Novel hybrid beam forming system and hybrid precoding method

Abstract

The invention discloses a novel hybrid beamforming system and a hybrid precoding method, and relates to the field of wireless communication, and the system comprises a transmitting end and a receiving end. The transmitting end comprises a baseband coding module, a radio frequency link, a front switch network, a phase shifter network, a rear switch network and a transmitting antenna array which are arranged in sequence; a digital-to-analog converter is arranged in the radio frequency link; the baseband coding module is used for carrying out digital pre-coding on a signal of a transmitting end, and the front switch network, the phase shifter network and the rear switch network are used for carrying out analog pre-coding on the signal of the transmitting end. According to the invention, the flexibility of hardware in hybrid beam forming can be improved, and the system power consumption is reduced.

Description

Technical Field

[0001] The present application relates to the field of wireless communications, and in particular to a novel hybrid beamforming system and a hybrid precoding method. Background Art

[0002] At present, beamforming technology can be divided into three categories: full digital beamforming, analog beamforming and digital-analog hybrid beamforming.

[0003] In the sub-6GHz frequency band, the Multiple Input Multiple Output (MIMO) system often uses the full digital beamforming technology. In essence, full digital beamforming is the practical application of digital signal processing. Its signal shaping weights and phase shifts are completely implemented by digital circuits. Therefore, it is necessary to configure an RF link in front of each antenna array element. The implementation of this solution requires a large number of hardware devices, which will bring significant cost and power consumption.

[0004] Analog beamforming relies entirely on RF domain processing to reduce the number of RF chains. It uses an analog phase shifter network in the analog domain to change the relative phase of signals fed to different antenna array elements, thereby transmitting or receiving the corresponding beam in a specified direction. The analog domain consists of an RF link and several analog phase shifters (PS). However, since it only contains one RF link, it is only suitable for single-user and single-data stream transmission, and cannot fully utilize the available spatial spectrum resources.

[0005] The number of analog phase shifters used in the hybrid beamforming architecture is proportional to the size of the antenna array and the number of RF links, resulting in high system power consumption and low system energy efficiency. The dynamic subarray hybrid beamforming architecture is a hot research architecture. This architecture pioneered the use of a low-power and low-cost RF switch network combined with a phase shifter network in the analog beamforming part. Although the dynamic subarray architecture reduces the number of phase shifters in the system, the mapping structure in which each antenna array element can only be connected to one RF link through one phase shifter is still somewhat fixed, and the power consumption in the Massive MIMO system is relatively high. Secondly, most of the current research focuses on a single problem in the hybrid beamforming hardware, and few studies have considered the impact of phase shifters with quantized phase and DAC / ADC resolution on the spectral efficiency of various hybrid beamforming architectures. In addition, the precoding algorithms of various hybrid beamforming architectures under different channel state information (CSI) are incomplete. Currently, the precoding algorithm design based on partial CSI is mainly aimed at the fully connected architecture and the partially connected architecture. For a more flexible hybrid beamforming architecture, it is necessary to design a corresponding precoding algorithm. Summary of the invention

[0006] The purpose of this application is to provide a novel hybrid beamforming system and a hybrid precoding method, which can improve the flexibility of hardware in hybrid beamforming and reduce system power consumption.

[0007] To achieve the above objectives, this application provides the following solutions:

[0008] In a first aspect, the present application provides a novel hybrid beamforming system, comprising: a transmitting end and a receiving end; the transmitting end comprises a baseband coding module, a radio frequency link, a front switch network, a phase shifter network, a rear switch network and a transmitting antenna array arranged in sequence; a digital-to-analog converter is arranged in the radio frequency link;

[0009] The baseband coding module is used to perform digital precoding on the signal at the transmitting end, and the front switch network, the phase shifter network and the rear switch network are used to perform analog precoding on the signal at the transmitting end.

[0010] In a second aspect, the present application provides a hybrid precoding method, which is applied to the above-mentioned novel hybrid beamforming system, and the hybrid precoding method includes:

[0011] Calculate the spectrum efficiency of the transmitter;

[0012] Calculate the total power consumption of the transmitter;

[0013] determining energy efficiency based on the spectral efficiency and the total power consumption;

[0014] Constructing an optimization problem with the goal of maximizing the energy efficiency;

[0015] The optimization problem is solved based on a block coordinate descent algorithm to obtain a hybrid precoding matrix and a resolution of a digital-to-analog converter in a radio frequency link, thereby completing hybrid precoding of a transmitting end signal.

