阅读说明：本技术 一种基于毫米波混合预编码结构的低功耗电力通信方法 (Low-power-consumption power communication method based on millimeter wave hybrid precoding structure ) 是由 黎琦 汤显斌 曾琦 李代敏 唐志敏 于 2021-07-19 设计创作，主要内容包括：一种基于毫米波混合预编码结构的低功耗电力通信方法,提出了一种开关与两级移相器级联的混合预编码结构,基于该结构公开一种混合预编码方法,包括以下步骤：给出本发明的开关与两级移相器级联的混合预编码结构,建立基于该结构的混合预编码数学模型；对信道矩阵奇异值分解求解全数字预编码矩阵,生成元素值为1的移相器模拟预编码矩阵；利用线性最小二乘方法优化数字预编码矩阵；利用混合交替最小化优化移相器模拟预编码矩阵；利用坐标下降法优化开关模拟预编码矩阵；利用上述得到的预编码矩阵完成混合预编码。本发明的混合预编码结构与方法,在具有较低的硬件成本的同时,可实现更高的电力通信可靠性。(A low-power consumption electric power communication method based on a millimeter wave hybrid precoding structure provides a hybrid precoding structure with a switch and two-stage phase shifters cascaded, and discloses a hybrid precoding method based on the structure, which comprises the following steps: the method comprises the steps of providing a mixed pre-coding structure formed by cascading a switch and a two-stage phase shifter, and establishing a mixed pre-coding mathematical model based on the structure; solving a full-digital pre-coding matrix by channel matrix singular value decomposition, and generating a phase shifter simulation pre-coding matrix with an element value of 1; optimizing a digital pre-coding matrix by using a linear least square method; simulating a precoding matrix by using a hybrid alternating minimization optimization phase shifter; optimizing a switch simulation pre-coding matrix by using a coordinate descent method; and completing mixed precoding by using the obtained precoding matrix. The hybrid precoding structure and the hybrid precoding method have lower hardware cost and can realize higher power communication reliability.)

1. A low-power-consumption power communication method based on a millimeter wave hybrid precoding structure is characterized in that the hybrid precoding structure is formed by cascading digital precoding and analog precoding, wherein the analog precoding comprises analog switch precoding and analog phase-shifting precoding; the communication method comprises the following concrete implementation steps:

(1) giving a hybrid pre-coding model of a millimeter wave low-power-consumption power communication system and a hybrid pre-coding structure formed by cascading switching two-stage phase shifters, and giving an optimization objective function of the system;

(2) solving a full-digital optimal pre-coding matrix through singular value decomposition of a channel matrix, and then generating a simulation phase-shifting pre-coding matrix with the modulus of 1; optimizing a digital pre-coding matrix by using a linear least square method; optimizing an analog phase-shifting precoding matrix with two-stage phase shifters by using hybrid alternating minimization;

(3) and on the basis of the previous step, optimizing the analog switch precoding matrix by using a coordinate descent method.

2. The millimeter wave hybrid precoding structure-based low-power consumption electric power communication method according to claim 1, wherein the hybrid precoding model has the following optimization objectives:

{F_A，F_S，F_D}＝min(||F_opt-F_AF_SF_D||_F)²

s.t.(||F_AF_SF_D||_F)²＝N_s

[F_S]_m,n∈{0,1}，m∈[1,N_t]，n∈[1,N_RF]

|[F_A]_i,j|＝1，i,j∈[1,N_t]

wherein Fopt represents a dimension of N_t×N_sA full-digital optimal pre-coding matrix; f_AWith a representation dimension of N_t×N_tThe analog phase-shifting precoding matrix of (a); f_SRepresents N_t×N_RFSimulating a switched precoding matrix; f_DRepresents N_RF×N_sA digital precoding matrix; [ F ]_S]_m,nE {0,1} represents that the value of the mth row and nth column elements of the analog switch precoding matrix is 0 or 1; (| | F)_AF_SF_D||_F)²＝N_sRepresenting a power constraint at the transmitting end of the base station; n is a radical of_sFor data stream number, |, denotes the modulus of the complex number; i | · | purple wind_FAn F norm representing a matrix; n is a radical of_tRepresents the number of transmit antennas; n is a radical of_RFRepresenting the number of radio frequency chains;

analog phase-shift precoding matrix F_AA precoding matrix composed of two-stage phase shifters, represented as:

