A wireless resource allocation method for a LEO satellite-ground fusion uplink communication system

Abstract

The application discloses a wireless resource allocation method for a LEO satellite-ground fusion uplink communication system, which comprises user base station selection, power and carrier allocation and power and carrier allocation methods for base station to satellite link, first establishes a system energy efficiency maximization model for subcarrier and power allocation; then quotes a slack variable to convert the model into a target function lower bound model; uses a symbolic function to couple power allocation variables and carrier allocation variables; and further decomposes the target function lower bound model into a base station access and carrier allocation optimization model and a power allocation optimization model; linear approximation method and successive convex approximation method are respectively used for solving, so that the wireless resource allocation of the uplink communication system is realized. By using the method, 0-1 integer optimization variables are removed, the calculation complexity is low, and higher uplink system energy efficiency is obtained.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of communication, and particularly relates to a wireless resource allocation method for a LEO satellite-ground integrated uplink communication system. BACKGROUND

[0002] In the existing research on the uplink transmission system of the LEO satellite-ground integrated network, most of them take the uplink throughput of the system as the optimization target, lack the research on the energy efficiency of the system, and part of the research also lacks the consideration of the offloading rate demand of the user; in addition, most of the research uses the traditional mixed integer programming scheme in the design and optimization of the whole system, and the algorithm complexity is relatively high, and most of the user base station selection problems also use the strategy of accessing nearby to process. Therefore, a joint selection and allocation scheduling scheme is needed, but designing an optimization scheme for jointly selecting user base stations, subcarriers and power allocation will introduce a large number of 0-1 integer variables, which also increases the difficulty of algorithm design. SUMMARY

[0003] The present application provides a wireless resource allocation method for a LEO satellite-ground integrated uplink communication system.

[0004] Technical scheme: The present application discloses a wireless resource allocation method for a LEO satellite-ground integrated uplink communication system, which is specifically as follows: step one: taking the maximum uplink system energy efficiency as the objective function, constructing a system energy efficiency maximization model A1 based on the joint optimization problem of the power allocation of the ground network user, the base station selection, the carrier allocation, and the base station power allocation and carrier selection of the base station to satellite link; and using the sign function to couple the power allocation variable and the carrier allocation variable to construct the constraint condition;

[0005] Step two: converting the optimization problem in step one into an approximate problem of the lower bound of the optimization objective function, so as to convert the model A1 into a model A2;

[0006] Step three: according to the constraint condition of the model A2, decomposing the lower bound approximate problem in the model A2 into two sub-problems of the user and base station power allocation and the carrier allocation and base station selection, introducing the relaxation variable, converting the non-convex constraint condition related to the user and base station power allocation in the model A2 into a convex constraint condition; constructing a linear iterative function to convert the non-convex constraint condition related to the carrier allocation and base station selection in the model A2 into a convex constraint condition, so as to convert the model A2 into a model A3;

[0007] Step four: iteratively solving the model A3 to obtain the final user carrier power allocation state p and the base station carrier power state q; p=[p 1 ,…,p u ,…,p U ] and q=[q1,…,qb ,…,q B ], where q b denotes the optimal carrier power allocation of base station b, p u denotes the carrier power allocation of user u; B denotes the total number of base stations, U denotes the total number of users, q b,m denotes the transmit power allocated by base station b on the mth subcarrier of the satellite; m = 1, 2, …, M; M denotes the total number of available subcarriers between the base station and the satellite; denotes the transmit power allocated by user u on the nth subcarrier of the bth base station, n = 1, 2, …, N, N denotes the total number of available subcarriers between the user and the base station.

