Precoding optimization method in RIS-assisted cellular-free large-scale MIMO and ISAC system

Abstract

The invention discloses a precoding optimization method in an RIS-assisted cellular-free large-scale MIMO and ISAC system, and the method comprises the steps: building a joint optimization problem about a precoding vector by taking the sum maximization of a multi-user weighted communication rate as an optimization target, and taking the constraint of the transmitting power of each access point and the constraint that the sensing signal to interference plus noise ratio is not lower than a preset threshold value as the constraint; performing conversion processing on the joint optimization problem through a fractional programming FP algorithm, and decoupling the joint optimization problem into a communication pre-coding sub-problem and a sensing pre-coding sub-problem based on an alternate optimization framework; and by alternately and iteratively solving the two sub-problems, updating and outputting the optimized communication pre-coding vector and the sensing pre-coding vector, and configuring to each access point. Compared with the prior art, the method has the advantages that joint optimization is performed on communication and sensing precoding through a method of combining fractional planning and alternate optimization, and the multi-user communication rate is improved on the premise of ensuring sensing performance and power constraint.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and in particular to a precoding optimization method in RIS-assisted non-cellular massive MIMO and ISAC systems. Background Technology

[0002] To meet the urgent demand for higher-performance wireless networks in the 6G era, Intelligent Metasurfaces (RIS) have been introduced into non-cellular massive MIMO and Integrated Communication and Sensing (ISAC) systems. In complex deployment environments, direct propagation paths between base stations and targets / users are often physically blocked by obstacles such as buildings and terrain, leading to interruptions in sensing and communication links. RIS can overcome these obstacles and collaboratively enhance communication and sensing capabilities by dynamically adjusting wireless channels and intelligently constructing reflection paths. In this architecture, precoding design is the core factor affecting the overall system performance. It needs to coordinate and schedule communication and sensing waveforms within a unified framework to address the challenges posed by distributed nodes, reconfigurable channels, and shared interference.

[0003] Existing precoding optimizations for RIS-assisted non-cellular massive MIMO and ISAC systems have the following limitations: First, the modeling completely separates communication and sensing functions, failing to fully consider the interference of sensing signals on communication users and the impact of communication signals on sensing echoes, resulting in a mismatch between the model and the actual propagation environment. Second, the use of simplified maximum-minimum fairness criteria or single performance indicators fails to achieve a balance between overall communication and sensing efficiency centered on the system weighted sum rate. Third, due to the non-convex objective function and constraint coupling, existing algorithms often have high complexity and poor convergence, making it difficult to support real-time configuration and large-scale deployment.

[0004] A search revealed Chinese Patent Publication No. CN119921816A, which discloses an optimization method for fluid antenna position and radar communication precoding matrix in URLLC scenarios. This method establishes an optimization model centered on maximizing the radar sensing signal-to-noise ratio, constrained by the total base station power, URLLC delay, and fluid antenna position. An alternating optimization algorithm is then used to jointly solve for the antenna position, radar precoding matrix, and communication precoding matrix, achieving coordinated design of communication and sensing functions in a single-base station scenario. However, the precoding optimization design of this method is tightly coupled to the mechanical position adjustment of the fluid antenna, resulting in complex engineering implementation and limited dynamics. Its focus on sensing performance as a single central objective makes it difficult to balance fairness and efficiency in multi-user communication rates. Furthermore, its overall alternating optimization process involves multiple strongly coupled variables and complex constraint handling, leading to significant computational overhead.

[0005] Therefore, optimizing communication and sensing precoding vectors in cellular-free massive MIMO and ISAC systems with physical obstructions to maximize communication rate while ensuring sensing performance is a technical problem that needs to be solved. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a precoding optimization method for RIS-assisted non-cellular massive MIMO and ISAC systems.

[0007] The objective of this invention can be achieved through the following technical solutions: According to a first aspect of the present invention, a precoding optimization method is provided in a RIS-assisted non-cellular massive MIMO and ISAC system, the system comprising multiple distributed access points controlled by a central processing unit, a smart surface, at least one user equipment, and at least one target to be sensed, the method comprising: With the goal of maximizing the multi-user weighted communication rate, and constrained by the transmit power of each access point and the requirement that the perceived signal-to-interference-plus-noise ratio is not lower than a preset threshold, a joint optimization problem for the precoding vector is established. The objective function of the joint optimization problem is transformed by fractional programming (FP) algorithm to obtain a smoothed objective function; Based on the smoothed objective function and the constraints, an alternating optimization framework is adopted to decouple the transformed joint optimization problem into a communication precoding subproblem and a sensing precoding subproblem, and solve them through an alternating iterative method. In each iteration, the sensing precoding vector is fixed to solve the communication precoding subproblem, and the communication precoding vector is fixed to solve the sensing precoding subproblem. Update and output the optimized communication precoding vector and perception precoding vector, and configure them to each access point.

