Super-large-scale MIMO wireless transmission method based on dual-time-scale CSI and distributed processing architecture

Abstract

The invention discloses a super-large-scale MIMO (Multiple Input Multiple Output) wireless transmission method based on dual-time scale CSI (Channel State Information) and a distributed processing architecture, which comprises the following steps of: firstly, acquiring user position information and statistical channel state information, secondly, based on the user position information, executing user selection in combination with an equivalent distance judgment criterion, and then, by using each distributed processing unit as a criterion, carrying out wireless transmission on each distributed processing unit by using a maximum user rate as a criterion. The method comprises the following steps: designing an analog beam forming matrix according to user statistical channel state information, processing uplink user signals in a distributed processing unit by adopting a linear receiver algorithm, and finally designing a merging coefficient by a central processing unit according to instantaneous channel state information and by taking maximization of a system data transmission rate as a target. And merging the user data from different distributed processing units. According to the method provided by the invention, the spectrum efficiency close to an all-digital architecture can be realized with relatively low complexity and hardware overhead, and the requirement on channel information acquisition can be effectively reduced by cooperatively utilizing instantaneous and statistical channel state information.

Description

Technical Field

[0001] The present invention relates to a very large-scale MIMO wireless transmission method based on dual-time-scale CSI and distributed processing architecture, and belongs to the technical field of wireless communications. Background Art

[0002] As a new generation of mobile communication technology for 2020 and beyond, the fifth generation of mobile communication technology has been commercialized worldwide on a large scale, but the depth and breadth of information interaction are still insufficient to meet the communication business needs of huge traffic and huge connections in the future. As an evolved version of large-scale MIMO technology, ultra-large-scale MIMO technology is expected to support space-division multiplexing transmission of massive users by configuring ultra-large-scale array antennas at the base station end, increasing the number of antennas, and deeply mining the spatial resource dimension, thereby achieving a significant improvement in spectrum efficiency.

[0003] However, the substantial increase in array size will bring huge hardware cost overhead if the all-digital architecture in traditional transceivers is used. Therefore, reducing the number of RF links and RF front-end cost at the base station end while ensuring the system spectrum efficiency has become an important challenge for the actual deployment of ultra-large-scale MIMO. From the perspective of transmission complexity, the substantial increase in array size will increase the near-field range accordingly. If the existing transmission scheme for far-field channel conditions is adopted, it will bring about a large performance loss. In order to compensate for the performance loss caused by the far-field transmission scheme, it is necessary to design the corresponding codebook and transmission scheme for the near-field channel conditions. Therefore, the more complex near-field channel conditions introduce additional complexity for the system's high-efficiency transmission. It can be seen that hardware cost and computational complexity have become bottleneck issues restricting the actual deployment of ultra-large-scale MIMO. In addition, as a prerequisite for transmission design, the channel estimation overhead of ultra-large-scale MIMO systems is large, which will seriously reduce the system transmission rate. Therefore, the cooperative transmission scheme based on instantaneous and statistical channel state information is expected to reduce the channel estimation overhead and improve the system spectrum efficiency.

[0004] To this end, designing a low-cost hardware architecture and low-complexity transmission scheme that adapts to the channel conditions of ultra-large-scale MIMO is conducive to leveraging the advantages of the ultra-large-scale MIMO wireless transmission system. Summary of the invention

[0005] The technical problem to be solved by the present invention is: to provide a very large-scale MIMO wireless transmission method based on dual-time-scale CSI and distributed processing architecture, to schedule users according to their near-field position information, to design DFT beamforming vectors, and to deploy low-complexity receiver solutions in distributed processing units and central processing units to achieve low-cost and low-complexity transmission.

[0006] The present invention adopts the following technical solutions to solve the above technical problems:

[0007] A very large-scale MIMO wireless transmission method based on dual-time-scale CSI and distributed processing architecture, wherein the very large-scale MIMO system includes a base station and several users, wherein the base station is configured with a distributed processing architecture, the base station includes a central processing unit and several distributed units, and there is a bidirectional interactive link between each distributed unit and the central processing unit; the distributed unit includes an antenna subarray, an analog beamformer, a radio frequency link and a digital beamformer connected in sequence; and the following steps are included:

[0008] Step 1: The base station obtains the location information and statistical channel state information of all users according to the uplink pilot signal;

[0009] Step 2: The base station selects users with good spatial separation characteristics to provide uplink transmission services based on the equivalent distance judgment criterion according to the location information of all users obtained in step 1;

[0010] Step 3: Each distributed unit of the base station designs an analog beamforming matrix for each user based on the statistical channel state information of each user selected in step 2 and taking maximizing the user rate as a criterion, so as to configure the analog beamformer;

[0011] Step 4: The digital beamformer of each distributed unit of the base station uses a linear receiver to process the uplink data of each user selected in step 2;

[0012] Step 5: The central processing unit of the base station designs a combining coefficient based on the instantaneous channel state information and takes maximizing the system data transmission rate as a criterion to combine the user data from each distributed unit.

