Symbiotic radio system design and optimization method based on multiple active intelligent reflection surfaces

Abstract

The invention discloses a symbiotic radio system design and optimization method based on multiple active intelligent reflection surfaces. According to the method, high-efficiency backscatter communication is realized while the performance of the main transmission system is enhanced by using the intelligent reflection surface. A plurality of active intelligent reflecting surfaces are configured in the system, the main transmission performance is enhanced, and meanwhile, the active intelligent reflecting surfaces serve as backscattering equipment in turn, and transmission of secondary information is achieved by conducting scattering modulation on incident main signals. The method aims at maximizing the main transmission rate of the system, and jointly optimizes the base station sending power, the active beam forming vector and the reflection coefficient of the intelligent reflection surface. An original non-convex optimization problem is decomposed into a plurality of sub-problems, semi-definite relaxation, continuous convex approximation and other methods are adopted for conversion, and finally an optimal solution is obtained through iterative solution. Compared with a scheme based on a passive intelligent reflection surface and an active intelligent reflection surface scheme based on phase random, a higher main transmission rate can be realized under the condition of ensuring the same secondary symbol signal to interference plus noise ratio.

Description

Technical Field

[0001] This invention relates to the field of information and communication, specifically to the construction of a symbiotic radio system based on multiple active intelligent reflective surfaces and the joint optimization design of the base station's transmit power, active beamforming vector, and the reflection coefficient of the active intelligent reflective surface. Background Technology

[0002] With the rapid development of future wireless communication technologies, the utilization of high-frequency spectrum resources has become crucial for improving wireless communication capacity. However, the inherent propagation characteristics of high-frequency signals limit their coverage and severely affect communication quality. Traditional solutions rely on dense deployment of base stations, access points, or relay equipment, but these face problems such as high cost and high energy consumption. In recent years, Intelligent Reflecting Surface (IRS), as a new type of wireless communication enabling technology, has been considered one of the key technologies for improving the spectrum efficiency and energy efficiency of 6-Generation (6G) cellular wireless networks by dynamically adjusting the phase of electromagnetic waves to reconstruct the wireless channel environment [Wu Q, Zhang S, Zheng B, et al. Intelligent reflecting surface-aided wireless communications: atutorial[J].IEEE Transactions on Communications, 2021, 69(5): 3313-3351.]. IRS achieves passive beamforming by dynamically adjusting the phase shift of the incident signal, but because it is a reflection and cannot amplify the signal, the attenuation of the reflection link is the product of the attenuation of the two links, resulting in a weak signal in the reflection link. Current research on IRS mainly focuses on using IRS to adjust the phase of the incident signal to assist system communication, including the application of IRS in physical layer security, multi-cell cooperation, multiple access, and UAV communication. Active IRS is a further evolution of IRS, which integrates low-power active amplifier circuits in the reflector unit. Although it increases circuit power consumption, it can effectively overcome path loss and improve the strength of the received signal because it can actively amplify the reflected signal. Related research shows that in high-frequency bands and long-distance communication scenarios, the application of active IRS can more effectively improve system performance compared with passive IRS [Zhang Z, Dai L, Ling L, et al. Active RISvs. passive RIS: which will prevail in 6G? [J]. IEEE Transactions on Communications, 2023, 71(3): 1707-1725.].

[0003] In recent years, with the increasing number of Internet of Things (IoT) devices, improving the energy efficiency and spectrum efficiency of IoT has become an important research issue in academia and industry. Backscatter communication is a low-power transmission technology that uses existing radio frequency signals in the environment and modulates information onto the incident radio frequency signal by controlling the reflection characteristics to achieve information transmission [Van H, Lu X, Dinh T, et al. Ambient backscatter communications: a contemporary survey [J]. IEEE Communications Surveys & Tutorials, 2018, 20(4): 2889-2922.]. Symbiotic radio (SR), as a backscatter communication technology that efficiently utilizes energy and spectrum resources [Liang Y, Zhang Q, Larsson E, et al. Symbiotic radio: cognitive backscattering communications for future wireless networks[J]. IEEE Transactions on Cognitive Communications and Networking, 2020, 6(4): 1242-1255.], can effectively alleviate spectrum scarcity and improve energy efficiency by modulating environmental radio frequency signals with a backscatter device (BD) to transmit its secondary information without using a dedicated radio frequency (RF) signal transmitter. SR systems typically include two modes: symbiotic and parasitic [[9] Long R, Liang Y, Guo H, et al. Symbiotic radio: a new communication paradigm for passive internet of things[J]. IEEE Internet of Things Journal, 2020, 7(2): 1350-1363.]. When the secondary symbol period is much larger than the primary symbol period, it is in symbiotic mode. In this case, the reflection link in the form of BD can provide multipath gain for the primary transmission. When the secondary symbol period is comparable to the primary symbol period, the receiver of the primary transmission system cannot utilize the scattering modulation signal transmitted on the reflection link, and the scattering modulation signal will interfere with the primary transmission signal. The design, optimization, and performance analysis of SR systems are the main research contents at present.

