A Low-Earth Orbit Constellation Beam Resource Allocation Method Based on Adaptive Evolutionary Algorithm

Abstract

This invention discloses a method for low-Earth orbit (LEO) constellation beam resource allocation based on an adaptive evolutionary algorithm. The implementation steps are as follows: A trained user demand prediction neural network is used to obtain user demand prediction results, anticipating fluctuations in user demand and improving the real-time performance and accuracy of resource allocation. Based on a preference list, the Gale-Shapley algorithm is used to match satellite beams and users in beam overlap areas. An objective function and resource allocation constraints are defined according to a satellite communication system model. Based on these, an adaptive evolutionary algorithm is proposed to jointly allocate power and bandwidth, achieving optimized beam resource allocation. This invention improves user service coverage and enhances the resource utilization rate of LEO satellites by matching the capacity provided by satellites with the capacity demanded by users.

Description

Technical Field

[0001] This invention belongs to the field of communication technology, and more specifically relates to a method for allocating low-Earth orbit (LEO) constellation beam resources based on an adaptive evolutionary algorithm in the field of satellite communication technology. This invention can be used to improve service coverage for giant LEO constellations catering to heterogeneous user needs. Based on user traffic demands, it dynamically manages the bandwidth and power resources of the constellation beams using an optimization algorithm, matching the provided traffic with the traffic required by users to achieve the goal of improving service coverage. Background Art

[0002] With the ever-increasing demand for global communication services, building mega-satellite constellation networks has become a key development trend for improving system performance and meeting user needs. Low-Earth orbit (LEO) satellites, with their near-Earth orbit advantages, can provide high-bandwidth, low-latency services to global users, demonstrating enormous application potential in fields such as the Internet of Things (IoT), intelligent transportation, telemedicine, and emergency communications. However, due to the high angular velocity and high-speed movement of LEO satellites, users within the coverage area constantly switch over time, and the distribution of user terminals within the satellite coverage area is usually uneven. This uneven distribution leads to significant differences in user demand across different beams, especially during peak hours, where some areas may face over-utilization of resources while others cannot fully utilize available resources. This unbalanced user distribution and dynamic traffic demands pose challenges to satellite communication in providing services. Simultaneously, the increased scale of satellites and users brings problems of high computational complexity and low resource utilization. Therefore, designing effective beam resource management strategies to match the provided traffic with the user's required traffic, thereby improving service coverage while ensuring onboard resource utilization, has become one of the most important problems that current satellite communication systems urgently need to solve.

[0003] The Xi'an Space Radio Technology Research Institute disclosed a multi-beam control method for low-Earth orbit (LEO) mobile communication satellite constellations in its patent application, "A Multi-Beam Control Method for LEO Mobile Communication Satellite Constellations" (application date: June 18, 2021, application number: CN 202110678636.1, authorization announcement number: CN113612512 B). Based on the characteristics of LEO satellite constellations, this method proposes a dynamic multi-beam control approach that can adjust the beam direction, power allocation, and spectrum resources of each satellite according to dynamic factors such as satellite position and communication time periods. During the operation of the satellite constellation, the communication frequency of each satellite is allocated and managed to avoid co-channel interference in overlapping coverage areas of the satellite group on the ground at any given time. This controls the beam coverage state of each satellite, reduces mutual interference of signals within the satellite coverage area, and simultaneously analyzes the C / I ratio of the satellite coverage area. By adopting a novel constellation beam management and control method, the problem of mutual interference of signals within the satellite coverage area can be mitigated while ensuring seamless coverage of the ground surface. However, this method still has two shortcomings. First, it does not take into account the real-world scenario of uneven distribution of ground users. In practical applications, the distribution of ground users often varies significantly, such as the difference between urban and remote areas. Second, the traffic demand of users in different regions also varies, and relying solely on factors such as satellite orbital position and communication time to adjust the beam may not achieve optimal resource allocation.

[0004] Chongqing University of Posts and Telecommunications disclosed a resource allocation method for a multi-beam satellite communication system based on the multi-agent A3C algorithm in its patent application, "Resource Allocation Method for Multi-Beam Satellite Communication System Based on Multi-Agent A3C Algorithm" (application date: August 30, 2023, application number: 202311105149.1, publication number: CN116896407 A). The specific steps of this method are as follows: First, establish a low-Earth orbit satellite communication system model and a ground service model; second, model the beam correlation variable scheme and power allocation variable scheme, and determine the satellite load balancing scheme through satellite channel modeling, modeling the system utility function and constraints; third, model the system state, actions, and rewards, construct and train a multi-agent A3C network, and determine the system beam hopping scheme and resource allocation strategy based on the multi-agent A3C algorithm. The drawback of this method is that the objective function only considers user satisfaction, which may lead to insufficient allocation of satellite resources, resulting in low satellite resource utilization. Furthermore, the algorithm training process requires a large amount of historical data and computing resources. The resource allocation strategy training process may be affected by factors such as the dynamic changes of satellite coverage area over time and convergence delay, which makes the training process more complex and the convergence speed slower, affecting the real-time performance and accuracy of the resource allocation strategy. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of the existing technology by proposing a low-Earth orbit constellation beam resource allocation method based on an adaptive evolutionary algorithm. This method aims to solve the problems of low service coverage and satellite resource utilization caused by the non-uniform distribution of ground users and the differentiation of user needs, as well as the lack of real-time performance and accuracy of beam resource allocation due to the dynamic changes of the satellite system and the convergence delay of the algorithm.

