A coordinated multi-point transmission method and system with intelligent reflective surface backscattering empowerment

Abstract

The present invention discloses a method and system for coordinated multi-point transmission enabled by intelligent reflector backscattering. A multi-cell MIMO downlink network model enabled by intelligent reflector backscattering is established. The method considers weighted sum rate maximization and maximum-minimum fairness problems. The method decomposes the weighted sum rate maximization problem into the weighted sum rate maximization problem of intelligent reflector backscattering and the problem of maximizing the received power of the intelligent reflector, and obtains a secure approximate solution. The method decomposes the maximum-minimum fairness problem into the problem of maximizing the minimum weighted SINR of all user devices and the problem of maximizing the received power of the intelligent reflector, and obtains a secure approximate solution. When a transmitter transmits information to a receiver via an intelligent reflector, the multipath cascade channels of multiple element units are merged into one channel to complete coordinated multi-point transmission. The method utilizes intelligent reflector backscattering to achieve coordinated multi-point transmission, reducing computational and control complexity and meeting different communication requirements.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communications, and in particular relates to a coordinated multi-point transmission method and system enabled by backscattering of intelligent reflecting surfaces. Background Art

[0002] With the explosive growth of mobile data services, wireless networks need to support an increasing number of user devices. One of the most important technologies for achieving this goal is network densification, achieved by deploying a large number of active base stations. However, dense base station deployment inevitably causes inter-cell interference, severely limiting the efficient use of spectrum resources. Furthermore, the large number of RF links deployed on base stations consumes significant transmit power and results in high hardware costs.

[0003] Coordinated multi-point (CoMP) transmission technology, a strategy for mitigating inter-cell interference, has been identified as a key technology for 5G and B5G wireless communications. This technology enables base stations to coordinate the transmission of multiple data streams through signal processing, significantly improving the quality of service for user devices at the cell edge. However, the energy consumption and hardware costs of active base stations remain significant challenges. When active base stations are densely deployed, energy supply and hardware budgets increase dramatically. To reduce hardware costs and achieve green communications, CoMP transmission technology requires further improvements in hardware devices and network architecture, in addition to algorithms. In recent years, emerging smart reflector technology has provided an excellent opportunity for the development of CoMP technology.

[0004] A smart reflector is a two-dimensional, artificially programmable metasurface composed of numerous low-cost, passive reflective elements. Each element can independently control the phase and amplitude of the incident electromagnetic wave in real time. Furthermore, the large number of elements provides a rich set of spatial degrees of freedom. Because smart reflectors have no or minimal radio frequency links, they consume very little power. As a green and economical passive component, smart reflectors are expected to help improve the spectrum and energy efficiency of multi-cell networks. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to address the deficiencies in the above-mentioned prior art and provide a coordinated multi-point transmission method and system enabled by backscattering of intelligent reflective surfaces, so as to realize multi-cell MIMO downlink network communication in a manner with low computational complexity and high spectrum energy efficiency.

[0006] The present invention adopts the following technical solutions:

[0007] A coordinated multi-point transmission method enabled by backscattering of intelligent reflective surfaces comprises the following steps:

[0008] S1. Establish a multi-cell MIMO downlink network model with intelligent reflector backscattering capability.

[0009] S2. Based on the multi-cell MIMO downlink network model established in step S1, model the weighted sum rate maximization problem and the maximum-minimum fairness problem respectively. By solving the minimum weighted SINR maximization problem and the smart reflector received power maximization problem for all user equipments, secure approximate solutions to the weighted sum rate maximization problem and the maximum-minimum fairness problem are obtained.

[0010] S3. In the multi-cell MIMO downlink network model established in step S1, when the active macrocell base station transmits information to the user equipment through the intelligent reflection surface, the beamforming vector and transmit power allocation of the macrocell base station and the beamforming vector of the intelligent reflection surface are set according to the secure approximate solution to the weighted sum rate maximization problem and the secure approximate solution to the maximum-minimum fairness problem obtained in step S2, respectively. Coordinated multi-point transmission is completed by merging the multipath cascade channels of multiple element units into one channel.

[0011] Specifically, in step S1, in a multi-cell MIMO downlink network model, the M antennas equipped by the active macrocell base station are divided into I groups, the i-th group of antennas is used to supply wireless energy to the i-th smart reflector surface, each group of antennas is a directional antenna, and each user equipment is equipped with N antennas. When the active macrocell base station provides energy supply for the backscattering of the smart reflector surface, a group of passive smart reflectors is used to replace the traditional active small cell base station to communicate with a group of user equipment; each smart reflector surface is uniformly composed of L passive reflector element units.

[0012] Furthermore, the signal to interference and noise ratio γ of the kth user equipment in the jth cell is jk and achievable rate R jk They are:

[0013]

[0014] R jk =log(1+γ jk )

[0015] in,(·) H is the Hermite matrix, θ jk is the signal x jk The beamforming vector, T jk,j is the channel gain matrix from the jth antenna group of the macrocell base station to the kth user equipment in the jth cell through the jth smart reflector, is an independent and identically distributed cyclic complex Gaussian random vector, G jk,i is the channel gain matrix from the i-th antenna group of the macrocell base station to the k-th user equipment in the j-th cell, wi is the beamforming vector of the i-th antenna group of the macrocell base station, T jk,i is the channel gain matrix from the i-th antenna of the macrocell base station to the k-th user equipment in the j-th cell through the i-th smart reflector, θ ir is the signal x ir The beamforming vector of .

[0016] Specifically, in step S2, the weighted sum rate maximization problem is decomposed into the weighted sum rate maximization problem of the backscattering of the smart reflector and the problem of maximizing the received power of the smart reflector after Lagrangian dual transformation. By solving the weighted sum rate maximization problem of the backscattering of the smart reflector and the problem of maximizing the received power of the smart reflector, a safe approximate solution to the weighted sum rate maximization problem is obtained. and as follows:

[0017] (P1.2)

[0018] st

[0019]

[0020] (P1.5)

[0021] st

[0022]

[0023]

[0024]

[0025] Where Tr(·) is the trace of the matrix, (·) H is the Hermite matrix, rank(·) is the rank of the matrix, H i is the channel gain matrix from the i-th antenna group of the macrocell base station to the i-th smart reflector surface, is the set of smart reflective surfaces, ω jk is the weighting factor of the kth UE in the jth cell, α jk is an auxiliary variable, β jk is an auxiliary variable, for Set, p j is the transmission power of the jth antenna group of the macrocell base station, P is the total transmission power of the macrocell base station, is the set of element units of the smart reflective surface, L is the number of element units of the smart reflective surface, is the set of user equipments in the jth cell.