[0004] In recent years, there have been some studies on the application of IRS in SR systems. There are two main application methods. One is that the IRS is used as a reflection device in the SR system to improve the performance of the main transmission and scattering transmission system communication. The other is that the IRS has a reflection function similar to BD. While using the IRS to improve the main transmission performance, it is also used as a BD to scatter and modulate the incident signal to generate a backscattered signal and transmit it to the scattering information receiver. The literature [Peng X, Tao Q, Gan X, et al. Intelligent reflecting surface-enhanced cell-free symbiotic radio systems[J].IEEE Internet of Things Journal,2023,10(22):19545-19557.] considers minimizing the transmission bit error rate of secondary symbols in a symbiotic mode SR system. Under the constraint of the main transmission rate, it jointly optimizes the active beamforming of the base station and the reflection coefficient matrix of the IRS. The literature [Long R,Liang Y,Di R,et al.Activeconfigurable intelligent surface-based symbiotic radio:boosting mutual benefits[J].IEEE Transactions on Cognitive Communications and Networking,2025,11(1):423-436.] focuses on SR systems using active IRS as BD. In the system symbiotic mode, it jointly optimizes the base station transmit power and the reflection coefficient matrix of the IRS to maximize the main transmission rate of the system. The optimization of the reflection coefficient matrix includes the optimization of the IRS reflection power. The literature [Hua M, Wu Q, Yang L, et al. A novel wireless communication paradigm for intelligent reflecting surface based symbiotic radio systems[J]. IEEE Transactions on Signal Processing, 2020, 70: 550-565.] considers the SR system in two different modes, symbiotic and parasitic, respectively. Under the requirement of minimum main transmission rate, it minimizes the bit error rate of secondary symbols by jointly optimizing the active beamforming of the base station and the reflection coefficient matrix of the IRS. The simulation results show that, since the main transmission in parasitic mode treats the reflected link signal as interference, the bit error rate in symbiotic mode is better than that in parasitic mode. Summary of the Invention

[0005] The purpose of this invention is to provide a design for a symbiotic radio system that combines an active smart reflector and backscattering technology when the user is simultaneously a primary and secondary information receiver, thereby improving the primary transmission rate. This scheme first jointly optimizes the base station transmit power, active beamforming vector, and reflection coefficient of the active smart reflector in different time slots of the system transmission frame to obtain the optimal primary transmission rate for each time slot, and then calculates the maximum average primary transmission rate.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: First, the system operates in parasitic mode. Within a system transmission frame, each active intelligent reflective surface takes turns acting as a reflector to send secondary symbols to the user in different time slots. The active intelligent reflective surface that does not send secondary symbols in the same time slot provides multipath gain for the main transmission. Further, an optimization problem is constructed with the goal of maximizing the main transmission rate of the system. Then, an iterative algorithm is proposed to alternately optimize the transmission power, the active beamforming vector, and the reflection coefficient of the active intelligent reflective surface until convergence is achieved.

[0007] The specific steps are as follows:

[0008] (1) Analyze the signal transmission process and construct a communication system model: Set up a symbiotic radio system with multiple active smart reflective surfaces in parasitic mode. The smart reflective surfaces are both devices that improve the main transmission performance and time-division reflective devices. The single antenna user is the receiver of both primary and secondary information. Set the channel in the system as a quasi-static flat fading channel. In one transmission frame, the smart reflective surfaces take turns as reflective devices and transmit secondary information to the receiver by scattering and modulating the incident signal. When not acting as a reflective device, they reflect the main signal to enhance the performance of the main transmission.

[0009] (2) Taking the maximization of the main transmission rate of the system as the optimization objective, an optimization mathematical model is constructed for the base station transmission power, active beamforming and intelligent reflective surface reflection coefficient under the constraints of the minimum secondary symbol rate and the amplification power of the intelligent reflective surface.

[0010] (3) The optimization problem is transformed into four sub-problems: optimizing the base station transmission power, optimizing active beamforming, optimizing the reflection coefficient of the smart reflective surface that transmits secondary symbols, and optimizing the reflection coefficient of the smart reflective surface that does not transmit secondary symbols. For each non-convex sub-problem, a convex approximation fitting method is used to transform it into a convex problem for solution.

[0011] (4) Use the alternating iterative algorithm to solve for the optimal transmit power, active beamforming vector, reflection coefficient of the smart reflector for transmitting secondary symbols, and reflection coefficient of the smart reflector for not transmitting secondary symbols.

[0012] Compared with existing related research, the present invention has the following beneficial technical effects: (1) Among the relevant literature on symbiotic radio system research that can be retrieved, such as [Hua M, Wu Q, Yang L, et al. A novel wireless communication paradigm for intelligent reflecting surface based symbiotic radio systems[J]. IEEE Transactions on Signal Processing, 2020, 70: 550-565.], most of the literature is based on the symbiotic radio system working in symbiotic mode and the reflection device is a traditional passive intelligent reflective surface for system performance analysis and rate optimization design. However, for wireless channels, the actual channel may have various uncertain path losses, and the reflection link has multiplicative attenuation, resulting in very weak reflection link signal received at the user. Therefore, the present invention considers the system working in parasitic mode and adopts an active intelligent reflective surface, which is more practical. (2) The present invention can improve system performance while ensuring the signal-to-interference-plus-noise ratio requirement of the secondary symbol. Simulation experiments show that the main transmission rate of the system of the present invention is significantly better than that of the traditional passive intelligent reflective surface scheme. (3) This invention maximizes the main transmission rate of the system by jointly optimizing the transmission power, active beamforming vector, and reflection coefficient of the active smart reflector under different time slots. To address the difficulty in finding the optimal solution, an iterative algorithm is proposed to alternately optimize the transmission power, active beamforming vector, and reflection coefficient of the active smart reflector until convergence is achieved. Attached Figure Description

