A Method for Passive IRS-Assisted Communication in a Joint Synaesthesia Cloud Access Network

Abstract

The present invention discloses a method for passive IRS-assisted communication in a joint communication and sensing cloud radio access network. The multi-antenna baseband processing unit pool (BBU) is responsible for generating communication and sensing signals, and at the same time completing communication and sensing integrated beamforming, and transmitting them to each multi-antenna remote radio head (RRH) through a wired fronthaul link. Each RRH transmits a beam to realize the sensing function for the unobstructed targets within its service range, and at the same time serves communication users. The passive IRS is mainly used to assist communication, and a reflection link is established between the RRH and the obstructed communication users within its service range for communication, so as to expand the coverage of the system.

Description

Technical Field

[0001] The present invention relates to the technical field of communication and sensing integration, and particularly relates to realizing communication and sensing integrated beamforming with the assistance of a passive intelligent reflecting surface in the context of a cloud radio access network, realizing the sensing function for unobstructed targets within its service range, serving communication users simultaneously, and reducing system complexity. Background Art

[0002] After future communication enters the post-5G and 6G eras, a series of emerging application scenarios will emerge, including autonomous driving, industrial automation, virtual reality, etc. The requirements imposed by these scenarios on the system are not limited to the mutual communication between devices, but also require the devices to have a certain sensing ability for the surrounding environment. The communication and sensing integration technology can not only meet the needs of new application scenarios for devices to have both communication and sensing capabilities, but also integrate communication and sensing systems, and achieve a certain sensing ability for targets while ensuring communication quality. The cloud radio access network (C-RAN) virtualizes the network functions of wireless access into virtualized functions and deploys them in a standard cloud environment, which can improve design flexibility and computing scalability, improve energy efficiency, and reduce integration costs. The intelligent reflecting surface (IRS) is a planar array composed of a large number of reconfigurable passive elements, and each element can independently introduce a certain phase shift in response to the incident signal, thereby cooperatively changing the propagation of the reflected signal and reconstructing the entire wireless channel environment to improve wireless communication performance. Summary of the Invention

[0003] The object of the present invention is to complete communication and sensing integrated beamforming through the communication and sensing integrated waveform design assisted by a passive intelligent reflecting surface (Passive-IRS) under the framework of the cloud radio access network (C-RAN), and at the same time use the intelligent reflecting surface technology to reduce system complexity.

[0004] The technical solution of the present invention is as follows:

[0005] A method for jointly assisting communication by a passive IRS in a communication and sensing cloud radio access network, maximizing the minimum waveform gain of each RRH in the direction of its respective area to be sensed, ensuring that the received SINR of each user is greater than a specific threshold, and optimizing the RRH transmitted signal and the passive IRS reflection matrix, specifically including the following steps:

[0006] 1.1) The passive IRS-assisted C-RAN downlink communication and sensing integrated system includes a BBU pool and L RRHs. Information is transmitted between the BBU pool and each RRH through a wired fronthaul link, and each RRH is equipped with N RA uniform linear array (ULA) composed of root antennas is responsible for sensing user information within the cell (i.e., for sensing a certain number of targets), and at the same time, it performs downlink communication with K single-antenna users with the assistance of I passive IRSs. Each passive IRS consists of M reflecting elements. The BBU pool knows the channel state information of all channels and the directions of the targets relative to each RRH.

[0007] 1.2) The BBU pool first generates signals where represents the signal sent to the l-th RRH, and the matrix represents the communication beamforming matrix. The vector represents the sensing signal, and the vector represents the signal sent to each communication user. Since the capacity of the fronthaul link is limited, the signal needs to be quantized and compressed before being transmitted to each RRH. The final set of signals sent to all RHHs is:

[0008]

[0009] where represents the set of quantization and compression noise for all RRHs, and it follows the distribution The covariance matrix of the quantized and compressed signal can be expressed as:

[0010] R = WW H + R0 + Ω,

[0011] 1.3) The compressed signal received by the l-th RRH through the forward compression link is expressed as

[0012]

[0013] where the matrix represents the selection matrix related to the l-th RRH. Specifically, this matrix has elements that are an identity matrix of dimension N only in the rows from (l - 1)N R + 1 to lN R rows, and all remaining elements are 0. R

[0014] 1.4) Let P R represent the transmit power limit of each RRH. The expression for the transmit power limit of the l-th RRH can be obtained as

[0015]

[0016] 1.5) For any communication user k, its received signal y k has the following expression

[0017] y k = (h k + g k ΨG)x + n k

[0018] = z k x + n k ,

[0019] where represents the direct link between all RRHs and user k, Ψ represents the reflection matrix of the passive IRS, represents the reflected link between the active IRS and user k, represents the additive white Gaussian noise received by the user. z k = (h k + g k ΨG) represents the combined channel composed of the direct and reflected links.

