Terahertz multilayer time delayer structure and hybrid precoding design method

Abstract

The invention provides a terahertz multi-layer time delayer structure and a hybrid precoding design method, which are used for solving the problems of large delay range time delayers, high hardware complexity and high power consumption in the traditional full-connection antenna structure scheme based on single-layer and double-layer time delayers. The method comprises the following steps: firstly, designing a terahertz sub-connection antenna structure based on multiple layers of time delayers, and deriving the number of time delayers and a discrete time delay set according to the number of time delayers and antennas in each layer; secondly, a multi-user analog precoding and digital precoding mixed model is provided based on the designed terahertz sub-connection antenna structure, and analog precoding is achieved by designing phase shifting of a phase shifter and delay of a time delayer; and finally, the sum rate of the terahertz sub-connection antenna structure is obtained through the hybrid model. Compared with a traditional scheme, the scheme provided by the invention can effectively reduce the number of time delayers in a large delay range and the hardware complexity on the premise of sacrificing a small amount of system performance.

Description

Technical Field

[0001] The present invention relates to the technical field of terahertz communication, and in particular to a terahertz multi-layer delay device structure and a hybrid precoding design method. Background Art

[0002] Terahertz (0.1~10THz) communication is an inevitable trend in the development of 6G wireless communication in the future due to its advantages such as ultra-wide bandwidth and ultra-high transmission rate. The shorter wavelength of terahertz signals allows base stations to deploy a large number of antennas within a limited physical size to form high-gain directional beams, effectively improving the strength of received signals. However, traditional all-digital beamforming technology requires each antenna to be connected to a radio frequency (RF) chain, and the deployment of large array antennas will lead to huge energy consumption and hardware costs. To this end, researchers proposed a hybrid analog / digital precoding technology based on high-dimensional analog precoding and low-dimensional digital precoding. It connects large array antennas to a small number of RF chains through phase shifters (PS), effectively reducing the number of RF chains and hardware costs. The hybrid precoding structure can be divided into a fully connected and a sub-connected structure. The sub-connected structure is simple and has high energy efficiency, but compared with the fully connected structure, there may be a certain rate loss. In a narrowband single-carrier system, asymptotically optimal system performance can be obtained by optimizing analog / digital precoding. However, due to the frequency-independent characteristics of the phase shifter, in broadband terahertz systems, the multi-carrier beams generated by large array antennas will be oriented in different directions with different carriers, a phenomenon known as beam dispersion, resulting in severe gain loss across the entire bandwidth of the array.

[0003] At present, some work has studied how to overcome the beam dispersion problem in broadband terahertz communications. The most direct method is to replace all phase shifters with frequency-dependent time delays (TD), but this requires a large number of time delays, which will lead to higher power consumption and hardware costs. Therefore, the researchers proposed to insert a limited number of time delays between the RF chain and the phase shifter, and point the multi-carrier beam to the target by jointly controlling the phase shift of the phase shifter and the delay of the time delay. Initially, the literature [ZHAI B, TANG A, PENGC, et al. SS-OFDMA: spatial-spread orthogonal frequency division multiple access for Terahertz networks [J]. IEEE Journal on Selected Areas in Communications, 2021, 39 (6): 1678-1692.] designed an antenna structure based on a serial time delay, but the number of time delays is the same as the number of antennas, and its hardware complexity is relatively high. To this end, the literature [GAO FF, WANG BL, XING CW, et al. Wideband beamforming for hybrid massive MIMO Terahertz communications [J]. IEEE Journal on Selected Areas in Communications, 2021, 39 (6): 1725-1740.] proposed a low-complexity hybrid beamforming scheme based on virtual subarrays and a time delayer. The former divides the entire antenna array into several virtual subarrays to form a beam, and the latter introduces a time delayer in the traditional architecture to form a beam pointing to the target direction, but the performance of the former is lower than that of the latter. The literature [TAN JB, DAI L L. Delay-phase precoding for THz massive MIMO with beamsplit [C] / / 2019 IEEE Global Communications Conference (GLOBECOM). Hawaii: IEEE, 2019: 1-6.] proposed a delay-phase precoding (DPP) technology based on a time delayer. By designing the delay of the delay network, a multi-carrier beam aligned with the target direction is generated over the entire bandwidth. However, the delay devices in the above two documents adopt a parallel cascade structure, each delay device needs to be configured separately, and the delay device must support a large delay range.To this end, the literature [Boljanovic V, Yan H, Lin CC, et al. Fast beam training with true-time-delay arrays in wideband mllimeter-wave systems [J]. IEEE Transactions on Circuits and Systems. I, Regular Papers, 2021, 68 (4): 1727-1739.] proposed a hybrid delay device architecture based on analog / digital delay devices, which reduced the required delay range. Then, the literature [Yan L, Han C, Yuan J. Energy-Efficient Dynamic-subarray with fixed true-time-delay design for terahertz wideband hybrid beamforming [J]. IEEE Journal on Selected Areas in Communications, 2022, 40 (10): 2840-2854.] proposed a dynamic subarray with Fixed True Time Delay (DS-FTTD) structure based on a fixed delay device. Compared with the existing complex and adjustable delay devices, the DS-FTTD structure has lower power consumption and hardware complexity. However, this solution has limited improvement on system performance. At the same time, all the above works assume that the delay device can provide a delay with infinite resolution, while the actual delay device can generally only set a finite number of discrete delays.

