A resource scheduling method for a cognitive backscattering communication system based on imperfect CSI

Abstract

The application belongs to the technical field of Internet of Things, and particularly relates to a resource scheduling method of a cognitive backscattering communication system based on non-perfect CSI, which comprises the following steps: establishing a signal transmission model of the cognitive backscattering communication network; considering the quality of service of the cognitive backscattering user, the backscattering coefficient constraint, the transmission time constraint and the maximum interference power constraint of the primary receiver, constructing a resource allocation problem of maximum secondary system throughput; establishing a multivariate coupled nonlinear robust resource allocation model; and converting the original problem into a deterministic convex optimization problem by using the worst criterion, continuous convex approximation, S-Procedure, variable substitution and alternating iteration method and solving the problem. By converting the system performance optimization problem into a deterministic convex optimization problem by using the worst criterion, continuous convex approximation, S-Procedure, variable substitution and alternating iteration method, the application has better throughput and robustness compared with the non-robust algorithm.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of Internet of Things, and particularly relates to a resource scheduling method of a cognitive backscatter communication system based on imperfect CSI. BACKGROUND

[0002] In future Internet of Things systems, the number of sensor nodes is increasing, leading to exponential growth in the demand for spectrum. In addition, due to the small size of Internet of Things devices, it is difficult to equip them with large-capacity batteries, resulting in limited running time of Internet of Things devices, so it is crucial to design spectrum-efficient and energy-saving Internet of Things communication technology. The cognitive backscatter communication system combines the advantages of cognitive radio and backscatter, and can effectively improve the spectrum efficiency and energy efficiency. The cognitive backscatter communication system is composed of a primary system and a secondary system, and the secondary system shares the spectrum resources of the primary user, effectively improving the spectrum efficiency of the communication system. Under the premise of ensuring the communication quality of the primary user, the secondary system uses the radio frequency signal of the primary system for wireless power supply, and modulates its own signal on the radio frequency source signal, so as to realize low-power reflection communication.

[0003] Many existing methods have studied cognitive backscatter communication networks, but most of them are based on single antenna and do not consider multiple antennas from the perspective of multiple antennas, however, in the existing B5G, multiple antennas have better communication performance. In addition, many existing methods only consider the system performance optimization problem under ideal channel conditions. However, in actual systems, it will be affected by channel estimation error, quantization error, reflection circuit nonlinear characteristics and feedback delay, making it difficult to obtain perfect channel state information, so that the designed algorithm is invalid in the actual system. Therefore, in order to facilitate practical engineering application, a method is needed to improve the throughput and robustness of the cognitive backscatter communication system. SUMMARY

[0004] To solve the above technical problems, the application provides a resource scheduling method of a cognitive backscatter communication system based on imperfect CSI, comprising the following steps:

[0005] S1: establishing a signal transmission model of the cognitive backscatter communication network;

[0006] S2: considering the quality of service of the cognitive backscatter user and the backscatter coefficient constraint, the transmission time constraint, and the maximum interference power constraint of the primary receiver, constructing a resource allocation problem of maximizing the secondary system throughput;

[0007] S3: based on the bounded channel uncertainty, converting the resource allocation problem of maximizing the secondary system throughput into a multivariate coupled nonlinear robust resource allocation problem;

[0008] S4: converting the multivariable coupled nonlinear robust resource allocation problem into a deterministic convex optimization problem by using worst-case criteria, successive convex approximation, S-Procedure, variable substitution and alternating iteration method, and solving by using convex optimization tools to obtain a resource scheduling scheme.

[0009] The present application has the advantages that: the present application converts the system performance optimization problem into a deterministic convex optimization problem by using worst-case criteria, successive convex approximation, S-Procedure, variable substitution and alternating iteration method, and has better throughput and robustness compared with non-robust algorithms. BRIEF DESCRIPTION OF DRAWINGS

[0010] Figure 1 A flow chart of the resource scheduling method for the cognitive backscatter communication system with imperfect CSI in the present application. DETAILED DESCRIPTION

[0011] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present application.

[0012] A resource scheduling method for a cognitive backscatter communication system based on imperfect CSI, as shown in Figure 1 , comprises the following steps.