[0016] According to the specific embodiments provided in this application, this application has the following technical effects:

[0017] (1) By introducing RF switch networks before and after the phase shifter (i.e., the front switch network and the rear switch network), the RF link is dynamically connected with the phase shifter and the antenna array. The front switch network and the rear switch network reduce the number of phase shifters used and can reasonably utilize the RF link according to the amount of transmitted data streams, thereby improving the energy efficiency of the system and reducing power consumption.

[0018] (2) A pre-switch network is inserted between the RF link and the phase shifter network. The pre-switch network can select the output port of the RF link. Its state determines whether the output signal of any RF link passes through the phase shifter connected to it, thereby flexibly controlling the connection between the phase shifter network and the RF link.

[0019] (3) A post-switch network is inserted between the phase shifter network and the antenna array. The post-switch network selects an antenna array element in the antenna array for connection and feeds the RF signal after phase shifting by the phase shifter to the selected antenna array element, thereby flexibly controlling the connection between the antenna array and the phase shifter network. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 A schematic diagram of the structure of a transmitting end of a novel hybrid beamforming system provided in one embodiment of the present application;

[0022] Figure 2 A schematic diagram of a flow chart of a hybrid precoding method provided in an embodiment of the present application;

[0023] Figure 3 A schematic diagram of a process for solving an optimization problem based on a block coordinate descent algorithm provided in an embodiment of the present application;

[0024] Figure 4 Schematic diagram of the relationship between spectrum efficiency and signal-to-noise ratio (SIGNAL-NOISE RATIO, SNR);

[0025] Figure 5 Schematic diagram of the relationship between energy efficiency and SNR;

[0026] Figure 6 is the energy efficiency and the number of RF links N tRF Schematic diagram of the relationship between;

[0027] Figure 7 Schematic diagram of the relationship between energy efficiency and estimation accuracy ξ;

[0028] Figure 8 N for some CSI s Schematic diagram of the relationship between spectrum efficiency and SNR when =1;

[0029] Fig. 9 N for some CSI sSchematic diagram of the relationship between energy efficiency and SNR when =1. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0031] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0032] In an exemplary embodiment, a novel hybrid beamforming system is provided, including: a transmitting end and a receiving end. Figure 1 As shown, the transmitting end includes a baseband coding module 1, a radio frequency link 2, a front switch network 3, a phase shifter network 4, a rear switch network 5 and a transmitting antenna array 6 which are arranged in sequence.

[0033] A digital to analog converter (DAC) is provided in the RF link 2, the front switch network 3 includes multiple front switches, the phase shifter network 4 includes multiple phase shifters, the rear switch network 5 includes multiple rear switches, and the transmitting antenna array 6 includes multiple antenna array elements.

[0034] The baseband coding module 1 is used to perform digital precoding on the signal at the transmitting end, and the pre-switch network 3, the phase shifter network 4 and the post-switch network 5 are used to perform analog precoding on the signal at the transmitting end.

[0035] After digital precoding, the signal enters multiple RF links 2, and a pre-switch network 3 is inserted between the RF link 2 and the phase shifter network 4. The pre-switch network 3 can select the output port of the RF link 2, and its state determines whether the output signal of any RF link 2 passes through the phase shifter connected to it, thereby flexibly controlling the connection between the phase shifter and the RF link 2.

[0036] A post-switch network 5 is inserted between the phase shifter network 4 and the antenna array 6. The post-switch network 5 selects an antenna array element for connection and feeds the RF signal after phase shifting by the phase shifter to the selected antenna array element, thereby flexibly controlling the connection between the antenna array 6 and the phase shifter network 4.

[0037] In an exemplary embodiment, based on the novel hybrid beamforming system proposed above, the present application further provides a hybrid precoding method for iteratively optimizing the digital precoding matrix, the analog precoding matrix and the resolution of the digital-to-analog converter, with the ultimate goal of maximizing the energy efficiency of the transmitter, that is, the ratio of the spectrum efficiency of the transmitter to the total power consumption of the transmitter. Figure 2 The main steps shown are as follows:

[0038] S1: Calculate the spectrum efficiency of the transmitter.

