Beamforming Optimization Method for Space-Based Internet of Things Access

Abstract

The present invention proposes a beamforming optimization method for space-based Internet of Things access. The method includes: receiving the signals received by the receiving satellite in the l-th time slot, and calculating the directions of arrival of all signals in the received collision signals; calculating the optimal pointing of the auxiliary beam generated in the m-th time slot, the optimal sidelobe level of the Bayliss window function, and the optimal auxiliary beam gain peak limit variable; generating the auxiliary beam pattern according to the beamforming parameters, the optimal pointing of the auxiliary beam, the optimal sidelobe level of the Bayliss window function, and the optimal auxiliary beam gain peak limit variable. The present invention effectively improves the capacity of the space-based communication system by increasing the number of main lobe roll-off bands and optimizing the receiving direction of the antenna pattern.

Description

Technical Field

[0001] The present invention belongs to the technical field of satellite communications, and in particular relates to a beamforming optimization method for space-based Internet of Things access. Background Art

[0002] Due to geographical conditions and economic cost constraints, ground-based IoT networks cannot provide communication services to remote areas such as deserts, oceans and forests. Low-orbit satellite IoT, with its global seamless coverage, strong anti-damage capability and low transmission delay, has become an important part of the future 6G integrated space, land, and sea intelligent connection of all things. However, due to the wide coverage of satellite IoT systems and the short burst characteristics of data collection services, terminals need to meet the characteristics of low power consumption and light control. This leads to a contradiction between terminal access constraints and the demand for large-capacity non-orthogonal multiple access, which makes it easy for a large number of data packets to collide, resulting in a rapid drop in system throughput.

[0003] There are many types of IoT services, but the ones suitable for satellite IoT services mainly include data collection and unmanned control services. At a transmission distance of hundreds of kilometers in low-orbit satellites, both power consumption and latency are significantly greater than those of ground networks, and a large number of data collection services have short burst characteristics. Conventional access methods based on resource reservation have extremely low resource utilization, while access methods based on dynamic applications have large signaling overhead, and latency and power consumption do not meet the requirements of satellite IoT. Random access technology seizes resources in a competitive manner, and users do not need to schedule, which improves system transmission efficiency and resource utilization, and can better adapt to satellite IoT short data packet burst services. However, due to the random transmission characteristics of random access terminals, when the load gradually increases, multiple terminals will access the system at the same time, and data packets at satellite access points will collide. However, the near-far effect of different terminals is not obvious, power control is not applicable, and collision signal separation conditions cannot be achieved, resulting in a sharp drop in actual throughput at medium and high loads. Therefore, how to create collision signal separation conditions, increase the probability of successful reception of collision data packets, and thus increase system throughput is a practical need to promote the application of satellite IoT. Summary of the invention

[0004] The main purpose of the present invention is to provide a beamforming optimization method for space-based Internet of Things access, which increases the number of main lobe roll-off bands, increases the success rate of collision signal separation as much as possible, improves system access performance, and effectively supports the large-capacity random access requirements of satellite Internet of Things terminals.

[0005] To achieve the above objectives, the present invention provides a beamforming optimization method for space-based Internet of Things access, comprising the following steps:

[0006] Step 1: The receiving end satellite The signal received in the time slot is ,in, , is the total number of time slots, and then collision signal detection is performed on each time slot. The collision signal received by the time slot with a detected collision signal is , where ;

[0007] Estimate the direction of arrival of all signals in the collision signal received by the satellite , where , is to detect the number of collision signals in the th time slot;

[0008] Step 2: According to the direction of arrival of the collision signal estimated in Step 1 , calculate the optimal pointing of the auxiliary beam generated in the th time slot , the optimal sidelobe level of the Bayliss window function and the optimal auxiliary beam gain peak limit variable ;

[0009] Step 3: According to the receiving beamforming parameters of the system main channel and the optimal pointing of the auxiliary beam obtained in Step 2 , the optimal sidelobe level of the Bayliss window function and the optimal auxiliary beam gain peak limit variable generate the auxiliary beam pattern ;

[0010] A further improvement of the present invention is that Step 1 further includes the following steps:

[0011] Step 11: According to the snapshot data received by the receiving beam array of the system main channel, sampling the received signal gives:

[0012] ,

[0013] where is the array data vector, , is the array noise vector, ; is the signal complex envelope vector, , is the th complex envelope of the signal source, is the array manifold matrix, , where is the th steering vector of the signal source;

[0014] Step 12, according to received signal vectors, the estimated value of the covariance matrix is:

[0015] ;

[0016] wherein, is the array data vector, denotes the conjugate transpose of the matrix;