[0013] S1: establishing a signal transmission model of the cognitive backscatter communication network;

[0014] S2: constructing a resource allocation problem of maximizing the secondary system throughput according to the signal transmission model of the cognitive backscatter communication network, considering the quality of service of the cognitive backscatter user, the backscatter coefficient constraint, the transmission time constraint and the maximum interference power constraint of the primary receiver;

[0015] S3: converting the resource allocation problem of maximizing the secondary system throughput into a multivariable coupled nonlinear robust resource allocation problem based on bounded channel uncertainty;

[0016] S4: converting the multivariable coupled nonlinear robust resource allocation problem into a deterministic convex optimization problem by using worst-case criteria, successive convex approximation, S-Procedure, variable substitution and alternating iteration method, and solving by using convex optimization tools to obtain a resource scheduling scheme.

[0017] The signal transmission model of the cognitive backscatter communication network comprises the following steps.

[0018] A primary base station with N antennas, M primary receivers, K cognitive backscatter users, a cognitive information receiver; and the primary user and secondary user sets are defined as and The primary base station transmits signals, the cognitive backscatter users share the primary user spectrum resources in the underlay spectrum sharing mode and reflect the backscatter signals in the throughput of the kth cognitive backscatter user in the backscatter phase, the cognitive backscatter users cause interference to the M primary receivers in the backscatter phase, the signal of the kth cognitive backscatter user is reflected to the cognitive information receiver in turn through the time division multiple access mode, the total transmission time is defined as T, and the transmission time of the kth cognitive backscatter user is τ k , and satisfies In this mode, the cognitive backscatter users can use the licensed spectrum for data transmission at the same time as the primary receivers. Therefore, the total interference of the cognitive backscatter users to any primary receiver is less than the interference temperature threshold. In order to improve the spectrum utilization and transmission efficiency, the cognitive backscatter users have the ability of backscatter circuit and spectrum sensing, and can transmit data to the cognitive information receiver through backscatter. It is assumed that all channels satisfy the block fading channel, that is, they remain unchanged within a small time frame and are time-varying in the whole time process.

[0019] The cognitive information receiver receives the signals of the primary base station and the kth cognitive backscatter user, including:

[0020]

[0021] wherein represents the signals of the primary base station and the kth cognitive backscatter user received by the cognitive information receiver, c k represents the signal of the kth cognitive backscatter user, has E[|s m v 2 ]=1 and E[|c k | 2 ]=1, E[ ] represents the expected value, s m represents the beamforming signal of the primary base station sent to the mth primary receiver, β k ∈[0,1] represents the backscatter coefficient of the kth cognitive backscatter user, f k ∈C N×1 is the channel vector of the primary base station to the kth cognitive backscatter user, h k represents the channel coefficient of the kth cognitive backscatter user to the cognitive information receiver, f p ∈C N×1 represents the channel vector of the primary base station to the cognitive information receiver, n represents the Gaussian white noise at the cognitive information receiver, satisfies n~CN(0,σ 2) denotes subject to additive white Gaussian noise with mean zero and variance Sigma 2 2 denotes interference at the cognitive information receiver, x P denotes the transmit signal of the primary base station, H denotes the conjugate transpose operation, C N×1 denotes an N x 1 dimensional complex vector.

[0022] The throughput of the k-th cognitive backscatter user in the backscatter phase, R

[0023]

[0024] where R k denotes the throughput of the k-th cognitive backscatter user in the backscatter phase, τ k denotes the time for the k-th cognitive backscatter user to perform backscatter, β k ∈ [0, 1] denotes the backscatter coefficient of the k-th cognitive backscatter user, f k ∈ C N×1 is the channel vector from the primary base station to the k-th cognitive backscatter user, h k denotes the channel coefficient from the k-th cognitive backscatter user to the cognitive information receiver, f p ∈ C N×1 denotes the channel vector from the primary base station to the cognitive information receiver, w m ∈ C N×1 denotes the beamforming vector from the primary base station to the m-th primary receiver, H denotes the conjugate transpose operation, M denotes the number of primary receivers, σ 2 denotes interference at the cognitive information receiver, Tr() denotes the trace of a matrix, C N×1 denotes an N x 1 dimensional complex vector.