F_A＝[g₁(f₁)^T,…,g_k(f_k)^T,…g_K(f_K)^T]^T，

wherein f is_kA first-stage analog precoding vector connected with a kth sub-antenna array; g_kA second-stage analog precoding vector connected with the kth sub-antenna array; k is the number of the sub antenna arrays; m is the number of antennas in each sub-antenna array, and N_t＝KM。

3. The method according to claim 1, wherein the channel matrix H is subjected to singular value decomposition:

H＝USV^Htaking the first N of the decomposed right singular matrix V_sTaking the column as a full-digital optimal pre-coding matrix, and then randomly generating an analog phase-shifting pre-coding matrix with each element modulus being 1; fixing an analog phase-shifting pre-coding matrix, and optimizing a digital pre-coding matrix by using a linear least square method, specifically solving the following steps: f_D＝(F_A)^HF_opt(ii) a Optimizing an analog phase-shifting precoding matrix with two-stage phase shifters by using a hybrid alternating minimization method; the solving steps are as follows:

(1) fixing a digital pre-coding matrix and an analog switch pre-coding matrix, and optimizing an analog phase-shifting pre-coding matrix; let F_G＝F_SF_D(ii) a Then use { F_A，F_S，F_D}＝min(||F_opt-F_AF_SF_D||_F)²Conversion to solution (| | F)_opt(F_G)^H-F_A||_F)²Analog phase-shift precoding matrix F at minimum value of_AA value of (d);

(2) will (| | F)_opt(F_G)^H-F_A||_F)²The formula is decomposed into K independent subproblems; for the kth sub-antenna array, the sub-problem may be expressed as:

F_A＝min(||[F_opt(F_G)^H]_k-g_k(f_k)^T||_F)²

wherein, [ F ]_opt(F_G)^H]_kExpressed as extraction of [ F_opt(F_G)^H](k-1) M +1 to kM; f. of_kRepresenting a first-stage analog precoding vector connected with a kth sub-antenna array; g_kA second-stage analog precoding vector connected with the kth sub-antenna array;

(3) fixed second-level analog precoding vector g_kTo solve the bestf_k(ii) a Using (| | g)_k||_F)²＝(||f_k||_F)²＝(N_t)²F is to be_A＝min(||[F_opt(F_G)^H]_k-g_k(f_k)^T||_F)²The square of the formula is expanded and then the sum f is removed_kAfter the irrelevant terms, the optimization objective translates into:

f_k＝min R_e{(([F_opt(F_G)^H]_k)^T(g_k)^*)^Hf_k}

wherein R is_e(. -) represents a new matrix obtained by taking the real part of each element of the matrix; (.)^*Expressing the new matrix obtained by conjugating each element of the matrix;

by the above formula f_k＝min R_e{(([F_opt(F_G)^H]_k)^T(g_k)^*)^Hf_kH, respectively optimizing f_kAnd dividing each element by the modulus of the corresponding element to obtain f after unitization_k；

(4) Fixed first-stage analog precoding vector f_kBy unitizing g_k＝[F_opt(F_G)^H]_kf_kTo solve for the optimal g_k。

4. The millimeter wave hybrid precoding structure-based low-power consumption electric power communication method of claim 1, wherein the coordinate descent method optimizes an analog switch precoding matrix as follows:

(1) a fixed digital precoding matrix and an analog phase-shift precoding matrix, let F_E＝(F_A)^HF_optExpressed as an equivalent optimal precoding matrix;

