WO2023087819A1 - 波束训练方法、第一节点、第二节点、通信系统及介质 - Google Patents

波束训练方法、第一节点、第二节点、通信系统及介质 Download PDF

Info

Publication number
WO2023087819A1
WO2023087819A1 PCT/CN2022/113557 CN2022113557W WO2023087819A1 WO 2023087819 A1 WO2023087819 A1 WO 2023087819A1 CN 2022113557 W CN2022113557 W CN 2022113557W WO 2023087819 A1 WO2023087819 A1 WO 2023087819A1
Authority
WO
WIPO (PCT)
Prior art keywords
sampling
sampling point
ris
field
node
Prior art date
Application number
PCT/CN2022/113557
Other languages
English (en)
French (fr)
Inventor
戴凌龙
巍秀红
赵亚军
郁光辉
段向阳
Original Assignee
中兴通讯股份有限公司
清华大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司, 清华大学 filed Critical 中兴通讯股份有限公司
Publication of WO2023087819A1 publication Critical patent/WO2023087819A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/06—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
    • H04B7/0613—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission
    • H04B7/0615—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal
    • H04B7/0617—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal for beam forming
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/06—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Definitions

  • the present application relates to the technical field of wireless communication networks, for example, to a beam training method, a first node, a second node, a communication system and a medium.
  • the intelligent metasurface can realize the intelligent regulation of the wireless communication environment through a large number of controllable units.
  • RIS can be deployed between the base station and the user, providing it with an additional reflection link to assist communication.
  • Such effective reflection beamforming requires accurate channel state information (Channel State Information, CSI).
  • CSI can be obtained by beam training. Beam training can be understood as performing a training process between the base station and the user through multiple directional beams to find the optimal beam. These directional beams are also called codewords, and multiple codewords together form a codebook.
  • the beam training codebook in the related art is based on the far-field channel model.
  • the increase in the number of elements (for example, from 256 to 1024) will lead to a very large array size, which changes the structure of the electromagnetic wave field, that is, the scatterers are basically in the near-field of ultra-large-scale RIS scope.
  • the beam training codebook based on the far-field channel model will cause serious performance loss in the ultra-large-scale RIS-assisted near-field communication.
  • the present application provides a beam training method, a first node, a second node, a communication system and a medium.
  • An embodiment of the present application provides a beam training method, including:
  • each group of sampling point pairs consists of a first sampling point and a second sampling point
  • the first sampling point is a candidate position of a scatterer between the first node and the RIS
  • the second sampling point is a candidate position of a scatterer between the RIS and the second node
  • Feedback information is received, where the feedback information includes information about an optimal beam.
  • the embodiment of the present application also provides a beam training method, including:
  • each group of sampling point pairs is composed of one of the current sampling point pairs Composed of a first sampling point and a second sampling point, the first sampling point is the candidate position of the scatterer between the first node and the RIS, and the second sampling point is between the RIS and the second node Candidate locations for scatterers;
  • the feedback information includes information about the optimal beam within the current sampling range
  • the embodiment of the present application also provides a beam training method, including:
  • the training symbols are received through the RIS, wherein the training symbols are sent according to the near-field codewords of each group of sampling point pairs within the sampling range, and the near-field codewords are constructed according to the array response vectors of the near-field concatenated channels of the RIS, and each group
  • the sampling point pair is composed of a first sampling point and a second sampling point, the first sampling point is a candidate position of a scatterer between the first node and the RIS, and the second sampling point is the Candidate locations for scatterers between the second nodes;
  • the feedback information includes information about the optimal beam within the sampling range.
  • the embodiment of the present application also provides a first node, including: a memory, a processor, and a computer program stored in the memory and operable on the processor, and the processor implements the above beam training method when executing the program .
  • the embodiment of the present application also provides a second node, including: a memory, a processor, and a computer program stored in the memory and operable on the processor, and the processor implements the above beam training method when executing the program .
  • the embodiment of the present application also provides a communication system, including: RIS, the first node provided in the embodiment of the present application, and the second node provided in the embodiment of the present application.
  • An embodiment of the present application also provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the program is executed by a processor, the foregoing beam training method is implemented.
  • FIG. 1 is a flowchart of a beam training method provided by an embodiment
  • FIG. 2 is a schematic diagram of an ultra-large-scale RIS-assisted communication system provided by an embodiment
  • FIG. 3 is a schematic diagram of an implementation of a super-large-scale RIS near-field codebook construction and beam training method provided by an embodiment
  • FIG. 4 is a flowchart of another beam training method provided by an embodiment
  • FIG. 5 is a schematic diagram of a near-field codebook and a hierarchical near-field codebook provided by an embodiment
  • FIG. 6 is a schematic diagram of a system attainable rate performance comparison between a near-field codebook and a hierarchical near-field codebook provided by an embodiment
  • FIG. 7 is a schematic diagram of beam training overhead comparison between a near-field codebook and a hierarchical near-field codebook provided by an embodiment
  • FIG. 8 is a schematic diagram of an implementation of a hierarchical near-field codebook construction and hierarchical beam training method provided by an embodiment
  • FIG. 9 is a flowchart of another beam training method provided by an embodiment.
  • FIG. 10 is a schematic structural diagram of a beam training device provided by an embodiment
  • Fig. 11 is a schematic structural diagram of another beam training device provided by an embodiment
  • Fig. 12 is a schematic structural diagram of another beam training device provided by an embodiment
  • FIG. 13 is a schematic diagram of a hardware structure of a first node provided by an embodiment
  • FIG. 14 is a schematic diagram of a hardware structure of a second node provided by an embodiment
  • Fig. 15 is a schematic structural diagram of a communication system provided by an embodiment.
  • the CSI acquisition may include two types of methods, one is explicit CSI acquisition, that is, channel estimation; the other is implicit CSI acquisition, that is, beam training.
  • the base station sends a pilot to the user through the RIS, and the user estimates the channel based on the received pilot. Since the RIS unit is usually passive, generally only the concatenated channel composed of the base station to the RIS and the RIS to the user can be estimated. Compared with the channel dimension from the base station to the user in the non-RIS system, the concatenated channel dimension is doubled. For example, when the number of RIS units is 256, the concatenated channel dimension is increased by 256 times compared with the original. High-dimensional concatenated channel estimation results in huge pilot overhead. For this type of explicit CSI acquisition method, since the RIS does not perform effective reflection beamforming before channel estimation, the received signal-to-noise ratio is generally very low. It is difficult for channel estimation to achieve satisfactory estimation performance under low received signal-to-noise ratio.
  • the second class of methods is beam training.
  • a beam training method is used to construct a codebook and find an optimal beam.
  • the physical angle of the channel path can be estimated instead of the entire channel.
  • Beam training refers to performing a training process between the base station and the user through multiple directional beams to find the optimal beam. These directional beams are also called codewords, which are defined in advance in the codebook used for beam training. After beam training, the physical angle of the channel path can be obtained equivalently.
  • beam training can directly implement an effective beamforming, thus avoiding estimating the entire channel at a lower received SNR. Combining beam training in RIS-assisted communication systems can be used for CSI acquisition.
  • RIS cascaded array response vector of RIS concatenated channels first a codebook containing multiple RIS directional beams (ie, codewords) can be designed, and then a training process is performed between the RIS and the user to find out the Optimal RIS directional beam.
  • a beam training scheme based on partial search can be used to reduce the search complexity.
  • the RIS auxiliary system faces serious "multiplicative fading" path loss, that is, the equivalent path loss of the base station-RIS-user reflection link depends on the base station to RIS and RIS to user The product of the path loss of the two links. Thanks to the advantages of low cost and low power consumption of RIS, RIS is more likely to develop into a super-large-scale RIS in the future, so as to obtain higher gain to make up for its severe path loss. From RIS to ultra-large-scale RIS, not only means the increase in the number of RIS units, but also will lead to changes in the structure of the electromagnetic wave field. The field structure of electromagnetic waves is divided into near field and far field.
  • the boundary between the far and near fields can be determined by the Rayleigh distance, which is proportional to the square of the array size.
  • the RIS beam training codebook is based on the far-field channel model. From RIS to very large-scale RIS, as the number of elements increases (for example, from 256 to 1024), the array size of very large-scale RIS becomes very large, and its Rayleigh distance increases accordingly, and the scatterers are more likely to be in the super-large Scale the near-field range of the RIS. Therefore, the beam training codebook based on the far-field channel model will cause severe performance loss in very large-scale RIS-assisted near-field communication.
  • a beam training method which can realize ultra-large-scale RIS near-field codebook design and beam training based on the near-field channel model, and improve the performance of RIS-assisted near-field communication; and on this basis, further A super-large-scale RIS hierarchical near-field codebook design and a hierarchical beam training method are proposed to reduce beam training overhead.
  • Figure 1 is a flow chart of a beam training method provided by an embodiment, which can be applied to a first node (such as a base station), as shown in Figure 1, the method provided by this embodiment includes step 110, step 120 and step 130.
  • a first node such as a base station
  • step 110 according to the array response vector of the near-field concatenated channel of the RIS, the near-field codeword of each group of sampling point pairs is constructed, wherein each group of sampling point pairs is composed of a first sampling point and a second sampling point , the first sampling point is a candidate position of the scatterer between the first node and the RIS, and the second sampling point is a candidate position of the scatterer between the RIS and the second node.
  • the first node and the second node may be understood as communication nodes for data communication, for example, the first node may be a base station, and the second node may be a user terminal (User Equipment, UE).
  • the first node is a base station and the second node is a user terminal as an example for illustration.
  • the smart metasurface namely RIS
  • the RIS may be a hyperscale RIS.
  • the cascaded channel may refer to a channel from the base station to the RIS and a channel from the RIS to the user, and the cascaded channel formed by these two channels.
  • the near-field cascaded channel can be understood as a cascaded channel in the near-field range of the RIS.
  • the array response vector can be used to characterize the response capability of the RIS array antenna to incoming waves in a certain direction.
  • the RIS can usually be deployed between the base station and the user terminal, and the RIS can provide an additional reflection link for the base station and the user terminal to assist communication.
  • a scatterer can correspond to a spatial position, which can be represented by three-dimensional coordinates.
  • Each set of sampling point pairs may consist of a first sampling point and a second sampling point, wherein the first sampling point may be a candidate position of a scatterer between the first node and the RIS, and there may be multiple first sampling points;
  • the second sampling point is a candidate position of the scatterer between the RIS and the second node, and there may be multiple second sampling points.
  • multiple spatial positions may be included between the first node (that is, the base station) and the RIS as candidate positions of the scatterer, and the first sampling point may be the three-dimensional coordinates corresponding to the candidate positions.
  • the RIS and the second node ie, the user terminal
  • the second sampling point may also include multiple spatial positions as candidate positions of the scatterer, and the second sampling point may be the three-dimensional coordinates corresponding to the candidate positions.
  • the first node (that is, the base station) can construct the near-field codeword of each group of sampling point pairs according to the array response vector of the near-field concatenated channel of the RIS, and each near-field codeword can depend on the distance between the first sampling point and the RIS The sum of the distance and the distance between the second sampling point and RIS.
  • a group of sampling point pairs can correspond to construct a near-field codeword, wherein the near-field codewords corresponding to different sampling point pairs may be the same or different.
  • step 120 the training symbols are sent through the RIS according to the near-field codewords of each group of sampling point pairs.
  • the training symbols may be understood as transmission symbols used for beam training.
  • the first node i.e., the base station
  • the second node i.e., the user terminal
  • the optimal beam that is, the optimal near-field codeword
  • information such as the index, identification or near-field codeword of the optimal beam can be fed back to the first node .
  • step 130 feedback information is received, where the feedback information includes information about the optimal beam.
  • multiple directional beams (directional beams can also be called codewords) that can be used for communication may be included between the first node and the second node, and the optimal beam can be understood as the multiple directional beams
  • the optimal directional beam among the multiple near-field codewords is the optimal near-field codeword.
  • the information of the optimal beam may refer to the subscript of the near-field codeword corresponding to the optimal beam. For example, after receiving the training symbols, the second node performs corresponding processing, and returns the processed feedback information to the first node, and the first node receives the feedback information returned by the second node, that is, the information of the optimal beam.
  • the method further includes:
  • Step 112 Construct a codebook according to the near-field codewords of each group of sampling point pairs, and the codebook includes a plurality of non-repetitive near-field codewords.
  • the codebook may be understood as a set of multiple near-field codewords. Different sampling point pairs may be constructed into the same near-field codeword, and the codebook includes multiple non-repetitive near-field codewords, that is, the near-field codewords contained in the codebook are different from each other. Therefore, in After constructing a near-field codeword for each set of sampling point pairs, the near-field codeword can be compared with all the near-field codewords contained in the current codebook. If the current codebook does not contain the near-field codeword , the near-field codeword can be added to the codebook; if the current codebook already contains the near-field codeword, the near-field codeword will not be added to the codebook. On this basis, continue to judge the next near-field codeword until the near-field codewords of all sampling point pairs are judged.
  • the method before step 110, the method further includes:
  • Step 101 Determine a first set of sampling points and a second set of sampling points, the first set of sampling points includes a plurality of first sampling points, and the second set of sampling points includes a plurality of second sampling points.
  • the first set of sampling points may refer to a set composed of multiple first sampling points; the second set of sampling points may refer to a set composed of multiple second sampling points.
  • a sampling range is respectively determined within the range of the channel (ie, the main path) between the first node and the RIS and within the range of the channel (ie, the main path) between the RIS and the second node , respectively perform sampling in the two sampling ranges to obtain a plurality of first sampling points and a plurality of second sampling points, and form a corresponding first sampling point set and a second sampling point set.
  • the sampling range can be set according to the actual situation, and the sampling method can be equal interval sampling or non-uniform sampling, etc., which is not limited here.
  • step 101 includes:
  • Step 1011 Sampling the sampling range according to the set sampling step to obtain a first set of sampling points and a second set of sampling points.
  • the first node may sample the sampling range according to the preset sampling step size, that is, set the sampling step size, so as to obtain the corresponding first sampling point set and the second sampling point set.
  • the near-field codeword of each group of sampling point pairs is associated with the sum of the distance between the first sampling point and the RIS and the distance between the second sampling point and the RIS.
  • the first sampling point and the second sampling point in each group of sampling point pairs correspond to a spatial position respectively
  • the near-field codeword of each group of sampling point pairs is based on the distance between the spatial position of the first sampling point and the RIS
  • the sum of the distance and the distance between the spatial position of the second sampling point and the RIS is determined.
  • step 120 includes:
  • Step 1210 traverse the near-field codewords of each group of sampling point pairs, set the reflection coefficient of the RIS to be equal to the currently traversed near-field codewords, and send training symbols based on the reflection coefficients through the RIS.
  • the first node traverses the near-field codewords of each group of sampling point pairs, and makes the reflection coefficient of the RIS equal to the currently traversed near-field codewords.
  • the training symbols are sent to the first node through the RIS based on the reflection coefficients.
  • Two nodes, so that the second node receives the training symbols sent according to each near-field codeword, and determines the optimal beam therefrom.
  • the beam training is exemplarily described below through different embodiments.
  • Fig. 2 is a schematic diagram of a very large-scale RIS-assisted communication system provided by an embodiment.
  • a base station with M antennas i.e., the first node
  • a single-antenna user i.e., the second node
  • the ultra-large-scale RIS provides a reflection link for it.
  • the ultra-large-scale RIS is placed on the xz plane, including N 1 ⁇ N 2 RIS units, and its center is located at the origin of the xyz three-dimensional coordinate system.
  • the base station sends a signal s to the user through the ultra-large-scale RIS, then the received signal r at the user side can be expressed as:
  • v ⁇ C M ⁇ 1 represents the beamforming vector on the base station side
  • G ⁇ C N ⁇ M represents the channel matrix from the base station to the very large-scale RIS
  • ⁇ C N ⁇ 1 represents the reflected beam multiplied by the reflection coefficient of the very large-scale RIS Shaped vector
  • h r ⁇ C N ⁇ 1 represents the channel from the user to the ultra-large-scale RIS
  • n represents Gaussian white noise with a mean value of 0 and a variance of ⁇ 2 .
  • the channel can be considered to be composed of main paths, and the purpose of beam training is to align the beam with the main path . Since the base station and the ultra-large-scale RIS are deployed in corresponding fixed positions, the channel G between the two does not change for a long time, so only the beam training on the ultra-large-scale RIS side can be considered, assuming that the beam on the base station side has been aligned main path.
  • the near-field channel model can be used to model the channel on the ultra-large-scale RIS side, namely
  • h r ⁇ r c(x r ,y r ,z r ),
  • ⁇ G and ⁇ r represent the path gain
  • (x G , y G , z G ) and (x r , y r , z r ) represent the main path between the base station and the hyperscale RIS and between the hyperscale RIS and the user, respectively
  • the coordinates of the scatterer corresponding to the main path ⁇ G represents the beam transmission angle of the base station side.
  • the base station still uses the far-field channel model to model
  • b( ⁇ G ) ⁇ C M ⁇ 1 represents the far-field array response vector
  • c(x, y, z) ⁇ C N ⁇ 1 represents the near-field array response vector.
  • the near-field array response vectors c(x G , y G , z G ) and c(x r , y r , z r ) can be expressed as:
  • the above received signal model can be transformed into:
  • ultra-large-scale RIS beam training is to find a suitable codeword for ⁇ , so that the energy of the received signal is maximized.
  • a near-field codebook can be constructed based on the near-field concatenated channel array response vectors to match the near-field channel model.
  • sampling step size (ie set step size) Sampling the entire sampling range considered to generate two sampling point sets ⁇ G and ⁇ r (namely the first sampling point set and the second sampling point set), which can be expressed as:
  • the near-field codebook W ⁇ C N ⁇ L can be obtained according to all the preserved near-field codewords.
  • Each column in the near-field codebook can represent a near-field codeword, and L can represent a total of L codewords.
  • this embodiment proposes a near-field beam training scheme, that is, a training process is performed between the ultra-large-scale RIS and the user, and a traversal search is performed on the near-field codebook W.
  • This process includes a total of L time slots.
  • the reflection coefficient of the super-large-scale RIS is equal to the l-th codeword in the near-field codebook
  • the base station transmits to the user via the super-large-scale RIS Training symbols
  • the user receives the training symbols, and records the codeword corresponding to the training symbols with the highest energy among the currently received training symbols.
  • the user feeds back the optimal codeword index (that is, the information of the optimal beam) to the base station through the ultra-large-scale RIS.
  • FIG. 3 is a schematic diagram of an implementation of a super-large-scale RIS near-field codebook construction and beam training method provided by an embodiment. As shown in FIG. 3, the implementation process of the method is as follows:
  • sampling step size i.e. set step size
  • ⁇ G and ⁇ r i.e. the first sampling point set and The second set of sampling points
  • A2 Take a sampling point (namely the first sampling point and the second sampling point) from the two sampling point sets ⁇ G and ⁇ r to form a set of sampling point pairs, according to the array response of the ultra-large-scale RIS near-field cascaded channel
  • the vector constructs the near-field codewords of each group of sampling point pairs, and each near-field codeword depends on the sum of the distances from two sampling points to the ultra-large-scale RIS (that is, the near-field codewords of each group of sampling point pairs are associated with the first The sum of the distance between the sampling point and the RIS and the distance between the second sampling point and the RIS);
  • A7 Construct a near-field codebook according to the near-field codewords of each group of sampling point pairs;
  • A8 Perform an ergodic search on the near-field codebook generated in A7.
  • the base station sends training symbols to the user via the ultra-large-scale RIS;
  • the user receives the training symbols, and records the near-field codeword corresponding to the training symbols with the highest energy among the currently received training symbols;
  • the user side can obtain the optimal near-field codeword, and the user side feeds back the optimal codeword index to the base station through the ultra-large-scale RIS.
  • FIG. 4 is a flowchart of another beam training method provided by an embodiment. As shown in FIG. 4, it is applied to a first node (such as a base station).
  • the method provided by this embodiment includes steps 210, 220, 230, and Step 240.
  • hierarchical beam training is performed on the constructed near-field codebook, for example, the beam training process of the constructed codebook is classified into multiple search stages, and each search stage performs corresponding subcode This is to search for the optimal codeword through traversal.
  • the search stage of each level corresponds to different sub-codebooks, and correspondingly corresponds to different sampling ranges and sampling steps.
  • the sampling range and sampling step size of each search stage can be based on the sampling range and sampling step size of the previous search stage. Taking the near-field codeword corresponding to the optimal beam in the current search stage as the center, the corresponding reduction is performed according to the set ratio.
  • step 210 in the current search phase, according to the array response vector of the near-field concatenated channel of the RIS, the near-field codewords of each group of sampling point pairs are constructed, wherein each group of sampling point pairs consists of a first A sampling point and a second sampling point, the first sampling point is the candidate position of the scatterer between the first node and the RIS, and the second sampling point is the scatterer between the RIS and the second node body candidate positions.
  • the first node constructs a near-field codeword for each group of sampling point pairs according to the array response vector of the near-field concatenated channel of the RIS.
  • step 220 the training symbols are sent through the RIS according to the near-field codewords of each group of sampling point pairs.
  • the first node sends training symbols to the second node through the RIS according to the near-field codewords of each group of sampling point pairs.
  • step 230 feedback information is received, where the feedback information includes information about the optimal beam within the current sampling range.
  • the first node receives the feedback information returned by the second node through the RIS, and the feedback information includes the information of the optimal beam within the current sampling range in the current search phase.
  • step 240 the current sampling range is updated according to the information of the optimal beam, and enters the next search stage, returning to the operation of constructing the near-field codeword of each group of sampling point pairs until the search of the optimal beam is satisfied stop condition.
  • the first node after receiving the feedback information, the first node takes the optimal beam information contained in the feedback information (that is, the near-field codeword corresponding to the optimal beam) as the center in the current sampling range, and the front and rear boundaries to the center
  • the distance is the set ratio of the currently set sampling step, such as one-half, and the range determined according to the center and the front and rear boundaries is the updated current sampling range.
  • the current sampling range after the above-mentioned update is used as the sampling range of the next search stage.
  • the current setting step of the setting ratio (the range of the setting ratio is greater than 0 and less than 1, The specific value can be flexibly set according to the actual situation) as the set sampling step size of the next search stage, enter the next search stage, and return to execute the operation of constructing the near-field codeword of each set of sampling point pairs until the optimal beam is satisfied
  • the search stop condition can be understood as the condition that the constructed near-field codebook is searched hierarchically until the classification is completed to obtain the optimal beam.
  • step 240 includes:
  • Step 2410 Set the current sampling range as the center of the near-field codeword corresponding to the optimal beam, and the distances from the front and rear borders to the center are both half of the set sampling step in the current search stage.
  • updating the current sampling range according to the information of the optimal beam may include that the first node sets the current sampling range to be centered on the near-field codeword corresponding to the optimal beam, and the front and rear boundaries to the center The distance is half of the set sampling step in the current search stage.
  • the method before step 210, the method further includes:
  • Step 201 Determine a first sampling point set and a second sampling point set within the current sampling range, the first sampling point set includes a plurality of first sampling points, and the second sampling point set includes a plurality of second sampling points.
  • step 201 includes:
  • Step 2011 Sampling the current sampling range according to the set sampling step in the current search stage to obtain the first set of sampling points and the second set of sampling points.
  • Step 241 Decrease the set sampling step size of the current search stage according to the set ratio, and use it as the set sampling step size of the next search stage.
  • the set sampling step size of the next search stage can be set to be smaller than that of the current search stage.
  • Set the sampling step size so the setting ratio range can be greater than 0 and less than 1.
  • the set sampling step size of the current search stage is reduced according to a set ratio, and used as the set sampling step size of the next search stage.
  • step 210 the method includes:
  • Step 211 Construct a codebook corresponding to the current sampling range according to the near-field codewords of each group of sampling point pairs within the current sampling range.
  • the near-field codeword of each group of sampling point pairs is associated with the sum of the distance between the first sampling point and the RIS and the distance between the second sampling point and the RIS within the current sampling range.
  • step 220 includes:
  • Step 2210 traverse the near-field codewords of each group of sampling point pairs within the current sampling range, set the reflection coefficient of the RIS to be equal to the currently traversed near-field codewords, and send training symbols based on the reflection coefficients through the RIS.
  • the hierarchical beam training method is exemplarily described below through different embodiments.
  • the near-field beam training method based on ergodic search in the above embodiment may cause relatively large beam training overhead.
  • this embodiment further proposes a hierarchical near-field codebook and a corresponding hierarchical beam training method.
  • the implementation process is as follows:
  • Fig. 5 is a schematic diagram of a near-field codebook and a hierarchical near-field codebook provided by an embodiment.
  • the entire beam training process can be divided into K-level search stages, and each level performs an ergodic search for the optimal beam for the corresponding sub-codebook, and the sub-codebooks corresponding to different levels of search stages
  • the sampling range of the codebook and the set sampling step size gradually become smaller.
  • the sampling range of the first-level sub-codebook is consistent with the sampling range of the near-field codebook in the above embodiment, that is:
  • the corresponding k-th sub-codebook W k is generated, and then W k is traversed to obtain the optimal near-field codeword corresponding to the k-th sub-codebook, That is, the optimal near-field codeword of the current search stage k can be expressed as Based on the optimal near-field codeword obtained in the k-level search stage, we update the sampling range of the k+1-th stage (that is, the next-level search stage) to Indicates the sampling range of the k+1-th level search stage as an example:
  • the user feeds back the subscript of the optimal near-field codeword searched in the K-th level sub-codebook as feedback information to the base station through the RIS.
  • the codebook size of each sub-codebook in the hierarchical codebook can be much smaller than the near-field codebook, so compared with the beam training based on ergodic search, the hierarchical beam training can greatly reduce the beam training overhead.
  • Fig. 6 is a schematic diagram of a comparison of system achievable rate performance between a near-field codebook and a hierarchical near-field codebook provided by an embodiment
  • Fig. 7 is a diagram of a near-field codebook and a hierarchical near-field codebook provided by an embodiment Schematic diagram of beam training overhead comparison.
  • the construction of two near-field codebooks (near-field codebook and hierarchical near-field codebook) and the corresponding beam training method provided by the embodiment of the present application are far-field beam training in ultra-large-scale RIS-assisted In the near-field communication, a better system reachable rate performance can be achieved.
  • the near-field codebook beam training proposed in this embodiment can greatly improve the system reachable rate performance compared with the far-field codebook beam training in the related art. About 50%.
  • this embodiment further proposes a hierarchical near-field codebook and a hierarchical search method, which gradually reduces the sampling range of the search and the sampling step of the search.
  • the hierarchical near-field The beam training overhead of the codebook is much smaller than the beam training overhead of the near-field codebook, and compared with the near-field codebook traversal search, the near-field beam training overhead can be greatly reduced by about 90%.
  • FIG. 8 is a schematic diagram of an implementation of a hierarchical near-field codebook construction and hierarchical beam training method provided by an embodiment. As shown in FIG. 8, the implementation process of the method is as follows:
  • the entire beam training process is divided into K-level search stages (corresponding to K-level sub-codebooks), and the sampling range is initialized (the sampling range of the first-level search stage is set to the entire sampling range of the near-field codebook in the above-mentioned embodiment) And set the sampling step size (it can be understood that the set sampling step size of the first-level search stage can be set to be larger);
  • the user side can obtain the optimal near-field codeword from the K-th sub-codebook, and the user feeds back the optimal codeword index to the base station through the RIS.
  • FIG. 9 is a flowchart of another beam training method provided by an embodiment. As shown in FIG. 9, this method can be applied to a second node, such as a user terminal.
  • the method provided by this embodiment includes steps 310, 320, and Step 330.
  • step 310 the training symbol is received through the RIS, wherein the training symbol is sent according to the near-field codeword of each group of sampling point pairs within the sampling range, and the near-field codeword is based on the array response of the near-field concatenated channel of the RIS Vector construction, each set of sampling point pair is composed of a first sampling point and a second sampling point, the first sampling point is the candidate position of the scatterer between the first node and the RIS, the second sampling point is a candidate location for a scatterer between the RIS and the second node.
  • step 320 an optimal beam is determined according to the received energy of the training symbols.
  • step 330 feedback information is sent, where the feedback information includes information about the optimal beam within the sampling range.
  • the second node receives the training symbols through the RIS, wherein the training symbols are sent by the first node through the RIS according to the near-field codewords of each group of sampling point pairs within the sampling range. Then, the second node determines the optimal beam according to the energy of the received training symbols, that is, selects the near-field codeword corresponding to the training symbol with the highest energy as the optimal beam. Finally, the second node sends feedback information to the base station through the RIS, where the feedback information includes information about the optimal beam obtained within the sampling range.
  • step 320 includes:
  • Step 3210 Use the near-field codeword corresponding to the received training symbol with the highest energy as the optimal beam.
  • the second node determines the optimal beam according to the energy of the received training symbols
  • the near-field codeword corresponding to the received training symbol with the highest energy is used as the optimal beam.
  • the second node records the near-field codeword corresponding to the training symbol with the largest energy; The size between the maximum energy of the training symbol, if the energy of the currently received training symbol is greater than the energy of the currently recorded training symbol, then select the near-field codeword corresponding to the currently received training symbol as the current optimal beam , to replace the previously recorded near-field codeword; otherwise, keep the recorded near-field codeword unchanged and wait for the next judgment.
  • Fig. 10 is a schematic structural diagram of a beam training device provided by an embodiment. As shown in Figure 10, the beam training device includes:
  • the first construction module 410 is configured to construct the near-field codeword of each group of sampling point pairs according to the array response vector of the near-field concatenated channel of the RIS, wherein each group of sampling point pairs consists of a first sampling point and a second sampling point Composed of sampling points, the first sampling point is a candidate position of a scatterer between the first node and the RIS, and the second sampling point is a candidate position of a scatterer between the RIS and the second node;
  • the first sending module 420 is configured to send training symbols through the RIS according to the near-field codewords of each group of sampling point pairs;
  • the first receiving module 430 is configured to receive feedback information, where the feedback information includes information about an optimal beam.
  • the first node constructs a near-field codeword according to the array response vector of the RIS near-field concatenated channel, and then sends a training symbol to the second node through the RIS to obtain feedback information based on the near-field codeword , to obtain the optimal beam, and according to the near-field codeword, a near-field codebook matching the near-field channel model can be constructed to realize beam training based on the near-field codebook in ultra-large-scale RIS-assisted near-field communication, and improve the efficiency of beam training in the system Achievable rate, and reduce the performance loss of beam training.
  • the device also includes:
  • the first codebook construction module is configured to construct a codebook according to the near-field codewords of each group of sampling point pairs, and the codebook includes a plurality of non-repetitive near-field codewords.
  • the device also includes:
  • the first sampling point set determination module is configured to determine the first sampling point set and the second sampling point set before constructing the near-field codeword of each group of sampling point pairs according to the array response vector of the near-field concatenated channel of the RIS,
  • the first sampling point set includes a plurality of first sampling points
  • the second sampling point set includes a plurality of second sampling points.
  • the sampling point set determination module includes:
  • the first sampling unit is configured to sample a sampling range according to a set sampling step to obtain the first set of sampling points and the second set of sampling points.