[0004] Current research has proposed a sparse RF chain antenna structure based on a single-layer delay element to overcome the problem caused by beam dispersion. However, the number of delay elements with a large delay range and the hardware complexity in the designed antenna structure are high, resulting in high hardware costs and power consumption. Summary of the invention

[0005] In view of the technical problems existing in the above-mentioned background technology, the present invention proposes a terahertz multi-layer time delay structure and a hybrid precoding design method, designs a terahertz sub-connected antenna structure based on the multi-layer time delay, and proposes a multi-user hybrid analog / digital precoding scheme based on the antenna structure. The scheme realizes analog precoding by designing the phase shift of the phase shifter and the delay of the time delay device, which effectively reduces the number of time delay devices with a large delay range and the hardware complexity compared with the traditional scheme.

[0006] In order to achieve the above object, the technical solution of the present invention is achieved as follows:

[0007] A terahertz multilayer delay structure and hybrid precoding design method, the steps are as follows:

[0008] Step 1: Design a terahertz sub-connected antenna structure based on multi-layer delay devices, and derive the number of delay device bits and discrete delay sets according to the number of delay devices and antennas in each layer;

[0009] Step 2: Based on the designed terahertz sub-connected antenna structure, a multi-user analog precoding and digital precoding hybrid model is proposed, and the sum rate of the terahertz sub-connected antenna structure is obtained through the hybrid model;

[0010] Specifically, first, the phase shift of the phase shifter is designed according to the array response vector at the center frequency, then the aperture transit time of the antenna array corresponding to the subarray where the delay device is located is used as the delay of the delay device, and the delay of the delay device is discretized based on the discrete delay set to obtain an analog precoding matrix; then, based on the analog precoding matrix, a low-complexity zero-forcing precoding technology is used to design a digital precoding matrix; finally, the sum rate of the terahertz sub-connected antenna structure is obtained according to the analog precoding matrix and the digital precoding matrix.

[0011] Preferably, the conventional terahertz sub-connection antenna structure comprises N T Antennas and N RF RF chains, each subarray contains N antennas, where N = N T / N RF ; If the base station serves K single-antenna users, the maximum number of users that can be served is K = N RF ; The distance between two adjacent antennas is d = c / 2f c =λ c / 2, where c is the speed of light, f c represents the center carrier frequency, λ c represents the center carrier wavelength;

[0012] The terahertz sub-connected antenna structure based on multi-layer delay devices refers to inserting n layers of delay devices between the RF chain and the phase shifter based on the traditional sub-connected antenna structure, wherein the number of single-antenna users served by the base station is the same as the number of RF chains. Therefore, it is stipulated that the sub-array where each RF chain is located serves one single-antenna user, and each user is located at a different azimuth angle θ k .

