A spectrum sharing method based on discrete beamforming of intelligent reflector

Abstract

To improve the efficiency of wireless spectrum sharing, the present invention provides a spectrum sharing method based on discrete beamforming of intelligent reflective surfaces. For secondary base stations, the main steps of this method are as follows: 1) obtaining channel state information; 2) establishing a joint optimization model; 3) optimizing the transmit power of the secondary base station; 4) optimizing the discrete beamforming vector of the intelligent reflective surface; 5) repeating steps 3) and 4) until convergence; 6) sharing the underlay spectrum of the secondary base station. In this method, the secondary base station continuously performs alternating iterative optimization of its transmit power and the discrete beamforming of the intelligent reflective surface until convergence. In the process of optimizing the discrete beamforming of the intelligent reflective surface, the quadratic fraction is converted into a linear expression to reduce the solution complexity, and then a binary search method is used for the solution. This optimization strategy can be widely applied to wireless communication scenarios assisted by intelligent reflective surfaces, not limited to the underlay spectrum sharing scenarios listed above.

Description

Technical Field

[0001] The present invention relates to the field of wireless spectrum sharing, and in particular to a method for suppressing interference between primary and secondary users in an underlay spectrum sharing system by using discrete beamforming of an intelligent reflective surface, thereby improving wireless spectrum utilization efficiency. Background Art

[0002] While promoting the mature application of 5G networks and the continuous evolution of 5G standards, academia and industry at home and abroad have also begun exploring the next generation of mobile wireless networks (6G). The prospect of 6G presents a grand vision of interconnecting everything on land, sea, air, and space, but it also faces the harsh reality of increasingly scarce communication resources, particularly spectrum. Faced with the mutually constraining requirements of 6G's higher speeds, lower energy consumption, and denser connectivity, simply extending existing 5G transmission technologies is no longer sustainable. Therefore, finding new solutions to overcome existing technological bottlenecks and further improve the efficiency of limited spectrum resources is a key topic in future wireless communication research.

[0003] Smart reflector technology has recently attracted considerable attention due to its unique wireless channel manipulation capabilities. Specifically, a smart reflector is an artificial composite electromagnetic material composed of a large number of subwavelength unit structures arranged periodically or aperiodically on a two-dimensional plane. Through reconfigurable electromagnetic parameters, the reflection characteristics of electromagnetic waves can be flexibly controlled to meet diverse application requirements, achieving a "smart reflection" effect. Smart reflectors provide an additional reflection path for user signal reception. Unlike other propagation paths in the environment, such as scattering and reflection, this reflection path is controllable. By dynamically adjusting the reflection coefficient of each reflector unit structure, the reflected signal from the smart reflector can be manipulated to combine in-phase or anti-phase with the remaining path signals, achieving signal enhancement or interference reduction. (Equivalently, the "smart reflection" properties of smart reflectors make the wireless propagation environment of electromagnetic signals intelligently controllable.) Spectrum sharing technology, on the other hand, has attracted extensive research due to its significant potential for improving spectrum utilization. Spectrum sharing methods include overlay and underlay. Underlay spectrum sharing allows unlicensed users to access the same spectrum simultaneously while meeting interference limits imposed by licensed users. This approach is both more aggressive and more efficient, maximizing spectrum sharing opportunities. However, in future wireless networks, denser node distribution will make interference mitigation more difficult, especially when adjacent nodes share spectrum due to the presence of strong interfering links. This significantly limits the future application of underlay spectrum sharing.

[0004] Fortunately, the channel manipulation capabilities of smart reflective surfaces offer a new approach to addressing the interference mitigation challenge in underlay spectrum sharing. Specifically, by manipulating reflected signals to eliminate interference, the constraints imposed by node location on the strength of interference channels can be effectively eliminated, significantly expanding the application scenarios of underlay spectrum sharing. Consequently, applying smart reflective surface technology to efficient spectrum sharing in next-generation wireless networks can effectively overcome the scalability, energy consumption, and cost constraints faced by existing transmission solutions, thereby resolving the fundamental conflict between the demand for higher performance metrics and the increasingly scarce spectrum resources in next-generation wireless networks.