[0020] Furthermore, for the passive IRS-assisted integrated sensing and communication system in C-RAN downlink, optimize the transmit signal covariance matrix R, the user transmit beamforming matrix W, and the passive IRS reflection matrix Ψ, aiming to achieve the minimum waveform gain in the direction of the maximum area to be sensed, under the constraints of the single RRH transmit power, multi-user SINR communication performance, and fronthaul link capacity. The specific steps are as follows:

[0021] 2.1) Each RRH is equipped with a uniform linear antenna array composed of N R antennas, and the direction vector of the RRH transmit signal can be modeled as follows

[0022]

[0023] where d RRH represents the spacing between adjacent RRH transmit antennas, and λ represents the wavelength of the transmit signal. The waveform gain of the l-th RRH for a target at a specific direction angle θ is expressed as:

[0024]

[0025] 2.2) Maximize the minimum waveform gain of each RRH in the direction of its respective area to be sensed. Since each RRH needs to sense T targets, the objective function can be expressed as

[0026]

[0027] where the set represents all the targets that need to be sensed by all RRHs.

[0028] 2.3) Meanwhile, to ensure the multi - user communication performance, the received SINR of each user is greater than a specific threshold. The expression form of the SINR constraint condition for the k - th user is

[0029]

[0030] where \(w\) i represents the i - th column of \(W\), that is, the beamforming vector of the i - th user, and \(\Gamma\) is the set minimum user SINR constraint.

[0031] 2.4) For the fronthaul compression link, the fronthaul link capacity constraint condition can be expressed as

[0032]

[0033] The capacity of each finite fronthaul link between the BBU and the RRH is \(C\) l , represents the mutual information between l and \(x\)

[0034] 2.5) The optimization problem can be expressed as:

[0035] (P3)

[0036] s.t.(C1):

[0037] (C2):

[0038] (C3)

[0039] (C4): \(R = WW\) H + \(R_0+\Omega\)

[0040] Furthermore, in the optimization problem, considering the non - convexity of the objective function and the constraints, and the coupling between the optimization variables, the original problem is decomposed into two sub - problems for solution during the solution process.

[0041] 3.1.1) Initialize the matrix \(R\) according to the RRH power limit in 1.4), and rewrite \(R\) in the form of a covariance matrix

[0042]

[0043] \(w\) k represents the k - th column of the communication beamforming matrix \(W\), which is the beamforming vector of user k.

[0044] represents the covariance matrix of each user's beamforming vector. According to the definition, \(R\) k should be of rank one.

[0045] 3.1.2) Rewrite the user SINR power limit condition in 2.3) according to the formula in 2.2)

[0046]

[0047]

[0048] The SINR constraints are related to the matrices R and R k The linear constraints of this expression and w i Not relevant.

[0049] 3.1.3) The formula in 2.4) uses the link capacity limit of the multivariate fronthaul link compression model,

[0050] log|Ω|≤log|Σ|+Tr|Σ -1 Ω|-N,

[0051] This inequality takes on an equal sign only when the condition Ω=Σ holds.

[0052] 3.1.4) In summary, the optimization problem in 2.5) can be rewritten as follows:

[0053] (P3.1)

[0054] st(C1):

[0055] (C2):

[0056] (C3):

[0057]

[0058] (C4):

[0059] (C5):

[0060] (C6):

[0061] (C7):

[0062] 3.2.1) According to sub-problem 1, the SINR threshold of each user is Γ. Sub-problem 2 is as follows: find the actual achievable rate R of the user k , maximize the difference between the achievable rate of each user and the minimum rate of each user, the objective function can be expressed as

[0063]

[0064] 3.2.2) Further, to ensure that R k is definitely greater than the minimum user - achievable rate, we introduce an auxiliary variable α, and at the same time require that this auxiliary variable be greater than or equal to 0. The formula in 3.2.1) is rewritten as

[0065]

[0066] R k -log(1 + Γ)≥α, α≥0

[0067] 3.2.3) According to the transmitted signal form in 1.3) and the channel expression form in 1.4), through the linear receiver u k the user signal y k is restored. The MSE of this process is denoted by the symbol e k and can be specifically expressed as

