A method and system for scheduling a satellite-ground laser link

Abstract

The application relates to the technical field of satellite-ground laser communication network resource scheduling, and provides a satellite-ground laser link scheduling method and system.The method comprises the following steps: generating a possible available random scene of each candidate link according to link availability prediction data; the available random scene comprises an availability realization value and a scene occurrence probability; calculating a weighted expected data volume of a corresponding scene based on the availability realization value and the scene occurrence probability of the random scene; taking the sum of the weighted expected data volumes of the scenes in a scheduling period as a target, taking the number of satellite-ground laser links established by a satellite, the number of satellite-ground laser links accessed by a ground station and whether the satellite and the ground station establish a satellite-ground laser link as constraint terms to construct a stochastic programming model; solving the stochastic programming model to obtain an optimal scheduling result and scheduling the satellite-ground laser link. The scheduling scheme can be optimized, the scheduling scheme can be improved in adaptability to a random environment, and system throughput loss can be reduced.

Description

Technical Field

[0001] This application relates to the technical field of resource scheduling in satellite-to-ground laser communication networks, and in particular to a satellite-to-ground laser link scheduling method and system. Background Technology

[0002] The rapid development of satellite remote sensing, space science exploration, satellite internet, and space-ground integrated networks has led to a continuous increase in the amount of data acquired and carried by satellites in orbit, putting traditional space-ground communication methods under increasing pressure for high-speed data backhaul. Space-ground laser communication, relying on its high data transmission rate, abundant spectrum resources, and narrow transmission beam, is suitable for connecting multiple satellites, multiple ground stations, and their laser communication terminals to form a high-speed space-ground data transmission network.

[0003] Once the ground station layout is predetermined, satellites cannot simultaneously establish laser links with any ground station at any given time. Instead, selection must be made based on link conditions and terminal resources within different scheduling time slots. Specifically, the actual link establishment combination to be executed must be determined from candidate satellite-to-ground laser links, taking into account factors such as the visibility duration between the satellite and the ground station, the link availability probability, the link transmission rate, and the concurrent link establishment capabilities of the laser communication terminals at both the satellite and ground station ends. Different link establishment combinations correspond to different data backhaul capabilities, thus affecting the total amount of data transmission the system can complete within the scheduling cycle.

[0004] Existing deterministic link scheduling methods typically assume that link availability in each future time slot can be accurately predicted, and use the predicted values ​​as fixed parameters input into the scheduling model. However, link availability is affected by cloud formation, movement, and dissipation, resulting in discrepancies between predicted values ​​and actual values ​​during the scheduling execution phase. If link establishment relationships are determined solely based on a single predicted value, a limited number of laser communication terminals may be assigned to candidate links with low actual availability, leading to a lack of adaptability of the scheduling scheme to the environment and causing a loss in system throughput. Summary of the Invention

[0005] Based on this, it is necessary to provide a satellite-to-ground laser link scheduling method and system to address the above-mentioned technical problems. This method integrates the link availability, effective visibility duration, and transmission rate of each candidate link in multiple scenarios into a scenario-weighted expected data volume. This can directly quantify the expected contribution of different candidate links to the system throughput, and use finite scenarios and their occurrence probabilities to describe the uncertainty of link availability. The expected system throughput after scenario probability weighting is used as the optimization objective to optimize the scheduling scheme, improve the adaptability of the scheduling scheme to the lack of random environment, and reduce the loss of system throughput.

[0006] Firstly, this application provides a satellite-to-ground laser link scheduling method, including: Obtain link availability prediction data for each candidate link, and generate possible random availability scenarios for each candidate link based on the link availability prediction data; the available random scenarios include availability realization values ​​and scenario occurrence probabilities; The weighted expected data volume of the corresponding scenario is calculated based on the availability realization value and the probability of scenario occurrence of random scenarios. With the goal of maximizing the sum of weighted expected data of scenarios within the scheduling cycle, a stochastic programming model is constructed with constraints including the number of satellite-to-ground laser links established by the satellite, the number of satellite-to-ground laser links accessed by the ground station, and whether satellite and ground station establish satellite-to-ground laser links. The stochastic programming model is solved to obtain the optimal scheduling result, and the satellite-to-ground laser link is scheduled based on the optimal scheduling result.

[0007] In one embodiment, the weighted expected data volume of the corresponding scenario is calculated based on the availability realization value of the random scenario and the probability of scenario occurrence, including: Get the Within the first time slot satellite and the first Effective visibility duration between ground stations and the corresponding transmission rate of the satellite-to-ground laser link ; Based on valid visibility duration and transmission rate Calculate the first Within the first time slot satellite and the first Data volume between ground stations ; Availability implementation value, scenario occurrence probability, and data volume based on random scenarios. Calculate the first Within the first time slot satellite and the first Weighted expected data volume between ground stations .