[0005] Currently, a small number of papers have studied intelligent reflector-assisted underlay spectrum sharing systems, exploring the core issue of intelligent reflector beamforming. In [Intelligent reflecting surface-assisted cognitive radio system, IEEE Transactions on Communications, vol. 69, no. 1, pp. 675–687, Jan. 2021.] and [Intelligent reflecting surface aided MIMO cognitive radio systems, IEEE Transactions on Vehicular Technology, vol. 69, no. 10, pp. 11445–11457, Oct. 2020.], intelligent reflector beamforming was designed for multiple-input single-output (MISO) and multiple-input multiple-output (MIMO) underlay spectrum sharing systems, respectively, with the goal of maximizing rate while considering transmit power constraints and interference power constraints. The paper [Robust beamforming design for intelligent reflecting surface aided cognitive radio systems with imperfect cascaded CSI, IEEE Transactions on Cognitive Communications and Networking, vol. 8, no. 1, pp. 186-201, Mar. 2021.] further considers the robust beamforming problem when perfect channel state information is not available, and jointly optimizes the beamforming of the secondary user transmitter and the beamforming of the smart reflector using an alternating iterative method. The paper [Reconfigurable intelligent surface-enabled spectrum-sharing communications, IEEE Wireless Communications Letters, vol. 11, no. 1, Jan. 2022.] considers using continuous interference cancellation technology at the receiver end and jointly designs the receive decoding order and the smart reflector beamforming to minimize the transmit power.

[0006] It's important to note that all of the aforementioned references assume that the reflection coefficient of the smart reflector is continuously and arbitrarily adjustable. However, in real systems, this is difficult to achieve due to the hardware complexity and cost constraints of the smart reflector's control circuitry. Therefore, the results in these references can only serve as an upper bound on the performance of smart reflector-assisted underlay spectrum sharing. Discrete beamforming, which considers the discrete reflection coefficient of the smart reflector, is more practical. Summary of the Invention

[0007] The purpose of this invention is to design an underlay spectrum sharing method based on discrete beamforming of smart reflectors. Under the condition of discrete reflection coefficients of smart reflectors, discrete beamforming can effectively eliminate interference between primary and secondary users and improve spectrum sharing efficiency. The main steps of this method are as follows:

[0008] Step 1: Obtain channel state information: The primary user (PU) transmits a pilot signal, and the primary base station (PBS) receives the pilot signal and calculates the direct channel h between the primary user and the primary base station. pp and reflection channel h prp The secondary base station (SBS) estimates the direct channel h between the primary user and the secondary base station based on the received pilot signal. ps and reflection channel h prs The secondary user (SU) transmits a pilot signal, and the primary base station estimates the direct channel h between the secondary user and the primary base station based on the received pilot signal. sp and reflection channel h srp The secondary base station estimates the direct channel h between the secondary user and the secondary base station based on the received pilot signal. ss and reflection channel h srs Estimation is performed; the primary base station sends the estimated channel state information to the secondary base station.

[0009] Step 2: Establish a joint optimization model: The transmission power of the primary base station and the secondary base station is recorded as p p and p s (p s The maximum value is P max , that is, p s ≤P max ), the noise power of the primary user and the secondary user are respectively recorded as and The minimum signal to interference and noise ratio required by the primary user is denoted as γ th , the smart reflector beamforming vector is recorded as v H =[v1,...,v N ](in is the reflection coefficient of the nth reflection unit of the smart reflection surface, θ nis the phase of the reflection coefficient of the nth reflection unit of the smart reflector, n=1,...,N). Through the joint optimization of the secondary base station transmit power and the smart reflector beamforming vector, it is necessary to ensure that the primary user signal to interference and noise ratio is higher than γ th At the same time, the secondary user's signal to interference and noise ratio is maximized (equivalent to maximizing the secondary user rate). Therefore, the secondary base station obtains the joint optimization problem shown in formula (1):

[0010]

[0011] Where Δθ = 2π / 2 Q is the interval of discrete phase values ​​of the reflection coefficient of each reflection unit, and Q is the order of discrete phase values ​​of the reflection coefficient of each reflection unit. Since the joint optimization problem shown in formula (3) is difficult to solve, the following steps 3 and 4 are used to alternately optimize the secondary base station transmit power and the smart reflector beamforming (initialize the smart reflector beamforming vector to v H =[1,...,1]).

[0012] Step 3: Optimize the secondary base station transmit power: For a given smart reflector beamforming vector v H , the joint optimization problem shown in formula (3) degenerates into the secondary base station transmission power optimization problem shown in formula (4):

[0013]

[0014] The closed-form solution to the problem shown in formula (4) is:

[0015]

[0016] Here, min{a,b} represents the smaller of a and b.