[0013] Preferably, the method for obtaining the simulated precoding matrix is:

[0014] After inserting n layers of delays, the phase shifter unit produces a frequency-independent phase shift for:

[0015]

[0016] Where p = 1, 2, ..., P, P represents the number of antennas connected to each delay in the outermost layer, T d is the time delay between two adjacent antennas, n RF =1,...,N RF , Indicates the nth RF The azimuth of the user served by each RF chain;

[0017] Frequency-dependent delay generated by the delay unit in the i-th layer of the n-layer delay scheme for:

[0018]

[0019] Among them, m i =1,2,...,M i , M i Represents the number of delay devices in the i-th layer, P = N / M1…M n Indicates the number of antennas connected to the last layer of delay devices;

[0020] right Discretization can be performed to obtain the discrete delay for:

[0021]

[0022] in, represents the discrete delay set of the i-th layer delay device, Represents the number of bits of the i-th layer delay, satisfying

[0023] The frequency-dependent phase shift produced by the delay unit for:

[0024]

[0025] Among them, f m represents the frequency of the mth subcarrier, represents the delay of the m1th (m1=1,2,...,M1) delay unit in the first layer, represents the delay of the m1th (m1=1,2,...,M1) delay unit in the second layer;

[0026] Total phase shift of antenna element The expression is:

[0027]

[0028] in, represents the frequency-independent phase shift produced by the phase shifter unit;

[0029] The analog precoding of a single RF chain It is expressed as:

[0030]

[0031] Therefore, the simulated precoding matrix of the n-layer delay scheme is The expression is:

[0032]

[0033] Preferably, the method for obtaining the digital precoding matrix is:

[0034] The equivalent channel vector of the kth user is calculated based on the obtained simulated precoding matrix F

[0035]

[0036] in, is the channel vector between the base station and the kth user;

[0037] Therefore, the equivalent channel matrix It can be expressed as:

[0038]

[0039] Among them, the superscript T represents transposition;

[0040] The digital precoding matrix is ​​obtained by using the zero-forcing precoding scheme

[0041]

[0042] Wherein, the superscript H represents the conjugate transpose;

[0043] Digital precoding matrix It can also be expressed as:

[0044]

[0045] Digital precoding matrix The elements in are normalized respectively, where the digital precoding of the k-th user is expressed as:

[0046]

[0047] Among them, ||·|| F represents the F-norm.

[0048] Preferably, the sum rate of the terahertz sub-connected antenna structure is expressed as:

[0049]

[0050] Among them, γ m,k is the signal-to-interference-noise ratio received by the kth user, expressed as:

[0051]

[0052] in, represents the variance of additive Gaussian white noise.

[0053] Beneficial effects of the present invention:

[0054] 1) The present invention proposes a terahertz sub-connected antenna structure based on multi-layer delay devices, which arranges the delay devices in multiple layers, and then designs the number of delay devices and discrete delay sets according to the number of delay devices and antennas placed in each layer; the proposed structure reduces the number of delay devices in a large delay range and the hardware complexity while effectively overcoming beam dispersion.

[0055] 2) A multi-user hybrid analog / digital precoding scheme is proposed, which realizes analog precoding by designing phase shifter phase shift and time delay, wherein the phase shifter phase shift is designed according to the array response vector at the center frequency, and the corresponding antenna array aperture transit time connected to the time delay is used as the time delay, which is then discretized; based on the designed analog precoding scheme, digital precoding is designed using zero-forcing (ZF) technology; compared with traditional schemes, the present invention effectively reduces the number of time delays and hardware complexity in a large delay range. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0057] Figure 1 It is a communication system based on terahertz sub-connected antenna structure.

[0058] Figure 2 Three different terahertz antenna structures; (a) is a phase shifter structure, (b) is a single-layer delay structure, and (c) is a double-layer delay structure.

[0059] Figure 3 It is a multi-layer delay device structure; (a) is a three-layer delay device structure, and (b) is a four-layer delay device structure.

[0060] Figure 4The invention relates to a sub-connected antenna structure communication system based on an n-layer time delay scheme.

[0061] Figure 5 This is the relationship between the sum rate and signal-to-noise ratio under different schemes.

[0062] Figure 6 This is the relationship between the sum rate and the signal-to-noise ratio under different delay step sizes.

[0063] Figure 7 The relationship between the achievable sum rate and signal-to-noise ratio under different delay structure. DETAILED DESCRIPTION