[0017] Step 13, calculate the beamforming weight according to the covariance matrix calculated in Step 12 as shown in the following formula:

[0018]

[0019] wherein, is the estimated value of the covariance matrix, is the steering vector, is the direction-of-arrival estimation variable, denotes the inverse of the matrix, denotes the conjugate transpose of the matrix;

[0020] Step 14, the array spatial spectrum function is:

[0021]

[0022] wherein, is the estimated value of the covariance matrix, is the steering vector, is the direction-of-arrival estimation variable, denotes the inverse of the matrix, denotes the conjugate transpose of the matrix, and perform a traversal search on to find the spectral peak ;

[0023] Step 15, through the search for the spectral peak in Step 14, obtain the direction-of-arrival estimates of all signals in the collision signal .

[0024] A further improvement of the present invention lies in that, in Step 14, the specific steps of performing a traversal search on to find the spectral peak are as follows:

[0025] Step a, traverse with a step size of and calculate the value of the array spatial spectrum corresponding to the th ​ ;

[0026] Step b, set the decision threshold of the spatial spectrum to , if , then execute Step c, otherwise, is a non-spectrum peak;

[0027] Step c, successively compare the spatial spectrum values of the three adjacent values 、 and that exceed the decision threshold in Step b. If and , then is an inflection point, is a spectrum peak, and output the current , otherwise, is not an inflection point, is a non-spectrum peak.

[0028] A further improvement of the present invention is that Step 2 further includes the following steps:

[0029] Step 21, according to the direction of arrival of the collision data packet and by means of the intermediate value of the power and power gain ratio, calculate the carrier-to-interference-plus-noise ratio of the collision signal in the auxiliary channel, expressed as:

[0030]

[0031] where the received power gain of the auxiliary beam is , is the optimal auxiliary beam gain peak limit variable, is the auxiliary beam steering vector, is the auxiliary beam array weight vector, is the conventional beam steering vector, is the beam pointing of the auxiliary beam, represents the conjugate transpose of the matrix; is the received gain of the th relatively strong collision signal in the auxiliary beam, is the received gain of the th relatively weak collision signal in the auxiliary beam; is the estimated value of the auxiliary channel noise power; the received gain of the main beam is , is the direction of arrival of the desired signal of the received beam in the main channel, is the received gain of the th collision signal in the main beam, is the The reception gain of a collision signal; For the corresponding th collision signal in the main channel, the estimated signal-to-noise ratio, For the corresponding th collision signal in the main channel, the estimated signal-to-noise ratio, Is the estimated value of the noise power of the collision signal in the main channel; Is the weight expression of the Bayliss window function, specifically as follows:

[0032]

[0033] Among them, The expression is specifically as follows:

[0034]

[0035] Among them, The expression is specifically as follows:

[0036]

[0037]

[0038] Is an intermediate variable used to calculate the weight. Among them, Is the number of equal sidelobe levels, Is the number of array elements, , where Is the sidelobe level. Bayliss gives the coefficients of a fourth-order polynomial, Are the polynomial coefficients;

[0039] Step 22: Taking the maximum of the carrier-to-interference-plus-noise ratio of the collision signal in Step 21 as the goal, establish a mathematical model to solve the optimal pointing of the th time slot auxiliary beam, the optimal sidelobe level of the Bayliss window function, and the optimal auxiliary beam gain peak limit variable , expressed as:

[0040]

[0041]

[0042]

[0043]

[0044]

[0045] Among them, is the number of time slot collision data packets in the th time slot, is the set of collision signal pair numbers, is the demodulation threshold, is the main lobe width of the conventional beam; is the lower bound of the constraint on the sidelobe level of the auxiliary beam R , is the upper bound of the constraint on the sidelobe level of the auxiliary beam , is the lower bound of the constraint on the peak gain limit variable of the auxiliary beam , is the upper bound of the constraint on the peak gain limit variable of the auxiliary beam . The optimization problem aims at maximizing the carrier-to-interference-plus-noise ratio (CINR) of the strong signal. Constraint 1 is the constraint condition for the CINR of the strong signal, that is, the CINR of the strong signal needs to be greater than the demodulation threshold. Constraint 2 is the constraint condition for the direction of the auxiliary beam, that is, the change in the direction of the auxiliary beam needs to be within the main lobe width range. Constraint 3 is the constraint condition for the sidelobe level of the auxiliary beam. Constraint 4 is the constraint condition for the peak power gain of the auxiliary beam, and its peak gain is restricted to be less than the peak power gain of the main beam;

[0046] Step 23: For the mathematical model established in Step 22, introduce the particle swarm optimization algorithm to solve it.