[0025] All cognitive backscatter users cause interference to the m-th primary receiver. To describe the interference level at the primary receiver side, the interference temperature model has been proposed early. The interference temperature model is mainly to ensure that the communication of the primary users is not affected when sharing resources, and it is necessary to ensure that the interference caused by all cognitive backscatter users to the m-th primary receiver satisfies the following constraint:

[0026]

[0027] where I m and denote the average interference power of all cognitive backscatter users to the m-th primary receiver and the maximum interference power threshold that the m-th primary receiver can tolerate, τ k denotes the time for the k-th cognitive backscatter user to perform backscatter, β k ​∈ [0, 1] represents the backscattering coefficient of the kth cognitive backscattering user, f k ∈ C N×1 is the channel vector from the primary base station to the kth cognitive backscattering user, h k represents the channel coefficient from the kth cognitive backscattering user to the cognitive information receiver, w m ∈ C N×1 represents the beamforming vector from the primary base station to the mth primary receiver, H represents the conjugate transpose operation, M represents the number of primary receivers, K represents the number of cognitive backscattering users, T represents the total transmission time, Tr( ) represents the trace of a matrix, h k,m represents the channel coefficient from the kth cognitive backscattering user to the mth primary receiver, C N×1 represents an N x 1 dimensional complex vector.

[0028] Considering the quality of service of cognitive backscattering users and the backscattering coefficient constraint, the transmission time constraint, and the maximum interference power constraint of primary receivers, a resource allocation problem of maximizing the secondary system throughput is constructed

[0029] Under perfect channel state information, the resource allocation problem of maximizing the secondary system throughput includes:

[0030]

[0031] Subject to

[0032]

[0033]

[0034]

[0035] C5: 0 ≤ β k ≤ 1

[0036] where P1 represents the resource allocation problem of maximizing the secondary system throughput, P max represents the maximum transmission power of the primary base station, R k represents the throughput of the kth cognitive backscattering user in the backscattering phase, represents the minimum throughput threshold of the kth cognitive backscattering user, C1 represents the maximum transmission power constraint of the primary base station, C2 represents the constraint of guaranteeing the minimum throughput of the kth cognitive backscattering user, C3 represents the constraint of guaranteeing the average maximum interference power of the primary receiver, C4 represents the total transmission time constraint, C5 represents the backscattering coefficient constraint, w m ∈ C N×1 represents the beamforming vector from the primary base station to the mth primary receiver, τ kdenotes the time when the kth cognitive backscatter user backscatters, β k denotes the backscattering coefficient of the kth cognitive backscatter user, I m and denote the average interference power of all cognitive backscatter users to the mth primary receiver and the maximum interference power threshold that the mth primary receiver can tolerate, respectively, M denotes the number of primary receivers, K denotes the number of cognitive backscatter users, T denotes the total transmission time, C N×1 denotes an N x 1 dimensional complex vector.

[0037] Due to the electromagnetic environment between the primary base station and the cognitive backscatter user and the cognitive receiver link is not known, resulting in inaccurate system channel state information, therefore, bounded channel uncertainty, including:

[0038]

[0039] where R F , R H , and R Q denote the uncertainty set of the channel from the primary base station to the cognitive information receiver, the kth cognitive backscatter user to the cognitive information receiver, and the kth cognitive backscatter user to the mth primary receiver, respectively, and denote the channel estimates of the channel from the primary base station to the cognitive information receiver, the kth cognitive backscatter user to the cognitive information receiver, and the kth cognitive backscatter user to the mth primary receiver, respectively, Δf p , Δh k , and Δh k,m denote the channel estimation errors of the channel from the primary base station to the cognitive information receiver, the kth cognitive backscatter user to the cognitive information receiver, and the kth cognitive backscatter user to the mth primary receiver, respectively, ψ k ≥ 0, and ∈ k ≥ 0 denote the upper bounds of the first, second, and third uncertainty parameters, respectively, f p ∈ C N×1 denotes the channel vector from the primary base station to the cognitive information receiver, h k denotes the channel coefficient of the kth cognitive backscatter user to the cognitive information receiver, h k,m denotes the channel coefficient of the kth cognitive backscatter user to the mth primary receiver, || || F denotes the F-norm, C N×1 denotes an N x 1 dimensional complex vector.