[0008] In one embodiment, the objective function of the stochastic programming model is:

[0009] The objective function satisfies the following constraints:

[0010]

[0011]

[0012] in, The set of scheduling slots within a scheduling period. ; For satellite collection, ; Gathering at the ground station, ; For the first Within the first time slot satellite and the first The weighted expected data volume for scenarios between ground stations; For the constructed binary link scheduling variable, when the first Within the first time slot satellite and the first When establishing a satellite-to-ground laser link between two ground stations, ,otherwise, ; The constraint on the number of satellite-to-ground laser links established for a satellite means that the number of satellite-to-ground laser links established simultaneously by each satellite in any time slot shall not exceed [a certain limit]. ; The constraint on the number of satellite-to-ground laser links accessed by a ground station means that the number of satellite-to-ground laser links simultaneously accessed by each ground station in any time slot shall not exceed [a certain limit]. ; Constraints are set for whether to establish a satellite-to-ground laser link between the satellite and the ground station, which are used to limit the link scheduling variable to a binary variable.

[0013] In one embodiment, solving a stochastic programming model includes: Constructing a scenario-based weighted expected data volume Weight matrix; the weight matrix in which the first... The element represents the first element. satellite and the first The weighted expected data volume for scenarios between ground stations; Constraints based on a stochastic programming model are used to select elements in the weight matrix to simulate the construction of a satellite-to-ground laser link; where the link scheduling variable corresponding to the selected element is set to... The number of elements selected in each row does not exceed Each satellite can establish a maximum of One laser communication link; the number of selected elements in each column does not exceed [number missing]. The maximum number of stations that can be built per ground station One laser communication link; The objective is to maximize the sum of the weighted expected data volume of the simulated satellite-to-ground laser link within the scheduling period.

[0014] In one embodiment, the solution is performed with the objective of maximizing the sum of expected data volumes of the simulated satellite-to-ground laser link scenario within the scheduling period, including: Each satellite will be established at most Each laser communication link and each ground station can establish a maximum of [number] links. The multi-terminal concurrent link establishment constraint represented by a laser communication link is transformed into a one-to-one matching constraint, which will be used to connect a laser communication link equipped with a laser communication link. The satellite equivalent of a laser communication terminal is Each satellite replica is equipped with a laser communication terminal; a satellite equipped with... The ground station of a laser communication terminal is equivalent to Each ground station replica is equipped with a laser communication terminal. Based on one-to-one matching constraints, the dimensions are... The weight matrix is ​​expanded to dimension 1. The target matrix; The target matrix is ​​solved using the Munkres matching algorithm.

[0015] In one embodiment, the dimension is determined based on a one-to-one matching constraint. The weight matrix is ​​expanded to dimension 1. The target matrix includes: Establish dimensions as matrix ,for , and ,Will The weight matrix of the first Copy each element to The There are _ ... Rearranged according to ground station terminal copy number The columns form a matrix. Column expansion results ; matrix Copy line by line Next, the dimension is formed as Target matrix .

[0016] In one embodiment, the target matrix is ​​solved using the Munkres matching algorithm, including: Construct an extended square matrix based on the target matrix, and then... Placed in front of the expanded array row, before Columns are arranged in order, and the remaining positions are filled with zero elements to obtain a square matrix. Among them, the order of the expanded square matrix ; Opponent Invert all elements of the matrix to construct the cost matrix; The Munkres matching algorithm is used to solve the problem of selecting elements from the cost matrix that minimizes the total cost, thus obtaining the optimal scheduling result.

[0017] Secondly, this application also provides a satellite-to-ground laser link scheduling system, comprising: The random scenario generation module is used to obtain the link availability prediction data of each candidate link and generate possible available random scenarios for each candidate link based on the link availability prediction data; the available random scenarios include availability realization values ​​and scenario occurrence probabilities. The weighted expected data volume calculation module is used to calculate the weighted expected data volume of the corresponding scenario based on the availability realization value and the probability of scenario occurrence of the random scenario. The model building module is used to construct a stochastic programming model with the goal of maximizing the sum of weighted expected data of the scenario within the scheduling period, and with constraints such as the number of satellite-to-ground laser links established by the satellite, the number of satellite-to-ground laser links accessed by the ground station, and whether the satellite and the ground station have established satellite-to-ground laser links. The result generation module is used to solve the stochastic programming model, obtain the optimal scheduling result, and schedule the satellite-to-ground laser link based on the optimal scheduling result.

