High-spectral-efficiency transmission method suitable for star cognitive network forward link

Abstract

The invention discloses a high-spectral-efficiency transmission method suitable for a star-sky cognitive network forward link, which comprises the following steps of: selecting two gateway stations with the best channel quality from a plurality of independent and unrelated gateway stations on the ground to carry out Alamouti space-time coding in a satellite and high-altitude platform frequency spectrum sharing scene, and communicating with a stationary orbit satellite by utilizing a terahertz frequency band. And the satellite determines a decoding sequence according to the channel state information of the user, and a robust beam forming scheme based on a virtual signal to interference plus noise ratio criterion is completed only under the imperfect CSI condition of the known high-altitude platform user. NOMA user combination is dynamically optimized through beam forming processing and a multi-user feedback mechanism, and time-frequency resource multiplexing maximization is achieved. According to the invention, aiming at a forward link of a starry sky cognitive network, space-time coding and a non-orthogonal multiple access technology are combined, so that spectrum coexistence of a satellite network and a high-altitude platform network is realized, and the spectrum efficiency of the system is remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communication and is applied to the forward link transmission of a star cognitive network in a multi-user scenario. Specifically, it relates to a high spectral efficiency transmission method applicable to the forward link of a star cognitive network. Background Art

[0002] Satellite communication technology uses satellites as space relay stations to achieve comprehensive interconnection among sea, land, air, and mobile users, and even between terrestrial networks and mobile networks. A geostationary communication system using beamforming technology can generate a large number of narrow spot beams, cover a large ground area with high gain, and adjust the beam shape as needed to achieve seamless coverage. At the same time, it can utilize the spatial diversity characteristics between different beams and adopt full frequency reuse to achieve high spectral efficiency. Compared with traditional terrestrial wireless systems, it has the advantages of wide coverage and being unaffected by geographical conditions, and plays an increasingly important role in the next-generation mobile communication system. At the same time, as a frontier evolution of satellite communication technology, the star cognitive network has shown significant advantages in improving spectral efficiency. By deeply integrating high-altitude platforms (such as stratospheric balloons, drones, etc.), the star cognitive network can dynamically adjust and optimize beamforming technology to achieve more refined spectrum management.

[0003] For the satellite link part of the star cognitive network, there will be a situation where a large number of users apply for services simultaneously. When a large number of users access the network simultaneously, it will impose a greater burden on the feeder link. Due to the increasingly scarce spectrum resources, traditional radio frequency links cannot effectively meet the ultra-reliable and large-capacity requirements for large-scale access. Therefore, it is necessary to consider a new communication method in the feeder link. In addition, when a geostationary satellite serves a large number of users, the interference between users will significantly reduce the communication capacity of the system. Theory and practice have shown that beamforming algorithms or user scheduling methods with the optimization goal of maximizing system capacity can effectively reduce interference between users, thereby improving the communication capacity of the system. However, this often requires obtaining the channel information of users. Due to the transmission delay between networks, it is difficult for the satellite system to obtain perfect channel state information of high-altitude platform users. Therefore, developing beamforming algorithms and low-complexity user scheduling methods based on imperfect channel information of high-altitude platform users can effectively improve the robustness of the star fusion system, thereby achieving effective beamforming and user scheduling, while increasing the ergodic capacity of the system and reducing system overhead.

[0004] Although the starry sky cognitive network can significantly improve the spectrum efficiency through its unique architecture and intelligent management mechanism, in the face of the urgent demands of future mobile communication systems for higher capacity and lower latency, a single technical means is often difficult to fully meet these requirements. Against this background, NOMA technology effectively improves the utilization rate of spectrum resources by allowing different users to transmit data non - orthogonally on the same spectrum resource. Compared with traditional orthogonal multiple access technologies, NOMA technology can support more users to access simultaneously within limited spectrum resources, significantly improving the user capacity and spectrum efficiency of the system. In addition, NOMA technology also uses advanced signal processing technologies to effectively separate and identify signals from different users at the receiving end, ensuring the reliability and stability of data transmission. Therefore, applying NOMA technology to the user link of the starry sky cognitive network can make full use of limited spectrum resources to achieve ultra - reliable and high - capacity transmission for large - scale access, providing strong technical support for the development of the starry sky cognitive network.