[0047] A further improvement of the present invention lies in that Step 23 is further implemented by the following steps:

[0048] Step S1: By default, the particle swarm optimization algorithm is a minimization problem. Map the negative value of the collision signal carrier-to-interference-plus-noise ratio in Step 21 to the objective function ;

[0049] Step S2: Initialize the particle swarm, set the size of the particle swarm to , the maximum number of iterations to . Each particle in the particle swarm includes three variables, namely the direction of the auxiliary beam , the sidelobe level of the auxiliary beam and the peak gain limit variable of the auxiliary beam . Among them, represents the number of the current particle, represents the iteration number of the particle, , , and create the initial particle swarm ;

[0050] Step S3: In Step S2, the direction of the auxiliary beam The constraint for is , where is the main lobe width of the conventional beam, and the side lobe level of the auxiliary beam The constraint for is , is the lower limit of the constraint of the side lobe level of the auxiliary beam R , is the upper limit of the constraint of the side lobe level of the auxiliary beam , is the variable of the peak gain limit of the auxiliary beam The lower limit of the constraint is the variable of the peak gain limit of the auxiliary beam The upper limit of the constraint, and the negative value of the carrier-to-interference-plus-noise ratio of the collision signal in step S1 is the objective function;

[0051] Step S4, initialize the position and velocity of each particle in the initial particle swarm created in step S2 , and calculate the fitness function value corresponding to each particle ;

[0052] Step S5, according to the fitness value of the particle obtained in step S4 is , initialize the local optimal position and the global optimal position of the particle;

[0053] Step S6, calculate the velocity and position of each particle in the iterative process of step S5. The velocity update formula is as follows:

[0054]

[0055] where is the inertia weight, which controls the search range of the particle; is the local learning factor, which controls the step size of the particle approaching its own local optimal position; is the global learning factor, which controls the step size of the particle approaching the global optimal position; are two random numbers, which enhance the diversity of the search; the position update formula is as follows:

[0056]

[0057] Step S7, calculate the fitness value at the latest position of each particle under the conditions of step S5 and step S6;

[0058] Step S8, if the current iteration number is less than the maximum iteration number , set the termination threshold , according to the fitness function value calculated from the initial particle swarm in step S4 and the fitness function value calculated from the current new generation of particle swarm in step S7 , determine whether the iteration converges: if , the iteration does not converge, execute step S9, otherwise, the iteration converges, execute step S10; if the iteration number reaches the maximum value , the iteration terminates, execute step S10, where the termination threshold is selected in the order of magnitude of , indicating taking the absolute value;

[0059] Step S9, let the fitness function value , and return to step S5 to execute again;

[0060] Step S10, output the negative value of the optimal fitness function value in the current particle swarm , that is the corresponding optimal individual .

[0061] A further improvement of the present invention is that step 3 further includes the following steps:

[0062] Step 31, according to the beamforming parameters received by the main channel of the system and the beam direction of the auxiliary beam obtained in step 2 and the sidelobe level of the Bayliss window function , obtain the array weight vector of the auxiliary beam ;

[0063] Step 32, according to the auxiliary beam gain peak limit variable obtained in step 22 and the optimal array weight vector of the auxiliary beam obtained in step 31, obtain the optimal auxiliary beam pattern , expressed as:

[0064] ,

[0065] where is the optimal auxiliary beam gain peak limit variable, is the auxiliary beam steering vector, , is the auxiliary beam array weight vector, is the conventional beam steering vector, is the beam direction of the auxiliary beam, , indicating taking the absolute value, Denotes the conjugate transpose of a matrix.

[0066] Advantages of the present invention: On the one hand, the present invention utilizes the advantages of the auxiliary beam to create a difference in the signal-to-noise ratio of the received signal without power control, enabling multi-user detection in the satellite Internet of Things scenario; on the other hand, the present invention utilizes the edge roll-off band of the main lobe of the auxiliary beam to achieve a difference in the signal-to-noise ratio of the received signal, enabling it to meet the power domain separation condition, improving the successful separation probability of collision signals, enhancing the system access throughput, and reducing the system access packet loss rate. Description of the Drawings

[0067] Figure 1 Is the flow block diagram for optimizing the pointing of the auxiliary receiving beam of the method of the present invention.

[0068] Figure 2 Is the auxiliary beam pattern optimized based on the Bayliss window function in the present invention.

[0069] Figure 3 Is the system throughput performance diagram under different angle measurement errors in the present invention.

[0070] Figure 4 Is the system throughput performance diagram under different amplitude-phase errors in the present invention. Detailed Implementation Manner

[0071] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be described in detail below with reference to the drawings and specific embodiments.