[0018] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the following steps: Obtain link availability prediction data for each candidate link, and generate possible random availability scenarios for each candidate link based on the link availability prediction data; the available random scenarios include availability realization values ​​and scenario occurrence probabilities; The weighted expected data volume of the corresponding scenario is calculated based on the availability realization value and the probability of scenario occurrence of random scenarios. With the goal of maximizing the sum of weighted expected data of scenarios within the scheduling cycle, a stochastic programming model is constructed with constraints including the number of satellite-to-ground laser links established by the satellite, the number of satellite-to-ground laser links accessed by the ground station, and whether satellite and ground station establish satellite-to-ground laser links. The stochastic programming model is solved to obtain the optimal scheduling result, and the satellite-to-ground laser link is scheduled based on the optimal scheduling result.

[0019] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, performs the following steps: Obtain link availability prediction data for each candidate link, and generate possible random availability scenarios for each candidate link based on the link availability prediction data; the available random scenarios include availability realization values ​​and scenario occurrence probabilities; The weighted expected data volume of the corresponding scenario is calculated based on the availability realization value and the probability of scenario occurrence of random scenarios. With the goal of maximizing the sum of weighted expected data of scenarios within the scheduling cycle, a stochastic programming model is constructed with constraints including the number of satellite-to-ground laser links established by the satellite, the number of satellite-to-ground laser links accessed by the ground station, and whether satellite and ground station establish satellite-to-ground laser links. The stochastic programming model is solved to obtain the optimal scheduling result, and the satellite-to-ground laser link is scheduled based on the optimal scheduling result.

[0020] This application employs the aforementioned satellite-to-ground laser link scheduling method and system, which has the following beneficial effects: 1. By combining the link availability, effective visibility duration, and transmission rate of each candidate link in multiple scenarios into scenario-weighted expected data volume, we can directly quantify the expected contribution of different candidate links to the system throughput. We can also use finite scenarios and their occurrence probabilities to describe the uncertainty of link availability, and use the scenario probability-weighted expected system throughput as the optimization objective to optimize the scheduling scheme, improve the scheduling scheme's adaptability to random environments, and reduce system throughput loss.

[0021] 2. Taking advantage of the fact that the scheduling model does not have cross-time slot coupling constraints, the scheduling cycle model is decomposed into multiple independent time slot sub-problems, reducing the scale of the model to be solved in a single run.

[0022] 3. The weight matrix is ​​copied in columns, the column structure is reorganized, and the rows are copied. The concurrent constraints of multiple laser communication terminals at the satellite end and the ground station end are transformed into one-to-one matching constraints in the extended matrix. The extended matrix is ​​completed into a square matrix using zero elements. By taking the negative of the revenue matrix, the problem of maximizing system throughput is transformed into a problem of minimizing cost. This allows the one-to-one optimal matching algorithm (Munkres matching algorithm) to be called for solving the problem, reducing the difficulty of solving the problem. Attached Figure Description

[0023] Figure 1 A flowchart of a satellite-to-ground laser link scheduling method in one embodiment; Figure 2 This is a comparison chart of system throughput for different link scheduling methods under different numbers of ground stations in one embodiment. Figure 3 This is a comparison chart of system throughput for different link scheduling methods under different upper limits of the number of laser communication terminals at different ground stations in one embodiment. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0025] This application provides a satellite-to-ground laser link scheduling method and system applied to a satellite-to-ground laser communication network consisting of multiple satellites, multiple ground stations, and a ground control center. Each satellite is equipped with at least one laser communication terminal, and each ground station is also equipped with at least one laser communication terminal. Satellites and ground stations establish satellite-to-ground laser communication links through these terminals. The ground control center acquires satellite orbit information, ground station location information, satellite-to-ground visibility window information, link transmission rate, and link availability prediction information, and generates a corresponding satellite-to-ground laser link scheduling scheme based on this information.

[0026] Firstly, based on the above, this application provides a satellite-to-ground laser link scheduling method, including: S100, obtain the link availability prediction data of each candidate link, and generate possible random availability scenarios for each candidate link based on the link availability prediction data; the available random scenarios include availability realization values ​​and scenario occurrence probabilities.

[0027] The link availability prediction data mentioned in this application is generated after predicting the link availability in each future time slot. It is presented in the form of a probability distribution, and the availability realization value and the probability of occurrence of random scenarios can be generated based on the probability distribution.

[0028] During link scheduling, a preset scheduling period is divided into multiple consecutive scheduling time slots, with each scheduling time slot serving as the basic time unit for link resource allocation. Assume a satellite-to-ground laser communication network includes M satellites and K ground stations, with the scheduling period divided into I scheduling time slots. Each satellite can establish a maximum of [number missing] simultaneous connections within any given time slot. Each ground station can connect to a satellite-to-ground laser link at a maximum of [number] simultaneous connections in any given time slot. A satellite-to-ground laser link.

