beneficial effects can be referred to the beneficial effects of the corresponding method provided above, and will not be repeated here repeat.

Specifically, the processing module 1122 is configured to control and manage actions of the terminal device 11 . The communication module 1123 is configured to support the terminal device 11 to execute the above functions of the receiving unit 1111 and the sending unit 1112 . The storage module 1121 is used for storing program codes and data of the terminal device.

Wherein, the processing module 1122 may be a processor or a controller, such as a central processing unit (central processing unit, CPU), a general processor, a digital signal processor (digital signal processor, DSP), an application-specific integrated circuit (application-specific integrated circuit, ASIC), field programmable gate array (field programmable gate array, FPGA) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It can implement or execute the various illustrative logical blocks, modules and circuits described in connection with the present disclosure. The processor may also be a combination of computing functions, for example, a combination of one or more microprocessors, a combination of DSP and a microprocessor, and so on. The communication module 1123 may be a transceiver, a transceiver circuit, Bluetooth, a network interface or a communication interface, and the like. The storage module 1121 may be a memory.

Specifically, the processing module 1122 may be the processor 180 in FIG. 2 , the communication module 1123 may be the RF circuit 110 in FIG. 2 , and the storage module 1121 may be the memory 120 in FIG. 2 .

When the processing module 1122 is a processor, the communication module 1123 is an RF circuit, and the storage module 1121 is a memory, the terminal device involved in this application may be the terminal device 11 shown in FIG. 11 .

Referring to FIG. 11 , the terminal device 11 includes: one or more processors 1132 , an RF circuit 1133 , a memory 1131 , a bus system 1134 , and one or more programs. Wherein, the RF circuit 1133, the processor 1132, and the memory 1131 are connected to each other through a bus system 1134; the bus system 1134 may be a standard bus for interconnecting peripheral components or an extended industry standard structure bus. The bus can be divided into address bus, data bus, control bus and so on. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus. The one or more programs are stored in the memory 1131, and the one or more programs include instructions, which when executed by the terminal device cause the terminal device to execute the relevant methods in any one of the drawings in FIG. 4 .

The present application also provides a computer storage medium storing one or more programs, and the one or more programs include instructions, and the instructions, when executed by the terminal device, cause the terminal device to execute the related method in any one of the drawings in FIG. 4 .

The present application also provides a computer program product containing instructions. When the computer program product is run on the terminal device, the terminal device is made to execute the related method in any one of the drawings in FIG. 4 .

An embodiment of the present application provides a system-on-a-chip, where the system-on-a-chip includes a processor, configured to support a terminal device in implementing the foregoing method, for example, receive a sounding beam from a network device. In a possible design, the chip system further includes a memory. The memory is used to store necessary program instructions and data of the terminal equipment. Of course, the memory may not be in the system-on-a-chip. The chip system may include a chip, an integrated circuit, or may include a chip and other discrete devices, which are not specifically limited in this embodiment of the present application.

Among them, the terminal equipment, computer storage medium, computer program product or chip system provided in this application are all used to execute the corresponding method provided above, therefore, the beneficial effects it can achieve can refer to the corresponding method provided above The beneficial effects in the method will not be repeated here.

The present application provides a network device configured to execute the above method. The embodiment of the present application may divide the network device into functional modules according to the above method example, for example, each functional module may be divided corresponding to each function, or two or more functions may be integrated into one processing module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software function modules. It should be noted that the division of modules in this application is schematic, and is only a logical function division, and there may be other division methods in actual implementation.

In the case of dividing each functional module corresponding to each function, FIG. 12 shows a possible structural diagram of the network device involved in the above embodiment. The network device 12 includes: a sending unit 1211 and a receiving unit 1212 . Each of the above units is used to support the network device to execute the related method in any one of the drawings in FIG. 4 . The network device provided in this application is used to implement the corresponding method provided above, therefore, its corresponding features and beneficial effects can be referred to the beneficial effects of the corresponding method provided above, and will not be repeated here repeat.

Exemplarily, the sending unit 1211 is configured to support the network device 12 to execute the processes S101 and S105 in FIG. 4 ; the receiving unit 1212 is configured to support the network device 12 to execute the processes S104 and S105 in FIG. 4 . Wherein, all relevant content of each step involved in the above-mentioned method embodiment can be referred to the function description of the corresponding function module, and will not be repeated here.

