Design method of low-complexity hybrid beamforming structure in massive antenna system

Abstract

This invention addresses the high complexity and hardware requirements of existing beamforming technologies in large-scale antenna systems by proposing a design method for a hybrid beamforming structure using low-precision phase shifters. This method utilizes a two-step update process to obtain analog and digital beamforming modules, thereby reducing algorithm implementation complexity. The specific process consists of two steps: the first step is to fix the structure of the digital beamforming module to eliminate multi-user interference; the second step is to update the phase of each phase shifter in the analog beamforming module using cyclic coordinate gradient descent based on the digital beamforming structure from the first step. Because the solution of the digital and analog beamforming modules does not require alternating optimization processes, the algorithm complexity is low. Furthermore, this invention can achieve good performance in a low-precision phase shifter network. Furthermore, this invention can meet the different performance requirements of each user while reducing transmit power.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a design method for low-complexity hybrid beamforming using a finite-precision phase shifter in a large-scale antenna system. Background Art

[0002] In multiple-input, multiple-output (MIMO) antenna systems, beamforming technology is often used to improve system performance. In traditional MIMO systems, the base station typically uses a fully digital beamforming structure, where each antenna corresponds to a separate RF link.

[0003] With the advancement of communication technology, massive MIMO systems have been proposed to address the reduced coverage area caused by the increase in communication frequency bands. However, as the number of antennas increases, the number of RF links used in fully digital beamforming structures also increases accordingly, significantly increasing hardware costs and energy consumption, making practical system deployment and application difficult. To address this cost issue, hybrid beamforming structures use low-dimensional digital beamforming modules in conjunction with analog beamforming modules, significantly reducing the number of required RF links.

[0004] Hybrid beamforming architectures can be further categorized as fully connected and partially connected. In a fully connected architecture, each RF link connects to all antennas. In a partially connected architecture, each RF link connects to a separate set of antennas. Therefore, a fully connected architecture achieves better system performance, but at the expense of hardware complexity and power consumption. A partially connected architecture offers a simpler hardware structure, but at the expense of some system performance.

[0005] Furthermore, in practical systems, lower-cost components are often selected while meeting user performance requirements. For example, finite-precision phase shifters are often used. This non-ideal phase can cause system performance loss, so this non-ideal factor should be considered during system design.

[0006] Hangzhou Dianzi University's invention application number 202010975925.3 discloses a design method for a hybrid beamforming structure. This method is applicable to single-user systems with multiple antennas and is used to improve spectral efficiency. The transmitter uses analog beamforming with a partially connected structure, determining the digital beamforming matrix using the least squares method and then the analog beamforming matrix using gradient descent. However, this method is not applicable to multi-user systems.

[0007] Wuhan University's invention, application number 202110813382.X, discloses a low-complexity hybrid beamforming architecture design method and system. This method, targeting a single-user system with multiple antennas, utilizes the EPI algorithm to obtain a hybrid beamforming matrix, achieving a good transmission rate. However, this method is primarily applicable to analog beamforming in fully connected structures and does not consider the limited precision of phase shifters in actual systems. Furthermore, this method is not applicable to multi-user systems. Summary of the Invention

[0008] The technical problem to be solved by the present invention is that in a large-scale antenna system, the existing beamforming technology has high complexity and high hardware requirements.

[0009] In order to solve the above technical problems, the technical solution of the present invention is to provide a design method of a low-complexity hybrid beamforming structure in a large-scale antenna system, characterized by comprising the following steps:

[0010] Step 1: Obtain the signal to be transmitted and channel state information;

[0011] Step 2: Determine the digital beamforming matrix structure and express it as a function of the channel and analog beamforming matrix, as shown in the following equation:

[0012]

[0013] Where: F BB represents the digital beamforming matrix; A = Diag[p1, p2, ..., p K ], Diag[·] represents the diagonal matrix, p k represents the transmission power corresponding to the kth user, k∈{1, 2, ..., K}, K represents the number of users in the system; F RF represents the analog beamforming matrix; H represents the channel matrix; (·) H represents the conjugate transpose of a matrix;

[0014] Step 3: The problem of minimizing the base station transmission power is expressed as:

[0015]

[0016]

[0017]

[0018] Where: ||·|| F represents the F norm of the matrix; BlkDiag[·] represents the block diagonal matrix; (F RF ) i represents the phase shifter vector connected to the i-th RF link, i∈{1, 2, ..., NRF}, N RF Indicates the number of radio frequency links used by the base station; represents the phase set that the phase shifter can take, and B represents the number of quantization bits of the phase shifter; [(F RF ) i ] j represents the phase shifter vector connected to the jth antenna of the i-th RF chain, j∈{1, 2, ..., M}, M represents the number of antennas connected to each RF chain;

[0019] The problem of minimizing the base station's transmit power is solved by using a cyclic coordinate gradient descent method, which updates the phase of each phase shifter in the analog beamforming matrix one by one, iterating until convergence.

