An IRS-aided multi-user interference channel sum-rate maximization phase optimization method

Abstract

The embodiment of the application provides a kind of IRS auxiliary multi-user interference channel sum rate maximization phase optimization method.Application in wireless communication technical field, this method is based on each transmitter to the channel state information of intelligent reflecting surface and the channel state information of intelligent reflecting surface to each receiver Establishment system sum rate maximization optimization problem;System sum rate maximization optimization problem is converted into the weighted interference leakage minimization problem under the constraint of unit module solution, and the initial reflection coefficient vector is obtained;With initial reflection coefficient vector as starting point, system sum rate maximization optimization problem is converted into the minimization problem on complex unit module manifold solution, and the final reflection coefficient vector is obtained;According to final reflection coefficient vector, the phase control instruction of intelligent reflecting surface is generated, and the phase control instruction is configured to each passive reflecting unit of intelligent reflecting surface.The method improves the convergence stability of IRS phase optimization, and improves the sum rate performance in multi-user interference channel.

Description

Technical Field

[0001] This application relates to the field of wireless communication technology, and in particular to a phase optimization method for maximizing the sum rate of an IRS-assisted multi-user interference channel. Background Technology

[0002] With the development of 5G and 6G mobile communication technologies, technologies such as Multiple-Input Multiple-Output (MIMO), multi-user access, and spectrum reuse have been widely used to improve the capacity and spectral efficiency of wireless communication systems. However, in scenarios where multiple users share the same spectrum resources, co-channel interference can occur between different users. Especially in multi-user interference channels, the desired link and the interfering link are coupled together, significantly limiting system performance and data rate. An Intelligent Reflecting Surface (IRS) is a programmable electromagnetic surface composed of a large number of low-power passive reflective elements. An IRS can reconstruct the wireless propagation environment by adjusting the phase of each reflective element, causing the incident wireless signal to be reflected according to a predetermined direction and phase relationship. Deploying an IRS between the transmitter and receiver can form an additional controllable reflection link, enhancing the desired signal at the target receiver and weakening the interfering signal at the non-target receiver. Therefore, phase optimization in multi-user single-input single-output (SISO) interference channels assisted by an IRS has become an important technical means to improve system performance and data rate.

[0003] Existing IRS phase optimization methods typically model the IRS phase design problem as a non-convex optimization problem and employ methods such as alternating optimization, iterative search after random initialization, Riemann conjugate gradient method, and continuous convex approximation method for solution. While these methods can yield feasible solutions under certain conditions, they still have the following shortcomings in scenarios with strong interference, high signal-to-noise ratio, or a large number of users:

[0004] On the one hand, the IRS phase optimization problem is highly non-convex, and the optimization result is sensitive to the initial phase. If random phase initialization or coarse initialization is used, subsequent iterations are prone to getting trapped in poor local solutions, resulting in some users' interference links not being effectively suppressed, and limited system and rate improvements. On the other hand, some existing initialization methods mainly focus on minimizing interference leakage, that is, minimizing the interference caused by non-target links to the receiver. However, in multi-user IRS-assisted interference channels, simply suppressing interference does not necessarily guarantee that the desired link will obtain sufficient effective signal gain. If the desired signal protection is ignored in the initialization phase, although the interference may be reduced, the effective received signal of the target user may also be weakened, thus affecting the subsequent system and rate optimization results. Summary of the Invention

[0005] This application provides a phase optimization method for maximizing the sum rate in an IRS-assisted multi-user interference channel, to address the problems of severe co-channel interference, strong non-convexity of the IRS phase optimization problem, and sensitivity of the optimization result to the initial phase in existing multi-user SISO interference channels. The method includes:

[0006] A smart reflector-assisted multi-user single-input single-output jamming channel system is constructed. The system includes: multiple single-antenna transmitters, multiple single-antenna receivers, and a smart reflector with multiple passive reflective elements.

[0007] Obtain the channel state information from each transmitter to the intelligent reflector and the channel state information from the intelligent reflector to each receiver. Based on the channel state information from each transmitter to the intelligent reflector and the channel state information from the intelligent reflector to each receiver, establish the system and rate maximization optimization problem.

[0008] The system and rate maximization optimization problem is transformed into a weighted interference leakage minimization problem under unity modulus constraints, and the initial reflection coefficient vector is obtained by solving the weighted interference leakage minimization problem;

[0009] Starting with the initial reflection coefficient vector, the system and rate maximization optimization problem is transformed into a minimization problem on a complex unity modulus manifold. Solving the minimization problem on the complex unity modulus manifold yields the final reflection coefficient vector.

[0010] The phase control command for the intelligent reflective surface is generated based on the final reflection coefficient vector, and the phase control command is configured to each passive reflective unit of the intelligent reflective surface.

