Intelligent reflection surface auxiliary backscattering-based common inductance computing resource allocation method and system

Abstract

The invention provides a method and a system for allocating intelligent reflecting surface auxiliary backscattering-based general sensing calculation resources. The method specifically comprises the following steps of: constructing a system model of an intelligent reflecting surface auxiliary backscattering-based general sensing calculation network; constructing an optimization problem of a user resource allocation strategy, a base station transmitting beam and an intelligent reflecting surface reflection coefficient; converting an original problem into a plurality of sub-optimization problems by using a fractional programming and block coordinate descent method; a user resource allocation strategy problem is converted into a linear programming problem and is solved based on CVX; solving the optimization problem of the reflection coefficient of the intelligent reflecting surface through a continuous convex approximation method based on a penalty function; and solving the optimization problem of the base station transmitting beam by combining maximization, minimization and positive semidefinite relaxation algorithms. According to the invention, the intelligent reflecting surface backscattering technology is adopted, the calculation efficiency of the system is improved, and the balance between the calculation efficiency and the sensing performance can be kept.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a method and system for allocating synaesthesia resources for intelligent reflective surface-assisted backscattering. Background Art

[0002] Smart reflective surface backscattering is a method of achieving low-power transmission and improving system confidentiality. The principle is to control the reflection coefficient of the circuit by adjusting its internal impedance, thereby changing the amplitude, frequency, phase, etc. of the RF signal from other devices or environments to achieve signal modulation and transmission.

[0003] Backscattering technology adjusts the phase of the reflective surface unit to achieve signal transmission through environmental reflection without an active signal source. In intelligent reflective surface backscattering technology, communication between devices does not rely on traditional wireless transmitters and receivers. The key is to use the intelligent control ability of the reflective surface to reflect or scatter the received signal back to the receiving device to achieve communication. The advantages of this technology include: backscattering technology usually does not require the high power consumption of traditional transmitting and receiving modules, so it can significantly reduce the energy consumption of the system; without the need for complex RF front-end design and signal processing, backscattering technology can transmit information through reflected and scattered signals in the environment.

[0004] As an emerging wireless communication technology, smart reflective surface backscattering technology can improve the coverage, energy efficiency, and data transmission performance of wireless communication systems by intelligently controlling the reflection of signals in the environment with its low power consumption, low complexity, and high adaptability to the environment. It is gradually becoming an important research direction in the fields of 5G, 6G, and the Internet of Things. However, there is still a throughput limit problem caused by round-trip path loss, the computing efficiency of the system needs to be improved, and the performance balance between computing communication and perception is difficult to ensure. Summary of the invention

[0005] The object of the present invention is to provide a method and system for allocating synaesthesia resources for intelligent reflector-assisted backscattering, which can effectively reduce system cost, transmit information in a low-power manner, and improve calculation efficiency while estimating radar target parameters.

[0006] The technical solution to achieve the purpose of the present invention is: a method for allocating synaesthesia resources for intelligent reflective surface assisted backscattering, comprising the following steps:

[0007] Step 1, constructing a system model of a synaesthesia computing network for intelligent reflective surface assisted backscattering;

[0008] Step 2: construct the optimization problem of user resource allocation strategy, base station transmission beam and smart reflective surface reflection coefficient;

[0009] Step 3: Use fractional programming and block coordinate descent methods to transform the original problem into multiple sub-optimization problems;

[0010] Step 4: Convert the user resource allocation strategy problem into a linear programming problem and solve it based on CVX;

[0011] Step 5, solving the optimization problem of the reflection coefficient of the smart reflection surface by a continuous convex approximation method based on a penalty function;

[0012] Step 6: Solve the optimization problem of the base station transmit beam by combining the maximization-minimization and semi-definite relaxation algorithms.

[0013] Furthermore, in step 1, a system model of the synaesthesia computing network for intelligent reflective surface-assisted backscattering is constructed, as follows:

[0014] Step 1.1: In the synaesthesia network system with intelligent reflective surface assisted backscattering, the number of antennas of K users and the number of antennas of the base station are 1 and M respectively, and the intelligent reflective surface has N passive reflective elements; the M antennas of the base station include M t Signal transmitting antennas, M c Communication signal receiving antenna and M r Radar signal receiving antenna;

[0015] The base station can transmit information and radar signals at the same time to realize communication and radar sensing functions. The transmitted signal x is expressed as:

[0016] x=W c s c +W r s r =Ws,(1)

[0017] in and Respectively represent the transmission signal matrix of communication and radar, w c,K and w r,Mt Represent the transmit beam vectors of communication and radar respectively; the covariance matrix R of the transmit signal is expressed as:

[0018]