(2) will be the formula | [ F | ]_A]_i,jCarrying out equivalent transformation on 1 and then expanding the square to obtain an optimization target f_kUpper bound (| | F)_E||_F)²-2R_e{tr(F_D(F_E)^HF_S)}+(||F_S||_F)²(ii) a Will optimize the objective f_kIn the upper bound (| F)_E||_F)²Remove and add (| | R)_e{F_E(F_D)^H}||_F)²A precoding matrix F for the analog switch is obtained_SThe specific function of the complete flat mode of (1) is as follows:

F_S＝min(||R_e{F_E(F_D)^H-F_S}||_F)²

thus, the optimal analog switch precoding matrix consists of the matrix R_e{F_E(F_D)^HAnd matrix F_SObtaining, i.e. traversing, the matrix R_e{F_E(F_D)^HJudging all elements of the F, if the element value is more than 0.5, F_SThe corresponding position element value is 1, otherwise it is 0.

5. The method of claim 1, wherein the hybrid precoding structure comprises a digital precoder, a radio frequency link, and a switched two-stage analog phase-shift cascade network; the digital precoder is connected with a switch two-stage analog phase-shift cascade network through a radio frequency link; the switch two-stage analog phase-shifting cascade network comprises a switch network, a first-stage analog phase-shifting network and a second-stage analog phase-shifting network; the switch network is sequentially cascaded with the first-stage analog phase-shifting network and the second-stage analog phase-shifting network.

Technical Field

The invention relates to a low-power-consumption power communication method based on a millimeter wave hybrid precoding structure, and belongs to the technical field of power communication.

Background

With the continuous development of power communication networks, the problems of high reliability operation and high power consumption of power systems gradually step into the field of vision of scholars.

The power communication network is used as an important infrastructure for bearing interaction of the smart grid and future energy information, and the requirements on reliability and low power consumption are higher and higher. Millimeter wave communication (30GHz-300GHz) is widely concerned by people due to rich frequency spectrum resources and ultrahigh transmission rate, so that many students combine millimeter wave communication technology with power communication networks and try to realize more reliable information interaction by using the rich frequency spectrum resources of millimeter waves. However, the rf link in the mm wave communication generates high communication cost, which makes it necessary to consider the power consumption of the system. A hybrid precoding communication scheme that employs a combination of analog and digital is one possible solution to this problem. The hybrid precoding scheme reduces the implementation cost and energy consumption of the system and provides an excellent basic environment for the development of the smart grid.

In summary, the reliability and low-energy-consumption technical research of the communication network in the smart grid is developed, the reliability guarantee problems of the physical layer, the network layer and the like of the power communication network can be solved in a targeted manner, and the method has great advantages in economic cost and network resource utilization and important practical application value. Therefore, how to design an analog network structure and provide a corresponding hybrid precoding method, and achieve higher spectral efficiency by using lower hardware cost and power consumption is one of the key problems to be solved urgently in low-power-consumption power communication.

Disclosure of Invention

The invention aims to solve the problems of reliability and high power consumption of a millimeter wave hybrid precoding structure, and provides a low-power-consumption electric power communication method based on the millimeter wave hybrid precoding structure.

The technical scheme of the invention comprises the following steps that:

s1, providing a hybrid precoding model of the millimeter wave low-power-consumption power communication system and a hybrid precoding structure of a switch two-stage simulation phase shifter cascade, and providing a target optimization function of the hybrid precoding model;

suppose N_sNumber of data streams, N, for information sending end of smart grid_RFRepresenting the number of radio frequency chains, N_tIndicating the number of transmit antennas. And the hybrid precoding mathematical model is represented by utilizing a minimum hybrid precoding matrix residual error criterion, and the target optimization function is as follows:

{F_A，F_S，F_D}＝min(||F_opt-F_AF_SF_D||_F)² (1)

wherein | · | purple sweet_FAn F norm representing a matrix; f_optWith a representation dimension of N_t×N_sA full-digital optimal pre-coding matrix; f_AIs N_t×N_tThe modulus of the element in the matrix is 1; f_SIs N_t×N_RFThe analog switch precoding matrix of (1) or (0) represents the connection and disconnection of the switch, and the disconnection of the phase shifter is controlled by the analog switch precoding matrix; f_DIs represented by N_RF×N_sThe digital precoding matrix of (a);

the three precoding matrices satisfy the power constraint condition of the base station transmitting end, and are expressed as: (| | F)_AF_SF_D||_F)²＝N_s；