[0026] Furthermore, the weighted sum rate maximization problem is:

[0027] (P1)

[0028] st

[0029]

[0030]

[0031] Where Θ and W represent θ respectively. jk and w i The set of θ jk is the signal x jk The beamforming vector, w i is the beamforming matrix of the i-th antenna group at the macrocell base station, ω jk is the weighting factor of the kth user equipment in the jth cell, R jk is the transmission rate of the kth user equipment in the jth cell, Tr(·) is the trace of the matrix, (·) H is the Hermite matrix, p i is the transmit power allocated to the i-th antenna group by the macrocell base station, is the set of smart reflective surfaces, P is the total transmit power of the macrocell base station, A collection of element units of the smart reflective surface.

[0032] Specifically, in step S2, the maximum-minimum fairness problem of step S2 is decomposed into the minimum weighted SINR maximization problem of all user devices and the smart reflector surface received power maximization problem after secondary conversion. By solving the minimum weighted SINR maximization problem of all user devices and the smart reflector surface received power maximization problem, a secure approximate solution to the maximum-minimum fairness problem is obtained as follows:

[0033] (P1.2)

[0034] st

[0035]

[0036] (P2.4)

[0037] st

[0038]

[0039]

[0040]

[0041]

[0042] τ>0

[0043] Where Tr(·) is the trace of the matrix, (·) H is the Hermite matrix, rank(·) is the rank of the matrix, H i is the channel gain matrix from the i-th antenna group of the macrocell base station to the i-th smart reflector surface, is the set of smart reflection surfaces, τ is the relaxation variable, β jk is an auxiliary variable, for gather, is the set of user equipment in the jth cell, p j is the transmission power of the jth antenna group of the macrocell base station, is a collection of element units of the smart reflective surface. is the set of smart reflective surfaces, P is the total transmit power of the macrocell base station, and L is the number of elements of the smart reflective surface.

[0044] Furthermore, the maximum-minimum fairness problem is:

[0045] (P2)

[0046] st

[0047]

[0048]

[0049] Where Θ and W represent θ respectively. jk and w i The collection of ω jk is the weighting factor of the kth user equipment in the jth cell, γ jk is the signal to interference and noise ratio of the kth user equipment in the jth cell, Tr(·) is the trace of the matrix, (·) H is the Hermite matrix, w i is the beamforming vector of the i-th antenna group of the macrocell base station, p i is the transmit power allocated to the i-th antenna group by the macrocell base station, P is the total transmit power of the macrocell base station, θ jk is the signal x jkThe beamforming vector of is a collection of smart reflective surfaces, A collection of element units of the smart reflective surface.

[0050] Specifically, in step S3, the smart reflector is equally divided into C parts, each part is L / C element units, and for the c-th element unit cluster of the i-th smart reflector, the cascade channel is expressed as F jk,ic Θ ic H ic , F jk,ic and H ic They refer to the channel gain from the cth element unit cluster of the i-th smart reflector to the kth user equipment in the j-th cell, and the channel gain from the active macrocell base station to the element unit cluster, Θ ic represents the diagonal reflection coefficient matrix of the element unit cluster, θ ic is the matrix Θ ic diagonal elements of ; and a received signal of the kth user equipment in the jth cell reflected by the cth element unit cluster of the i-th smart reflection surface.

[0051] Furthermore, based on the multi-cell MIMO downlink network model established in step S1, or meeting the network's requirement of improving the maximum weighted sum rate of the network, or meeting the fairness requirement between user devices, when all element unit clusters of an intelligent reflecting surface communicate collaboratively with K user devices in the cell it serves, multiple element unit clusters jointly perform backscattering.

[0052] Another technical solution of the present invention is a coordinated multi-point transmission system enabled by backscattering of intelligent reflective surfaces, comprising:

[0053] The network module establishes a multi-cell MIMO downlink network model enabled by intelligent reflector backscattering;

[0054] The problem module, based on the multi-cell MIMO downlink network model established by the network module, models the weighted sum rate maximization problem and the maximum-minimum fairness problem respectively. By solving the minimum weighted SINR maximization problem for all user devices and the received power maximization problem for the smart reflector, it obtains a secure approximate solution to the weighted sum rate maximization problem and a secure approximate solution to the maximum-minimum fairness problem.

[0055] The transmission module is based on the multi-cell MIMO downlink network model established by the network module. When the active macrocell base station transmits information to the user equipment through the intelligent reflection surface, the beamforming vector and transmit power allocation of the macrocell base station and the beamforming vector of the intelligent reflection surface are set according to the secure approximate solution of the weighted sum rate maximization problem and the secure approximate solution of the maximum and minimum fairness problem obtained by the problem module. By merging the multipath cascade channels of multiple element units into one channel, coordinated multi-point transmission is completed.

[0056] Compared with the prior art, the present invention has at least the following beneficial effects:

[0057] The present invention provides a coordinated multi-point transmission method enabled by backscattering of intelligent reflecting surfaces. A mathematical analysis model for a real-world scenario is established by establishing a multi-cell MIMO downlink network model enabled by backscattering of intelligent reflecting surfaces; mathematical modeling is performed considering the weighted sum rate maximization problem and the maximum-minimum fairness problem; the weighted sum rate maximization problem and the maximum-minimum fairness problem are solved to obtain approximate solutions to the two optimization problems; when an active macrocell base station transmits information to a user device through an intelligent reflecting surface, if the requirement of improving the weighted sum rate of the network is met, the beamforming vector and transmit power distribution of the macrocell base station and the beamforming vector of the intelligent reflecting surface are set according to the safe approximate solution of the weighted sum rate maximization problem; if the requirement of fairness between users is met, the beamforming vector and transmit power distribution of the macrocell base station and the beamforming vector of the intelligent reflecting surface are set according to the safe approximate solution of the maximum-minimum fairness problem; the multipath cascade channels of multiple element units are merged into one channel to complete the coordinated multi-point transmission, and by performing clustering of the element units of the intelligent reflecting surface, the calculation and control complexity of the system is reduced to form coordinated multi-point transmission enabled by backscattering of the intelligent reflecting surface.