  • the near-field codeword of each group of sampling point pairs is associated with the sum of the distance between the first sampling point and the RIS and the distance between the second sampling point and the RIS.
  • the first sending module 420 includes:
  • the first traversal unit is configured to traverse the near-field codeword of each group of sampling point pairs, make the reflection coefficient of the RIS equal to the currently traversed near-field codeword, and send training symbols based on the reflection coefficient through the RIS.
  • the beam training device proposed in this embodiment belongs to the same idea as the beam training method proposed in the above embodiment.
  • Fig. 11 is a schematic structural diagram of a beam training device provided by an embodiment. As shown in Figure 11, the beam training device includes:
  • the second construction module 510 is configured to construct the near-field codeword of each group of sampling point pairs according to the array response vector of the near-field cascaded channel of the smart metasurface RIS in the current search phase, wherein each group of sampling point pairs is composed of the current A first sampling point and a second sampling point within the sampling range, the first sampling point is the candidate position of the scatterer between the first node and the RIS, the second sampling point is the Candidate locations for scatterers between the second nodes;
  • the second sending module 520 is configured to send training symbols through the RIS according to the near-field codewords of each group of sampling point pairs;
  • the second receiving module 530 is configured to receive feedback information, where the feedback information includes information about the optimal beam within the current sampling range;
  • the update module 540 is configured to update the current sampling range according to the information of the optimal beam, and enter the next search stage, and return to execute the operation of constructing the near-field codeword of each group of sampling point pairs until the optimal beam is satisfied. Search for stop conditions.
  • the beam training apparatus of this embodiment can further reduce the beam training overhead on the basis of the near-field codebook through the hierarchical near-field codebook and the corresponding hierarchical beam training.
  • the update module 540 includes:
  • a range setting unit configured to set the current sampling range as centered on the near-field codeword corresponding to the optimal beam, and the distance from the front and rear boundaries to the center is equal to the set sampling step of the current search stage Half.
  • the device also includes:
  • the second sampling point set determination module is configured to determine the first sampling point in the current sampling range before constructing the near-field codeword of each group of sampling point pairs according to the array response vector of the near-field cascaded channel of the smart metasurface RIS.
  • the second sampling point set determination module includes:
  • the second sampling unit is configured to sample the current sampling range according to the set sampling step of the current search stage to obtain the first sampling point set and the second sampling point set.
  • the device also includes:
  • the sampling step determination module is configured to reduce the current search rate according to a set ratio after updating the current sampling range according to the information of the optimal beam and before returning to the operation of constructing the near-field codeword of each group of sampling point pairs.
  • the set sampling step of the stage is used as the set sampling step of the next search stage.
  • the device also includes:
  • a codebook corresponding to the current sampling range is constructed according to the near-field codewords of each group of sampling point pairs within the current sampling range.
  • the second codebook construction module is configured to construct a codebook corresponding to the current sampling range according to the near-field codewords of each group of sampling point pairs within the current sampling range.
  • the near-field codeword of each group of sampling point pairs is associated with the distance between the first sampling point and the RIS and the distance between the second sampling point and the RIS within the current sampling range. The sum of the distances between.
  • the second sending module 520 includes:
  • the second traversal unit is configured to traverse the near-field codewords of each group of sampling point pairs in the current sampling range, make the reflection coefficient of the RIS equal to the near-field codewords currently traversed, and use the RIS based on the reflection Coefficients send training symbols.
  • the beam training device proposed in this embodiment belongs to the same idea as the beam training method proposed in the above embodiment.
  • Fig. 12 is a schematic structural diagram of a beam training device provided by an embodiment. As shown in Figure 12, the beam training device includes:
  • the symbol receiving module 610 is configured to receive training symbols through the intelligent metasurface RIS, wherein the training symbols are sent according to the near-field codewords of each group of sampling point pairs in the sampling range, and the near-field codewords are based on the
  • the array response vector of the near-field cascaded channel is constructed, and each set of sampling point pairs is composed of a first sampling point and a second sampling point, and the first sampling point is a candidate for a scatterer between the first node and the RIS position, the second sampling point is a candidate position of a scatterer between the RIS and the second node;
  • the beam determination module 620 is configured to determine an optimal beam according to the energy received from the training symbols;
  • the information sending module 630 is configured to send feedback information, where the feedback information includes information about the optimal beam within the sampling range.
  • the second node can determine the optimal beam according to the energy of the received training symbol by receiving the training symbol sent by the first node, and feed back the information of the optimal beam to the first node through the RIS.
  • the beam determination module 620 includes:
  • the beam determining unit is configured to use the near-field codeword corresponding to the received training symbol with the highest energy as the optimal beam.
  • the beam training device proposed in this embodiment belongs to the same idea as the beam training method proposed in the above embodiment.
  • the embodiment of the present application also provides a communication node, where the communication node may be a first node, for example, the first node may be a base station, an access point, or a transmission point.
  • Fig. 13 is a schematic diagram of the hardware structure of a first node provided by an embodiment. As shown in Fig. 13, the first node provided by the present application includes a memory 720, a processor 710, and a A computer program running on 710, the processor 710 implements the beam training method described above when executing the program.
  • the first node may also include a memory 720; there may be one or more processors 710 in the first node, and one processor 710 is taken as an example in FIG. 13; the memory 720 is configured to store one or more programs; the one The one or more programs are executed by the one or more processors 710, so that the one or more processors 710 implement the beam training method as described in the embodiment of the present application.
  • the first node further includes: a communication device 730 , an input device 740 and an output device 750 .
  • the processor 710, the memory 720, the communication device 730, the input device 740, and the output device 750 in the first node may be connected via a bus or in other ways. In FIG. 13, connection via a bus is taken as an example.
  • the input device 740 can be configured to receive input numbers or character information, and generate key signal input related to user settings and function control of the first node.
  • the output device 750 may include a display device such as a display screen.
  • Communications device 730 may include a receiver and a transmitter.
  • the communication device 730 is configured to perform information sending and receiving communication according to the control of the processor 710 .
  • the memory 720 can be configured to store software programs, computer-executable programs and modules, such as the program instructions/modules corresponding to the beam training method described in the embodiment of the present application (for example, as shown in FIG. 10
  • the memory 720 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application program required by at least one function; the data storage area may store data created according to the use of the first node, and the like.
  • the memory 720 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid-state storage devices. In some instances, the memory 720 may further include memory located remotely relative to the processor 710, and these remote memories may be connected to the first node through a network. Examples of the aforementioned networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
  • the embodiment of the present application also provides a communication node, where the communication node may be a second node, for example, the second node may be a user terminal (User Equipment, UE).
  • Fig. 14 is a schematic diagram of the hardware structure of a second node provided by an embodiment. As shown in Fig. 14, the second node provided by this application includes a memory 820, a processor 810, and a A computer program running on 810, the processor 810 implements the above-mentioned beam training method when executing the program.
  • the second node may also include a memory 820; there may be one or more processors 810 in the second node, and one processor 810 is taken as an example in FIG. 14; the memory 820 is configured to store one or more programs; the one The one or more programs are executed by the one or more processors 810, so that the one or more processors 810 implement the beam training method as described in the embodiment of the present application.
  • the second node further includes: a communication device 830 , an input device 840 and an output device 850 .
  • the processor 810, the memory 820, the communication device 830, the input device 840, and the output device 850 in the second node may be connected through a bus or in other ways. In FIG. 14, connection through a bus is taken as an example.
  • the input device 840 can be configured to receive input numbers or character information, and generate key signal input related to user settings and function control of the second node.
  • the output device 850 may include a display device such as a display screen.
  • Communications device 830 may include a receiver and a transmitter.
  • the communication device 830 is configured to perform information sending and receiving communication according to the control of the processor 810 .
  • the memory 820 can be configured to store software programs, computer-executable programs and modules, such as the program instructions/modules corresponding to the beam training method described in the embodiment of the present application (for example, shown in FIG. 12 symbol receiving module 610, beam determining module 620 and information sending module 630) in the beam training device.
  • the memory 820 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application program required by at least one function; the data storage area may store data created according to the use of the second node, and the like.
  • the memory 820 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid-state storage devices.
  • the memory 820 may further include memory located remotely relative to the processor 810, and these remote memories may be connected to the second node through a network. Examples of the aforementioned networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
  • FIG. 15 is a schematic structural diagram of a communication system provided by an embodiment. As shown in FIG. 15 , the communication system includes: RIS 910, a first node 920, and a second node 930.
  • the first node 920 constructs the near-field codewords of each group of sampling point pairs according to the array response vector of the near-field concatenated channel of the RIS 910, and respectively according to the near-field codewords of each group of sampling point pairs, through RIS 910 sends training symbols to second node 930.
  • the second node 930 receives the training symbol through the RIS 910, and determines the optimal beam according to the energy of the received training symbol, and sends the information of the optimal beam as feedback information to the first node 920 through the RIS 910.
  • the first node 920 receives the feedback information.
  • a communication system provided in this embodiment may be configured to execute the beam training method provided in any of the foregoing embodiments, and has corresponding functions and beneficial effects.
  • An embodiment of the present application further provides a storage medium, where the storage medium stores a computer program, and when the computer program is executed by a processor, the beam training method described in any one of the embodiments of the present application is implemented.
  • the beam training method includes: according to the array response vector of the near-field cascaded channel of the smart metasurface RIS, constructing the near-field codeword of each group of sampling point pairs, wherein each group of sampling point pairs consists of a first sampling point and a Composed of second sampling points, the first sampling point is a candidate position of a scatterer between the first node and the RIS, and the second sampling point is a candidate position of a scatterer between the RIS and the second node; Sending training symbols through the RIS according to the near-field codewords of each group of sampling point pairs; receiving feedback information, where the feedback information includes optimal beam information.
  • the beam training method includes: in the current search stage, according to the array response vector of the near-field cascaded channel of the smart metasurface RIS, constructing the near-field codeword of each group of sampling point pairs, wherein each group of sampling point pairs is composed of the current sampling point
  • the first sampling point is a candidate position of the scatterer between the first node and the RIS
  • the second sampling point is the Candidate positions of scatterers between the two nodes
  • sending training symbols through the RIS according to the near-field codewords of each group of sampling point pairs receiving feedback information, the feedback information including the optimal beam within the current sampling range information; update the current sampling range according to the information of the optimal beam, and enter the next search stage, return to execute the operation of constructing the near-field codeword of each group of sampling point pairs, until the search stop condition of the optimal beam is satisfied .
  • the beam training method includes: receiving training symbols through the intelligent metasurface RIS, wherein the training symbols are sent according to the near-field codewords of each group of sampling point pairs in the sampling range, and the near-field codewords are based on the intelligent metasurface RIS.
  • the array response vector of the near-field cascaded channel is constructed, and each set of sampling point pairs is composed of a first sampling point and a second sampling point, and the first sampling point is a candidate for a scatterer between the first node and the RIS position, the second sampling point is the candidate position of the scatterer between the RIS and the second node; determine the optimal beam according to the energy received from the training symbol; send feedback information, the feedback information includes the sampling range Information about the optimal beam within.
  • the computer storage medium in the embodiments of the present application may use any combination of one or more computer-readable media.
  • the computer readable medium may be a computer readable signal medium or a computer readable storage medium.
  • a computer-readable storage medium may be, for example, but not limited to: an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any combination thereof. More specific examples (non-exhaustive list) of computer-readable storage media include: electrical connections with one or more conductors, portable computer disks, hard disks, Random Access Memory (RAM), read-only memory (Read Only Memory, ROM), Erasable Programmable Read Only Memory (EPROM), flash memory, optical fiber, portable CD-ROM, optical storage device, magnetic storage device, or any suitable combination of the above .
  • a computer readable storage medium may be any tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device.
  • a computer readable signal medium may include a data signal carrying computer readable program code in baseband or as part of a carrier wave. Such propagated data signals may take many forms, including but not limited to: electromagnetic signals, optical signals, or any suitable combination of the foregoing.
  • a computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. .
  • Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wires, optical cables, radio frequency (Radio Frequency, RF), etc., or any suitable combination of the above.
  • any appropriate medium including but not limited to: wireless, wires, optical cables, radio frequency (Radio Frequency, RF), etc., or any suitable combination of the above.
  • Computer program codes for performing the operations of the present application may be written in one or more programming languages or combinations thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional A procedural programming language, such as the "C" language or similar programming language.
  • the program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server.
  • the remote computer can be connected to the user computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (such as through the Internet using an Internet service provider). connect).
  • LAN local area network
  • WAN wide area network
  • connect such as AT&T, MCI, Sprint, EarthLink, MSN, GTE, etc.
  • user terminal covers any suitable type of wireless user equipment, such as a mobile phone, a portable data processing device, a portable web browser or a vehicle-mounted mobile station.
  • the various embodiments of the present application can be implemented in hardware or special purpose circuits, software, logic or any combination thereof.
  • some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software, which may be executed by a controller, microprocessor or other computing device, although the application is not limited thereto.
  • Embodiments of the present application may be realized by a data processor of a mobile device executing computer program instructions, for example in a processor entity, or by hardware, or by a combination of software and hardware.
  • Computer program instructions may be assembly instructions, Instruction Set Architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or written in any combination of one or more programming languages source or object code.
  • ISA Instruction Set Architecture
  • Any logic flow block diagrams in the drawings of the present application may represent program steps, or may represent interconnected logic circuits, modules and functions, or may represent a combination of program steps and logic circuits, modules and functions.
  • Computer programs can be stored on memory.
  • the memory may be of any type suitable for the local technical environment and may be implemented using any suitable data storage technology, such as but not limited to Read-Only Memory (ROM), Random Access Memory (RAM), Optical Memory devices and systems (Digital Video Disc (DVD) or Compact Disk (CD), etc.
  • Computer-readable media may include non-transitory storage media.
  • Data processors may be any Types, such as but not limited to general-purpose computers, special-purpose computers, microprocessors, digital signal processors (Digital Signal Processing, DSP), application-specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable logic devices (Field-Programmable Gate Array , FGPA) and processors based on multi-core processor architectures.
  • DSP Digital Signal Processing
  • ASIC Application Specific Integrated Circuit
  • FGPA programmable logic devices
  • processors based on multi-core processor architectures.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