[0029] That is, the set of scheduling time slots is represented as:

[0030] The satellite set is represented as:

[0031] The set of ground stations is represented as:

[0032] Considering the randomness inherent in the availability of satellite-to-ground laser links, in actual satellite-to-ground laser communication, although candidate links may meet the geometric visibility conditions between the satellite and the ground station, the actual availability of the link may be affected by factors such as clouds and weather. Therefore, directly determining the availability of a link based on a single link availability prediction result can easily lead to scheduling results being affected by prediction errors. Therefore, this embodiment does not directly use a single deterministic link availability parameter, but instead constructs multiple possible random link availability scenarios for each candidate link.

[0033] Specifically, for any candidate satellite-to-ground laser link, based on link availability prediction data, multiple possible link availability states for that candidate link are determined, and each of the different link availability states is used as a corresponding available random field. The scenario set is then represented as follows:

[0034] set up Indicates the first In the first scenario, the... Within the first time slot satellite and the first The availability of links between the ground stations is achieved, and ;set up Let represent the probability of the corresponding scenario occurring, satisfying:

[0035] When it's impossible to obtain differentiated probability information between different scenarios, the same probability of occurrence can be set for each random scenario. That is, when each scenario occurs with equal probability, .

[0036] In one embodiment, calculating the weighted expected data volume of the corresponding scenario based on the availability realization value of the random scenario and the probability of scenario occurrence includes: obtaining the first... Within the first time slot satellite and the first Effective visibility duration between ground stations and the corresponding transmission rate of the satellite-to-ground laser link Based on the effective visibility duration and transmission rate Calculate the first Within the first time slot satellite and the first Data volume between ground stations Availability implementation value, scenario occurrence probability, and data volume based on random scenarios; Calculate the first Within the first time slot satellite and the first Weighted expected data volume between ground stations .

[0037] For the Within the first time slot satellite and the first A ground station was established to determine the effective visible duration of candidate satellite-to-ground laser links between the two. And obtain the transmission rate of the satellite-to-ground laser chain corresponding to the candidate link. If a candidate link meets the satellite-to-ground visibility condition, the amount of data that the candidate link can transmit under fully available conditions is determined based on the effective visibility duration and transmission rate. Therefore, a data volume representing the basic data transmission capacity of the candidate link can be established. In one implementation, the data volume coefficient can be determined based on the product of the effective visibility duration and the transmission rate; that is, the longer the effective visibility duration and the higher the transmission rate, the greater the amount of data that the corresponding candidate link can transmit. When there is no effective visibility window between the satellite and the ground station within the corresponding scheduling time slot, the data volume coefficient of the corresponding candidate link is set to zero.

[0038] Specifically, the first The amount of data between the two within a time slot is represented as follows:

[0039] S200 calculates the weighted expected data volume of the corresponding scenario based on the availability realization value of the random scenario and the probability of scenario occurrence.

[0040] Based on the probability of scenario occurrence, scenario link availability, and candidate link data volume coefficient, calculate the first... Within the first time slot satellite and the first Weighted expected data volume between ground stations :

[0041] in, This represents the expected data transmission volume of a candidate link across all random scenarios, assuming the candidate link is scheduled. As shown in the formula above, the scenario-weighted expected data volume considers the candidate link's basic data transmission capacity, link availability under different random scenarios, and the probability of each random scenario occurring. When a candidate link has high availability in multiple random scenarios, its scenario-weighted expected data volume is high; conversely, when a candidate link has high availability only in a few low-probability scenarios and low availability in other high-probability scenarios, its scenario-weighted expected data volume decreases accordingly. Through calculation... This compresses multiple link availability scenarios into a deterministic expected return coefficient that can be directly incorporated into the link scheduling model.

[0042] S300 aims to maximize the sum of weighted expected data of scenarios within the scheduling cycle, and constructs a stochastic programming model with constraints such as the number of satellite-to-ground laser links established by the satellite, the number of satellite-to-ground laser links accessed by the ground station, and whether the satellite and the ground station establish satellite-to-ground laser links.

[0043] In one embodiment, the objective function of the stochastic programming model is:

[0044] The objective function satisfies the following constraints:

[0045]

[0046]

[0047] in, The set of scheduling slots within a scheduling period. ; For satellite collection, ; Gathering at the ground station, ; For the first Within the first time slot satellite and the first The weighted expected data volume for scenarios between ground stations; For the constructed binary link scheduling variable, when the first Within the first time slot satellite and the first When establishing a satellite-to-ground laser link between two ground stations, ,otherwise, ; The constraint on the number of satellite-to-ground laser links established for a satellite means that the number of satellite-to-ground laser links established simultaneously by each satellite in any time slot shall not exceed [a certain limit]. ; The constraint on the number of satellite-to-ground laser links accessed by a ground station means that the number of satellite-to-ground laser links simultaneously accessed by each ground station in any time slot shall not exceed [a certain limit]. ; Constraints are set for whether to establish a satellite-to-ground laser link between the satellite and the ground station, which are used to limit the link scheduling variable to a binary variable.