[0008] Further, the expression of the step of maximizing the system energy efficiency model A1 is as follows:

[0009]

[0010] where R u denotes the data offloading rate that can be achieved by the u th user, denotes the data offloading rate that can be achieved by user u on the nth subcarrier of the bth base station, The expression of R B0 is the bandwidth of each subband, B0 = B c / N, B c is the OFDM frequency band bandwidth of the ground transmission system, is the transmit power allocated by user u on the nth subcarrier of the bth base station, is the channel gain coefficient of user u on the nth subcarrier of the bth base station, The expression of R is the path loss exponent, is the small-scale Rayleigh fading, d u,b is the distance between the user and the base station; is the Gaussian white noise power of the ground transmission system; p0 is the rated transmit power of each user; sgn(.) is the sign function, is the indicator variable of user u using the nth subcarrier of the bth base station, if user u uses the nth subcarrier of the bth base station, otherwise D0 is the preset lower limit of the user data offloading rate, q0 is the rated transmit power of each base station, sgn(q b,m ) = y b,m , y b,man indicator variable for the bth base station using the mth subcarrier of the satellite; R b a data offloading rate achievable by the bth base station, R b The expression of A2 is: C b a total backhaul link capacity achievable by the base station b to the satellite.

[0011] Further, the expression of A2 is:

[0012]

[0013] where t is the lower bound of the square of the uplink communication system energy efficiency, and v is the upper bound of the square of the total base station power.

[0014] Further, the non-convex constraint condition in A2 related to the power allocation of the user and the base station in step three is converted into a convex constraint condition, which is:

[0015] Step 3.1: lower bound relaxation variable is adopted The constraint condition C1.1 is converted into C1.1.1 and C1.1.2:

[0016]

[0017] where, SINR achievable by the user u on the nth subcarrier of the bth base station,

[0018] Step 3.2: lower bound relaxation variable is adopted The constraint conditions C1.1.1 and C1.1.2 are converted into C1.1.3 and

[0019] C1.1.4:

[0020]

[0021] Step 3.3: the non-convex constraint condition C1.1.3 is converted into the convex constraint condition C1.1.5:

[0022]

[0023] where, (i) represents a convex upper bound, i represents the i-th iteration when the user and base station power allocation problem in A3 is solved iteratively, and

[0024] Step 3.4: upper bound relaxation variable is introduced The non-convex constraint condition C1.1.2 is converted into C1.1.6 and C1.1.7

[0025]

[0026] Step 3.5: Introducing upper bound slack variables Transforming the non-convex constraint condition C1.1.6 into a convex constraint condition C1.1.8:

[0027]

[0028] Step 3.6: Transforming the non-convex constraint condition C4.1 into a convex constraint condition C4.1.1:

[0029]

[0030] Step 3.7: Transforming the non-convex constraint condition C7.1 into a convex constraint condition C7.1.1:

[0031]

[0032] where g b,m represents the channel gain coefficient of the b-th base station on the m-th subcarrier of the satellite, G s and G r represent the antenna gains of the base station and the satellite, respectively, λ m represents the wavelength of the m-th subcarrier, d b represents the distance between the base station and the satellite, β b,m represents the small-scale Rayleigh fading, B1 = B ka / M, and B ka is the OFDM frequency band bandwidth of the satellite transmission system; represents the Gaussian white noise power of the satellite transmission system.

[0033] Further, the lower linear iteration function expression established in step three is as follows:

[0034]

[0035] where τ1 and τ2 are constant regularization factors, c1 and c2 are constants, and the superscript k is the k-th iteration when solving the carrier allocation and base station problem in model A3;

[0036] Transforming the non-convex constraint condition related to carrier allocation and base station selection in model A2 into a convex constraint condition, specifically:

[0037] Transforming the non-convex constraint condition C3.1 of user base station selection into a convex constraint condition C3.1.1:

[0038]

[0039] The non-convex constraint condition C2.1 of the user carrier allocation is converted into a convex constraint condition C2.1.1:

[0040] The non-convex constraint condition C6.1 of the base station carrier allocation is converted into a convex constraint condition C6.1.1:

[0041] Further, the expression of the model A3 is:

[0042]

[0043] Wherein, δ is a matrix composed of a matrix composed of a matrix composed of , τ1 and τ2 are constant regularization factors, c1 and c2 are constants, and the superscript k is the kth iteration for solving the carrier allocation and base station problem in the model A3.