[0058] Furthermore, considering the communication entities in the network model, in the multi-cell MIMO downlink network model, the M antennas equipped by the active macrocell base station are divided into I groups. The i-th group of antennas is used to supply wireless energy to the i-th smart reflector surface. Each group of antennas is a directional antenna, and each user device is equipped with N antennas. When the active macrocell base station provides energy for the backscattering of the smart reflector surface, a group of passive smart reflectors is used to replace the traditional active small cell base station to communicate with a group of user devices. Each smart reflector surface is uniformly composed of L passive reflective element units, which can realize its own information transmission by modulating and backscattering the signal from the macrocell base station.

[0059] Furthermore, by calculating the signal-to-interference-and-noise ratio γ of the kth user equipment in the jth cell jk and achievable rate R jk Establish a mathematical model of the communication rate of any user device in the network.

[0060] Furthermore, the goal of the weighted sum rate maximization problem is to maximize the maximum weighted sum rate of the entire network and the minimum weighted signal-to-interference-and-noise ratio of all user devices by jointly optimizing the beamforming vectors of the macrocell base station and the smart reflector, as well as the transmit power allocation of the macrocell base station, under the constraint of the total transmit power. The non-convex weighted sum rate maximization problem is decomposed to make it easy to solve. Using an algorithm composed of mathematical means, the weighted sum rate maximization problem of the backscattering of the smart reflector, the problem of maximizing the received power of the smart reflector, the problem of maximizing the minimum weighted SINR of all user devices, and the problem of maximizing the received power of the smart reflector are respectively optimized in several steps of alternating optimization or quadratically constrained quadratic programming to obtain an approximate solution to the initial problem.

[0061] Furthermore, by establishing a mathematical model of the weighted sum rate maximization problem, the weighted sum rate maximization problem is converted into a mathematical problem, which is convenient for calculation.

[0062] Furthermore, the goal of the maximum-minimum fairness problem is to maximize the maximum weighted sum rate of the entire network and the minimum weighted signal-to-interference-and-noise ratio of all user devices by jointly optimizing the beamforming vectors of the macrocell base station and the smart reflector, as well as the transmit power allocation of the macrocell base station, under the constraint of the total transmit power. The maximum-minimum fairness problem is decomposed to make it easy to solve the problem of maximizing the weighted sum rate of backscattering of the smart reflector and the problem of maximizing the received power of the smart reflector, as well as the problem of maximizing the minimum weighted SINR of all user devices and the problem of maximizing the received power of the smart reflector.

[0063] Furthermore, by establishing a mathematical model of the maximum-minimum fairness problem, the maximum-minimum fairness problem is converted into a mathematical problem, which is easy to calculate.

[0064] Furthermore, an element unit clustering scheme for the smart reflective surface is designed, which divides the smart reflective surface equally into C parts, each part consists of L / C element units. For the c-th element unit cluster of the i-th smart reflective surface, the communication performance is improved, and a low-complexity element unit clustering method is used to achieve backscattering.

[0065] Furthermore, based on the established multi-cell MIMO downlink network model, or to meet the network's requirements for improving the maximum weighted sum rate, or to meet the fairness requirements between user devices, the designed intelligent reflector element unit clustering scheme is implemented to reduce the complexity of solving the weighted sum rate maximization problem and the maximum and minimum fairness problem.

[0066] In summary, the method of the present invention can meet the requirement of improving the maximum sum rate of the network or the fairness requirement between user equipments, and improve the spectrum efficiency of the multi-cell network in a low-power and low-complexity manner.

[0067] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 Schematic diagram of a multi-cell MIMO downlink network model enabling IRS backscattering;

[0069] Figure 2 Schematic diagram of the multipath channel model for IRS and element unit clustering scheme;

[0070] Figure 3 Schematic diagram of the spatial distribution of MBS, IRS and UE;

[0071] Figure 4 Schematic diagram of convergence behavior under one simulation, where (a) is WSR and (b) is MMR;

[0072] Figure 5 Schematic diagram of the comparison between the non-clustering scheme and the multi-element unit clustering scheme (with different numbers of clusters), where (a) is WSR and (b) is MMR;

[0073] Figure 6 Schematic diagram of the impact of the number of IRS elements on communication performance, where (a) is WSR and (b) is MMR;

[0074] Figure 7 Schematic diagram of the impact of MBS total transmit power on communication performance, where (a) is WSR and (b) is MMR;

[0075] Figure 8 Graph showing the impact of the number of UEs and the number of UE antennas in a single cell on communication performance, where (a) is WSR with I=2 and N=2, (b) is WSR with I=2 and N=8, (c) is WSR with I=4 and N=2, (d) is MMR with I=2 and N=2, (e) is MMR with I=2 and N=8, and (f) is MMR with I=4 and N=2.

[0076] Figure 9 Flowchart for the implementation of the present invention. DETAILED DESCRIPTION

[0077] The present invention provides a coordinated multi-point transmission method enabled by backscattering of intelligent reflecting surfaces, which utilizes backscattering of intelligent reflecting surfaces to realize coordinated multi-point transmission and is a brand-new communication architecture.

[0078] Specifically, consider a multi-cell MIMO network, where an active macrocell base station is used to transmit energy signals. At the same time, each smart reflector acts as a small cell base station, which can realize its own information transmission by modulating and backscattering the signal from the macrocell base station.

[0079] To meet diverse communication requirements in this coordinated multi-point transmission mode, two optimization problems are considered: weighted sum rate maximization and maximum-minimum fairness. To find their optimal solutions, the beamforming vectors for the macrocell base station and all smart reflectors, as well as the transmit power allocation, are optimized using Lagrangian dual transformations and an alternating method.

[0080] In addition, an element unit clustering scheme is used to reduce the computational and control complexity of the system, where each element cluster can work like an element unit and all clusters can communicate with users collaboratively.