[0072] Here, it should be noted that in order to avoid obscuring the present invention with unnecessary details, only the structures and / or processing steps closely related to the solution of the present invention are shown in the drawings, while other details less related to the present invention are omitted.

[0073] In addition, it should also be noted that the term "comprises", "comprising", or any other variation thereof is intended to cover a non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or device.

[0074] The beamforming optimization method for space-based Internet of Things access of the present invention mainly includes the following steps:

[0075] Step 1, the signal received by the receiving end satellite in the th time slot is , where , is the total number of time slots, and then collision signal detection is performed on each time slot. The collision signal received by the time slot with a detected collision signal is , where ;

[0076] Estimate the direction of arrival of all signals in the collision signal received by the satellite , where , is the number of collision signals detected in the th time slot;

[0077] Step 2: According to the direction of arrival of the collision signal estimated in Step 1 , calculate the optimal pointing of the auxiliary beam generated in the th time slot, the optimal sidelobe level of the Bayliss window function, and the optimal auxiliary beam gain peak limit variable ;

[0078] Step 3: According to the receiving beamforming parameters of the system main channel and the optimal pointing of the auxiliary beam obtained in Step 2 , the optimal sidelobe level of the Bayliss window function, and the optimal auxiliary beam gain peak limit variable generate the auxiliary beam pattern

[0079] Combined with Figures 1 to 4 to describe the present invention in detail.

[0080] Step 1: The signal received by the receiving satellite in the th time slot is , where , is the total number of time slots, and then collision signal detection is performed on each time slot. The collision signal received by the time slot with a detected collision signal is , where ;

[0081] Estimate the direction of arrival of all signals in the collision signal received by the satellite , where , is the number of collision signals detected in the th time slot

[0082] In practical applications, there are various methods for estimating the direction of arrival (DOA), such as the Capon algorithm, MUSIC algorithm, etc. In step 1, the Capon algorithm is used as an example to illustrate the specific process of DOA estimation. Under conditions allowing, other methods can also be used for DOA estimation.

[0083] The specific steps in step 1 are as follows:

[0084] Step 11, the received beam of the system main channel is , and based on the snapshots of data received by this received beam array, the received signal can be sampled as follows:

[0085] ,

[0086] where is the array data vector, , is the array noise vector, ; is the signal complex envelope vector, , is the th complex envelope of the signal source, is the array manifold matrix, , where is the th steering vector of the signal source;

[0087] Step 12, based on the received signal vectors, the estimated value of the covariance matrix is:

[0088] ;

[0089] where is the array data vector, represents the conjugate transpose of the matrix;

[0090] Step 13, calculate the beamforming weight as shown in the following formula:

[0091]

[0092] where is the estimated value of the covariance matrix, is the steering vector, is the DOA estimation variable, represents the inverse of the matrix, represents the conjugate transpose of the matrix;

[0093] Step 14, the array spatial spectrum function is:

[0094]

[0095] where, is the estimated value of the covariance matrix, is the steering vector, is the direction of arrival (DOA) estimation variable, represents the inverse of the matrix, represents the conjugate transpose of the matrix. Perform a traversal search on to find the spectral peak ;

[0096] Step 15, through the search for the spectral peak in Step 14, obtain the direction of arrival estimates of all signals in the collision signal . In Step 14, the specific steps for performing a traversal search on to find the spectral peak are as follows:

[0097] Step a, traverse with a step size of and calculate the value of the array spatial spectrum corresponding to the rd ; ;

[0098] Step b, set the decision threshold of the spatial spectrum to . If , then execute Step c; otherwise, is a non-spectral peak value;

[0099] Step c, compare the three adjacent values , , and of the spatial spectrum value exceeding the decision threshold in Step b in sequence. If and , then is an inflection point, is the spectral peak value, and output the current ; otherwise, is not an inflection point,

[0100] Step 2, based on the direction of arrival of the collision signal estimated in Step 1 , calculate the optimal pointing of the auxiliary beam generated in the th time slot, the optimal sidelobe level of the Bayliss window function, and the optimal auxiliary beam gain peak limit variable​ 。

[0101] In practical applications, there are various methods for solving optimization problems, such as genetic algorithms, particle swarm optimization algorithms, ant colony algorithms, etc. In this step, the particle swarm optimization algorithm is used as an example to illustrate the specific process of solving the optimal direction of the auxiliary beam and the maximum carrier-to-interference-plus-noise ratio of the collision signal. Under condition that permits, other methods can also be used for solving.