[0019] in Expressed as the covariance matrix of the communication transmission beam, and(·) H They represent the expectation and conjugate transpose of the matrix respectively; and They are respectively represented as the channels from the base station transmitting antenna to the k-th smart reflecting surface, from the k-th smart reflecting surface to the base station communication signal receiving antenna and to the radar signal receiving antenna and to the user;

[0020] The steering vectors from the base station transmitting antenna to the lth radar target, the lth radar target to the communication signal receiving antenna and the radar signal receiving antenna are expressed as:

[0021]

[0022]

[0023] in It is expressed as the target angle of the lth radar target, λ and d are the wavelength and distance of two adjacent transmitting / receiving antennas respectively;

[0024] When the transmitted signal passes through the smart transmitting surface, the signal will be modulated and carry the user's data information. The k-th signal modulation method on the smart reflecting surface is expressed as:

[0025]

[0026] in, represents the kth modulated signal, Θ k is the reflection coefficient matrix of the kth smart reflector and is expressed as Where diag(·) represents the diagonal matrix of the vector, so the received signal y of user k is k It is expressed as:

[0027]

[0028] in Represented as zero mean and variance Sorta distribution User k receives noise, σ k It is represented as the noise variance received by user k; the received signal to interference and noise ratio SINR of user k k It is expressed as:

[0029]

[0030] where ||·|| represents the second norm, It is represented as the effective channel from user k to the base station, and the signal y received by the base station communication signal receiving antenna c It is expressed as:

[0031]

[0032] in Expressed as the complex reflection coefficient of target l, n c represents the noise received by the base station, The SINR at the base station is expressed as:

[0033]

[0034] where |·| represents the norm, and the radar echo signal Y received by the base station r It is expressed as:

[0035]

[0036] in Denoted as zero mean and variance σ 2 Sorta distribution The noise at the base station, Represented as a matrix set of steering vectors for the radar signal receiving antenna, Represented as a set of steering vector matrices for the communication signal receiving antenna, In order to obtain an unbiased estimate of the target's arrival angle parameters, the radar target angle The Cramer-Rao bound is the optimal parameter estimate, and the estimation accuracy is It is expressed as:

[0037]

[0038] Tr(·) and Re(·) represent the trace and real part of the matrix, respectively. represents the partial derivative of a function, (·) -1 represents the inverse of a matrix; And they are respectively expressed as:

[0039]

[0040] in(·) T Indicates transposition;

[0041] In order to achieve accurate estimation of the arrival angle of all targets, The Cramer-Rao bound of should be less than an upper bound, namely

[0042]

[0043] where γ l Expressed as the upper bound of CRB for radar target l, where Represents a collection of radar targets;

[0044] In addition, the rate at which users compute at the edge is represented by R c,k =Blog2(1+SINR c,k ), where B is the bandwidth, so the number of bits transmitted by the user uplink is expressed as C EC,k =t EC R c,k , where t EC It is expressed as the communication time, and the corresponding edge computing energy consumption is expressed as E EC =tEC Nμ, μ represents the power consumption of each smart reflector; the number of bits calculated locally by user k is expressed as in Denotes the local computing time of user k, f k and c k They represent the frequency and cycle number of the user's local calculation respectively, and the energy consumption of the local calculation is expressed as ∈ k Represents the energy consumption coefficient; define ω k Expressed as the proportion allocated to edge computing, define η k It is expressed as the computational efficiency of user k, which is defined as the ratio of the number of bits calculated by user k to the computational energy consumption, expressed as:

[0045]

[0046] Furthermore, in step 2, the optimization problem of user resource allocation strategy, base station transmission beam and smart reflector reflection coefficient is constructed as follows:

[0047] In order to maximize the computational efficiency of each user, the transmit beam W at the base station and the reflection coefficient θ of the smart reflector are optimized by joint optimization. k , calculate the mode allocation ratio ω k And the time t of edge computing and local computing EC ,t LC,k , and constructed the following optimization problem:

[0048]

[0049] where Γ k represents the lower limit of the communication service quality of user k, and They represent the energy consumption thresholds of edge computing and local computing of user k respectively; constraint C1 ensures the user's communication quality requirements, C2 guarantees the radar perception performance of each target, C3 limits the transmission power at the base station, C4 is expressed as the reflection coefficient constraint at the smart reflection surface, C5 ensures that the proportion of computing tasks selected by the user is within a reasonable range, C6 represents the limit of the user's edge computing time and local computing time, and C7 is the energy constraint for each user.