[0006] The technical approach of this invention is as follows: Based on the basic parameters of the low-Earth orbit constellation communication link, a satellite communication system model is established. Addressing the issues of insufficient real-time performance and accuracy in beam resource allocation caused by dynamic changes in the satellite system and convergence delays, a user demand prediction neural network is trained using a training set constructed from historical user terminal locations and user service demand information to obtain user demand prediction results. This allows for the early prediction of user demand fluctuations, improving the real-time performance and accuracy of resource allocation. Regarding the issues of low user service coverage and low satellite resource utilization, in the beam domain, the Gale-Shapley algorithm is used to match overlapping satellites with user terminals based on a preference list to obtain a matching list. In the user domain, based on the user demand prediction results and the matching list, combined with the defined system model objective function and resource allocation constraints, an adaptive evolutionary algorithm is used to jointly allocate power and bandwidth to determine the optimal beam resource allocation strategy. This ensures that the satellite communication capacity provided by the satellite beam matches user demand, improving user service coverage and minimizing the waste of satellite resources.

[0007] To achieve the above objectives, the specific implementation steps of the present invention include the following:

[0008] Step 1: Establish a satellite communication system model based on the orbital information of the low-Earth orbit constellation to be allocated beam resources and the basic parameters of the communication link;

[0009] Step 2: Based on historical user terminal locations and user business demand information, construct a training set of sample data, and use the user demand prediction neural network trained on the training set to obtain user demand prediction results.

[0010] Step 3: Obtain the location information of user terminals in the beam overlap area between different satellites, and use the Gale-Shapley algorithm to match the satellites in the overlap area with the user terminals to generate a matching list;

[0011] Step 4: Define the objective function and resource allocation constraints based on the satellite communication system model;

[0012] Step 5: Based on the user demand prediction results and the matching list, determine the optimal beam resource allocation strategy by jointly allocating power and bandwidth through an adaptive evolutionary algorithm.

[0013] Furthermore, the satellite communication system model specifically includes:

[0014] The first step is to establish the gain for each channel: h b,u (t)=a b,c (t)·G b,u (θ t )·G b,u (θ r ), where h b,u (t) represents the channel gain from the b-th beam to the u-th user in the t-th time slot, where b = 1, 2, ..., B, B represents the total number of beams for a single satellite, u = 1, 2, ..., U, U represents the total number of user terminals, t = 1, 2, ..., T, T represents the total number of time slots (50 ms each) for the low-Earth orbit constellation system with resources to be allocated, where the ground area served by the satellite system remains unchanged within each time slot, and a b,c (t) represents the free space loss from the b-th beam to the c-th user in the t-th time slot, G b,u (.), G b,u (.) represent the transmit gain and receive antenna gain between the b-th beam and the u-th user, respectively, and θ t and θ r These represent the angles of the receiving end and the transmitting end, respectively.

[0015] The second step is to establish the signal-to-interference-plus-noise ratio (SIR) for each beam to each user: Among them, SINR b,u i represents the signal-to-interference-plus-noise ratio from the b-th beam to the u-th user. b,u p represents the channel allocation coefficient from the b-th beam to the u-th user. b,u p represents the transmit power allocated to the u-th user by the b-th beam. j,u Let Nj represent the transmit power allocated to the u-th user by the j-th beam, and N0 represent the noise power spectral density. total h represents the total bandwidth of each satellite. j,u This represents the channel gain when the j-th beam reaches the u-th user in the t-th time slot;

[0016] The third step is to establish the satellite communication capacity provided by each satellite to each user: in, B represents the communication capacity provided by the b-th beam of the n-th satellite to the n-th user. n,b,u This represents the bandwidth allocated to the u-th user by the b-th beam of the n-th satellite, where n = 1, 2, ..., N, and N represents the total number of satellites in the low-Earth orbit constellation.

[0017] Furthermore, the training set of the sample data includes the location information of all user terminals in the low-Earth orbit constellation system with resources to be allocated, and the tag data of user service requirements; the user terminal location information is a three-dimensional data matrix, where the rows of the user terminal location information matrix represent the number of time slots before the time slot for predicting user demand, the columns represent the number of each user terminal, and the depth is the three-dimensional position coordinates of the user terminal in the Cartesian coordinate system in each time slot; the tag data of user service requirements is a two-dimensional matrix, where the rows of the matrix represent the number of time slots for predicting user demand in the current low-Earth orbit constellation system with resources to be allocated, the columns represent the number of each user terminal, and each element value in the matrix represents a user service requirement.

[0018] Furthermore, the structure of the user demand prediction neural network is composed of a first fully connected layer, a hidden layer, and a second fully connected layer connected in series. The number of input features of the first fully connected layer is set to 3×V based on the size of the user terminal location information, where V equals the total number of user terminals U. The number of input channels equals the number of time slots for the user demand to be predicted. The number of output neurons is set to 9×V. The hidden layer consists of three identical structures: a first hidden layer, a second hidden layer, and a third hidden layer connected in series. Each hidden layer consists of a first fully connected layer, a first activation function layer, a second fully connected layer, and a second activation function layer connected in series. The output neurons of the first and second fully connected layers of the first hidden layer are... The number of neurons is set to 8×V and 7×V respectively; the number of output neurons in the first and second fully connected layers of the second hidden layer is set to 6×V and 5×V respectively; the number of output neurons in the third hidden layer is set to 4×V, and the number of output neurons in the second fully connected layer is set to 3×V; the number of input channels in the first and second fully connected layers of the first to third hidden layers is set to 1; the first activation function layer of the first to third hidden layers is implemented using the Tanh activation function, and the second activation function is implemented using the sigmoid activation function; based on the output feature tensor of the third hidden layer, the number of input features in the second fully connected layer is set to 3×V, the number of input channels is 1, and the number of output neurons is set to V.