[0102] The specific implementation of this step is as follows:

[0103] Step 21, according to the direction of arrival of the collision data packet and calculate the carrier-to-interference-plus-noise ratio of the collision signal in the auxiliary channel by means of the intermediate value of the power and power gain ratio , which can be expressed as:

[0104]

[0105] Among them, the receiving power gain of the auxiliary beam is , is the peak limit variable of the optimal auxiliary beam gain, is the steering vector of the auxiliary beam, is the array weight vector of the auxiliary beam, is the conventional beam steering vector, is the beam direction of the auxiliary beam, represents the conjugate transpose of the matrix; is the receiving gain of the th relatively strong collision signal in the auxiliary beam, is the receiving gain of the th relatively weak collision signal in the auxiliary beam. is the estimated value of the auxiliary channel noise power. The receiving gain of the main beam is , is the direction of arrival of the desired signal of the receiving beam in the main channel, is the receiving gain of the th collision signal in the main beam, is the receiving gain of the th collision signal in the main beam. is the estimated signal-to-noise ratio corresponding to the th collision signal in the main channel, is the estimated signal-to-noise ratio corresponding to the th collision signal in the main channel, is the estimated value of the collision signal noise power in the main channel. is the weight expression of the Bayliss window function, which is specifically as follows:

[0106]

[0107] Among them, the expression is specifically as follows:

[0108]

[0109] Among them, the expression is specifically as follows:

[0110]

[0111]

[0112] is an intermediate variable used to calculate the weight value. Among them, is the number of equal sidelobe levels, is the number of array elements, , where is the sidelobe level. Bayliss gives the coefficients of a fourth-order polynomial as shown in Table 1:

[0113] Table 1 Coefficients and sidelobe levels of the Bayliss pattern Coefficients and sidelobe levels

[0114]

[0115] Among them, is the polynomial coefficient, and the corresponding coefficients are shown in Table 2:

[0116] Table 2 Polynomial coefficients

[0117]

[0118] Step 22: Taking the maximum of the carrier-to-interference-plus-noise ratio of the collision signal in Step 21 as the goal, establish a mathematical model to solve the optimal pointing of the auxiliary beam in the -th time slot, the optimal sidelobe level of the Bayliss window function, and the optimal auxiliary beam gain peak limit variable , which can be expressed as:

[0119]

[0120]

[0121]

[0122]

[0123]

[0124] Among them, is the number of time slot collision data packets in the th time slot, is the set of collision signal pair numbers, is the demodulation threshold, is the main lobe width of the conventional beam; is the lower bound of the constraint on the sidelobe level of the auxiliary beam R , is the upper bound of the constraint on the sidelobe level of the auxiliary beam , is the lower bound of the constraint on the peak gain limit variable of the auxiliary beam , is the upper bound of the constraint on the peak gain limit variable of the auxiliary beam . The optimization problem aims to maximize the carrier-to-interference-plus-noise ratio (CINR) of the strong signal. Constraint 1 is the constraint condition for the CINR of the strong signal, that is, the CINR of the strong signal needs to be greater than the demodulation threshold. Constraint 2 is the constraint condition for the direction of the auxiliary beam, that is, the change in the direction of the auxiliary beam needs to be within the main lobe width range. Constraint 3 is the constraint condition for the sidelobe level of the auxiliary beam. Constraint 4 is the constraint condition for the peak power gain of the auxiliary beam, and its peak gain is restricted to be less than the peak power gain of the main beam;

[0125] Step 23. For the mathematical model established in Step 22, a particle swarm optimization algorithm is introduced for solution, including the following steps:

[0126] Step S1. By default, the particle swarm optimization algorithm is a minimization problem. The negative value of the collision signal CINR in Step 21 is mapped to the objective function ;

[0127] Step S2. Initialize the particle swarm, set the size of the particle swarm to , the maximum number of iterations to . Each particle in the particle swarm includes three variables, namely the direction of the auxiliary beam , the sidelobe level of the auxiliary beam and the peak gain limit variable of the auxiliary beam . Among them, represents the number of the current particle, represents the iteration number of the particle, , , and create the initial particle swarm ;

[0128] Step S3. The constraint on the direction of the auxiliary beam in Step S2 is , where is the main lobe width of the conventional beam and the sidelobe level of the auxiliary beam The constraint for is that the peak gain limit variable of the auxiliary beam The constraint for , is the sidelobe level of the auxiliary beam R The lower limit of the constraint for is the sidelobe level of the auxiliary beam The upper limit of the constraint for is the peak gain limit variable of the auxiliary beam The lower limit of the constraint for is the peak gain limit variable of the auxiliary beam The upper limit of the constraint for. The negative value of the carrier-to-interference-plus-noise ratio of the collision signal in step S1 is the objective function;

[0129] Step S4, initialize the positions of each particle in the initial particle swarm created in step 2 and velocities , and calculate the fitness function value corresponding to each particle ;