[0005] Regarding the above two aspects, the present invention provides a high - spectral - efficiency and robust transmission method applicable to the forward link of the starry sky cognitive network. Summary of the Invention

[0006] Technical Problems:

[0007] In the existing starry sky cognitive network, the traditional radio - frequency feeding link cannot effectively meet the capacity requirements for large - scale access. Studying a new feeding - link transmission method is one problem that the present invention needs to solve; on the other hand, due to the information transmission delay between networks, it is difficult for satellites to obtain perfect channel state information of high - altitude platform users. Studying a robust beamforming algorithm and user scheduling method for satellites to only know the imperfect CSI of high - altitude platform users to improve the system capacity while reducing system overhead is another problem that the present invention needs to solve.

[0008] Technical Solution: A data transmission method applied to the forward link of the starry sky cognitive network. Under the condition that the satellite network and the high - altitude platform network share spectrum resources, select the 2 gateway stations with the best channel quality for Alamouti space - time coding, and communicate with the geostationary satellite through the terahertz band. The satellite decodes the signals from the gateway stations; the satellite determines the decoding order on the user side according to the ground - user channel state information, and completes the design of the beamforming scheme based on the virtual signal - to - interference - plus - noise ratio criterion under the condition of ensuring the quality of service of high - altitude platform users; the geostationary satellite performs beamforming processing on all signals and then sends them to the corresponding users, and performs user scheduling according to the feedback information of each user to obtain the optimal NOMA user combination on the same time - frequency resource.

[0009] Specifically, it includes the following steps:

[0010] Step 1. In the scenario of spectrum coexistence between satellite networks and high-altitude platform networks, select the 2 gateway stations with the best channel quality from multiple independent and uncorrelated gateway stations on the ground for Alamouti space-time coding, and further use terahertz technology to communicate with the geostationary satellite. The geostationary satellite decodes the signals from the gateway stations using the decode-and-forward protocol;

[0011] Step 2. The geostationary satellite determines the decoding order on the user side according to the channel state information of the users and performs superposition coding in the power domain;

[0012] Step 3. Under the condition of ensuring the quality of service of high-altitude platform users, construct an optimization problem with the goal of maximizing the ergodic capacity of the satellite network, and use the virtual signal-to-interference-plus-noise ratio criterion to transform the optimization problem into a solvable form;

[0013] Step 4. Use the tight bound method to handle the channel error from the satellite to the high-altitude platform, and use the generalized Rayleigh quotient theorem to solve for the beamforming weight vector and transmit power;

[0014] Step 5. Based on the proposed beamforming scheme, the satellite performs beamforming processing on all signals and then sends them to each user, and performs user scheduling according to the 1-bit information fed back by each user to obtain the optimal NOMA user combination and beamforming weight vector under the same time-frequency resource conditions, realizing the maximization of the system capacity.

[0015] First, select the 2 gateway stations with the best channel quality from multiple independent and uncorrelated gateway stations on the ground for Alamouti space-time coding, and further use terahertz technology to connect to the geostationary satellite. The received signal of the geostationary satellite can be expressed as:

[0016]

[0017] where, y i represents the signal received by the geostationary satellite, I1 and I2 are terahertz channel coefficients, s i is the transmitted data symbol vector, n i is the additive white Gaussian noise, and i represents the gateway station serial number.

[0018] The geostationary satellite determines the decoding order on the user side according to the channel state information of the users and performs superposition coding in terms of power; then the transmitted signal of the geostationary satellite can be expressed as:

[0019]

[0020] where, x represents the transmitted signal of the geostationary satellite, w k is the normalized beamforming weight vector for serving the k-th user, P kDenote the transmit power for serving the \(k\) -th user; \(k\) represents the serial number of the served user, \(k = 1,\cdots,K\), and \(K\) represents the number of users served by the satellite.