[0050] Furthermore, in step 3, fractional programming and block coordinate descent methods are used to transform the original problem into multiple sub-optimization problems, as follows:

[0051] Due to the strong coupling between the fraction of the objective function and the optimization variable, the original optimization problem is a classic non-convex optimization problem. This fractional programming problem is first solved by implementing the Dinkelbach method, which is expressed as

[0052]

[0053] where ξ* represents the optimal value that satisfies the fractional programming update, (·) * Represents the optimal value of the variable. Since ξ* cannot be obtained in advance, an update parameter ξ is used to replace ξ * , therefore, problem (18) is equivalently transformed as follows

[0054]

[0055] In order to solve the coupled optimization variables, the block coordinate descent (BCD) technology is used to solve each sub-problem separately; the penalty continuous convex approximation (SCA) method is used to optimize the smart reflector reflection coefficient sub-problem, and the maximum-minimization combined with semi-positive definite programming algorithm is used to optimize the transmission beamforming matrix.

[0056] Furthermore, in step 4, the user resource allocation strategy problem is transformed into a linear programming problem and solved based on CVX, as follows:

[0057] When fixed ω k , W, θ k When there are variables, the optimization problem of time allocation can be expressed as:

[0058]

[0059] Where ζ is the auxiliary variable introduced, d k =ω k (R c,k -ξNμ), The optimization problem of time allocation is a linear programming problem, so it is solved by CVX tools; k The optimization problem is also a linear programming problem and is solved in the same way.

[0060] Furthermore, in step 5, the optimization problem of the reflection coefficient of the smart reflection surface is solved by a continuous convex approximation method based on a penalty function, as follows:

[0061] Fixing other variables, the optimization problem of the reflection coefficient of the smart reflector can be reformulated as:

[0062]

[0063] Among them, α k , k and β k is the auxiliary variable introduced. By transforming the variables, we get:

[0064]

[0065] in, and

[0066]

[0067] in Indicates redefining the variables. Based on the above transformation, the optimization problem is:

[0068]

[0069] where rank(·) represents the rank of the matrix. Then, a continuous convex approximation scheme based on a penalty function is used to optimize the problem of (27), and the objective function is expanded by a first-order Taylor expansion. The optimization problem is expressed as:

[0070]

[0071] Where δ is the penalty factor, the problem in equation (27) is transformed into the convex problem (28) and can be solved. The optimization variables are The reflection coefficient of the optimal smart reflective surface is restored.

[0072] Furthermore, in step 6, the optimization problem of the base station transmit beam is solved by combining the maximization-minimization and semidefinite relaxation algorithms, as follows:

[0073] By fixing other variables, the optimization problem of the base station transmit beam is expressed as

[0074]

[0075] The terms in brackets in the objective function can be restated as:

[0076]

[0077] in The original problem is expressed as:

[0078]

[0079] in The second non-convex constraint is a convex difference programming problem. In order to solve the convex difference programming problem, the convex difference programming problem is reformulated as a series of problems using the maximum minimization algorithm and optimized until convergence is achieved;

[0080] Transform using the first-order Taylor expansion:

[0081]

[0082] Therefore, problem (33) can be transformed into finding R kA new problem with feasible solutions to R

[0083]

[0084] By ignoring the rank-one constraint, the optimal communication beam is expressed as in According to Cholesky decomposition, the radar transmit beam is obtained

[0085] A system for allocating synaesthesia resources for intelligent reflective surface assisted backscattering, the system is used to implement the method for allocating synaesthesia resources for intelligent reflective surface assisted backscattering, the system comprises a first module to a sixth module, and the functions of each module are as follows:

[0086] The first module is to construct a system model of the synaesthesia computing network with intelligent reflective surface assisted backscattering;

[0087] The second module constructs the optimization problem of user resource allocation strategy, base station transmission beam and smart reflector reflection coefficient;

[0088] In the third module, fractional programming and block coordinate descent methods are used to transform the original problem into multiple sub-optimization problems;

[0089] The fourth module transforms the user resource allocation strategy problem into a linear programming problem and solves it based on CVX;

[0090] The fifth module solves the optimization problem of the reflection coefficient of the smart reflective surface through a continuous convex approximation method based on a penalty function;

[0091] The sixth module solves the optimization problem of base station transmit beam by combining maximization-minimization and semi-definite relaxation algorithms.

[0092] A mobile terminal comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the program, the method for allocating synaesthesia resources for intelligent reflective surface-assisted backscattering is implemented.

[0093] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps in the method for allocating synaesthesia resources for intelligent reflective surface-assisted backscattering.

[0094] Compared with the prior art, the present invention has the following significant advantages: (1) The present invention utilizes the intelligent reflective surface backscattering technology to solve the throughput upper limit problem caused by round-trip path loss in a low-power manner; (2) The proposed continuous convex approximation method based on penalty function and maximum minimization method can significantly improve the computational efficiency of the system and ensure the performance balance between computing, communication and perception. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0096] Figure 1 The present invention is a flow chart of the synaesthesia resource allocation method for intelligent reflective surface assisted backscattering.

[0097] Figure 2 This is a model diagram of the synaesthesia computing system with intelligent reflective surface assisted backscattering according to an embodiment of the present invention.