In the case of using an integrated unit, FIG. 13 shows a possible structural diagram of the network device involved in the above embodiment. The network device 12 includes: a storage module 1221 , a processing module 1222 , and a communication module 1223 . The above-mentioned modules are used to support the network device to execute the related method in any one of the drawings in FIG. 4 . The network device provided in this application is used to implement the corresponding method provided above, therefore, its corresponding features and beneficial effects can be referred to the beneficial effects of the corresponding method provided above, and will not be repeated here repeat.

Specifically, the processing module 1222 is used to control and manage the actions of the network device 12 . The communication module 1223 is configured to support the network device 12 to execute the functions of the sending unit 1211 and the receiving unit 1212 described above. The storage module 1221 is used for storing program codes and data of network devices.

Wherein, the processing module 1222 may be a processor or a controller, such as a central processing unit (central processing unit, CPU), a general processor, a digital signal processor (digital signal processor, DSP), an application-specific integrated circuit (application-specific integrated circuit, ASIC), field programmable gate array (Field programmable gate array, FPGA) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It can implement or execute the various illustrative logical blocks, modules and circuits described in connection with the present disclosure. The processor may also be a combination of computing functions, for example, a combination of one or more microprocessors, a combination of DSP and a microprocessor, and so on. The communication module 1223 may be a transceiver, a transceiver circuit, Bluetooth, a network interface or a communication interface, and the like. The storage module 1221 may be a memory.

Specifically, the processing module 1222 may be the processor in the BBU 1201 in FIG. 3 , the communication module 1223 may be the RF circuit in the RRU 1202 in FIG. 3 , and the storage module 1221 may be the memory in the BBU 1201 in FIG. 3 .

When the processing module 1222 is a processor, the communication module 1123 is an RF circuit, and the storage module 1221 is a memory, the network device involved in this application may be the network device 12 shown in FIG. 14 .

Referring to FIG. 14 , the network device 12 includes: a processor 1231 , a memory 1232 , a bus system 1233 , an RF circuit 1234 , an optical fiber 1235 , a coaxial cable 1236 , an antenna 1237 , and one or more programs. Wherein, the processor 1231 and the memory 1232 of the BBU 1201 are connected to each other through a bus system 1233 . The RF circuit 1234 in the RRU 1202 is connected to the BBU 1201 through an optical fiber 1235 . The RF circuit 1234 in the RRU 1202 is connected to the antenna 1237 through a coaxial cable 1236 . The aforementioned bus system may be a standard bus for interconnecting peripheral components or an extended industry standard structure bus. The bus can be divided into address bus, data bus, control bus and so on. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus. Wherein the one or more programs are stored in the memory, the one or more programs include instructions, and the instructions, when executed by the network device, cause the network device to perform the relevant methods in any one of Figure 4 .

The present application also provides a computer storage medium storing one or more programs, and the one or more programs include instructions, and the instructions, when executed by the network device, cause the network device to execute the related method in any one of the drawings in FIG. 4 .

The present application also provides a computer program product containing instructions. When the computer program product is run on the network device, the network device is made to execute the related method in any one of the drawings in FIG. 4 .

An embodiment of the present application provides a chip system, the chip system includes a processor, configured to support the network device to implement the above information indication method, for example, to send the first indication information to the terminal device, the first indication information is to indicate the first time resource instructions for the . In a possible design, the chip system further includes a memory. The memory is used to store necessary program instructions and data of the network equipment. Of course, the memory may not be in the system-on-a-chip. The chip system may include a chip, an integrated circuit, or may include a chip and other discrete devices, which are not specifically limited in this embodiment of the present application.

Wherein, the network device, computer storage medium, computer program product or chip system provided in this application are all used to execute the corresponding method provided above, therefore, the beneficial effects it can achieve can refer to the corresponding method provided above The beneficial effects in the method will not be repeated here.

It should be understood that, in various embodiments of the present application, the sequence numbers of the above-mentioned processes do not mean the order of execution, and the execution order of the processes should be determined by their functions and internal logic, and should not be used in the embodiments of the present application. The implementation process constitutes any limitation.

Those skilled in the art can appreciate that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art may use different methods to implement the described functions for each specific application, but such implementation should not be regarded as exceeding the scope of the present application.