本申请提供一种波束训练方法、第一节点、第二节点、通信系统及介质。该方法根据智能超表面RIS的近场级联信道的阵列响应矢量，构建多组采样点对中每组采样点对的近场码字，其中，每组采样点对由一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置；分别根据每组采样点对的近场码字，通过所述RIS发送训练符号；接收反馈信息，所述反馈信息包括最优波束的信息。

Description

波束训练方法、第一节点、第二节点、通信系统及介质 技术领域
本申请涉及无线通信网络技术领域，例如涉及一种波束训练方法、第一节点、第二节点、通信系统及介质。
背景技术
智能超表面(Reconfigurable Intelligent Surface，RIS)可以通过大量可控单元实现对无线通信环境的智能调控。一般而言，RIS可以被部署在基站和用户之间，为其提供一条额外的反射链路来辅助通信。通过对大量RIS单元的合理调控，RIS能够以较低的成本和功耗提供较高的反射波束增益。这种有效的反射波束赋形需要准确的信道状态信息(Channel State Information，CSI)。可以通过波束训练获取CSI。波束训练可以理解为基站和用户之间通过多个指向性波束执行一个训练过程以寻找出最优波束，这些指向性波束又称为码字，多个码字共同构成码本。相关技术中的波束训练码本是基于远场信道模型。
然而，从RIS到超大规模RIS，单元数目的增加(比如，从256到1024)会导致阵列尺寸变得非常大，这使得电磁波场结构发生改变，即散射体基本是处于超大规模RIS的近场范围。基于远场信道模型的波束训练码本，会在超大规模RIS辅助的近场通信中造成严重的性能损失。
发明内容
本申请提供一种波束训练方法、第一节点、第二节点、通信系统及介质。
本申请实施例提供一种波束训练方法，包括：
根据RIS的近场级联信道的阵列响应矢量，构建多组采样点对中每组采样点对的近场码字，其中，每组采样点对由一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置；
分别根据每组采样点对的近场码字，通过所述RIS发送训练符号；
接收反馈信息，所述反馈信息包括最优波束的信息。
本申请实施例还提供了一种波束训练方法，包括：
在当前搜索阶段，根据RIS的近场级联信道的阵列响应矢量，构建多组采样点对中每组采样点对的近场码字，其中，每组采样点对由当前采样范围内的一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置；
分别根据每组采样点对的近场码字，通过所述RIS发送训练符号；
接收反馈信息，所述反馈信息包括所述当前采样范围内的最优波束的信息；
根据所述最优波束的信息更新所述当前采样范围，并进入下一搜索阶段，返回执行构建多组采样点对中每组采样点对的近场码字的操作，直至满足最优波束的搜索停止条件。
本申请实施例还提供了一种波束训练方法，包括：
通过RIS接收训练符号，其中，所述训练符号根据采样范围内每组采样点对的近场码字发送，所述近场码字根据RIS的近场级联信道的阵列响应矢量构建，每组采样点对由一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置；
根据接收所述训练符号的能量确定最优波束；
发送反馈信息，所述反馈信息包括所述采样范围内最优波束的信息。
本申请实施例还提供了一种第一节点，包括：存储器、处理器以及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时实现上述的波束训练方法。
本申请实施例还提供了一种第二节点，包括：存储器、处理器以及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时实现上述的波束训练方法。
本申请实施例还提供了一种通信系统，包括：RIS、如本申请实施例提供的第一节点以及如本申请实施例提供的第二节点。
本申请实施例还提供了一种计算机可读存储介质，计算机可读存储介质上存储有计算机程序，该程序被处理器执行时实现上述的波束训练方法。
附图说明
图1为一实施例提供的一种波束训练方法的流程图；
图2为一实施例提供的一种超大规模RIS辅助的通信系统的示意图；
图3为一实施例提供的一种超大规模RIS近场码本构建和波束训练方法的实现示意图；
图4为一实施例提供的另一种波束训练方法的流程图；
图5为一实施例提供的一种近场码本和分级近场码本的示意图；
图6为一实施例提供的一种近场码本和分级近场码本的系统可达速率性能对比的示意图；
图7为一实施例提供的一种近场码本和分级近场码本的波束训练开销对比的示意图；
图8为一实施例提供的一种分级近场码本构建和分级波束训练方法的实现示意图；
图9为一实施例提供的又一种波束训练方法的流程图；
图10为一实施例提供的一种波束训练装置的结构示意图；
图11为一实施例提供的另一种波束训练装置的结构示意图；
图12为一实施例提供的另一种波束训练装置的结构示意图；
图13为一实施例提供的一种第一节点的硬件结构示意图；
图14为一实施例提供的一种第二节点的硬件结构示意图；
图15为一实施例提供的一种通信系统的结构示意图。
具体实施方式
下面结合附图和实施例对本申请进行说明。可以理解的是，此处所描述的具体实施例仅仅用于解释本申请，而非对本申请的限定。需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互任意组合。另外还需要说明的是，为了便于描述，附图中仅示出了与本申请相关的部分而非全部结构。
CSI获取可以包括两类方法，一是显式的CSI获取，即信道估计；二是隐式的CSI获取，即波束训练。在第一类方法中，基站通过RIS向用户发送导频，用户基于接收的导频估计信道。由于RIS单元通常是无源的，一般只能估计基站到RIS和RIS到用户这两段信道组成的级联信道。相比于无RIS系统中基站到用户的信道维度，级联信道维度成倍提高，比如当RIS单元数目为256时，级联信道维度相比原来提高了256倍。高维级联信道估计导致巨大的导频开销。对于这类显式的CSI获取方法而言，由于在进行信道估计之前RIS并未进行有效的反射波束赋形，接收信噪比一般很低。在较低的接收信噪比下，信道估计难以达到令人满意的估计性能。
第二类方法是波束训练。本实施例采用波束训练的方法构建码本并找到最优波束。在此过程中，可以估计信道路径的物理角度而非整个信道。波束训练是指基站和用户之间通过多个指向性波束执行一个训练过程以寻找出最优波束。这些指向性波束又称为码字，提前定义于波束训练用到的码本中。在波束训练之后，可以等效地得到信道路径的物理角度。相比于信道估计，波束训练可以直接实现一个有效的波束赋形，从而避免了在较低的接收信噪比下来估计整个信道。在RIS辅助通信系统中结合波束训练可以用于CSI获取。其核心思想是基于RIS级联信道的级联阵列响应矢量，首先可以设计一个包含多个RIS指向性波束(即码字)的码本，然后在RIS和用户之间执行一个训练过程以寻找出最优的RIS指向性波束。通过考虑RIS级联阵列响应矢量主要是由在RIS侧的角度差来决定，可以利用基于部分搜索的波束训练方案来降低搜索复杂度。
然而，随着RIS单元数目越来越多，RIS辅助系统面临严重的“乘性衰落”路损，即基站-RIS-用户这个反射链路的等效路损取决于基站到RIS和RIS到用户这两段链路的路损之积。得益于RIS的低成本、低功耗优势，未来RIS更有可能发展为超大规模RIS，以获取更高的增益来弥补其严重的路损。从RIS到超大规模RIS，不仅意味着RIS单元数目的增加，还将导致电磁波场结构的改变。电磁波的场结构分为近场和远场，远近场的分界线可以由瑞利距离决定，瑞利距离正比于阵列尺寸的平方。在RIS辅助系统中，如果阵列尺寸不是很大，瑞利距离比较小，散射体一般假设分布于RIS的远场范围，则RIS波束训练码本是基于远场信道模型。而从RIS到超大规模RIS，随着单元数目的增加(比如，从256到1024)，超大规模RIS的阵列尺寸变得非常大，其瑞利距离也随之增加，散射体更有可能处于超大规模RIS的近场范围。因此，基于远场信道模型的波束训练码本将在超大规模RIS辅助的近场通信中造成严重的性能损失。
在本申请实施例中，提供了一种波束训练方法，能够基于近场信道模型实现超大规模RIS近场码本设计和波束训练，提高RIS辅助的近场通信性能；并在此基础上进一步还提出超大规模RIS分级近场码本设计和分级波束训练方法以降低波束训练开销。
图1为一实施例提供的一种波束训练方法的流程图，该方法可应用在第一节点(如基站)，如图1所示，本实施例提供的方法包括步骤110、步骤120和步骤130。
在步骤110中，根据RIS的近场级联信道的阵列响应矢量，构建每组采样点对的近场码字，其中，每组采样点对由一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置。
本实施例中，第一节点和第二节点可以理解为用于数据通信的通信节点，例如第一节点可以为基站，第二节点可以为用户终端(User Equipment，UE)。在本实施例中，以第一节点为基站，第二节点为用户终端为例进行示例性说明。
本实施例中，智能超表面，即RIS可以理解为一种智能的、具有特殊功能的超材料表面结构。RIS可以为超大规模RIS。级联信道可以指基站到RIS的信道和RIS到用户的信道，这两段信道所组成的级联信道。近场级联信道可以理解为处于RIS近场范围的级联信道。阵列响应矢量可以用于表征RIS阵列天线对某一方向来波的响应能力。
RIS通常可以被部署在基站和用户终端之间，RIS可以为基站和用户终端提供一条额外的反射链路以用于辅助通信。基站和RIS、RIS和用户终端之间存在散射体，其中散射体可以指造成电磁波空间传播轨迹发生偏移的结构体。散射体可以对应一个空间位置，可用三维坐标表征该空间位置。
每组采样点对可以由一个第一采样点和一个第二采样点组成，其中，第一采样点可以为第一节点和RIS之间散射体的候选位置，第一采样点可以为多个；第二采样点为RIS和第二节点之间散射体的候选位置，第二采样点可以为多个。例如，第一节点(即基站)和RIS之间可以包括多个空间位置作为散射体的候选位置，第一采样点可以为该候选位置对应的三维坐标。RIS和第二节点(即用户终端)之间同样可以包括多个空间位置作为散射体的候选位置，第二采样点可以为该候选位置对应的三维坐标。
第一节点(即基站)可以根据RIS的近场级联信道的阵列响应矢量，构建每组采样点对的近场码字，每个近场码字可取决于第一采样点和RIS之间距离与第二采样点和RIS之间距离的和。一组采样点对可对应构建一个近场码字，其中，不同的采样点对所对应的近场码字可能相同，也可能不同。
在步骤120中，分别根据每组采样点对的近场码字，通过所述RIS发送训练符号。
本实施例中，训练符号可以理解为用于波束训练的传输符号。第一节点(即基站)可以分别根据每组采样点对的近场码字，即控制RIS采用相应的指向性波束，发送训练符号至第二节点(即用户终端)，第二节点通过接收多个训练符号并记录其中接收能量最高的训练符号，可以确定最优波束(即最优的近场码字)，并将最优波束的索引、标识或近场码字等信息反馈给第一节点。
在步骤130中，接收反馈信息，所述反馈信息包括最优波束的信息。
本实施例中，第一节点和第二节点之间可以包括多个可用于通信的指向性波束(指向性波束又可称为码字)，其中最优波束可以理解为该多个指向性波束中最优的一个指向性波束，即多个近场码字中最优的一个近场码字。最优波束的信息可以指最优波束所对应的近场码字下标。例如，第二节点在接收到训练符号之后，进行相应的处理，并将处理得到的反馈信息返回至第一节点，第一节点接收第二节点所返回的反馈信息，即最优波束的信息。
在一实施例中，在步骤110之后，该方法还包括：
步骤112：根据每组采样点对的近场码字构建码本，码本包括多个不重复的近场码字。
本实施例中，码本可以理解为多个近场码字所构成的集合。不同的采样点对可能会被构建成相同的近场码字，而码本中包括多个不重复的近场码字，即码本中所包含的近场码字互不相同，因此，在对每组采样点对构建一个近场码字之后，可以先将该近场码字分别与当前码本中所包含的所有近场码字进行对比，若当前码本不包含该近场码字，则可以将该近场码字加入码本中；若当前码本已经包含该近场码字，则该近场码字不再加入码本中。在此基础上，继续进行下一个近场码字的判断，直至将所有采样点对的近场码字判断完毕。
在一实施例中，在步骤110之前，该方法还包括：
步骤101：确定第一采样点集合和第二采样点集合，第一采样点集合中包括多个第一采样点，第二采样点集合中包括多个第二采样点。