[0098] Figure 3 This is a curve diagram of the calculation efficiency of an embodiment of the present invention as the Cramer-Rao bound threshold changes.

[0099] Figure 4 This is a graph showing the relationship between the calculation efficiency and the number of smart reflective surface units according to an embodiment of the present invention.

[0100] Figure 5 FIG. 4 is a diagram showing the relationship between the beam pattern and the angle according to an embodiment of the present invention. DETAILED DESCRIPTION

[0101] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0102] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0103] This embodiment provides a method for allocating synaesthesia resources using intelligent reflector-assisted backscattering, wherein intelligent reflector-assisted backscattering is used to assist in uplink communication modulation and retransmission of original signals. Specifically, the base station communicates with the user at the same time and senses the target based on the echo signal. The user uses intelligent reflector-assisted backscattering technology to passively offload the re-tuned data to the base station. Figure 1 shown.

[0104] Combination Figure 1 The present invention provides a method for allocating synaesthesia resources for intelligent reflective surface-assisted backscattering, comprising the following steps:

[0105] Step 1, constructing a system model of a synaesthesia computing network for intelligent reflective surface assisted backscattering;

[0106] Step 2: construct the optimization problem of user resource allocation strategy, base station transmission beam and smart reflective surface reflection coefficient;

[0107] Step 3: Use fractional programming and block coordinate descent methods to transform the original problem into multiple sub-optimization problems;

[0108] Step 4: Convert the user resource allocation strategy problem into a linear programming problem and solve it based on CVX;

[0109] Step 5, solving the optimization problem of the reflection coefficient of the smart reflection surface by a continuous convex approximation method based on a penalty function;

[0110] Step 6: Solve the optimization problem of the base station transmit beam by combining the maximization-minimization and semi-definite relaxation algorithms.

[0111] As a specific example, in step 1, a system model of a synaesthesia computing network with intelligent reflective surface assisted backscattering is constructed, combined with Figure 2 , as follows:

[0112] Step 1.1: In the synaesthesia network system with intelligent reflective surface assisted backscattering, the number of antennas of K users and the number of antennas of the base station are 1 and M respectively, and the intelligent reflective surface has N passive reflective elements; the M antennas of the base station include M t Signal transmitting antennas, M c Communication signal receiving antenna and M r Radar signal receiving antenna;

[0113] The base station can transmit information and radar signals at the same time to realize communication and radar sensing functions. The transmitted signal x is expressed as:

[0114] x=W c s c +W r s r =Ws,(1.1)

[0115] in and Represent the transmission signal matrices of communication and radar respectively, w c,K and Represent the transmit beam vectors of communication and radar respectively. The covariance matrix R of the transmit signal is expressed as:

[0116]

[0117] in Expressed as the covariance matrix of the communication transmission beam, and(·) H They represent the expectation and conjugate transpose of the matrix respectively. and They are respectively represented as the channels from the base station transmitting antenna to the kth smart reflecting surface, from the kth smart reflecting surface to the base station communication signal receiving antenna and to the radar signal receiving antenna and to the user.

[0118] The steering vectors from the base station transmitting antenna to the lth radar target, the lth radar target to the communication signal receiving antenna and the radar signal receiving antenna are expressed as:

[0119]

[0120] in It is represented as the target angle of the lth radar target, λ and d are the wavelength and distance of two adjacent transmitting / receiving antennas respectively. When the transmitted signal passes through the smart transmitting surface, the signal will be modulated and carry the user's data information. The modulation method of the kth signal on the smart reflecting surface is expressed as:

[0121]

[0122] in, represents the kth modulated signal, Θ k is the reflection coefficient matrix of the kth smart reflector and is expressed as Where diag(·) represents the diagonal matrix of the vector, so the received signal y of user k is k It is expressed as:

[0123]

[0124] in Represented as zero mean and variance Sorta distribution User k receives noise, σ k The received signal to interference and noise ratio SINR of user k is expressed as the noise variance received by user k. k It is expressed as:

[0125]

[0126] where ||·|| represents the second norm, It is represented as the effective channel from user k to the base station, and the signal y received by the base station communication signal receiving antenna c It is expressed as:

[0127]

[0128] in Expressed as the complex reflection coefficient of target l, n c represents the noise received by the base station, The SINR at the base station is expressed as:

[0129]

[0130] where |·| represents the norm, and the radar echo signal Y received by the base station r It is expressed as:

[0131]

[0132] in Denoted as zero mean and variance σ 2 Follow the normal distribution The noise at the base station, Represented as a matrix set of steering vectors for the radar signal receiving antenna, Represented as a set of steering vector matrices for the communication signal receiving antenna, In order to obtain an unbiased estimate of the target's arrival angle parameters, the radar target angle The Cramer-Rao bound is the optimal parameter estimate, and the estimation accuracy is It is expressed as:

[0133]

[0134] Tr(·) and Re(·) represent the trace and real part of the matrix, respectively. represents the partial derivative of a function, (·) -1 It means to find the inverse of the matrix. And they are respectively expressed as:

[0135]

[0136] in(·) T In order to accurately estimate the arrival angle of all targets, The Cramer-Rao bound of should be less than an upper bound, namely

[0137]

[0138] where γ l Expressed as the upper bound of CRB for radar target l, where Represents a collection of radar targets;

[0139] In addition, the rate at which users compute at the edge is represented by R c,k =Blog2(1+SINR c,k ), where B is the bandwidth, so the number of bits transmitted by the user uplink is expressed as C EC,k =t EC R c,k , where t EC It is expressed as the communication time, and the corresponding edge computing energy consumption is expressed as E EC =t ECNμ, μ represents the power consumption of each smart reflector; the number of bits calculated locally by user k is expressed as in Denotes the local computing time of user k, f k and c k They represent the frequency and cycle number of the user's local calculation respectively, and the energy consumption of the local calculation is expressed as ∈ k Represents the energy consumption coefficient. Definition ω k Expressed as the proportion allocated to edge computing, define η k It is expressed as the computational efficiency of user k, which is defined as the ratio of the number of bits calculated by user k to the computational energy consumption, expressed as:

[0140]

[0141] As a specific example, in step 2, the optimization problem of user resource allocation strategy, base station transmission beam and smart reflective surface reflection coefficient is constructed as follows:

[0142] In order to maximize the computational efficiency of each user, the transmit beam W at the base station and the reflection coefficient θ of the smart reflector are optimized by joint optimization. k , calculate the mode allocation ratio ω k And the time t of edge computing and local computing EC ,t LC,k , and constructed the following optimization problem:

[0143]

[0144] where Γ k represents the lower limit of the communication service quality of user k, and They represent the energy consumption thresholds of edge computing and local computing of user k respectively. Constraint C1 ensures the user's communication quality requirements, C2 guarantees the radar perception performance of each target, C3 limits the transmission power at the base station, C4 is expressed as the reflection coefficient constraint at the smart reflection surface, C5 ensures that the proportion of user-selected computing tasks is within a reasonable range, C6 represents the limit of the user's edge computing time and local computing time, and C7 is the energy constraint for each user.

[0145] As a specific example, in step 3, the original problem is transformed into multiple sub-optimization problems using fractional programming and block coordinate descent methods, as follows:

[0146] Due to the strong coupling between the fraction of the objective function and the optimization variable, the original optimization problem is a classic non-convex optimization problem. This fractional programming problem is first solved by implementing the Dinkelbach method, which is expressed as

[0147]

[0148] where ξ* represents the optimal value that satisfies the fractional programming update, (·) * Represents the optimal value of the variable. Since ξ* cannot be obtained in advance, an update parameter ξ is used to replace ξ * , therefore, problem (1.18) is equivalently transformed as follows

[0149]

[0150] In order to solve the coupled optimization variables, the block coordinate descent (BCD) technology is used to solve each sub-problem separately. In addition, the penalty continuous convex approximation (SCA) method is used to optimize the smart reflector reflection coefficient sub-problem. At the same time, the maximum-minimization combined with the semi-positive definite programming algorithm is used to optimize the transmission beamforming matrix.

[0151] As a specific example, in step 4, the user resource allocation strategy problem is converted into a linear programming problem and solved based on CVX, as follows:

[0152] When fixed ω k , W, θ k When there are variables, the optimization problem of time allocation can be expressed as:

[0153]

[0154] Where ζ is the auxiliary variable introduced, d k =ω k (R c,k -ξNμ), The optimization problem of time allocation is a linear programming problem, so it is solved by CVX tools; similarly, the optimization problem of ω k The optimization problem is also a linear programming problem and is solved in the same way.

[0155] As a specific example, in step 5, the optimization problem of the reflection coefficient of the smart reflection surface is solved by a continuous convex approximation method based on a penalty function, as follows:

[0156] Fixing other variables, the optimization problem of the reflection coefficient of the smart reflector can be reformulated as:

[0157]

[0158] Among them, α k , k and β k is the auxiliary variable introduced. By transforming the variables, we get:

[0159]

[0160]

[0161] in, and

[0162]

[0163] in Indicates redefining the variables. Based on the above transformation, the optimization problem is:

[0164]

[0165] Where rank(·) represents the rank of the matrix. Then, a continuous convex approximation scheme based on a penalty function is used to optimize the problem of (1.27), and the objective function is expanded by a first-order Taylor expansion. The optimization problem is expressed as:

[0166]

[0167] Where δ is the penalty factor, the problem of formula (1.27) is transformed into a convex problem (1.28) and can be solved. The optimization variable is The reflection coefficient of the optimal smart reflective surface is restored.