[0017] Step 4: Optimize the discrete beamforming vector of the smart reflector: Substitute the secondary base station transmit power p obtained in step 3 into s Substitute into formula (3), introduce the intermediate variable t, convert the difficult fractional expression into a quadratic expression, and perform the following variable substitution: The optimization problem can be rewritten as:

[0018]

[0019] Let G s =H ss -tH ps , G p =H pp -tH sp , but

[0020]

[0021]

[0022] Through the above transformation, the quadratic expression in formula (6) is transformed into a linear expression, that is, the original integer nonlinear programming problem is transformed into an integer linear programming problem that is easy to solve. At the same time, observe f p and f s It can be seen that the optimization variable has changed from the phase of the reflection coefficient to the cosine and sine of the phase and the cosine and sine of the phase difference between any two reflection coefficients. Based on the phase periodicity, the vector Θ is defined as [0, Δθ, 2Δθ, ..., (2 M -1)θ], then θ i ,θ j -θ i are all elements in Θ. Then, the cosine function vector and sine vector are defined as follows:

[0023]

[0024] but j,cosθ i and cos(θ j -θ i ) are all ξ c An element in, sinθ i and sin(θ j -θ i ) are all ξ s An element in. Define SOS1 b Type (a binary sequence with only one element set to 1 and the rest set to 0) variable z i,j , w i and ρ i,j , then there are the following constraints:

[0025]

[0026] The above relationships are all linear equations. Then, the optimal discrete beamforming vector v that maximizes the objective function in formula (6) can be obtained by bisection search. H .

[0027] Step 5: Repeat steps 3 and 4 until the secondary user's signal to interference and noise ratio converges.

[0028] Step 6: Underlay spectrum sharing transmission of the secondary base station: The intelligent reflector sets the reflection coefficient of each reflector unit according to the optimized discrete beamforming vector. The secondary base station uses the optimized power p s Send a signal to the secondary user.

[0029] The present invention provides an underlay spectrum sharing method based on discrete beamforming of intelligent reflective surfaces, which utilizes the channel control capability of intelligent reflective surfaces to eliminate interference between primary and secondary users in the underlay spectrum sharing system, thereby improving the efficiency of wireless spectrum utilization. By jointly optimizing the secondary base station transmission power and the discrete beamforming of the intelligent reflective surface, the method can effectively suppress interference between primary and secondary users and improve spectrum sharing performance. In the process of optimizing the discrete beamforming of the intelligent reflective surface, the method converts the signal-to-interference-and-noise ratio from a quadratic fraction to a linear expression, converting the originally difficult-to-solve nonlinear integer programming problem into an integer linear programming problem that is easier to solve, thereby achieving efficient solution of the discrete reflection coefficient of the intelligent reflective surface. This optimization strategy can be widely applied to the design of discrete beamforming of intelligent reflective surfaces in different scenarios, not limited to the underlay spectrum sharing scenarios listed above. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flowchart for implementing a spectrum sharing method based on discrete beamforming of intelligent reflective surfaces.

[0031] Figure 2 It is an underlay spectrum sharing model assisted by intelligent reflective surfaces.

[0032] Figure 3 It is a flow chart of the binary search for discrete beamforming vectors of intelligent reflectors.

[0033] Figure 4 This is a schematic diagram of how the user rate changes with the maximum transmit power of the secondary base station under different spectrum sharing methods. DETAILED DESCRIPTION

[0034] According to the attached Figure 1 A flowchart of an implementation method for underlay spectrum sharing based on smart reflector discrete beamforming is provided. The implementation steps of underlay spectrum sharing proposed in the present invention mainly include:

[0035] Step 1: Obtain channel status information: Figure 2 The proposed smart reflector-assisted underlay spectrum sharing model consists of a primary base station (PBS), a primary user (PU), a secondary base station (SBS), a secondary user (SU), and a smart reflector (IRS). The smart reflector has N reflective units. The four reflection channels, PBS-smart reflector-PU, PBS-smart reflector-SU, SB-smart reflector-SU, SB-smart reflector-PU, and SB-smart reflector-SU, are denoted as h. prp =h pr diag(h rp ),h prs =hpr diag(h rs ),h srp =h sr diag(h rp ) and h srs =h sr diag(h rs First, the primary user transmits a pilot signal, and the primary base station modulates the direct channel h between the primary user and the primary base station based on the received pilot signal. pp and reflection channel h prp The secondary base station estimates the direct channel h between the primary user and the secondary base station based on the received pilot signal. ps and reflection channel h prs Then, the secondary user transmits a pilot signal, and the primary base station estimates the direct channel h between the secondary user and the primary base station based on the received pilot signal. sp and reflection channel h srp The secondary base station estimates the direct channel h between the secondary user and the secondary base station based on the received pilot signal. ss and reflection channel h srs Finally, the primary base station sends the estimated channel state information to the secondary base station.

