XL-MIMO near-field CSI estimation method based on spatial geometric distributed processing

Abstract

The invention discloses an XL-MIMO near-field CSI estimation method based on spatial geometric distributed processing, which comprises the following steps that: a single sub-array in a base station firstly estimates a part of CSI by using an uplink pilot signal, the part of CSI comprises a direction angle and time delay of one or more propagation paths in a single sub-array channel, and then the part of CSI is estimated by using the estimated time delay information of the single sub-array; the method comprises the following steps of: performing orthogonal decoupling on multipath receiving signals of other sub-arrays, estimating direction angles of one or more propagation paths in channels of other sub-arrays, and finally, jointly estimating distance information of one or more propagation paths by a base station by using the estimated direction angles in different sub-arrays on the basis of a spatial geometrical relationship, and reconstructing a full-dimensional XL-MIMO near-field channel. The invention provides an XL-MIMO system full-dimensional near-field CSI estimation method, and particularly solves the problem that the calculation complexity of obtaining full-dimensional near-field CSI of a super-large-scale MIMO system is too high.

Description

Technical Field

[0001] The invention relates to an XL-MIMO near-field CSI estimation method based on spatial geometric distributed processing, belonging to the technical field of full-dimensional near-field CSI estimation. Background Art

[0002] In the sixth generation of wireless communication systems, the requirements for regional throughput and spectral efficiency will be greatly improved, which is not supported by the fifth generation of wireless communication technology. In order to achieve higher spectral efficiency, ultra-large-scale multiple-input multiple-output (XL-MIMO) is considered to be a promising technology for future wireless communication systems. In future wireless communication systems, base stations (BS) use thousands or even more antennas, which can provide stronger beamforming gain and higher spectral efficiency. In order to fully utilize the performance advantages brought by XL-MIMO, it is necessary to obtain channel state information (CSI) in advance. However, the greatly increased number of antennas will lead to high computational complexity overhead. Since the near-field characteristics in the XL-MIMO system cannot be ignored, how to estimate the near-field CSI under limited computational complexity conditions needs to be studied and explored. Therefore, the estimation of near-field CSI has become an important problem that needs to be solved in the XL-MIMO system.

[0003] The CSI estimation method based on compressed sensing is a mainstream algorithm suitable for near-field CSI acquisition in XL-MIMO systems. In XL-MIMO systems, the near-field channel is modeled as a spherical wavefront, in which the signals received by different receiving antennas have phase differences caused by different transmission distances. The base station uses a full sampling dictionary in the polar coordinate domain to detect sparse multipaths in the near-field channel, and extracts multiple paths in sequence through an iterative algorithm. However, the algorithm based on compressed sensing needs to establish a dictionary dimension that is proportional to the cube of the number of antennas, which will lead to unacceptable computational complexity overhead in actual wireless communication systems. In addition, the greatly increased number of antennas will lead to high hardware costs. In order to ensure the low-cost requirements of actual wireless communication systems, the BS will deploy a hybrid precoding architecture, connecting a large number of antennas only through a small number of RF chains, resulting in a reduction in the dimension of the received signal, making it difficult to obtain full-dimensional CSI. Therefore, the acquisition of full-dimensional near-field CSI in hybrid architecture ultra-large-scale MIMO systems has become a bottleneck.

[0004] Preliminary research has been conducted on the acquisition of full-dimensional near-field CSI in hybrid-architecture ultra-large-scale MIMO systems. The currently widely adopted strategy is to evenly divide the full-dimensional array into several sub-arrays, each of which independently processes the received signal and reconstructs part of the CSI information. The BS reconstructs the full-dimensional CSI through the partial CSI estimated by several sub-arrays. However, in actual XL-MIMO systems, the number of sub-arrays is still large, which will lead to high computational complexity. In addition, multiple sub-arrays independently process the received signal, ignoring the correlation between sub-arrays and the near-field characteristics of the full-dimensional array. The estimation accuracy of the full-dimensional CSI is limited and cannot meet the requirements of low computational complexity and high-precision full-dimensional near-field CSI estimation in actual hybrid-architecture XL-MIMO systems.

