Two-dimensional beam search method based on smart reflector surface

Abstract

This invention belongs to the field of 6G communication technology, specifically a two-dimensional beam search method based on a smart reflector. Assume the hierarchical codebook on the one-dimensional smart reflector adopts an x-ary tree structure. Based on this hierarchical codebook, this invention includes two methods: the JS beam search method and the DWS beam search method. The former integrates the hierarchical codebooks in two directional spaces into a new two-dimensional hierarchical codebook. When each of the two directional spaces uses an x-ary tree hierarchical codebook, the JS scheme synthesizes an x-ary tree hierarchical codebook. 2 - A codebook in the form of a cross-tree; the latter first uses omnidirectional beaming in one direction, and then uses this omnidirectional beaming and codewords in the other direction to form new codebooks, realizing beam search in the second direction space. This invention solves the problem of how to integrate the joint search of two directions of intelligent reflectors, which can be selected or combined for different application scenarios, increasing the design freedom of beam training and meeting the requirements of application scenarios for latency and communication performance.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of 6G communication, and particularly relates to a two-dimensional beam search method based on an intelligent reflecting surface. BACKGROUND

[0002] With the proposal of millimeter wave and terahertz, the scale of the array becomes larger and larger. This is mainly due to two reasons: on the one hand, due to the serious attenuation of high-frequency signals, it is necessary to increase the array scale to provide sufficient beamforming gain; on the other hand, due to the short wavelength of millimeter wave, more antennas can be integrated in the same space range. Especially since the proposal of intelligent reflecting surface, the cost of the antenna array is getting lower and lower, which makes the scale of the antenna array larger and larger. The intelligent reflecting surface is a new type of passive array device proposed in 6G. Since it has no radio frequency link, its manufacturing cost is lower; and since the intelligent reflecting surface is mainly used in the millimeter wave frequency band, the size of the antenna is smaller due to the shorter wavelength, and more antennas can be integrated in a limited range. Since the intelligent reflecting surface has no radio frequency link, it can provide less gain than the relay device under the same array size, which often requires a larger array to provide sufficient reflected power. In summary, due to the low cost and application needs, the array size of the intelligent reflecting surface is generally very large, that is, the number of antennas on the intelligent reflecting surface is often very large. However, a super large scale array will form a pencil wave with very narrow beam width. The pencil wave has the characteristics of large beamforming gain, but small direction coverage range. The smaller coverage range leads to a large time delay in the beam alignment process to search for the accurate beam. In practical applications, we want to retain the advantages of large pencil wave gain, and want to avoid the defect of long search time. At the same time, since the intelligent reflecting surface serves a cascaded channel, not a single-hop channel like a traditional array. Therefore, the beam search work on the intelligent reflecting surface needs to consider both the incident channel direction and the reflected channel direction, which also brings certain difficulties to the beam search work. SUMMARY

[0003] The purpose of the application is to provide a two-dimensional beam search method based on an intelligent reflecting surface, to solve the problem of two-dimensional spatial joint beam search of the intelligent reflecting surface.

[0004] The two-dimensional beam search method based on the intelligent reflecting surface provided by the application includes two kinds: JS beam search method and DWS beam search method according to the hierarchical codebook on the intelligent reflecting surface. Assuming that Ξ is a hierarchical codebook on a one-dimensional intelligent reflecting surface (ULA), the codebook adopts an x-ary tree structure, where x can be any integer greater than 2, the codebook has S layers, and s is the layer number, that is: s = 1, 2, …, S; the s-th layer codebook is denoted as Ξ s , and the code word is:

[0005] (1) JS beam search method (JS scheme for short); Specifically, the hierarchical codebooks in two direction spaces are integrated into a new two-dimensional hierarchical codebook; when the hierarchical codebooks in two direction spaces are x-ary tree respectively, the JS scheme synthesizes x 2 ary tree codebook; Specifically, the JS method is to combine the code words in two direction spaces of each layer into two two-dimensional arrays, and then apply this codebook to search.