[0008] As a preferred technical solution, the optimization objective is: , in, Let be the weight coefficient of the i-th user device. Let be the signal-to-interference-plus-noise ratio (SIR) of the i-th user equipment. Let be the precoding vector for the transmission access point to user equipment i. This is the pre-coded vector for the transmission access point to sense the target.

[0009] As a preferred technical solution, the access point transmit power constraint is: for the k-th transmit access point, its total transmit power does not exceed a preset maximum transmit power threshold. .

[0010] As a preferred technical solution, the joint optimization problem is solved using the fractional programming (FP) algorithm, specifically including: Introducing the first auxiliary variable Second auxiliary variable The optimization objective is transformed into a smoothed objective function. : , in, Let i be the weighting coefficient for user device i. Let be the equivalent communication channel vector for the i-th user. The total number of user devices. Let be the precoding vector for the transmission access point to user equipment i. This is the precoded vector for the transmission access point to sense the target. and The power allocation factor, This represents noise power.

[0011] As a preferred technical solution, in each iteration, the first auxiliary variable Second auxiliary variable Update based on the precoding vector of the current iteration, specifically as follows: , , in, Let be the weight coefficient of the i-th user device. Let be the signal-to-interference-plus-noise ratio (SIR) of the i-th user equipment. For sensing precoded vectors, Let i be the communication precoding vector of the i-th user equipment. Let i be the total channel vector of the i-th user equipment. and The power allocation factor, This represents noise power.

[0012] As a preferred technical solution, solving the perceptual precoding sub-problem specifically includes: Under the condition of fixed communication precoding vector, based on the smoothed objective function We construct a sensing precoding subproblem with the objective of minimizing the interference term of the sensing signal received by each user equipment, while satisfying the transmit power constraint and the sensing signal-to-interference-plus-noise ratio constraint. The non-convex constraints in the perceptual precoding subproblem are transformed into convex constraints using a first-order Taylor expansion. A convex optimization solver is used to solve the transformed perceptual precoding subproblem, resulting in an updated perceptual precoding vector.

[0013] As a preferred technical solution, solving the communication precoding subproblem specifically includes: Under the condition of fixed sensing precoding vectors, the communication precoding vectors of all user equipment are combined into a communication precoding matrix; Based on the smoothed objective function We construct a communication precoding subproblem with the objective of maximizing the difference between the linear and quadratic terms, while satisfying transmit power constraints and sensing signal-to-interference-plus-noise ratio constraints. The communication precoding subproblem is solved using a convex optimization solver to obtain an updated communication precoding matrix, from which the communication precoding vectors of each user equipment are recovered.

[0014] As a preferred technical solution, the objective function of the communication precoding subproblem is in the form of: ,in, For communication precoding matrix, and This is a matrix related to the channel and auxiliary variables.

[0015] As a preferred technical solution, the iteration termination condition for the alternating optimization is: the difference between the objective function values ​​obtained from two consecutive iterations is less than a preset tolerance, or the number of iterations reaches the preset maximum number of iterations.

[0016] As a preferred technical solution, before starting the alternating iterative optimization, the method further includes initializing the sensing precoding vector and the communication precoding vector; wherein, the sensing precoding vector is initialized as a beamforming vector pointing in the direction of the target to be sensed, and the communication precoding vector is initialized as a maximum ratio transmission precoding vector based on the channel state information of each user.

[0017] Compared with the prior art, the present invention has the following advantages: 1. This invention addresses non-cellular massive MIMO and ISAC systems with physical obstruction under RIS assistance. By optimizing precoding with the goal of maximizing multi-user weighted communication rate and constrained by ensuring the sensing signal-to-interference-plus-noise ratio and distributed power limitations, the multi-user weighted communication rate is effectively improved while ensuring sensing performance.

[0018] 2. In solving complex non-convex optimization problems, this invention adopts an algorithm framework that combines fractional programming and alternating optimization. First, the non-convex fractional terms in the objective function are transformed into a smooth form that is easy to handle using the fractional programming algorithm. Then, the original coupled problem is decoupled into two relatively independent and iteratively solvable subproblems, communication precoding and sensing precoding, through the alternating optimization framework. This reduces the algorithm complexity and improves the optimization efficiency and feasibility.