[0040] The resource allocation problem of maximizing the secondary system throughput is converted into a multivariable coupled nonlinear robust resource allocation problem, including:

[0041]

[0042] subject to C1, C4, C5

[0043]

[0044]

[0045] C6: Δf p ∈ R F , Δh k ∈ R H , Δh k,m ∈ R Q

[0046] where P2 denotes a nonlinear robust resource allocation problem with multivariable coupling, C1 denotes the maximum transmit power constraint of the primary base station, C2 denotes the minimum throughput constraint of the k-th cognitive backscatter user, C3 denotes the average maximum interference power constraint of the primary receiver, C4 denotes the total transmission time constraint, C5 denotes the backscatter coefficient constraint, C6 denotes the set of uncertain parameters, denotes the minimum throughput constraint with uncertainty, denotes the average maximum interference power constraint of the primary receiver with uncertainty, R k denotes the throughput of the k-th cognitive backscatter user in the backscatter phase, denotes the minimum throughput threshold of the k-th cognitive backscatter user, w m ∈ C N×1 denotes the beamforming vector of the primary base station to the m-th primary receiver, τ k denotes the time of the k-th cognitive backscatter user for backscatter, β k ∈ [0, 1] denotes the backscatter coefficient of the k-th cognitive backscatter user, I m and denote the average interference power of all cognitive backscatter users to the m-th primary receiver and the maximum interference power threshold that the m-th primary receiver can tolerate, respectively, Δf p , Δh k and Δh k,m denote the channel estimation error from the primary base station to the cognitive information receiver, the k-th cognitive backscatter user to the cognitive information receiver, and the k-th cognitive backscatter user to the m-th primary receiver, respectively, R F , R H , R Qdenote the sets of uncertainties from the primary base station to the cognitive information receiver, the kth cognitive backscatter user to the cognitive information receiver, and the kth cognitive backscatter user to the mth primary receiver, respectively, R sum denotes the total throughput of the system, C N×1 denotes an N x 1 dimensional complex vector.

[0047] The multivariable coupled nonlinear robust resource allocation problem is converted into a deterministic convex optimization problem by using the worst-case criterion, successive convex approximation, S-Procedure, variable substitution and alternating iteration method, and solved by using convex optimization tools to obtain the resource scheduling scheme, specifically including:

[0048] S41: The channel uncertainty in the multivariable coupled nonlinear robust resource allocation problem is processed by using the worst-case criterion, successive convex approximation, S-Procedure method, and converted into a deterministic problem;

[0049] S42: The deterministic problem is converted into two sub-problems by using the alternating iteration method, the rank-one constraint is discarded, and auxiliary variables are introduced based on the variable substitution method to process the existing coupling variable constraints;

[0050] S43: The two sub-problems are solved by using convex optimization tools to obtain the beamforming vector, reflection coefficient and backscatter transmission time, i.e., to obtain the resource scheduling scheme; the convex optimization tool is the CXV toolbox.

[0051] The channel uncertainty in the multivariable coupled nonlinear robust resource allocation problem is processed by using the worst-case criterion, successive convex approximation, S-Procedure method, and converted into a deterministic problem, including:

[0052] The channel uncertainty and non-convexity in the objective function and C2 are processed, and the following variable relaxations are considered

[0053]

[0054]

[0055]

[0056] where θ k , φ k and r k are relaxation variables;

[0057] However, is still a non-convex constraint, in order to process the non-convex constraint, the successive convex approximation method and Taylor series expansion are adopted, and then is approximated as:

[0058]

[0059] where, and are slack variables. and are the last iteration of and

[0060] Definition then can be rewritten as:

[0061]

[0062] where,

[0063] Theorem 1 (S-procedure) method: Definition where i = {1, 2}, A i ∈ C N×N , b i ∈ C N×1 , x ∈ C N×1 , c i ∈ R, if and only if there exists an auxiliary variable λ ≥ 0, holds, then:

[0064]

[0065] According to Theorem 1, the above formula is transformed into a semi-positive constraint form:

[0066]

[0067] where λ1≥ 0 is the introduced slack variable.