[0013] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0014] 1. The present invention can adapt to ultra-large-scale MIMO near-field channel characteristics at a relatively low hardware cost through a distributed processing architecture and a distributed processing algorithm.

[0015] 2. The present invention only utilizes the user's location information and the statistical channel state information of the selected user, effectively reducing the complexity of the method while achieving a near-all-digital architecture and instantaneous channel state information performance.

[0016] 3. The present invention collaboratively utilizes statistical and instantaneous dual-time-scale channel state information, effectively reducing the requirements for channel information acquisition, avoiding excessive channel information acquisition overhead, and is conducive to improving spectrum efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of the architecture configuration of a very large-scale MIMO system based on a distributed processing architecture proposed by the present invention;

[0018] Figure 2 It is a flow chart of the ultra-large-scale MIMO wireless transmission method based on dual-time-scale CSI and distributed processing architecture of the present invention. DETAILED DESCRIPTION

[0019] 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.

[0020] The present invention proposes a super-large-scale MIMO wireless transmission method based on dual-time-scale CSI and distributed processing architecture, which schedules users according to their near-field position information, designs analog beamforming matrices, and deploys low-complexity receiver solutions in distributed processing units and central processing units, which can achieve low-cost and low-complexity transmission. At the same time, the statistical and instantaneous dual-time-scale channel state information are used in a coordinated manner, which effectively reduces the requirements for channel information acquisition, avoids excessive channel information acquisition overhead, and is conducive to improving spectrum efficiency.

[0021] The configuration of base stations and users is as follows Figure 1 As shown in Figure 1, the ultra-large-scale MIMO system works in time-division duplex mode, where the base station adopts a distributed processing architecture, which is divided into a distributed unit and a central processing unit; the distributed unit includes an antenna array, an analog beamformer, a radio frequency link, and a digital beamformer, and the central processing unit includes a low-cost processor; there is a bidirectional interactive link between the distributed unit and the central processing unit.

[0022] The number of base station antennas is N, which are divided into B sub-arrays, and the number of antennas in each sub-array is N. B , configure N RF RF links, N B The antenna elements are first connected to the analog beamformer and then connected to N RF The analog beamformer can be implemented by a DFT matrix, and the digital beamformer can adopt a linear receiver solution, such as maximum ratio combining or zero-forcing receiver. The user uses a single antenna or multiple antennas.

[0023] The channel between the base station and the user can be modeled based on the Rice channel model as follows:

[0024]

[0025] Among them, κ u represents the Rice factor, and denote the line-of-sight and non-line-of-sight components, r u and represents the distance and azimuth of user u relative to the center of the ultra-large-scale array, represents the near-field array response vector, where the nth element can be modeled as λ represents the carrier wavelength, Indicates that the coordinates are The distance between the user and the nth antenna, represents the spatial correlation matrix corresponding to the non-line-of-sight component, g u is a complex Gaussian vector, β u The large-scale fading of the user channel can be modeled at the distributed unit level as:

[0026]

[0027] Among them, β u,b represents the large-scale fading coefficient from user u to subarray b, which can be further expressed as β u,b ∝1 / r u,b , r u,b represents the distance from user u to the center of subarray b. The non-line-of-sight component of the user channel is described by the spatial correlation matrix, which can be further expressed as:

[0028]

[0029] Among them, θ u represents the azimuth of the non-line-of-sight component, and Respectively represent the angular range and energy distribution of the non-line-of-sight component. Then the covariance matrix of the user channel (i.e., statistical channel state information) can be expressed as:

[0030]

[0031] in,

[0032] Taking distributed unit b as an example, the analog beamforming matrix can be expressed as F b ; The digital beamforming matrix can be written as W b , under the maximum ratio combining receiving algorithm, Under the zero-forcing receiving algorithm, we have Among them, H eq,b =F b H b , H b =[h 1,b ,...,h U,b ].

[0033] Based on the above description, if Figure 2 As shown, the wireless transmission method proposed in the present invention specifically includes the following steps:

[0034] Step 1, channel information acquisition phase: The base station obtains the location information of all users in the cell and statistical channel state information based on the uplink pilot signal, including the distance {r u,b} b＝1,...,B , the azimuth from the user to the center of each subarray and the channel covariance matrix {Θ u,bb} b＝1,...,B , Θ u,bb is the matrix Θ u The (b,b)th submatrix satisfies

[0035] Step 2, user selection stage: The base station selects U users with good spatial separation characteristics to provide uplink transmission services based on the location information of all users in the cell and the equivalent distance judgment criterion, thereby eliminating interference between multiple users.