[0011] Optionally, the establishment of the system and rate maximization optimization problem based on the channel state information from each transmitter to the smart reflector and the channel state information from the smart reflector to each receiver includes:

[0012] Construct the cascaded channel vector from the transmitter to the receiver via the intelligent reflector based on the channel state information from each transmitter to the intelligent reflector and the channel state information from the intelligent reflector to each receiver.

[0013] Based on the cascaded channel vectors from each transmitter to each receiver via the intelligent reflector, the transmit power of each user, and the noise power at the receiver, the maximum information transmission rate and system sum rate of each user are determined. Using the reflection coefficient vector of the intelligent reflector as the optimization variable, and under the condition that each passive reflection unit of the intelligent reflector satisfies the unit mode constraint, the optimization problem of maximizing system sum rate is established.

[0014] Optionally, constructing the cascaded channel vector from the transmitter to the receiver via the smart reflector based on the channel state information from each transmitter to the smart reflector and the channel state information from the smart reflector to each receiver includes:

[0015] Let the first The channel vector from each transmitter to the smart reflector is represented as follows: Intelligent reflective surface to the first The channel vector of each receiver is represented as follows: The reflection coefficient vector of the intelligent reflective surface is expressed as: ;

[0016] in, Represents a set of N-dimensional complex column vectors; Let represent the phase shift of the nth intelligent reflective surface passive reflective element, satisfying the unity modulus constraint: ;

[0017] According to the Channel vectors from transmitter to smart reflector and intelligent reflective surface to the first Channel vectors of each receiver Construct the first The transmitter is reflected by the smart reflector to the first... cascaded channel vectors of multiple receivers Its expression is:

[0018] ;

[0019] in, This means placing the vectors on the diagonal to form a diagonal matrix. Indicates complex conjugation. Represents the Hadamard element-wise product; when hour, Indicates the first The transmitter to the first The desired concatenated channel vector of each receiver; when hour, Indicates the first The transmitter to the first Interference concatenation channel vectors formed by multiple receivers.

[0020] Optionally, the step of determining the maximum information transmission rate and system rate for each user based on the cascaded channel vector from each transmitter to each receiver via the intelligent reflector, the transmit power of each user, and the noise power at the receiver, and using the reflection coefficient vector of the intelligent reflector as the optimization variable, establishes a system rate maximization optimization problem under the condition that each passive reflection unit of the intelligent reflector satisfies the unity modulus constraint, including:

[0021] The signal-to-interference-plus-noise ratio (SIR) of each user is determined based on the cascaded channel vectors from each transmitter to each receiver via the smart reflector, the transmit power of each user, and the noise power at the receiver. The signal-to-interference-plus-noise ratio (SIR) for an individual user is expressed as:

[0022] ;

[0023] in, Indicates the first Transmit power of each user Indicates the first Transmit power of each user Indicates the noise power at the receiving end;

[0024] The maximum data transmission rate for each user is determined based on the signal-to-interference-plus-noise ratio (SINR) of each user. The maximum information transmission rate for a single user is expressed as:

[0025] ;

[0026] The system and rate are determined based on the maximum information transmission rate of each user. Represented as:

[0027] ;

[0028] With the reflection coefficient vector of the intelligent reflective surface To optimize the variables, a smart reflector phase optimization problem is constructed with the objective of maximizing the system and rate. The expression is:

[0029] .

[0030] Optionally, the transformation of the system and rate maximization optimization problem into a weighted disturbance leakage minimization problem under unit modulus constraints includes:

[0031] Based on the cascaded channel vector Constructing the interference leakage equivalent correlation matrix Equivalent correlation matrix of the expected signal The expressions are as follows:

[0032] ;

[0033] ;

[0034] Introducing non-negative signal protection weights Construct a weighted matrix The expression is:

[0035] ;

[0036] Based on weighted matrix The weighted perturbation leakage minimization problem under the unit modulus constraint is expressed as:

[0037] .

[0038] Optionally, the initial reflection coefficient vector is obtained by solving the weighted interference leakage minimization problem, including:

[0039] For a finite candidate set Any signal protection weight in ,make And select parameters satisfy:

[0040] ;

[0041] in, Representation matrix The largest eigenvalue;

[0042] With a preset initial point Begin MM phase projection iteration, at the... In the next iteration, the auxiliary vector is calculated, expressed as:

[0043] ;

[0044] in, For the first Auxiliary vector for the next iteration Represents the identity matrix. For the first The reflection coefficient vector of the next iteration;

[0045] According to the The auxiliary vector in the next iteration updates the reflection coefficient vector, expressed as follows:

[0046] ;

[0047] in, For the first The reflection coefficient vector of the next iteration. This indicates taking the phase element by element. This indicates that the phase is mapped to the complex unit circle;

[0048] Repeat the MM phase projection iteration until the first-stage convergence threshold is met or the maximum number of iterations in the first stage is reached, thus obtaining the signal protection weights. The corresponding reflection coefficient vector ;

[0049] For a finite set of candidates All signal protection weights Repeat the above process, and adjust according to the system and rate. Select the optimal signal protection weight :

[0050] ;

[0051] Optimal signal protection weight The corresponding reflection coefficient vector is used as the initial reflection coefficient vector.