[0044] Further, the step four is specifically:

[0045] Step 4.1: initialize the iteration number, accuracy, user power and base station power; initialize the user power as p0 / N and the base station power as q0 / M, and the initialized user power and base station power satisfy the constraint conditions C1, C2.1.1, C3.1.1, C5.1, C6.1.1;

[0046] Step 4.2: solve the transformed model A3 by using the successive convex approximation method to obtain the ith iteration result (i) , (i) , (i) , (i) ;

[0047] Step 4.3: substitute (i) , (i) , (i) , (i) into the model A3 to perform the next iteration until the objective function value of the model A3 no longer changes, the iteration is terminated, and the initial values of p and q are obtained based on the optimal * , * , * , * , and go to step 4.4:

[0048] Step 4.4: solve the model A3 by using the linear approximation method to obtain the kth iteration result p (k) and q (k) ;

[0049] Step 4.5: substitute p (k)and q (k) Substitute into model A3, carry out next iteration, when the result of iteration is not changed, terminate iteration, output the optimal value of p and q.

[0050] Further, the step four is specifically: when the step 4.1 initialization, t (0) The initial value is:

[0051]

[0052] Wherein, the superscript 0 represents the initial value.

[0053] Beneficial effect: the present application removes 0-1 integer optimization variable, has low calculation complexity, and obtains higher uplink system energy efficiency. DETAILED DESCRIPTION

[0054] The illustrative embodiments of the present application and the description thereof are used to explain the present application, and do not constitute an improper limitation on the present application.

[0055] The present application provides a wireless resource allocation method for LEO satellite-ground fusion uplink communication system. The method takes single LEO satellite multi-base station uplink OFDM communication system as background, and takes maximum uplink system energy efficiency as objective function, studies the joint optimization problem of ground network user power allocation, base station selection, carrier allocation, and base station power allocation and carrier selection of base station to satellite link. Firstly, the mathematical optimization model corresponding to the problem is derived, the mathematical optimization problem is converted into an approximate problem of lower bound of optimization objective function by using relaxation variable, and the relationship between user and base station power allocation variable and carrier allocation variable is established by using symbolic function; according to the constraint condition of the problem, the lower bound approximate problem is decomposed into two sub-problems of user and base station power allocation and carrier allocation and base station access for iterative solution; the base station and user power allocation sub-problem is iteratively solved by using successive convex approximation algorithm; the carrier allocation and base station access sub-problem is iteratively solved by using a low complexity linear approximation method.

[0056] In order to better illustrate the method of the present application, the following will be described in more detail in conjunction with embodiments:

[0057] Consider a single LEO satellite uplink OFDM transmission system with U users and B ground base stations. The uplink communication system includes two parts: ground transmission system and base station to satellite transmission system. Each base station is equipped with an isotropic omnidirectional transmitting antenna, and the rated transmitting power of each base station is q0, and the rated transmitting power of each user is p0; U users are randomly distributed in the area covered by B base stations, each user can only access one base station and allocate a subcarrier to unload data; the base station sends the data unloaded to the local user to the LEO satellite through the backhaul link between the base station and the satellite, and a subcarrier can only be allocated to one base station; the transmission link between the user and the base station uses C band, B c is the OFDM frequency band bandwidth of the ground transmission system, and N is the number of available subcarriers between the user and the base station; the transmission link between the base station and the satellite uses Ka band, B ka is the OFDM frequency band bandwidth of the satellite transmission system, and M is the number of available subcarriers between the base station and the satellite.

[0058] First, in the ground transmission system, let (b, n) represent the nth subcarrier of the bth base station, and let represent the indication variable of the user u using the nth subcarrier of the bth base station, if represent using, otherwise Let x = [x 1 ,…,x u ,…,x U ] represent the carrier allocation state vector of all users, and the carrier allocation state vector x u of the user u has the following expression:

[0059]

[0060] Let p = [p 1 ,…,p u ,…,p U ] represent the carrier power allocation state vector of the user, and the carrier power allocation state vector p u of the user u has the following expression:

[0061]

[0062] wherein, represents the transmitting power allocated by the user u on the nth subcarrier of the bth base station. is the channel gain coefficient of the user u on the nth subcarrier of the bth base station, is a complex number, which has the following expression:

[0063]

[0064] wherein, α represents the path loss index, is a complex Gaussian variable with zero mean, real and imaginary parts statistically independent and variance one, representing small-scale Rayleigh fading, d u,b denotes the distance between the user and the base station.