本实施例中，第一采样点集合可以指由多个第一采样点所构成的集合；第二采样点集合可以指由多个第二采样点所构成的集合。例如，以信道由主要路径组成为例，在第一节点和RIS之间的信道(即主要路径)范围内和RIS和第二节点之间的信道(即主要路径)范围内分别确定一个采样范围，在这两个采样范围内分别进行采样以得到多个第一采样点和多个第二采样点，并构成对应的第一采样点集合和第二采样点集合。需要说明的是，采样范围可以根据实际情况进行设定，采样方式可以是等间隔采样或非均匀采样等，此处对此不作限定。
在一实施例中，步骤101，包括：
步骤1011：根据设定采样步长对采样范围进行采样，得到第一采样点集合和第二采样点集合。
本实施例中，第一节点可以根据所预先设置的采样步长，即设定采样步长对采样范围进行采样，以得到对应的第一采样点集合和第二采样点集合。
在一实施例中，每组采样点对的近场码字关联于第一采样点到RIS之间的距离与第二采样点到RIS之间的距离之和。
本实施例中，每组采样点对中的第一采样点和第二采样点分别对应一个空间位置，每组采样点对的近场码字根据第一采样点的空间位置到RIS之间的距离与第二采样点的空间位置到RIS之间的距离之和确定。
在一实施例中，步骤120，包括：
步骤1210：遍历每组采样点对的近场码字，令RIS的反射系数等于当前遍历的近场码字，并通过RIS基于反射系数发送训练符号。
本实施例中，第一节点遍历每组采样点对的近场码字，并令RIS的反射系数等于当前遍历的近场码字，在此基础上，通过RIS基于反射系数发送训练符号至第二节点，以使第二节点接收到根据每个近场码字发送的训练符号，从中确定最优波束。
以下通过不同实施例对波束训练进行示例性说明。
本申请实施例针对超大规模RIS辅助的通信系统，提出一种近场码本设计和波束训练方法。图2为一实施例提供的一种超大规模RIS辅助的通信系统的示意图。如图2所示，假设一个M天线的基站(即第一节点)在一个超大规模RIS辅助的情况下与单天线用户(即第二节点)进行通信，基站与用户之间的直接链路被障碍物遮挡，超大规模RIS为其提供了一条反射链路。其中超大规模RIS放置在x-z平面上，共包括N 1×N 2个RIS单元，其中心位于x-y-z三维坐标系的原点。
考虑下行传输，基站经超大规模RIS向用户发送信号s，则用户侧的接收信号r可以表示为：
Figure PCTCN2022113557-appb-000001
其中v∈C M×1表示基站侧的波束赋形矢量；G∈C N×M表示基站到超大规模RIS的信道矩阵；θ∈C N×1表示由超大规模RIS的反射系数组乘反射波束赋形向量；h r∈C N×1表示用户到超大规模RIS的信道；n表示高斯白噪声，均值为0，方差为σ 2。考虑到信道常由主要路径(或者少数路径)组成，尤其是毫米波或者太赫兹等高频场景下，本实施例中可考虑信道由主要路径组成，波束训练的目的是将波束对准主要路径。由于基站和超大规模RIS部署在相应的固定位置后，二者之间的信道G在较长时间内均不发生变化，故可仅考虑超大规模RIS侧的波束训练，假设基站侧波束已经对准主要路径。
由于超大规模RIS阵列尺寸较大，散射体更容易处于其近场范围。故可以用近场信道模型来建模超大规模RIS侧的信道，即
G＝α Gc(x G，y G，z G)b T(φ G)，
h r＝α rc(x r，y r，z r)，
其中α G和α r表示路径增益，(x G，y G，z G)和(x r，y r，z r)分别表示基站和超大规模RIS之间主要路径和超大规模RIS和用户之间主要路径所对应散射体的坐标，φ G表示基站侧波束发送角度。注意基站侧仍用远场信道模型来建模，b(φ G)∈C M×1表示远场阵列响应矢量，c(x，y，z)∈C N×1表示近场阵列响应矢量。可以理解的是，为方便描述，以下所描述的距离和坐标参数均被载波波长归一化，两个RIS单元之间的水平或者垂直距离设为d，则第(n 1，n 2)个RIS单元在x-y-z三维坐标系的坐标可以被表示为：
((n 1-(N 1+1)/2)d，0，(n 2-(N 1+1)/2)d)，
其中n 1＝1，2，L，N 1，n 2＝1，2，L，N 2。则近场阵列响应矢量c(x G，y G，z G)和c(x r，y r，z r)可以分别表示为：
Figure PCTCN2022113557-appb-000002
Figure PCTCN2022113557-appb-000003
其中
Figure PCTCN2022113557-appb-000004
表示基站和 超大规模RIS之间的散射体到第(n 1，n 2)个RIS单元的距离；
Figure PCTCN2022113557-appb-000005
表示超大规模RIS和用户之间的散射体到第(n 1，n 2)个RIS单元的距离。(x G，y G，z G)和(x r，y r，z r)满足：
Figure PCTCN2022113557-appb-000006
Figure PCTCN2022113557-appb-000007
考虑到基站侧已做好波束赋形，即
Figure PCTCN2022113557-appb-000008
即上述接收信号模型可以转化为：
Figure PCTCN2022113557-appb-000009
其中
Figure PCTCN2022113557-appb-000010
表示超大规模RIS近场级联信道，α＝α Gα r表示等效路径增益，
Figure PCTCN2022113557-appb-000011
表示级联信道阵列响应矢量，
Figure PCTCN2022113557-appb-000012
表示基站侧等效的发送信号。
Figure PCTCN2022113557-appb-000013
的具体形式可表示为：
Figure PCTCN2022113557-appb-000014
其中D(n 1，n 2)＝D G(n 1，n 2)+D r(n 1，n 2)表示近场级联信道阵列响应矢量的等效距离。
超大规模RIS波束训练目的是为θ找到合适的码字，使得接收信号的能量最大。在此之前，可基于近场级联信道阵列响应矢量构建近场码本，以匹配近场信道模型。
本实施例中构建近场码本的过程如下：
根据一定的采样步长(即设定步长)
Figure PCTCN2022113557-appb-000015
对所考虑的整个采样范围进行采样，生成两个采样点集合Φ G和Φ r(即第一采样点集合和第二采样点集合)，可分别表示为：
Figure PCTCN2022113557-appb-000016
Figure PCTCN2022113557-appb-000017
遍历Φ G和Φ r，每次从两个采样点集合中各取一个采样点(即第一采样点和第二采样点)，组成一个采样点对，按照如下公式构建近场码字：
Figure PCTCN2022113557-appb-000018
其中
Figure PCTCN2022113557-appb-000019
Figure PCTCN2022113557-appb-000020
和
Figure PCTCN2022113557-appb-000021
分别表示为：
Figure PCTCN2022113557-appb-000022
Figure PCTCN2022113557-appb-000023
需要说明的是，可能不同的采样点对会产生同样的近场码字，故在遍历过程中，每次新生成的近场码字需要和之前生成的所有近场码字不同时才可以加入码本中。在此基础上，根据所有保留下来的近场码字可以得到近场码本W∈C N×L，近场码本中的每一列都可表示一个近场码字，L可表示一共有L个码字。
基于所构建的近场码本，本实施例提出一种近场波束训练方案，即在超大规模RIS和用户之间执行一个训练过程，对近场码本W展开遍历搜索。该过程共包括L个时隙，在第l个时隙中(l∈L)，令超大规模RIS的反射系数等于近场码本中的第l个码字，基站经超大规模RIS向用户发送训练符号，用户接收训练符号，并记录当前所接收的训练符号中能量最大的训练符号对应的码字。最后，用户通过超大规模RIS反馈最优的码字下标(即最优波束的信息)至基站。
本实施例中，提供一种超大规模RIS近场码本构建和波束训练方法。图3为一实施例提供的一种超大规模RIS近场码本构建和波束训练方法的实现示意图，如图3所示，该方法的实现过程如下：
A1.在x-y-z三维坐标系中根据一定的采样步长(即设定步长)对所考虑的整个采样范围进行采样，生成两个采样点集合Φ G和Φ r(即第一采样点集合和第二采样点集合)，其中这两个采样点集合分 别对应于基站和超大规模RIS之间散射体的候选位置和超大规模RIS和用户之间散射体的候选位置；
A2.从两个采样点集合Φ G和Φ r中各取一个采样点(即第一采样点和第二采样点)构成一组采样点对，根据超大规模RIS近场级联信道的阵列响应矢量构建每组采样点对的近场码字，每个近场码字取决于两个采样点分别到超大规模RIS的距离之和(即每组采样点对的近场码字关联于第一采样点到RIS之间的距离与第二采样点到RIS之间的距离之和)；
A3.判断当前新构建的近场码字是否不在当前码本中，基于当前新构建的近场码字不在当前码本中的判断结果，执行A4；基于当前新构建的近场码字在当前码本中的判断结果，执行A5，继续进行判断下一个近场码字是否在当前码本中；
A4.将当前所构建的近场码字放在当前码本中；
A5.跳过；
A6.判断两个采样点集合所组成的采样点对是否都遍历完毕，基于两个采样点集合所组成的采样点对都遍历完毕的判断结果，执行A7；基于两个采样点集合所组成的采样点对未遍历完毕的判断结果，返回执行A2；
A7.根据每组采样点对的近场码字构建近场码本；
A8.对A7中生成的近场码本进行遍历搜索，在第l次搜索中，令超大规模RIS的反射系数等于近场码本中的第l个码字(即当前遍历搜索的近场码字)，基站经超大规模RIS向用户发送训练符号；
A9.用户接收训练符号，并记录当前所接收的训练符号中能量最大的训练符号对应的近场码字；
A10.判断近场码本中的近场码字是否都遍历搜索完毕，基于近场码本中的近场码字都遍历搜索完毕的判断结果，执行A11；基于近场码本中的近场码字未遍历搜索完毕的判断结果，返回执行A8直至遍历完近场码本中的所有近场码字；
A11.用户侧可以得到最优的近场码字，用户侧通过超大规模RIS反馈最优的码字下标至基站。
本申请实施例还提供一种波束训练方法。图4为一实施例提供的另一种波束训练方法的流程图，如图4所示，应用在第一节点(如基站)，本实施例提供的方法包括步骤210、步骤220、步骤230和步骤240。
本实施例中，对所构建的近场码本进行分级波束训练，例如，对所构建码本的波束训练过程进行分级，即分级为多个搜索阶段，每级搜索阶段都对相应的子码本进行遍历搜索最优码字。每级的搜索阶段对应不同的子码本，相应的对应不同的采样范围和采样步长，每级搜索阶段的采样范围和采样步长可以基于上一级搜索阶段的采样范围和采样步长，以当前搜索阶段最优波束对应的近场码字为中心，按照设定比例进行相应的缩小。
在步骤210中，在当前搜索阶段，根据RIS的近场级联信道的阵列响应矢量，构建每组采样点对的近场码字，其中，每组采样点对由当前采样范围内的一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置。
本实施例中，在当前搜索阶段，第一节点根据RIS的近场级联信道的阵列响应矢量，构建每组采样点对的近场码字。
在步骤220中，分别根据每组采样点对的近场码字，通过所述RIS发送训练符号。
本实施例中，第一节点分别根据每组采样点对的近场码字，通过RIS发送训练符号至第二节点。
在步骤230中，接收反馈信息，所述反馈信息包括所述当前采样范围内的最优波束的信息。
本实施例中，第一节点通过RIS接收第二节点所返回的反馈信息，反馈信息包括当前搜索阶段的当前采样范围内的最优波束的信息。
在步骤240中，根据所述最优波束的信息更新所述当前采样范围，并进入下一搜索阶段，返回执行构建每组采样点对的近场码字的操作，直至满足最优波束的搜索停止条件。
本实施例中，第一节点在接收到反馈信息之后，在当前采样范围中以反馈信息中所包含的最优波束信息(即最优波束对应的近场码字)为中心，前后边界到中心的距离均为当前设定采样步长的设定比例，如二分之一，据此根据中心和前后边界所确定的范围即为更新后的当前采样范围。在此基础上，将上述更新后的当前采样范围作为下一搜索阶段的采样范围，此时还可以将设定比例的当前设定步长(该设定比例的范围为大于0且小于1，具体数值可根据实际情况进行灵活设定)作为下一搜索阶段的设定采样步长，进入下一搜索阶段，返回执行构建每组采样点对的近场码字的操作，直至满足最优波束的搜索停止条件。搜索停止条件可以理解为将所构建的近场码本进行分级搜索直至分级完毕，得到最优波束的条件。
在一实施例中，步骤240，包括：
步骤2410：将当前采样范围设置为以最优波束对应的近场码字为中心，且前后边界到所述中心的距离均为当前搜索阶段的设定采样步长的二分之一。