[0130] Step S5, according to the fitness values of the particles obtained in step S4, which are , initialize the local optimal position of the particle and the global optimal position ;

[0131] Step S6, calculate the velocities and positions of each particle during the iteration process of step S5. The velocity update formula is as follows:

[0132]

[0133] Among them, is the inertia weight, which controls the search range of the particle; is the local learning factor, which controls the step size for the particle to approach its local optimal position; is the global learning factor, which controls the step size for the particle to approach the global optimal position; are two random numbers, which enhance the diversity of the search. The position update formula is as follows:

[0134]

[0135] Step S7, calculate the fitness value at the latest position of each particle under the conditions of step S5 and step S6 ;

[0136] Step S8, if the current iteration number is less than the maximum iteration number Set a termination threshold and the fitness function value calculated based on the initial particle swarm in step S4 and the fitness function value calculated for the current new generation of particle swarm in step S7 to determine whether the iteration converges: If , the iteration does not converge, execute step S9; otherwise, the iteration has converged, execute step S10; if the number of iterations reaches the maximum value , the iteration terminates, execute step S10, where the termination threshold is selected at the order of magnitude of , where | | represents taking the absolute value;

[0137] Step S9: Let the fitness function value , and return to step S5 to execute again;

[0138] Step S10: Output the negative value of the optimal fitness function value in the current particle swarm, that is , and the corresponding optimal individual . .

[0139] Step 3: Generate an auxiliary beam pattern based on the receiving beamforming parameters of the system main channel and the optimal pointing of the auxiliary beam obtained in step 2 , the optimal sidelobe level of the Bayliss window function and the optimal auxiliary beam gain peak limit variable . .

[0140] Step 3 further includes the following steps:

[0141] Step 31: Obtain the array weight vector of the auxiliary beam based on the receiving beamforming parameters of the system main channel and the beam pointing of the auxiliary beam obtained in step 22 , the sidelobe level of the Bayliss window function . ;

[0142] Step 32: Obtain the optimal auxiliary beam pattern based on the optimal auxiliary beam gain peak limit variable obtained in step 22 and the optimal array weight vector of the auxiliary beam obtained in step 31, which can be expressed as:

[0143] ,

[0144] where is the optimal auxiliary beam gain peak limit variable, is the auxiliary beam steering vector, ,​ is the auxiliary beam array weight vector, is the conventional beam steering vector, is the beam direction of the auxiliary beam, represents taking the absolute value, represents the conjugate transpose of the matrix.

[0145] The effect of the present invention can be further verified by the following simulations.

[0146] Experimental scenario:

[0147] Set the number of array elements of the Balissy window function to 32, the number of zeros in the sidelobe region to 4, and the sidelobe level to 30; set the carrier frequency of the scenario to 2 GHz, the bandwidth to 20 kHz, the number of array elements of the conventional beam to 32, the element spacing to half of the wavelength, the satellite-ground link distance to 1000 km, the equivalent noise temperature to 290 K, the terminal transmission power to -10 dBW, the terminal transmission gain to 0 dBi, and the carrier-to-interference-plus-noise ratio thresholds for the collision signals to be separable to 8 dB, 9 dB, and 10 dB respectively.

[0148] Experimental content and results:

[0149] Experiment 1, generate the auxiliary beam pattern optimized based on the weights of the Bayliss window function.

[0150] Figure 2 is the auxiliary beam pattern generated after optimizing the weights of the Bayliss window function. It can be seen from the simulation results that, compared with the conventional beam pattern, the antenna pattern optimized by the Bayliss window function forms a depression in the specified direction, which can increase the power difference between the useful signal and the collision signal, so that there is a greater probability that the collision signal can be successfully separated and received.

[0151] Experiment 2, verify the system throughput performance under different angle measurement errors.

[0152] Figure 3 is the performance of the system throughput. It can be seen from the simulation results that, compared with the traditional slotted ALOHA random access technology, introducing the auxiliary beam optimization method based on the weight optimization of the Bayliss window function to separate the collision signals can effectively improve the system throughput, and there is still a greater probability that the collision signals can be successfully separated and received in the presence of angle measurement errors.

[0153] Experiment 3, verify the system throughput performance under different amplitude-phase errors.

[0154] Figure 4It is the performance of system throughput. It can be seen from the simulation results that, compared with the traditional slotted ALOHA random access technology, introducing the auxiliary beam optimization method based on the weight optimization of the Bayliss window function to separate the collision signals can effectively improve the system throughput, and there is still a high probability that the collision signals can be successfully separated and received in the presence of angle measurement errors.