[0039] Consider using Figure 1 The downlink transmission scenario of a single-user MIMO communication system with a novel hybrid beamforming system described in t antennas, the transmitter sends N s The transmitter is equipped with N tRF In the above communication system, suppose the data stream symbol vector to be transmitted is of dimension N s ×1, assuming that it satisfies the widely used normalized signal power, that is:

[0040]

[0041] The transmission signal first passes through the digital precoding (baseband coding) matrix F BB Perform digital signal processing, After that, the simulated precoding matrix F RF To adjust the phase, In this hybrid beamforming system, the simulated precoding matrix F RF It can be expressed as:

[0042] F RF =S 2 F ps S 1 (2)

[0043] in, It is the front switch matrix responsible for connecting the phase shifter with the RF link. To provide an adjustable phase shifter matrix for phase shifting, is the post-switch matrix responsible for connecting the phase shifter to the antenna array element. There is also a constraint on the total power at the transmitting end, namely:

[0044]

[0045] DAC devices in practical applications have certain resolution limitations and introduce quantization errors, and higher-resolution DACs will result in higher power consumption. Considering the use of a linear additive quantization noise model to represent the quantization error, the signal at the DAC output port after digital precoding can be expressed as:

[0046] Q(F BB x)=ΔF BB x+ε b (4)

[0047] ε b is additive quantization noise, which obeys complex Gaussian distribution. It is worth noting that the mainstream DAC resolution in beamforming technology is about 8 bits. At this time, ε b The non-zero elements of the covariance matrix are in the order of 10-5 and can be ignored. Δ represents the multiplication distortion parameter caused by the quantization error of the DAC on each RF link. The diagonal element Δ i,i It can be expressed as:

[0048]

[0049] b i is the resolution of the DAC in the i-th RF chain, assuming that the resolution range of all DACs is the same:

[0050] b i ∈[b min ,b max ],i=1,2,...,N tRF (6)

[0051] The phase shifters used extensively in the analog precoding part do not actually have a high resolution. Assuming that the phase shifters used in the hybrid beamforming system have a quantization resolution q, let Φ represent the phase shift it can provide, then:

[0052]

[0053] In the proposed new hybrid beamforming system, F ps is a diagonal matrix, and the diagonal elements can be expressed as:

[0054]

[0055] Considering the above practical hardware limitations, let the signal received at the receiving end be y. Here, assuming that the CSI is known and complete, y can be expressed as:

[0056]

[0057] Where H is the dimension N r ×N t The channel state information matrix is: ρ is the average received power, n is the additive white Gaussian noise vector, and each element is independent and obeys the same complex Gaussian distribution Assume the decoding matrix of the receiving end is The received signal can be expressed as:

[0058]

[0059] At this time, the spectrum efficiency can be expressed as:

[0060]

[0061] S2: Calculate the total power consumption of the transmitter.

[0062] Here, we use the extended Saleh-Valenzuela channel model, which can more accurately describe the channels in the centimeter wave, millimeter wave, terahertz and other frequency bands. Specifically, the channel matrix It can be expressed as:

[0063]

[0064] Among them, γ is the normalization factor, which can be expressed as L represents the number of multipaths in the channel, α i represents the complex path gain of the i-th multipath, is a diagonal matrix describing the multipath complex gain, and the i-th diagonal element is α i .a ri and a ti They represent the array response vectors of the receiving end and the transmitting end under the i-th multipath, and are matrices A and r and A t The array response vector here is only a function related to the antenna array. Without loss of generality, consider using a uniform planar array (UPA) as the structure of the antenna array in the hybrid beamforming system. Then the array response vector at the transmitting end can be described as

[0065]

[0066] Where k = 2π / λ, λ is the carrier wavelength, d is the spacing between adjacent antenna array elements, which is set to λ / 2, and φ and θ are the departure azimuth and elevation directions of a multipath relative to the antenna array plane, respectively. and are the array element indexes of the antenna array on the y-axis and z-axis respectively, and the total number of transmitting antennas is N t Since the antenna array at the receiving end also adopts the UPA architecture, the array response vector is similar to equation (14). At this time, the total number of receiving antennas is N r .

[0067] In general, the total power consumption of the hybrid beamforming system transmitter consists of digital precoding power consumption, analog precoding power consumption, and RF signal radiation power consumption. For digital precoding power consumption, this part of power consumption is mainly generated by baseband signal processing. Here, it can be assumed that the digital precoding power consumption remains unchanged, and P is used BB To express.