本实施例中，在当前搜索阶段，根据最优波束的信息更新当前采样范围可以包括，第一节点将当前采样范围设置为以最优波束对应的近场码字为中心，且前后边界到中心的距离均为当前搜索阶段的设定采样步长的二分之一。
在一实施例中，在步骤210之前，该方法还包括：
步骤201：确定当前采样范围内的第一采样点集合和第二采样点集合，第一采样点集合中包括多个第一采样点，第二采样点集合中包括多个第二采样点。
在一实施例中，步骤201，包括：
步骤2011：根据当前搜索阶段的设定采样步长对当前采样范围进行采样，得到第一采样点集合和第二采样点集合。
在一实施例中，在根据最优波束的信息更新当前采样范围之后、返回执行构建每组采样点对的近场码字的操作之前，还包括：
步骤241：按照设定比例减小当前搜索阶段的设定采样步长，作为下一搜索阶段的设定采样步长。
本实施例中，可以理解的是，在下一搜索阶段的采样范围逐渐变小，即小于当前搜索阶段采样范围的情况下，下一搜索阶段的设定采样步长可以设置为小于当前搜索阶段的设定采样步长，故设定比例范围可以为大于0且小于1。例如，在根据最优波束的信息更新当前采样范围之后，按照设定比例减小当前搜索阶段的设定采样步长，作为下一搜索阶段的设定采样步长。
在一实施例中，在步骤210之后，该方法包括：
步骤211：根据当前采样范围内的每组采样点对的近场码字，构建当前采样范围对应的码本。
在一实施例中，每组采样点对的近场码字关联于当前采样范围内第一采样点到RIS之间的距离与第二采样点到RIS之间的距离之和。
在一实施例中，步骤220，包括：
步骤2210：遍历当前采样范围内每组采样点对的近场码字，令RIS的反射系数等于当前遍历的近场码字，并通过RIS基于反射系数发送训练符号。
以下通过不同实施例对分级波束训练方法进行示例性说明。
上述实施例中基于遍历搜索的近场波束训练方法可能会造成较大的波束训练开销，为避免该情况，本实施例进一步提出分级近场码本和对应的分级波束训练方法。实现过程如下：
图5为一实施例提供的一种近场码本和分级近场码本的示意图。如图5所示，在分级波束训练方法中，可以将整个波束训练过程分为K级搜索阶段，每级都对相应的子码本进行最优波束的遍历搜索，不同级搜索阶段对应的子码本的采样范围和设定采样步长逐渐变小。第1级子码本的采样范围与上述实施例中近场码本的采样范围一致即：
Figure PCTCN2022113557-appb-000024
其中设定采样步长设为Δ 1＝AΔ，其中A为一个大于1的标量。
之后，在第k个阶段，首先根据当前搜索阶段的采样范围和设定采样步长生成两个采样点集合
Figure PCTCN2022113557-appb-000025
和
Figure PCTCN2022113557-appb-000026
基于此根据上述实施例中近场码本的构建方法生成对应的第k级子码本W k，之后对W k进行遍历搜索以得到第k级子码本对应的最优近场码字，即当前搜索阶段k的最优近场码字，可表示为
Figure PCTCN2022113557-appb-000027
基于第k级搜索阶段得到的最优近场码字，我们对第k+1级(即下一级搜索阶段)的采样范围进行更新，以
Figure PCTCN2022113557-appb-000028
表示第k+1级搜索阶段的采样范围为例：
Figure PCTCN2022113557-appb-000029
其中
Figure PCTCN2022113557-appb-000030
为第k级的设定采样步长。同时，可将设定采样步长更新为第k级的设定采样步长的δ(0＜δ＜1)倍，即Δ k+1＝δΔ k作为第k+1级搜索阶段的设定采样步长。K级搜索阶段完成后，用户将在第K级子码本中搜索到的最优近场码字下标作为反馈信息通过RIS反馈至基站。通过上述方法，分级码本中每级子码本的码本尺寸都可远小于近场码本，故相比基于遍历搜索的波束训练而言，分级波束训 练可大幅降低波束训练开销。
图6为一实施例提供的一种近场码本和分级近场码本的系统可达速率性能对比的示意图图7为一实施例提供的一种近场码本和分级近场码本的波束训练开销对比的示意图。如图6所示，本申请实施例所提供的两种近场码本构建(即近场码本和分级近场码本)和对应的波束训练方法相比远场波束训练在超大规模RIS辅助的近场通信中可以达到更优的系统可达速率性能，可见，本实施例所提出的近场码本波束训练可比相关技术中的远场码本波束训练在系统可达速率性能上大幅提升约50％。在此基础上，为降低波束训练开销，本实施例还进一步提出分级近场码本和分级搜索方法，逐级缩小搜索的采样范围和搜索的采样步长，如图7所示，分级近场码本的波束训练开销远小于近场码本的波束训练开销，相比近场码本遍历搜索可大幅降低近场波束训练开销约90％。
本实施例中，一种超大规模RIS分级近场码本构建和对应的分级波束训练方法。图8为一实施例提供的一种分级近场码本构建和分级波束训练方法的实现示意图，如图8所示，该方法实现过程如下：
B1.将整个波束训练过程分为K级搜索阶段(对应K级子码本)，初始化采样范围(即将第1级搜索阶段的采样范围设置为上述实施例中近场码本的整个采样范围)和设定采样步长(可以理解的是，第1级搜索阶段的设定采样步长可以设为较大的)；
B2.在第k个搜索阶段，根据当前的采样范围和设定采样步长生成两个采样点集合
Figure PCTCN2022113557-appb-000031
和
Figure PCTCN2022113557-appb-000032
B3.按照上述实施例中步骤A2-A7所述方法构建生成第k级子码本(此处对此不作具体描述)；
B4.按照上述实施例中步骤A8-A11所述方法得到第k级子码本对应的最优近场码字(此处对此不作具体描述)；
B5.根据B4得到的第k级子码本对应的最优近场码字更新当前的采样范围，即当前的采样范围变为以当前最优近场码字为中心，前后各取当前设定采样步长的一半(即将当前采样范围设置为以当前最优波束对应的近场码字为中心，且前后边界到中心的距离均为当前搜索阶段的设定采样步长的二分之一)；
B6.更新当前设定采样步长为原来的δ倍，0＜δ＜1；
B7.判断K级搜索阶段是否全部完成，基于K级搜索阶段全部完成的判断结果，执行B8；基于K级搜索阶段未全部完成的判断结果，返回执行B2，直至完成K级搜索；
B8.用户侧可以从第K级子码本中得到最优的近场码字，用户通过RIS反馈最优的码字下标至基站。
本申请实施例还提供一种波束训练方法。图9为一实施例提供的又一种波束训练方法的流程图，如图9所示，该方法可应用在第二节点，如用户终端，本实施例提供的方法包括步骤310、步骤320和步骤330。
在步骤310中，通过RIS接收训练符号，其中，所述训练符号根据采样范围内每组采样点对的近场码字发送，所述近场码字根据RIS的近场级联信道的阵列响应矢量构建，每组采样点对由一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置。
在步骤320中，根据接收所述训练符号的能量确定最优波束。
在步骤330中，发送反馈信息，所述反馈信息包括所述采样范围内最优波束的信息。
本实施例中，首先，第二节点通过RIS接收训练符号，其中训练符号为第一节点根据采样范围内每组采样点对的近场码字通过RIS所发送的。然后，第二节点根据所接收训练符号的能量确定最优波束，即选取能量最高的训练符号对应的近场码字为最优波束。最后，第二节点通过RIS发送反馈信息至基站，其中反馈信息包括采样范围内所得到的最优波束的信息。
在一实施例中，步骤320，包括：
步骤3210：将接收到的能量最大的训练符号对应的近场码字作为最优波束。
本实施例中，在第二节点根据接收训练符号的能量确定最优波束的过程中，将接收到的能量最大的训练符号对应的近场码字作为最优波束。示例性的，第二节点子接收训练符号的过程中，记录能量最大的训练符号对应的近场码字；在每次接收到训练符号时，判断当前所接收到的训练符号的能量与记录的训练符号的最大能量之间的大小，若当前所接收到的训练符号的能量大于当前记录的训练符号的能量，则选取当前所接收到的训练符号对应的近场码字作为当前的最优波束，替换掉之前所记录的近场码字；否则保持记录的近场码字不变以等待下次的判断。
本申请实施例还提供一种波束训练装置。图10为一实施例提供的一种波束训练装置的结构示意图。 如图10所示，所述波束训练装置包括：
第一构建模块410，设置为根据RIS的近场级联信道的阵列响应矢量，构建每组采样点对的近场码字，其中，每组采样点对由一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置；
第一发送模块420，设置为分别根据每组采样点对的近场码字，通过所述RIS发送训练符号；
第一接收模块430，设置为接收反馈信息，所述反馈信息包括最优波束的信息。
本实施例的波束训练装置，第一节点根据RIS近场级联信道的阵列响应矢量构建近场码字，在此基础上根据近场码字通过RIS发送训练符号至第二节点以得到反馈信息，获取最优波束，根据近场码字可以构建匹配近场信道模型的近场码本，实现在超大规模RIS辅助的近场通信中基于近场码本的波束训练，提高波束训练在系统的可达速率，以及降低波束训练的性能损失。
在一实施例中，该装置还包括：
第一码本构建模块，设置为根据每组采样点对的近场码字构建码本，所述码本包括多个不重复的近场码字。
在一实施例中，该装置还包括：
第一采样点集合确定模块，设置为在根据RIS的近场级联信道的阵列响应矢量，构建每组采样点对的近场码字之前，确定第一采样点集合和第二采样点集合，所述第一采样点集合中包括多个第一采样点，所述第二采样点集合中包括多个第二采样点。
在一实施例中，采样点集合确定模块包括：
第一采样单元，设置为根据设定采样步长对采样范围进行采样，得到所述第一采样点集合和所述第二采样点集合。
在一实施例中，每组采样点对的近场码字关联于所述第一采样点到所述RIS之间的距离与所述第二采样点到所述RIS之间的距离之和。
在一实施例中，第一发送模块420包括：
第一遍历单元，设置为遍历每组采样点对的近场码字，令所述RIS的反射系数等于当前遍历的近场码字，并通过所述RIS基于所述反射系数发送训练符号。
本实施例提出的波束训练装置与上述实施例提出的波束训练方法属于同一构思，未在本实施例中详尽描述的技术细节可参见上述任意实施例，并且本实施例具备与执行波束训练方法相同的有益效果。
本申请实施例还提供一种波束训练装置。图11为一实施例提供的一种波束训练装置的结构示意图。如图11所示，所述波束训练装置包括：
第二构建模块510，设置为在当前搜索阶段，根据智能超表面RIS的近场级联信道的阵列响应矢量，构建每组采样点对的近场码字，其中，每组采样点对由当前采样范围内的一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置；
第二发送模块520，设置为分别根据每组采样点对的近场码字，通过所述RIS发送训练符号；
第二接收模块530，设置为接收反馈信息，所述反馈信息包括所述当前采样范围内的最优波束的信息；
更新模块540，设置为根据所述最优波束的信息更新所述当前采样范围，并进入下一搜索阶段，返回执行构建每组采样点对的近场码字的操作，直至满足最优波束的搜索停止条件。
本实施例的波束训练装置，通过分级近场码本和对应的分级波束训练，能够在近场码本的基础上进一步降低波束训练的开销。
在一实施例中，更新模块540，包括：
范围设置单元，设置为将所述当前采样范围设置为以所述最优波束对应的近场码字为中心，且前后边界到所述中心的距离均为当前搜索阶段的设定采样步长的二分之一。
在一实施例中，该装置还包括：
第二采样点集合确定模块，设置为在根据智能超表面RIS的近场级联信道的阵列响应矢量，构建每组采样点对的近场码字之前，确定所述当前采样范围内的第一采样点集合和第二采样点集合，所述第一采样点集合中包括多个第一采样点，所述第二采样点集合中包括多个第二采样点。
在一实施例中，第二采样点集合确定模块，包括：
第二采样单元，设置为根据当前搜索阶段的设定采样步长对所述当前采样范围进行采样，得到所述第一采样点集合和所述第二采样点集合。
在一实施例中，该装置还包括：
采样步长确定模块，设置为在根据所述最优波束的信息更新所述当前采样范围之后、返回执行构建每组采样点对的近场码字的操作之前，按照设定比例减小当前搜索阶段的设定采样步长，作为所述下一搜索阶段的设定采样步长。
在一实施例中，该装置还包括：
根据所述当前采样范围内的每组采样点对的近场码字，构建所述当前采样范围对应的码本。
第二码本构建模块，设置为根据所述当前采样范围内的每组采样点对的近场码字，构建所述当前采样范围对应的码本。
在一实施例中，每组采样点对的近场码字关联于所述当前采样范围内所述第一采样点到所述RIS之间的距离与所述第二采样点到所述RIS之间的距离之和。