Those skilled in the art can clearly understand that for the convenience and brevity of the description, the specific working process of the above-described system, device and unit can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here.

In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods may be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or May be integrated into another system, or some features may be ignored, or not implemented. In another point, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of devices or units may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit.

In the above embodiments, all or part of them may be implemented by software, hardware, firmware or any combination thereof. When implemented using a software program, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on the computer, the processes or functions according to the embodiments of the present application will be generated in whole or in part. The computer can be a general purpose computer, a special purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from a website, computer, server, or data center Transmission to another website site, computer, server or data center by wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.). The computer-readable storage medium may be any available medium that can be accessed by a computer, or may be a data storage device including one or more servers, data centers, etc. that can be integrated with the medium. The available medium may be a magnetic medium (such as a floppy disk, a hard disk, or a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid state disk (solid state disk, SSD)), etc.

The above is only a specific implementation of the application, but the scope of protection of the application is not limited thereto. Anyone familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the application. Should be covered within the protection scope of this application. Therefore, the protection scope of the present application should be determined by the protection scope of the claims.

Claims (29)

1. A downlink beam training method, comprising: The terminal device receives a sounding beam from the network device, the sounding beam is modulated by an antenna weight vector, and the antenna weight vector includes a fixed codebook vector and a randomly generated random codebook vector; The terminal device constructs an observation matrix according to the antenna weight vector; The terminal device performs signal reconstruction on a sparse vector of a beam search energy space through the observation matrix, and calculates beam pair information according to the sparse vector, wherein the beam pair information includes beam information at a sending end; The terminal device sends the sending end beam information to the network device. 2. The method according to claim 1, wherein the observation matrix is: Among them, b _r,i ＝W _r (W _r W _r ^H ) ^-1 u _r,i , b _t,i ＝W _t (W _t W _t ^H ) ^-1 u _t,i , vec() means the matrix becomes a vector, W _r is the antenna weight vector of the beam at the receiving end during actual transmission, W _t is the antenna weight vector of the beam at the sending end during actual transmission, is the i-th receiving antenna weight vector, u _t,i is the i-th transmitting antenna weight vector, () ^H means conjugate transpose, () ^-1 means matrix inverse. 3. The method according to claim 2, wherein the beam pair information includes optimal beam pair information, and the terminal device performs signal reconstruction on the sparse vector of the beam search energy space through the observation matrix, and Solving the beam pair information according to the sparse vector includes: Accurately estimate the position x of the maximum value in each component of the sparse vector q of the observation matrix θ according to any one of the following formulas: or or Among them, θ _χ represents the χth column of the observation matrix θ, q=(θ ^H θ) ^-1 θ ^H h, Indicates to take the corresponding χ when the calculation result of the following formula is the minimum value, Indicates that the χ corresponding to the calculation result of the following formula is the maximum value, || || ₂ indicates the 2-norm, || || indicates the modulus, || || ² indicates the modulus square, and h is after vectorization The multipath channel response element of ; The optimal beam pair information (k _opt , l _opt ) is obtained according to the following formula: l _opt = x/K _r , k _opt = x - K _r l _opt , Among them, K _r is the number of beams at the receiving end, l _opt is the beam number of the optimal sending end, and k _opt is the beam number of the optimal receiving end. 4. The method according to claim 3, wherein the beam pair information further includes candidate beam pair information, and the method further comprises: The terminal device makes an accurate estimate of the position x of the maximum value of the remaining non-zero value components in q according to the formula, and obtains the candidate beam pair information (k _oth , _{loth ), where l oth is} the candidate beam pair Select the beam number of the sending end, k _oth is the beam number of the alternative receiving end. 5. The method according to any one of claims 1-4, wherein the beam information at the transmitting end includes a codebook vector corresponding to the beam at the transmitting end or a sequence number of the beam at the transmitting end. 6. The method according to any one of claims 1-4, wherein the sending end beam information includes optimal sending end beam information and alternative sending end beam information. 