[0068] For the power consumption of the analog precoding part, such as Figure 1 As shown in the figure, after the signal is processed by the baseband, it is first converted into an RF analog signal through the RF link. Then the phase is adjusted through the analog precoding part. The analog precoding part is mainly composed of a front switch network, a phase shifter network and a rear switch network. The front switch can select the output port of the RF link, and its state determines whether the output signal of any RF link passes through the phase shifter connected to it. The rear switch selects the antenna array element for connection, and feeds the RF signal after phase shifting by the phase shifter to the selected antenna array element, and then sends it to the channel after passing through the power amplifier (PA) in front of the antenna array element. In practice, except that the power consumption of the DAC will change with the resolution used, the remaining RF hardware components can be assumed to have fixed power consumption.

[0069] In the above hardware, the power consumption of the PA before the antenna array element is P PA It can be expressed in linear form as the transmit power:

[0070]

[0071] P t represents the transmit power of the antenna array element, ρ Pa Indicates the conversion efficiency of the power amplifier.

[0072] DAC power consumption P DAC is the sampling frequency f s and the converter quality factor FoM, and increases exponentially with the increase of its resolution b, which can be expressed as:

[0073] P DAC =FoM×f s ×2 b =P D ×2 b (16)

[0074] P D is the equivalent power of a single bit of the digital-to-analog converter.

[0075] In this architecture, the connection between the RF link and the antenna array element is determined by the front switch and the rear switch. When a path is activated, the RF link and the antenna array element on the path will consume their rated power. Conversely, if there is no signal path on a RF link or antenna array element, it can be turned off to save power. RF It can be expressed as:

[0076]

[0077] Among them, P PS is the power consumption of a single phase shifter, N ps is the number of phase shifters, P SW is the power consumption of a single RF switch, P RFC is the power consumption of a single RF link, c is a function indicating whether the RF link is connected, C is an index set of connected RF links in the system, and r is a function indicating whether the antenna array elements are connected.

[0078] in And there are:

[0079]

[0080] S in the formula 1i,n and S 2n,i Respectively represent S 1 The sum of the elements in row i and column n of S 2 The element at row n and column i.

[0081] For the RF signal radiation power consumption, it is mainly composed of transmission power. The actual transmission power is mainly provided by the antenna array at the transmitting end. It can be considered that the power provided by a single antenna array element is stable. In the proposed architecture, not all antenna array elements are connected, so the RF signal radiation power consumption P Tr It can be expressed as:

[0082]

[0083] Among them, F RF To simulate the precoding matrix, the subscript F represents the Frobenius norm.

[0084] The total power consumption P can be expressed as:

[0085]

[0086] Among them, P con is the constant power consumption value in the total power consumption, P 1 (S 1 ,b) is the total power consumption and S 1 Power consumption related to b, P 2 (S2 ) is the total power consumption and S 2 Regarding power consumption.

[0087] P con =P BB +P PS N ps +2P SW N ps (twenty two)

[0088]

[0089] in, Indicates the resolution of the DAC on the RF link.

[0090] S3: Determine energy efficiency based on spectral efficiency and total power consumption.

[0091] The optimization goal is to maximize the transmit energy efficiency E of the proposed new hybrid beamforming system, which can well represent the trade-off between power consumption and performance of the communication system. Specifically, the energy efficiency E can be defined as the ratio of the spectrum efficiency R to the total power consumption P:

[0092]

[0093] The spectrum efficiency R is given by equation (12), and the total power consumption P is given by equation (21), then:

[0094]

[0095] S4: Construct an optimization problem with the goal of maximizing energy efficiency.

[0096] This application decouples the optimization problem of the joint transmitter and receiver, and only considers the precoding matrix design of the transmitter. Assuming perfect decoding at the receiver, the index of the precoding matrix of the transmitter can be equivalent to the spectral efficiency by the mutual information I of the signal on the channel:

[0097]

[0098] To further simplify the problem, consider fixing the connection relationship between the phase shifter network and the antenna array, which can be specifically expressed as:

[0099]

[0100] At this time there are:

[0101]

[0102] In summary, considering only the precoding design at the transmitter, the equivalent energy efficiency It can be expressed as:

[0103]

[0104] In fact Figure 1 In the analog precoding part described, the "selection right" of the front switch network and the rear switch network is on the phase shifter network side. The "single-pole multi-throw" type of RF switch is used in the proposed new hybrid beamforming system, which means that each phase shifter can only select one RF link to connect at the front end and one antenna array element to connect at the back end. For the front switch matrix S 1 For example:

[0105] ||s 1i,: || 0 =1,i=1,2...,N ps (31)

[0106] Where S 1i,: Indicates S 1 The i-th row of .