在一实施例中，第二发送模块520，包括：
第二遍历单元，设置为遍历所述当前采样范围内每组采样点对的近场码字，令所述RIS的反射系数等于当前遍历的近场码字，并通过所述RIS基于所述反射系数发送训练符号。
本实施例提出的波束训练装置与上述实施例提出的波束训练方法属于同一构思，未在本实施例中详尽描述的技术细节可参见上述任意实施例，并且本实施例具备与执行波束训练方法相同的有益效果。
本申请实施例还提供一种波束训练装置。图12为一实施例提供的一种波束训练装置的结构示意图。如图12所示，所述波束训练装置包括：
符号接收模块610，设置为通过智能超表面RIS接收训练符号，其中，所述训练符号根据采样范围内每组采样点对的近场码字发送，所述近场码字根据智能超表面RIS的近场级联信道的阵列响应矢量构建，每组采样点对由一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置；
波束确定模块620，设置为根据接收所述训练符号的能量确定最优波束；
信息发送模块630，设置为发送反馈信息，所述反馈信息包括所述采样范围内最优波束的信息。
本实施例的波束训练装置，第二节点通过接收第一节点发送的训练符号，根据所接收训练符号的能量能够确定最优波束，并将最优波束的信息通过RIS反馈至第一节点。
在一实施例中，波束确定模块620，包括：
波束确定单元，设置为将接收到的能量最大的训练符号对应的近场码字作为最优波束。
本实施例提出的波束训练装置与上述实施例提出的波束训练方法属于同一构思，未在本实施例中详尽描述的技术细节可参见上述任意实施例，并且本实施例具备与执行波束训练方法相同的有益效果。
本申请实施例还提供了一种通信节点，该通信节点可以是第一节点，例如第一节点可以为基站、接入点或传输点等。图13为一实施例提供的一种第一节点的硬件结构示意图，如图13所示，本申请提供的第一节点，包括存储器720、处理器710以及存储在存储器720上并可在处理器710上运行的计算机程序，处理器710执行所述程序时实现上述的波束训练方法。
第一节点还可以包括存储器720；该第一节点中的处理器710可以是一个或多个，图13中以一个处理器710为例；存储器720设置为存储一个或多个程序；所述一个或多个程序被所述一个或多个处理器710执行，使得所述一个或多个处理器710实现如本申请实施例中所述的波束训练方法。
第一节点还包括：通信装置730、输入装置740和输出装置750。
第一节点中的处理器710、存储器720、通信装置730、输入装置740和输出装置750可以通过总线或其他方式连接，图13中以通过总线连接为例。
输入装置740可设置为接收输入的数字或字符信息，以及产生与第一节点的用户设置以及功能控制有关的按键信号输入。输出装置750可包括显示屏等显示设备。
通信装置730可以包括接收器和发送器。通信装置730设置为根据处理器710的控制进行信息收发通信。
存储器720作为一种计算机可读存储介质，可设置为存储软件程序、计算机可执行程序以及模块，如本申请实施例所述波束训练方法对应的程序指令/模块(例如，附图10所示的波束训练装置中的第一构建模块410、第一发送模块420和第一接收模块430；或者附图11所示的波束训练装置中的第二构建模块510、第二发送模块520、第二接收模块530和更新模块540)。存储器720可包括存储程序区和存储数据区，其中，存储程序区可存储操作系统、至少一个功能所需的应用程序；存储数据区可存储根据第一节点的使用所创建的数据等。此外，存储器720可以包括高速随机存取存储器，还可以包括非易失性存储器，例如至少一个磁盘存储器件、闪存器件、或其他非易失性固态存储器件。在一些实例中，存储器 720可进一步包括相对于处理器710远程设置的存储器，这些远程存储器可以通过网络连接至第一节点。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
本申请实施例还提供了一种通信节点，该通信节点可以是第二节点，例如第二节点可以为用户终端(User Equipment，UE)。图14为一实施例提供的一种第二节点的硬件结构示意图，如图14所示，本申请提供的第二节点，包括存储器820、处理器810以及存储在存储器820上并可在处理器810上运行的计算机程序，处理器810执行所述程序时实现上述的波束训练方法。
第二节点还可以包括存储器820；该第二节点中的处理器810可以是一个或多个，图14中以一个处理器810为例；存储器820设置为存储一个或多个程序；所述一个或多个程序被所述一个或多个处理器810执行，使得所述一个或多个处理器810实现如本申请实施例中所述的波束训练方法。
第二节点还包括：通信装置830、输入装置840和输出装置850。
第二节点中的处理器810、存储器820、通信装置830、输入装置840和输出装置850可以通过总线或其他方式连接，图14中以通过总线连接为例。
输入装置840可设置为接收输入的数字或字符信息，以及产生与第二节点的用户设置以及功能控制有关的按键信号输入。输出装置850可包括显示屏等显示设备。
通信装置830可以包括接收器和发送器。通信装置830设置为根据处理器810的控制进行信息收发通信。
存储器820作为一种计算机可读存储介质，可设置为存储软件程序、计算机可执行程序以及模块，如本申请实施例所述波束训练方法对应的程序指令/模块(例如，附图12所示的波束训练装置中的符号接收模块610、波束确定模块620和信息发送模块630)。存储器820可包括存储程序区和存储数据区，其中，存储程序区可存储操作系统、至少一个功能所需的应用程序；存储数据区可存储根据第二节点的使用所创建的数据等。此外，存储器820可以包括高速随机存取存储器，还可以包括非易失性存储器，例如至少一个磁盘存储器件、闪存器件、或其他非易失性固态存储器件。在一些实例中，存储器820可进一步包括相对于处理器810远程设置的存储器，这些远程存储器可以通过网络连接至第二节点。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
本申请实施例还提供了一种通信系统。图15为一实施例提供的一种通信系统的结构示意图，如图15所示，该通信系统包括：RIS 910、第一节点920以及第二节点930。
本实施例中，第一节点920根据RIS 910的近场级联信道的阵列响应矢量，构建每组采样点对的近场码字，并分别根据每组采样点对的近场码字，通过RIS 910发送训练符号至第二节点930。第二节点930通过RIS 910接收训练符号，并根据接收训练符号的能量确定最优波束，将最优波束的信息作为反馈信息通过RIS 910发送至第一节点920。第一节点920接收反馈信息。
本实施例提供的一种通信系统可以设置为执行上述任意实施例提供的波束训练方法，具备相应的功能和有益效果。
本申请实施例还提供一种存储介质，所述存储介质存储有计算机程序，所述计算机程序被处理器执行时实现本申请实施例中任一所述的波束训练方法。该波束训练方法，包括：根据智能超表面RIS的近场级联信道的阵列响应矢量，构建每组采样点对的近场码字，其中，每组采样点对由一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置；分别根据每组采样点对的近场码字，通过所述RIS发送训练符号；接收反馈信息，所述反馈信息包括最优波束的信息。该波束训练方法，包括：在当前搜索阶段，根据智能超表面RIS的近场级联信道的阵列响应矢量，构建每组采样点对的近场码字，其中，每组采样点对由当前采样范围内的一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置；分别根据每组采样点对的近场码字，通过所述RIS发送训练符号；接收反馈信息，所述反馈信息包括所述当前采样范围内的最优波束的信息；根据所述最优波束的信息更新所述当前采样范围，并进入下一搜索阶段，返回执行构建每组采样点对的近场码字的操作，直至满足最优波束的搜索停止条件。该波束训练方法，包括：通过智能超表面RIS接收训练符号，其中，所述训练符号根据采样范围内每组采样点对的近场码字发送，所述近场码字根据智能超表面RIS的近场级联信道的阵列响应矢量构建，每组采样点对由一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置；根据接收所述训练符号的能量确定最优波 束；发送反馈信息，所述反馈信息包括所述采样范围内最优波束的信息。
本申请实施例的计算机存储介质，可以采用一个或多个计算机可读的介质的任意组合。计算机可读介质可以是计算机可读信号介质或者计算机可读存储介质。计算机可读存储介质例如可以是，但不限于：电、磁、光、电磁、红外线、或半导体的系统、装置或器件，或者任意以上的组合。计算机可读存储介质的更具体的例子(非穷举的列表)包括：具有一个或多个导线的电连接、便携式计算机磁盘、硬盘、随机存取存储器(Random Access Memory，RAM)、只读存储器(Read Only Memory，ROM)、可擦式可编程只读存储器(Erasable Programmable Read Only Memory，EPROM)、闪存、光纤、便携式CD-ROM、光存储器件、磁存储器件、或者上述的任意合适的组合。计算机可读存储介质可以是任何包含或存储程序的有形介质，该程序可以被指令执行系统、装置或者器件使用或者与其结合使用。
计算机可读的信号介质可以包括在基带中或者作为载波一部分传播的数据信号，其中承载了计算机可读的程序代码。这种传播的数据信号可以采用多种形式，包括但不限于：电磁信号、光信号或上述的任意合适的组合。计算机可读的信号介质还可以是计算机可读存储介质以外的任何计算机可读介质，该计算机可读介质可以发送、传播或者传输用于由指令执行系统、装置或者器件使用或者与其结合使用的程序。
计算机可读介质上包含的程序代码可以用任何适当的介质传输，包括但不限于：无线、电线、光缆、无线电频率(Radio Frequency，RF)等等，或者上述的任意合适的组合。
可以以一种或多种程序设计语言或其组合来编写用于执行本申请操作的计算机程序代码，所述程序设计语言包括面向对象的程序设计语言，诸如Java、Smalltalk、C++，还包括常规的过程式程序设计语言，诸如“C”语言或类似的程序设计语言。程序代码可以完全地在用户计算机上执行、部分地在用户计算机上执行、作为一个独立的软件包执行、部分在用户计算机上部分在远程计算机上执行、或者完全在远程计算机或服务器上执行。在涉及远程计算机的情形中，远程计算机可以通过任意种类的网络，包括局域网(LAN)或广域网(WAN)，连接到用户计算机，或者，可以连接到外部计算机(例如利用因特网服务提供商来通过因特网连接)。
以上所述，仅为本申请的示例性实施例而已，并非用于限定本申请的保护范围。
本领域内的技术人员应明白，术语用户终端涵盖任何适合类型的无线用户设备，例如移动电话、便携数据处理装置、便携网络浏览器或车载移动台。
一般来说，本申请的多种实施例可以在硬件或专用电路、软件、逻辑或其任何组合中实现。例如，一些方面可以被实现在硬件中，而其它方面可以被实现在可以被控制器、微处理器或其它计算装置执行的固件或软件中，尽管本申请不限于此。
本申请的实施例可以通过移动装置的数据处理器执行计算机程序指令来实现，例如在处理器实体中，或者通过硬件，或者通过软件和硬件的组合。计算机程序指令可以是汇编指令、指令集架构(Instruction Set Architecture，ISA)指令、机器指令、机器相关指令、微代码、固件指令、状态设置数据、或者以一种或多种编程语言的任意组合编写的源代码或目标代码。
本申请附图中的任何逻辑流程的框图可以表示程序步骤，或者可以表示相互连接的逻辑电路、模块和功能，或者可以表示程序步骤与逻辑电路、模块和功能的组合。计算机程序可以存储在存储器上。存储器可以具有任何适合于本地技术环境的类型并且可以使用任何适合的数据存储技术实现，例如但不限于只读存储器(Read-Only Memory，ROM)、随机访问存储器(Random Access Memory，RAM)、光存储器装置和系统(数码多功能光碟(Digital Video Disc，DVD)或光盘(Compact Disk，CD)等。计算机可读介质可以包括非瞬时性存储介质。数据处理器可以是任何适合于本地技术环境的类型，例如但不限于通用计算机、专用计算机、微处理器、数字信号处理器(Digital Signal Processing，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、可编程逻辑器件(Field-Programmable Gate Array，FGPA)以及基于多核处理器架构的处理器。