[0081] See also Figure 9 The present invention provides a coordinated multi-point transmission method with intelligent reflective surface backscattering enablement, comprising the following steps:

[0082] S1. Establish a multi-cell MIMO downlink network model with intelligent reflector backscattering capability.

[0083] See also Figure 1 , describes a multi-cell multiple-input multiple-output (MISO) downlink network enabled by intelligent reflecting surface (IRS) backscatter.

[0084] Among them, the active macro cell base station (MBS) provides energy for the backscatter communication of the IRS, and each passive IRS replaces the traditional active small cell base station (SBS) to communicate with a group of user equipment (UE). The M antennas equipped by the MBS are divided into I groups. Antenna group M i Used to supply wireless energy to the i-th IRS. To improve energy conversion efficiency from the MBS to the IRS, each antenna group uses directional antennas. All IRSs share the same hardware architecture, consisting of L passive reflective elements. By independently adjusting the reflection phase and amplitude of each IRS element in real time, the ambient energy signal from the MBS can be modulated into a signal carrying new information for backscattering. All UEs are equipped with N antennas.

[0085] make and They refer to the IRS set, the UE set served by the i-th IRS, and the element unit set of the IRS respectively. Refers to the channel gain matrix from the i-th IRS to the k-th UE in the j-th cell, and Denote the channel gain matrices from the i-th antenna group of the MBS to the i-th IRS and the k-th UE in the j-th cell, respectively. It is assumed that channel and control information is smoothly exchanged between the MBS and all IRSs via dedicated high-speed links. Another assumption is that all channel state information is complete and that all links obey a quasi-static channel model.

[0086] When the MBS broadcasts its energy-carrying signal, the radio frequency (RF) signal incident on the IRS is modulated and converted into a signal carrying new information. The mathematical expression of this process is:

[0087]

[0088] Among them, s and x ir represent the original signal and the modulated signal for the rth UE in the i-th cell, respectively. vector w i is the beamforming vector of the i-th antenna group of MBS.

[0089] Easy to find, and

[0090] Among them, θ i represents the backscattering vector of the i-th IRS, θ ir is the signal x ir The beamforming vector of θ i s to Represents the modulation process of IRS.

[0091] Specifically, Refers to θ i The lth element of Depend on and It consists of two parts, namely

[0092]

[0093] in, and They can control the reflection and modulation of the incident signal respectively.

[0094] For simplicity, let Only use For signal modulation, considering all elements of IRS, the backscatter vector θ i Obviously, it integrates the two functions of reflection beamforming and signal modulation. Because the reflection coefficient of the IRS element unit is not greater than 1, there is in[·] l,l Represents the i-th diagonal element of the matrix.

[0095] For the kth UE in the i-th cell, its received signal is expressed as:

[0096]

[0097] Among them, n jk are independent and identically distributed cyclically symmetric complex Gaussian random vectors, Because MBS uses a directional antenna design, the interference from MBS at the UE is generally very small.

[0098] In this system model, the signal to interference plus noise ratio (SINR) of the kth UE in the jth cell is expressed as:

[0099]

[0100] The data rate achievable by the kth UE in the jth cell is R jk =log(1+γ jk ).

[0101] S2. Based on the multi-cell MIMO downlink network model established in step S1, model the weighted sum rate maximization problem and the maximum-minimum fairness problem respectively. By solving the minimum weighted SINR maximization problem and the smart reflector received power maximization problem for all user equipments, secure approximate solutions to the weighted sum rate maximization problem and the maximum-minimum fairness problem are obtained.

[0102] Based on the established system model, in order to meet different communication requirements, the present invention will consider two optimization problems, namely, the weighted sum rate (WSR) maximization problem and the maximum-minimum fairness problem.

[0103] S201, Weighted Sum Rate Maximization Problem

[0104] For the weighted sum rate maximization problem, the goal is to maximize the weighted sum rate of the entire network by jointly optimizing the beamforming vectors of MBS and IRS, as well as the transmit power allocation of MBS, under the constraint of total transmit power.

[0105] The optimization problem is described as

[0106] (P1)

[0107] st

[0108]

[0109]

[0110] Where Θ and W represent θ respectively. jk and w i A collection of ω. jk represents the weighting factor of the kth UE in the jth cell. P is the total transmit power of the MBS, where p i Represents the transmit power allocated by MBS to the i-th antenna group.

[0111] It should be noted that the information exchange between MBS and all IRSs is smooth enough, so the parameters Θ, W and p i Can be jointly optimized to maximize network weight and rate.

[0112] S202: Maximizing weighted rate

[0113] In the optimization problem (P1), the optimization variables Θ, W and p i are coupled, and the objective function is The problem is difficult to solve because it is a weighted sum of logarithmic functions. This section will give a feasible method to solve problem (P1).

[0114] First, the logarithmic function is transformed into a new objective function using Lagrange dual transformation;

[0115] Then, the new objective function is decomposed into two sub-problems, which are used to maximize the received power of IRS and the weighted sum rate of backscatter communication respectively; by solving the two sub-problems, a secure approximate solution to problem (P1) is obtained.

[0116] S2021. Objective function conversion

[0117] Since the objective function is a weighted sum of logarithmic functions, solving the non-convex problem (P1) is more difficult. In order to facilitate the solution, the Lagrange dual transformation is used to transform the logarithmic function. Specifically, the auxiliary variable α is introduced jk , then the weighted logarithmic function and can be rewritten as

[0118]

[0119] Using the quadratic transformation, the objective function is further expressed as

[0120]

[0121] Among them, β jk is an auxiliary variable, A ik and B ik They are

[0122] A jk =T jk,j θ jk

[0123]

[0124] According to the objective function f(W,Θ,p i ,α jk ,β jk ), problem (P1) can be rewritten as

[0125] (P1.1)

[0126] st

[0127]

[0128]

[0129]

[0130] However, this problem is still difficult to solve directly. Therefore, we will give an approximate solution in the next subsection.

[0131] S2022, Problem Optimization

[0132] As mentioned earlier, the MBS antennas are deployed in a directional manner. Based on this design, the system can achieve at least two benefits. First, the use of directional antennas at the MBS enhances the received power of each IRS; second, interference from the MBS can be effectively suppressed. Furthermore, optimizing the beamforming design at the MBS helps improve the weighted sum rate to maximize the received power at each IRS. Taking these factors into account, we propose to approximate the problem (P1.1) by using two easily tractable subproblems. Compared to the commonly used alternating optimization method, the method proposed in this invention can guarantee convergence and reduce computational complexity. Finally, a suboptimal solution to the initial problem (P1) is obtained.