[0068]

[0069] where w k is the user - transmitted beamforming vector of the k - th user after restoration,

[0070] denotes the signal obtained by the actual user

[0071] 3.2.4) According to the two formulas in 3.2.2) and 3.2.3), the user - achievable rate R k problem can be equivalently expressed as

[0072]

[0073] where w k is a constant auxiliary variable, and when maximizing R k the variables w k and u k satisfy

[0074]

[0075]

[0076] 3.2.5) Substitute the formula in 3.2.4) into and expand the composite channel z k and separating the part related to the IRS reflection matrix, we can get:

[0077]

[0078] where further define the following variables

[0079]

[0080]

[0081]

[0082]

[0083]

[0084] 3.2.6) According to the variables defined in the formula of 3.2.6) and the expanded expression form of we can write the expression form related to the maximum R in the formula of 3.2.6) k and the IRS reflection matrix as

[0085]

[0086] Define the variable

[0087] 3.2.7) In summary, the original optimization problem expressed in the formula of 3.2.3) can be rewritten in a form related to the IRS reflection matrix

[0088] (P3.3)

[0089] s.t.(C1):

[0090] (C2):

[0091] (C3):

[0092] (C4):

[0093] 3.2.8) Only the constraint condition in the formula of 3.2.7) is non-convex. By using the SDR technique to omit this constraint condition, the final solvable convex form of the optimization problem formula can be obtained

[0094]

[0095] s.t.(P3.3)(C1)-(C3)

[0096] And solve this optimization problem through CVX to obtain the final optimization result. The final optimization result needs to go through a sufficient number of Gaussian randomizations to obtain a sub-optimal solution for the current problem, and use this sub-optimal solution to update the comprehensive channel z k and return to step 3.2.3) until the optimization converges.

[0097] The beneficial effects of the present invention are as follows: By means of comprehensive beamforming, communication and sensing integration is achieved, enabling the sensing function for unobstructed targets within its service range while serving communication users, providing a reference for future communication. BRIEF DESCRIPTION OF THE DRAWINGS

[0098] Figure 1 It is a waveform diagram of the RRH transmission signal under different numbers of antennas of the present invention;

[0099] Figure 2 It is a schematic diagram of the system layout of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0100] A method for passive IRS-assisted communication in a joint communication and sensing cloud radio access network, maximizing the minimum waveform gain of each RRH in the direction of its respective area to be sensed, ensuring that the received SINR of each user is greater than a specific threshold, and optimizing the RRH transmission signal and the passive IRS reflection matrix, specifically including the following steps:

[0101] 1.1) The passive IRS-assisted C-RAN downlink communication and sensing integrated system includes a BBU pool and L RRHs. Information is transmitted between the BBU pool and each RRH through a wired fronthaul link. Each RRH is equipped with a uniform linear antenna array (ULA) composed of N R antennas, responsible for sensing user information within the cell and simultaneously performing downlink communication with K single-antenna users with the assistance of I passive IRSs. Each passive IRS consists of M reflection units. The BBU pool knows the state information of all channels and the approximate directions of the targets relative to each RRH.

[0102] 1.2) The BBU pool first generates a signal where represents the signal sent to the l-th RRH, and the matrix represents the communication beamforming matrix. The vector represents the sensing signal, and the vector represents the signal sent to each communication user. Due to the limited capacity of the fronthaul link, the signal needs to be quantized and compressed before being transmitted to each RRH. The final set of signals sent to all RHHs is:

[0103]

[0104] where represents the set of quantization and compression noise for all RRHs, and simultaneously follows the distribution The covariance matrix of the quantized and compressed signal can be expressed as:

[0105] R = WW H + R0 + Ω,

[0106] 1.3) The compressed signal received by the l-th RRH through the forward compression link is expressed as

[0107]

[0108] where the matrix represents the selection matrix related to the l-th RRH. Specifically, only the elements from the ((l - 1)N R + 1)-th row to the lN R -th row of this matrix are an identity matrix of dimension N R , and all the remaining elements are 0.

[0109] 1.4) Let P R represent the transmit power limit of each RRH. The transmit power limit expression of the l-th RRH can be obtained

[0110]

[0111] 1.5) For any communication user k, its received signal y k has the following expression

[0112] y k = (h k + g k ΨG)x + n k

[0113] = z k x + n k ,

[0114] where represents the direct link between all RRHs and user k, Ψ represents the reflection matrix of the passive IRS, represents the reflection link between the active IRS and user k, represents the additive white Gaussian noise received by the user. z k = (h k + g k ΨG) represents the combined channel composed of the direct and reflection links.