[0019] 3. This invention proposes a convex approximation method based on first-order Taylor expansion to address the non-convex constraints that still exist in the decoupled subproblems. This transforms the subproblems into standard convex optimization problems, which can be solved efficiently using mature convex optimization solvers, ensuring the engineering feasibility and computational robustness of the algorithm. Attached Figure Description

[0020] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a schematic diagram of the signal transmission model of the RIS-assisted non-cellular MIMO and ISAC system of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0022] Example 1: like Figure 1 As shown, this invention provides a precoding vector optimization method, which is applied to RIS-assisted non-cellular massive MIMO and ISAC systems.

[0023] like Figure 2 As shown, each access point (AP) in this system can act as either an ISAC transmitter or receiver. The number of transmitting and receiving APs are denoted as follows: and Each AP is equipped with M Each AP has one transmit or receive antenna. All APs are centrally controlled by a central processing unit (CPU). This unit also houses a RIS controller for adjusting the RIS phase shift, with a phase shift matrix of... .

[0024] A centralized, non-cellular, massive MIMO system is used for communication and sensing processing. Each ISAC transmitting AP provides joint service by transmitting centralized precoded signals. A single-antenna user equipment (UE); simultaneously, a sensing symbol sequence that shares waveform and time-frequency resources with communication symbols can be integrated during transmission. In this case, Each transmitting AP will jointly focus its detection beam on potential targets at known locations; on the other hand... Each receiving AP synchronously receives the sensing echo, but since all APs are physically obstructed from the target, target sensing relies on RIS.

[0025] In this system model, Representing the Zero-mean downlink unit power communication symbols for each user equipment, i.e. Sensing signals They are also assumed to have zero mean and unit power, i.e. Assuming all signals are independent, this means that for any All meet Therefore, at time From the launch access point Transmission signal sent It can be represented as: (1) Among them, for , and These represent the transmission access points respectively. To user equipment Communication precoding vectors and from the access point The emitted perceptual precoding vector; This represents the given power allocation coefficient.

[0026] Each user equipment is based on all The channel state information (CSI) of each distributed transmit antenna is used for joint selection. Therefore, each transmit AP selects the user equipment. The precoding vectors can be concatenated into a single unified precoding vector: (2) Similarly, the precoding vectors of each transmitting AP for target sensing can be concatenated into a unified precoding vector: (3) Taking advantage of the independence of communication signals and sensing signals, the access point The average transmitted signal can be expressed as: (4) Obviously, each AP has a transmission power limit, that is... .

[0027] Communication channel model: Based on the above system, communication performance is measured by the multi-user weighted sum rate. Consider a wide-area environment where the channel remains constant within each coherent block. Let... Indicates from the access point To user equipment direct channel, Indicates the process from RIS to user equipment The channel, Indicates from the access point The channel to RIS, then the total communication channel It can be represented as: (5) All of the network One transmitting antenna to user equipment The communication channel is defined as The channel can be represented as: (6) Since APs are widely distributed within the region, it is assumed that all of the above communication channels are Ricean fading channels.

[0028] Based on communication channel modeling, user equipment The signal received at point can be represented as: (7) in, For the required signal; Interference from other user equipment; Interference from sensors; user equipment The receiver noise at the location is caused by This indicates that the user equipment The downlink SINR of the communication can be expressed as: (8) in, The desired useful signal power received by user equipment i; The total received power generated at user equipment i by the communication signals of all user equipments; To sense the interference power generated by the signal at user equipment i; This represents the Gaussian white noise power of the user receiver.

[0029] Using Shannon's formula, the achievable spectral efficiency of this user equipment can be expressed as: (9) Sensing channel modeling: Sensing performance is constrained by ensuring the lowest possible signal-to-interference-plus-noise ratio (SNR). Assuming the AP antennas are arranged with a half-wavelength spacing, the corresponding azimuth angle... and elevation angle The antenna array response vector can be expressed as : (10) in, Spatial frequency factor, For phase coefficients, This represents the phase difference between adjacent antenna elements.

[0030] Therefore, the transmitter access point can be defined. -RIS-Target-RIS-Receiver Access Point The reflective perception path, denoted as Its expression is: (11) in, Indicates access point from the transmitter To the receiving end access point The radar cross section (RCS) corresponding to the reflection path is assumed to be a constant. and These represent the azimuth and elevation angles from the target position to the RIS position, respectively. and These represent the azimuth and elevation angles from the RIS position to the target position, respectively. and These represent the RIS and the receiver access point, respectively. Between, and between the RIS and the transmitter access point Beamforming matrix of the sensing paths between; Let be the overall channel gain, and define it as: (12) in, Indicates the carrier wavelength. Indicates the transmitter access point Distance from the target location Indicates the receiving end access point Distance to the target location.