[0020] Step 4: Obtain a digital beamforming matrix based on the updated analog beamforming matrix and calculate the transmit power.

[0021] Preferably, in the large-scale antenna system, the number of radio frequency links at the transmitting end is equal to or unequal to the number of transmission data streams.

[0022] Preferably, the digital beamforming module in the hybrid beamforming structure adopts a zero-forcing algorithm.

[0023] Preferably, in step 3, updating the phase of the current phase shifter in the analog beamforming matrix comprises the following steps:

[0024] Step 301: Select the current phase shifter and fix the phases of other phase shifters to be constant;

[0025] Step 302: Get the 2 corresponding to all possible values ​​of the current phase shifter. B an analog beamforming matrix;

[0026] Step 303: The digital beamforming module uses a zero-forcing algorithm to calculate the digital beamforming corresponding to each analog beamforming matrix;

[0027] Step 304: Calculate 2 based on the possible values ​​of the phase shifter. B possible transmit power values;

[0028] Step 305: Select the phase corresponding to the minimum transmit power, and complete the update of the selected phase shifter.

[0029] Preferably, the analog beamforming structure in the hybrid beamforming structure may be a partially connected analog beamforming structure.

[0030] Another technical solution of the present invention is to provide an application of the design method of a low-complexity hybrid beamforming structure in the above-mentioned large-scale antenna system, which is characterized by being applicable to multi-user scenarios and being able to meet the different service quality requirements of each user.

[0031] This paper addresses the high complexity and hardware requirements of existing beamforming technologies in large-scale antenna systems by proposing a design method for a hybrid beamforming structure using low-precision phase shifters. This method utilizes a two-step update process to obtain analog and digital beamforming modules, thereby reducing algorithm implementation complexity. The specific process consists of two steps: the first step fixes the structure of the digital beamforming module to eliminate multi-user interference; the second step updates the phase of each phase shifter in the analog beamforming module using a cyclic coordinate descent method based on the digital beamforming structure from the first step.

[0032] Compared with the existing technical solutions, the present invention has the following characteristics:

[0033] (1) Since the solution of digital and analog beamforming modules does not require an alternating optimization process, the algorithm complexity is low;

[0034] (2) Compared with the method of directly quantizing the optimized phase shifter phase under the assumption of infinite precision phase shifter, the technical solution proposed in the present invention can achieve lower transmission power while meeting the same system performance;

[0035] (3) Compared with the exhaustive search method, the complexity of the method proposed in the present invention is greatly reduced, while at the same time the loss of system performance is relatively small;

[0036] (4) In the design of digital and analog beamforming modules, the influence of finite-precision phase shifters in the analog beamforming matrix is ​​considered.

[0037] The method provided by the present invention can achieve good performance in low-precision phase shifter networks. Furthermore, while reducing transmit power, it can meet the varying performance requirements of individual users. Furthermore, the method is applicable to phase shifters of varying precision and exhibits rapid convergence. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A schematic diagram of a transmitting end of a partially connected hybrid beamforming structure to which the present invention is applicable.

[0039] Figure 2 It shows how the transmission power of the solution proposed in the present invention changes with user performance requirements when a 4-bit quantized phase shifter is used.

[0040] Figure 3 Indicates the change in transmit power at the base station when phase shifters of different accuracies are used. DETAILED DESCRIPTION

[0041] Below in conjunction with specific embodiment, further set forth the present invention.Should be understood that these embodiments are only used to illustrate the present invention and are not used in limiting the scope of the present invention.In addition, should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms fall equally within the scope limited by the appended claims of the application.

[0042] In a large-scale antenna system, the present invention reduces the transmission power of a transmitter using a partially connected hybrid beamforming structure. At the same time, the present invention also solves the problem of system performance loss caused by the use of finite-precision phase shifters in analog beamforming modules.