[0021] Under the interference power constraint of the satellite on the high - altitude platform users, the design of the beamforming algorithm based on the virtual signal - to - interference - plus - noise ratio (SINR) criterion is completed. First, with the interference power of the satellite on the high - altitude platform users as the constraint condition, an optimization model is established under the condition that only the imperfect channel state information (CSI) of the high - altitude platform users is known; then, a beamforming scheme based on the virtual SINR criterion is designed, and the beamforming weight vector \(\mathbf{w}\) is obtained by iterative solution. k The established optimization model is shown as follows:

[0022]

[0023] where \(\mathbf{w}\) k is the beamforming weight vector for serving the \(k\) -th user, \(P\) k denotes the transmit power for serving the \(k\) -th user; \(k\) represents the serial number of the served user, \(k = 1,\cdots,K\), and \(K\) represents the number of users served by the satellite; \(\mathbf{h}\) k represents the channel vector between the satellite and the \(k\) -th user, \(\sigma\) 2 denotes the variance of additive white Gaussian noise. \(\mathbf{g}\) l represents the channel vector between the satellite and the \(l\) -th high - altitude platform user, \(\gamma\) th represents the interference power threshold of the satellite on the high - altitude platform users, \(l\) represents the serial number of the users served by the high - altitude platform, \(l = 1,\cdots,L\), and \(L\) represents the number of users served by the high - altitude platform, and \(E[\cdot]\) represents the mathematical expectation.

[0024] Then, a beamforming scheme based on the virtual SINR criterion is designed, and the beamforming weight vector is obtained by iterative solution. The specific steps are as follows:

[0025] Use Jensen's inequality to approximately transform the objective function, and obtain:

[0026]

[0027] Then, using the principle of virtual SINR, the problem of maximizing the ergodic capacity is transformed into the problem of maximizing the virtual SINR of each user:

[0028]

[0029] where is the channel covariance matrix between the satellite and the \(k\) -th user, is the channel covariance matrix between the satellite and the \(l\) -th high - altitude platform user, \(\mathbf{G}\) l is the available covariance matrix, \(\Delta\mathbf{G}\) l is the error of the channel covariance matrix, satisfying \(\|\Delta\mathbf{G}\| \)l ||≤α l ,μ k.j is a weighting coefficient, w k is the normalized beamforming weight vector for serving the k-th user, and P k represents the transmit power for serving the k-th user; k represents the serial number of the served user, k = 1,..., K, where K represents the number of users served by the satellite, and σ 2 represents the variance of additive white Gaussian noise, and g l represents the channel vector between the satellite and the l-th high-altitude platform user, and γ th represents the interference power threshold of the satellite for the high-altitude platform user, l represents the serial number of the user served by the high-altitude platform, l = 1,..., L, where L represents the number of users served by the high-altitude platform, and E[.] represents the mathematical expectation. First, the compact bound method is used to handle the channel error from the satellite to the high-altitude platform, and the constraint condition is transformed into:

[0030]

[0031] Then, an alternating iteration method is adopted to solve the above problem. First, fix the weighting coefficient, and use the maximum Rayleigh quotient to solve for w k and P k :

[0032]

[0033] where, A(φ k ) = a(φ k )a H (φ k ), I is the identity matrix, φ k is the angle of arrival of the k-th user's signal, c is the speed of light, G R is the antenna gain, f c is the carrier frequency and represents the distance between the k-th user and the center of the point beam n, r k is the gain vector of the on-board antenna point beam, and ν max (.) represents the principal eigenvector of the matrix. Then, fix the weight vector and use the gradient descent method to solve for μ k,j :

[0034]

[0035] where, Q k is the coefficient matrix corresponding to the k-th user, and 1 K-k is a vector of all 1s.

[0036] Finally, the optimal beamforming weight vector is obtained by alternating solutions.

[0037] When a large number of users access the network simultaneously, the interference among users will reduce the system throughput, and the system needs to reasonably arrange users, that is, user scheduling. In the satellite-aided cognitive network, based on the proposed beamforming scheme, the geostationary satellite first performs beamforming processing on all signals and then sends them to each user, and conducts user scheduling according to the 1-bit information fed back by each user, searching for the optimal NOMA user combination and beamforming weight vector on the same time-frequency resources to maximize the system capacity.