[0168] As a specific example, in step 6, the optimization problem of the base station transmit beam is solved by combining the maximization-minimization and semidefinite relaxation algorithms, as follows:

[0169] By fixing other variables, the optimization problem of the base station transmit beam is expressed as

[0170]

[0171] The terms in brackets in the objective function can be restated as:

[0172]

[0173] in The original problem is expressed as:

[0174]

[0175] in The second non-convex constraint is a convex difference programming problem. In order to solve the convex difference programming problem, the convex difference programming problem is reformulated as a series of problems using the maximum minimization algorithm and optimized until convergence is achieved. Specifically, the first-order Taylor expansion is used to transform

[0176]

[0177] Therefore, problem (1.33) can be transformed into finding R k A new problem with feasible solutions to R

[0178]

[0179] By ignoring the rank-one constraint, the optimal communication beam is expressed as in According to Cholesky decomposition, the radar transmit beam is obtained

[0180] Figure 3 The relationship between computation and sensing performance is described, and as the Cramer-Rao bound increases, the constraints on sensing capabilities are relaxed, allowing more power to be allocated to users. In addition, increasing the number of smart reflector units increases the degree of freedom, thereby improving computational efficiency while maintaining the same Cramer-Rao bound.

[0181] Figure 4 The graph shows how the computational efficiency performance varies with the number of reflective elements of the smart reflector. The computational efficiency of the proposed method improves with the increase in the number of reflective elements of the smart reflector, which is attributed to the enhanced signal power reception and the increase in spatial degrees of freedom. In addition, the performance of the random phase shift smart reflector method is the worst, and its computational efficiency is significantly lower than that of the smart reflector-assisted backscattering algorithm, which shows that smart reflector-assisted backscattering can effectively enhance computational efficiency.

[0182] Figure 5 The beam patterns for different Cramer-Rao thresholds are shown. As shown, the lower the Cramer-Rao threshold, the more focused the beam is in the direction of the radar target. Conversely, as the Cramer-Rao threshold increases, the sensing beam weakens and the power in the direction of the user increases, indicating enhanced communication performance at the expense of reduced sensing capability. In addition, the power allocated to the interfering signal is kept low, which helps reduce communication interference.

[0183] The present invention also provides a system for allocating synaesthesia resources for intelligent reflective surface assisted backscattering, which is used to implement the method for allocating synaesthesia resources for intelligent reflective surface assisted backscattering. The system includes a first module to a sixth module, and the functions of each module are as follows:

[0184] The first module is to construct a system model of the synaesthesia computing network with intelligent reflective surface assisted backscattering;

[0185] The second module constructs the optimization problem of user resource allocation strategy, base station transmission beam and smart reflector reflection coefficient;

[0186] In the third module, fractional programming and block coordinate descent methods are used to transform the original problem into multiple sub-optimization problems;

[0187] The fourth module transforms the user resource allocation strategy problem into a linear programming problem and solves it based on CVX;

[0188] The fifth module solves the optimization problem of the reflection coefficient of the smart reflective surface through a continuous convex approximation method based on a penalty function;

[0189] The sixth module solves the optimization problem of base station transmit beam by combining maximization-minimization and semi-definite relaxation algorithms.

[0190] The present invention also provides a mobile terminal, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the synaesthesia resource allocation method for intelligent reflective surface assisted backscattering when executing the program.

[0191] The present invention also provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps in the synaesthesia resource allocation method for intelligent reflective surface-assisted backscattering are implemented.

[0192] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

[0193] It should be understood that in order to simplify the present invention and help those skilled in the art understand the various aspects of the present invention, in the above description of the exemplary embodiments of the present invention, the various features of the present invention are sometimes described in a single embodiment or described with reference to a single figure. However, the present invention should not be interpreted as the features included in the exemplary embodiments are all necessary technical features of the patent claims.

Claims

1. A method for allocating synaesthesia resources for intelligent reflective surface-assisted backscattering, characterized in that: The following steps are involved: Step 1, constructing a system model of a synaesthesia computing network for intelligent reflective surface assisted backscattering; Step 2: construct the optimization problem of user resource allocation strategy, base station transmission beam and smart reflective surface reflection coefficient; Step 3: Use fractional programming and block coordinate descent methods to transform the original problem into multiple sub-optimization problems; Step 4: Convert the user resource allocation strategy problem into a linear programming problem and solve it based on CVX; Step 5, solving the optimization problem of the reflection coefficient of the smart reflection surface by a continuous convex approximation method based on a penalty function; Step 6: Solve the optimization problem of the base station transmit beam by combining the maximization-minimization and semi-definite relaxation algorithms.