[0048] S400 solves the stochastic programming model to obtain the optimal scheduling result and schedules the satellite-to-ground laser link based on the optimal scheduling result.

[0049] In one embodiment, solving a stochastic programming model includes: constructing a scenario-based weighted expected data set. Weight matrix; the weight matrix in which the first... The element represents the first element. satellite and the first The weighted expected data volume of the scenario between the two ground stations; the constraint terms of the stochastic programming model are used to select elements in the weight matrix to simulate the construction of the satellite-to-ground laser link; wherein, the link scheduling variable corresponding to the selected element is set to The number of elements selected in each row does not exceed Each satellite can establish a maximum of One laser communication link; the number of selected elements in each column does not exceed [number missing]. The maximum number of stations that can be built per ground station A laser communication link is constructed; the objective is to maximize the sum of the weighted expected data volume of the simulated satellite-to-ground laser link within the scheduling period.

[0050] Since neither the objective function nor the constraints contain cross-slot coupling terms, the scheduling period model can be decomposed into... Each time slot is a separate subproblem. The optimal solution obtained by solving each time slot subproblem separately is consistent with the optimal solution obtained by directly solving the entire scheduling cycle model.

[0051] This application is directed to the first Each time slot is constructed based on the scenario-weighted expected data volume of all candidate links. weight matrix :

[0052] matrix The The element represents the first element. satellite and the first The expected amount of data that a ground station can obtain when establishing a link. If a candidate link cannot be established within the corresponding time slot, the corresponding matrix element is set to zero.

[0053] Considering that the objective function and constraints of the stochastic programming model do not have cross-slot coupling terms, and the optimal solution obtained based on each slot is consistent with the optimal solution obtained based on the entire scheduling cycle, it is only necessary to construct the optimization problem for a single slot. For the th Each time slot, input is weight matrix The first in the matrix The element represents the first element. satellite and the first The expected amount of scenario probability weighted data between the ground stations, i.e. .

[0054] In the weight matrix Select an element and set the link scheduling variable corresponding to the selected element to... The number of elements selected in each row does not exceed [a certain limit]. Each satellite can establish a maximum of One laser communication link; the number of selected elements in each column does not exceed [number missing]. The maximum number of stations that can be built per ground station The solution is to solve for a simulated satellite-to-ground laser communication link, with the objective of maximizing the sum of weighted expected data of the simulated satellite-to-ground laser link within the scheduling period.

[0055] In one embodiment, the solution aims to maximize the sum of expected data volumes of the simulated satellite-to-ground laser links within the scheduling period, including: establishing a maximum of [number missing] data points per satellite. Each laser communication link and each ground station can establish a maximum of [number] links. The multi-terminal concurrent link establishment constraint represented by a laser communication link is transformed into a one-to-one matching constraint, which will be used to connect a laser communication link equipped with a laser communication link. The satellite equivalent of a laser communication terminal is Each satellite replica is equipped with a laser communication terminal; a satellite equipped with... The ground station of a laser communication terminal is equivalent to Each ground station replica is equipped with a laser communication terminal; dimensions are determined based on one-to-one matching constraints. The weight matrix is ​​expanded to dimension 1. The target matrix is ​​obtained; the Munkres matching algorithm is used to solve for the target matrix.

[0056] In one embodiment, the dimension is determined based on a one-to-one matching constraint. The weight matrix is ​​expanded to dimension 1. The target matrix includes: establishing a dimension of matrix ,for , and ,Will The weight matrix of the first Copy each element to The There are 10 locations; where i is the i-th time slot; rearranged according to the ground station terminal copy number. The columns form a matrix. Column expansion results ; to matrix Copy line by line Next, the dimension is formed as Target matrix .

[0057] Since each satellite may be equipped with multiple laser communication terminals, and each ground station may also be equipped with multiple laser communication terminals, the original weight matrix does not correspond to a standard one-to-one matching problem. For example, a satellite equipped with P laser communication terminals can participate in at most P links simultaneously within a scheduling time slot, while a standard one-to-one matching model typically requires a node to participate in at most one match. Therefore, in order to solve the above link scheduling problem using a one-to-one matching algorithm, this embodiment expands the dimension of the weight matrix.