[0038] The beneficial technical effects achieved by the present invention are as follows:

[0039] The present invention adopts the architecture that combines the feeder link Alamouti space-time coding with terahertz technology and the user link NOMA technology. Under the condition of only knowing the non-perfect CSI of the high-altitude platform users, a robust beamforming algorithm suitable for NOMA transmission is designed. At the same time, multi-network cooperation is realized through spectrum sharing between the satellite and the high-altitude platform, which can not only significantly improve the communication capacity of the satellite network, but also guarantee the quality of service of the high-altitude platform users and meet the requirements of ultra-large capacity and high-reliability communication. Description of the Drawings

[0040] Figure 1 It is a schematic diagram of the satellite-aided cognitive network in the embodiment of the present invention;

[0041] Figure 2 It is a flowchart of the specific implementation manner of the present invention. Specific Implementation Manner

[0042] The technical solution of the present invention will be further described below with reference to the drawings. The following embodiments are only used to illustrate the present invention more clearly.

[0043] Embodiment, as Figure 1 shown, the present invention is applied to the data transmission scenario of the forward link of the satellite-aided cognitive network. In this scenario, two gateway stations with the best channel quality are selected from M independent and uncorrelated gateway stations on the ground for Alamouti space-time coding, and terahertz technology is used to communicate with the geostationary satellite. The geostationary satellite is configured with a multi-feed single-reflector antenna to generate N beams to serve K single-antenna users, and the high-altitude platform uses multicast technology to serve L single-antenna users. In the present invention, the satellite only knows the non-perfect channel state information of the high-altitude platform users.

[0044] A data transmission method applied to the forward link of a geostationary satellite system includes the following steps:

[0045] (1) Select two gateway stations with the best channel quality from multiple independent and uncorrelated gateway stations on the ground for Alamouti space-time coding, and use terahertz technology to communicate with the geostationary satellite. Then the received signal of the geostationary satellite can be expressed as:

[0046]

[0047] Among them, y i represents the signal received by the geostationary satellite, G t and G r are respectively the transmission gain of the gateway station and the reception gain of the satellite; η p and l s are respectively the antenna pointing loss and the free space loss, is the fading coefficient, following the α-μ distribution; s i is the transmitted data symbol vector, n i is the additive white Gaussian noise, and i represents the gateway station serial number.

[0048] (2) The geostationary satellite determines the decoding order on the user side according to the channel state information of the user, and performs superposition coding in the power domain:

[0049] In the user link, the geostationary satellite generates N beams to serve K users. Assuming the signal sent to the kth user is x k , k = 1, 2,..., K, satisfying E[|x k | 2 = 1, and its corresponding beamforming weight vector and transmission power are w k and P k , then the signal received by the kth user is:

[0050]

[0051] Among them, n 2,k is the additive white Gaussian noise, and h k is the channel vector between the geostationary satellite and the kth user. Consider the following geostationary satellite channel model:

[0052]

[0053] Among them, G R is the reception gain of the user, ρ k is the channel fading coefficient of the kth user, following the shadowed Rice distribution; r k is the gain vector of the on-board antenna spot beam, r max is the maximum beam gain of the spot beam. Then the beam gain between user k and beam n can be calculated by the following formula:

[0054]

[0055] Among them, J1(·) and J3(·) are the first-order and third-order Bessel functions of the first kind respectively; u kn= 2.07123sinφ kn / sinφ 3dB where φ kn denotes the angle between the incident direction of the k user signals and the center of the point beam n, and φ 3dB is the angle corresponding to the 3 dB attenuation of the beam gain relative to the maximum gain.

[0056] is the steering vector between the satellite and the k-th user, and its l-th element can be expressed as:

[0057]

[0058] where c is the speed of light, f c is the carrier frequency, and d l represents the distance from the user to the l-th satellite antenna; e is the natural constant and j is the imaginary unit.