[0036] Specifically, based on the iterative method, the equivalent distance is initialized as u=1,...,U tot , b = 1, ..., B. For the i-th iteration, the selected users are based on the following criteria:

[0037]

[0038] For the (i+1)th iteration, the equivalent distance needs to be updated first. When the distributed unit adopts a zero-forcing receiver, the interference between different users is eliminated, and the equivalent distance of each user is consistent with the actual distance, that is, When the distributed unit adopts the maximum ratio combining receiver, it can be updated according to the following criteria:

[0039]

[0040] in, C(·) and S(·) represent the Fresnel integral function, I l,u,b and Δ u,b They are defined as:

[0041]

[0042] Among them, θ l represents the spatial azimuth corresponding to the selected DFT beam, which can be assumed to be closest to the actual user azimuth, d A represents the antenna spacing, and λ represents the carrier wavelength. The algorithm is iterated until i=N RF The algorithm terminates when U=N RF .

[0043] Step 3, analog beamforming stage: Based on the maximization of user rate, each distributed processing unit of the base station designs the analog beamforming matrix of each user according to the statistical channel state information of the selected U users for configuring the RF front end.

[0044] Specifically, the RF front end of each distributed processing unit of the base station allocates a DFT beam to each user from the predefined DFT beam. When the distributed processing units are configured with the maximum ratio combining and zero forcing reception algorithms respectively, the approximate expression of the uplink rate of user u can be written as:

[0045]

[0046] in,

[0047]

[0048] ζ u,b represents the signal combining coefficient, η u,b =|ζ u,b | 2 , and The matrices are and It can be seen that when analog beamforming is performed in distributed processing units, the system capacity is mainly related to tr(Q b Θ u,bb ). Therefore, taking maximizing the user rate as the criterion, the simulated beamforming vector (i.e., DFT beam number) of user u in the distributed processing unit b is:

[0049]

[0050] Where U is N B dimensional discrete Fourier matrix.

[0051] Step 4, distributed processing stage: each distributed processing unit of the base station uses a linear receiver to process uplink user data. The base station can use maximum ratio combining or zero forcing reception algorithm to process user data.

[0052] Step 5, central merging stage: The central processing unit of the base station designs the merging coefficient based on the instantaneous channel state information and the maximization of the user rate, and merges the user data from each distributed processing unit.

[0053] Taking user u as an example, the assumed merging coefficient is {ζ u,b} b＝1,...,B , when the distributed processing unit adopts maximum ratio combining or zero-forcing receiver, the combining coefficient satisfy:

[0054] ζ u =c(Σ \u +N u ) -1 χ u

[0055] Where c is the normalization constant, w b,u is the matrix W b The u-th row of the matrix Matrix N u is a diagonal matrix, satisfying N u =diag(‖h u,1 F1‖ 2 ,...,‖h u,B F B ‖ 2 ).

[0056] When the distributed processing unit adopts a ZF receiver, the combining coefficient can also be calculated by the following simplified criteria:

[0057]

[0058] Among them, γ u,b,ZF =|n u,b | 2 ,

[0059] 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 very large-scale MIMO wireless transmission method based on dual-time-scale CSI and distributed processing architecture, characterized in that: The ultra-large-scale MIMO system includes a base station and several users, wherein the base station is configured with a distributed processing architecture, the base station includes a central processing unit and several distributed units, and there is a bidirectional interactive link between each distributed unit and the central processing unit; the distributed unit includes an antenna subarray, an analog beamformer, a radio frequency link and a digital beamformer connected in sequence; and the following steps are included: Step 1: The base station obtains the location information and statistical channel state information of all users according to the uplink pilot signal; Step 2: The base station selects users with good spatial separation characteristics to provide uplink transmission services based on the equivalent distance judgment criterion according to the location information of all users obtained in step 1; Step 3: Each distributed unit of the base station designs an analog beamforming matrix for each user based on the statistical channel state information of each user selected in step 2 and taking maximizing the user rate as a criterion, so as to configure the analog beamformer; Step 4: The digital beamformer of each distributed unit of the base station uses a linear receiver to process the uplink data of each user selected in step 2; Step 5: The central processing unit of the base station designs a combining coefficient based on the instantaneous channel state information and takes maximizing the system data transmission rate as a criterion to combine the user data from each distributed unit.