[0058] In this embodiment, the core purpose of dimensional expansion is not to change the link revenue between the original satellite and the ground station, but to perform equivalent splitting of the original network nodes, thereby transforming the multi-terminal concurrent link establishment constraint into a one-to-one matching constraint between multiple node replicas.

[0059] To convert the aforementioned multi-terminal concurrent connection establishment constraints into one-to-one matching constraints, a chip equipped with... The satellite equivalent of a laser communication terminal is Each satellite replica is equipped with a laser communication terminal; a satellite equipped with... The ground station of a laser communication terminal is equivalent to Each ground station replica is equipped with a laser communication terminal. This application expands the dimensions of the weight matrix; the steps for expanding the dimensions of the weight matrix are as follows: Step 1: Copy and expand column elements. Establish a dimension of... matrix .for , and The original weight matrix is ​​changed to the first... Copy each element to The The element mapping is as follows: (Number of positions)

[0060] Step two, for the matrix Perform column rearrangement to generate column-by-column expansion results That is, for , and , matrix The Column mapping to matrix The The column and element mapping are as follows:

[0061] Step 3, for the matrix Perform row copy expansion. That is, for , and , matrix Copy line by line Next, the dimension is formed as matrix The element mapping is as follows:

[0062] In one embodiment, solving the target matrix using the Munkres matching algorithm includes: constructing an extended square matrix based on the target matrix, and then... Placed in front of the expanded array row, before Columns are arranged in order, and the remaining positions are filled with zero elements to obtain a square matrix. Among them, the order of the expanded square matrix ; against the formation Invert all elements to construct the cost matrix; use the Munkres matching algorithm to solve the problem of selecting elements from the cost matrix that minimizes the total cost, and obtain the optimal scheduling result.

[0063] One-to-one matching algorithms require that the number of rows and columns of the input matrix be the same. In this application, the order D of the extended square matrix is... and The maximum value in That is, the order D of the extended matrix is ​​calculated as follows:

[0064] matrix A square matrix with an embedding dimension of D×D In the middle. Then, for and ,have:

[0065] For square array Fill all positions except the embedded element positions with zero elements.

[0066] Zero-fill elements correspond to virtual satellite replicas or virtual ground station replicas, generating no expected data volume. Their corresponding revenue elements are all set to zero, thus not altering the weight relationships between the original candidate links. Therefore, virtual nodes are only used to satisfy the matrix dimension requirements of the one-to-one matching algorithm and do not change the original revenue relationship between the actual satellites and ground stations. After obtaining the matching results, the matching relationships associated with virtual nodes are not restored to the actual link scheduling relationships.

[0067] After the above dimensional expansion, the revenue matrix corresponding to the current scheduling slot is obtained. Since the optimization objective of this embodiment is to maximize the sum of the scenario-weighted expected data volume corresponding to the selected link, and the Munkres matching algorithm is used to solve for the matching result with the minimum total cost, the revenue matrix is ​​further converted into a cost matrix. Specifically, the cost matrix can be constructed by negating each element in the revenue matrix. :

[0068] Through the above transformation, the problem of maximizing profit is transformed into a problem of minimizing cost. With the candidate matching positions remaining unchanged, the position with the larger profit value corresponds to the smaller cost value. Therefore, the matching relation that maximizes the total profit is the same as the matching relation that minimizes the total cost.

[0069] The obtained cost matrix is ​​input into the Munkres matching algorithm, which establishes a one-to-one matching relationship between the row and column sets of the expanded matrix. During the matching process, each actual satellite replica is matched with at most one ground station replica, and each actual ground station replica is matched with at most one satellite replica; for virtual rows or virtual columns generated by matrix filling, the matching results do not correspond to actual communication links.

[0070] Specifically, the cost matrix is ​​input into the Munkres matching algorithm. The algorithm first performs row reduction on the cost matrix so that each row contains at least one zero element; then it performs column reduction so that each column contains at least one zero element; then it covers all zero elements with the minimum number of horizontal and vertical lines. When the number of covering lines is less than the order of the matrix, the smallest uncovered element is determined, the minimum value is subtracted from the uncovered elements, and the minimum value is added to the intersection of the covering lines. The zero-element covering and matrix adjustment are repeated until a complete one-to-one match is obtained.

[0071] set up Representing the cost matrix The matching results in; where, Indicates the first One time slot; The row of the cost matrix represents the first row. One element; The column representing the cost matrix is ​​the first column. Each element. Indicates the satellite copy number, Let represent the ground station replica number. For the original m-th satellite and k-th ground station, check the matching results of each corresponding satellite replica and ground station replica in turn. There exists at least one pair of p and q such that:

[0072] At this point, the original link scheduling variable is set to:

[0073] When no matching result exists, Set to 0.

[0074] If the position of the virtual row or column formed by zero padding does not correspond to the actual link, it will be discarded when restoring the original link relationship.