[0064] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0065] In order to overcome the influence of beam dispersion in terahertz communication, the present invention proposes a terahertz sub-connected antenna structure based on multi-layer delayers, which is aimed at the problems of large number of delayers with large delay range, high hardware complexity and high power consumption in the traditional fully connected antenna structure scheme based on single-layer and double-layer delayers. The structure divides the delayers into multiple layers for arrangement, and then designs the number of delayer bits and discrete delay sets according to the number of delayers and the number of antennas placed in each layer. The proposed structure reduces the number of delayers with large delay range and the hardware complexity under the premise of effectively overcoming beam dispersion. At the same time, the present invention proposes a multi-user hybrid analog / digital precoding scheme based on the designed antenna structure. The scheme realizes analog precoding by designing phase shifter phase shift and delayer delay, wherein the phase shifter phase shift is designed according to the array response vector at the center frequency, and the corresponding antenna array aperture transit time connected to the delayer is used as the delayer delay, which is then discretized. Then, based on the designed analog precoding scheme, the digital precoding is designed using the Zero-Forcing (ZF) technology and normalized. Although the performance of the solution proposed in the present invention is lower than that of the traditional solution, it also effectively reduces the number of delay devices with a large delay range and the hardware complexity.

[0066] Traditional terahertz sub-connection antenna structure such as Figure 1 As shown, including N T Antennas and N RF RF chains, each subarray contains N antennas, where N = N T / N RF Assuming that the base station serves K single-antenna users, the maximum number of users that can be served is K = N RFThe distance between two adjacent antennas is d = c / 2f c =λ c / 2, where c is the speed of light, f c and λ c In order to overcome frequency selective fading, the system adopts Orthogonal Frequency Division Multiplexing (OFDM) technology, then the frequency f at the mth (m = 1, ..., M) subcarrier m It can be expressed as:

[0067]

[0068] Wherein, B represents bandwidth and M represents the number of subcarriers.

[0069] The signal received by the kth (k=1,...,K)th user on the mth subcarrier can be expressed as:

[0070]

[0071] Among them, ρ m,k is the power transmitted by the base station for the mth subcarrier of the kth user, is the channel vector between the base station and the kth user, is the analog precoding matrix generated by the constant modulus phase shifter, is the baseband digital precoding vector of the mth subcarrier of the kth user, is the digital precoding matrix at the mth subcarrier, s m,k is the signal sent by the base station to the kth user, It is Additive White Gaussian Noise (AWGN).

[0072] Since the indirect path loss of the terahertz signal is much greater than the direct path, the present invention only considers the direct path, and the channel vector can be expressed as:

[0073] h m,k =aa(f m ,θ k ) (3)

[0074] Where α is the channel gain, θ k ∈[-p / 2,π / 2] is the azimuth of the kth user, is the array response vector at the base station, expressed as:

[0075]

[0076] Among them, nT =1,...,N T .

[0077] For the convenience of analysis, it is assumed that there is only one RF chain at the base station connecting N antennas through phase shifters, serving a single-antenna user with an azimuth angle of θ0. Figure 2 Three traditional sparse RF chain antenna structures and their system performance are shown.

[0078] 1) Phase shifter solution

[0079] The structure of terahertz antenna based on phase shifter is as follows: Figure 2 As shown in (a), in order to make the beam toward θ0, the nth (n=1,...,N) phase shifter is at the center frequency f c The phase shift ψ n for:

[0080] ψ n =2πf c (n-1)T d sinθ0=ω c τ n (5)

[0081] Among them, T d =d / c is the time delay between two adjacent antennas, ω c is the center frequency f c The corresponding angular frequency, τ n is the transmission delay difference between the first antenna and the nth antenna, which can be expressed as:

[0082] τ n =(n-1)T d sinθ0 (6)

[0083] Analog precoding with phase shifter scheme It is expressed as:

[0084]

[0085] When the frequency is f c Change to f m When , the ideal phase to be compensated should be:

[0086]

[0087] Since the phase shifter is a frequency-independent device, the phase produced by the phase shifter does not change when the frequency changes. Therefore, when the carrier frequency changes, there will be a phase difference △ψ between the ideal phase and the actual phase. n ,Right now:

[0088]

[0089] The frequency is f c Change f m The phase difference generated when will cause the beam direction to deviate from θ0 to θ0′. Calculation shows that θ0′ is:

[0090]

[0091] Therefore, the phase difference increases with the increase of bandwidth and the number of antennas. Since beam splitting is essentially caused by the propagation delay of the antenna array aperture, the key to solving beam dispersion is to compensate for the propagation delay.

[0092] 2) Single-layer delay solution

[0093] A common method to compensate for propagation delay is to introduce a time delay device, which generates different time delays at different carriers using a frequency-dependent time delay device. Figure 2 As shown in (b), N antennas are evenly divided into U sub-arrays using a time delay device, where each sub-array contains S = N / U antennas.