[0019] Furthermore, the training of the user demand prediction neural network refers to inputting the training set into the user demand prediction neural network, using the gradient descent method to iteratively update the network parameters until the network loss function converges, thereby obtaining the trained user demand prediction neural network.

[0020] The network loss function is: Where L(.) represents the loss function, θ represents the network optimization parameters of the user demand prediction neural network at the current iteration, and M represents the number of users waiting for resource allocation at the current iteration. This represents the user demand predicted by the user demand prediction neural network at the current iteration. Indicates and The corresponding actual user needs.

[0021] Furthermore, the steps for matching satellites and user terminals in different satellite beam overlap areas using the Gale-Shapley algorithm are as follows:

[0022] The first step is to establish a user terminal preference list and a satellite preference list for the beam overlap area. The user terminal preference list includes the matching satellite number covering each user within each beam overlap area. The satellite preference list includes the terminal numbers of all users covered by each satellite within each beam overlap area. Each user terminal in the satellite preference list is sorted in ascending order by its number, and the smallest satellite number is stored in the user terminal preference list.

[0023] The second step is to calculate the signal strength of each satellite for each user within each beam overlap region: Among them, ML q,n,u d represents the signal strength of the nth satellite to the uth user within the qth beam overlap region. n,u This represents the distance from the nth satellite to the uth user within the beam overlap area;

[0024] The third step is to compare the nth beam with each beam overlap region, for all satellites covering the u-th user terminal, in order of arrangement. u,i The signal strength of the n'th satellite to the u-th user stored in WL(u) and WL(u) are given by the following: Then use the nth u,i Satellite number n u,i Replace the satellite number stored in the preference list WL(u) with n', delete the u-th user terminal number in the preference list WS(n) of the n-th satellite, and add the deleted u-th user terminal number to the n-th satellite. u,i In the preference list of satellites, if Then no operation is performed, and the final satellite channel gain for the u-th user terminal stored in WL(u) is added to the matching list;

[0025] The fourth step involves iterating through each user terminal in each overlapping area and performing the operation in step three. This yields a matching list for each overlapping area, consisting of a two-dimensional matrix. The rows of this matching list represent the satellite number of each beam overlap area, the columns represent the number of each user terminal in the beam overlap area, and each element value in the matrix represents the matching channel gain between the satellite and the user terminal.

[0026] Furthermore, the objective function and resource allocation constraints are defined based on the satellite communication system model:

[0027] The objective function is: Wherein, USC represents the function that minimizes the maximum unmet service coverage of user demand in the low-Earth orbit constellation beams, that is, minimizing the maximum unmet service coverage of user demand. and The gap This represents the satellite communication capacity provided by the b-th beam of the n-th satellite to the u-th user. This represents the communication capacity required by the u-th user;

[0028] The resource allocation constraints include: channel allocation state constraints, bandwidth constraints, angle constraints from the user to the beam center, maximum allocation power constraints, and beam service user number constraints. The channel allocation state constraint stipulates that the channel allocation state for each user of each beam can only be allocated or not allocated. The bandwidth constraint stipulates that the total transmission bandwidth allocated to users within the coverage area of ​​each beam does not exceed the beam's maximum transmission bandwidth. The angle constraint stipulates that the angle from each user to the beam center within the coverage area of ​​each beam is less than 3dB of the beam's angle. The maximum allocation power constraint stipulates that the total transmission power allocated to all users within the coverage area of ​​each beam does not exceed the beam's maximum available transmission power. The beam service user number constraint stipulates that the number of users served by each beam does not exceed the maximum allowed number of users served by the beam.

[0029] Furthermore, the step of jointly allocating power and bandwidth using an adaptive evolutionary algorithm is as follows:

[0030] The first step is to randomly generate a set of resource pre-allocation strategy matrices R = {r1, r2, ..., r...} based on resource allocation constraints. Z} as the parent generation, where Z represents the total number of randomly generated resource pre-allocation strategy matrices;

[0031] The second step is to establish the fitness function of the genetic algorithm as: f(r z )=-USC(r z ), where f(r) z () represents the fitness value of the z-th resource pre-allocation strategy matrix predicted based on the trained neural network. This represents the z-th resource pre-allocation strategy matrix. and Let z1 and z2 represent the power allocation matrix and bandwidth allocation matrix, respectively. The power allocation matrix and bandwidth allocation matrix contain the power and bandwidth allocated to each user within the coverage area of ​​each satellite beam in the z-th resource pre-allocation strategy. The superscript T indicates the transpose operation.