[0155] The above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A beamforming optimization method for space-based Internet of Things access, characterized in that, Including the following steps: Step 1, the signal received by the receiving end satellite in the th time slot is , where ， is the total number of time slots. Then, collision signal detection is performed on each time slot, and the collision signal received by the time slot with a detected collision signal is , where ; Estimate the collision signals received by the satellite The direction of arrival of all signals in , where ， is to detect the number of collision signals in the th time slot; Step 2, based on the direction of arrival of the collision signal estimated in Step 1 , calculate the optimal pointing of the auxiliary beam generated in the th time slot , the optimal sidelobe level of the Bayliss window function and the optimal auxiliary beam gain peak limit variable ; Step 2 further includes the following steps: Step 21, according to the direction of arrival of the collision data packet and calculate the carrier-to-interference-plus-noise ratio of the collision signal in the auxiliary channel by means of the intermediate value of the power and power gain ratio , expressed as: Among them, the auxiliary beam receiving power gain is , is the optimal auxiliary beam gain peak limit variable, is the auxiliary beam steering vector, is the auxiliary beam array weight vector, is the conventional beam steering vector, is the beam direction of the auxiliary beam, represents the conjugate transpose of the matrix; is the reception gain of the th power-stronger collision signal in the auxiliary beam, is the reception gain of the th power-weaker collision signal in the auxiliary beam; is the estimated value of the auxiliary channel noise power; the main beam receiving gain is , is the incoming wave direction of the desired signal of the main channel receiving beam, is the reception gain of the th collision signal in the main beam, is the reception gain of the th collision signal in the main beam; is the signal-to-noise ratio estimated for the th corresponding collision signal in the main channel, is the signal-to-noise ratio estimated for the th corresponding collision signal in the main channel, is the estimated value of the collision signal in the main channel noise power; is the weight expression of the Bayliss window function, specifically as follows: Among them, The expression is specifically as follows: Among them, The expression is specifically as follows: is an intermediate variable used to calculate the weight value, where, is the number of equal sidelobe levels, is the number of array elements, , where is the sidelobe level. Bayliss gave the coefficients of a fourth-order polynomial, are the polynomial coefficients; Step 22, taking the maximum of the carrier-to-interference-plus-noise ratio of the collision signal in Step 21 as the objective, establish a mathematical model to solve for the optimal pointing of the auxiliary beam in the nth time slot, the optimal sidelobe level of the Bayliss window function and the optimal auxiliary beam gain peak limit variable , which is expressed as: , denoted as: Among them, is the number of time slot collision data packets in the th time slot, is the set of collision signal pair numbers, is the demodulation threshold, is the main lobe width of the conventional beam; is the lower bound of the constraint on the sidelobe level of the auxiliary beam R , is the upper bound of the constraint on the sidelobe level R of the auxiliary beam, is the lower bound of the constraint on the auxiliary beam gain peak limit variable , is the upper bound of the constraint on the auxiliary beam gain peak limit variable . The optimization problem aims to maximize the carrier-to-interference-plus-noise ratio (CINR) of the strong signal. Constraint 1 is the limitation condition of the CINR of the strong signal, that is, the CINR of the strong signal needs to be greater than the demodulation threshold. Constraint 2 is the limitation condition of the direction of the auxiliary beam, that is, the change in the direction of the auxiliary beam needs to be within the main lobe width range. Constraint 3 is the limitation condition of the sidelobe level of the auxiliary beam. Constraint 4 is the limitation condition of the peak power gain of the auxiliary beam, and its gain peak is restricted to be less than the peak power gain of the main beam; Step 23, for the mathematical model established in Step 22, introduce the particle swarm optimization algorithm to solve it; Step 3: Generate the auxiliary beam pattern based on the beamforming parameters received by the system main channel and the optimal pointing of the auxiliary beam obtained in Step 2 , the optimal sidelobe level of the Bayliss window function , and the optimal auxiliary beam gain peak limit variable . .

2. The beamforming optimization method for space-based Internet of Things access according to claim 1, characterized in that: Step 1 further includes the following steps: Step 11, the main channel receiving beam of the system is , according to the snapshot data received by this uplink receiving beam array, sampling the received signal gives: ， Among them, is the array data vector, , is the array noise vector, ; is the signal complex envelope vector, , is the th complex envelope of the th signal source, where is the th steering vector of the th signal source; Step 12, according to received signal vectors, the estimated value of the covariance matrix is: ； Among them, is the array data vector, indicating the matrix conjugate transpose; Step 13: Calculate the beamforming weights based on the covariance matrix obtained in Step 12 As shown in the following formula: Among them, is the estimated value of the covariance matrix, is the steering vector, is the direction-of-arrival estimation variable, represents the inverse of the matrix, represents the conjugate transpose of the matrix; Step 14, the array spatial spectrum function is: Among them, is the estimated value of the covariance matrix, is the steering vector, is the direction-of-arrival estimation variable, represents the inverse of the matrix, represents the conjugate transpose of the matrix, and performs a traversal search to find the spectral peak ; Step 15, through the search for spectral peaks in Step 14 to obtain the direction-of-arrival estimates of all signals in the collision signal .