[0013] Figure 1 This is a communication system model of the present invention;

[0014] Figure 2 This is the convergence process of the alternating iterative algorithm of this invention;

[0015] Figure 3 The variation of average main transmission rate with the maximum transmission power of the base station;

[0016] Figure 4 The average main transmission rate varies with the number of antennas N;

[0017] Figure 5 The variation of average primary transmission rate with secondary symbol signal-to-interference-plus-noise ratio threshold;

[0018] Figure 6 The average primary transmission rate varies with the number of IRS K;

[0019] Figure 7 The variation of the average main transmission rate with the number of reflection units Q. Detailed Implementation

[0020] Consider as Figure 1 The transmission system model shown consists of a base station with N antennas, a single-antenna user, and K active IRSs equipped with Q reflector elements. It is assumed that the channel in the system is a quasi-static flat fading channel. Let the direct link channel coefficient vector from the base station to the user be denoted as... The channel coefficient matrix from the base station to the k-th (k∈{1,2,…,K}) IRS is: The channel coefficient vector from the k-th IRS to the user is

[0021] A single-antenna user is simultaneously the receiver of primary and secondary information. The SR system operates in parasitic mode, with the primary symbol period and secondary symbol period being the same. Assume the symbols transmitted by the base station are s[n], with a mean of 0 and a variance of 1. For simplicity, the symbol number "[n]" will be omitted in the following description. A transmission frame consists of K equal-length time slots, denoted by t, where t∈{1,2,…,K}. Each time slot contains one IRS acting as a BD, transmitting secondary messages to the receiver by scattering and modulating the signal transmitted by the base station. K IRSs take turns acting as BDs, and the remaining K-1 IRSs act as auxiliary reflection devices to improve the performance of the primary transmission. Let the beamforming vector of the base station in the t-th time slot be denoted as... Transmission power is P (t) The reflection coefficient matrix of the k-th IRS is in Let represent the reflection coefficient of the q-th reflecting unit, where q∈{1,2,…,Q}. Without loss of generality, the t-th IRS in the t-th time slot is taken as the BD. The transmitted signal of the base station in the t-th time slot is...

[0022]

[0023] The t-th IRS, acting as a BD, reflects the incident signal from the base station back to the receiver after scattering and modulation. Let c be the secondary symbol representing unit power. The reflected signal of the t-th IRS can be expressed as...

[0024]

[0025] in, This indicates the thermal noise generated by the amplifier circuitry inside the active IRS. The symbol CN indicates that it follows a complex Gaussian distribution. This represents thermal noise power.

[0026] The remaining K-1 IRSs act as reflectors, amplifying and reflecting the incident signal to provide multipath gain for the main transmission. The reflected signal of the k-th IRS can be expressed as...

[0027]

[0028] The k-th IRS reflected signal can be represented as

[0029]

[0030] The amplification power of the k-th active IRS is

[0031]

[0032] Here, the symbol || denotes the Euclidean norm of a vector. The user's received signal can be represented as...

[0033]

[0034] Among them, z u The noise at the user's location is complex Gaussian white noise with variance of

[0035] The user needs to decode the main signal and a signal containing secondary information. First, the main signal is decoded. Due to the parasitic mode, the signal reflected by the t-th IRS in the t-th time slot is the secondary modulated signal, which acts as interference when decoding the main signal. The signals reflected by the remaining IRSs are multipath transmission signals of the main signal. The signal-to-interference-plus-noise ratio (SIR) for decoding the main signal is...

[0036]

[0037] Therefore, the transmission rate of the main information is

[0038]

[0039] After decoding the main signal, the receiver can use Successive Interference Cancellation (SIC) to cancel the main signal from the direct link, thus obtaining...

[0040]

[0041] The average signal-to-interference-plus-noise ratio of the secondary signal is

[0042]

[0043] Wherein, the symbol E represents the averaging operation, and the average main transmission rate within a transmission frame is...

[0044]

[0045] The power consumption of an active IRS is

[0046]

[0047] Among them, P c and P dc η represents the power consumption of the control circuit, amplifier circuit, and DC bias circuit of each reflector unit of the active IRS, respectively, and η represents the power amplification efficiency of the active IRS.

[0048] This paper considers optimizing the base station transmit power, active beamforming vector, and reflection coefficient matrices of K IRSs to maximize the main transmission rate while ensuring the minimum signal-to-noise ratio of the secondary received signal. Since the optimizations of the K time slots are independent, the rate of each time slot can be optimized independently. Therefore, the optimization problem in the t-th time slot can be expressed as:

[0049]

[0050] C2:||w (t) ||=1

[0051] C3:P (t) ≤P max

[0052]

[0053] k∈{1,2,...,K},q∈{1,2,...,Q},t∈{1,2,...,K}

[0054] Where, γ min It is the minimum required signal-to-interference-plus-noise ratio (SIN / N) of the secondary symbol, P. max P IRS and These are the maximum values ​​of the base station transmit power, the active intelligent reflector amplification power, and the reflection coefficient amplitude, respectively; C1 constrains the minimum signal-to-noise ratio of the secondary signal, C2 constrains the active beamforming vector length to 1, C3 is the maximum transmit power constraint of the base station, C4 is the maximum amplification power constraint of the active intelligent reflector, and C5 is the reflection coefficient amplitude constraint of the active intelligent reflector.