[0032] The third step is to establish the crossover probability function and the mutation probability function as follows:

[0033]

[0034] Among them, P c p m These represent the crossover probability function and the mutation probability function, respectively. These represent the maximum and minimum crossover probabilities, respectively. Let f represent the maximum and minimum values ​​of the mutation probability, exp represent the exponential operation with base e, c0 represent the linear parameter with a value of 9.903438, and f max This represents the value with the highest fitness among the current parent generations. f represents the larger fitness value among the two resource allocation strategy vectors involved in the crossover. m f represents the fitness value of the current resource allocation strategy vector. avg f represents the average fitness value of all resource allocation strategy vectors in the parent generation. z This represents the fitness value of the current z-th resource allocation strategy vector;

[0035] The fourth step involves selecting the parent generation based on the fitness value using the selection probability, calculating the crossover probability function, and performing pairwise crossovers on the resource allocation strategy vectors of the selected parent generation to obtain the crossover parent generation. Then, the mutation probability function is used to calculate the mutation probability, and the resource allocation strategy vectors of the crossover parent generation are mutated to obtain the final generation of resource pre-allocation strategy matrix R' as the new parent generation. Based on the user demand predicted by the user demand prediction neural network, the fitness of each resource allocation strategy vector in R' is calculated.

[0036] Fifth, repeat step four until the fitness function of the adaptive evolutionary algorithm converges, thus obtaining the optimal resource allocation strategy.

[0037] The advantages of this invention compared to the prior art are as follows:

[0038] First, this invention trains a user demand prediction neural network using a training set constructed from historical user terminal locations and user service demand information to obtain user demand prediction results and predict changes in user demand in a timely manner. In the beam domain, the Gale-Shapley algorithm is used to match overlapping satellites and user terminals based on a preference list to obtain a matching list. In the user domain, based on the user demand prediction results and the matching list, combined with the defined objective function and resource allocation constraints, an adaptive evolutionary algorithm is used to jointly allocate power and bandwidth to determine the optimal beam resource allocation strategy. The proposed objective function comprehensively considers the satellite communication capacity provided by the satellite beam, the matching degree of user demand, and the utilization rate of satellite resources. It overcomes the problems of low user service coverage, low satellite resource utilization, and insufficient real-time performance and accuracy of beam resource allocation caused by the differentiated distribution of user demand, dynamic changes in the satellite system, and algorithm convergence delay in the prior art. It ensures user service coverage and resource utilization while improving the real-time performance and accuracy of allocation. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of a satellite communication system according to an embodiment of the present invention;

[0040] Figure 2 This is a flowchart of the beam resource allocation method based on the adaptive evolutionary algorithm of the present invention;

[0041] Figure 3 The figure shows the simulation results of this invention. Detailed Implementation

[0042] The present invention will now be further described with reference to the accompanying drawings.

[0043] Reference Figure 1 The satellite communication system of the present invention will be further described below.

[0044] The low-Earth orbit satellite communication system used in this example consists of 720 low-Earth orbit multi-beam satellites with an orbital altitude of 1200km. The satellites use the Ka band (17.7~21.2GHz), and each satellite's onboard multi-beam transmitter transmits a total of 16 beams. The total bandwidth of the system is 40MHz, and the maximum gain of the satellite's transmit antenna is 58.5dBi.

[0045] Reference Figure 2 The specific steps of the beam resource allocation method of the present invention will be further described below.

[0046] Step 1: Establish a satellite communication system model based on the orbital information of the low-Earth orbit constellation to be allocated beam resources and the basic parameters of the communication link.

[0047] In an embodiment of the present invention, a satellite communication scenario is built using Python based on the parameters of the low-Earth orbit constellation. After generating a beam, the orbital information and basic parameters of the communication link of the low-Earth orbit constellation are obtained.

[0048] The process of establishing the satellite communication system model is as follows:

[0049] The first step is to establish the gain for each channel as follows:

[0050] h b,u (t)=a b,c (t)·G b,u (θ t )·G b,u (θ r )

[0051] Among them, h b,u (t) represents the channel gain from the b-th beam to the u-th user in the t-th time slot, where b = 1, 2, ..., 16, u = 1, 2, ..., U, t = 1, 2, ..., T, and T represents the total number of time slots (50 ms each) that divide the total time of the low-Earth orbit constellation system to be allocated resources. The ground area served by the satellite system remains unchanged within each time slot. b,c (t) represents the free space loss from the b-th beam to the c-th user in the t-th time slot, G b,u (.), G b,u (.) represent the transmit gain and receive antenna gain between the b-th beam and the u-th user, respectively, and θ t and θ r These represent the angles of the receiving end and the transmitting end, respectively.

[0052] The second step is to establish the signal-to-interference-plus-noise ratio (SIR) for each beam to each user:

[0053]

[0054] Among them, SINR b,u i represents the signal-to-interference-plus-noise ratio from the b-th beam to the u-th user. b,u p represents the channel allocation coefficient from the b-th beam to the u-th user. b,u p represents the transmit power allocated to the u-th user by the b-th beam. j,u Let Nj represent the transmit power allocated to the u-th user by the j-th beam, and N0 represent the noise power spectral density. total h represents the total bandwidth of each satellite. j,u This represents the channel gain when the j-th beam reaches the u-th user in the t-th time slot.

[0055] The third step is to establish the satellite communication capacity provided by each satellite to each user:

[0056]

[0057] in, B represents the communication capacity provided by the b-th beam of the n-th satellite to the n-th user. n,b,u This represents the bandwidth allocated to the u-th user by the b-th beam of the n-th satellite, where n = 1, 2, ..., N, and N represents the total number of satellites in the low-Earth orbit constellation.