[0115] For the proposed passive IRS-assisted integrated communication and sensing system in the C-RAN downlink, optimizing the transmit signal covariance matrix R, the user transmit beamforming matrix W, and the passive IRS reflection matrix Ψ, with the goal of achieving the minimum waveform gain in the direction of the maximum area to be sensed, under the constraints of the transmit power of a single RRH, the multi-user SINR communication performance, and the fronthaul link capacity. The specific steps are as follows:

[0116] 2.1) Each RRH is equipped with a uniform linear antenna array consisting of N R antennas, and the direction vector of the signal transmitted by the RRH can be modeled as follows:

[0117]

[0118] where d RRH represents the spacing between adjacent RRH transmitting antennas, and λ represents the wavelength of the transmitted signal. The waveform gain of the l-th RRH for a target at a specific direction angle θ is expressed as:

[0119]

[0120] 2.2) Maximize the minimum waveform gain of each RRH in the direction of its respective area to be sensed. Each RRH needs to sense T targets, so the objective function can be expressed as

[0121]

[0122] where the set represents all the targets that all RRHs need to sense.

[0123] 2.3) At the same time, to ensure multi-user communication performance, the received SINR of each user is greater than a specific threshold. The expression form of the SINR constraint condition of the k-th user is

[0124]

[0125] where w i represents the i-th column in W, that is, the beamforming vector of the i-th user, and Γ is the set minimum user SINR constraint.

[0126] 2.4) For the fronthaul compression link, the fronthaul link capacity constraint condition can be expressed as

[0127]

[0128] The capacity of each finite fronthaul link between the BBU and the RRH is C l , represents the mutual information between l and x

[0129] 2.5) The optimization problem can be expressed as:

[0130] (P3)

[0131] s.t. (C1):

[0132] (C2):

[0133] (C3)

[0134] (C4):R = WW H +R0+Ω

[0135] For the proposed optimization problem, considering the non - convexity of the objective function and constraints, and the coupling among various optimization variables, the original problem is decomposed into two sub - problems for solution during the solving process.

[0136] 3.1.1) Initialize the matrix R according to the RRH power limit in 1.4), and rewrite R in the form of a covariance matrix

[0137]

[0138] w k denotes the k - th column of the communication beamforming matrix W, corresponding to the beamforming vector of user k. represents the covariance matrix of each user's beamforming vector. According to the definition, R k should be of rank one.

[0139] 3.1.2) Rewrite the user SINR power limit condition in 2.3) according to the formula in 2.2)

[0140]

[0141]

[0142] The SINR limit condition is a linear constraint condition with respect to matrices R and R k and is independent of w i at the same time.

[0143] 3.1.3) The formula in 2.4) adopts the link capacity limit of the multi - hop fronthaul link compression model,

[0144] log|Ω| ≤ log|Σ|+Tr|Σ -1 Ω| - N,

[0145] This inequality takes the equal sign only when the condition Ω = Σ holds.

[0146] 3.1.4) In summary, the optimization problem in 2.5) is rewritten in the following form:

[0147] (P3.1)

[0148] s.t.(C1):

[0149] (C2):

[0150] (C3):

[0151]

[0152] (C4):

[0153] (C5):

[0154] (C6):

[0155] (C7):

[0156] 3.2.1) According to sub - problem 1, the SINR threshold for each user is Γ. The sub - problem 2 is as follows: find the actual achievable rate R of the user k , maximize the difference between the achievable rate of each user and the minimum rate of each user. The objective function can be expressed as

[0157]

[0158] 3.2.2) Further, to ensure that R k is definitely greater than the minimum achievable rate of the user, we introduce an auxiliary variable α, and at the same time require that this auxiliary variable is greater than or equal to 0. The formula in 3.2.1) is rewritten as

[0159]

[0160] R k −log(1 + Γ)≥α, α≥0

[0161] 3.2.3) According to the transmitted signal form in 1.3) and the channel expression form in 1.4), through the linear receiver u k restore the user signal y k The MSE of this process is denoted by the symbol e k and can be specifically expressed as

[0162]

[0163] where w k is the user transmit beamforming vector of the k - th user after restoration,