[0005] In summary, how to obtain high-precision hybrid architecture ultra-large-scale MIMO full-dimensional near-field CSI with lower computational complexity has become a difficult problem that needs to be solved urgently in the sixth-generation mobile communications. Summary of the invention

[0006] The technical problem to be solved by the present invention is: to provide an XL-MIMO near-field CSI estimation method based on spatial geometric distributed processing, in which the central subarray estimates partial CSI with relatively low computational complexity, the other subarrays use the estimation information of the central subarray to perform multipath orthogonal decoupling and estimate the direction angles of multiple paths respectively, and the spatial geometric relationship between multiple subarrays is used to jointly estimate the distance, so as to realize the reconstruction of the full-dimensional near-field CSI of the hybrid architecture XL-MIMO system with relatively low computational complexity, while ensuring the high-precision characteristics of the reconstructed channel.

[0007] The present invention adopts the following technical solutions to solve the above technical problems:

[0008] An XL-MIMO near-field CSI estimation method based on spatial geometry distributed processing comprises the following steps:

[0009] Step 1: In a very large-scale MIMO system, the entire base station array is divided into a plurality of sub-arrays, one of which is selected as a central sub-array and four other sub-arrays, the four other sub-arrays are evenly distributed in the entire array, and the central sub-array has the same dimension as the four other sub-arrays;

[0010] Step 2: The central subarray receives the all-one uplink pilot signal sent by the user equipment through a hybrid precoding architecture, processes and calculates the received signal, and obtains an estimated number of central subarray propagation paths and an estimated result of the directional angle and delay parameters of each path;

[0011] Step 3: for each other subarray, use the estimated number of propagation paths of the central subarray and the delay parameter of each path to perform multipath orthogonal decoupling on the received signals of the other subarrays, and estimate the direction angle of one or more propagation paths in the other subarrays through the decoupled multipath signals;

[0012] Step 4: Using the directional information of each path estimated by the central subarray and the directional information of each path estimated by the four other subarrays, the distance information of all paths in the entire array is jointly estimated based on the spatial geometric relationship, and the path gain is estimated based on the distance information, thereby reconstructing the full-dimensional XL-MIMO near-field channel.

[0013] A computer device comprises a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein when the processor executes the computer program, the steps of the XL-MIMO near-field CSI estimation method based on spatial geometric distributed processing are implemented.

[0014] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the XL-MIMO near-field CSI estimation method based on spatial geometric distributed processing are implemented.

[0015] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0016] 1. The present invention utilizes the phase controllable characteristics of the hybrid precoding architecture and sets the analog phase shift matrix as an angle domain undersampling dictionary by adjusting the phase. It only requires low computational complexity to estimate the number of propagation paths, azimuth, delay and gain of the central subarray, effectively reducing the pilot overhead of CSI estimation and overcoming the impact of the hybrid precoding architecture on channel sparsity.

[0017] 2. After the central subarray estimates the number of multiple paths and delay information in the channel, the present invention utilizes the common delay characteristics between different subarrays to orthogonally decouple the multipath receiving signals of other subarrays, thereby solving the problem of disordered estimation paths of multiple subarrays and large interference between paths.

[0018] 3. After obtaining the angle information of multiple subarrays, the present invention utilizes the spatial geometric relationship between the multiple subarrays, jointly derives the distance information of multiple paths and re-estimates the near-field gain, thereby ensuring the accuracy of full-dimensional near-field channel reconstruction of the ultra-large-scale MIMO system and greatly reducing the computational complexity required for channel reconstruction. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a flow chart of an XL-MIMO near-field CSI estimation method based on spatial geometric distributed processing of the present invention;

[0020] Figure 2 This is a model diagram of the sub-array hybrid architecture XL-MIMO system;

[0021] Figure 3 is the angle domain under-sampling point dictionary map of the hybrid precoding architecture;

[0022] Figure 4 It is a spatial geometric relationship diagram between multiple sub-arrays. DETAILED DESCRIPTION

[0023] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.