[0058] Step 2: Based on historical user terminal locations and user business demand information, construct a training set of sample data, and use the user demand prediction neural network trained on the training set to obtain user demand prediction results.

[0059] The steps of the user demand prediction neural network in predicting user demand are as follows:

[0060] Step 1: Obtain user service requirements within the beam coverage area based on the satellite communication system scenario. In this embodiment of the invention, the user service requirements are the product of the current user's area traffic concurrency rate and the population density within the user's area. The traffic concurrency rate is referenced from the 2021 traffic concurrency rate dataset, and the population density is mainly referenced from the Grid Population (GPW) v4 dataset. By gridding, each 1°×1° area is regarded as a cell, and the population data is regarded as the population density of that cell.

[0061] Step 2 involves data processing, specifically processing the user terminal locations and user service requirements information from the previous five time slots to obtain a training set.

[0062] Step 3: Input the training set into the user demand prediction neural network, and use the gradient descent method to iteratively update the network parameters until the network loss function converges, thus obtaining the trained user demand prediction neural network. The user demand prediction neural network is then used to obtain the predicted user demand.

[0063] Step 3: Obtain the location information of user terminals in the beam overlap area between different satellites, and use the Gale-Shapley algorithm to match the satellites in the overlap area with the user terminals to generate a matching list.

[0064] The steps for matching overlapping area satellites and user terminals based on the Gale-Shapley algorithm are as follows:

[0065] Step 1: Establish user terminal preference list and satellite preference list for beam overlap area respectively.

[0066] Step 2: Calculate the signal strength of each satellite to each user within each beam overlap area.

[0067] Step 3: Within each beam overlap area, compare the signal strength of each satellite to the user in the order of satellite arrangement to complete the user and satellite matching list.

[0068] Step 4: Define the objective function and resource allocation constraints based on the satellite communication system model.

[0069] The objective function in the embodiments of the present invention is:

[0070]

[0071] Wherein, USC represents the function that minimizes the maximum unmet service coverage of user demand in the low-Earth orbit constellation beams, that is, minimizing the maximum unmet service coverage of user demand. and The gap This represents the satellite communication capacity provided by the b-th beam of the n-th satellite to the u-th user. This represents the communication capacity required by the u-th user.

[0072] In the embodiments of the present invention, the resource allocation constraints include: channel allocation state constraints, bandwidth constraints, angle constraints from the user to the beam center, maximum allocation power constraints, and beam service user number constraints.

[0073] The channel allocation state constraint states that for each beam, the channel allocation state for each user can only be either allocated or not allocated.

[0074] The bandwidth constraint states that the total transmission bandwidth allocated to users within the coverage area of ​​each beam shall not exceed the maximum transmission bandwidth of the beam.

[0075] The angle constraint from the user to the beam center is that the angle from each user to the beam center within the coverage area of ​​each beam is less than 3dB of the beam.

[0076] The maximum allocated power constraint is that the total transmission power allocated to all users within the coverage area of ​​each beam does not exceed the maximum available transmission power of the beam.

[0077] The constraint on the number of users served by each beam is that the number of users served by each beam shall not exceed the maximum number of users allowed to serve by the beam.

[0078] Step 5: Based on the user demand prediction results and the matching list, determine the optimal beam resource allocation strategy by jointly allocating power and bandwidth through an adaptive evolutionary algorithm.

[0079] The steps of the adaptive evolutionary algorithm in the embodiments of the present invention for jointly allocating power and bandwidth are as follows:

[0080] Step 1: Based on resource allocation constraints, randomly generate a set of resource pre-allocation strategy matrices R = {r1, r2, ..., r...} Z} as the parent, where Z represents the total number of randomly generated resource pre-allocation strategy matrices.

[0081] Step 2, establish the fitness function for the adaptive evolutionary algorithm as follows:

[0082] f(r z )=-USC(r z )

[0083] Where, f(r) z () represents the fitness value of the z-th resource pre-allocation strategy matrix predicted based on the trained neural network. This represents the z-th resource pre-allocation strategy matrix. and Let z1 and z2 represent the power allocation matrix and bandwidth allocation matrix, respectively. The power allocation matrix and bandwidth allocation matrix contain the power and bandwidth allocated to each user within the coverage area of ​​each satellite beam in the z-th resource pre-allocation strategy. The superscript T indicates the transpose operation.

[0084] Step 3, establish the crossover probability function and the mutation probability function as follows:

[0085]

[0086] Among them, P c p m These represent the crossover probability function and the mutation probability function, respectively. These represent the maximum and minimum crossover probabilities, respectively. Let f represent the maximum and minimum values ​​of the mutation probability, exp represent the exponential operation with base e, c0 represent the linear parameter with a value of 9.903438, and f max This represents the value with the highest fitness among the current parent generations. f represents the larger fitness value among the two resource allocation strategy vectors involved in the crossover. m f represents the fitness value of the current resource allocation strategy vector. avg f represents the average fitness value of all resource allocation strategy vectors in the parent generation. z This represents the fitness value of the current z-th resource allocation strategy vector.

[0087] Step 4: Select the parent generation based on the fitness value using the selection probability to obtain the selected parent generation. Calculate the crossover probability function and perform pairwise crossovers on the resource allocation strategy vectors of the selected parent generation to obtain the crossover parent generation. Then, calculate the mutation probability function and perform mutation operations on the resource allocation strategy vectors of the crossover parent generation to obtain the final generation of resource pre-allocation strategy matrix R' as the new parent generation. Calculate the fitness of each resource allocation strategy vector in R' based on the user demand predicted by the user demand prediction neural network.