[0075] After obtaining the one-to-one matching results of the expanded array, this application maps the replica matching relationships back to the original satellite and ground station numbers based on the satellite replica number and ground station replica number corresponding to the matching position. For example, when a satellite replica corresponding to a certain satellite matches a ground station replica corresponding to a certain ground station, the replica matching relationship is restored to the satellite-to-ground laser link between the corresponding original satellite and the original ground station. Matching relationships formed with virtual satellite replicas or virtual ground station replicas are discarded during the restoration process.

[0076] After restoring the original satellite-ground station relationship, the number of matches between each original satellite and each original ground station is counted. Since each original satellite corresponds to P satellite replicas, and each satellite replica participates in at most one match, the number of matches for the same original satellite in the current scheduling slot will not exceed P; similarly, the number of matches for the same original ground station in the current scheduling slot will not exceed Q. Therefore, the restored link scheduling result naturally satisfies the terminal resource constraints of the satellite and ground stations in the original stochastic programming model.

[0077] Perform the one-to-one matching process described above for each time slot to obtain all link scheduling variables. The sum of the optimal matching weights for each time slot is the maximum expected system throughput within the scheduling period. This is based on link scheduling variables. This function outputs a list of link establishment objects for both the satellite and ground stations within each scheduling time slot, the expected data transmission volume for each time slot, and the maximum expected system throughput within the scheduling period. Here, the maximum expected system throughput is the data volume at which the sum of the weighted expected data volumes for each scenario reaches its maximum within the system scheduling period.

[0078] The ground control center generates link establishment, maintenance, and release commands based on the time slot start time, link establishment target, effective visibility duration, and terminal number, and issues these commands to the satellite and ground station before the start of the corresponding time slot. Under rolling scheduling conditions, the scenario set, scenario probability, and weight matrix are updated based on the latest link availability prediction data, and a new link scheduling plan for subsequent time slots is generated.

[0079] In one simulation setting, the number of satellites Number of scenes The upper limit of the number of laser communication terminals at both the satellite and ground stations meets the requirements. Number of ground stations Set to respectively , , , and Comparing the method of this invention with existing deterministic methods, greedy heuristic algorithms, simulated annealing algorithms, and the results are as follows: Figure 2 As shown.

[0080] As the number of ground stations increases, the number of candidate links available to the system also increases, leading to an overall improvement in system throughput. Existing deterministic methods do not comprehensively consider stochastic scenarios related to link availability, resulting in lower system throughput compared to the method described in this application. Greedy heuristic algorithms construct scheduling schemes by successively selecting the candidate link with the highest current benefit, which may limit subsequent link combinations due to local selections; simulated annealing algorithms rely on random search, making it difficult to stably approximate the optimal solution for large-scale scheduling problems within a finite number of iterations. This application solves for the link combination in each time slot through dimensional expansion and optimal matching, achieving the highest system throughput for all ground station numbers.

[0081] In another simulation setting, the number of ground stations Number of satellites Number of scenes Maximum number of laser communication terminals at ground stations from Increase to The system throughput of different scheduling methods is as follows: Figure 3 As shown.

[0082] along with As the number of satellites increases, the ground station can simultaneously access more satellites, and the overall system throughput shows a non-decreasing trend. After increasing to a certain value, the system throughput gradually saturates. This is because the total number of links that can be established in any given time slot is limited. and minimum value Limitations; when ground station terminal resources no longer constitute a bottleneck, continue to increase... No longer generating significant throughput gains. The method described in this application is applicable to various... The throughput gains under each value are no less than those of other comparative methods.

[0083] Secondly, this application also provides a satellite-to-ground laser link scheduling system, comprising: The random scenario generation module is used to obtain the link availability prediction data of each candidate link and generate possible available random scenarios for each candidate link based on the link availability prediction data; the available random scenarios include availability realization values ​​and scenario occurrence probabilities. The weighted expected data volume calculation module is used to calculate the weighted expected data volume of the corresponding scenario based on the availability realization value and the probability of scenario occurrence of the random scenario. The model building module is used to construct a stochastic programming model with the goal of maximizing the sum of weighted expected data of the scenario within the scheduling period, and with constraints such as the number of satellite-to-ground laser links established by the satellite, the number of satellite-to-ground laser links accessed by the ground station, and whether the satellite and the ground station have established satellite-to-ground laser links. The result generation module is used to solve the stochastic programming model, obtain the optimal scheduling result, and schedule the satellite-to-ground laser link based on the optimal scheduling result.