[0024] like Figure 1 As shown, the present invention proposes an XL-MIMO near-field CSI estimation method based on spatial geometric distributed processing. The base station divides the entire array into several sub-arrays, each sub-array processes the received signal separately, and after a single sub-array estimates part of the CSI, other sub-arrays use common delay orthogonal decoupling of multipath received signals, and estimate the directional angle information of each sub-array through a small amount of computational complexity. The base station uses the angle information of multiple sub-arrays to jointly estimate the distance information of multiple paths of the entire array based on the spatial geometric relationship, and reconstructs the full-dimensional near-field channel; specifically, the following steps are included:

[0025] (1) The user equipment (UE) sends an uplink pilot signal. The BS single central subarray uses the received uplink pilot signal to estimate the uplink CSI, that is, the information of one or more propagation paths, including but not limited to: the direction angle and delay of the propagation path.

[0026] The strategies for dividing the base station subarrays include but are not limited to the following forms: the entire array is evenly divided into several subarrays of the same dimension, the entire array is divided into five subarrays of the same dimension in the center and four corners, and other subarrays are divided arbitrarily.

[0027] The user needs to send multiple all-one pilot signals. The number of pilots is proportional to the number of sampling points in the angle domain undersampling dictionary, which is the number of subarray antennas divided by 4 in a subarray single RF chain system.

[0028] (2) The other subarrays of the base station use the delay information estimated by the central subarray to perform orthogonal decoupling of the received signal multipaths, obtain interference-free received signals of L paths for each subarray, and estimate the directional angle information of multiple paths in different subarrays through the decoupled multipath signals.

[0029] (3) The BS estimates the distance information of the L paths of the full-dimensional channel based on the azimuth information of the L paths estimated by the central subarray and the azimuth information of the L paths estimated by multiple other subarrays, and reconstructs the full-dimensional near-field channel.

[0030] The CSI used by the base station to reconstruct the full-dimensional near-field channel includes but is not limited to: the azimuth and delay of the L propagation paths estimated by the central subarray, and the distance and gain of the L propagation paths jointly estimated by multiple subarrays.

[0031] Example

[0032] In the hybrid architecture ultra-large-scale MIMO system, the BS is equipped with a uniform planar array (UPA) with N antennas and N r Row N c The whole array is evenly divided into K sub-arrays, and the number of antennas in each sub-array is N. s , N sr Row N sc Column, where N sr =N sc , each subarray antenna is connected to a single RF chain, and the user equipment (UE) adopts a single antenna configuration. In order to detect and transmit data, the base station needs to obtain the full-dimensional near-field CSI and reconstruct the full-dimensional near-field channel H. This embodiment will reconstruct H, including the following steps:

[0033] Step 1: User equipment sends N s / 4 all-one uplink pilot signals, such as Figure 2 As shown, each subarray of the ultra-large-scale UPA receives an uplink pilot signal through a hybrid precoding architecture, where the phase of the hybrid precoding architecture is set to an angle domain undersampling dictionary, such as Figure 3 As shown in the shaded part, the signal model received by the central sub-array at this time is:

[0034]

[0035] in, is the signal received by the central subarray, is the uplink channel between the user equipment and the central subarray, is the angle domain undersampling dictionary designed by phase adjustment for the hybrid precoding architecture, P is the user equipment transmit power, is the noise, M is the number of transmitted subcarriers, and the uplink central subarray channel has the following expression:

[0036]

[0037] Where L is the number of propagation paths, g cen,lis the gain of the lth path in the central subarray channel, a(·) is the central subarray response vector, p(·) is the multi-subcarrier frequency domain steering vector, are the azimuth and elevation angles of the lth path in the central subarray channel, τ l is the delay of the lth path;

[0038] The base station uses the received uplink pilot signal, sets the hybrid precoding angle domain undersampling dictionary, and adopts the compressed sensing algorithm to perform the received signal Y cen After processing and calculation, we get The propagation path The results of group angle and delay parameter estimation, each group of parameter estimation includes