[0055] The optimization problem P0 contains multiple optimization variables and is non-convex, making direct solution difficult. Therefore, we consider decomposing the original problem into a transmission power P. (t) Active beamforming vector w (t) Reflection coefficient matrix and The four subproblems are optimized, and then the subproblems are transformed using the Successive Convex Approximation (SCA) and Semidefinite Relaxation (SDR) methods. Finally, the solution to the original problem is obtained by iteratively solving the problem.

[0056] (1) In the active beamforming vector w (t) and the reflection coefficient matrix of the smart reflective surface When fixed, optimize the transmission power P (t) To maximize the system's main transmission rate, i.e.

[0057]

[0058] stC1,C3,C4

[0059] The objective function in the above equation includes the optimization variable P. (t) Given the fractional form of the objective function, consider approximating it using its first-order Taylor expansion, and then obtain a solution that approximates the original problem through SCA. The first-order Taylor expansion of the objective function is:

[0060]

[0061] In the above formula, Let be the point of Taylor series expansion, use Replace, get

[0062]

[0063] stC1,C3,C4

[0064] The optimization problem described above is a standard convex optimization problem with a unique optimal solution. Therefore, it can be solved using the interior point method or the convex optimization toolbox. yes The approximation, the accuracy of the approximation, and the expansion point. Regarding this, to improve accuracy, when using SCA iterative optimization, given... After initializing the values, P is obtained by solving an approximate convex problem. (t) The suboptimal solution is used as the first-order Taylor expansion point of the objective function in the next iteration. This process is repeated cyclically to gradually approach the optimal solution until the approximate solution meets the required tolerance value.

[0065] (2) At the transmission power P (t) and the reflection coefficient matrix of the smart reflective surface When fixed, optimize the active beamforming vector w (t) To maximize the system's main transmission rate, i.e.

[0066]

[0067] stC1,C2,C4

[0068] Because the objective function in the above formula Includes variable w (t) The fractional and quadratic terms are in form, and the subproblem is nonconvex. Define variables. W (t) =w (t) (w (t) ) H , Where k∈{1,2,…,K}. The optimization variable w in the above equation... (t) Using matrix W (t) Replacement, thereby optimizing the active beamforming vector w (t) The equivalence problem of the subproblems can be restated as follows:

[0069]

[0070] C2.2:Tr(W (t) ) = 1

[0071]

[0072] k∈{1,2,…,K}

[0073] C2.4:rank(W (t) ) = 1

[0074]

[0075] The problem above is still a non-convex problem, which can be approximately transformed into a convex problem through a first-order Taylor expansion. The first-order Taylor expansion of the objective function in the above equation is:

[0076]

[0077] In the above formula, Let be the Taylor series expansion point of the objective function in the equivalent problem of subproblem P2. The objective function of the equivalent problem of subproblem P2 at the Taylor expansion point The value at that location, Therefore, the equivalent problem of subproblem P2 can be transformed into the following approximate form:

[0078]

[0079] stC2.1,C2.2,C2.3,C2.4,C2.5

[0080] If W is removed (t) With rank-1 constraints, the optimization problem in the above equation is a standard convex optimization problem with a unique optimal solution, and therefore can be solved using the interior-point method or the convex optimization toolbox; due to the relaxation of W... (t) Given the rank-1 constraint, the solution (W) (t) )opt The rank may not be 1, if (W (t) ) opt The rank is 1 for (W) (t) ) opt The eigenvectors obtained after eigenvalue decomposition are W. (t) The solution (W) (t) ) opt If (W) (t) ) opt If the rank is not 1, first check (W) (t) ) opt Perform eigenvalue decomposition to obtain (W) (t) ) opt =UΛU H ,in It is a unitary matrix. Given a diagonal matrix composed of its eigenvalues, randomly generate multiple N-dimensional complex Gaussian random vectors r, each with a mean of 0 and a variance of 1. By w (t) =UΛ 1 / 2 r obtains multiple w that satisfy the conditions. (t) Substituting the solution into subproblem P2 of the design and optimization method for a symbiotic radio system based on multiple active intelligent reflective surfaces as described in claim 8, we can maximize w on the left side of constraint C1. (t) As an approximate solution (w) (t) ) opt For the obtained suboptimal solution, SCA iterative loop is used. In each iteration, the solution obtained from the previous iteration is used as the expansion point. This process is repeated until the optimal solution is gradually approached, until the required tolerance value (3) is met at the transmission power P. (t) and active beamforming vector w (t) When fixed, optimize the reflection coefficient matrix of the smart reflective surface for transmitting secondary symbols. To maximize the system's main transmission rate, i.e.