[0088] Step 6: Repeat step 4 of this step until the fitness function of the adaptive evolutionary algorithm converges, thus obtaining the optimal resource allocation strategy.

[0089] The effects of the present invention will be further described below with reference to simulation experiments.

[0090] 1. Simulation experimental conditions:

[0091] The software platform for the simulation experiment of this invention is: Windows 11 operating system and Matlab 2018b.

[0092] In the simulation experiment scenario of this invention, the low-Earth orbit constellation is the OneWeb constellation, with an orbital altitude of 1200km and 16 beams per satellite. The user demand traffic is mainly referenced from the Grid Population (GPW) v4 dataset and the 2021 business concurrency rate dataset.

[0093] 2. Simulation content and result analysis.

[0094] This invention employs simulation experiments to optimize three different algorithms, obtaining the standard deviation of user satisfaction and beam resource utilization for each algorithm under varying total service demand. Then, using Matlab 2018b, line graphs of the standard deviation of user satisfaction for the three beam resource allocation algorithms under different total service demand are plotted, as shown below. Figure 3 As shown in (a), and the beam resource utilization histograms of the three beam resource allocation algorithms under different total service demand conditions, as shown in (a), Figure 3 As shown in (b).

[0095] The three beam resource allocation algorithms refer to:

[0096] The first method is based on a randomly assigned low-Earth orbit constellation beam resource allocation algorithm. It uses the user demand prediction neural network proposed in step 2 of the present invention to obtain the user demand prediction results. For satellites and user terminals in different satellite beam overlap areas, the matching is performed using the Gale-Shapley algorithm proposed in step 3. For each user terminal, beam resources are randomly allocated using a random function.

[0097] The second method is a low-Earth orbit constellation beam resource allocation algorithm based on a genetic algorithm. This method uses the user demand prediction neural network trained in step 2 of the present invention to obtain user demand prediction results. For satellites and user terminals in different satellite beam overlap areas, the Gale-Shapley algorithm proposed in step 3 is used for matching. A genetic algorithm is used to simulate the natural selection process, randomly generating multiple beam resource allocation strategies as a beam resource allocation strategy set. The fitness of each strategy is evaluated, and strategies with higher fitness are selected for crossover and mutation to generate a new generation of strategy sets. This evaluation and selection process is repeated until the convergence condition is met to determine the optimal beam resource allocation strategy.

[0098] The third method is the low-orbit constellation beam resource allocation algorithm based on adaptive evolutionary algorithm proposed in this invention. It uses the user demand prediction neural network training proposed in step 2 of the embodiment of this invention to obtain the user demand prediction result. It uses the Gale-Shapley algorithm proposed in step 3 to match satellites and user terminals in different satellite beam overlap areas. Then, it uses the adaptive evolutionary algorithm proposed in step 5 of this invention to jointly allocate power and bandwidth to determine the optimal beam resource allocation strategy.

[0099] The following is combined Figure 3 The simulation effects of the present invention will be further described.

[0100] Figure 3 In (a), the horizontal axis represents the total user data traffic demand in Gbps, and the vertical axis represents the standard deviation of user satisfaction. Figure 3 (b) The horizontal axis represents the total user traffic demand in Gbps, and the vertical axis represents the satellite resource utilization rate. The light gray rectangles correspond to the histogram of the low-Earth orbit constellation beam resource allocation algorithm based on random allocation, the dark gray rectangles correspond to the histogram of the low-Earth orbit constellation beam resource allocation algorithm based on genetic algorithm, and the black rectangles correspond to the histogram of the low-Earth orbit constellation beam resource allocation algorithm based on adaptive evolution algorithm proposed in this invention.

[0101] Depend on Figure 3As can be seen, when the total user traffic demand is below 800Gbps, the three algorithms are essentially the same due to abundant resources, and the standard deviation of user satisfaction is not significantly different across the three algorithms. However, when the total user traffic demand exceeds 800Gbps, the satellite beam system resources become less abundant, and the difference in user demand increases. The low-Earth orbit constellation beam resource allocation algorithm based on the adaptive evolutionary algorithm achieves significantly higher user service coverage and better resource utilization than the other two. When the total user traffic demand reaches 1200Gbps, the method of this invention improves the resource utilization rate to 88.78%. When the total user traffic demand exceeds 1320Gbps, the total user traffic demand begins to exceed the total satellite beam system resources, and the improvement effect on resource utilization begins to weaken. Figure 3 The results shown demonstrate that the method proposed in this invention maximizes user service coverage while also improving satellite resource utilization.

[0102] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0103] The contents not described in detail in this invention are common knowledge to those skilled in the art.