[0006] (2) DWS beam search method (DWS scheme for short); Specifically, first use an omnidirectional beam in one direction, then use this omnidirectional beam and the code words in another direction to form a new codebook, and use this new codebook to search one direction space; after the search of one direction space is completed, use the beam finally selected in this direction space and the codebook in the second direction space to form a new two-dimensional codebook, and use this new two-dimensional codebook to search the beam in the second direction space; in this process, the bifurcation number of the two-dimensional hierarchical codebook is consistent with the bifurcation number of the original one-dimensional codebook, that is, the original one-dimensional hierarchical codebook is x-bifurcation, and the two-dimensional hierarchical codebook synthesized by using the DWS scheme is also x-bifurcation.

[0007] The following takes x=2 as an example for introduction, and the relevant conclusions can be easily extended to the case where x is other integers.

[0008] (1) JS beam search method, see Figure 2 ; The specific process is as follows:

[0009] (1) According to the number of array elements of the intelligent reflecting surface, determine the number of layers S of the one-dimensional hierarchical codebook used; if the intelligent reflecting surface has M×N reflecting elements, that is, the intelligent reflecting surface is a square matrix with M rows and N columns, then Where min(·) is the smaller of the two numbers, represents the floor;

[0010] (2) Form two-dimensional code words using one-dimensional code words; the two-dimensional code words of the first layer (i.e. s=1) are denoted as: They can be obtained in the following way:

[0011]

[0012] Where, is the Kronecker product;

[0013] The two-dimensional code words of other layers can be obtained in the same way, and thus a four-ary tree form two-dimensional codebook is obtained;

[0014] (3) Use the two-dimensional code book formed above to search for beams; the specific search process is: first, use the four code words of the first layer to respectively configure the smart reflecting surface, that is, the configuration matrix of the smart reflecting surface is respectively:

[0015]

[0016] wherein diag(·) represents putting a vector on the diagonal of a diagonal matrix; then, the optimal code word is selected according to the power of the received signal; then, the search of the next layer code book is entered, and only the sub-code words of the optimal code word of the previous layer need to be searched in the next layer search; until the code words in the last layer code book are searched.

[0017] (2) The DWS beam search method, see Figure 2 ; the specific process is as follows:

[0018] (1) Determine the direction space to be searched first, if M≤N, search in the elevation direction space corresponding to each column of the smart reflecting surface first;

[0019] (2) Make an omnidirectional code word in another direction space: wherein n = 1, 2, …, N;

[0020] (3) Use the omnidirectional code word and the one-dimensional layered code book in the first search direction to form a two-dimensional layered code book; for example By analogy, each original code word is combined with ξ0 to form a new two-dimensional code word; at this time, the new two-dimensional layered code book is still in the form of a binary tree;

[0021] (4) Use the two-dimensional layered code book to search for beams until the optimal code word is selected.

[0022] (5) Use and the layered code book in the second direction space to form a new two-dimensional layered code book; for example:

[0023] (6) Use the new two-dimensional layered code book to complete the beam search in the second direction space until the optimal code word is selected.

[0024] Further explanation is as follows:

[0025] In the JS scheme, since the two direction spaces are searched synchronously, the width of the intermediate beam in the two direction spaces is the same, and such a result can be used for communication transmission; therefore, the JS scheme can interrupt the search process when the search is not completed, and directly use the beam of the intermediate result for communication; such a property makes the beam search scheme design more flexible.

[0026] In the DWS scheme, since the search in one direction is completed first and then the search in the second direction, the width of the intermediate result beam in the two directions is not consistent, and if the intermediate result beam is used for communication, it will bring greater interference to other users.

[0027] The JS scheme can only realize the branching number structure of the square of an integer, such as a 4-ary tree and a 9-ary tree, and thus is not flexible in designing the branching number of a two-dimensional hierarchical codebook; and the branching number of the two-dimensional hierarchical codebook obtained by the DWS scheme is consistent with that of a one-dimensional hierarchical codebook, and thus has greater flexibility in designing the codebook structure.

[0028] The JS scheme is generally applied to an intelligent reflecting surface in a square matrix case; and the DWS codebook has no such limitation.

[0029] The JS scheme and the DWS scheme are not in an either-or relationship, and the two can be combined to form a hybrid scheme; one scheme can be used first, and the second scheme is switched to complete the subsequent search.