[0065] The second step is to establish the relationship between the user carrier allocation variable and the carrier power allocation variable by using the sign function sgn(·):

[0066]

[0067] The achievable signal-to-interference-and-noise ratio of the user u on the nth subcarrier of the bth base station is denoted as The expression of is:

[0068]

[0069] wherein, denotes the sum of the noise power of the co-channel interference caused to the nth subcarrier of the user u, denotes the Gaussian white noise power of the ground transmission system.

[0070] The third step is to denote the achievable data offloading rate of the user u on the nth subcarrier of the bth base station as The expression of is:

[0071]

[0072] wherein, B0=B c / N denotes the bandwidth of each subband; the achievable data offloading rate of the uth user is R u , R u The expression is:

[0073]

[0074] The achievable data offloading rate R b of the bth base station has the following expression:

[0075]

[0076] The fourth step is in the transmission system from the base station to the satellite, y is the carrier allocation state vector of the base station, y=[y1,…,y b ,…,y B ], y b is the carrier allocation state vector of the bth base station, y b =[y b,1 ,…,y b,m ,…,y b,M ] -1 ; y b,mis the indicator variable for the base station b to use the mth subcarrier of the satellite, y b,m = 1 means to use, otherwise y b,m = 0; q is the carrier power allocation state vector of the base station, q = [q1,..., q b ,..., q B ], q b is the carrier power allocation state vector of the base station b, q b = [q b,1 ,..., q b,m ,..., q b,M ] -1 ; g b,m is the channel gain coefficient of the mth subcarrier of the satellite for the bth base station, g b,m is a complex number, which has the following representation:

[0077]

[0078] where G s and G r represent the antenna gains of the base station and the satellite, respectively, λ m represents the wavelength of the mth subcarrier, β b,m represents the small-scale Rayleigh fading, and d b represents the distance between the base station and the satellite.

[0079] In the fifth step, the relationship between the base station carrier allocation variable and the carrier power allocation variable is established using the sign function sgn(·):

[0080] y b,m = sgn(q b,m )

[0081] The achievable backhaul link capacity of the base station b on the mth subcarrier of the satellite is C b,m , which has the following representation:

[0082]

[0083] where B1 = B ka / M represents the bandwidth of each subband; represents the Gaussian white noise power of the satellite transmission system.

[0084] The total achievable backhaul link capacity C b of the base station b to the satellite has the following expression:

[0085]

[0086] In the sixth step, all base stations have no buffer for uploading data, so the data offloading rate R band the total backhaul link capacity C between them b The following constraints are satisfied between the total backhaul link capacity C and the total backhaul link capacity C between them:

[0087]

[0088] A system energy efficiency maximization model A1 for subcarrier and power allocation is established, and the model A1 is expressed as:

[0089]

[0090] The model A1 is simplified as follows:

[0091]

[0092] Constraint condition C1 represents that the sum of the transmission power of all users is not greater than p0; constraint condition C2 represents that the subcarrier of any base station can only be allocated to one user; constraint condition C3 represents that one user can only access one base station and allocate one subcarrier; constraint condition C4 represents that the data offloading rate of the user cannot be lower than the preset lower limit; constraint condition C5 represents that the total transmission power of any base station is not greater than q0; constraint condition C6 represents that one subcarrier can only be allocated to one base station; constraint condition C7 represents that the data offloading rate R of the user collected by the bth ground base station is not greater than the total link capacity C between the bth ground base station and the satellite b The total link capacity C between the bth ground base station and the satellite b .

[0093] The model A1 is solved, and the specific steps are as follows:

[0094] Step 1: Introducing the lower bound t of the square of the system energy efficiency and the upper bound v of the square of the total power of the base station, converting the model A1 into a model A2, and the expression of the model A2 is:

[0095]

[0096] Limited to:

[0097]

[0098] The model A2 is simplified as:

[0099]

[0100]

[0101] Step 2: Power allocation of the user and the base station;

[0102] Step 2-1: The lower bound of the relaxation variable Transform the constraint C1.1 into C1.1.1 and C1.1.2:

[0103]

[0104] Step 2-2: Use lower bound relaxation variable Transform the constraints C1.1.1 and C1.1.2 into C1.1.3 and C1.1.4:

[0105]

[0106] Where C1.1.4 is a convex constraint, the inequality right side expression of constraint C1.1.3 is non-convex, so C1.1.3 is a non-convex constraint.