[0052] Optionally, the step of transforming the system and rate maximization optimization problem into a minimization problem on a complex unity modulus manifold, starting from the initial reflection coefficient vector, and solving the minimization problem on the complex unity modulus manifold to obtain the final reflection coefficient vector includes:

[0053] The set of reflection coefficient vectors satisfying the unity modulus constraint can be represented as a complex unity modulus manifold:

[0054] ;

[0055] The system and rate maximization optimization problem is transformed into a minimization problem on a complex unity modulus manifold, expressed as:

[0056] ;

[0057] Calculate the objective function using the initial reflection coefficient vector as the initial point. The Euclidean gradient is calculated and projected onto the complex unity modulus manifold at the current point. The tangent space at the point is used to obtain the objective function. The Riemann gradient;

[0058] Based on the objective function The Riemann gradient is used to iteratively solve the minimization problem on the complex unity modulus manifold, and the search direction is determined based on the tangent space shift and the difference in Riemann gradient between two adjacent iterations.

[0059] The step size is determined by backtracking search, the reflection coefficient vector is updated according to the search direction and step size, and the updated reflection coefficient vector is remapped to the complex unity modulus manifold by element-wise normalization operation.

[0060] When the preset second-stage convergence threshold is met or the maximum number of iterations in the second stage is reached, the final reflection coefficient vector is output.

[0061] This application provides a sum-rate maximization phase optimization method for IRS-assisted multi-user interference channels. This method is applied to a multi-user single-input single-output interference channel system comprising multiple single-antenna transmitters, multiple single-antenna receivers, and a smart reflector. Based on channel state information, a cascaded channel from the transmitter to the receiver via the smart reflector is constructed. A sum-rate maximization optimization problem is established under the condition that the reflection coefficients of the smart reflector satisfy the unity-mode constraint. A two-stage smart reflector phase optimization algorithm is used to solve this non-convex optimization problem. In the first stage, an initial reflection coefficient vector that balances interference suppression and desired signal protection is constructed using a signal protection adaptive weighted interference leakage minimization method. In the second stage, starting from this initial reflection coefficient vector, the sum-rate objective is directly refined on the complex unity-mode manifold using the Riemann finite memory algorithm to obtain the final reflection coefficient vector, which is then configured into each passive reflection element of the smart reflector. This method improves the convergence stability of the smart reflector phase optimization and enhances the sum-rate performance in multi-user interference channels. Attached Figure Description

[0062] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0063] Figure 1 This is a system model diagram of an IRS-assisted multi-user SISO interference channel provided in an embodiment of this application;

[0064] Figure 2 This is a flowchart illustrating an IRS-assisted multi-user interference channel sum rate maximization phase optimization method provided in an embodiment of this application.

[0065] Figure 3 This is a diagram showing the relationship between the system and the rate and transmission power provided in the embodiments of this application;

[0066] Figure 4 This is a diagram showing the relationship between the system and rate provided in the embodiments of this application and the number of IRS reflective units N;

[0067] Figure 5 This is a diagram showing the relationship between the system and the rate and the number of users K provided in the embodiments of this application.

[0068] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0069] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0070] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein.

[0071] In this application, the terms "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0072] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0073] Figure 2 This is a flowchart illustrating a method for maximizing the sum and rate phase optimization of an IRS-assisted multi-user interference channel, as provided in an embodiment of this application. Figure 2 As shown in the figure, this embodiment provides a phase optimization method for maximizing the sum rate of an IRS-assisted multi-user interference channel, which includes:

[0074] S1: Construct a smart reflector-assisted multi-user single-input single-output jamming channel system.

[0075] Figure 1 This is a system model diagram of an IRS-assisted multi-user SISO interference channel provided in an embodiment of this application. For example... Figure 1 As shown, the system includes K single-antenna transmitters, K single-antenna receivers, and an IRS with N passive reflector elements. The K transmitters and K receivers transmit signals on the same time-frequency resource, therefore the... When a receiver receives the desired signal from its corresponding transmitter, it is simultaneously subjected to co-channel interference from signals transmitted by other transmitters.