[0133] Sub-question 1

[0134] The goal of the first sub-problem is to maximize the received power of each IRS, which is given by

[0135]

[0136] st

[0137] make This problem is equivalent to being converted into

[0138]

[0139] st

[0140] Obviously, It is obtained by solving the following equivalent problem.

[0141]

[0142] st

[0143] To make this problem tractable, it is lifted to higher dimensions using a semidefinite relaxation.

[0144] definition This problem can be rewritten as

[0145] (P1.2)

[0146] st

[0147]

[0148] Ignore rank constraints (P1.2) can be relaxed into a convex semidefinite programming problem, which can then be solved using the existing CVX software. By using singular value decomposition or Gaussian randomization, a rank-1 solution can be recovered.

[0149] Sub-question 2

[0150] The goal of the second sub-problem is to maximize the weighted sum rate by jointly optimizing the transmit power allocation on the MBS and the passive beamforming on all IRSs. The problem is expressed as

[0151] (P1.3)

[0152] st

[0153]

[0154]

[0155] Next, an alternating solution process is used to optimize the variables Θ, p i , α jkand β jk , to solve Problem (P1.2). This process is divided into two steps.

[0156] Step 1: Given Θ and p i , we respectively jk and β jk Find the partial derivative and get the optimal and

[0157] make

[0158]

[0159]

[0160] Then, get the best and for:

[0161]

[0162]

[0163] Step 2:

[0164] Given α jk and β jk , the objective function of problem (P1.3) is simplified to:

[0165]

[0166] definition and refer to Then, the objective function can be further reformulated as:

[0167]

[0168] Based on this new objective function, subproblem (P1.3) is reformulated as

[0169] (P1.4)

[0170] st

[0171]

[0172]

[0173] According to the quadratic constraint quadratic programming,

[0174]

[0175]

[0176] in, and

[0177]

[0178] Therefore, subproblem (P1.4) can be equivalently remodeled as

[0179] (P1.5)

[0180] st

[0181]

[0182]

[0183]

[0184] Ignore rank 1 constraint Then the problem becomes a convex problem. According to the optimal solution of the problem The rank 1 solution can be recovered by using singular value decomposition or Gaussian randomization method. Thus, we can get θ jk .

[0185] S203. Maximin fairness problem

[0186] For the maximum-minimum fairness problem, we aim to maximize the minimum weighted SINR of all UEs by jointly optimizing the beamforming vectors of MBS and all IRSs, as well as the transmit power allocation of MBS, under the total power constraint. This problem is modeled as

[0187] (P2)

[0188] st

[0189]

[0190]

[0191] The maximum and minimum fairness optimization design is to ensure fairness to each UE to a certain extent. jk The maximum and minimum SINR under γ jk Function, by the relation R jk =log(1+γ jk ), the weighting factor ω of the (j,k)th UE can be calculated jk The maximum-min rate (MMR) under the

[0192] S204: Maximization of minimum weighted SINR

[0193] An alternating optimization method is presented to solve the max-min fairness problem (P2).

[0194] First, the objective function in the form of maximum and minimum ratio is equivalently converted into a form that is easy to solve using quadratic transformation;

[0195] Then, the transformed new problem can be decomposed into two sub-problems to maximize the received power at the IRS and the minimum weighted SINR of all UEs respectively;

[0196] By solving the two subproblems, a safe approximate solution to problem (P2) can be obtained.

[0197] S2041, Problem Conversion

[0198] Since the objective function is in the form of maximum-minimum ratio, the non-convex optimization problem (P2) is not easy to solve. Introducing the slack variable τ, the problem (P2) is reformulated as

[0199] (P2.1)

[0200] st

[0201]

[0202]

[0203]

[0204] Use the secondary transform to convert SINRγ jk The equivalent conversion is:

[0205]

[0206] A jk =T jk,j θ jk

[0207]

[0208] For the first constraint, we have:

[0209] min

[0210]

[0211]

[0212] Then, the problem (P2) is further reformulated as:

[0213] (P2.2)

[0214] st

[0215]

[0216]

[0217]

[0218] However, this problem is still difficult to solve directly. In the next section, we give an approximate solution to this problem.

[0219] S2042, Problem Optimization

[0220] The problem (P2.2) is approximated by two subproblems. The first subproblem is the same as the first subproblem in the weighted rate maximization optimization process. Therefore, we only focus on the second subproblem here. The second subproblem is expressed as

[0221] (P2.3)

[0222] st

[0223]

[0224]

[0225] For this subproblem, β jk , Θ and p i The optimization was carried out in two steps.

[0226] Step 1: Given W and Θ, the optimal make

[0227]

[0228] Then, the optimal for

[0229]

[0230] Step 2: When W and β are given jk , the first constraint of subproblem (P2.3) can be further reformulated as

[0231]

[0232] definition and make for A collection of .

[0233] According to equations (5) to (8), problem (P2.3) can be remodeled as

[0234] (P2.4)

[0235] st

[0236]

[0237]

[0238]

[0239]

[0240] τ>0

[0241] Ignore rank-1 constraints This problem is a convex problem that is easy to solve. According to its optimal solution By using singular value decomposition or Gaussian randomization method, the rank 1 solution can be recovered. In view of this, we can get θ jk .

[0242] S3. In the multi-cell MIMO downlink network model established in step S1, when the active macrocell base station transmits information to the user equipment through the intelligent reflection surface, the beamforming vector and transmit power allocation of the macrocell base station and the beamforming vector of the intelligent reflection surface are set according to the secure approximate solution to the weighted sum rate maximization problem and the secure approximate solution to the maximum-minimum fairness problem obtained in step S2, respectively. Coordinated multi-point transmission is completed by merging the multipath cascade channels of multiple element units into one channel.

[0243] Increasing the number of IRS elements provides more spatial degrees of freedom and higher reflected power, thereby improving communication performance. However, this advantage comes at the cost of increased computational and control complexity. Therefore, this section will employ a low-complexity element clustering approach to achieve backscattering. The key idea behind element clustering is to treat multiple elements as a single entity.