[0164] represents the signal obtained by the actual user

[0165] 3.2.4) According to the two formulas in 3.2.2) and 3.2.3), the user achievable rate R k problem can be equivalently expressed as

[0166]

[0167] where \(w\) k is a constant auxiliary variable, maximizing \(R\) k when the variables \(w\) k and \(u\) k satisfy

[0168]

[0169]

[0170] 3.2.5) Substitute the formula in 3.2.4) into and expand the composite channel \(z\) k and separate the part related to the IRS reflection matrix to obtain

[0171]

[0172] where Further define the following variables

[0173]

[0174]

[0175]

[0176]

[0177]

[0178] 3.2.6) According to the variables defined in the formula of 3.2.6) and the expanded expression form of we can write the maximum \(R\) in the formula of 3.2.6) k and the expression form related to the IRS reflection matrix as

[0179]

[0180] Define the variable

[0181] 3.2.7) In summary, the original optimization problem represented by the formula in 3.2.3) can be rewritten in a form related to the IRS reflection matrix

[0182] (P3.3)

[0183] s.t. (C1):

[0184] (C2):

[0185] (C3):

[0186] (C4):

[0187] 3.2.8) 3.2.7) In the formula, only The constraint condition is non-convex. By using the SDR technique to omit this constraint condition, the final solvable convex-form optimization problem formula can be obtained.

[0188]

[0189] s.t. (P3.3) (C1)-(C3)

[0190] And the final optimization result is obtained by solving this optimization problem through CVX. The final optimization result needs to go through a sufficient number of Gaussian randomizations to obtain a sub-optimal solution for the current problem, and this sub-optimal solution is used to update the comprehensive channel z. k And return to step 3.2.3) until the optimization converges.

Claims

1. A method for passive IRS-assisted communication in a joint tactile Internet of Things cloud radio access network, characterized by maximizing the minimum waveform gain of each RRH in the direction of its respective area to be sensed, ensuring that the received SINR of each user is greater than a specific threshold, and optimizing the RRH transmitted signal and the passive IRS reflection matrix, specifically including the following steps: 1.1) The passive IRS-assisted C-RAN downlink communication and sensing integrated system includes a BBU pool and L RRHs; information is transmitted between the BBU pool and each RRH through a wired fronthaul link. Each RRH is equipped with a uniform linear antenna array ULA composed of N R antennas, responsible for sensing user information and simultaneously performing downlink communication with K single-antenna users with the assistance of I passive IRSs; each passive IRS is composed of M reflection units; the BBU pool knows the state information of all channels and the directions of the targets relative to each RRH. 1.2) The BBU pool first generates a signal where represents the signal sent to the l-th RRH, the matrix represents the communication beamforming matrix; the vector represents the sensing signal, and the vector represents the signal sent to each communication user; due to the limited capacity of the fronthaul link, the signal needs to be quantized and compressed before being transmitted to each RRH; the final signal set sent to all RHHs is: where represents the quantization and compression noise set for all RRHs, and simultaneously follows the distribution R0 represents the covariance matrix of the sensing signal, and the covariance matrix of the quantized and compressed signal is expressed as: R = WW H + R0 + Ω, 1.3) The compressed signal received by the l-th RRH through the forward compression link is expressed as: where the matrix represents the selection matrix related to the l-th RRH. Specifically, only the elements from the ((l - 1)N R + 1)-th row to the lN R -th row of this matrix are an identity matrix of dimension N R , and all the remaining elements are 0; 1.4) Let P R represent the transmit power limit of each RRH, and the transmit power limit expression of the l-th RRH is obtained: 1.5) For any communication user k, its received signal y k has the following expression: y k =(h k +g k ΨG)x + n k = z k x + n k , Among them represents the direct link between all RRHs and user k, Ψ represents the reflection matrix of the passive IRS, represents the reflection link between the active IRS and user k, represents the additive white Gaussian noise received by the user; z k =(h k +g k ΨG) represents the combined channel composed of the direct and reflection links.