[0094] Define the transmission delay difference from the first delay device to the uth (u=1,...,U)th delay device as τ u , then:

[0095] τ u =(u-1)ST d sinθ0(11)

[0096] Among them, when θ0∈[-p / 2,π / 2], the transmission delay range is τ u ∈[0,(U-1)ST d ]. For the single-layer delay scheme, the phase at each antenna unit is controlled by a frequency-independent phase shifter and a frequency-dependent delay. m At , the phase shift produced by the sth (s=1,...,S)th phase shifter in the uth subarray is:

[0097] ψ u,s =2πf m (u-1)ST d sinθ0+2πf c (s-1)T d sinθ0 (12)

[0098] Analog precoding under single-layer delay scheme It can be expressed as:

[0099]

[0100] In order to form a center frequency f cPointing to a consistent beam, when the carrier frequency becomes f m When , the ideal phase to be compensated is:

[0101]

[0102] Therefore, the phase difference △ψ between the ideal phase and the actual phase under the single-layer delay scheme is u,s for:

[0103]

[0104] It can be observed that the phase error is related to the subarray aperture, that is, the phase error will increase with the increase of the number of antennas in each subarray. Therefore, in order to reduce the phase error and alleviate the beam dispersion effect, the number of subarrays, that is, the number of delay devices, can be increased, but this will increase the complexity of the antenna structure.

[0105] 3) Double-layer delay solution

[0106] In the terahertz antenna structure based on the time delay device, the time delay devices are cascaded and have the same delay range. For large antenna arrays, a single-layer time delay device solution needs to use a time delay device with a large delay range, which increases the hardware cost. To this end, the present invention proposes a double-layer time delay device antenna structure, which can reduce the number of cascaded time delay devices, that is, the number of time delay devices with a large delay range. The terahertz sparse RF chain antenna structure solution based on the double-layer time delay device is as follows: Figure 2 As shown in (c), the RF chain connects H time delays, each time delay of the first layer is connected to L time delays, and each time delay of the second layer is connected to P = N / (HL) antennas through a phase shifter. Then the transmission delay difference τ of the hth (h = 1, ..., H)th time delay unit in the first layer is h for:

[0107] τ h =(h-1)LPT d sinθ0 (16)

[0108] Define the transmission delay difference τ of the lth (l=1,...,L)th delay device in the second layer connected to the hth delay device in the first layer h,l for:

[0109] τ h,l =(l-1)PT d sinθ0(17)

[0110] From formulas (16) and (17), it can be obtained that when θ0∈[-p / 2,π / 2], the transmission delay ranges are τ h ∈[0,(H-1)LPT d ],τ h,l ∈[0,(L-1)PTd ]. Therefore, for the dual-layer delay scheme, the delay range of the second layer delay is τ h,l Always less than the delay range τ of the first layer delay device h . Then at frequency f m At , the phase shift generated by the p(1,...,P)th phase shifter connected to the lth time delay unit of the second layer connected to the hth time delay unit of the first layer is:

[0111] ψ h,l,p =2πf m (h-1)LPT d sinθ0+2πf m (l-1)PT d sinθ0+2πf c (p-1)T d sinθ0 (18)

[0112] Therefore, the analog precoding under the double-layer delay scheme It can be expressed as:

[0113]

[0114] In order to ensure that the formed beam does not deviate from the center frequency f c , when the carrier frequency becomes f m When , the ideal phase to be compensated should be:

[0115]

[0116] Therefore, the phase difference between the ideal phase and the actual phase under the double-layer delay scheme is:

[0117]

[0118] 4) Terahertz multilayer delay device structure

[0119] In order to further reduce the number of delay devices with a large delay range and the hardware complexity, and to further reduce the overall hardware cost of the system, the present invention proposes a multi-layer delay device solution based on the above, such as Figure 3 As shown, the three-layer and four-layer delay solutions are taken as examples, assuming that the base station has only one RF chain.