[0039] Step 2: The base station uses the estimated number of central subarray propagation paths and Path delay parameters The received signals of other subarrays (four corner subarrays) are subjected to multipath orthogonal decoupling. Based on the characteristics that different subarrays have the same number of paths and delay parameters, the received signals of the kth (k=1,…,4) subarray can be multipath orthogonally decoupled as follows:

[0040]

[0041] in, represents the decoupled k-th subarray The signal is received by the paths, Receive the signal for the kth subarray, is the kth channel in the kth subarray The gain of the path, is the kth channel in the kth subarray The azimuth and elevation angles of the paths,

[0042] After the k-th subarray multipath orthogonal decoupling receives the signal, k = 1, ..., 4, each path channel uses a compressed sensing algorithm to estimate the azimuth and elevation angle of the path. The decoupling of the paths can ensure that the estimated paths between different subarrays have a matching order and can remove the interference between multipaths. Each set of estimated parameters includes

[0043]

[0044] Step 3: The base station estimates multiple subarrays After obtaining the azimuth and elevation information of each path, the spatial geometric relationship between different sub-arrays is used to combine the estimated angles of the four sub-arrays at the four corners of the overall array and the central sub-array to deduce the Path distance parameter information, such as Figure 4 As shown, the distance parameter information of each path can be expressed as:

[0045]

[0046] in represents the pseudo-inverse operation, represents the distance from the user (direct line of sight path) / scatterer (scattering path) in the lth path to the center of the full-dimensional subarray, d is the spacing between adjacent base station antennas, set to half a wavelength, and the equivalent angle vector It can be expressed as:

[0047]

[0048] in Equivalent Angle Definition Figure 4 As shown, the path gain is re-estimated using the least squares method The BS uses the angle, delay, distance, and gain estimated by the central subarray to reconstruct the full-dimensional near-field channel between the user and the base station:

[0049]

[0050] where w(·) is the near-field response vector of the full-dimensional UPA array, To reconstruct the full-dimensional near-field channel between the user and the base station.

[0051] Based on the same inventive concept, an embodiment of the present application provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the aforementioned XL-MIMO near-field CSI estimation method based on spatial geometric distributed processing are implemented.

[0052] Based on the same inventive concept, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the aforementioned XL-MIMO near-field CSI estimation method based on spatial geometry distributed processing are implemented.

[0053] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0054] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0055] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0056] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0057] The above embodiments are only for illustrating the technical idea of ​​the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the present invention.

Claims

1. A XL-MIMO near-field CSI estimation method based on spatial geometry distributed processing, characterized in that: The steps include: Step 1: In a very large-scale MIMO system, the entire base station array is divided into a plurality of sub-arrays, one of which is selected as a central sub-array and four other sub-arrays, the four other sub-arrays are evenly distributed in the entire array, and the central sub-array has the same dimension as the four other sub-arrays; Step 2: The central subarray receives the all-one uplink pilot signal sent by the user equipment through a hybrid precoding architecture, processes and calculates the received signal, and obtains an estimated number of central subarray propagation paths and an estimated result of the directional angle and delay parameters of each path; Step 3: for each other subarray, use the estimated number of propagation paths of the central subarray and the delay parameter of each path to perform multipath orthogonal decoupling on the received signals of the other subarrays, and estimate the direction angle of one or more propagation paths in the other subarrays through the decoupled multipath signals; Step 4: Using the directional information of each path estimated by the central subarray and the directional information of each path estimated by the four other subarrays, the distance information of all paths in the entire array is jointly estimated based on the spatial geometric relationship, and the path gain is estimated based on the distance information, thereby reconstructing the full-dimensional XL-MIMO near-field channel.

2. The XL-MIMO near-field CSI estimation method based on spatial geometry distributed processing according to claim 1, characterized in that: In step 1, the division strategy of dividing the entire base station array into a number of sub-arrays includes but is not limited to the following forms: evenly dividing the entire base station array into a number of sub-arrays of the same dimension; and dividing the entire base station array into a center and four corners with a total of five sub-arrays of the same dimension, and dividing the other sub-arrays arbitrarily.