[0068] To deal with the channel uncertainty, the worst-case method is adopted, then

[0069]

[0070] To deal with the channel uncertainty in C3, the Cauchy-Schwarz inequality is used for transformation, then

[0071]

[0072]

[0073]

[0074]

[0075] ​The deterministic problem is converted into two sub-problems by using alternating optimization; based on variable substitution method, auxiliary variables are introduced, The existing coupling variable constraints are processed;

[0076] After converting the objective function and all uncertainty constraints into deterministic form, the deterministic robust resource allocation problem is obtained as:

[0077]

[0078] Subject to C4,C5

[0079]

[0080]

[0081]

[0082]

[0083] C7:Rank(W m )=1

[0084] Where W m denotes the product of beamforming vector sent by the main base station to the mth main receiver and its conjugate transpose, denotes the maximum transmit power deterministic constraint of the main base station, denotes the minimum throughput deterministic constraint of the kth cognitive backscatter user, denotes the average maximum interference power deterministic constraint to guarantee the main receiver, C4 denotes the total transmission time constraint, and C5 denotes the backscatter coefficient constraint, denotes the constraint of W m , C7 denotes the rank one constraint of W m , P max denotes the maximum transmit power of the main base station, τ k denotes the time for the kth cognitive backscatter user to backscatter, r k denotes the rate of the kth cognitive backscatter user, denotes the minimum throughput threshold of the kth cognitive backscatter user, β k ∈[0,1] denotes the backscatter coefficient of the kth cognitive backscatter user, denotes the channel estimation value from the kth cognitive backscatter user to the mth main receiver, ∈ k denotes the upper bound of the third uncertainty parameter, f k denotes the channel vector from the main base station to the kth cognitive backscatter user, denotes the maximum interference power threshold that the mth primary receiver can tolerate, Tr() denotes the trace of a matrix, M denotes the number of primary receivers, K denotes the number of cognitive backscatter users, T denotes the total time of transmission, Λ and Θ denote the first and second slack variables respectively, and Rank() denotes the rank of a matrix.

[0085] The deterministic problem is converted into two sub-problems by using the alternating iteration method, the rank-one constraint is discarded, and auxiliary variables are introduced based on the variable substitution method The existing coupling variable constraints are processed, including:

[0086] Sub-problem 1: fixing τ k and β k , solving W m , Λ and Θ:

[0087]

[0088] Subject to

[0089] wherein, denotes the maximum transmit power deterministic constraint of the primary base station, denotes the minimum throughput deterministic constraint of the kth cognitive backscatter user, denotes the average maximum interference power deterministic constraint for guaranteeing the primary receivers, C4 denotes the total time of transmission constraint, and C5 denotes the backscatter coefficient constraint, denotes the constraint of W m , C7 denotes the rank-one constraint of W m , and W m denotes the product of the beamforming vector sent by the primary base station to the mth primary receiver and the conjugate transpose thereof, τ k denotes the time of backscatter performed by the kth cognitive backscatter user, r k denotes the rate of the kth cognitive backscatter user, K denotes the number of cognitive backscatter users, and Λ and Θ denote the first and second slack variables respectively.

[0090] The rank-one constraint is discarded by using the semidefinite relaxation method:

[0091]

[0092] Subject to

[0093] wherein P3 denotes the deterministic problem after the rank-one constraint is discarded, denotes the maximum transmit power deterministic constraint of the primary base station, denotes the minimum throughput deterministic constraint of the kth cognitive backscatter user, denotes the average maximum interference power deterministic constraint for guaranteeing the primary receivers, W m denotes the constraint of W m denotes the product of the beamforming vector sent by the primary base station to the mth primary receiver and its conjugate transpose, τ k denotes the time of the kth cognitive backscatter user to perform backscatter, r k denotes the rate of the kth cognitive backscatter user, Λ, Θ denote the first and second slack variables, respectively, and K denotes the number of cognitive backscatter users.

[0094] Subproblem 2: Fixing W m , Λ, and Θ, solve τ k and β k :

[0095]

[0096] subject to

[0097] wherein, denotes the minimum throughput certainty constraint of the kth cognitive backscatter user, denotes the average maximum interference power guarantee certainty constraint for the primary receivers, C4 denotes the total transmission time constraint, and C5 denotes the backscatter coefficient constraint, τ k denotes the time of the kth cognitive backscatter user to perform backscatter, r k denotes the rate of the kth cognitive backscatter user, β k ∈ [0, 1] denotes the backscatter coefficient of the kth cognitive backscatter user, and K denotes the number of cognitive backscatter users.