[0081]

[0082] C1,C5

[0083] The objective function of subproblem P3 above contains the optimization variable w. (t) The problem is nonconvex in both fractional and quadratic form. This can be addressed by introducing auxiliary variables. (The symbol diagvec indicates taking the diagonal elements of the diagonal matrix to form a column vector) The equivalent problem of subproblem P3 can be represented as follows:

[0084]

[0085] C3.4:rank(V t(t) ) = 1

[0086]

[0087] in, M is a diagonal matrix formed by the vectors of the back-reflection link channel coefficients. The symbol diag indicates that the elements of the vectors are used as diagonal elements to form a diagonal matrix. t =G t G t H ,

[0088] and The problem above is still a non-convex problem, which can be approximately transformed into a convex problem through a first-order Taylor expansion. The first-order Taylor expansion of the objective function is:

[0089]

[0090] In the formula Let be the Taylor series expansion point of the objective function of the equivalent problem of the subproblem P3. For the objective function at the expansion point The value at that location, Therefore, the equivalent problem of subproblem P3 can be approximately transformed into the following form

[0091]

[0092] stC3.1,C3.2,C3.3,C3.4,C3.5

[0093] Remove the V t (t) After the rank-1 constraint, the above equation becomes a convex problem, which can be solved using tools such as CVX; if the optimal solution (V) is obtained... t (t) ) opt Satisfying the rank-1 constraint, for (V t (t) ) opt The eigenvectors obtained after eigenvalue decomposition are v t (t) The solution (v) t (t) ) opt , by (v t (t) ) opt The diagonal matrix formed by the elements is the optimal reflection coefficient matrix. If the optimal solution (V) is obtained t (t) ) optIf the rank-1 constraint is not satisfied, a similar method is used to obtain the optimal (v) t (t) ) opt Then by (v t (t) ) opt The optimal reflection coefficient matrix is ​​obtained by constructing a diagonal matrix from the elements.

[0094] (4) At the transmission power P (t) and active beamforming vector w (t) When fixed, optimize the reflection coefficient matrix of a smart reflective surface that does not send secondary symbols. To maximize the system's main transmission rate, i.e.

[0095]

[0096] C1,C5

[0097] The objective function of subproblem P4 above contains the optimization variable w. (t) The problem is nonconvex in both fractional and quadratic form. This can be addressed by introducing auxiliary variables. The equivalent problem of obtaining subproblem P4 can be represented as:

[0098]

[0099] in, The symbol blkdiag represents matrix G. k Add 0 to the diagonal element;

[0100] The problem above is still a non-convex problem, which can be approximately transformed into a convex problem through a first-order Taylor expansion. The first-order Taylor expansion of the objective function is:

[0101]

[0102] In the above formula Let represent the Taylor series expansion point of the objective function of the equivalent problem of subproblem P4. The objective function at the reference point The value at that location, The equivalent problem of subproblem P4 can be transformed into

[0103]

[0104] stC4.1,C4.2,C4.3,C4.4,C4.5

[0105] Remove The rank-1 constraint above is a convex problem, which can be solved using tools such as CVX. If the optimal solution is obtained... Satisfying the rank-1 constraint, for The eigenvectors obtained after eigenvalue decomposition are Solution If the optimal solution (V) is obtained k (t) ) opt If the rank-1 constraint is not satisfied, a similar method is used to obtain the optimal value. To obtain the optimal Then, we obtain the following formula:

[0106]

[0107] in and They represent and The i-th element, arg(x), represents the phase angle of the complex number x, yielding... Then The elements are used as diagonal elements to obtain the reflection coefficient matrix.

[0108] This invention addresses the optimization problem of maximizing the average main transmission rate (MPR) under different time slots t. It jointly optimizes the base station transmit power, beamforming vector, and IRS reflection coefficient matrix under each time slot to maximize the MPR. After obtaining the maximum MPR for each time slot, the maximum average MPR of the system is then calculated. The algorithm for solving the optimization problem P0 under different time slots t is shown in Algorithm 1. In the symbols, the superscript {j} represents the value after the j-th iteration optimization, and the superscript 0 represents the initial value. At the beginning of the algorithm, initialization is performed... First, let's define the Taylor expansion point. Initialization settings Then, the subproblems are solved to obtain suboptimal solutions; this is the first round of iteration. In the j-th iteration, the Taylor expansion point is... Set to the value obtained from the (j-1)th iteration optimization Solving the subproblems separately yields the results. The main transmission rate was then calculated based on the results of this iterative optimization. In the alternating iterative optimization in Algorithm 1, the value of the main transmission rate remains monotonically increasing. The algorithm converges when the change in the transmission rate obtained from the previous two iterations is less than a preset threshold.

[0109] Algorithm 1: Algorithm for Alternating Iterative Solution of Optimization Problem P0

[0110] (1) Initialize parameters: j = 0, k = 0, ε,

[0111] (2)while;

[0112] (3)j=j+1.

[0113] (4) Let w (t) =(w (t) ) {j-1} , Solving the equivalent problem of subproblem P1 yields (P (t) ) (j) .

[0114] (5) Let P (t) =(P (t) ) {j} , Solving the equivalent problem of subproblem P2 yields (W) (t) ) {j} .

[0115] (6) For (W) (t) ) {j} Eigenvalue decomposition yields (w) (t) ) {j} .