[0076] In this embodiment, the direct link between the transmitter and receiver is blocked, ignored, or not considered an effective transmission link in the phase optimization model. The transmitted signal is transmitted through a cascaded reflection link of transmitter-IRS-receiver. The IRS enhances the desired signal of the target user and suppresses interference signals generated by other users by adjusting the phase of each passive reflection unit.

[0077] S2: Obtain the channel state information from each transmitter to the intelligent reflector and the channel state information from the intelligent reflector to each receiver. Based on the channel state information from each transmitter to the intelligent reflector and the channel state information from the intelligent reflector to each receiver, establish the system and rate maximization optimization problem.

[0078] Specifically, let the first The channel vectors from each transmitter to the IRS are: IRS to the The channel vectors of each receiver are ,in Represents the set of N-dimensional complex column vectors. The IRS reflection coefficient vector is represented as:

[0079] ;

[0080] in, Let represent the phase shift of the nth IRS passive reflector element, satisfying the unity mode constraint: .

[0081] Based on the channel vector from the transmitter to the IRS and the channel vector from the IRS to the receiver, construct the first... The transmitter is reflected by the IRS to the first Cascaded channel vectors of multiple receivers:

[0082] ;

[0083] in, This means placing the vectors on the diagonal to form a diagonal matrix. Indicates complex conjugation. Represents the Hadamard element-wise product; when hour, Indicates the first The transmitter to the first The desired concatenated channel vector of each receiver; when hour, Indicates the first The transmitter to the first Interference concatenation channel vectors formed by multiple receivers.

[0084] No. The received signal at each receiver is represented as follows:

[0085] ;

[0086] in, Indicates the first The baseband signal transmitted by each transmitter satisfies and , Indicates the first The baseband signal transmitted by each transmitter Indicates the first Additive white Gaussian noise at the receiver This indicates the conjugate transpose.

[0087] Based on the above received signal model, the first Signal-to-interference-plus-noise ratio per user Represented as:

[0088] ;

[0089] in, Indicates the first Transmit power of each user Indicates the first Transmit power of each user Indicates the noise power at the receiving end; the first The reachable rate for a user is represented as:

[0090] ;

[0091] System and rate Represented as:

[0092] .

[0093] Therefore, the IRS phase optimization problem in this embodiment is to maximize the system and rate while ensuring that the IRS reflection coefficient satisfies the unity-mode constraint. The optimization variable for this problem is the IRS reflection coefficient vector. Its constraints are The problem has n = 1, 2, ..., N. Because this problem is non-convex, this invention employs a two-stage phase optimization method to solve it.

[0094] S3: The system and rate maximization optimization problem is transformed into a weighted interference leakage minimization problem under unit modulus constraints. The initial reflection coefficient vector is obtained by solving the weighted interference leakage minimization problem.

[0095] Specifically, based on the concatenated channel vector Constructing the interference leakage equivalent correlation matrix Equivalent correlation matrix of the expected signal :

[0096] ;

[0097] .

[0098] Introducing non-negative signal protection weights And construct the weighted matrix:

[0099] .

[0100] when When the optimization objective is primarily used to suppress total interference leakage; when At the same time, the optimization objective is to introduce a desired signal protection term while suppressing interference leakage, thereby avoiding excessive weakening of the desired link signal during the initialization phase.

[0101] For a given Solve the following weighted perturbation leakage minimization problem under the unit modulus constraint:

[0102]

[0103] This optimization problem is a non-convex quadratic programming problem with a unit modulus constraint; for a finite candidate set any of ,make Choose parameter μ to satisfy ,in Representation matrix The largest eigenvalue. In practical implementation, to reduce computational complexity, we can let ,in This represents the Frobenius norm.

[0104] For that The corresponding IRS reflection coefficient vector is solved by MM phase projection iteration, which is a phase-constrained iterative solution method based on the Majorization-Minimization optimization framework.

[0105] First set the initial point Preferably, Depend on The eigenvector corresponding to the largest eigenvalue of the unit norm get:

[0106] ;

[0107] .

[0108] In the t-th iteration, calculate the auxiliary vector:

[0109] ;

[0110] Then update the IRS reflection coefficient vector as follows:

[0111] ;

[0112] in, Represents the identity matrix. This indicates taking the phase element by element. This means mapping the phase to the complex unit circle so that the updated IRS reflection coefficient vector satisfies the unit mode constraint.

[0113] For the present Repeat the above MM phase projection iteration until the first-stage convergence condition is met or the maximum number of iterations in the first stage is reached, to obtain the result. The corresponding IRS reflection coefficient vector The convergence condition for the first stage can be set as follows:

[0114] ;

[0115] in, , This is the convergence threshold for the first stage. In one specific embodiment, Pick Maximum number of iterations in the first phase Take 100 to 300 times.