[0043] Specifically, the present invention provides a hybrid beamforming design method using a low-precision phase shifter, comprising:

[0044] Establish the problem of minimizing base station transmit power in a multi-user massive antenna system, where the base station uses hybrid beamforming to preprocess the transmitted signal. Model the problem of minimizing base station transmit power, where the digital beamforming architecture at the transmitter uses a zero-forcing algorithm to eliminate interference between multiple users, and determine the functional relationship between the digital and analog beamforming matrices.

[0045] Considering the complexity of actual systems and hardware loss, the analog beamforming module uses low-precision phase shifters. Given the digital beamforming structure, the phase of each phase shifter in the analog beamforming matrix is ​​updated using a cyclic coordinate gradient descent algorithm. Furthermore, to further reduce hardware complexity, the analog beamforming module adopts a partially connected architecture.

[0046] In this embodiment, an optimization model for minimizing transmit power is established, as shown in the following formula:

[0047] minP T

[0048] Among them, P T Indicates the transmit power of the base station. ||·|| F represents the F norm of the matrix, F RF represents the analog beamforming matrix, F BB represents the digital beamforming matrix.

[0049] In this embodiment, for the partially connected structure, each RF link is connected to a group of independent antennas, so the simulated beamforming matrix F RFThe first constraint is satisfied as follows:

[0050]

[0051] Where BlkDiag[·] represents the block diagonal matrix, (F RF ) i represents the phase shifter vector connected to the i-th RF link, i∈{1, 2, ..., N RF}, N RF Indicates the number of radio frequency links used by the base station.

[0052] Furthermore, in some connection structures, the base station's transmission power P T It can be further simplified to:

[0053]

[0054] Where M represents the number of antennas connected to each RF link.

[0055] Furthermore, for a phase shifter with finite precision, (F RF ) i The second constraint condition shown in the following formula must also be met:

[0056]

[0057] in, represents the phase set that the phase shifter can take, B represents the number of quantization bits of the phase shifter, i∈{1, 2, ..., N RF}, j∈{1, 2, ..., M}.

[0058] Furthermore, to better serve multiple users, each user must meet their own minimum service quality requirements as shown in the third constraint condition below:

[0059] SINR k ≥γ k

[0060] Among them, γ k represents the SINR of the kth user k The lowest threshold value of , k∈{1, 2, ..., K}, K represents the number of users in the system.

[0061] At the same time, SINR k It can be further expressed as:

[0062]

[0063] Among them, h k represents the channel between the base station and the kth user, (F BB )k represents the kth column in the digital beamforming matrix, σ 2 represents the variance of the noise.

[0064] Furthermore, in order to ensure the performance of each user in the multi-user system, the design of the digital beamforming part is mainly used to eliminate the interference between multiple users. Therefore, the zero-forcing algorithm is used to achieve it. BB It can be expressed as:

[0065]

[0066] in,(·) H Represents the conjugate transpose of the matrix, A=Diag[p1,p2,…,p K ], Diag[·] represents the diagonal matrix, p k represents the transmission power corresponding to the kth user, k∈{1, 2, ..., K}, and H represents the channel matrix.

[0067] After knowing the structure of the digital beamforming module, the problem of minimizing the base station's transmit power can be further simplified to:

[0068]

[0069]

[0070]

[0071] The above problem is solved by using cyclic coordinate gradient descent to update the phase of each phase shifter in the analog beamforming, iterating until convergence. Therefore, the design of the hybrid beamforming structure includes the following steps:

[0072] Step 1: Initialize the phase of each phase shifter in the analog beamforming so that its value is in the discrete phase set middle.

[0073] Step 2: Update the phases of all phase shifters in sequence. The update of each phase shifter phase includes the following steps:

[0074] Step 201: Select any phase shifter and fix the phases of other phase shifters to be constant;

[0075] Step 202: Get the 2 corresponding to all possible values ​​of the phase shifter. B an analog beamforming matrix;

[0076] Step 203: The digital beamforming module uses a zero-forcing algorithm to calculate the digital beamforming corresponding to each analog beamforming matrix;

[0077] Step 204: Calculate 2 based on the possible values ​​of the phase shifter.B possible transmit power values;

[0078] Step 205: Select the phase corresponding to the minimum transmit power, and complete the update of the selected phase shifter.

[0079] Step 3: Execute step 2 above until the set accuracy is reached and obtain F RF .