[0116] (7) Let P (t) =(P (t) ) {j} , w (t) =(w (t) ) {j} Solving the equivalent problem of subproblem P3 yields V. t (t) .

[0117] (8) For V t (t) By performing eigenvalue decomposition and transforming it into a diagonal matrix, we can obtain...

[0118] (9) Let P (t) =(P (t) ) {j} w (t) =(w (t) ) {j} , Solving the equivalent problem of subproblem P4 yields

[0119] (10) Eigenvalue decomposition yields Then solve for Finally, after converting it into a diagonal matrix, we can obtain...

[0120] (11) Update the objective function value

[0121] (12)

[0122] (13) Output

[0123] The invention will now be described in further detail with reference to the accompanying drawings. Unless otherwise specified, the number of antennas at the base station N = 3, the number of reflection elements in the IRS Q = 40, the number of IRSs K = 3, the minimum signal-to-interference-plus-noise ratio (SNR) requirement for the secondary symbol is 5 dB, the maximum transmit power of the base station is 30 dBm, and the maximum amplification power consumption of the active IRS is 20 dBm. The location coordinates of the base station are (0,0,30), the location coordinates of the user are (0,40,0), and the location coordinates of the three IRSs are (5,20,15), (10,20,15), and (15,20,15), respectively, in meters (m). This invention assumes that the channel between nodes in the system is a Ricean fading channel, and the channel fading includes path loss (large-scale fading) and small-scale fading. The path loss model is as follows: d0 = 1m is the reference distance, d is the distance between system nodes, Γ0 = -30dB is the path loss at the reference distance d0 = 1m, and κ is the path loss factor. BR,k d represents the distance from the base station to the k-th IRS. BU d represents the distance from the base station to the user. RU,k This represents the distance from the k-th IRS to the user. In the simulation, the path loss factor from the base station to the user is set to 3.5, and the path loss factors from the base station to the IRS and from the IRS to the user are set to 2.5. The small-scale fading factor is a random variable following a complex Gaussian distribution with a mean of 0 and a unit variance. The noise power is... IRS control circuit power consumption and bias circuit power consumption P c =P dc = -3dBm, active IRS power amplification efficiency η = 0.95, maximum reflection coefficient The convergence factor ε in Algorithm 1 is 10. -3 .

[0124] The simulation simultaneously presents three comparative schemes for performance comparison analysis with the present invention. Comparison Scheme 1 – Passive IRS Scheme: Except that the IRS is passive, it is the same as the present invention. The amplitude of the IRS reflection coefficient is less than or equal to 1, and the power consumption of the IRS control circuit is (P c +P dc Q, the maximum transmission power of the base station is KP IRS +P maxThe total maximum power of the system is the same as that of the proposed solution. Comparison Scheme 2 – Active IRS Random Phase Shift Scheme: The amplitude of each reflection element of the active IRS is set to... The phase shift is randomly selected within the range of [-π, +π], and the variable transmission power P is optimized using the same alternating iterative algorithm as in this invention. (t) and beamforming vector w (t) Comparison with Scheme 3 – Single-Antenna BD Scheme: In this scheme, the system uses a single-antenna BD with a reflection coefficient amplitude less than or equal to 1. When each BD takes turns reflecting the signal transmitted by the base station to transmit secondary information, inactive BDs do not reflect. The maximum transmission power of the base station is set to KP. IRS +P max +(P c +P dc Q, using the same algorithm as this invention to optimize P (t) w (t) And the reflection coefficient of BD. Multiple sets of channel samples are randomly generated. The optimization problem is solved under each set of channel samples to obtain the main transmission rate value. Finally, the average transmission rate is obtained by averaging the main transmission rates under all channel samples.

[0125] Figure 2 The convergence process of the alternating iterative algorithm of the present invention under different maximum base station transmit power and different numbers of IRS is presented. As can be seen from the figure, under multiple different samples, the main transmission rate of the proposed algorithm gradually increases with the number of iterations, and converges after 5 iterations, indicating that the algorithm proposed in this invention has good convergence.

[0126] Figure 3The figure shows the relationship between the average main transmission rate and the maximum transmit power of the base station for each scheme. The results show that the main transmission rate of all schemes increases with the increase of the maximum transmit power of the base station. The main transmission rate of the scheme proposed in this invention is significantly higher than that of the passive IRS scheme and the single-antenna BD scheme. This is because the active IRS used in this invention can enhance the signal strength of the reflected link. Compared with the passive IRS and single-antenna BD schemes, the IRS, which does not transmit secondary information, can provide higher gain for the main transmission, thus achieving a higher main transmission rate. Simultaneously, the average main transmission rate of the scheme proposed in this invention is also higher than that of the random phase scheme, indicating that the optimization of the IRS reflection coefficient phase in the proposed scheme can also improve the main transmission rate. Simulation results also show that the passive IRS scheme, utilizing the IRS as a reflection device, provides multipath gain for the main transmission, so its average main transmission rate is slightly higher than that of the single-antenna BD scheme. However, due to multiplicative attenuation, the signal strength of the reflected link is relatively small, resulting in a smaller difference in the average main transmission rate. Furthermore, although the average main transmission rate of the random phase scheme is lower than that of the scheme of this invention, it is still significantly higher than that of the passive IRS scheme and the single-antenna BD scheme. This shows that the scheme of this invention using active IRS has a significant advantage in improving the main transmission rate.