[0059] (3) Under the condition of ensuring the quality of service for high-altitude platform users, an optimization problem with the goal of maximizing the satellite network ergodic capacity is constructed, and the virtual signal-to-interference-plus-noise ratio (SINR) criterion is used to transform the optimization problem into a solvable form: 3.1) With the goal of maximizing the ergodic capacity and subject to the interference power from the satellite to the high-altitude platform users, an optimization model is established as follows:

[0060]

[0061] where w k is the beamforming weight vector for serving the k-th user, P k represents the transmit power for serving the k-th user; k represents the serial number of the served user, k = 1, …, K, where K is the number of users served by the satellite; h k represents the channel vector between the satellite and the k-th user, and σ 2 represents the variance of the additive white Gaussian noise. g l represents the channel vector between the satellite and the l-th high-altitude platform user, and γ th represents the interference power threshold from the satellite to the high-altitude platform user, l represents the serial number of the users served by the high-altitude platform, l = 1, …, L, where L is the number of users served by the high-altitude platform, and E[.] represents the mathematical expectation.

[0062] 3.2) Using Jensen's inequality to transform the objective function, we get:

[0063]

[0064] 3.3) Using the principle of virtual SINR, the problem of maximizing the ergodic capacity is transformed into the problem of maximizing the virtual SINR for each user:

[0065]

[0066] Among them, is the channel covariance matrix between the satellite and the k-th user, is the channel covariance matrix between the satellite and the l-th high-altitude platform user, is the available covariance matrix, ΔG l is the error of the channel covariance matrix, satisfying ||ΔG l || ≤ α l , μ k.j is the weighting coefficient, w k is the normalized beamforming weight vector for serving the k-th user, P k represents the transmit power for serving the k-th user; k represents the serial number of the served user, k = 1, …, K, K represents the number of users served by the satellite, σ 2 represents the variance of additive white Gaussian noise, g l represents the channel vector between the satellite and the l-th high-altitude platform user, γ th represents the interference power threshold of the satellite for high-altitude platform users, l represents the serial number of the user served by the high-altitude platform, l = 1, …, L, L represents the number of users served by the high-altitude platform, and E[.] represents the mathematical expectation.

[0067] (4) First, use the compact bound method to handle the channel error from the satellite to high-altitude platform users, and then use the generalized Rayleigh quotient theorem to solve for the beamforming weight vector and transmit power:

[0068] 4.1) For the constraint conditions, to maximize the left side of the inequality, the problem is reconstructed as:

[0069]

[0070] The Lagrangian dual function of the above problem can be written as:

[0071]

[0072] where λ and Z are Lagrangian variables. Taking the derivative of the above with respect to ΔG l gives Substituting it back into the Lagrangian dual function gives:

[0073]

[0074] The above can be equivalently written as

[0075]

[0076] Taking the derivative of the objective function of the above problem with respect to λ equal to 0 gives the optimal λ lFor Substituting it in, we can get:

[0077]

[0078] Because and Z l ≥ 0, so when the problem reaches the optimum, Z l = 0. Therefore, the original constraint can be transformed into:

[0079]

[0080] 4.2) Use the alternating iteration method to solve the above problem. First, fix the weighting coefficient μ k,j , and use the maximum Rayleigh quotient to solve the beamforming weight vector and transmit power, as shown in the following formula:

[0081]

[0082] Where A(φ k ) = a(φ k )a H (φ k ), I is the identity matrix, φ k is the arrival angle of the k-th user signal, c is the speed of light, G R is the antenna gain, f c is the carrier frequency, and d kn represents the distance between the k-th user and the center of the point beam n, r k is the gain vector of the spaceborne antenna point beam. ν max (.) represents the principal eigenvector of the matrix.

[0083] 4.3) Fix the beamforming weight vector w k , and use the gradient descent method to solve the weighting coefficient μ k,j . First, take the derivatives of steps 3.2) and 4.2) to obtain the following formula:

[0084]

[0085] Among them, D j = P j |a H (φ j )w j | 2 , σ 2 represents the variance of the additive white Gaussian noise, φ k is the arrival angle of the k-th user signal, c is the speed of light, G R is the antenna gain, fc is the carrier frequency and d kn represents the distance between the k-th user and the center of the point beam, r n is the gain vector of the spaceborne antenna point beam. The following equations are obtained by combining them: k is the vector all ones; Q

[0086]

[0087] The above equation can be further written as:

[0088]

[0089] where, 1 K-k is the vector all ones; Q k is the coefficient matrix corresponding to the k-th user and can be expressed as