3. The XL-MIMO near-field CSI estimation method based on spatial geometry distributed processing according to claim 1, characterized in that: The specific process of step 2 is as follows: The central subarray receives the all-one uplink pilot signal sent by the user equipment through a hybrid precoding architecture. The phase of the hybrid precoding architecture is set to an angle domain undersampling dictionary. The number of all-one uplink pilot signals is one-fourth the number of central subarray antennas. The number of pilots of the all-one uplink pilot signal is proportional to the number of sampling points of the angle domain undersampling dictionary. The signal model received by the central subarray is: Among them, Y cen is the signal received by the central subarray, P is the transmit power of the user equipment, is the angle domain undersampling dictionary of hybrid precoding implemented by analog phase shifter, H cen is the uplink channel between the user equipment and the central subarray, N is the noise, H cen The expression is as follows: Where L is the number of propagation paths, g cen,l is the gain of the lth path in the central subarray channel, a(·) is the central subarray response vector, p(·) is the multi-subcarrier frequency domain steering vector, θ cen,l are the azimuth and elevation angles of the lth path in the central subarray channel, τ l is the delay of the lth path; The central subarray uses the received all-one uplink pilot signal and the angle domain undersampling dictionary implemented by the analog phase shifter in the hybrid precoding architecture to use the compressed sensing algorithm to calibrate the received signal Y cen Processing and calculation are performed to obtain the estimated number of central subarray propagation paths And the estimated results of the direction angle and delay parameters of each path are the first estimated central subarrays. The azimuth, elevation and delay of each path.

4. The XL-MIMO near-field CSI estimation method based on spatial geometry distributed processing according to claim 1, characterized in that: The specific process of step 3 is as follows: Using the estimated number of central subarray propagation paths and the delay parameters for each path The multipath orthogonal decoupling of the received signals of other subarrays is performed. Based on the characteristics that different subarrays have the same number of paths and delay parameters, the received signals of the kth other subarrays are multipath orthogonally decoupled as follows: in, represents the kth subarray decoupled from the other subarrays The path receives the signal, Y k is the kth other subarray received signal, p(·) is the multi-subcarrier frequency domain steering vector, is the kth channel in the other subarray The gain of the path, P is the transmit power of the user equipment, is the angle-domain undersampling dictionary of hybrid precoding implemented by analog phase shifter, a(·) is the central subarray response vector, are the kth channels in the other subarrays. The azimuth and elevation angles of the paths, k = 1,...,4, After the kth other sub-array multipath orthogonal decoupling receives the signal, each path channel uses the compressed sensing algorithm to estimate the azimuth and elevation angle of the path, that is, are the kth estimated subarrays respectively. The azimuth and elevation angles of the paths.

5. The XL-MIMO near-field CSI estimation method based on spatial geometry distributed processing according to claim 1, characterized in that: The specific process of step 4 is as follows: Using the azimuth information of each path estimated by the central subarray and the azimuth information of each path estimated by the four other subarrays, the distance parameter information of all paths of the full array is jointly estimated based on the spatial geometric relationship. And re-estimate the gain of each path using the least squares method According to distance Gain The angle and delay estimated by the central subarray reconstruct the full-dimensional XL-MIMO near-field channel between the user and the base station: in, To reconstruct the full-dimensional near-field channel between the user and the base station, The number of propagation paths estimated for the central subarray, For the The gain of each path, w(·) is the near-field response vector of the full-dimensional subarray, Indicates is the distance from the user / scatterer in the path to the center of the full-dimensional subarray, p(·) is the multi-subcarrier frequency domain steering vector, are the first estimated central subarrays. The azimuth, elevation and delay of each path.

6. A computer device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that: When the processor executes the computer program, the steps of the XL-MIMO near-field CSI estimation method based on spatial geometric distributed processing are implemented as described in any one of claims 1 to 5.

7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the XL-MIMO near-field CSI estimation method based on spatial geometric distributed processing are implemented as described in any one of claims 1 to 5.

Images