[0031] Based on sensor channel modeling, the receiver access point The sensed signal received at the point can be represented in the following form: (13) in, For the target reflected signal, To prevent communication interference in the sensor reflection path, It is Gaussian noise.

[0032] Furthermore, the system contains a total of There are 10 sensing access points, therefore the total received sensing signal can be expressed as: (14) The overall sensor signal-to-interference-plus-noise ratio can be expressed as: (15) in, The number of APs that can be launched; The number of receiving APs; The power scaling factor assigned to the sensed signal; For the first A precoded vector is used by the transmitting AP for sensing, which is used to focus the beam toward the target direction; For from the first The first transmission AP to the first The sensing channel matrix of each receiving AP; receiving access point The noise of each antenna at that location is caused by express.

[0033] The optimization problem is defined as follows: Based on the above communication objectives and sensing constraints, and considering the power limitations of the access point (AP), the core optimization problem to be solved by this invention is as follows: (16) in, These are the weighting coefficients; This indicates that each AP has a transmit power limit; This indicates that the sensing capability of the ISAC system needs to be guaranteed. This represents the lower limit of the signal-to-interference-plus-noise ratio (SIR).

[0034] Optimization algorithm design: To solve the aforementioned non-convex optimization problem, this invention employs an alternating optimization algorithm based on fractional programming (FP). First, fractional programming is used to process the logarithmic and fractional terms in the objective function, transforming them into an easier-to-process form. Then, through the alternating optimization framework, the coupled joint optimization problem is decoupled into two sub-problems concerning perceptual precoding and communication precoding, and solved iteratively.

[0035] Specifically, the algorithm starts from the original objective function of the optimization problem: The original objective function is: (17) Applying the FP algorithm to the above equation, where: (18) (19) (20) Therefore, It is considered as the new, smoothed objective function.

[0036] For constraint (16.a), an AP selection matrix is ​​introduced for description: (twenty one) Therefore, (16.a) can be rewritten as: (twenty two) The constraint (16.b) can be rewritten as: (twenty three) in: (twenty four) Therefore, the new subproblem is: (25) Further decoupling into two sub-problems: (26) (27) Among them, matrix It consists of a global sensing channel. Constraint (26.b) is a non-convex quadratic constraint, obtained through the previous iteration point. A first-order Taylor expansion is performed at the point to obtain its conservative linear lower bound approximation, thereby transforming the perceptual precoding subproblem (26) into a convex problem, which can be solved using standard convex optimization tools such as the interior point method (e.g., CVX).

[0037] Problem (27) still cannot be solved directly using CVX. The objective function of problem (27) is further transformed into: (28) in, and The known matrix consists of the channel and auxiliary variables: in This represents the Kronecker product of matrices.

[0038] The constraint (27.a) can be rewritten as: (29) The constraint (27.b) can be rewritten as: (30) Therefore, problem (27) can be rewritten as (31) Problem (31) can be solved using the CVX toolbox. The Central Processing Unit (CPU) will... , Distributed to each transmission access point.

[0039] The overall algorithm for optimizing the precoding vector is shown in Table 1 below: Table 1 This invention provides a systematic method for effectively co-optimizing precoding to balance communication and sensing performance in a RIS-assisted non-cellular large-scale MIMO-ISAC system. This method transforms a complex non-convex problem into a series of efficiently solvable convex subproblems through fractional programming and alternating optimization, ensuring the feasibility and convergence of the algorithm and providing key technical support for the practical deployment of integrated sensing and communication networks.

[0040] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A precoding optimization method in a RIS-assisted non-cellular massive MIMO and ISAC system, the system comprising multiple distributed access points controlled by a central processing unit, a smart surface, at least one user equipment, and at least one target to be sensed, characterized in that, The method includes: With the goal of maximizing the multi-user weighted communication rate, and constrained by the transmit power of each access point and the requirement that the perceived signal-to-interference-plus-noise ratio is not lower than a preset threshold, a joint optimization problem for the precoding vector is established. The objective function of the joint optimization problem is transformed by fractional programming (FP) algorithm to obtain a smoothed objective function; Based on the smoothed objective function and the constraints, an alternating optimization framework is adopted to decouple the transformed joint optimization problem into a communication precoding subproblem and a sensing precoding subproblem, and solve them through an alternating iterative method. In each iteration, the sensing precoding vector is fixed to solve the communication precoding subproblem, and the communication precoding vector is fixed to solve the sensing precoding subproblem. Update and output the optimized communication precoding vector and perception precoding vector, and configure them to each access point.