[0080] Step 4: Based on H and F RF , according to the zero-forcing algorithm, calculate F BB .

[0081] Step 5: Calculate the transmit power.

[0082] Specifically, take the downlink multi-user system as an example, where the base station is equipped with N t Transmit antennas, the number of RF chains is N RF Each RF link connects M antennas. The system has U users, each with a single antenna. Assume the base station uses a uniform linear array antenna and the channel is multipath, with L = 50 paths.

[0083] Figure 2 For N t =64, N RF =U=4, B=4, the transmission power in the solution proposed by the present invention is related to the user performance requirements. Figure 2 It can be seen that the technical solution proposed in the present invention can achieve better system performance in a system using a finite precision phase shifter. Compared with the case of using an infinite precision phase shifter, the solution proposed in the present invention has a 2dB transmission power loss after 4-bit quantization. In addition, the solution proposed in the present invention can effectively save transmission power compared with other methods. k =10dB, the transmission power of the solution proposed in the present invention can save at least 5dB.

[0084] Figure 3 For N t =64, N RF =U=4,γ k =10dB, the transmission power changes with the phase shifter accuracy. Figure 3 As can be seen, to achieve the same quality of service, the base station's required transmit power decreases as phase shifter accuracy improves. Furthermore, the performance of a system using a 4-bit phase shifter is very similar to that of an 8-bit phase shifter. Therefore, a 4-bit phase shifter can be used in practical systems to achieve better system performance while reducing hardware requirements.

Claims

1. A design method for a low-complexity hybrid beamforming structure in a large-scale antenna system, characterized by: The following steps are involved: Step 1: Obtain the signal to be transmitted and channel state information; Step 2: Determine the digital beamforming matrix structure and express it as a function of the channel and analog beamforming matrix, as shown in the following equation: Where: F BB represents the digital beamforming matrix; Λ=Diag[p1,p2,…,p K ], Diag[·] represents the diagonal matrix, p k represents the transmission power corresponding to the kth user, k∈{1,2,…,K}, K represents the number of users in the system; F RF represents the analog beamforming matrix; H represents the channel matrix; (·) H represents the conjugate transpose of a matrix; Step 3: The problem of minimizing the base station transmission power is expressed as: Where: ||·|| F represents the F norm of the matrix; BlkDiag[·] represents the block diagonal matrix; (F RF ) i represents the phase shifter vector connected to the i-th RF link, i∈{1,2,…,N RF }, N RF Indicates the number of radio frequency links used by the base station; represents the phase set that the phase shifter can take, and B represents the number of quantization bits of the phase shifter; [(F RF ) i ] j represents the phase shifter vector connected to the jth antenna of the i-th RF chain, j∈{1,2,…,M}, M represents the number of antennas connected to each RF chain; The base station transmit power minimization problem is solved using a cyclic coordinate gradient descent method to update the phase of each phase shifter in the analog beamforming matrix one by one, iterating until convergence. Updating the phase of the current phase shifter in the analog beamforming matrix includes the following steps: Step 301: Select the current phase shifter and fix the phases of other phase shifters to be constant; Step 302: Obtain the 2 corresponding to the possible values ​​of the phase of the current phase shifter. B an analog beamforming matrix; Step 303: The digital beamforming module in the hybrid beamforming structure uses a zero-forcing algorithm to calculate a digital beamforming matrix corresponding to each analog beamforming matrix. Step 304: Calculate 2 based on the possible values ​​of the phase of the current phase shifter. B possible transmit power values; Step 305: Select the phase that minimizes the transmission power, and complete the update of the phase of the current phase shifter; Step 4: Obtain a digital beamforming matrix based on the updated analog beamforming matrix and calculate the transmit power.

2. The method for designing a low-complexity hybrid beamforming structure in a large-scale antenna system according to claim 1, characterized in that: In the large-scale antenna system, the number of radio frequency links at the transmitting end is equal to or unequal to the number of transmission data streams.

3. The method for designing a low-complexity hybrid beamforming structure in a large-scale antenna system according to claim 1, characterized in that: The analog beamforming structure in the hybrid beamforming structure is a partially connected analog beamforming structure.

4. An application method of the design method of a low-complexity hybrid beamforming structure in a large-scale antenna system according to claim 1, characterized in that: It is suitable for multi-user scenarios and can meet the different service quality requirements of each user.

Images