Claims (20)

  1. 一种波束训练方法，包括：
    根据智能超表面RIS的近场级联信道的阵列响应矢量，构建多组采样点对中每组采样点对的近场码字，其中，所述每组采样点对由一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置；
    分别根据所述每组采样点对的近场码字，通过所述RIS发送训练符号；
    接收反馈信息，所述反馈信息包括最优波束的信息。
  2. 根据权利要求1所述的方法，还包括：
    根据所述每组采样点对的近场码字构建码本，所述码本包括多个不重复的近场码字。
  3. 根据权利要求1所述的方法，在所述根据智能超表面RIS的近场级联信道的阵列响应矢量，构建多组采样点对中每组采样点对的近场码字之前，还包括：
    确定第一采样点集合和第二采样点集合，所述第一采样点集合中包括多个第一采样点，所述第二采样点集合中包括多个第二采样点。
  4. 根据权利要求3所述的方法，其中，所述确定第一采样点集合和第二采样点集合，包括：
    根据设定采样步长对采样范围进行采样，得到所述第一采样点集合和所述第二采样点集合。
  5. 根据权利要求1所述的方法，其中，所述每组采样点对的近场码字关联于所述第一采样点到所述RIS之间的距离与所述第二采样点到所述RIS之间的距离之和。
  6. 根据权利要求1所述的方法，其中，所述分别根据所述每组采样点对的近场码字，通过所述RIS发送训练符号，包括：
    遍历所述每组采样点对的近场码字，令所述RIS的反射系数等于当前遍历的近场码字，并通过所述RIS基于所述反射系数发送训练符号。
  7. 一种波束训练方法，包括：
    在当前搜索阶段，根据智能超表面RIS的近场级联信道的阵列响应矢量，构建多组采样点对中每组采样点对的近场码字，其中，所述每组采样点对由当前采样范围内的一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置；
    分别根据所述每组采样点对的近场码字，通过所述RIS发送训练符号；
    接收反馈信息，所述反馈信息包括所述当前采样范围内的最优波束的信息；
    根据所述最优波束的信息更新所述当前采样范围，并进入下一搜索阶段，返回执行构建多组采样点对中每组采样点对的近场码字的操作，直至满足最优波束的搜索停止条件。
  8. 根据权利要求7所述的方法，其中，所述根据所述最优波束的信息更新所述当前采样范围，包括：
    将所述当前采样范围设置为以所述最优波束对应的近场码字为中心，且前后边界到所述中心的距离分别为所述当前搜索阶段的设定采样步长的二分之一。
  9. 根据权利要求7所述的方法，在所述根据智能超表面RIS的近场级联信道的阵列响应矢量，构建多组采样点对中每组采样点对的近场码字之前，还包括：
    确定所述当前采样范围内的第一采样点集合和第二采样点集合，所述第一采样点集合中包括多个第一采样点，所述第二采样点集合中包括多个第二采样点。
  10. 根据权利要求9所述的方法，其中，所述确定所述当前采样范围内的第一采样点集合和第二采样点集合，包括：
    根据所述当前搜索阶段的设定采样步长对所述当前采样范围进行采样，得到所述第一采样点集合和所述第二采样点集合。
  11. 根据权利要求8所述的方法，在所述根据所述最优波束的信息更新所述当前采样范围之后、返回执行构建多组采样点对中每组采样点对的近场码字的操作之前，还包括：
    按照设定比例减小所述当前搜索阶段的设定采样步长，作为所述下一搜索阶段的设定采样步长。
  12. 根据权利要求7所述的方法，还包括：
    根据所述当前采样范围内的每组采样点对的近场码字，构建所述当前采样范围对应的码本。
  13. 根据权利要求7所述的方法，其中，所述每组采样点对的近场码字关联于所述当前采样范围内所述第一采样点到所述RIS之间的距离与所述第二采样点到所述RIS之间的距离之和。
  14. 根据权利要求7所述的方法，其中，所述分别根据所述每组采样点对的近场码字，通过所述RIS发送训练符号，包括：
    遍历所述当前采样范围内每组采样点对的近场码字，令所述RIS的反射系数等于当前遍历的近场码字，并通过所述RIS基于所述反射系数发送训练符号。
  15. 一种波束训练方法，包括：
    通过智能超表面RIS接收训练符号，其中，所述训练符号根据采样范围内多组采样点对中的每组采样点对的近场码字发送，所述近场码字根据RIS的近场级联信道的阵列响应矢量构建，所述每组采样点对由一个第一采样点和一个第二采样点组成，所述第一采样点为第一节点和所述RIS之间散射体的候选位置，所述第二采样点为所述RIS和第二节点之间散射体的候选位置；
    根据接收所述训练符号的能量确定最优波束；
    发送反馈信息，所述反馈信息包括所述采样范围内最优波束的信息。
  16. 根据权利要求15所述的方法，其中，所述根据接收所述训练符号的能量确定最优波束，包括：
    将接收到的能量最大的训练符号对应的近场码字作为最优波束。
  17. 一种第一节点，包括存储器、处理器以及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时实现如权利要求1-14中任一项所述的波束训练方法。
  18. 一种第二节点，包括存储器、处理器以及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时实现如权利要求15-16中任一项所述的波束训练方法。
  19. 一种通信系统，包括：智能超表面RIS、如权利要求17所述的第一节点以及如权利要求18所述的第二节点。
  20. 一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时实现如权利要求1-16中任一所述的波束训练方法。
PCT/CN2022/113557 2021-11-19 2022-08-19 波束训练方法、第一节点、第二节点、通信系统及介质 WO2023087819A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111399156.8A CN116155332A (zh) 2021-11-19 2021-11-19 波束训练方法、第一节点、第二节点、通信系统及介质
CN202111399156.8 2021-11-19

Publications (1)

Publication Number Publication Date
WO2023087819A1 true WO2023087819A1 (zh) 2023-05-25

Family

ID=86349380

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/113557 WO2023087819A1 (zh) 2021-11-19 2022-08-19 波束训练方法、第一节点、第二节点、通信系统及介质

Country Status (2)

Country Link
CN (1) CN116155332A (zh)
WO (1) WO2023087819A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117335848B (zh) * 2023-11-06 2024-04-16 国家工业信息安全发展研究中心 一种超大规模mimo空间非平稳信道的波束训练方法

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111245492A (zh) * 2020-01-10 2020-06-05 北京邮电大学 基于接收功率排序的联合波束训练和智能反射面选择方法
CN111917448A (zh) * 2020-08-13 2020-11-10 深圳大学 一种毫米波通信的波束训练方法、装置、系统及存储介质
CN112564752A (zh) * 2020-11-13 2021-03-26 西安电子科技大学 一种优化稀疏天线激活可重构智能表面辅助通信方法
CN112887002A (zh) * 2021-01-13 2021-06-01 之江实验室 一种智能反射面辅助通信的波束域信道角度估计方法
CN113179232A (zh) * 2021-04-22 2021-07-27 南通大学 一种基于深度学习的无源智能反射表面的信道估计方法
CN113489553A (zh) * 2021-07-06 2021-10-08 东南大学 一种智能反射面反射系数与偏置电压关系的测量方法
CN113765581A (zh) * 2021-09-27 2021-12-07 北京理工大学 基于压缩感知与波束对齐的ris快时变信道估计方法

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111245492A (zh) * 2020-01-10 2020-06-05 北京邮电大学 基于接收功率排序的联合波束训练和智能反射面选择方法
CN111917448A (zh) * 2020-08-13 2020-11-10 深圳大学 一种毫米波通信的波束训练方法、装置、系统及存储介质
CN112564752A (zh) * 2020-11-13 2021-03-26 西安电子科技大学 一种优化稀疏天线激活可重构智能表面辅助通信方法
CN112887002A (zh) * 2021-01-13 2021-06-01 之江实验室 一种智能反射面辅助通信的波束域信道角度估计方法
CN113179232A (zh) * 2021-04-22 2021-07-27 南通大学 一种基于深度学习的无源智能反射表面的信道估计方法
CN113489553A (zh) * 2021-07-06 2021-10-08 东南大学 一种智能反射面反射系数与偏置电压关系的测量方法
CN113765581A (zh) * 2021-09-27 2021-12-07 北京理工大学 基于压缩感知与波束对齐的ris快时变信道估计方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
MOHAMED A. ELMOSSALLAMY; HONGLIANG ZHANG; LINGYANG SONG; KARIM G. SEDDIK; ZHU HAN; GEOFFREY YE LI: "Reconfigurable Intelligent Surfaces for Wireless Communications: Principles, Challenges, and Opportunities", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, 201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA, NY 14853, 3 May 2020 (2020-05-03), 201 Olin Library Cornell University Ithaca, NY 14853 , XP081658426 *

Also Published As

Publication number Publication date
CN116155332A (zh) 2023-05-23

Similar Documents

Publication Publication Date Title
CN113726383B (zh) 一种智能反射面辅助的无线通信系统
CN102164374B (zh) 波束搜索处理方法、装置和系统
KR101576915B1 (ko) 낮은 복잡도의 공동의 유출 억압 기법을 사용하는 통신 시스템
US20230198605A1 (en) Relay-aided intelligent reconfigurable surfaces
JP2007159130A (ja) 分散型アンテナ移動通信システムにおける上り受信方法及び装置
WO2023087819A1 (zh) 波束训练方法、第一节点、第二节点、通信系统及介质
JP2022546279A (ja) 無線通信を使用した環境情報の生成
JP2010166316A (ja) Ｍｉｍｏ通信システム
CN113196679A (zh) 深度学习特征波束成形
WO2021068194A1 (zh) 天线信号处理模型的训练方法、装置、天线及存储介质
US11881916B2 (en) Precoding method and apparatus
CN116056118A (zh) 基于主被动混合智能超表面的无线通信传输方法及系统
WO2021088501A1 (zh) 一种码本确定方法及装置
CN114499612A (zh) 波束赋形的处理方法、装置及可读存储介质
CN103259580B (zh) 一种多数据流波束赋形向量确定方法及装置
KR101251958B1 (ko) 증폭 전달 기법을 사용하는 다중 안테나 릴레이 시스템에서 전처리 장치 및 방법
CN112152689A (zh) 波束发送控制方法、装置及发送端
US9847799B2 (en) Method of providing communication service and apparatus for performing the method
CN117880817B (zh) 无人机轨迹与波束成形向量确定方法、装置和电子设备
KR102140758B1 (ko) 상관관계 추정 메커니즘을 가지는 컴퓨팅 시스템 및 그의 동작 방법
CN113852972B (zh) 一种基于波束共享的高速移动终端波束调度方法
WO2023030020A1 (zh) 信号中继、指示方法、装置、中继节点、第一网元及介质
CN115442828A (zh) 一种基站、智能反射面和用户之间的关联匹配方法
WO2024027652A1 (zh) 传输参数确定方法、装置、网络侧设备及介质
WO2023125424A1 (zh) 解码方法、设备及可读存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22894361

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2022894361

Country of ref document: EP

ENP Entry into the national phase

Ref document number: 2022894361

Country of ref document: EP

Effective date: 20240328