3. The beamforming optimization method for space-based Internet of Things access according to claim 2, wherein: Step 14 further includes the following steps: Step a, with a step size traverse and calculate the value of the array spatial spectrum corresponding to the th ; ; Step b, set the decision threshold of the spatial spectrum to ; if , then execute step c, otherwise, is a non-spectrum peak value; Step c, successively take the spatial spectrum values exceeding the decision threshold in step b of the three adjacent values , and for comparison. If and , then is an inflection point, is the spectral peak value, and output the current . Otherwise, is not an inflection point, is not the spectral peak value.

4. The beamforming optimization method for space-based Internet of Things access according to claim 1, characterized in that: Step 23 is further implemented by the following steps: Step S1, the particle swarm optimization algorithm defaults to a minimization problem. Map the negative value of the carrier-to-interference-plus-noise ratio of the collision signal in step 21 to the objective function ; ; Step S2, initialize the particle swarm, set the size of the particle swarm to , and the maximum number of iterations to . Each particle in the particle swarm includes three variables, namely the auxiliary beam pointing , the auxiliary beam sidelobe level and the variable for limiting the peak gain of the auxiliary beam . Among them, represents the number of the current particle, represents the iteration number of the particle, , , and create the initial particle swarm ; Step S3, the constraint on the pointing direction of the auxiliary beam in Step S2 is where , and is the main lobe width of the conventional beam, and the constraint on the sidelobe level of the auxiliary beam is . The constraint on the auxiliary beam gain peak limit variable is . is the lower limit of the constraint on the sidelobe level R of the auxiliary beam, . is the lower limit of the constraint on the auxiliary beam gain peak limit variable . is the upper limit of the constraint on the auxiliary beam gain peak limit variable . The negative value of the carrier-to-interference-plus-noise ratio of the collision signal in Step S1 is the objective function; Step S4, initialize the positions of each particle in the initial particle swarm created in Step 2 and the velocities of each particle , and calculate the fitness function value corresponding to each particle ; Step S5. Based on the fitness value of the particles obtained in step S4, , initialize the local optimal position of the particles and the global optimal position ; Step S6, calculate the velocity and position of each particle during the iteration process of step S5 and position , and the velocity update formula is as follows: Among them, is the inertial weight, which controls the search range of the particle; is the local learning factor, which controls the step size for the particle to approach its own local optimal position; is the global learning factor, which controls the step size for the particle to approach the global optimal position; are two random numbers, which enhance the diversity of the search; The formula for position update is as follows: ; Step S7, calculate the fitness value at the latest position of each particle under the conditions of Step S5 and Step S6 ; Step S8, if the current iteration number is less than the maximum iteration number , set the termination threshold , based on the fitness function value calculated from the initial particle swarm in step S4 and the fitness function value calculated from the current new generation of particle swarm in step S7 , determine whether the iteration converges: if , then the iteration does not converge, execute step S9, otherwise, the iteration has converged, execute step S10; if the iteration number reaches the maximum value , then the iteration terminates, execute step S10, where the termination threshold is selected in the order of magnitude of indicating taking the absolute value; Step S9, set the fitness function value , and return to step S5 to execute again; Step S10, output the negative value of the optimal fitness function value in the current particle swarm, that is, the corresponding optimal individual .​ 5. The beamforming optimization method for space-based Internet of Things access according to claim 1, wherein: Step 3 further includes the following steps: Step 31: Obtain the array weight vector of the auxiliary beam based on the beamforming parameters received by the system main channel and the beam direction of the auxiliary beam obtained in Step 22 and the sidelobe level of the Bayliss window function to obtain the array weight vector of the auxiliary beam ; Step 32, according to the auxiliary beam gain peak limit variable obtained in Step 22 and the optimal auxiliary beam array weight vector obtained in Step 31, obtain the optimal auxiliary beam pattern , which is expressed as: ， Among them, is the optimal auxiliary beam gain peak limit variable, is the auxiliary beam steering vector, , is the auxiliary beam array weight vector, is the conventional beam steering vector, is the beam direction of the auxiliary beam, represents taking the absolute value, represents the conjugate transpose of the matrix.

Images