[0107] Step 2-3: For the inequality right side expression of constraint C1.1.3 The convex upper bound Θ of the i-th iteration on its domain t ≥ 0, v ≥ 0 (i) Expressed as:

[0108]

[0109] At this time t and v represent the optimal value.

[0110] Transform the non-convex constraint C1.1.3 into the convex constraint C1.1.5:

[0111]

[0112] Step 2-4: Use upper bound relaxation variable Transform the non-convex constraint C1.1.2 into C1.1.6 and C1.1.7:

[0113]

[0114] Where C1.1.7 is a convex constraint, and C1.1.6 is a non-convex constraint.

[0115] Step 2-5: Use upper bound relaxation variable Express the convex upper bound of the inequality right side expression of the non-convex constraint C1.1.6 as:

[0116]

[0117] Where i is the iteration number, And is the result of the i-th iteration; at this time Indicates the optimal value.

[0118] Step 2-6: Transform the non-convex constraint C1.1.6 into the convex constraint C1.1.8:

[0119]

[0120] Step 2-7: Convert the non-convex constraint condition C4.1 into a convex constraint C4.1.1:

[0121]

[0122] Step 2-8: Convert the non-convex constraint condition C7.1 into a convex constraint C7.1.1:

[0123]

[0124] Step 3: Linear approximation method for user base station access, user and base station subcarrier allocation;

[0125] Step 3-1: Establish the following linear iterative function:

[0126]

[0127] Where τ1 and τ2 are constant regularization factors, c1 and c2 are constants, k is the iteration number, And The optimization result of the kth iteration, when the iteration converges, the function left and right sides are equal.

[0128] Step 3-2: Convert the non-convex constraint condition C3.1 of user base station selection into a convex constraint condition C3.1.1:

[0129]

[0130] Step 3-3: Convert the non-convex constraint condition C2.1 of user subcarrier allocation into a convex constraint condition C2.1.1:

[0131]

[0132] Step 3-4: Convert the non-convex constraint condition C6.1 of base station subcarrier allocation into a convex constraint condition C6.1.1:

[0133]

[0134] Step 3-5: Convert the model A2 into model A3 by the above steps, and the expression of the model A3 is:

[0135]

[0136]

[0137] δ is The matrix composed of φ is The matrix composed of Γ is The matrix formed.

[0138] Step 4: Iteratively solve model A3 to obtain the optimal solution of the lower bound t of the square of system energy efficiency, and then determine the carrier power allocation state vector p of the user and the carrier power allocation state vector q of the base station in the constraints of model A1, so as to realize the wireless resource allocation of the satellite-ground integrated uplink communication system.

[0139] Step 4-1: Initialize the number of iterations, accuracy, user and base station power, etc.;

[0140] Step 4-1-1: Initialize the number of iterations k = 0, i = 0, and the precision ζ = 10. -3 ω=10 -2 ;

[0141] Step 4-1-2: Initialize user power and base station power to satisfy constraints C1, C2.1.1, C3.1.1, C5.1, and C6.1.1;

[0142] Step 4-1-3: Initialize user power as p0 / N and base station power as q0 / M;

[0143] Step 4-2: According to Solving for the initial value γ using the expression (0) γ is The matrix formed is given an initial value δ by constraint C1.1.2. (0) δ is The matrix formed by this is an initial value indicated by the superscript 0.

[0144] Step 4-3: Solve for the initial value Γ according to constraint C1.1.7 (0) Γ is The matrix formed;

[0145] Step 4-4: Solve for the initial value v according to constraint C1.2 (0) ;

[0146] Steps 4-5: Solve for the initial value t (0) Solve it as follows:

[0147]

[0148] Substitute the result v(0) from step 4-4 into the above equation to solve for the initial value t. (0) ;

[0149] Steps 4-6: Iteratively solve model A3. In the i-th iteration, the solution process is as follows:

[0150] Step 4-6-1: solve the transformed convex optimization problem model A3 by using the method of successive convex approximation, and obtain the i-th iteration result t (i) (i) (i) (i) wherein t represents the result value obtained in the i-th iteration;