[0090]

[0091] Using the above formula, the beamforming weight vector and the weighting coefficient are obtained by alternating iteration. The specific iterative algorithm is as follows:

[0092] a) Input {a(φ k ), δ, σ 2 )}, where δ is the iteration accuracy value, and σ 2 is the noise variance;

[0093] b) Initialize the counting variable t = 0;

[0094] c) Initialize the weighting coefficient

[0095] d) Iteration:

[0096] i: Update

[0097] using the maximum Rayleigh quotient method in step 3.4);

[0098] ii: Calculate iii: Update the weighting coefficient using the formula

[0099] iv: Make the following judgment: If then return to the previous step for loop iteration; otherwise, end the loop and stop the iteration;

[0100] f) Output the beamforming weight vector w k and the transmit power P k ;

[0101] (5) There is interference among users who simultaneously apply for network access, which will reduce the system ergodic capacity. Therefore, based on the above beamforming scheme, the geostationary satellite first performs beamforming on all signals and then sends them to the corresponding users, and conducts user scheduling according to the 1-bit information fed back by each user, searching for the optimal NOMA user combination and beamforming weight vector under the same time-frequency resource conditions. The specific steps of user scheduling are as follows:

[0102] 5.1) First, use the beamforming scheme in step (3) to calculate the beamforming weight vectors for serving each user and perform beamforming;

[0103] 5.2) Each user makes a judgment and gives a 1-bit feedback based on the received signal: if the communication requirements can be met, feedback "1"; otherwise, feedback "0";

[0104] 5.3) The geostationary satellite conducts user scheduling according to the obtained feedback information and selects the users that meet the communication requirements;

[0105] 5.4) Repeat steps 4.1)-4.3) until the system ergodic capacity no longer changes under the time-frequency resource conditions;

[0106] 5.5) Repeat the above process on other time-frequency resources until time-frequency resources are allocated to all users.

[0107] The technical means disclosed in the solution of the present invention are not limited to those disclosed in the above embodiments, but also include technical solutions composed of any combination of the above technical features. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements are also regarded as the protection scope of the present invention.

Claims

1. A high spectral efficiency transmission method suitable for a forward link of a starry sky cognitive network, characterized in that: Under the condition that the satellite network and the high-altitude platform network share spectrum resources, the two gateway stations with the best channel quality are selected for Alamouti space-time coding, and communicate with the geostationary orbit satellite through the terahertz frequency band. The satellite decodes the signals from the gateway station; the satellite determines the user-side decoding order according to the channel state information of the ground users, and completes the beamforming scheme design based on the virtual signal-to-interference-and-noise ratio criterion while ensuring the service quality of the high-altitude platform users; the geostationary orbit satellite performs beamforming processing on all signals and sends them to the corresponding users, and performs user scheduling according to the feedback information of each user to obtain the optimal NOMA user combination on the same time-frequency resources.

2. The high spectral efficiency transmission method suitable for the forward link of the starry sky cognitive network according to claim 1, characterized in that: The specific steps include: Step 1. Under the condition that the satellite network and the high-altitude platform network share spectrum resources, two gateways with the best channel quality are selected from multiple independent and unrelated gateways on the ground for Alamouti space-time coding, and further terahertz technology is used to communicate with the geostationary orbit satellite. The satellite uses a decoding and forwarding protocol to decode the signal from the gateway in the future; Step 2. The geostationary orbit satellite determines the user-side decoding order according to the user channel state information and performs superposition coding in the power domain; Step 3. Under the condition of ensuring the service quality of users on the high-altitude platform, an optimization problem with the goal of maximizing the satellite network traversal capacity is constructed, and the optimization problem is transformed into a solvable form using the virtual signal-to-interference-noise ratio criterion; Step 4. Use the compact bound method to process the channel error from the satellite to the high-altitude platform, and use the generalized Rayleigh quotient theorem to solve the beamforming weight vector and the transmit power; Step 5. Based on the proposed beamforming scheme, the satellite performs beamforming on all signals and sends them to each user. It also schedules users according to the 1-bit information fed back by each user to obtain the optimal NOMA user combination and beamforming weight vector under the same time-frequency resource conditions, thereby maximizing the system capacity.