Claims

1. A method for allocating low-Earth orbit constellation beam resources based on an adaptive evolutionary algorithm, characterized in that, An objective function model for low-Earth orbit satellite beam resource allocation is constructed. An adaptive adjustment algorithm is used to predict user demand and channel conditions, and then allocate beam bandwidth and power. The steps of this allocation method are as follows: Step 1: Establish a satellite communication system model based on the orbital information of the low-Earth orbit constellation to be allocated beam resources and the basic parameters of the communication link; Step 2: Based on historical user terminal locations and user business demand information, construct a training set of sample data, and use the user demand prediction neural network trained on the training set to obtain user demand prediction results. Step 3: Obtain the location information of user terminals in the beam overlap area between different satellites, and use the Gale-Shapley algorithm to match the satellites in the overlap area with the user terminals to generate a matching list; Step 4: Define the objective function and resource allocation constraints based on the satellite communication system model: The objective function is: Wherein, USC represents the function that minimizes the maximum unmet service coverage of user demand in the low-Earth orbit constellation beams, that is, minimizing the maximum unmet service coverage of user demand. and The gap This represents the satellite communication capacity provided by the b-th beam of the n-th satellite to the u-th user. This represents the communication capacity required by the u-th user; The resource allocation constraints include: channel allocation state constraints, bandwidth constraints, angle constraints from the user to the beam center, maximum allocation power constraints, and beam service user number constraints. The channel allocation state constraint states that for each beam, the channel allocation state for each user can only be either allocated or not allocated. The bandwidth constraint states that the total transmission bandwidth allocated to users within the coverage area of ​​each beam shall not exceed the maximum transmission bandwidth of the beam. The angle constraint from the user to the beam center is that the angle from each user to the beam center within the coverage area of ​​each beam is less than 3dB of the beam. The maximum allocated power constraint is that the total transmission power allocated to all users within the coverage area of ​​each beam does not exceed the maximum available transmission power of the beam. The constraint on the number of users served by each beam is that the number of users served by each beam shall not exceed the maximum number of users allowed to serve by the beam. Step 5: Based on the user demand prediction results and the matching list, determine the optimal beam resource allocation strategy by jointly allocating power and bandwidth using an adaptive evolutionary algorithm; the steps are as follows: The first step is to randomly generate a set of resource pre-allocation strategy matrices R = {r1, r2, ..., r...} based on resource allocation constraints. Z } as the parent generation, where Z represents the total number of randomly generated resource pre-allocation strategy matrices; The second step is to establish the fitness function for the adaptive evolutionary algorithm as follows: f(r z )=-USC(r z ) Where, f(r) z () represents the fitness value of the z-th resource pre-allocation strategy matrix predicted based on the trained neural network. This represents the z-th resource pre-allocation strategy matrix. and Let z1 and z2 represent the power allocation matrix and bandwidth allocation matrix, respectively. The power allocation matrix and bandwidth allocation matrix contain the power and bandwidth allocated to each user within the coverage area of ​​each satellite beam in the z-th resource pre-allocation strategy. The superscript T indicates the transpose operation. The third step is to establish the crossover probability function and the mutation probability function as follows: Among them, P c p m These represent the crossover probability function and the mutation probability function, respectively. These represent the maximum and minimum crossover probabilities, respectively. Let f represent the maximum and minimum values ​​of the mutation probability, exp represent the exponential operation with the natural constant e as the base, c0 represent the linear parameter with a value of 9.903438, and f max This represents the value with the highest fitness among the current parent generations. f represents the larger fitness value among the two resource allocation strategy vectors involved in the crossover. m f represents the fitness value of the current resource allocation strategy vector. avg f represents the average fitness value of all resource allocation strategy vectors in the parent generation. z This represents the fitness value of the current z-th resource allocation strategy vector; The fourth step involves selecting the parent generation based on the fitness value using the selection probability. Then, the crossover probability function is used to calculate the crossover probability. The resource allocation strategy vectors of the selected parent generation are crossed pairwise to obtain the crossover parent generation. Next, the mutation probability function is used to calculate the mutation probability. The resource allocation strategy vectors of the crossover parent generation are mutated to obtain the final new generation resource pre-allocation strategy matrix R′ as the new parent generation. The fitness of each resource allocation strategy vector in R′ is calculated based on the user demand predicted by the user demand prediction neural network. Fifth, repeat step four until the fitness function of the adaptive evolutionary algorithm converges, thus obtaining the optimal resource allocation strategy.

2. The method for allocating low-Earth orbit constellation beam resources based on adaptive evolutionary algorithm according to claim 1, characterized in that, The satellite communication system model described in step 1 specifically includes: The first step is to establish the gain for each channel as follows: h b,u (t)=a b,c (t)·G b,u (i t )·G b,u (i r ) Among them, h b,u (t) represents the channel gain from the b-th beam to the u-th user in the t-th time slot, where b = 1, 2, ..., B, B represents the total number of beams for a single satellite, u = 1, 2, ..., U, U represents the total number of user terminals, t = 1, 2, ..., T, T represents the total number of time slots (50 ms each) for the low-Earth orbit constellation system with resources to be allocated, where the ground area served by the satellite system remains unchanged within each time slot, and a b,c (t) represents the free space loss from the b-th beam to the c-th user in the t-th time slot, G b,u (.), G b,u (.) represent the transmit gain and receive antenna gain between the b-th beam and the u-th user, respectively, and θ t and θ r These represent the angles of the receiving end and the transmitting end, respectively. The second step is to establish the signal-to-interference-plus-noise ratio (SIR) for each beam to each user: Among them, SINR b,u i represents the signal-to-interference-plus-noise ratio from the b-th beam to the u-th user. b,u p represents the channel allocation coefficient from the b-th beam to the u-th user. b,u p represents the transmit power allocated to the u-th user by the b-th beam. j,u Let Nj represent the transmit power allocated to the u-th user by the j-th beam, and N0 represent the noise power spectral density. total h represents the total bandwidth of each satellite. j,u This represents the channel gain when the j-th beam reaches the u-th user in the t-th time slot; The third step is to establish the satellite communication capacity provided by each satellite to each user: in, B represents the communication capacity provided by the b-th beam of the n-th satellite to the n-th user. n,b,u This represents the bandwidth allocated to the u-th user by the b-th beam of the n-th satellite, where n = 1, 2, ..., N, and N represents the total number of satellites in the low-Earth orbit constellation.