[0244] See also Figure 2 ,When a transmitter transmits information to a receiver through IRS, the multipath cascade channels of multiple element units are ,merged into one channel.

[0245] Specifically, the IRS is equally divided into C parts, each part is L / C element units, and for the c-th element unit cluster of the i-th IRS, the cascade channel can be expressed as F jk,ic Θic H ic ,in and They refer to the channel gain from the cth element unit cluster of the i-th IRS to the kth UE in the j-th cell, and the channel gain from the MSB to the element unit cluster. ic represents the diagonal reflection coefficient matrix of the element unit cluster, θ ic is the matrix Θ ic The received signal reflected by the cth element unit cluster of the i-th IRS to the kth UE in the j-th cell is expressed as:

[0246]

[0247] From Equation (8), it is easy to find that the system is equivalent to I × C sub-IRSs. In other words, each element unit cluster corresponds to a sub-IRS.

[0248] When all element unit clusters of an IRS cooperate to communicate with K UEs in the cell it serves, multiple element unit clusters can jointly perform backscattering. In this process, each element unit cluster is equivalent to a large element unit. Specifically, when the signal goes from the MBS to a certain UE and passes through the cluster, the transmission signals on all paths are superimposed and are therefore regarded as a multipath cascade channel. Considering the beamforming of the MBS and the backscattering of the element unit clusters, the cascade channel of the cth cluster on the i-th IRS is described as F jk,ic diag{H ic w i}θ ic .make in is the phase compensation vector, set to 1.

[0249] In addition,

[0250]

[0251]

[0252] Then, formula (5) can be reformulated as:

[0253]

[0254] It is easy to observe that the form of Equation (9) is the same as that of Equation (1). Obviously, when element unit clustering is used, a similar method is adopted as above to solve the weighted sum rate maximization problem and the maximum-minimum fairness problem, so the solution process is not repeated here.

[0255] When element unit clustering is used to implement backscattering, the computational and control complexity can be reduced. The more element units form a cluster, the lower the complexity of the optimization problem.

[0256] In another embodiment of the present invention, a collaborative multi-point transmission system enabled by backscattering of intelligent reflecting surfaces is provided. The system can be used to implement the above-mentioned collaborative multi-point transmission method enabled by backscattering of intelligent reflecting surfaces. Specifically, the collaborative multi-point transmission system enabled by backscattering of intelligent reflecting surfaces includes a network module, a problem module and a transmission module.

[0257] The network module establishes a multi-cell MIMO downlink network model enabled by intelligent reflector backscattering.

[0258] The problem module, based on the multi-cell MIMO downlink network model established by the network module, models the weighted sum rate maximization problem and the maximum-minimum fairness problem respectively. By solving the minimum weighted SINR maximization problem for all user devices and the received power maximization problem for the smart reflector, it obtains a secure approximate solution to the weighted sum rate maximization problem and a secure approximate solution to the maximum-minimum fairness problem.

[0259] The transmission module is based on the multi-cell MIMO downlink network model established by the network module. When the active macrocell base station transmits information to the user equipment through the intelligent reflection surface, the beamforming vector and transmit power allocation of the macrocell base station and the beamforming vector of the intelligent reflection surface are set according to the secure approximate solution of the weighted sum rate maximization problem and the secure approximate solution of the maximum and minimum fairness problem obtained by the problem module. By merging the multipath cascade channels of multiple element units into one channel, coordinated multi-point transmission is completed.

[0260] The following numerical simulations evaluate the achievable communication performance of an IRS backscatter-enabled multi-cell MIMO downlink network. In addition to the proposed IRS backscatter-enabled CoMP scheme, several other schemes are presented for comparison.

[0261] Joint-OP: This figure refers to the IRS backscatter-enabled CoMP solution proposed in this invention. In this solution, the active beamforming of the MBS is optimized first, and then the MBS transmit power and the passive beamforming of all IRSs are jointly optimized.

[0262] Reflection-BF: This illustration shows a simplified CoMP optimization scheme with IRS backscattering enabled. In this scheme, only the passive beamforming of all IRSs is optimized, while the MBS beamformer is randomly generated, and the transmit power of different MBS antenna groups is equally distributed.

[0263] Active-BS: This figure illustrates a beamforming solution within a traditional CoMP framework. In this solution, active antennas function as SBSs to communicate with UEs. In this framework, the SBS beamforming design is optimized.

[0264] In the simulation, it is assumed that all channel gains obey the Rice distribution, the Rice factor is represented by k, and d is used. x and d0 represent the transmission distance and reference distance respectively, and the channel loss is modeled as PL = PL0-25lg(d x Because the incident signal of the IRS is only reflected at the front half of the IRS, a channel gain of 3 dBi is considered for compensation. Figure 3 A possible spatial distribution of MBS, IRS, and UE is given. The MBS is located at the origin of the coordinate system, and the distance between the two IRSs and the MBS is d. For simplicity, the number of UEs served by each IRS is the same, and the locations of all UEs are uniformly and randomly distributed within the cell radius r. In addition, the number of antennas in each antenna group of the MBS is the same. Because the MBS adopts a directional antenna design, the antenna gain from each antenna group of the MBS to the corresponding IRS is set to 30 dBi. At the same time, each UE can only receive one percent of the signal power from the MBS. Table 1 lists some important parameter values ​​in the simulation. However, in some calculations, some of these parameters can be used as variables or take other values. Note: For given parameter values, after solving the maximum and minimum fairness problem, the MMR can be calculated and given by the simulation results.

[0265] Table 1. Simulation parameter values

[0266]

[0267]

[0268] Figure 4 It shows the convergence behavior of all schemes in a multi-cell MIMO downlink network in one simulation. Figure 4 (a) and Figure 4 It can be clearly seen in (b) that for the weighted sum rate problem and the maximum-minimum fairness problem, the proposed optimization scheme converges very quickly.

[0269] Figure 5 A comparison was made between a non-clustered scheme and a multi-element clustering scheme (with varying numbers of clusters). Simulation results show that the achievable WSR and MMR increase with more clusters. Furthermore, regardless of whether element clustering is employed, the communication performance of Joint-OP outperforms that of Reflection-BF.