[0036] Step 2: Establish a joint optimization model: The transmission power of the primary base station and the secondary base station is recorded as p p and p s (p s The maximum value is P max , that is, p s ≤P max ), the noise power of the primary user and the secondary user are respectively recorded as and The minimum signal to interference and noise ratio required by the primary user is denoted as γ th , the smart reflector beamforming vector is recorded as v H =[v1,...,v N ](in is the reflection coefficient of the nth reflection unit of the smart reflection surface, θ n is the phase of the reflection coefficient of the nth reflection unit of the smart reflector, n=1,...,N). Through the joint optimization of the secondary base station transmit power and the smart reflector beamforming vector, it is necessary to ensure that the primary user signal to interference and noise ratio is higher than γ th At the same time, the secondary user's signal to interference and noise ratio is maximized (equivalent to maximizing the secondary user rate). Therefore, the secondary base station obtains the joint optimization problem shown in formula (3):

[0037]

[0038] Among them, and The Gaussian white noise power of the primary user and the secondary user is Δθ=2π / 2 Q is the interval of discrete phase values ​​of the reflection coefficient of each reflection unit, and Q is the order of discrete phase values ​​of the reflection coefficient of each reflection unit. Since the joint optimization problem shown in formula (3) is difficult to solve, the following steps 3 and 4 are used to alternately optimize the secondary base station transmit power and the smart reflector beamforming (initialize the smart reflector beamforming vector to v H =[1,...,1]).

[0039] Step 3: Optimize the secondary base station transmit power: For a given smart reflector beamforming vector v H , the joint optimization problem shown in formula (3) degenerates into the secondary base station transmission power optimization problem shown in formula (4):

[0040]

[0041] The closed-form solution to the problem shown in formula (4) is:

[0042]

[0043] Here, min{a,b} represents the smaller of a and b.

[0044] Step 4: Optimize the discrete beamforming vector of the smart reflector: Substitute the secondary base station transmit power p obtained in step 3 into s Substitute into formula (3), introduce the intermediate variable t, convert the difficult fractional expression into a quadratic expression, and perform the following variable substitution: The optimization problem of formula (3) can be rewritten as:

[0045]

[0046] Let G s =H ss -tH ps , G p =H pp -tH sp ,but

[0047]

[0048]

[0049] Through the above transformation, the quadratic expression in formula (6) is transformed into a linear expression, that is, the original integer nonlinear programming problem is transformed into an integer linear programming problem that is easy to solve. At the same time, observe f p and f sIt can be seen that the optimization variable has changed from the phase of the reflection coefficient to the cosine and sine of the phase and the cosine and sine of the phase difference between any two reflection coefficients. Based on the phase periodicity, the vector Θ is defined as [0, Δθ, 2Δθ, ..., (2 M -1)θ], then θ i ,θ j -θ i are all elements in Θ. Then, the cosine function vector and sine vector are defined as follows:

[0050]

[0051] but j,cosθ i and cos(θ j -θ i ) are all ξ c An element in, sinθ i and sin(θ j -θ i ) are all ξ s An element in. Define SOS1 b Type (a binary sequence with only one element set to 1 and the rest set to 0) variable z i,j , w i and ρ i,j , then there are the following constraints:

[0052]

[0053] The above relations are all linear equations. So far, the quadratic inequality constraint in formula (12) has been converted into a linear inequality constraint. Figure 3 The given binary search method obtains the optimal discrete beamforming vector v that maximizes the objective function t in formula (4) H , the specific implementation steps include:

[0054] Determine the lower bound t of the objective function t in formula (4) low and upper bound t up : The secondary user signal to interference and noise ratio is non-negative, so t≥0, that is, the lower bound of t is t low =0; When the primary base station transmit power is 0, the secondary user is minimally interfered with by the primary base station, and the secondary user's signal-to-interference-and-noise ratio can be maximized. That is, the upper bound of t is