[0098] Based on the variable substitution method, auxiliary variables are introduced to handle the existing coupling variable constraints:

[0099]

[0100] subject to

[0101]

[0102]

[0103] wherein P4 denotes the certainty problem after handling the existing coupling variable constraints, denotes the minimum throughput certainty constraint of the kth cognitive backscatter user after variable substitution, denotes the average maximum interference power guarantee certainty constraint for the primary receivers after variable substitution, and C4 denotes the total transmission time constraint, denotes the backscatter coefficient determinism constraint after variable substitution, τ k denotes the time when the kth cognitive backscatter user performs backscatter, r k denotes the rate of the kth cognitive backscatter user, denotes the minimum throughput threshold of the kth cognitive backscatter user, θ k denotes the third slack variable, denotes the channel estimate of the kth cognitive backscatter user to the cognitive information receiver, denotes the second uncertainty parameter upper bound, f k denotes the channel vector of the primary base station to the kth cognitive backscatter user, w m beamforming vector of the primary base station to the mth primary receiver, denotes the channel estimate of the kth cognitive backscatter user to the mth primary receiver, k denotes the third uncertainty parameter upper bound, T denotes the total time of transmission, denotes the maximum interference power threshold that the mth primary receiver can tolerate, K denotes the number of cognitive backscatter users, M denotes the number of primary receivers, Tr() denotes the trace of a matrix.

[0104] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary and that changes in form and detail can be made without departing from the spirit and scope of the application, which is defined by the following claims and their equivalents.