The analog phase-shifting precoding matrix is a precoding matrix composed of two-stage phase shifters and can be expressed as F_A＝[g₁(f₁)^T，…，g_K(f_K)^T]^T(ii) a Wherein f is_kA first-stage analog precoding vector connected with a kth sub-antenna array; g_kA second-stage analog precoding vector connected with the kth sub-antenna array; k is the number of sub-antenna arrays, M is the number of antennas in each sub-antenna array, and N_t＝KM。

S2, singular value decomposition is carried out on the channel matrix H: H-USV^HTaking the first N of the right singular matrix V_sColumn as full digital optimum pre-compilingA code matrix, and then a simulation pre-coding matrix with the modulus of 1 is randomly generated; optimizing a digital precoding matrix by using a linear least square method, wherein the concrete solution is as follows: f_D＝(F_A)^HF_opt(ii) a An analog precoding matrix with two-stage phase shifters is optimized using a hybrid alternating minimization method.

The solving steps are as follows:

s21, fixing a digital pre-coding matrix and an analog switch pre-coding matrix, and optimizing an analog phase-shifting pre-coding matrix; let F_G＝F_SF_DThen equation (1) can be converted to solve (| | F)_opt(F_G)^H-F_A||_F)²Analog precoding matrix F at minimum value of_AAnd (4) taking values.

S22, converting the formula (| | F)_opt(F_G)^H-F_A||_F)²Decomposed into K independent sub-problems, which for the K-th sub-antenna array can be expressed as:

F_A＝min(||[F_opt(F_G)^H]_k-g_k(f_k)^T||_F)² (2)

wherein [ F_opt(F_G)^H]_kExpressed as extraction of [ F_opt(F_G)^H]And (k-1) M +1 to kM.

S23, fixing the second-stage phase shifter simulation precoding vector g_kSolving for the best f_k. Using (| | g)_k||_F)²＝(||f_k||_F)²＝(N_t)²Expanding equation (2) and then removing_kAfter the independent item, the optimization objective is converted into

f_k＝minR_e{(([F_opt(F_G)^H]_k)^T(g_k)^*)^Hf_k} (3)

Wherein R is_eThe expression is a new matrix obtained by taking the real parts of the elements of the matrix, (. DEG)^*Expressed as a matrix of elementsThe new matrix obtained after the yoke. Therefore, f can be optimized separately by equation (3)_kAnd dividing each element by the modulus of the corresponding element to obtain f after unitization_k。

S24, fixing the first-stage phase shifter simulation precoding vector f_kBy unitizing g_k＝[F_opt(F_G)^H]_kf_kTo solve for the optimal g_k。

S3, optimizing the analog switch precoding matrix by using a coordinate descent method based on the step S2, the method comprising:

s31, performing equivalent transformation on the fixed digital precoding matrix and the analog phase-shifting precoding matrix according to the formula (1), and then expanding the square to obtain the upper bound (| | F) of the optimization target_E||_F)²-2R_e{tr(F_D(F_E)^HF_S)}+(||F_S||_F)²。

S32, removing the upper bound of the optimization target (| F)_E||_F)²And plus (| | R)_e{F_E(F_D)^H}||_F)²A precoding matrix F for the analog switch can be obtained_SThe specific function of the complete flat mode of (1) is as follows:

F_S＝min(||R_e{F_E(F_D)^H-F_S}||_F)² (4)

thus, the optimal analog switch precoding matrix may be represented by the matrix R_e{F_E(F_D)^HAnd matrix F_SObtaining, i.e. traversing, the matrix R_e{F_E(F_D)^HJudging all elements of the plant, if the elements are more than 0.5, F_SThe corresponding position element value is 1, otherwise it is 0.