2. The method for allocating synaesthesia resources for intelligent reflective surface-assisted backscattering according to claim 1 is characterized in that: In step 1, the system model of the synaesthesia computing network with intelligent reflective surface assisted backscattering is constructed as follows: Step 1.1: In the synaesthesia network system with intelligent reflective surface assisted backscattering, the number of antennas of K users and the number of antennas of the base station are 1 and M respectively, and the intelligent reflective surface has N passive reflective elements; the M antennas of the base station include M t Signal transmitting antennas, M c Communication signal receiving antenna and M r Radar signal receiving antenna; The base station can transmit information and radar signals at the same time to realize communication and radar sensing functions. The transmitted signal x is expressed as: x=W c s c +W r s r =Ws (1) in and Respectively represent the transmission signal matrix of communication and radar, w c,K and Represent the transmit beam vectors of communication and radar respectively; the covariance matrix R of the transmit signal is expressed as: in Expressed as the covariance matrix of the communication transmission beam, and(·) H They represent the expectation and conjugate transpose of the matrix respectively; and They are respectively represented as the channels from the base station transmitting antenna to the k-th smart reflecting surface, from the k-th smart reflecting surface to the base station communication signal receiving antenna and to the radar signal receiving antenna and to the user; The steering vectors from the base station transmitting antenna to the lth radar target, the lth radar target to the communication signal receiving antenna and the radar signal receiving antenna are expressed as: in It is expressed as the target angle of the lth radar target, λ and d are the wavelength and distance of two adjacent transmitting / receiving antennas respectively; When the transmitted signal passes through the smart transmitting surface, the signal will be modulated and carry the user's data information. The k-th signal modulation method on the smart reflecting surface is expressed as: in, represents the kth modulated signal, Θ k is the reflection coefficient matrix of the kth smart reflector and is expressed as Where diag(·) represents the diagonal matrix of the vector, so the received signal y of user k is k It is expressed as: in Represented as zero mean and variance Follow the normal distribution User k receives noise, σ k It is represented as the noise variance received by user k; the received signal to interference and noise ratio SINR of user k k It is expressed as: where ||·|| represents the second norm, It is represented as the effective channel from user k to the base station, and the signal y received by the base station communication signal receiving antenna c It is expressed as: where θ l Expressed as the complex reflection coefficient of target l, n c represents the noise received by the base station, The SINR at the base station is expressed as: where |·| represents the norm, and the radar echo signal Y received by the base station r It is expressed as: in Denoted as zero mean and variance σ 2 Sorta distribution The noise at the base station, Represented as a matrix set of steering vectors for the radar signal receiving antenna, Represented as a set of steering vector matrices for the communication signal receiving antenna, In order to obtain an unbiased estimate of the target's arrival angle parameters, the radar target angle The Cramer-Rao bound is the optimal parameter estimate, and the estimation accuracy is It is expressed as: Tr(·) and Re(·) represent the trace and real part of the matrix, respectively. represents the partial derivative of a function, (·) -1 represents the inverse of a matrix; And they are respectively expressed as: in(·) T Indicates transposition; In order to achieve accurate estimation of the arrival angle of all targets, The Cramer-Rao bound of should be less than an upper bound, namely where γ l Expressed as the upper bound of CRB for radar target l, where Represents a collection of radar targets; In addition, the rate at which users compute at the edge is represented by R c,k =Blog2(1+SINR c,k ), where B is the bandwidth, so the number of bits transmitted by the user uplink is expressed as C EC,k =t EC R c,k , where t EC It is expressed as the communication time, and the corresponding edge computing energy consumption is expressed as E EC =t EC Nμ, μ represents the power consumption of each smart reflector; the number of bits calculated locally by user k is expressed as in Denotes the local computing time of user k, f k and c k They represent the frequency and cycle number of the user's local calculation respectively, and the energy consumption of the local calculation is expressed as ∈ k Represents the energy consumption coefficient; define ω k Expressed as the proportion allocated to edge computing, define η k It is expressed as the computational efficiency of user k, which is defined as the ratio of the number of bits calculated by user k to the computational energy consumption, expressed as:

3. The method for allocating synaesthesia resources for intelligent reflective surface-assisted backscattering according to claim 2 is characterized in that: In step 2, the optimization problem of user resource allocation strategy, base station transmission beam and smart reflector reflection coefficient is constructed as follows: In order to maximize the computational efficiency of each user, the transmit beam W at the base station and the reflection coefficient θ of the smart reflector are optimized by joint optimization. k , calculate the mode allocation ratio ω k And the time t of edge computing and local computing EC ,t LC,k , and constructed the following optimization problem: where Γ k represents the lower limit of the communication service quality of user k, and They represent the energy consumption thresholds of edge computing and local computing of user k respectively; constraint C1 ensures the user's communication quality requirements, C2 guarantees the radar perception performance of each target, C3 limits the transmission power at the base station, C4 is expressed as the reflection coefficient constraint at the smart reflection surface, C5 ensures that the proportion of computing tasks selected by the user is within a reasonable range, C6 represents the limit of the user's edge computing time and local computing time, and C7 is the energy constraint for each user.