[0127] Figure 4 The figure shows the changes in the average main transmission rate of the proposed scheme and various comparative schemes as the number of base station antennas N increases. As can be seen from the figure, the average main transmission rate of all schemes increases with the increase of the number of antennas N. Comparing the curves of different schemes, it can be seen that, with the increase of the number of antennas N, the proposed scheme and the phase random scheme show a greater trend in the increase of the average main transmission rate compared to other schemes. The passive IRS and single-antenna BD schemes show a relatively slower growth trend with the increase of the number of antennas, and gradually tend to stabilize. This is because the active IRS used in the proposed scheme results in a higher signal strength in the reflected link. Under the same conditions, the main transmission rate changes more significantly with the increase of the number of antennas, and therefore the growth rate is relatively faster.

[0128] Figure 5 The average primary transmission rate is given as a function of the secondary symbol signal-to-interference-plus-noise ratio threshold γ. min The graph shows the variation of the secondary symbol signal-to-interference-plus-noise ratio (SIN / NDR) threshold γ. min As γ gradually increases, the average main transmission rate of all schemes decreases. This is because when γ... min When the signal is increased, the BD transmitting the secondary symbol requires a higher reflection coefficient amplitude to meet the secondary signal-to-interference-plus-noise ratio (SINR) requirement. However, the signal reflected by the BD in parasitic mode is interference for the main transmission, and increased interference leads to a decrease in the main transmission rate. Comparing the curves of different schemes shows that, compared with passive IRS and single-antenna BD schemes, the main transmission rate of the scheme in this invention increases with the SINR threshold γ. minThe decreasing trend is more gradual, and the decreasing trends of the main transmission rate are roughly the same for passive IRS and single-antenna BD schemes. This indicates that the solution of this invention can be applied to scenarios with higher signal-to-noise ratio requirements. The solution of this invention can meet higher signal-to-interference-plus-noise ratio requirements by increasing the amplitude of the reflection coefficient within the allowable range of active IRS amplification power consumption.

[0129] Figure 6 The diagram shows the variation of the main transmission rate with the number of IRS, K. ​​It can be seen that the main transmission rate of the proposed scheme and the random phase scheme increases rapidly with the increase of IRS, while the main transmission rate of the passive IRS scheme also increases, but at a slower rate. As the number of IRS increases, the reflection link gain of the main transmission signal increases. The scheme using active IRS in this invention amplifies the reflected signal, resulting in a faster increase in the reflection link signal strength and a correspondingly faster increase in the main transmission rate compared to the passive IRS scheme. In contrast, the main transmission rate remains unchanged when the number of BDs in the single-antenna BD scheme is increased. In the single-antenna BD scheme, only one BD reflects the signal at a time; the other BDs do not reflect the signal. Therefore, increasing the number of BDs does not lead to a change in the main transmission performance.

[0130] Figure 7 The variation of the average main transmission rate with increasing number of reflection units Q per IRS is presented. The main transmission rate of all IRS schemes increases with increasing number of reflection units, with the scheme of this invention and the phase-random active IRS scheme showing the highest increase rate, while the passive IRS scheme shows a relatively lower increase rate. This is because, as Q increases, although IRS interference in the secondary symbol increases, the gain provided by reflections from other IRSs also increases. In the passive IRS scheme, although interference in the reflection link where BD is located increases with increasing Q, the gain of reflection links in other IRSs also increases, so the average main transmission rate also increases with increasing Q. However, the multiplicative attenuation of the reflection links has a relatively small impact on the main transmission rate increase.

Claims

1. A design and optimization method for a symbiotic radio system based on multiple active intelligent reflective surfaces, characterized in that: Includes the following steps: (1) System design scheme: a symbiotic radio system with multiple active smart reflective surfaces, in which the smart reflective surfaces are both devices that improve the main transmission performance and time-division scattering devices. The user is simultaneously the receiver of the main and secondary information. Within a transmission frame, each smart reflective surface takes turns as a scattering device, transmitting secondary information in a parasitic manner by scattering and modulating the incident signal. When not acting as a reflective device, it reflects the main signal to enhance the performance of the main transmission. The main symbol rate is the same as the secondary symbol rate. (2) Construct an optimization mathematical model with the main transmission rate as the optimization objective, the secondary symbol minimum signal-to-interference-plus-noise ratio, and the base station transmit power, active beamforming and smart reflective surface reflection coefficient under the constraints of base station and smart reflective surface amplification power; (3) The optimization problem is decomposed into four sub-problems: optimizing the base station transmission power, base station active beamforming, the reflection coefficient of the intelligent reflective surface as a backscattering device, and the reflection coefficient of other intelligent reflective surfaces. Then, each non-convex problem is transformed into a convex problem. (4) Solve each optimization subproblem by alternating iterations to obtain an approximate solution to the original optimization problem.