2. The precoding optimization method for a RIS-assisted cellular-free massive MIMO and ISAC system according to claim 1, characterized in that, The optimization objective is: ， in, Let be the weight coefficient of the i-th user device. Let be the signal-to-interference-plus-noise ratio (SIR) of the i-th user equipment. Let be the precoding vector for the transmission access point to user equipment i. This is the pre-coded vector for the transmission access point to sense the target.

3. The precoding optimization method for a RIS-assisted non-cellular massive MIMO and ISAC system according to claim 1, characterized in that, The transmit power constraint for the access point is as follows: for the k-th transmit access point, its total transmit power does not exceed a preset maximum transmit power threshold. .

4. The precoding optimization method for a RIS-assisted non-cellular massive MIMO and ISAC system according to claim 1, characterized in that, The objective function of the joint optimization problem is transformed using the fractional programming (FP) algorithm, specifically including: Introducing the first auxiliary variable Second auxiliary variable The optimization objective is transformed into a smoothed objective function. : ， in, Let i be the weighting coefficient for user device i. Let be the equivalent communication channel vector for the i-th user. The total number of user devices. Let be the precoding vector for the transmission access point to user equipment i. This is the precoded vector for the transmission access point to sense the target. and The power allocation factor, This represents noise power.

5. The precoding optimization method for a RIS-assisted non-cellular massive MIMO and ISAC system according to claim 4, characterized in that, In each iteration, the first auxiliary variable Second auxiliary variable Update based on the precoding vector of the current iteration, specifically as follows: ， ， in, Let be the weight coefficient of the i-th user device. Let be the signal-to-interference-plus-noise ratio (SIR) of the i-th user equipment. Let be the precoding vector for the transmission access point to user equipment i. This is the precoded vector for the transmission access point to sense the target. Let be the equivalent communication channel vector for the i-th user. and The power allocation factor, This represents noise power.

6. The precoding optimization method for a RIS-assisted non-cellular massive MIMO and ISAC system according to claim 4, characterized in that, Solving the aforementioned perceptual precoding subproblem specifically includes: Under the condition of fixed communication precoding vector, based on the smoothed objective function We construct a sensing precoding subproblem with the objective of minimizing the interference term of the sensing signal received by each user equipment, while satisfying the transmit power constraint and the sensing signal-to-interference-plus-noise ratio constraint. The non-convex constraints in the perceptual precoding subproblem are transformed into convex constraints using a first-order Taylor expansion. A convex optimization solver is used to solve the transformed perceptual precoding subproblem, resulting in an updated perceptual precoding vector.

7. The precoding optimization method for a RIS-assisted non-cellular massive MIMO and ISAC system according to claim 4, characterized in that, Solving the aforementioned communication precoding subproblem specifically includes: Under the condition of fixed sensing precoding vectors, the communication precoding vectors of all user equipment are combined into a communication precoding matrix; Based on the smoothed objective function We construct a communication precoding subproblem with the objective of maximizing the difference between the linear and quadratic terms, while satisfying transmit power constraints and sensing signal-to-interference-plus-noise ratio constraints. The communication precoding subproblem is solved using a convex optimization solver to obtain an updated communication precoding matrix, from which the communication precoding vectors of each user equipment are recovered.

8. The precoding optimization method for a RIS-assisted non-cellular massive MIMO and ISAC system according to claim 7, characterized in that, The objective function of the second sub-optimization problem is in the form of: ,in, For communication precoding matrix, and This is a matrix related to the channel and auxiliary variables.

9. The precoding optimization method for a RIS-assisted non-cellular massive MIMO and ISAC system according to claim 1, characterized in that, The iteration termination condition for the alternating optimization is: the difference between the objective function values ​​obtained from two consecutive iterations is less than a preset tolerance, or the number of iterations reaches the preset maximum number of iterations.

10. A precoding optimization method for a RIS-assisted cellular-free massive MIMO and ISAC system according to claim 1, characterized in that, Before starting the alternating iterative optimization, the method further includes initializing the sensing precoding vector and the communication precoding vector; wherein, the sensing precoding vector is initialized as a beamforming vector pointing in the direction of the target to be sensed, and the communication precoding vector is initialized as a maximum ratio transmission precoding vector based on the channel state information of each user.

Images