[0055] Step 11: Fixing the Smart Reflector Discrete Beamforming v H , secondary base station transmission power p s The closed-form solution is

[0056]

[0057] Among them, min{a,b} represents the smaller of a and b, p p The main base station transmit power, P max is the maximum transmission power of the secondary base station, h pp and h prp are the direct channel and the reflected channel between the primary user and the primary base station, respectively, h sp and h srp They are the direct channel and the reflected channel between the secondary user and the primary base station respectively. The secondary base station receives the pilot signal. is the noise power of the primary user, γ th The minimum signal to interference and noise ratio required by the primary user, v H is the beamforming vector of the smart reflector.

[0058] Step 12: Fix the secondary base station transmit power p s , optimize the smart reflector discrete beamforming by binary search v H .

[0059] The specific process is:

[0060] Step 121: Determine the lower bound t of the secondary user signal to interference and noise ratio t low and upper bound t up : The secondary user signal to interference and noise ratio is non-negative, so t≥0, that is, the lower bound of t is t low =0; When the primary base station transmit power is 0, the secondary user is minimally interfered with by the primary base station, and the secondary user's signal to interference and noise ratio is maximized. That is, the upper bound of t is

[0061] Step 122: Let t = (t low +t up ) / 2, to test whether the optimization problem shown in formula (2) has a solution:

[0062]

[0063] In order to simplify the expression, the matrix expression G is defined as s =p s [h srs ,h ss ][h srs ,h ss ] T -tp p [h prs ,h ps ][h prs ,h ps ] T and G p =pp [h prp ,h pp ][h prp ,h pp ] T -tp s [h srp ,h sp ][h srp ,h sp ] T , and the vector expression ξ c =[1,cosΔθ,cos(2Δθ)...,cos((2 Q -1)Δθ)]、ξ s =[1,sinΔθ,sin(2Δθ)...,sin((2 Q -1)Δθ)] and Θ=[0,Δθ,2Δθ,...,(2 M -1)θ](where Δθ=2π / 2 Q is the interval of discrete phase values ​​of reflection coefficient of each reflection unit, Q is the order of discrete phase values ​​of reflection coefficient of each reflection unit); i and j represent the subscripts of the corresponding matrix or vector; G s (i,j) and G p (i,j) represent the matrix expression G s and G p The i-th row and j-th column element of G s (i,i) and G p (i,i) represent the matrix expression G s and G p The i-th row and i-th column element of G s (i,N+1) and G p (i, N+1) represent the matrix expression G s and G p The element in row i and column N+1; z i,j and ρ i,j Represents the SOS1 corresponding to the i-th row and j-th column of the three-dimensional matrix variables z and ρ respectively b Type variable (a binary sequence with only one element being 1), w i Indicates SOS1 corresponding to the i-th row of the two-dimensional matrix variable w b Type variables.

[0064] Step 123: If the problem shown in formula (2) has a solution, let t low =t; otherwise, let t up =t;

[0065] Step 124: Repeat steps 122 and 123 until t up =t low ≤10＝3 , get the optimal discrete beamforming vector v H .

[0066] Step 4.3: If the problem shown in formula (2) has a solution, let t low -t; otherwise, let t up =t.

[0067] Step 4.4: Repeat steps 4.2 and 4.3 until t up -t low ≤10 -3 , we get the optimal discrete beamforming vector v that maximizes the objective function t in formula (10) H .

[0068] Step 5: Repeat steps 3 and 4 until the secondary user's signal to interference and noise ratio converges.

[0069] Step 6: Underlay spectrum sharing transmission of the secondary base station: The intelligent reflector sets the reflection coefficient of each reflector unit according to the optimized discrete beamforming vector. The secondary base station uses the optimized power p s Send a signal to the secondary user.