[0107] The optimization problem of maximizing energy efficiency can be formulated as:

[0108]

[0109] in represents the collection space of binary "01" matrices, Indicates F ps The possible value set space of , the phase of its diagonal elements satisfies the constraints in equation (7) and the constant modulus constraint. i is the i-th element in b. From formula (5), we can see that Δ can be expressed as a function of b.

[0110] S5: Solve the optimization problem based on the block coordinate descent algorithm, obtain the hybrid precoding matrix and the resolution of the digital-to-analog converter in the RF link, and complete the hybrid precoding of the transmitter signal. The hybrid precoding matrix includes a digital precoding matrix, a front switch matrix, and an adjustable phase shifter matrix.

[0111] Step S5 specifically includes:

[0112] S51: Divide the optimization problem into two sub-problems, namely the first sub-problem and the second sub-problem.

[0113] It is worth noting that the optimization problem (32) has an objective function in the form of a fraction, the denominator is a function of a discrete variable, and the optimization variable has a non-convex binary constraint, which makes the solution of this problem challenging. Therefore, this application proposes a hybrid precoding algorithm based on the idea of ​​Block Coordinate Descent (BCD) for this problem.

[0114] The core idea of ​​the BCD algorithm is to divide the optimization variables into multiple blocks, and optimize only the variables of one block in each iteration, while keeping the variables of other blocks fixed. The complex optimization problem is converted into a series of sub-problems. The variables in problem (32) are independent of each other, which is more suitable for using the BCD idea for iterative solution, such as Figure 3 shown.

[0115] The independent variables of this optimization problem are mainly F BB , S 1 、F ps and b. It is necessary to consider how to reasonably divide the block variables to simplify the problem, and the key point is how to convert S 1 This application considers the hybrid precoding design problem under the dynamic subarray structure, with the goal of maximizing the spectrum efficiency. The simulation results show that when the spectrum efficiency is used as the optimization objective function, the number of non-zero rows in the simulated precoding matrix is ​​always N. s , indicating that only N is used s RF links, to ensure multi-stream communication, there are N s ≤N tRF , so N s Combined with the proposed new hybrid beamforming system, if the objective function is to maximize the spectral efficiency, S 1 , then |‖c(S 1 )‖0=N s Therefore, the variables can be divided into two groups: F BB , S 1 and F ps As a group, the objective function is to maximize the mutual information of the transmitter, and b is a separate group. Solve the problem with the previous group of variables fixed. At this time, problem (32) can be divided into two sub-problems:

[0116]

[0117] as well as,

[0118]

[0119] S52: The first sub-problem is equivalent to an optimization problem of minimizing the Euclidean distance between the hybrid precoding matrix and the optimal precoding matrix, which is determined as the third sub-problem.

[0120] For the first sub-problem (33), it can be equivalent to minimizing the hybrid precoding matrix and the optimal precoding matrix F opt The optimization problem of the Euclidean distance between is:

[0121]

[0122] Here F optIt can be obtained by performing SVD decomposition on the channel matrix H. In this way, the optimization problem (32) can be equivalent to the iterative solution of the second subproblem (34) and the third subproblem (35). The specific iterative solution algorithm based on BCD is shown in Table 1.

[0123] Table 1

[0124]

[0125]

[0126] S53: Solve the second sub-problem by using a traversal method to obtain the resolution of the digital-to-analog converter in the radio frequency link.

[0127] Considering the solution of the second subproblem (34), the value of the independent variable b is discrete, which means that it is feasible to apply the search method to solve it. In particular, if the traversal method is applied to solve it, since b has N tRF elements, only need to perform (b max -b min +1)×N tRF The number of calculations is repeated, and the energy efficiency under each condition is compared. Finally, the value under the highest energy efficiency is taken as the solution. This is desirable for solving the problem. Therefore, the traversal method is used to solve the second sub-problem (34).

[0128] S54: Solve the third sub-problem using an alternating minimization method to obtain a hybrid precoding matrix.