[0151] Step 4-6-2: t (i+1) = t (i) , v (i+1) = v (i) , δ (i+1) = δ (i) , Γ (i+1) = Γ (i) , i = i + 1; substitute the updated numerical values into the problem model A3, replace the parameters, and calculate the objective function value of model A3 again. When the objective function value of model A3 no longer changes, the iteration method is terminated, and the results p and q are output;

[0152] Step 4-6-3: solve the base station selection and carrier allocation problem in the convex optimization problem model A3 by using the method of linear approximation iteration; obtain the k-th iteration result p (k) and q (k) ;

[0153] Step 4-6-4: p (k+1) = p (k) , q (k+1) = q (k) , k = k + 1; substitute the updated numerical values into model A3 again, replace the parameters, and when the iteration power result no longer changes, the iteration method is terminated, and the final result value is output.

[0154] In addition, it should be noted that the various specific technical features described in the above specific embodiments can be combined in any appropriate manner without contradiction. In order to avoid unnecessary repetition, the present application will not make further description on various possible combinations.​​​

Claims

1. A method for allocating radio resources in a LEO satellite-terrestrial converged uplink communication system, the uplink communication system comprising a terrestrial transmission system and a satellite-terrestrial transmission system, characterized in that, The method specifically comprises the following steps: Step one: taking the maximum uplink system energy efficiency as an objective function, constructing a system energy efficiency maximization model A1 based on the joint optimization problem of power allocation of ground network users, base station selection, carrier allocation, and base station power allocation and carrier selection of the base station to satellite link; and using a sign function to couple the power allocation variable and the carrier allocation variable to construct a constraint condition; Step two: converting the optimization problem in step one into an approximate problem of the lower bound of the optimization objective function, thereby converting the model A1 into a model A2; Step three: according to the constraint condition of the model A2, decomposing the lower bound approximate problem in the model A2 into two sub-problems of user and base station power allocation and carrier allocation and base station selection, introducing a relaxation variable, converting the non-convex constraint condition related to the user and base station power allocation in the model A2 into a convex constraint condition; constructing a linear iterative function to convert the non-convex constraint condition related to the carrier allocation and base station selection in the model A2 into a convex constraint condition, thereby converting the model A2 into a model A3; Step 4: Iteratively solve model A3 to obtain the final user carrier power allocation state p and base station carrier power state q; p = [p 1 ,…,p u ,…,p U ],q=[q1,…,q b ,…,q B ], where q b p represents the preferred carrier power allocation for base station b. u This represents the carrier power allocation for user u; B represents the total number of base stations, and U represents the total number of users. Pb,n(u) denotes the transmit power allocated to user u at the nth subcarrier of the bth base station, n = 1, 2,..., N, N denotes the total number of available subcarriers between the user and the base station.

2. The method of claim 1, wherein, The expression of the system energy efficiency maximization model A1 in the step one is as follows: wherein R u denotes the data offloading rate achievable by the u-th user, denotes the data offloading rate achievable by the u-th user at the b-th base station on the n subcarriers, The expression for B0 represents the bandwidth of each sub-band, B0 = B c / N, B c is the OFDM frequency band bandwidth of the terrestrial transmission system, represents the transmit power allocated to user u on the nth subcarrier of the bth base station, represents the channel gain coefficient of user u on the nth subcarrier of the bth base station, The expression of is: α represents the path loss exponent, represents the small-scale Rayleigh fading, d u,b represents the distance between the user and the base station; represents the Gaussian white noise power of the terrestrial transmission system; p0 represents the rated transmit power of each user; sgn(.) is the sign function, represents the indicator variable of user u using the nth subcarrier of the bth base station, if represents that user u uses the nth subcarrier of the bth base station, otherwise D0 represents the preset lower limit of the data offloading rate of the user, q0 is the rated transmit power of each base station, and sgn(q b,m ) = y b,m , y b,m is the indicator variable of the bth base station using the mth subcarrier of the satellite; R b represents the data offloading rate that the bth base station can achieve, R b The expression of is: C b represents the total backhaul link capacity that the base station b to the satellite can achieve; q b,m represents the transmit power allocated to the base station b on the mth subcarrier of the satellite; m = 1, 2, …, M; M represents the total number of available subcarriers between the base station and the satellite.