[0070] According to the attached Figure 1 The implementation process of the present invention is further explained by taking a smart reflector with 40 reflector units as an example. The parameter settings are as follows: carrier frequency 750MHz, carrier wavelength 0.4m, smart reflector unit spacing 0.15m, number of smart reflector units N=30, arranged in 3 rows and 10 columns, Gaussian white noise power of primary user and secondary user receivers The three-dimensional coordinates of the primary base station, secondary base station, primary user, and secondary user are (50,0,0), (50,200,0), (1,98,0), and (1,102,0), respectively. The smart reflector is deployed on the yoz plane, and the three-dimensional coordinates of its center are (0,100,2). The channel between any two points (including the base station, user, and smart reflector) is represented by h ab =L0d ab -c g ab , where L0 = 30dB is the path loss at the reference distance (1m), d ab is the distance between two points, g ab For the channel with small scale fading, g ab It obeys a complex Gaussian distribution with a mean of 0 and a variance of 1, c = 3; for the smart reflector-primary user / secondary user channel, g abContains only line-of-sight (LoS) channel components, c = 2. Using the method provided by the present invention (assuming that the reflection coefficient phase discrete value is 0 or π, attached Figure 4 The attached Figure 4 The secondary user rate varies with the maximum transmit power p of the secondary base station. max In order to verify the effectiveness of the proposed method, this section gives two other Underlay spectrum sharing methods: the smart reflector continuous phase beamforming method (the reflection coefficient of the smart reflector is continuously adjustable, which can provide higher beamforming gain, and Figure 4 The method of spectrum sharing without intelligent reflector (without deploying intelligent reflector to assist spectrum sharing, the interference to primary users can only be suppressed by controlling the transmit power of secondary base stations, with the addition of Figure 4 (Note: The following table is marked as “without smart reflective surface”). Figure 4 The transmitting power of the main base station is p p =20dBm, the minimum signal-to-interference-and-noise ratio requirement for the primary user is γ th =10dB, we can observe that: 1) Due to the limited phase control accuracy under the discrete reflection coefficient condition, the performance of the method provided by the present invention is lower than the ideal "continuous beamforming" method, but the method of the present invention can still achieve much better performance than the "no smart reflector" method, so that the secondary user can achieve a higher transmission rate, which shows that deploying smart reflectors (whether continuous beamforming or discrete beamforming) in the underlay spectrum sharing system can effectively improve the spectrum sharing efficiency; 2) In the method of the present invention, the secondary user rate can be adjusted with the maximum transmit power p of the secondary base station. max In the “no smart reflector” method, the secondary user rate increases with the maximum transmit power p of the secondary base station. max When it increases to a certain level, saturation occurs and the rate no longer increases. The reason is that in the "no intelligent reflector" method, the transmission of the secondary base station will always cause interference to the primary user, so its actual transmission power is limited by the minimum signal to interference and noise ratio of the primary user. Even if the secondary base station has more power available, it must control the actual transmission power within a certain range. Therefore, the secondary user rate does not change with the p max On the contrary, in the method of the present invention, since the reflection channel provided by the intelligent reflection surface can effectively offset the interference of the secondary base station to the primary user, the secondary user can increase with p max Increase and continuously improve the transmit power, thereby continuously improving the secondary user rate. In summary, the method of the present invention can obtain the channel control gain of the smart reflector through the discrete beamforming design of the smart reflector, effectively improving the spectrum sharing efficiency, and can also greatly reduce the hardware complexity and hardware cost of the smart reflector, making it more in line with actual system requirements.

Claims

1. A spectrum sharing method based on smart reflector discrete beamforming, characterized by: The main steps of implementing the method include: Step 1: Obtain channel state information; Step 2: Establish an optimization model; Step 3: Optimize the secondary base station transmit power: For a given smart reflector beamforming vector v H , get the secondary base station transmission power p s The closed-form solution of Step 4: Optimize the discrete beamforming vector of the smart reflector: Based on the secondary base station transmit power p obtained in step 3 s , the optimal discrete beamforming vector v is obtained by binary search H ; Step 5: Repeat steps 3 and 4 until the secondary user's signal-to-interference-and-noise ratio converges. Step 6: Underlay spectrum sharing transmission of the secondary base station: The intelligent reflector sets the reflection coefficient of each reflector unit according to the optimized discrete beamforming vector. The secondary base station uses the optimized power p s Send a signal to the secondary user.

2. The spectrum sharing method according to claim 1, wherein: Step 1: Obtain channel state information; the specific process is: the primary user transmits a pilot signal, and the primary base station adjusts the direct channel h between the primary user and the primary base station according to the received pilot signal. pp and reflection channel h prp The secondary base station estimates the direct channel h between the primary user and the secondary base station based on the received pilot signal. sp and reflection channel h srp The secondary user transmits a pilot signal, and the primary base station estimates the direct channel h between the secondary user and the primary base station based on the received pilot signal. ps and reflection channel h prs The secondary base station estimates the direct channel h between the secondary user and the secondary base station based on the received pilot signal. ss and reflection channel h srs Estimation is performed; the primary base station sends the estimated channel state information to the secondary base station.