7. The method according to any one of claims 1-4, wherein the beam pair information further includes receiver beam information, and the method further comprises: The terminal device sends the receiving end beam information to the network device. 8. A downlink beam training method, comprising: The network device sends a sounding beam, and the sounding beam is modulated by an antenna weight vector, and the antenna weight vector includes a fixed codebook vector and a randomly generated random codebook vector, and the antenna weight vector is used to construct an observation matrix, and the observation matrix performing signal reconstruction on a sparse vector of a beam search energy space, where the sparse vector is used to obtain beam pair information, where the beam pair information includes beam information at a transmitting end; The network device receives the sending end beam information from the terminal device. 9. The method according to claim 8, wherein the observation matrix is: Among them, b _r,i ＝W _r (W _r W _r ^H ) ^-1 u _r,i , b _t,i ＝W _t (W _t W _t ^H ) ^-1 u _t,i , vec() means the matrix becomes a vector, W _r is the antenna weight vector of the beam at the receiving end during actual transmission, W _t is the antenna weight vector of the beam at the sending end during actual transmission, is the i-th receiving antenna weight vector, u _t,i is the i-th transmitting antenna weight vector, () ^H means conjugate transpose, () ^-1 means matrix inverse. 10. The method according to claim 9, wherein the beam pair information includes optimal beam pair information (k _opt , l _opt ): l _opt = x/K _r , k _opt = x - K _r l _opt , Among them, K _r is the number of beams at the receiving end, l _opt is the beam number of the optimal sending end, k _opt is the beam number of the optimal receiving end, and χ is the position of the maximum value among the components of the sparse vector q of the observation matrix θ ; χ is obtained by making an accurate estimate according to any of the following formulas: or or Among them, θ _χ represents the χth column of the observation matrix θ, q=(θ ^H θ) ^-1 θ ^H h, Indicates to take the corresponding χ when the calculation result of the following formula is the minimum value, Indicates that the χ corresponding to the calculation result of the following formula is the maximum value, || || ₂ indicates the 2-norm, || || indicates the modulus, || || ² indicates the modulus square, and h is after vectorization The multipath channel response element of . 11. The method according to claim 10, wherein the beam pair information further includes candidate beam pair information (k _oth , l _oth ), where _loth is the beam sequence number of the candidate transmitter, and k _oth is The beam number of the candidate receiving end, and the candidate beam pair information (k _oth , _loth ) is obtained by accurately estimating the position χ of the maximum value among the remaining non-zero value components in q according to the formula. 12. The method according to any one of claims 8-11, wherein the beam information at the transmitting end includes a codebook vector corresponding to the beam at the transmitting end or a sequence number of the beam at the transmitting end. 13. The method according to any one of claims 8-11, wherein the beam information of the sending end includes beam information of an optimal sending end and beam information of an alternative sending end. 14. The method according to any one of claims 8-11, wherein the beam pair information further includes receiver beam information, and the method further comprises: The network device receives the receiver beam information from the terminal device. 15. A terminal device, characterized in that it comprises: A receiving unit, configured to receive a sounding beam from a network device, where the sounding beam is modulated by an antenna weight vector, and the antenna weight vector includes a fixed codebook vector and a randomly generated random codebook vector; a construction unit, configured to construct an observation matrix from the antenna weight vectors received by the receiving unit; The solving unit is used to perform signal reconstruction on the sparse vector of the beam search energy space with the observation matrix constructed by the construction unit, and solve the beam pair information according to the sparse vector, wherein the beam pair information includes the beam information of the transmitting end; A sending unit, configured to send the beam information of the sending end solved by the solving unit to the network device. 16. The terminal device according to claim 15, wherein the observation matrix is: Among them, b _r,i ＝W _r (W _r W _r ^H ) ^-1 u _r,i , b _t,i ＝W _t (W _t W _t ^H ) ^-1 u _t,i , vec() means the matrix becomes a vector, W _r is the antenna weight vector of the beam at the receiving end during actual transmission, W _t is the antenna weight vector of the beam at the sending end during actual transmission, is the i-th receiving antenna weight vector, u _t,i is the i-th transmitting antenna weight vector, () ^H means conjugate transpose, () ^-1 means matrix inverse. 