The hybrid precoding structure comprises a digital precoder, a radio frequency link and a switch two-stage analog phase-shifting cascade network; the digital precoder is connected with a switch two-stage analog phase-shift cascade network through a radio frequency link; the switch two-stage analog phase-shifting cascade network comprises a switch network, a first-stage analog phase-shifting network and a second-stage analog phase-shifting network; the switch network is sequentially cascaded with the first-stage analog phase-shifting network and the second-stage analog phase-shifting network.

The switching two-stage phase shifter cascade hybrid precoding structure has the advantages that the number of phase shifters required is less than that of fully-connected hybrid precoding, and requirements of lower hardware cost and power consumption of power communication are met; the switch two-stage phase shifter cascade structure provided by the invention realizes dynamic connection between the radio frequency link and the phase shifter through the switch network, so that data information on the radio frequency link can be shared by different antenna array elements, an important basis is provided for improving the reliability of a power communication system, and energy information interaction is better realized.

Drawings

FIG. 1 is a block diagram of a millimeter wave low power consumption power communication hybrid precoding architecture embodying the present invention;

FIG. 2 is a schematic diagram of a cascaded hybrid precoding structure of the switched two-stage phase shifter in FIG. 1;

fig. 3 is a schematic diagram of a hybrid precoding method based on a switching two-stage phase shifter cascade hybrid precoding structure according to an embodiment of the present invention.

Detailed Description

The invention is further explained by a specific implementation mode, the switch two-stage phase shifter cascade hybrid pre-coding structure provided by the invention is used for establishing a mathematical model of a millimeter wave low-power consumption electric power communication system of the structure; the proposed architecture achieves an increase in system gain and energy efficiency while maintaining high reliability.

The embodiment provides a low-power-consumption power communication method based on a millimeter wave hybrid precoding structure, and provides a hybrid precoding scheme based on a switch two-stage phase shifter cascade hybrid precoding structure.

Fig. 1 shows a millimeter wave low power consumption electric power communication hybrid precoding structure, which includes a digital precoder, a radio frequency link, and a switched two-stage analog phase shift cascade network. A transmitting end of the hybrid precoding structure performs hybrid precoding on data streams by using a radio frequency link, a switch and a phase shifter, wherein the radio frequency link forms a digital precoding network, the switch and the phase shifter form an analog precoding network, and the analog precoding network comprises a two-stage cascaded phase shifting network; a switched two-stage analog phase-shifting cascaded network is shown in fig. 2.

The embodiment effectively reduces the hardware cost of the system through the switch and the two-stage phase shifter network, and realizes the dynamic sharing of the radio frequency link and the large-scale antenna system.

The hybrid precoding method based on the switching two-stage phase shifter cascade hybrid precoding structure comprises the following steps:

s1, establishing a switch two-stage phase shifter cascade hybrid pre-coding mathematical model, representing the hybrid pre-coding mathematical model by using a minimum hybrid pre-coding matrix residual error criterion, and optimizing the target as follows:

{F_A，F_S，F_D}＝min(||F_opt-F_AF_SF_D||_F)²

s.t.(||F_AF_SF_D||_F)²＝N_s

[F_S]_m,n∈{0,1}，m∈[1,N_t]，n∈[1,N_RF]

|[F_A]_i,j|＝1，i,j∈[1,N_t] (5)

wherein, F_optWith a representation dimension of N_t×N_sA full-digital optimal pre-coding matrix; f_A、F_S、F_DRespectively representing a dimension of N_t×N_tOf the analog phase-shift precoding matrix, N_t×N_RFAnalog switch precoding matrix and N_RF×N_sA digital precoding matrix; [ F ]_S]_m,nThe value of the ith row and the nth column of the analog switch precoding matrix is 0 or 1, wherein 0 represents that the switch is disconnected, and 1 represents that the switch is connected; (| | F)_AF_SF_D||_F)²＝N_sRepresenting a power constraint at the transmitting end of the base station; suppose N_sData stream number of information sending terminal of intelligent power grid- | non-visual phosphor_FRespectively representing the modulo of the complex number and the F-norm of the matrix.