Claims

1. A method for resource scheduling of a cognitive backscattering communication system based on imperfect CSI, characterized in that, Comprise: S1: establish a signal transmission model of cognitive backscatter communication network; The signal transmission model of the cognitive backscatter communication network comprises: A main base station with N antennas, M main receivers, K cognitive backscatter users, and a cognitive information receiver; the main base station transmits signals, the cognitive backscatter users share the main user spectrum resources in a downcast spectrum sharing mode and reflect backscatter signals in the backscatter phase, the cognitive backscatter users in the backscatter phase cause interference to the M main receivers, and the signals of the K cognitive backscatter users are reflected to the cognitive information receiver in turn through time division multiple access; The cognitive information receiver receives signals of the main base station and the K cognitive backscatter users, comprising: wherein denotes the signal received by the cognitive information receiver from the primary base station and the kth cognitive backscatter user, c k denotes the signal received by the cognitive information receiver from the kth cognitive backscatter user, has E[|s m | 2 ]=1 and E[|c k | 2 ]=1, E[] denotes the expected value, s m denotes the beamformed signal transmitted by the primary base station to the mth primary receiver, β k denotes the backscatter coefficient of the kth cognitive backscatter user, f k ∈ C N×1 is the channel vector from the primary base station to the kth cognitive backscatter user, h k denotes the channel coefficient from the kth cognitive backscatter user to the cognitive information receiver, f p ∈ C N×1 denotes the channel vector from the primary base station to the cognitive information receiver, n denotes the Gaussian white noise at the cognitive information receiver, satisfies n ~ CN(0, σ 2 ), σ 2 denotes the interference at the cognitive information receiver, x P denotes the transmit signal of the primary base station, H denotes the conjugate transpose operation, C N＝1 denotes the complex vector of N=1 dimension; The throughput of the K cognitive backscatter user in the backscatter phase, comprising: wherein R k represents the throughput of the kth cognitive backscatter user in the backscatter phase, τ k represents the time for the kth cognitive backscatter user to perform backscattering, w m ∈ C N＝1 represents the beamforming vector sent by the primary base station to the mth primary receiver, M represents the number of primary receivers, and Tr() represents the trace of a matrix. The interference of all cognitive backscatter users in the network to the M main receivers satisfies the following constraints: where I m and respectively represent the average interference power of all cognitive backscatter users to the mth primary receiver and the maximum interference power threshold that the mth primary receiver can tolerate, K represents the number of cognitive backscatter users, T represents the total transmission time, h k,m represents the channel coefficient of the kth cognitive backscatter user to the mth primary receiver; S2: According to the signal transmission model of the cognitive backscatter communication network, considering the quality of service of the cognitive backscatter user and the backscatter coefficient constraint, the transmission time constraint, and the maximum interference power constraint of the main receiver, a resource allocation problem of maximizing the secondary system throughput is constructed; The resource allocation problem of maximizing the secondary system throughput constructed, comprising: subject to C5:0≤β k ≤1 where P1 represents a resource allocation problem for maximizing the throughput of the secondary system, P2 represents a resource allocation problem for maximizing the throughput of the primary system, max represents the maximum transmit power of the primary base station, represents the minimum throughput threshold of the kth cognitive backscatter user, C1 represents the maximum transmit power constraint of the primary base station, C2 represents the minimum throughput constraint of the kth cognitive backscatter user, C3 represents the average maximum interference power constraint of the primary receiver, C4 represents the total transmission time constraint, and C5 represents the backscatter coefficient constraint. S3: Based on the bounded channel uncertainty, the resource allocation problem of maximizing the secondary system throughput is converted into a multivariable coupled nonlinear robust resource allocation problem; The resource allocation problem of maximizing the secondary system throughput is converted into a multivariable coupled nonlinear robust resource allocation problem, comprising: Subject to C1, C4, C5 C6: Δf p ∈R F , Δh k ∈R H , Δh k,m ∈R Q where P2 denotes a multivariable coupled nonlinear robust resource allocation problem, C6 denotes a set of uncertain parameters, denotes a minimum throughput constraint with uncertainty, denotes an average maximum interference power constraint of the primary receivers with uncertainty, denotes a minimum throughput threshold of the kth cognitive backscatter user, Af p , Ah k , and Ah k,m denote channel estimation errors from the primary base station to the cognitive information receiver, the kth cognitive backscatter user to the cognitive information receiver, and the kth cognitive backscatter user to the mth primary receiver, respectively, F , R H , R Q denote sets of uncertainties from the primary base station to the cognitive information receiver, the kth cognitive backscatter user to the cognitive information receiver, and the kth cognitive backscatter user to the mth primary receiver, respectively, sum denotes the total throughput of the system; S4: The multivariable coupled nonlinear robust resource allocation problem is converted into a deterministic convex optimization problem by using the worst criterion, continuous convex approximation, S-Procedure, variable substitution and alternating iteration method, and solved by using a convex optimization tool to obtain a resource scheduling scheme; S41: The channel uncertainty in the multivariable coupled nonlinear robust resource allocation problem is converted into a deterministic problem by using the worst criterion, continuous convex approximation, and S-Procedure method; S42: convert the deterministic problem into two sub-problems by using an alternating iterative method, drop the rank-one constraint and introduce auxiliary variables based on variable substitution method handle the existing coupling variable constraints; S43: Two sub-problems are solved by using a convex optimization tool to obtain beamforming vectors, reflection coefficients, and backscatter transmission time, i.e. to obtain a resource scheduling scheme; The channel uncertainty in the multivariable coupled nonlinear robust resource allocation problem is converted into a deterministic problem by using the worst criterion, continuous convex approximation, and S-Procedure method, comprising: Subject to C4, C5 W m ≥0 C7: Rank(W m ) = 1 where W m denotes the product of the beamforming vector transmitted by the primary base station to the mth primary receiver and the conjugate transpose of it, denotes the maximum transmit power determinism constraint of the primary base station, denotes the minimum throughput determinism constraint of the kth cognitive backscatter user, denotes the average maximum interference power determinism constraint of the primary receivers, denotes the constraint of W m , C7 denotes the constraint of W m , rank-one constraint, r k denotes the rate of the kth cognitive backscatter user, denotes the channel estimation value of the kth cognitive backscatter user to the mth primary receiver, ∈ k denotes the third uncertainty parameter upper bound, Λ, Θ denote the first and second relaxation variables, respectively, and Rank() denotes the rank of a matrix.

2. The resource scheduling method for a cognitive backscattering communication system based on non- perfect CSI according to claim 1, wherein, The bounded channel uncertainty comprises: where R F , R H , and R Q denote the uncertainty sets of the channel from the primary base station to the cognitive information receiver, the k-th cognitive backscatter user to the cognitive information receiver, and the k-th cognitive backscatter user to the m-th primary receiver, respectively, and denote the channel estimates of the channel from the primary base station to the cognitive information receiver, the k-th cognitive backscatter user to the cognitive information receiver, and the k-th cognitive backscatter user to the m-th primary receiver, respectively, Δf p , Δh k , and Δh k,m denote the channel estimation errors of the channel from the primary base station to the cognitive information receiver, the k-th cognitive backscatter user to the cognitive information receiver, and the k-th cognitive backscatter user to the m-th primary receiver, respectively, Ψ k ≥ 0, and ∈ k ≥ 0 denote the first, second, and third uncertainty parameter upper bounds, respectively, f p ∈ C N×1 denotes the channel vector from the primary base station to the cognitive information receiver, h k denotes the channel coefficient of the k-th cognitive backscatter user to the cognitive information receiver, h k,m denotes the channel coefficient of the k-th cognitive backscatter user to the m-th primary receiver, ||| F denotes the F-norm, and C N×1 denotes an N x 1 dimensional complex vector.