3. The method of Claim 2, wherein, The expression of the model A2 is as follows: Wherein, t is the lower bound of the square of the uplink communication system energy efficiency, and v is the upper bound of the square of the total base station power.

4. The method of Claim 3, wherein, The non-convex constraint condition related to the user and base station power allocation in the model A2 is converted into a convex constraint condition in the step three, which is specifically as follows: Step 3.1 : Using lower bound relaxation variables Transforming constraint C1.1 into C1.1.1 and C1.1.2: wherein, Sb,n(u) denotes the signal-to-interference ratio achievable by the user u on the nth subcarrier at the bth base station, Step 3.2: Lower bounding variables using the relaxation Transforming constraints C1.1.1 and C1.1.2 into C1.1.3 and C1.1.4: Step 3.3: converting the non-convex constraint condition C1.1.3 into a convex constraint condition C1.1.5: where Θ (i) denotes the convex upper bound, i denotes the i-th iteration in the iterative solution of the user-to-base power allocation problem in model A3, and Step 3.4: Introducing upper bound slack variables Transforming the non-convex constraint C1.1.2 into C1.1.6 and C1.1.7 Step 3.5: Introducing upper bound slack variables Transforming the non-convex constraint C1.1.6 into the convex constraint C1.1.8: Step 3.6: converting the non-convex constraint condition C4.1 into a convex constraint condition C4.1.1: Step 3.7: converting the non-convex constraint condition C7.1 into a convex constraint condition C7.1.1: where g b,m represents the channel gain coefficient of the b-th base station on the m-th subcarrier of the satellite, G s and G r respectively represent the antenna gain of the base station and the satellite, λ m represents the wavelength of the m-th subcarrier, d b represents the distance between the base station and the satellite, β b,m represents the small-scale Rayleigh fading, B1=B ka / M, B ka is the OFDM frequency band bandwidth of the satellite transmission system; represents the Gaussian white noise power of the satellite transmission system.

5. The method of claim 3, wherein the method further comprises: The expression of the lower linear iterative function established in the step three is as follows: Wherein, τ1 and τ2 are constant regularization factors, c1 and c2 are constants, and superscript k is the kth iteration when solving the carrier allocation and base station problem in the model A3; The non-convex constraint condition related to the carrier allocation and base station selection in the model A2 is converted into a convex constraint condition, which is specifically as follows: The non-convex constraint condition C3.1 of the user base station selection is converted into a convex constraint condition C3.1.1: The non-convex constraint condition C2.1 of the user carrier allocation is converted into a convex constraint condition C2.1.1: The non-convex constraint condition C6.1 of the base station carrier allocation is converted into a convex constraint condition C6.1.1:

6. The method of Claim 4, wherein, The expression of the model A3 is as follows: where δ is is a matrix composed of is a matrix composed of is a matrix composed of τ1 and τ2 are constant regularization factors, c1 and c2 are constants, and the superscript k is the kth iteration in solving the base station problem in model A3.

7. The method of Claim 6, wherein, The step four is specifically as follows: Step 4.1: initializing the iteration number, precision, user power and base station power; initializing the user power as p0 / N and the base station power as q0 / M, and the initialized user power and base station power satisfy the constraint conditions C1, C2.1.1, C3.1.1, C5.1, C6.1.1; Step 4.2: Solve the transformed model A3 using the successive convexification approach to obtain the ith iteration result t (i) ,v (i) ,δ (i) ,Γ (i) ; Step 4.3: Substitute t (i) ,v (i) ,δ (i) ,Γ (i) into model A3, and perform the next iteration until the objective function value of model A3 does not change anymore, the iteration terminates, and the optimal t * ,v * ,δ * ,Γ * are obtained, and go to Step 4.4: Step 4.4: Model A3 is iteratively solved using a linear approximation method; the kth iteration result p is obtained (k) and q (k) ; Step 4.5: Substitute p (k) and q (k) into model A3, and perform the next iteration. Terminate the iteration when the result of the iteration does not change, and output the optimal values of p and q.

8. The method of Claim 7, wherein, The step four is specifically: the initial value of t (0) is: Wherein, superscript 0 represents the initial value.

Images