2. The ultra-large-scale MIMO wireless transmission method based on dual-time-scale CSI and distributed processing architecture according to claim 1, characterized in that: The number of base station antennas is N, which are divided into B antenna subarrays, and the number of distributed units is B; In each distributed unit, the number of antenna units in the antenna subarray is N. B , configure N RF RF links, N B The antenna units are connected to the analog beamformer and connected to N RF RF links, the digital beamformer is connected to each RF link, and there is a two-way interactive link between the digital beamformer and the central processing unit; the analog beamformer is implemented through a discrete Fourier transform matrix, the digital beamformer adopts a maximum ratio combining receiver or a zero-forcing receiver, and the user adopts a single antenna or multiple antennas; The channel between the base station and the user is established based on the Rice channel model, which is expressed as follows: Among them, h u represents the channel between the base station and user u, κ u represents the Rice factor, r u and denote the distance and azimuth of user u relative to the center of the ultra-large-scale array, respectively. represents the near-field array response vector, where the nth element is modeled as represents the imaginary unit, λ represents the carrier wavelength, Indicates that the coordinates are The distance between the user and the nth antenna, g u is a complex Gaussian vector; β u Representing the large-scale fading vector of the user channel, it is modeled at the distributed unit level as: Among them, β u,b represents the large-scale fading coefficient from user u to antenna subarray b, b=1,...,B, Indicates length N B A vector of all 1s; The spatial correlation matrix corresponding to the non-line-of-sight component is expressed as: Among them, θ u represents the azimuth of the non-line-of-sight component, and f(θ u ) represent the angular range and energy distribution of the non-line-of-sight component, b(r u ,θ u ) represents the near-field array response vector, and the superscript H represents the conjugate transpose of the matrix or vector.

3. The ultra-large-scale MIMO wireless transmission method based on dual-time-scale CSI and distributed processing architecture according to claim 2, characterized in that: The specific process of step 2 is as follows: Initialize the equivalent distance to r u,b ,u=1,...,U tot , r u,b represents the distance from user u to the center of antenna subarray b, U tot is the number of all users; users with good spatial separation characteristics are selected based on an iterative method. For the i-th iteration, the selected users are based on the following criteria: Among them, u (i) represents the user selected in the i-th iteration, represents the equivalent distance of the i-th iteration; Before each iteration, the equivalent distance is updated. In the i+1th iteration, if the digital beamformer uses a zero-forcing receiver, the equivalent distance of each user is consistent with the actual distance, that is, represents the equivalent distance of the i+1th iteration; If the digital beamformer uses a maximum ratio combining receiver, the equivalent distance is updated according to the following formula: in, C(·) and S(·) both represent Fresnel integral functions, represents the imaginary unit, represents the azimuth angle from user u to the center of antenna subarray b, θ represents the spatial azimuth angle corresponding to the selected discrete Fourier beam, and d A Indicates the antenna spacing; When i=N RF The iteration ends when the number of selected users U = N RF .

4. The ultra-large-scale MIMO wireless transmission method based on dual-time-scale CSI and distributed processing architecture according to claim 3, characterized in that: The specific process of step 3 is as follows: When the digital beamformer in the distributed unit adopts a maximum ratio combining receiver or a zero forcing receiver, the uplink rate expressions of user u are as follows: where tr(·) represents the trace of the matrix, The definition is as follows: They represent the uplink rate of user u when using the maximum ratio combining receiver and the zero forcing receiver, respectively, p represents the user's transmit power, ζ u,b represents the signal combining coefficient, η u,b =|ζ u,b | 2 , Indicates u,b′ The conjugate of F b and F b′ are the simulated beamforming matrices for the bth and b′th distributed units, Θ u,bb , Θ u,bb′ , Θ u,b′b are the channel covariance matrices Θ for user u respectively. u The (b,b), (b,b′), (b′,b)th submatrices, Θ j,bb′ is the channel covariance matrix Θ of user j j The (b,b′)th submatrix, and The matrices are and The (b,b)th submatrix of Taking maximizing the user rate as the criterion, the simulated beamforming vector (i.e., DFT beam number) of user u in the bth distributed unit is: Where U is N B dimensional discrete Fourier matrix.

5. The ultra-large-scale MIMO wireless transmission method based on dual-time-scale CSI and distributed processing architecture according to claim 4, characterized in that: The specific process of step 5 is as follows: When the digital beamformer in the distributed unit adopts a maximum ratio combining receiver or a zero forcing receiver, the combining coefficient satisfy: g u =c(Σ \u +N u ) -1 x u Among them, u,1 , u,B represents the signal combining coefficient of user u in the central processing unit for distributed units 1 and B, c represents the normalization constant, w b,u represents the digital beamforming vector corresponding to user u in distributed unit b, F b represents the analog beamforming matrix, h j,b represents the channel vector between user j and distributed unit b, N u is a diagonal matrix, satisfying N u =diag(‖h u,1 F1‖ 2 ,...,‖h u,B F B ‖ 2 ), b=1,...,B,h u,1 、h u,B denote the channel vectors between user u and distributed units 1 and B respectively; When the digital beamformer in the distributed unit adopts a zero-forcing receiver, the simplified calculation formula of the combining coefficient is as follows: Among them, γ u,b,ZF =|n u,b | 2 ,

Images