[0129] As for the solution of the third subproblem (35), it is still difficult to solve the problem directly due to the coupling between the related non-convex constraints of the variables and the matrix. Therefore, we consider using the idea of ​​alternating minimization to solve it, that is, alternately fixing one variable to optimize another. Among the three variables, F BB The constraints are the least, and the solution can be given priority. During analysis, Δ can be approximated as the unit matrix. 1 With F ps , and at the same time impose a semi-unitary constraint on it to facilitate the solution. The third subproblem (35) can be expressed as:

[0130]

[0131] This problem is an orthogonal Pluck problem, and its solution is:

[0132]

[0133] Considering the fixed F BB Solve for S in the case of 1 With F ps , then the optimization problem is:

[0134]

[0135] Among them, F RF =S 2 F ps S 1 , noting that:

[0136]

[0137] The first and second terms of polynomial (39) are both constants, so minimizing polynomial (39) is equivalent to maximizing its last term, which is equivalent to the following problem:

[0138]

[0139] Let F = F BB F opt H =[F 1 ,F 2 ,…,F Nt ],S 1i,: Indicates S 1 The i-th row of S 2i,j For S 2 The element in row i and column j is introduced The objective function of problem (40) can be transformed into:

[0140]

[0141] It is easy to see that for any i, S xi are independent, so they can be decomposed into N ps The same sub-problem:

[0142]

[0143] make Then we have:

[0144]

[0145] Obviously S xi is an element of The remaining elements of the sparse row vector are all zero. We can use the greedy idea to solve problem (42): we can first find ΣF j The element with the largest absolute value in S is indexed as n. xi Re(ΣF j ) is the largest, then Φ i Should meet:

[0146]

[0147] Due to Φ iAs in problem (42), the quantized phase constraint is used, so it is necessary to select the one closest to equation (44) in the phase set as Φ i The value of i ', at this time S xi It can be expressed as:

[0148]

[0149] After all problems of the form (42) are solved, we can get S x , from which S can be recovered 1 With F ps After solving problem (38), the variable value can be fixed and problem (36) can be solved until the condition for jumping out of the iteration is met. After the iteration is completed, F BB Normalize to satisfy the power constraint:

[0150]

[0151] The hybrid precoding method provided in this application has better energy efficiency than existing representative architectures and precoding algorithms under both complete and partial channel state information, and the spectrum efficiency loss is less than 20% compared to the fully connected architecture. The simulation results are as follows: Figure 4-Figure 9 shown. Figure 4 It shows that when the SNR is greater than 0 dB, the spectrum efficiency of the method proposed in this application is substantially higher than the minimum target spectrum efficiency, which is set to 80% of the spectrum efficiency achieved by the OMP-based precoding algorithm under the fully connected architecture; Figure 5 The energy efficiency of each architecture and algorithm under different SNRs is demonstrated. Under all SNRs, the architecture and algorithm proposed in this application perform best in terms of energy efficiency. Figure 6 The relationship between energy efficiency and the number of RF chains shown in shows that the energy efficiency of the algorithm proposed in this application is always the highest in various cases; Figure 7 When ξ=0.6, the energy efficiency of the algorithm proposed in this application is about 3.4 bits / Hz / J, while the energy efficiency under accurate CSI is about 3.7 bits / Hz / J, and the loss is about 8%, indicating that the algorithm proposed in this application is highly robust to CSI estimation errors; Figure 8 In the figure, the spectrum efficiency of the precoding algorithm proposed in this application under some CSIs is higher than the minimum target spectrum efficiency in the figure; Figure 8In the results, the energy efficiency of the proposed algorithm under each SNR in the case of partial CSI is slightly lower than that under full CSI, but the difference is only 0.1 bits / Hz / J, and the result is still significantly higher than the performance of the other baseline algorithms (under full CSI). This is mainly because the new hybrid beamforming system proposed in this application can reasonably utilize the RF link according to the amount of transmitted data streams. At the same time, the new hybrid beamforming system and hybrid precoding method further optimize the resolution of the DAC and reduce the number of phase shifters used through two switches, all of which improve the energy efficiency of the system.

[0152] This application combines a more comprehensive and practical simulation hardware limitation to derive the problem of maximizing energy efficiency in a narrowband single-user scenario, and designs a hybrid precoding algorithm based on complete channel state information, so that the system can achieve a certain balance between power consumption and spectrum efficiency. The low power consumption of the proposed structure and the effectiveness of the algorithm are proved by analyzing the simulation results under the open source channel simulation software NYUSIM channel.