2. The design and optimization method for a symbiotic radio system based on multiple active intelligent reflective surfaces according to claim 1, characterized in that: In the mathematical model of the optimization problem described in step (2), the signal-to-interference-plus-noise ratio (SINR) of the decoded main signal in the t-th time slot is: Among them, w (t) For the active beamforming vector of the base station, Let P be the reflection coefficient matrix of the k-th smart reflective surface. (t) For base station transmission power, Direct link channel coefficient vector from base station to user Let be the channel coefficient matrix from the base station to the k-th smart reflective surface. The channel coefficient vector from the k-th IRS to the user, where K is the number of smart reflectors, Q is the number of reflectors per smart reflector, and N is the number of base station antennas. User-side Gaussian white noise variance Let || be the thermal noise power of the intelligent reflective surface, and || denote the Euclidean norm of the vector; the main information transmission rate is... The average signal-to-interference-plus-noise ratio of the secondary signal is The optimization problem is C2:||in (t) ||=1 C3:P (t) ≤P max k∈{1,2,...,K},q∈{1,2,...,Q} t∈{1,2,...,K} Where, γ min It is the minimum required signal-to-interference-plus-noise ratio (SIN / N) of the secondary symbol, P. max P IRS and These are the maximum values ​​of the base station transmit power, the active intelligent reflective surface amplification power, and the reflection coefficient amplitude, respectively. Let be the amplification power of the k-th active IRS.

3. The design and optimization method for a symbiotic radio system based on multiple active intelligent reflective surfaces according to claim 1, characterized in that: Step (3) involves decomposing the optimization problem into four sub-problems and transforming each sub-problem into a convex sub-problem, specifically including: (1) Transmit power P (t) Optimization subproblems: stC1,C3,C4 The above problem is non-convex; by using continuous convex approximation, the atomic problem P1 can be transformed into... stC1,C3,C4 in: In the above formula: For the objective function The first-order Taylor expansion point; the optimization problem P1 is a standard convex optimization problem with a unique optimal solution, and therefore can be solved by the interior point method or the convex optimization toolbox. (2) Active beamforming vector w (t) Optimization subproblems: stC1,C2,C4 Subproblem P2 is nonconvex; construct optimization variable W. (t) =w (t) (w (t) ) H It can be transformed into continuous convex approximation and semidefinite relaxation. C2.2:Tr(W (t) )=1 C2.4:rank(W (t) )=1 In the above formula, symbol Denotes the Hadamard product of matrices, here For Υ1(W (t) The first-order Taylor expansion point, (3) Reflection coefficient matrix of the intelligent reflective surface as a backscattering device Optimization subproblems: s.t.P t (t) ≤P IRS C1,C5 Subproblem P3 is nonconvex; construct optimization variables. definition M is a diagonal matrix formed by the vectors of the back-reflection link channel coefficients. The symbol diag indicates that the elements of the vectors are used as diagonal elements to form a diagonal matrix. t =G t G t H , Π t (t) =diag(H t w (t) )×(diag(H t w (t) )) H ,and Using continuous convex approximation and positive semidefinite relaxation, the atomic problem P3 can be transformed into the following convex problem: C3.4:rank(V t (t) )=1 In the above formula: in, For Υ2(V t (t) The first-order Taylor series expansion point, (4) Other intelligent reflective surface reflection coefficient matrices The optimization subproblem is C1,C5 Subproblem P4 is still nonconvex; define variables. The symbol diagvec represents taking the diagonal elements of a diagonal matrix to form a column vector. Using continuous convex approximation and positive semidefinite relaxation, the atomic problem P4 can be transformed into the following convex problem: In the above formula: express The first-order Taylor expansion point, The symbol blkdiag represents matrix G. k Add 0 to the diagonal element.

4. The design and optimization method for a symbiotic radio system based on multiple active intelligent reflective surfaces according to claim 1, characterized in that: The iterative solution algorithm for obtaining an approximate solution to the original optimization problem by iteratively solving each optimization subproblem in step (4) is as follows: (1) First, initialize the transmit power, beamforming vector, and reflection coefficient matrix as (P (t) ) {0} , (w (t) ) {0} , (2) In the j-th iteration, the beamforming vector and reflection coefficient matrix are fixed at w. (t) =(w (t) ) {j-1} and Solving the transformed subproblem P1 yields (P (t) ) (j) ; (3) With the base station's transmit power and reflection coefficient matrix fixed at P (t) =(P (t) ) {j} and Solving the transformed subproblem P2 yields (W) (t) ) {j} And obtained through eigenvalue decomposition (w) (t) ) {j} ; (4) The reflection coefficient matrix of the smart reflective surface with fixed transmission power, beamforming vector, and no secondary symbols is P. (t) =(P (t) ) {j} w (t) =(w (t) ) {j} and Solving the transformed subproblem P3 yields (V) t (t) ) {j} And obtain it by eigenvalue decomposition and then transforming it into a diagonal matrix. (5) The reflection coefficient matrix of the smart reflective surface with transmitted power, beamforming vector, and transmitted secondary symbols is fixed at P. (t) =(P (t) ) {j} w (t) =(w (t) ) {j} and Solving the transformed subproblem P4 yields the following results. And obtain by eigenvalue decomposition and then transforming into a diagonal matrix (6) Substitute the optimal transmission power, beamforming vector and reflection coefficient matrix obtained in this round of optimization into the main transmission rate expression to calculate the main transmission rate. If the relative difference between the transmission rate obtained in this round and the transmission rate obtained in the previous round is less than the preset threshold, the iteration converges and the optimization ends; otherwise, iterative optimization continues.

Images