[0120] The three-layer delay antenna structure scheme is as follows Figure 3 As shown in (a), the RF chain is connected to O1 delays, each delay in the first layer is connected to O2 delays, each delay in the second layer is connected to O3 delays, and each delay in the third layer is connected to Q = N / O1O2O3 antennas. Phase shift of antenna units in the three-layer delay scheme It can be expressed as:

[0121]

[0122] Among them, when θ0∈[-p / 2,π / 2], the transmission delay range is

[0123] The four-layer delay antenna structure scheme is as follows Figure 3 (b) shows that the RF chain is connected to V1 delays, each delay in the first layer is connected to V2 delays, each delay in the second layer is connected to V3 delays, each delay in the third layer is connected to V4 delays, and each delay in the fourth layer is connected to W = N / V1V2V3V4 antennas. The phase of the antenna unit in the four-layer delay scheme It can be expressed as:

[0124]

[0125] Among them, when θ0∈[-p / 2,π / 2], the transmission delay range is

[0126] If the actual hardware conditions and performance allow, assume that there are n layers of delay devices, the first layer has M1 delay devices, and each subsequent layer has M2,...,M n Each delay is connected to each delay in the previous layer, and each delay in the last layer is connected to P = N / M1...M n antennas, the phase shift of the antenna unit in the n-layer delay scheme can be obtained as:

[0127]

[0128] Among them, when θ0∈[-p / 2,π / 2], the transmission delay range of the i-th (i=1,...,n) layer delay device is expressed as:

[0129]

[0130] correspond The simulated beam can be expressed as:

[0131]

[0132] Assume that the number of bits of the delay in the i-th layer of the n-layer delay scheme is Then the discrete delay set of the i-th layer delay device is expressed as:

[0133]

[0134] The delay of the i-th layer delay device should satisfy:

[0135]

[0136] Therefore, the number of delay bits in layer i is The following conditions are met:

[0137]

[0138] 5) Hybrid precoding design based on multi-layer delay

[0139] Based on the traditional sub-connected antenna structure, n-layer delay devices are inserted between the RF chain and the phase shifter. The sub-connected antenna structure based on the n-layer delay device solution is as follows: Figure 4 As shown. The number of single-antenna users served by the base station is the same as the number of RF chains, so it is assumed that the subarray where each RF chain is located serves a single-antenna user, and each user is located at a different azimuth angle θ k .

[0140] a) Analog precoding

[0141] To compensate for the array gain loss caused by beam dispersion, beamforming should generate beams aligned with the target physical direction on all subcarriers. The analog precoding of the entire antenna array under the delayer solution includes the frequency-independent analog beams generated by the phase shifter and the frequency-dependent analog beams generated by the delayer. Based on the hardware structure of the delayer and considering the finite resolution delay, the delayer needs to be discretized when designing analog precoding.

[0142] The frequency-independent phase shift produced by the phase shifter unit in the n-layer delay scheme for:

[0143]

[0144] Among them, n RF =1,...,N RF ,

[0145] Frequency-dependent delay generated by the delay unit in the i-th layer of the n-layer delay scheme for:

[0146]

[0147] right Discretization can be performed to obtain the discrete delay for:

[0148]

[0149] The frequency-dependent phase shift produced by the delay unit for:

[0150]

[0151] Total phase shift of antenna element The expression can be:

[0152]

[0153] The analog precoding of a single RF chain It can be expressed as:

[0154]

[0155] Therefore, the simulated precoding matrix of the n-layer delay scheme is The expression is:

[0156]

[0157] b) Digital precoding

[0158] After obtaining the simulated precoding matrix F, the equivalent channel vector of the kth user can be obtained: for:

[0159]

[0160] Therefore, the equivalent channel matrix It can be expressed as:

[0161]

[0162] The digital precoding matrix can be obtained by using the zero-forcing precoding scheme for:

[0163]

[0164] Digital precoding matrix It can also be written as follows:

[0165]

[0166] Normalization can be performed to obtain the digital precoding of the kth user:

[0167]

[0168] When the carrier frequency is f m When γ m,k is the signal to interference plus noise ratio (SINR) received by the kth user, expressed as:

[0169]

[0170] The sum rate

[31] The expression is:

[0171]

[0172] Simulation Analysis

[0173] The simulation results are given to compare the system performance under different schemes. The specific simulation parameters are shown in Table 1.

[0174] Table 1 Simulation parameters

[0175]

[0176] The complexity of different delay schemes is compared, as shown in Table 1. It can be found that the number of delays with a large delay range decreases as the number of delay layers increases. In addition, as the number of delay layers increases, additional delays with a small delay range are introduced, and their delay range is much smaller than that of delays with a large delay range, so the hardware cost can still be effectively reduced. In addition, when the number of delay layers increases, the total number of delay bits will also decrease, which will reduce the hardware complexity of the delay network.