[0270] Figure 6 and Figure 7 The figures show how the number of IRS elements and the total MBS transmit power affect the performance of an IRS backscatter-enabled multi-cell MIMO downlink network. From these two figures, we can observe that as the number of IRS elements or the total MBS transmit power increases, the WSR and MMR achieved for both Joint-OP and Reflection-BF also increase.

[0271] In addition, when an IRS is divided into multiple clusters, the WSR and MMR can be made higher. On the other hand, the achievable performance of joint-OP and reflection-BF is evaluated by comparing with active-BS. It is not difficult to find that with the increase of the number of IRS elements and the total transmit power of MBS, the WSR and MMR of P a =9dBi or P a =18dBi, the WSR and MMR of the combined-OP and reflected-BF can reach or exceed that of the active-BS.

[0272] Figure 8 This figure shows the impact of the number of UEs in a single cell on communication performance in a multi-cell MIMO downlink network enabled by IRS backscatter. It is clear that WSR improves as the number of UEs in a single cell increases.

[0273] In summary, the present invention provides a method and system for coordinated multi-point transmission using backscattering enabled by intelligent reflective surfaces, which has the following features:

[0274] The IRS backscatter-enabled CoMP approach for multi-cell MIMO networks involves deploying an active MBS to transmit energy signals. Simultaneously, each IRS acts as an SBS, modulating and reflecting the signals from the MBS to transmit its own information.

[0275] Based on the established model, the weighted sum rate maximization problem and the maximum-minimum fairness problem are solved to meet different communication requirements. Specifically, based on the established IRS backscatter-enabled multi-cell MIMO downlink network model, two optimization problems are considered to meet different communication requirements: weighted sum rate maximization and maximum-minimum fairness. To find the optimal solutions to these two problems, Lagrangian dual transformation and alternating methods are used to optimize the beamforming vectors of the MBS and IRS, as well as the MBS power allocation.

[0276] A clustering scheme for element units has been designed to reduce computational and control complexity. Specifically, each element unit cluster can function as a single element unit, and all clusters can coordinate communication with the UE. By combining all transmitted signals into a single cluster, and then using methods similar to non-clustered schemes, it can solve the weighted sum rate maximization problem and the maximum-minimum fairness problem.

[0277] The above content is only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A coordinated multi-point transmission method enabled by backscattering of intelligent reflective surfaces, characterized in that: The following steps are involved: S1. Establish a multi-cell MIMO downlink network model with intelligent reflector backscattering capability. S2. Based on the multi-cell MIMO downlink network model established in step S1, the weighted sum rate maximization problem and the maximum-minimum fairness problem are modeled respectively. By solving the minimum weighted SINR maximization problem of all user devices and the smart reflector receiving power maximization problem, a secure approximate solution to the weighted sum rate maximization problem and a secure approximate solution to the maximum-minimum fairness problem are obtained. For the weighted sum rate maximization problem, after Lagrangian dual transformation, it is decomposed into the weighted sum rate maximization problem of smart reflector backscattering and the smart reflector receiving power maximization problem. By solving the weighted sum rate maximization problem of smart reflector backscattering and the smart reflector receiving power maximization problem, a secure approximate solution to the weighted sum rate maximization problem is obtained. and as follows: Where Tr(·) is the trace of the matrix, (·) H is the Hermite matrix, rank(·) is the rank of the matrix, H i is the channel gain matrix from the i-th antenna group of the macrocell base station to the i-th smart reflector surface, is the set of smart reflective surfaces, ω jk is the weighting factor of the kth UE in the jth cell, α jk is an auxiliary variable, β jk is an auxiliary variable, for Set, p j is the transmission power of the jth antenna group of the macrocell base station, P is the total transmission power of the macrocell base station, is the set of element units of the smart reflective surface, L is the number of element units of the smart reflective surface, is the set of user equipments in the jth cell; The maximum-minimum fairness problem is decomposed into the minimum weighted SINR maximization problem for all user devices and the received power maximization problem for the smart reflector after secondary transformation. By solving the minimum weighted SINR maximization problem for all user devices and the received power maximization problem for the smart reflector, a secure approximate solution to the maximum-minimum fairness problem is obtained as follows: Where Tr(·) is the trace of the matrix, (·) H is the Hermite matrix, rank(·) is the rank of the matrix, H i is the channel gain matrix from the i-th antenna group of the macrocell base station to the i-th smart reflector surface, is the set of smart reflection surfaces, τ is the relaxation variable, β jk is an auxiliary variable, for gather, is the set of user equipment in the jth cell, p j is the transmission power of the jth antenna group of the macrocell base station, is a collection of element units of the smart reflective surface. is the set of smart reflective surfaces, P is the total transmit power of the macrocell base station, and L is the number of elements of the smart reflective surface; S3. In the multi-cell MIMO downlink network model established in step S1, when the active macrocell base station transmits information to the user equipment through the intelligent reflection surface, the beamforming vector and transmit power allocation of the macrocell base station and the beamforming vector of the intelligent reflection surface are set according to the secure approximate solution to the weighted sum rate maximization problem and the secure approximate solution to the maximum-minimum fairness problem obtained in step S2, respectively. Coordinated multi-point transmission is completed by merging the multipath cascade channels of multiple element units into one channel.

2. The coordinated multi-point transmission method using backscattering enabled by smart reflective surfaces according to claim 1, characterized in that: In step S1, in a multi-cell MIMO downlink network model, the M antennas equipped by the active macrocell base station are divided into I groups. The i-th group of antennas is used to supply wireless energy to the i-th smart reflector surface. Each group of antennas is a directional antenna. Each user equipment is equipped with N antennas. When the active macrocell base station provides energy for the backscattering of the smart reflector surface, a group of passive smart reflectors is used to replace the traditional active small cell base station to communicate with a group of user equipment. Each smart reflector surface is uniformly composed of L passive reflector element units.