4. The method for allocating synaesthesia resources for intelligent reflective surface-assisted backscattering according to claim 3 is characterized in that: In step 3, fractional programming and block coordinate descent are used to transform the original problem into multiple sub-optimization problems, as follows: Due to the strong coupling between the fraction of the objective function and the optimization variable, the original optimization problem is a classic non-convex optimization problem. This fractional programming problem is first solved by implementing the Dinkelbach method, which is expressed as where ξ* represents the optimal value that satisfies the fractional programming update, (·) * Represents the optimal value of the variable. Since ξ* cannot be obtained in advance, an update parameter ξ is used to replace ξ * , therefore, problem (18) is equivalently transformed as follows In order to solve the coupled optimization variables, the block coordinate descent (BCD) technology is used to solve each sub-problem separately; the penalty continuous convex approximation (SCA) method is used to optimize the smart reflector reflection coefficient sub-problem, and the maximum-minimization combined with semi-positive definite programming algorithm is used to optimize the transmission beamforming matrix.

5. The method for allocating synaesthesia resources for intelligent reflective surface-assisted backscattering according to claim 4, characterized in that: In step 4, the user resource allocation strategy problem is transformed into a linear programming problem and solved based on CVX, as follows: When fixed ω k , W, θ k When there are variables, the optimization problem of time allocation can be expressed as: Where ζ is the auxiliary variable introduced, d k =ω k (R c,k -ξNμ), The optimization problem of time allocation is a linear programming problem, so it is solved by CVX tools; k The optimization problem is also a linear programming problem and is solved in the same way.

6. The method for allocating synaesthesia resources for intelligent reflective surface-assisted backscattering according to claim 5, characterized in that: In step 5, the optimization problem of the reflection coefficient of the smart reflective surface is solved by a continuous convex approximation method based on a penalty function, as follows: Fixing other variables, the optimization problem of the reflection coefficient of the smart reflector can be reformulated as: Among them, α k , k and β k is the auxiliary variable introduced. By transforming the variables, we get: in, and in Indicates redefining the variables. Based on the above transformation, the optimization problem is: where rank(·) represents the rank of the matrix. Then, a continuous convex approximation scheme based on a penalty function is used to optimize the problem of (27), and the objective function is expanded by a first-order Taylor expansion. The optimization problem is expressed as: Where δ is the penalty factor, the problem in equation (27) is transformed into the convex problem (28) and can be solved. The optimization variables are The reflection coefficient of the optimal smart reflective surface is restored.

7. The method for allocating synaesthesia resources for intelligent reflective surface-assisted backscattering according to claim 6, characterized in that: In step 6, the optimization problem of the base station transmit beam is solved by combining the maximization-minimization and semidefinite relaxation algorithms, as follows: By fixing other variables, the optimization problem of the base station transmit beam is expressed as The terms in brackets in the objective function can be restated as: in The original problem is expressed as: in The second non-convex constraint is a convex difference programming problem. In order to solve the convex difference programming problem, the convex difference programming problem is reformulated as a series of problems using the maximum minimization algorithm and optimized until convergence is achieved; Transform using the first-order Taylor expansion: Therefore, problem (33) can be transformed into finding R k New Problems with Feasible Solutions of R By ignoring the rank-one constraint, the optimal communication beam is expressed as in According to Cholesky decomposition, the radar transmit beam is obtained 8. A synaesthesia resource allocation system assisted by backscattering with intelligent reflective surface, characterized in that: The system is used to implement the synaesthesia resource allocation method for intelligent reflective surface assisted backscattering according to any one of claims 1 to 7. The system includes the first module to the sixth module, and the functions of each module are as follows: The first module is to construct a system model of the synaesthesia computing network with intelligent reflective surface assisted backscattering; The second module constructs the optimization problem of user resource allocation strategy, base station transmission beam and smart reflector reflection coefficient; In the third module, fractional programming and block coordinate descent methods are used to transform the original problem into multiple sub-optimization problems; The fourth module transforms the user resource allocation strategy problem into a linear programming problem and solves it based on CVX; The fifth module solves the optimization problem of the reflection coefficient of the smart reflective surface through a continuous convex approximation method based on a penalty function; The sixth module solves the optimization problem of base station transmit beam by combining maximization-minimization and semi-definite relaxation algorithms.

9. A mobile terminal comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the synaesthesia resource allocation method for intelligent reflective surface assisted backscattering according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in the method for allocating synaesthesia resources for intelligent reflective surface-assisted backscattering as claimed in any one of claims 1 to 7 are implemented.

Images