[0177] Table 2 Comparison of the complexity of different delay schemes

[0178]

[0179] Figure 5 The relationship between the sum rate and the signal-to-noise ratio (SNR) under different schemes is shown, where the signal-to-noise ratio is defined as Set the delay step D = 0.4 / f c , the parameters of the delay scheme are according to Table 2. Figure 5 It can be found that the sum rate increases with the increase of signal-to-noise ratio. In addition, the sum rate under different delay schemes decreases with the increase of the number of delay layers. When the signal-to-noise ratio is 20dB, compared with the ideal situation, the sum rate under the phase shifter scheme decreases by about 62.5%, while the sum rates under different delay schemes decrease by about 19.4%, 22.9%, 24.8%, and 34.9%, respectively. Therefore, it can be observed that inserting a delay in the traditional hybrid precoding structure can effectively alleviate the beam dispersion effect. In addition, it can be seen from the data in Table 1 that the total number of bits and the number of delays in the large delay range of different delay schemes decrease with the increase of the number of delay layers. Therefore, the multi-layer delay scheme effectively reduces the hardware cost at the expense of a small amount of sum rate.

[0180] Figure 6The relationship between the sum rate and the signal-to-noise ratio under different delay step sizes is shown. Here, only the phase shifter, single-layer and three-layer delay schemes are considered. Among them, the single-layer delay scheme sets U = 32, the three-layer delay scheme sets O1 = 8, O2 = 2, O3 = 2, and the number of bits is set to b U =8, b O1 =8, b O2 =4, b O3 = 2. The delay step D includes the ideal case (delay device is continuous), 0.45 / f c , 0.55 / f c Three situations. It can be found that the achievable sum rate of the single-layer and three-layer delay schemes decreases with the increase of the delay step length. The reason is that the larger the delay step length, the worse the accuracy of the delay, which leads to beam misalignment and affects the system performance. In addition, it can be found that when the delay step length is small, the performance of the three-layer delay scheme is slightly lower than that of the single-layer delay scheme, but the number of delays in the large delay range and the total number of delay bits are also greatly reduced compared to the single-layer delay scheme, which shows the superiority of the multi-layer delay scheme. In addition, due to the severe beam dispersion effect, the performance of the phase shifter scheme is the worst.

[0181] Figure 7 The relationship between different delay structures and rates and signal-to-noise ratios under the three-layer delay scheme is shown. Assume D = 0.4 / f c , and the single delay scheme with U=32 is set as the upper limit of performance, and the classic phase shifter scheme is used as the benchmark. In the three-layer delay scheme, four cases are set, namely: O1=4, O2=4, O3=2; O1=8, O2=2, O3=2; O1=4, O2=2, O3=2; O1=2, O2=2, O3=2. The number of outermost subarray antennas P in these four cases is 4, 4, 8, and 16 respectively. It can be observed from the figure that the achievable sum rate decreases with the increase of the number of subarray antennas P. In addition, when the number of subarray antennas P remains unchanged, the achievable sum rate is basically the same, but the number of large delay range delays in the cases of O1=4, O2=4, and O3=2 is less than that in the cases of O1=8, O2=2, and O3=2. Therefore, considering the performance and hardware complexity comprehensively, the structure of the multi-layer delay scheme should appropriately reduce the number of delays in the first layer. In addition, when P = 4, the achievable sum rate of the three-layer delay solution is only slightly lower than that of the single-layer delay solution, which effectively proves that the multi-layer delay solution can achieve better performance with lower hardware complexity. When P is further reduced, the achievable sum rate of the three-layer delay solution is further reduced, but the total number of bits is also greatly reduced. Therefore, when the number of antennas remains unchanged, the purpose of reducing hardware costs can be achieved by appropriately increasing the number of antennas P connected to the outermost delay.

[0182] Although the traditional single-layer delay scheme can effectively overcome beam dispersion, the number of delays in a large delay range is large, resulting in high hardware costs. To this end, the present invention proposes a terahertz sub-connected antenna structure based on a multi-layer delay and a hybrid precoding design scheme thereof. The simulation results show that compared with the single-layer delay scheme, although the sum rate of the multi-layer delay scheme is reduced, the number of delays in a large delay range and the number of delay bits are effectively reduced, which reduces the overall hardware cost. Therefore, the multi-layer delay scheme achieves the purpose of reducing hardware costs by sacrificing a small amount of system performance.