3. The method of Claim 1, wherein The alternating iterative method is used to transform the deterministic problem into two sub-problems, the rank-one constraint is dropped and auxiliary variables are introduced based on the variable substitution method handling the existing coupling variable constraints, including: Sub-problem 1: Fix τ k and β k , solve for W m , Λ, and Θ: subject to C7 wherein, C1 represents a maximum transmit power determinism constraint of the primary base station, C2 represents a minimum throughput determinism constraint of the kth cognitive backscatter user, C3 represents an average maximum interference power determinism constraint guaranteeing the primary receivers, C4 represents a total transmission time constraint, and C5 represents a backscatter coefficient constraint, C6 represents a constraint of W m C7 represents a rank one constraint of W m C8 represents a constraint of W m C9 represents a product of a beamforming vector transmitted by the primary base station to the mth primary receiver and a conjugate transpose thereof, τ k C10 represents a time for backscattering by the kth cognitive backscatter user, r k C11 represents a rate of the kth cognitive backscatter user, K represents a number of cognitive backscatter users, and Λ, Θ represent first and second slack variables, respectively; The rank-one constraint is discarded by using a semidefinite relaxation method: subject to wherein P3 represents a deterministic problem after dropping the rank-one constraint, represents a maximum transmit power deterministic constraint of the primary base station, represents a minimum throughput deterministic constraint of the kth cognitive backscatter user, represents an average maximum interference power deterministic constraint to guarantee the primary receiver, represents a constraint of W m represents a constraint of W m represents a product of a beamforming vector transmitted by the primary base station to the mth primary receiver and a conjugate transpose thereof, τ k represents a time for backscattering by the kth cognitive backscatter user, r k represents a rate of the kth cognitive backscatter user, Λ, Θ represent first and second relaxation variables, respectively, and K represents a number of cognitive backscatter users. Sub-problem 2: Fix W m , A, and Θ, solve for τ k and β k : subject to C4, C5 wherein, denotes the minimum throughput certainty constraint for the kth cognitive backscatter user, denotes the average maximum interference power certainty constraint to guarantee the primary receiver, C4 denotes the total transmission time constraint, C5 denotes the backscatter coefficient constraint, τ k denotes the time for the kth cognitive backscatter user to perform backscatter, r k denotes the rate of the kth cognitive backscatter user, β k denotes the backscatter coefficient of the kth cognitive backscatter user, ∈ [0, 1], K denotes the number of cognitive backscatter users; Based on the variable substitution method, auxiliary variables are introduced Handling the existing coupling variable constraints: subject to where P4 represents the deterministic problem after processing the coupling variable constraint, represents the minimum throughput deterministic constraint of the kth cognitive backscatter user after variable substitution, represents the average maximum interference power deterministic constraint of the main receiver after variable substitution, C4 represents the total transmission time constraint, represents the backscatter coefficient deterministic constraint after variable substitution, τ k represents the time for the kth cognitive backscatter user to perform backscatter, r k represents the rate of the kth cognitive backscatter user, represents the minimum throughput threshold of the kth cognitive backscatter user, θ k represents the third slack variable, represents the channel estimation value of the kth cognitive backscatter user to the cognitive information receiver, represents the upper bound of the second uncertainty parameter, f k represents the channel vector of the main base station to the kth cognitive backscatter user, w m the beamforming vector of the main base station sent to the mth main receiver, the channel estimation value of the kth cognitive backscatter user to the mth main receiver, ∈ k represents the upper bound of the third uncertainty parameter, T represents the total transmission time, represents the maximum interference power threshold that the mth main receiver can tolerate, K represents the number of cognitive backscatter users, M represents the number of main receivers, and Tr() represents the trace of a matrix.

Images