[0030] Advantages of the present application:

[0031] (1) The present application directly uses two one-dimensional hierarchical codebooks to synthesize a two-dimensional hierarchical codebook, and does not use an azimuth angle to describe a two-dimensional direction; thus the design structure is clear, and the use is more convenient;

[0032] (2) The two schemes proposed in the present application can be selected or freely combined according to different application scenarios, thereby increasing the design freedom of the beam training scheme;

[0033] (3) The search scheme in the present application does not depend on the number of array elements of an intelligent reflecting surface; when the number of array elements is unknown or multiple pieces of intelligent reflecting surfaces with different numbers of array elements need to be designed uniformly, the present application can still be used;

[0034] In addition, the two schemes proposed in the present application can also derive other hybrid schemes, and the search can be stopped when the beam search is not completed, and the intermediate result is used for communication, which makes the beam training design more flexible, and in actual application, reasonable selection can be made according to actual requirements to meet the requirements of the application scenario in terms of time delay, communication performance and the like. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1 FIG. 1 is a process schematic diagram of a JS beam scheme.

[0036] Figure 2 FIG. 2 is a process schematic diagram of a DWS beam scheme. DETAILED DESCRIPTION

[0037] The advantages and effects of the present application can be easily understood by those skilled in the art from the disclosure of the specification. Although the description of the present application will be introduced in combination with the preferred embodiments, it does not mean that the features of the present application are limited to the embodiments. On the contrary, the purpose of introducing the application in combination with the embodiments is to cover other options or modifications that can be extended based on the claims of the present application. In order to provide a deep understanding of the present application, many specific details will be included in the following description. The present application can also be implemented without using these details. In addition, in order to avoid confusion or obscure the focus of the present application, some specific details will be omitted in the description. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0038] It should be noted that in the present specification, similar reference numbers and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0039] The terms "first", "second", and the like are only used for differentiation in description, and cannot be understood as indicating or implying relative importance.

[0040] It should be noted that in the description of the present embodiments, unless otherwise explicitly specified and limited, the number of layers and the number of branches used by the codebook can be freely changed; for example, if "a codebook search of three layers in a ternary tree structure" is used in an example, it can be changed to other branch numbers or other search layer numbers according to actual needs when applied.

[0041] For example, in a system with multiple smart reflecting surfaces and multiple users, it is assumed that the user has selected the corresponding smart reflecting surface to serve him, and there is no interference between the smart reflecting surfaces. However, due to the different environments of each user, the noise power in each cascaded channel (the cascaded channel composed of a base station, a smart reflecting surface, and a user) is different. Now each smart reflecting surface needs to search as few times as possible so that the signal-to-noise ratio of the received signal of each user meets the requirements. It is assumed that the smart reflecting surfaces used in this system are all surface arrays with an array element spacing of half a wavelength and an array element number of 100x100.

[0042] First, the ternary tree structure search is the most efficient, so the codebook of the ternary tree structure is used for search as much as possible. Because the JS scheme cannot realize the two-dimensional codebook of the ternary tree structure, the present application uses the DWS scheme to synthesize the two-dimensional codebook of the ternary tree structure for beam search. Second, the DWS scheme does not have the property of being able to interrupt the search and directly use the intermediate result for communication because the two directions are not searched at the same time. Therefore, after a certain number of searches, the JS scheme needs to be converted to complete the subsequent search until the signal-to-noise ratio of the received signal of the user meets the requirements.

[0043] We first use the codebook in the form of ternary tree to complete the three-layer search in each direction, i.e. 18 searches, by using the DWS scheme. Then, it is judged whether the obtained beam satisfies the requirement of the received signal signal-to-noise ratio. If not, one layer is searched by using the JS algorithm, and the above judgment is repeated. If it still does not satisfy, one layer is searched by using the JS algorithm again. At this time, the power concentration degree of the obtained beam has reached the performance limit of the 100×100-dimensional intelligent reflecting surface. If it still does not satisfy the requirement of the received signal signal-to-noise ratio, other intelligent reflecting surfaces need to be jointly assisted.