[0116] For candidate set All of them Repeat the above process to obtain multiple candidate IRS reflection coefficient vectors. Then, each Substitute the system and rate function Evaluate and select:

[0117] .

[0118] Finally, the optimal signal protection weights are determined. The corresponding IRS reflection coefficient vector is used as the initial point for the second stage of optimization:

[0119] .

[0120] Understandably, in the first stage, a signal protection weighted target is constructed by using the interference leakage equivalent correlation matrix and the desired signal equivalent correlation matrix, and the signal protection weight is adaptively selected from a finite candidate set, thereby obtaining a high-quality initial reflection coefficient vector that balances interference suppression and desired signal enhancement.

[0121] In one specific embodiment, the candidate set It can be set to:

[0122] .

[0123] In another embodiment, the candidate set It can also be set to normalized form:

[0124]

[0125] in, Represents the trace of a matrix. To prevent positive numbers with a denominator of zero.

[0126] S4: Starting with the initial reflection coefficient vector, the system and rate maximization optimization problem is transformed into a minimization problem on a complex unity modulus manifold. Solving the minimization problem on the complex unity modulus manifold yields the final reflection coefficient vector.

[0127] Specifically, the initial reflection coefficient vector obtained in step S3 above Starting from a point, the system and rate objectives are refined and optimized on a complex unity manifold. The feasible set of IRS reflection coefficient vectors is represented as:

[0128]

[0129] The system and rate maximization problem can be equivalently transformed into a minimization problem on a complex unity modulus manifold:

[0130] .

[0131] For the Individual users, define total received power and interference plus noise power :

[0132]

[0133] Then the first The rate for a single user can be expressed as:

[0134] .

[0135] achievable The Euclidean gradient is:

[0136]

[0137] Based on the optimization problem objective function Euclidean gradient It can be represented as:

[0138]

[0139] Project the Euclidean gradient onto the complex unity modulus manifold at the current point. The Riemann gradient is obtained from the tangent space at the given point:

[0140] ;

[0141] in, Indicates taking the real part, This represents the final reflection coefficient vector.

[0142] Based on the objective function Riemann gradient The optimization problem is solved iteratively using the complex unity modulus manifold Riemann finite memory Broyden-Fletcher-Goldfarb-Shanno algorithm. Record the first The next iteration point is Save the displacement-gradient difference pairs of the most recent m steps. (subscript) (This represents the historical iteration step number). Wherein, the displacement vector... Defined as the actual tangent space shift after contraction mapping between two adjacent iteration points, i.e. ,in For the first The step size of the next iteration. For the first The search direction for the next iteration. This represents the actual tangential spatial displacement. For vector transport operators. Gradient difference vector. Defined as the difference between the Riemann gradient of the current iteration point and the previous iteration point, i.e. By performing two recursive operations on the displacement-gradient difference pair, the inverse Hessian action in the tangent space of the complex unity modulus manifold is approximately constructed, thus obtaining the current iteration point. Search direction The recursive process is as follows:

[0143] First recursion (downward loop): from arrive Calculate in sequence initial value Second recursion (upward loop): from arrive Calculate sequentially: initial value The search direction was finally obtained. .

[0144] It should be noted that all inner products in the two recursions mentioned above... Both vector transport and vector transport are at the current point. It is executed in the tangent space, therefore the historical displacement vector needs to be... and gradient difference vector Projected onto the current point using the vector transport operator. Corresponding tangent space: For complex unit module manifolds Vector transport operator Defined as the orthogonal projection of the tangent space vectors: ,in This operator maps the tangent vector to tangent space .

[0145] Preferably, the finite memory parameter m is between 5 and 20; in this embodiment, m is 10.

[0146] Get search direction Then, a backtracking search is used to determine the step size. The backtracking search is based on the Armijo criterion: starting from the initial step size... In the beginning, if If this is not true, then use the contraction factor. Reduce the step size, that is Until the above conditions are met. .in, For shrinking mapping, To fully reduce the parameters.

[0147] In this embodiment, the initial step size is 1, and the step size shrinkage factor is... Take 0.5 to fully reduce the parameter. Pick .

[0148] Determine step size Then, the update on the manifold is completed by shrinking the map: For complex unit module manifolds Shrinking Map Specifically, element-wise normalization is used:

[0149]

[0150] This operation ensures that the IRS reflection coefficient remains on the complex unity manifold after each update.

[0151] When the second-stage convergence condition is met or the maximum number of iterations in the second stage is reached, the final IRS reflection coefficient vector is output. The convergence condition for the second stage is set as follows: ;

[0152] in, This is the convergence threshold for the second stage. In this embodiment, Pick Maximum number of iterations in the second stage Take 100 to 200 times.