The analog phase-shifting precoding matrix is a precoding matrix composed of two-stage phase shifters and can be expressed as F_A＝[g₁(f₁)^T,…,g_k(f_k)^T…,g_K(f_K)^T]^TWherein f is_kA first-stage analog precoding vector connected with a kth sub-antenna array; g_kIs a second-stage analog precoding vector connected to the kth sub-antenna array, K is the number of sub-antenna arrays, M is the number of antennas in each sub-antenna array, and N is_t＝KM。

S2, singular value decomposition is carried out on the channel matrix: H-USV^HFront N of right singular matrix V after singular value decomposition of channel matrix H_sThe columns are used as a full digital optimal precoding matrix, and then an analog phase-shifting precoding matrix with the modulus of 1 is randomly generated. Fixing an analog phase-shifting pre-coding matrix, and optimizing a digital pre-coding matrix by using a linear least square method, wherein the concrete solution is as follows: f_D＝(F_A)^HF_opt. An analog phase-shifting precoding matrix with two-stage phase shifters is optimized using a hybrid alternating minimization method.

The solving step comprises the following steps:

s21, fixing the analog digital precoding matrix and the analog switch precoding matrix, and optimizing the analog phase-shifting precoding matrix. From equation (5):

F_A＝min(||F_opt-F_AF_SF_D||_F)²＝min(||F_opt(F_G)^H-F_A||_F)² (6)

wherein F_G＝F_SF_DDimension of (A) is N_t×N_s。

S22, decomposing equation (6) into K independent sub-problems, which can be expressed as:

F_A＝min(||[F_opt(F_G)^H]_k-g_k(f_k)^T||_F)² (7)

wherein, [ F ]_opt(F_G)^H]_kExpressed as extraction of [ F_opt(F_G)^H]Is N in the dimension of (k-1) M +1 to kM_t×N_tA sub-matrix.

S23, fixing the second-stage phase shifter simulation precoding vector g_kSolving for the best f_k(ii) a The property and equation of the trace is solved by utilizing the F norm square of the matrix and the matrix (| | g)_k||_F)²＝(||f_k||_F)²＝(N_t)²After the problem (7) is expanded and the irrelevant items are removed, the optimization target problem is converted into:

f_k＝min R_e{(([F_opt(F_G)^H]_k)^T(g_k)^*)^Hf_k} (8)

wherein R is_eThe expression is a new matrix obtained by taking the real parts of the elements of the matrix, (. DEG)^*Is expressed as a new matrix obtained by conjugating each element of the matrix. Therefore, f can be optimized separately by equation (8)_kAnd dividing each element by the modulus of the corresponding element to obtain f after unitization_k。

S24, fixing the first-stage phase shifter simulation precoding vector f_kBy unitizing g_k＝[F_opt(F_G)^H]_kf_kTo solve for the optimal g_k。

S3, based on the step S2, the method for optimizing the analog switch precoding matrix by using the coordinate descent method is as follows:

a fixed digital precoding matrix and an analog phase-shift precoding matrix, let F_E＝(F_A)^HF_optExpressed as an equivalent optimal precoding matrix; the equation (5) is subjected to equivalent transformation and then the square is expanded, so that the upper bound (| | F) of the optimization target can be obtained_E||_F)²-2R_e{tr(F_D(F_E)^H F_S)}+(||F_S||_F)²(ii) a Will optimize the target upper bound (| F)_E||_F)²Remove and add (| | R)_e{F_E(F_D)^H}||_F)²Obtaining a precoding matrix F for the analog switch_SThe specific function of the complete flat mode of (1) is as follows:

F_S＝min(||R_e{F_E(F_D)^H}-F_S||_F)² (9)

thus, an optimal analog switch precoding matrix, which may be represented by matrix R_e{F_E(F_D)^HIs greater than 0.5, i.e. the matrix R is traversed_e{F_E(F_D)^HJudging all elements of the plant, if the elements are more than 0.5, F_SThe corresponding position element value is 1, otherwise it is 0.