[0183] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A terahertz multilayer delay structure and a hybrid precoding design method, characterized in that: The steps are as follows: Step 1: Design a terahertz sub-connected antenna structure based on multi-layer delay devices, and derive the number of delay device bits and discrete delay sets according to the number of delay devices and antennas in each layer; Step 2: Based on the designed terahertz sub-connected antenna structure, a multi-user analog precoding and digital precoding hybrid model is proposed, and the sum rate of the terahertz sub-connected antenna structure is obtained through the hybrid model; Specifically, first, the phase shift of the phase shifter is designed according to the array response vector at the center frequency, then the aperture transit time of the antenna array corresponding to the subarray where the delay device is located is used as the delay of the delay device, and the delay of the delay device is discretized based on the discrete delay set to obtain an analog precoding matrix; then, based on the analog precoding matrix, a low-complexity zero-forcing precoding technology is used to design a digital precoding matrix; finally, the sum rate of the terahertz sub-connected antenna structure is obtained according to the analog precoding matrix and the digital precoding matrix.

2. The terahertz multilayer delay structure and hybrid precoding design method according to claim 1, characterized in that: The traditional terahertz sub-connection antenna structure includes N T Antennas and N RF RF chains, each subarray contains N antennas, where N = N T / N RF ; If the base station serves K single-antenna users, the maximum number of users that can be served is K = N RF ; The distance between two adjacent antennas is d = c / 2f c =λ c / 2, where c is the speed of light, f c represents the center carrier frequency, λ c represents the center carrier wavelength; The terahertz sub-connected antenna structure based on multi-layer delay devices refers to inserting n layers of delay devices between the RF chain and the phase shifter based on the traditional sub-connected antenna structure, wherein the number of single-antenna users served by the base station is the same as the number of RF chains. Therefore, it is stipulated that the sub-array where each RF chain is located serves one single-antenna user, and each user is located at a different azimuth angle θ k .

3. The terahertz multilayer delay structure and hybrid precoding design method according to claim 2, characterized in that: The method for obtaining the simulated precoding matrix is: After inserting n layers of delays, the phase shifter unit produces a frequency-independent phase shift for: Where p = 1, 2, ..., P, P represents the number of antennas connected to each delay in the outermost layer, T d is the time delay between two adjacent antennas, n RF =1,...,N RF , Indicates the nth RF The azimuth of the user served by each RF chain; Frequency-dependent delay generated by the delay unit in the i-th layer of the n-layer delay scheme for: Among them, m i =1,2,...,M i , M i Represents the number of delay devices in the i-th layer, P = N / M1…M n Indicates the number of antennas connected to the last layer of delay devices; right Discretization processing can be performed to obtain the discrete delay for: in, represents the discrete delay set of the i-th layer delay device, Represents the number of bits of the i-th layer delay, satisfying The frequency-dependent phase shift produced by the delay unit for: Among them, f m represents the frequency of the mth subcarrier, represents the delay of the m1th (m1=1,2,...,M1) delay unit in the first layer, represents the delay of the m1th (m1=1,2,...,M1)th delay unit in the second layer; Total phase shift of antenna element The expression is: in, represents the frequency-independent phase shift produced by the phase shifter unit; The analog precoding of a single RF chain It is expressed as: Therefore, the simulated precoding matrix of the n-layer delay scheme is The expression is:

4. The terahertz multilayer delay structure and hybrid precoding design method according to claim 3, characterized in that: The method for obtaining the digital precoding matrix is: The equivalent channel vector of the kth user is calculated based on the obtained simulated precoding matrix F in, is the channel vector between the base station and the kth user; Therefore, the equivalent channel matrix It can be expressed as: Among them, the superscript T represents transposition; The digital precoding matrix is ​​obtained by using the zero-forcing precoding scheme Wherein, the superscript H represents the conjugate transpose; Digital precoding matrix It can also be expressed as: Digital precoding matrix The elements in are normalized respectively, where the digital precoding of the kth user is expressed as: Among them, ||·|| F represents the F-norm.

5. The terahertz multilayer delay structure and hybrid precoding design method according to claim 4, characterized in that: The sum rate of the terahertz sub-connected antenna structure is expressed as: Among them, γ m,k is the signal-to-interference-noise ratio received by the kth user, expressed as: in, represents the variance of additive Gaussian white noise.

Images