[0044] Although the present application has been illustrated and described with reference to certain preferred embodiments thereof, it should be understood that the foregoing is intended to cover all modifications and variations of this application that are within the scope of the present application, including the scope of the appended claims. It is intended that changes in form and detail not be regarded as a departure from the spirit, but as necessary to adapt the present application to particular situations.

Claims

1. A smart reflector based two-dimensional beam search method, characterized in that, According to the hierarchical codebook on the intelligent reflecting surface, including two kinds: JS beam search method, DWS beam search method; assuming For a hierarchical codebook on a one-dimensional intelligent reflecting surface, the codebook adopts The structure of a fork tree, wherein Is an arbitrary integer greater than 2, the codebook has Layers, The layer number is, that is: The first Layer codebook is denoted as The code word therein is: Specifically: (1) JS beam search method, referred to as JS scheme; for the case that the number of bifurcations of the hierarchical codebook is an integer square: The two-directional space layered codebook is integrated into a new two-dimensional layered codebook; when the two-directional space respectively uses a layered codebook in the form of a fork tree - a layered codebook in the form of a fork tree, the JS scheme synthesizes a codebook in the form of a fork tree - a layered codebook in the form of a fork tree, the JS scheme synthesizes a codebook in the form of a fork tree; specifically, the JS method divides the code words in each layer of the two-directional space into two two-combined groups to form a codebook on a two-dimensional array, and then applies the codebook to search. (2) DWS beam search method, referred to as DWS scheme; First, an omnidirectional beam is used in one direction, then the omnidirectional beam and the code word in another direction are used to form a new codebook, and the new codebook is used to search the space in one direction, and after the space in one direction is searched, the beam finally selected in the space in one direction is used to form a new two-dimensional codebook with the codebook in the space in the second direction; the new two-dimensional codebook is used to search the beam in the space in the second direction.

2. The two-dimensional beam search method of claim 1, wherein, For the case of The specific flow of the JS beam search method is as follows:​ (1) According to the number of array elements of the intelligent reflecting surface, the number of layers of the one-dimensional layered codebook used is determined If the intelligent reflecting surface has reflective elements, that is, the intelligent reflecting surface is a square matrix with rows and columns, wherein, is the smaller of the two numbers, represents the floor function; (2) One-dimensional code words are used to form two-dimensional code words; the two-dimensional code words of the first layer are respectively denoted as: They are respectively obtained by the following ways: , ， ， ； wherein is a Kronecker product; The two-dimensional code words of other layers can be obtained in the same way, and thus a two-dimensional codebook in the form of a quadtree is obtained; (3) The two-dimensional codebook formed above is used to search the beam, and the specific search process is as follows: first, the four code words in the first layer are used to configure the intelligent reflecting surface respectively, that is, the configuration matrix of the intelligent reflecting surface is respectively: ， wherein, represents putting the vector onto the diagonal of the diagonal matrix; then selecting the optimal code word from among according to the power of the received signal; then entering the search of the next layer codebook, and only searching the sub-code words of the optimal code word of the previous layer during the next layer search; until the code words in the last layer codebook are searched. For other cases of the same applies analogously.

3. The two-dimensional beam search method of claim 2, wherein, For the case of the specific procedure of the DWS beam search method is as follows: (1) determine the direction space to search first, if , search in the elevation direction space corresponding to each column of the smart reflective surface first; (2) making an omni-directional codeword in another direction space: , ; (3) using the omni-directional code word and the one-dimensional layered codebook in the first search direction to form a two-dimensional layered codebook, specifically, By analogy, each of the original code words and combined to form a new two-dimensional code word; at this time, the new two-dimensional layered codebook is still in the form of a binary tree; (4) Use the two-dimensional layered codebook for beam search until a beam is selected ; at which point the search in the first direction spatial domain is complete; (5) Use In combination with the layered codebook in the second direction space, a new two-dimensional layered codebook is formed, specifically, ; and the like. (6) Using the new two-dimensional layered codebook, complete the beam search in the second direction space until the optimal code word is selected ; For other cases of the analogy holds.

4. The two-dimensional beam search method of claim 3, wherein, JS scheme and DWS scheme are used alone or JS scheme and DWS scheme are used in combination, JS scheme is used first, and then DWS scheme is used.

Images