[0153] In this embodiment, the second stage starts with the initial reflection coefficient vector and directly optimizes the original sum and rate objectives on the complex unity modulus manifold. It also utilizes the finite memory displacement-gradient difference in the Riemann finite memory Broyden-Fletcher-Goldfarb-Shanno algorithm to approximate the inverse Hessian action, thereby improving the quality of the search direction and the convergence stability.

[0154] S5: Generate the phase of the smart reflective surface based on the final reflection coefficient vector, and configure the phase to each passive reflective unit of the smart reflective surface.

[0155] Specifically, the phase information of each complex-valued reflection coefficient in the vector is extracted, the continuous phase values ​​are mapped to discrete phase levels supported by the hardware, and the phase command is sent to each passive reflection unit through a wired / wireless control link, thus completing the phase calibration and synchronization configuration of the entire array.

[0156] like Figure 1 As shown, in a specific simulation embodiment, the IRS uses a uniform rectangular array, with its array plane located in the x=0 plane of a three-dimensional Cartesian coordinate system, and the reference point set to (0,0,0)m. K single-antenna transmitters and K single-antenna receivers are all located on a horizontal plane at z=-20m. The K single-antenna transmitters are randomly distributed in a rectangular region [5,45]×[-45,-5], and the K single-antenna receivers are randomly distributed in a rectangular region [5,45]×[5,45]. The transmitter-IRS link and the IRS-receiver link adopt a Ricean fading channel model. The system bandwidth is set to 10MHz, and the noise power spectral density is set to -170dBm / Hz. It is assumed that the communication optimization control equipment can obtain the channel state information of the transmitter-IRS link and the IRS-receiver link, that the signals transmitted by each user are independent, and that the transmission power of each user can be the same or configured according to a preset power.

[0157] In different embodiments, the number of IRS reflector units N can be 64, 128, 144, 256, or 512; the number of users K can be 2 to 20; and the transmit power of each user... The value can be selected within the range of 20dBm to 40dBm. Specifically, in one embodiment, N=144 and K=8, used to analyze the system and rate variations with transmit power; in another embodiment, K=6. =35dBm, used to analyze the system and rate as the number of IRS reflective units N changes.

[0158] The present invention provides an IRS-assisted multi-user interference channel sum rate maximization phase optimization method, which includes a first stage of signal protection adaptive weighted interference leakage minimization initialization and a second stage of complex unity modulus manifold Riemann finite memory Broyden-Fletcher-Goldfarb-Shanno refinement optimization, wherein the signal protection adaptive weighted interference leakage minimization is denoted as SP-AWILM and the Riemann finite memory Broyden-Fletcher-Goldfarb-Shanno is denoted as RLBFGS.

[0159] Figure 3 and Figure 4 The demonstrated algorithm performance includes benchmark methods, ablation comparison methods, and the complete method of this invention. The benchmark methods include random initialization (Random init.), random initialization combined with the Riemann conjugate gradient algorithm (Random init.+RCG), and random initialization combined with the successive convex approximation algorithm (Random init.+SCA). The ablation comparison methods include random initialization combined with the RLBFGS refinement method (Random init.+RLBFGS) and the SP-AWILM initialization method alone. The complete method of this invention is a two-stage IRS phase optimization method using SP-AWILM-RLBFGS. Through the above comparisons, the contributions of the RLBFGS refinement stage, the SP-AWILM initialization stage, and the two-stage joint structure to the sum and rate improvement can be verified.

[0160] Please see Figure 3 , Figure 3 This is a diagram showing the relationship between the system and the rate of transmission power provided in the embodiments of this application. From... Figure 3It can be seen that the summation rate of the random initialization method, Random initialization, remains consistently low. Random initialization + RCG and Random initialization + SCA show significant improvements over Random initialization, indicating that subsequent iterative optimization based on random initial phases can improve IRS phase configuration, but its performance is still affected by the random initial point. Random initialization + RLBFGS is generally superior to Random initialization + RCG and Random initialization + SCA, indicating that under the same random initialization conditions, RLBFGS refinement stage utilizes the finite memory shift-gradient difference to approximate the inverse Hessian action, which can obtain a more efficient search direction, thereby improving the summation rate. When SP-AWILM is used alone, its summation rate is significantly higher than that of Random initialization and continues to increase with the increase of transmission power, indicating that SP-AWILM initialization can protect the desired signal while suppressing interference leakage, providing a higher quality initial solution for IRS phase design. Furthermore, the complete method of this invention, SP-AWILM-RLBFGS, achieves the highest sum and rate across the entire transmission power range, demonstrating that combining high-quality initialization of SP-AWILM with refined optimization of RLBFGS can more effectively improve the system sum and rate under conditions of high transmission power and strong co-channel interference.