[0153] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0154] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A novel hybrid beamforming system, characterized in that: include: A transmitting end and a receiving end; the transmitting end comprises a baseband coding module, a radio frequency link, a front switch network, a phase shifter network, a rear switch network and a transmitting antenna array arranged in sequence; a digital-to-analog converter is arranged in the radio frequency link; The baseband coding module is used to perform digital precoding on the signal at the transmitting end, and the front switch network, the phase shifter network and the rear switch network are used to perform analog precoding on the signal at the transmitting end.

2. The novel hybrid beamforming system according to claim 1, characterized in that: The pre-switch network includes a plurality of pre-switches, the phase shifter network includes a plurality of phase shifters, the post-switch network includes a plurality of post-switches, and the transmitting antenna array includes a plurality of antenna array elements.

3. A hybrid precoding method, characterized in that: The hybrid precoding method is applied to the novel hybrid beamforming system according to any one of claims 1 to 2, and the hybrid precoding method comprises: Calculate the spectrum efficiency of the transmitter; Calculate the total power consumption of the transmitter; determining energy efficiency based on the spectral efficiency and the total power consumption; Constructing an optimization problem with the goal of maximizing the energy efficiency; The optimization problem is solved based on the block coordinate descent algorithm to obtain the hybrid precoding matrix and the resolution of the digital-to-analog converter in the radio frequency link, thereby completing the hybrid precoding of the transmitting end signal; the hybrid precoding matrix includes a digital precoding matrix, a front switch matrix and an adjustable phase shifter matrix.

4. The hybrid precoding method according to claim 3, characterized in that: The calculation formula of the spectrum efficiency is: Where R is the spectrum efficiency, N s ×N s The identity matrix, N s is the number of data streams sent from the transmitter to the receiver, ρ is the average received power, σ n is the standard deviation of additive white Gaussian noise, W is the decoding matrix at the receiving end, H is the channel state information matrix, S1 is the front switch matrix, S2 is the rear switch matrix, and F ps is the adjustable phase shifter matrix, Δ is the multiplication distortion parameter caused by the quantization error of the digital-to-analog converter on each RF link, and F BB is the digital precoding matrix.

5. The hybrid precoding method according to claim 4, characterized in that: The total power consumption of the novel hybrid beamforming system includes: digital precoding power consumption, analog precoding power consumption and radio frequency signal radiation power consumption.

6. The hybrid precoding method according to claim 5, characterized in that: The calculation formula of the simulated precoding power consumption is: Among them, P RF To simulate the precoding power consumption, P PS is the power consumption of a single phase shifter, N ps is the number of phase shifters, P SW is the power consumption of a single RF switch, P RFC is the power consumption of a single RF link, c is the function of whether the RF link is connected, P D is the equivalent power of a single bit of the digital-to-analog converter, P t is the transmit power of the antenna array element, ρ Pa is the conversion efficiency of the power amplifier, b i is the resolution of the digital-to-analog converter in the i-th RF link, C is the index set of the connected RF links in the system, and r is a function of whether the antenna array element is connected.

7. The hybrid precoding method according to claim 6, characterized in that: The calculation formula of the radio frequency signal radiation power consumption is: Among them, P Tr is the RF signal radiation power consumption, F RF To simulate the precoding matrix, the subscript F represents the Frobenius norm, N t is the number of antenna array elements in the antenna array.

8. The hybrid precoding method according to claim 7, characterized in that: The total power consumption is calculated as follows: Where P is the total power consumption, P BB is the power consumption of digital precoding, P con is the constant power consumption value in the total power consumption, P1(S1,b) is the power consumption related to S1 and b in the total power consumption, and P2(S2) is the power consumption related to S2 in the total power consumption.

9. The hybrid precoding method according to claim 8, characterized in that: The expression of the optimization problem is:

10. The hybrid precoding method according to claim 9, characterized in that: The optimization problem is solved based on the block coordinate descent algorithm to obtain a hybrid precoding matrix, which specifically includes: Divide the optimization problem into two sub-problems, namely a first sub-problem and a second sub-problem; The first sub-problem is equivalent to an optimization problem of minimizing the Euclidean distance between the hybrid precoding matrix and the optimal precoding matrix, which is determined as the third sub-problem; Solve the second sub-problem by using a traversal method to obtain the resolution of the digital-to-analog converter in the radio frequency link; The third sub-problem is solved by using an alternating minimization method to obtain a hybrid precoding matrix.

Images