3. The method for allocating low-Earth orbit constellation beam resources based on adaptive evolutionary algorithm according to claim 1, characterized in that, The training set of sample data in step 2 includes the historical terminal locations, user service demand information, and tag data of user service demands for all users in the low-Earth orbit constellation system with resources to be allocated. The user terminal location information is a three-dimensional data matrix, where the rows of the matrix represent the number of time slots before the time slot for the user demand to be predicted, the columns represent the number of each user terminal, and the depth is the three-dimensional position coordinates of the user terminal in the Cartesian coordinate system within each time slot. The tag data of user service demands is a two-dimensional matrix, where the rows of the matrix represent the number of time slots for the user demand to be predicted in the current low-Earth orbit constellation system with resources to be allocated, the columns represent the number of each user terminal, and each element value in the matrix represents the user service demand.

4. The method for low-Earth orbit constellation beam resource allocation based on adaptive evolutionary algorithm according to claim 3, characterized in that, The user demand prediction neural network described in step 3 consists of a first fully connected layer, a hidden layer, and a second fully connected layer connected in series. The number of input features in the first fully connected layer is set to 3×V based on the size of the user terminal location information, where V equals the total number of user terminals U. The number of input channels equals the number of time slots for the user demand to be predicted. The number of output neurons is set to 9×V. The hidden layer consists of three identical structures: a first hidden layer, a second hidden layer, and a third hidden layer connected in series. Each hidden layer consists of a first fully connected layer, a first activation function layer, a second fully connected layer, and a second activation function layer connected in series. The output neurons of the first and second fully connected layers of the first hidden layer are... The number of neurons in the first and second fully connected layers of the second hidden layer is set to 8×V and 7×V respectively; the number of output neurons in the first and second fully connected layers of the second hidden layer is set to 6×V and 5×V respectively; the number of output neurons in the third hidden layer is set to 4×V, and the number of output neurons in the second fully connected layer is set to 3×V; the number of input channels in the first and second fully connected layers of the first to third hidden layers is set to 1; the first activation function layer of the first to third hidden layers is implemented using the Tanh activation function, and the second activation function is implemented using the Sigmoid activation function; based on the output feature tensor of the third hidden layer, the number of input features in the second fully connected layer is set to 3×V, the number of input channels is 1, and the number of output neurons is set to V.

5. The method for allocating low-Earth orbit constellation beam resources based on an adaptive evolutionary algorithm according to claim 4, characterized in that, Step 3, which involves training the user demand prediction neural network, means inputting the training set into the user demand prediction neural network, using gradient descent to iteratively update the network parameters until the network loss function converges, thus obtaining the trained user demand prediction neural network.

6. The method for allocating low-Earth orbit constellation beam resources based on adaptive evolutionary algorithm according to claim 5, characterized in that, The network loss function is as follows: Where L(.) represents the loss function, θ represents the network optimization parameters of the user demand prediction neural network at the current iteration, and M represents the number of users waiting for resource allocation at the current iteration. This represents the user demand predicted by the user demand prediction neural network at the current iteration. Indicates and The corresponding actual user needs.

7. The method for allocating low-Earth orbit constellation beam resources based on an adaptive evolutionary algorithm according to claim 2, characterized in that, The steps in step 3, which involve using the Gale-Shapley algorithm to match satellites with user terminals in different satellite beam overlap areas, are as follows: The first step is to establish a user terminal preference list and a satellite preference list for the beam overlap area, respectively; the user terminal preference list includes the matching satellite number covering each user within each beam overlap area; The satellite preference list includes the terminal numbers of all users covered by each satellite within each beam overlap area; Sort each user terminal in the satellite preference list in ascending order by number, and store the smallest satellite number in the user terminal preference list; The second step is to calculate the signal strength of each satellite for each user within each beam overlap region: Among them, ML q,n,u d represents the signal strength of the nth satellite to the uth user within the qth beam overlap region. n,u This represents the distance from the nth satellite to the uth user within the beam overlap area; The third step is to compare the nth beam with each beam overlap region, for all satellites covering the u-th user terminal, in order of arrangement. u,i The signal strength of the nth satellite to the uth user stored in WL(u) and WL(u) are given by the following: Then use the nth u,i Satellite number n u,i Replace the satellite number stored in the preference list WL(u) with n′, delete the u-th user terminal number in the preference list WS(n) of the n-th satellite, and add the deleted u-th user terminal number to the n-th satellite. u,i In the preference list of satellites, if Then no operation is performed, and the channel gain of the satellite for the u-th user terminal that is finally stored in WL(u) is added to the matching list; The fourth step involves iterating through each user terminal in each overlapping area and performing the operation in step three. This yields a matching list for each overlapping area, consisting of a two-dimensional matrix. The rows of this matching list represent the satellite number of each beam overlap area, the columns represent the number of each user terminal in the beam overlap area, and each element value in the matrix represents the matching channel gain between the satellite and the user terminal.

Images