[0161] Please see Figure 4 , Figure 4 This is a graph showing the relationship between the system and rate of the embodiment of this application and the number of IRS reflective units N. From... Figure 4 It can be seen that as the number of IRS reflection units N increases, the sum rate of all methods except Random init. shows an upward trend, indicating that increasing the IRS size can provide more adjustable phase degrees of freedom, thereby enhancing the system's ability to control the desired link and interference link. Random init.+RLBFGS outperforms Random init.+RCG and Random init.+SCA, further verifying that the RLBFGS refinement stage has better optimization capabilities than conventional first-order search or approximate optimization methods. When SP-AWILM is used alone, its sum rate continuously increases with the increase of N, indicating that SP-AWILM can effectively coordinate the newly added IRS reflection units, forming a more reasonable phase configuration between interference leakage suppression and desired signal protection. The complete method of this invention, SP-AWILM-RLBFGS, achieves the highest sum rate for all values ​​of N, indicating that SP-AWILM initialization and RLBFGS refinement have complementary effects, and can make fuller use of the phase degrees of freedom brought about by the increase in the number of IRS reflection units.

[0162] Please see Figure 5 , Figure 5This is a graph showing the relationship between the system and speed, and the number of users K, as provided in the embodiments of this application. From... Figure 5 It can be seen that, under different numbers of IRS reflector units N, the system speed and rate initially increase and then decrease with the increase of the number of users K. When K is small, the addition of new users can bring multi-user multiplexing gain, thus increasing the system speed and rate. As K further increases, multi-user co-channel interference is significantly enhanced, and the limited number of IRS reflector units cannot simultaneously meet the desired signal enhancement and interference suppression requirements of all users, so the system speed and rate begin to decrease. Meanwhile, the larger the number of IRS reflector units N, the higher the overall system speed and rate, and the number of users corresponding to the peak rate shifts towards a larger K. This indicates that a larger-scale IRS can provide more degrees of freedom in phase modulation, improving the system's ability to serve more users.

[0163] This invention combines SP-AWILM initialization and RLBFGS refinement, enabling the IRS phase design process to achieve good interference suppression and signal protection during the initialization stage, and to directly optimize the system and rate targets using quasi-Newton curvature information during the refinement stage.

[0164] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0165] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method for maximizing the sum rate and phase optimization in an IRS-assisted multi-user interference channel, characterized in that, The method includes: A smart reflector-assisted multi-user single-input single-output jamming channel system is constructed. The system includes: multiple single-antenna transmitters, multiple single-antenna receivers, and a smart reflector with multiple passive reflective elements. Obtain the channel state information from each transmitter to the intelligent reflector and the channel state information from the intelligent reflector to each receiver. Based on the channel state information from each transmitter to the intelligent reflector and the channel state information from the intelligent reflector to each receiver, establish the system and rate maximization optimization problem. The system and rate maximization optimization problem is transformed into a weighted interference leakage minimization problem under unity modulus constraints, and the initial reflection coefficient vector is obtained by solving the weighted interference leakage minimization problem; Starting with the initial reflection coefficient vector, the system and rate maximization optimization problem is transformed into a minimization problem on a complex unity modulus manifold. Solving the minimization problem on the complex unity modulus manifold yields the final reflection coefficient vector. The phase control command for the intelligent reflective surface is generated based on the final reflection coefficient vector, and the phase control command is configured to each passive reflective unit of the intelligent reflective surface.

2. The method according to claim 1, characterized in that, The system and rate maximization optimization problem established based on the channel state information from each transmitter to the smart reflector and the channel state information from the smart reflector to each receiver includes: Construct the cascaded channel vector from the transmitter to the receiver via the intelligent reflector based on the channel state information from each transmitter to the intelligent reflector and the channel state information from the intelligent reflector to each receiver. Based on the cascaded channel vectors from each transmitter to each receiver via the intelligent reflector, the transmit power of each user, and the noise power at the receiver, the maximum information transmission rate and system sum rate of each user are determined. Using the reflection coefficient vector of the intelligent reflector as the optimization variable, and under the condition that each passive reflection unit of the intelligent reflector satisfies the unit mode constraint, the optimization problem of maximizing system sum rate is established.

3. The method according to claim 1, characterized in that, The construction of the cascaded channel vector from the transmitter to the receiver via the smart reflector, based on the channel state information from each transmitter to the smart reflector and the channel state information from the smart reflector to each receiver, includes: Let the first The channel vector from each transmitter to the smart reflector is represented as follows: Intelligent reflective surface to the first The channel vector of each receiver is represented as follows: The reflection coefficient vector of the intelligent reflective surface is expressed as: ; in, Represents a set of N-dimensional complex column vectors; Let represent the phase shift of the nth intelligent reflective surface passive reflective element, satisfying the unity modulus constraint: ; According to the Channel vectors from transmitter to smart reflector and intelligent reflective surface to the first Channel vectors of each receiver Construct the first The transmitter is reflected by the smart reflector to the first... cascaded channel vectors of multiple receivers Its expression is: ； in, This means placing the vectors on the diagonal to form a diagonal matrix. Indicates complex conjugation. Represents the Hadamard element-wise product; when hour, Indicates the first The transmitter to the first The desired concatenated channel vector of each receiver; when hour, Indicates the first The transmitter to the first Interference concatenation channel vectors formed by multiple receivers.