3. The high spectral efficiency transmission method suitable for the forward link of the starry sky cognitive network according to claim 2, characterized in that: In step 1, the received signal of the geostationary orbit satellite is expressed as: Among them, y i represents the signal received by the geostationary satellite, I1 and I2 are the terahertz channel coefficients, s i is the transmitted data symbol vector, n i is additive white Gaussian noise, and i represents the gateway station number.

4. The high spectral efficiency transmission method applicable to the forward link of the starry sky cognitive network according to claim 2, characterized in that: The step 2 specifically includes: The satellite determines the decoding order on the user side according to the obtained channel gain and performs superposition coding on the power; the transmission signal of the geostationary orbit satellite is expressed as: Where x represents the transmission signal of the geostationary orbit satellite, w k is the normalized beamforming weight vector serving the kth user, P k represents the transmission power serving the kth user; k represents the serial number of the served user, k=1,…,K, K represents the number of users served by the satellite.

5. The high spectral efficiency transmission method applicable to the forward link of the starry sky cognitive network according to claim 2, characterized in that: The step 3 specifically includes: 4.1) First, with the goal of maximizing the satellite network traversal capacity and the satellite interference power to the high-altitude platform users as the constraint, an optimization model is established under the condition that only the imperfect CSI of the high-altitude platform users is known; then, a beamforming scheme based on the virtual signal-to-interference-noise ratio criterion is designed, and the beamforming weight vector w is obtained by iterative solution. k ; The optimization model constructed is shown as follows: Among them, w k is the normalized beamforming weight vector serving the kth user, P k represents the transmission power serving the kth user; k represents the serial number of the served user, k=1,...,K, K represents the number of users served by the satellite; h k represents the channel vector between the satellite and the kth user, σ 2 represents the variance of additive Gaussian white noise, g l represents the channel vector between the satellite and the lth high-altitude platform user, γ th represents the interference power threshold of the satellite to the high-altitude platform users, l represents the serial number of the user served by the high-altitude platform, l=1,...,L, L represents the number of users served by the high-altitude platform, and E[] represents the mathematical expectation; 4.2) Using the Qinsheng inequality to approximate the objective function, we get: 4.3) Using the principle of virtual signal to noise ratio, the ergodic capacity problem is transformed into the problem of maximizing the virtual signal to noise ratio of each user. The transformed problem is expressed as: in, is the channel covariance matrix from the satellite to the kth user, is the channel covariance matrix from the satellite to the lth high-altitude platform user, G l is the available covariance matrix, ΔG l is the error of the channel covariance matrix, satisfying ||ΔG l ||≤α l , μ k.j is the weighting coefficient, 6. The high spectral efficiency transmission method applicable to the forward link of the starry sky cognitive network according to claim 2, characterized in that: The step 4 specifically includes: 5.1) Using the compact bound method to process the channel error from the satellite to the high-altitude platform, the constraints are transformed into: 5.2) Fix the weighting coefficient and use the maximum Rayleigh quotient method to solve w k With P k : in, I is the identity matrix, φ k is the arrival angle of the kth user signal, c is the speed of light, G R is the antenna gain, f c is the carrier frequency and represents the distance between the kth user and the center of spot beam n, r k is the gain vector of the satellite antenna spot beam; 5.3) Fix the weight vector and use the gradient descent method to solve μ k,j : in, Q k is the coefficient matrix corresponding to the kth user, 1 K-k is a vector whose elements are all 1; 5.4) Alternately solve step 5.2) and step 5.3) to obtain the optimal beamforming weight vector.

7. The high spectral efficiency transmission method applicable to the forward link of the starry sky cognitive network according to claim 1, characterized in that: When a large number of users access the network at the same time, interference between users will reduce the system throughput. The system needs to make reasonable arrangements for users, namely user scheduling. In the space cognitive network, based on the proposed beamforming scheme, the satellite first beamforms all user information under the condition of meeting the service quality of high-altitude platform network users, and schedules users according to the 1-bit information fed back by each user, searching for the optimal NOMA user combination and beamforming weight vector on the same time-frequency resources to maximize the system capacity.

Images