3. The spectrum sharing method according to claim 2, wherein: Step 2: Establish an optimization model; the specific process is: the transmission power of the primary base station and the secondary base station are respectively recorded as p p and p s , p s The maximum value is P max , that is, p s ≤P max , the noise power of the primary user and the secondary user are respectively recorded as and The minimum signal to interference and noise ratio required by the primary user is denoted as γ th , the smart reflector beamforming vector is recorded as v H =[v1,...,v N ],in is the reflection coefficient of the nth reflection unit of the smart reflection surface, θ n is the phase of the reflection coefficient, n=1,...,N; the primary user signal to interference noise ratio is higher than γ th As the constraint condition, the signal to interference and noise ratio of the secondary user Maximization is the goal, and the secondary base station transmission power p is established s and smart reflector beamforming vector v H The joint optimization model of where h ss and h srs are the direct channel and the reflected channel between the secondary user and the secondary base station, respectively, h ps and h prs They are the direct channel and the reflected channel between the primary user and the secondary base station respectively.

4. The spectrum sharing method according to claim 3, wherein: The secondary base station transmit power p is established by an alternating iterative method s and smart reflector beamforming vector v H The joint optimization model of , the specific process is: Step 11: Fixing the Smart Reflector Discrete Beamforming v H , secondary base station transmission power p s The closed-form solution is Among them, min{a,b} represents the smaller of a and b, p p The main base station transmit power, P max is the maximum transmission power of the secondary base station, h pp and h prp are the direct channel and the reflected channel between the primary user and the primary base station, respectively, h sp and h srp They are the direct channel and the reflected channel between the secondary user and the primary base station respectively. The secondary base station receives the pilot signal. is the noise power of the primary user, γ th The minimum signal to interference and noise ratio required by the primary user, v H Beamforming vector for smart reflector; Step 12: Fix the secondary base station transmit power p s , optimize the smart reflector discrete beamforming by binary search v H ; Step 13: Repeat steps 11 and 12 until convergence.

5. The spectrum sharing method according to claim 4, wherein: In step 12, the discrete beamforming of the smart reflector v is optimized by binary search. H The specific process is: Step 121: Determine the lower bound t of the secondary user signal to interference and noise ratio t low and upper bound t up : The secondary user signal to interference and noise ratio is non-negative, so t≥0, that is, the lower bound of t is t low =0; When the primary base station transmit power is 0, the secondary user is minimally interfered with by the primary base station, and the secondary user's signal to interference and noise ratio is maximized. That is, the upper bound of t is Step 122: Let t = (t low +t up ) / 2, to test whether the optimization problem shown in formula (2) has a solution: In order to simplify the expression, the matrix expression G is defined as s =p s [h srs ,h ss ][h srs ,h ss ] T -tp p [h prs ,h ps ][h prs ,h ps ] T and G p =p p [h prp ,h pp ][h prp ,h pp ] T -tp s [h srp ,h sp ][h srp ,h sp ] T , and the vector expression ξ c =[1,cosΔθ,cos(2Δθ)...,cos((2 Q -1)Δθ)]、ξ s =[1,sinΔθ,sin(2Δθ)...,sin((2 Q -1)Δθ)] and Θ=[0,Δθ,2Δθ,...,(2 Q -1)Δθ], where Δθ = 2π / 2 Q is the interval of discrete phase values ​​of reflection coefficient of each reflection unit, Q is the order of discrete phase values ​​of reflection coefficient of each reflection unit; i and j represent the subscripts of the corresponding matrix or vector; G s (i,j) and G p (i,j) represent the matrix expression G s and G p The i-th row and j-th column element of G s (i,i) and G p (i,i) represent the matrix expression G s and G p The i-th row and i-th column element of G s (i,N+1) and G p (i, N+1) represent the matrix expression G s and G p The element in row i and column N+1; z i,j and ρ i,j Represents the SOS1 corresponding to the i-th row and j-th column of the three-dimensional matrix variables z and ρ respectively b Type vector, w i Represents the i-th row SOS1 of the two-dimensional matrix w b type vector; Step 123: If the problem shown in formula (2) has a solution, let t low =t; otherwise, let t up =t; Step 124: Repeat steps 122 and 123 until t up -t low ≤10 -3 , get the optimal discrete beamforming vector v H .

Images