17. The terminal device according to claim 16, wherein the solving unit is specifically used for: Accurately estimate the position x of the maximum value in each component of the sparse vector q of the observation matrix θ according to any one of the following formulas: or, or, Among them, θ _χ represents the χth column of the observation matrix θ, q=(θ ^H θ) ^-1 θ ^H h, Indicates to take the corresponding χ when the calculation result of the following formula is the minimum value, Indicates that the χ corresponding to the calculation result of the following formula is the maximum value, || || ₂ indicates the 2-norm, || || indicates the modulus, || || ² indicates the modulus square, and h is after vectorization The multipath channel response element of ; The optimal beam pair information (k _opt , l _opt ) is obtained according to the following formula: l _opt = x/K _r , k _opt = x - K _r l _opt , Among them, K _r is the number of beams at the receiving end, l _opt is the beam number of the optimal sending end, and k _opt is the beam number of the optimal receiving end. 18. The terminal device according to claim 17, wherein the beam pair information further includes candidate beam pair information, and the solving unit is further configured to: Accurately estimate the position χ of the maximum value among the remaining non-zero value components in q in turn according to the formula, and obtain the candidate beam pair information (k _oth , _loth ), where _loth is the candidate beam at the sending end serial number, k _oth is the beam serial number of the alternative receiving end. 19. The terminal device according to any one of claims 15-18, wherein the beam information of the transmitter includes a codebook vector corresponding to the beam of the transmitter or a sequence number of the beam of the transmitter. 20. The terminal device according to any one of claims 15-18, wherein the beam information of the sending end includes beam information of an optimal sending end and beam information of an alternative sending end. 21. The terminal device according to any one of claims 15-18, wherein the beam pair information further includes receiver beam information, and the sending unit is further configured to: Send the receiving end beam information to the network device. 22. A network device, comprising: The network device sends a sounding beam, and the sounding beam is modulated by an antenna weight vector, and the antenna weight vector includes a fixed codebook vector and a randomly generated random codebook vector, and the antenna weight vector is used to construct an observation matrix, and the observation matrix performing signal reconstruction on a sparse vector of a beam search energy space, where the sparse vector is used to obtain beam pair information, where the beam pair information includes beam information at a transmitting end; The network device receives the sending end beam information from the terminal device. 23. The network device according to claim 22, wherein the observation matrix is: Among them, b _r,i ＝W _r (W _r W _r ^H ) ^-1 u _r,i , b _t,i ＝W _t (W _t W _t ^H ) ^-1 u _t,i , vec() means the matrix becomes a vector, W _r is the antenna weight vector of the beam at the receiving end during actual transmission, W _t is the antenna weight vector of the beam at the sending end during actual transmission, is the i-th receiving antenna weight vector, u _t,i is the i-th transmitting antenna weight vector, () ^H means conjugate transpose, () ^-1 means matrix inverse. 24. The network device according to claim 23, wherein the beam pair information includes optimal beam pair information (k _opt , l _opt ): l _opt = x/K _r , k _opt = x - K _r l _opt , Among them, K _r is the number of beams at the receiving end, l _opt is the beam number of the optimal sending end, k _opt is the beam number of the optimal receiving end, and χ is the position of the maximum value among the components of the sparse vector q of the observation matrix θ ; χ is obtained by making an accurate estimate according to any of the following formulas: or, or, Among them, θ _χ represents the χth column of the observation matrix θ, q=(θ ^H θ) ^-1 θ ^H h, Indicates to take the corresponding χ when the calculation result of the following formula is the minimum value, Indicates that the χ corresponding to the calculation result of the following formula is the maximum value, || || ₂ indicates the 2-norm, || || indicates the modulus, || || ² indicates the modulus square, and h is after vectorization The multipath channel response element of . 25. The network device according to claim 24, wherein the beam pair information further includes candidate beam pair information (k _oth , _loth ), wherein, _loth is the beam sequence number of the candidate transmitter, and k _oth is the beam number of the candidate receiving end, and the candidate beam pair information (k _oth , _loth ) is obtained by accurately estimating the position χ of the maximum value among the remaining non-zero value components in q according to the formula. 26. The network device according to any one of claims 16-25, wherein the beam information of the transmitter includes a codebook vector corresponding to the beam of the transmitter or a sequence number of the beam of the transmitter. 27. The network device according to any one of claims 16-25, wherein the beam information of the sending end includes information about an optimal sending end beam and information about an alternative sending end beam. 28. The network device according to any one of claims 16-25, wherein the beam pair information further includes receiving end beam information, and the receiving unit is further used for: receiving the receiver beam information from the terminal device. 29. A storage medium, on which a computer program is stored, characterized in that, when the computer program is executed by a processor, the method according to any one of claims 1-7 is implemented, or, any one of claims 8-14 is implemented. one of the methods described.