4. The method according to claim 3, characterized in that, The maximum information transmission rate and system rate for each user are determined based on the cascaded channel vector from each transmitter to each receiver via the intelligent reflector, the transmit power of each user, and the noise power at the receiver. Using the reflection coefficient vector of the intelligent reflector as the optimization variable, and under the condition that each passive reflection unit of the intelligent reflector satisfies the unit modulus constraint, a system rate maximization optimization problem is established, including: The signal-to-interference-plus-noise ratio (SIR) of each user is determined based on the cascaded channel vectors from each transmitter to each receiver via the smart reflector, the transmit power of each user, and the noise power at the receiver. The signal-to-interference-plus-noise ratio (SIR) for an individual user is expressed as: ； in, Indicates the first Transmit power of each user Indicates the first Transmit power of each user Indicates the noise power at the receiving end; The maximum data transmission rate for each user is determined based on the signal-to-interference-plus-noise ratio (SINR) of each user. The maximum information transmission rate for a single user is expressed as: ； The system and rate are determined based on the maximum information transmission rate of each user. Represented as: ； With the reflection coefficient vector of the intelligent reflective surface To optimize the variables, a smart reflector phase optimization problem is constructed with the objective of maximizing the system and rate. The expression is: 。 5. The method according to claim 3, characterized in that, The transformation of the system and rate maximization optimization problem into a weighted disturbance leakage minimization problem under unit modulus constraints includes: Based on the cascaded channel vector Constructing the interference leakage equivalent correlation matrix Equivalent correlation matrix of the expected signal The expressions are as follows: ； ； Introducing non-negative signal protection weights Construct a weighted matrix The expression is: ； Based on weighted matrix The weighted perturbation leakage minimization problem under the unit modulus constraint is expressed as: 。 6. The method according to claim 5, characterized in that, Solving the weighted interference leakage minimization problem yields the initial reflection coefficient vector, which includes: For a finite candidate set Any signal protection weight in ,make And select parameters satisfy: ； in, Representation matrix The largest eigenvalue; With a preset initial point Begin MM phase projection iteration, at the... In the next iteration, the auxiliary vector is calculated, expressed as: ； in, For the first Auxiliary vector for the next iteration Represents the identity matrix. For the first The reflection coefficient vector of the next iteration; According to the The auxiliary vector in the next iteration updates the reflection coefficient vector, expressed as follows: ； in, For the first The reflection coefficient vector of the next iteration. This indicates taking the phase element by element. This indicates that the phase is mapped to the complex unit circle; Repeat the MM phase projection iteration until the first-stage convergence threshold is met or the maximum number of iterations in the first stage is reached, thus obtaining the signal protection weights. The corresponding reflection coefficient vector ; For a finite set of candidates All signal protection weights Repeat the above process, and adjust according to the system and rate. Select the optimal signal protection weight : ； Optimal signal protection weight The corresponding reflection coefficient vector is used as the initial reflection coefficient vector.

7. The method according to claim 4, characterized in that, The process begins by using the initial reflection coefficient vector as a starting point, transforming the system and rate maximization optimization problem into a minimization problem on a complex unity modulus manifold. Solving this minimization problem on the complex unity modulus manifold yields the final reflection coefficient vector, including: The set of reflection coefficient vectors satisfying the unity modulus constraint can be represented as a complex unity modulus manifold: ； The system and rate maximization optimization problem is transformed into a minimization problem on a complex unity modulus manifold, expressed as: ； Calculate the objective function using the initial reflection coefficient vector as the initial point. The Euclidean gradient is calculated and projected onto the complex unity modulus manifold at the current point. The tangent space at the point is used to obtain the objective function. The Riemann gradient; Based on the objective function The Riemann gradient is used to iteratively solve the minimization problem on the complex unity modulus manifold, and the search direction is determined based on the tangent space shift and the difference in Riemann gradient between two adjacent iterations. The step size is determined by backtracking search, the reflection coefficient vector is updated according to the search direction and step size, and the updated reflection coefficient vector is remapped to the complex unity modulus manifold by element-wise normalization operation. When the preset second-stage convergence threshold is met or the maximum number of iterations in the second stage is reached, the final reflection coefficient vector is output.

Images