2. The method for passive IRS-assisted communication in a joint communication and sensing cloud radio access network according to claim 1, wherein For the passive IRS-assisted C-RAN downlink communication and sensing integrated system, optimize the transmit signal covariance matrix R, the user transmit beamforming matrix W, and the passive IRS reflection matrix Ψ. With the goal of achieving the minimum waveform gain in the direction of the maximum area to be sensed, under the constraints of the transmit power of a single RRH, the multi-user SINR communication performance, and the fronthaul link capacity, the specific steps are as follows: 2.1) Each RRH is equipped with a uniform linear antenna array consisting of N R antennas, and the direction vector of the signal transmitted by the RRH is modeled as follows: where d RRH represents the spacing between adjacent RRH transmit antennas, and λ represents the wavelength of the transmitted signal; the waveform gain of the target of the l-th RRH at a specific direction angle θ is expressed as: 2.2) Maximize the minimum waveform gain of each RRH in the direction of its respective area to be sensed. Since each RRH needs to sense T targets, the objective function is expressed as: where the set represents all the targets that all RRHs need to sense; 2.3) At the same time, to ensure the multi-user communication performance, the received SINR of each user is greater than a specific threshold. The expression form of the SINR constraint condition for the k-th user is: where w i represents the i-th column in W, i.e., the beamforming vector of the i-th user, and Γ is the set minimum user SINR limit; 2.4) For the fronthaul compression link, the fronthaul link capacity constraint condition is expressed as: The capacity of the limited fronthaul link between each BBU and RRH is C l , denotes the mutual information between l and x 2.5) The optimization problem is expressed as: (C4): R = WW H + R0 + Ω。 3. A method for passive IRS-assisted communication in a joint sensing and communication cloud radio access network according to claim 2, wherein For the above optimization problem, considering the non-convexity of the objective function and the constraints, and the coupling between the optimization variables, the original problem is decomposed into two sub-problems for solution during the solution process; 3.1.1) Initialize the matrix R according to the RRH power constraint in 1.4), and rewrite R in the form of a covariance matrix: w k represents the k-th column of the communication beamforming matrix W, which is the beamforming vector for user k; represents the covariance matrix of each user's beamforming vector. According to the definition, R k is of rank one; 3.1.2) Rewrite the user SINR power constraint condition in 2.3) according to the formula in 2.2): The SINR constraint is a linear constraint on matrices R and R k and this expression form has nothing to do with w i ; 3.1.3) The formula in 2.4) adopts the link capacity constraint of the multi-source fronthaul link compression model. log|Ω|≤log|Σ|+Tr|Σ -1 Ω|-N, This inequality takes the equal sign only when the condition Ω = Σ holds; 3.1.4) To sum up, the optimization problem in 2.5) is rewritten in the following form: (C5): R≥0, Ω≥0 3.2.1) The SINR threshold for each user obtained according to sub-problem 1 is Γ. The sub-problem 2 is as follows: find the actual achievable rate R of the user k , maximize the difference between the achievable rate of each user and the minimum rate of each user, and the objective function is expressed as: 3.2.2) To ensure that R k must be greater than the minimum user - achievable rate, an auxiliary variable α is introduced. At the same time, it is required that this auxiliary variable is greater than or equal to 0, and the formula in 3.2.1) is rewritten as: R k -log(1 + Γ) ≥ α, α ≥ 0 3.2.3) According to the transmitted signal form in 1.3) and the channel expression form in 1.4), the user signal y is restored through the linear receiver u k for the user signal y k The MSE of this process is denoted by the symbol e k and is specifically expressed as: where w k is the user transmit beamforming vector of the k-th user after restoration, represents the signal obtained by the actual user; 3.2.4) According to the two formulas in 3.2.2) and 3.2.3), the user achievable rate R k The problem is equivalently expressed as: where w k is a constant auxiliary variable that maximizes R k When the variable w k and u k satisfy: 3.2.5) Substitute the formula in 3.2.4) into and expand the composite channel \(z\) k to obtain, by separating the part related to the IRS reflection matrix: Among them Re represents taking the modulus, and further defines the following variables: 3.2.6) According to the variables defined in the formula of 3.2.6) and the expanded expression form of write the maximum R in the formula of 3.2.6) k and the expression related to the IRS reflection matrix as: Define variables 3.2.7) To sum up, the original optimization problem expressed by the formula in 3.2.3) is rewritten in a form related to the IRS reflection matrix: 3.2.8) 3.2.7) Only in the formula The constraint condition is non-convex. Using the SDR technology, the constraint condition is omitted to obtain the final solvable convex form of the optimization problem formula: s.t.(P3.3)(C1)-(C3) And solve this optimization problem through CVX to obtain the final optimization result; the final optimization result needs to be subjected to Gaussian randomization to obtain a suboptimal solution to the current problem, and use this suboptimal solution to update the comprehensive channel z k And return to step 3.2.3) until the optimization converges.

Images