3. The coordinated multi-point transmission method using backscattering enabled by intelligent reflective surfaces according to claim 2, characterized in that: The signal-to-interference-and-noise ratio γ of the kth user equipment in the jth cell jk and achievable rate R jk They are: R jk =log(1+γ jk ) in,(·) H is the Hermite matrix, θ jk is the signal x jk The beamforming vector, T jk,j is the channel gain matrix from the jth antenna of the macrocell base station to the kth user equipment in the jth cell through the jth smart reflector, is an independent and identically distributed cyclic complex Gaussian random vector, G jk,i is the channel gain matrix from the i-th antenna group of the macrocell base station to the k-th user equipment in the j-th cell, w i is the beamforming vector of the i-th antenna group of the macrocell base station, T jk,i is the channel gain matrix from the i-th antenna group of the macrocell base station to the k-th user equipment in the j-th cell through the i-th smart reflector, θ ir is the signal x ir The beamforming vector of .

4. The coordinated multi-point transmission method using backscattering enabled by intelligent reflective surfaces according to claim 1, characterized in that: In step S2, the weighted sum rate maximization problem is: Where Θ and W represent θ respectively. jk and w i The set of θ jk is the signal x jk The beamforming vector, w i is the beamforming matrix of the i-th antenna group at the macrocell base station, ω jk is the weighting factor of the kth user equipment in the jth cell, R jk is the transmission rate of the kth user equipment in the jth cell, Tr(·) is the trace of the matrix, (·) H is the Hermite matrix, p i is the transmit power allocated to the i-th antenna group by the macrocell base station, is the set of smart reflective surfaces, P is the total transmit power of the macrocell base station, A collection of element units of the smart reflective surface.

5. The coordinated multi-point transmission method using backscattering enabled by smart reflective surfaces according to claim 1, characterized in that: In step S2, the maximum-minimum fairness problem is: Where Θ and W represent θ respectively. jk and w i The collection of ω jk is the weighting factor of the kth user equipment in the jth cell, γ jk is the signal to interference and noise ratio of the kth user equipment in the jth cell, Tr(·) is the trace of the matrix, (·) H is the Hermite matrix, w i is the beamforming vector of the i-th antenna group of the macrocell base station, p i is the transmit power allocated to the i-th antenna group by the macrocell base station, P is the total transmit power of the macrocell base station, θ jk is the signal x jk The beamforming vector of is a collection of smart reflective surfaces, A collection of element units of the smart reflective surface.

6. The coordinated multi-point transmission method using backscattering enabled by smart reflective surfaces according to claim 1, characterized in that: In step S3, the smart reflector is divided into C parts, each part is L / C element units, and for the c-th element unit cluster of the i-th smart reflector, the cascade channel is represented as F jk,ic Θ ic H ic , F jk,ic and H ic They refer to the channel gain from the cth element unit cluster of the i-th smart reflector to the kth user equipment in the j-th cell, and the channel gain from the active macrocell base station to the element unit cluster, Θ ic represents the diagonal reflection coefficient matrix of the element unit cluster, θ ic is the matrix Θ ic The diagonal elements of A received signal of the kth user equipment in the jth cell is reflected by the cth element unit cluster of the i-th smart reflection surface.

7. The coordinated multi-point transmission method using backscattering enabled by smart reflective surfaces according to claim 6, characterized in that: Based on the multi-cell MIMO downlink network model established in step S1, either the network's maximum weighted sum rate requirement is met, or the fairness requirement between user devices is met. When all element unit clusters of a smart reflective surface communicate cooperatively with K user devices in the cell it serves, multiple element unit clusters jointly perform backscattering.

8. A coordinated multi-point transmission system enabled by intelligent reflective surface backscattering, characterized in that: include: The network module establishes a multi-cell MIMO downlink network model enabled by intelligent reflector backscattering; The problem module is a multi-cell MIMO downlink network model established based on the network module. It models the weighted sum rate maximization problem and the maximum-minimum fairness problem respectively. By solving the minimum weighted SINR maximization problem of all user devices and the smart reflector receiving power maximization problem, a secure approximate solution to the weighted sum rate maximization problem and a secure approximate solution to the maximum-minimum fairness problem are obtained. For the weighted sum rate maximization problem, after Lagrange dual transformation, it is decomposed into the weighted sum rate maximization problem of smart reflector backscattering and the smart reflector receiving power maximization problem. By solving the weighted sum rate maximization problem of smart reflector backscattering and the smart reflector receiving power maximization problem, a secure approximate solution to the weighted sum rate maximization problem is obtained. and as follows: Where Tr(·) is the trace of the matrix, (·) H is the Hermite matrix, rank(·) is the rank of the matrix, H i is the channel gain matrix from the i-th antenna group of the macrocell base station to the i-th smart reflector surface, is the set of smart reflective surfaces, ω jk is the weighting factor of the kth UE in the jth cell, α jk is an auxiliary variable, β jk is an auxiliary variable, for Set, p j is the transmission power of the jth antenna group of the macrocell base station, P is the total transmission power of the macrocell base station, is the set of element units of the smart reflective surface, L is the number of element units of the smart reflective surface, is the set of user equipments in the jth cell; The maximum-minimum fairness problem is decomposed into the minimum weighted SINR maximization problem for all user devices and the received power maximization problem for the smart reflector after secondary transformation. By solving the minimum weighted SINR maximization problem for all user devices and the received power maximization problem for the smart reflector, a secure approximate solution to the maximum-minimum fairness problem is obtained as follows: Where Tr(·) is the trace of the matrix, (·) H is the Hermite matrix, rank(·) is the rank of the matrix, H i is the channel gain matrix from the i-th antenna group of the macrocell base station to the i-th smart reflector surface, is the set of smart reflection surfaces, τ is the relaxation variable, β jk is an auxiliary variable, for gather, is the set of user equipment in the jth cell, p j is the transmission power of the jth antenna group of the macrocell base station, is a collection of element units of the smart reflective surface. is the set of smart reflective surfaces, P is the total transmit power of the macrocell base station, and L is the number of elements of the smart reflective surface; The transmission module is based on the multi-cell MIMO downlink network model established by the network module. When the active macrocell base station transmits information to the user equipment through the intelligent reflection surface, the beamforming vector and transmit power allocation of the macrocell base station and the beamforming vector of the intelligent reflection surface are set according to the secure approximate solution of the weighted sum rate maximization problem and the secure approximate solution of the maximum and minimum fairness problem obtained by the problem module. By merging the multipath cascade channels of multiple element units into one channel, coordinated multi-point transmission is completed.

Images