[0084] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0085] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0086] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0087] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0088] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0089] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0090] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A satellite-to-ground laser link scheduling method, characterized in that, include: Obtain link availability prediction data for each candidate link, and generate possible random availability scenarios for each candidate link based on the link availability prediction data; the available random scenarios include availability realization values ​​and scenario occurrence probabilities; The weighted expected data volume of the corresponding scenario is calculated based on the availability realization value and the probability of scenario occurrence of random scenarios. With the goal of maximizing the sum of weighted expected data of scenarios within the scheduling cycle, a stochastic programming model is constructed with constraints including the number of satellite-to-ground laser links established by the satellite, the number of satellite-to-ground laser links accessed by the ground station, and whether satellite and ground station establish satellite-to-ground laser links. The stochastic programming model is solved to obtain the optimal scheduling result, and the satellite-to-ground laser link is scheduled based on the optimal scheduling result.

2. The method according to claim 1, characterized in that, The weighted expected data volume for a given scenario is calculated based on the availability realization value and the probability of scenario occurrence in a random scenario, including: Get the Within the first time slot satellite and the first Effective visibility duration between ground stations and the corresponding transmission rate of the satellite-to-ground laser link ; Based on valid visibility duration and transmission rate Calculate the first Within the first time slot satellite and the first Data volume between ground stations ; Availability implementation value, scenario occurrence probability, and data volume based on random scenarios. Calculate the first Within the first time slot satellite and the first Weighted expected data volume between ground stations .

3. The method according to claim 1, characterized in that, The objective function of the stochastic programming model is: The objective function satisfies the following constraints: in, The set of scheduling slots within a scheduling period. ; For satellite collection, ; Gathering at the ground station, ; For the first Within the first time slot satellite and the first The weighted expected data volume for scenarios between ground stations; For the constructed binary link scheduling variable, when the first Within the first time slot satellite and the first When establishing a satellite-to-ground laser link between two ground stations, ,otherwise, ; The constraint on the number of satellite-to-ground laser links established for a satellite means that the number of satellite-to-ground laser links established simultaneously by each satellite in any time slot shall not exceed [a certain limit]. ; The constraint on the number of satellite-to-ground laser links accessed by a ground station means that the number of satellite-to-ground laser links simultaneously accessed by each ground station in any time slot shall not exceed [a certain limit]. ; Constraints are set for whether to establish a satellite-to-ground laser link between the satellite and the ground station, which are used to limit the link scheduling variable to a binary variable.

4. The method according to any one of claims 1 to 3, characterized in that, Solving stochastic programming models includes: Construction of scenario-based weighted expected data volume Weight matrix; the weight matrix in which the first... The element represents the first element. satellite and the first The weighted expected data volume for scenarios between ground stations; Constraints based on a stochastic programming model are used to select elements in the weight matrix to simulate the construction of a satellite-to-ground laser link; where the link scheduling variable corresponding to the selected element is set to... The number of elements selected in each row does not exceed Each satellite can establish a maximum of One laser communication link; the number of selected elements in each column does not exceed [number missing]. The maximum number of stations that can be built per ground station One laser communication link; The objective is to maximize the sum of the weighted expected data volume of the simulated satellite-to-ground laser link within the scheduling period.

5. The method according to claim 4, characterized in that, The solution aims to maximize the sum of expected data volumes for the simulated satellite-to-ground laser link scenario within the scheduling period, including: Each satellite will be established at most Each laser communication link and each ground station can establish a maximum of [number] links. The multi-terminal concurrent link establishment constraint represented by a laser communication link is transformed into a one-to-one matching constraint, which will be used to connect a laser communication link equipped with a laser communication link. The satellite equivalent of a laser communication terminal is Each satellite replica is equipped with a laser communication terminal; a satellite equipped with... The ground station of a laser communication terminal is equivalent to Each ground station replica is equipped with a laser communication terminal. Based on one-to-one matching constraints, the dimensions are... The weight matrix is ​​expanded to dimension 1. The target matrix; The target matrix is ​​solved using the Munkres matching algorithm.

6. The method according to claim 5, characterized in that, Based on one-to-one matching constraints, the dimensions are... The weight matrix is ​​expanded to dimension 1. The target matrix includes: Establish dimensions as matrix ,for , and ,Will The weight matrix of the first Copy each element to The There are _ ... Rearranged according to ground station terminal copy number The columns form a matrix. Column expansion results ; matrix Copy line by line Next, the dimension is formed as Target matrix .

7. The method according to claim 5, characterized in that, The solution is obtained using the Munkres matching algorithm based on the target matrix, including: Construct an extended square matrix based on the target matrix, and then... Placed in front of the expanded array row, before Columns are arranged in order, and the remaining positions are filled with zero elements to obtain a square matrix. Among them, the order of the expanded square matrix ; Opponent Invert all elements of the matrix to construct the cost matrix; The Munkres matching algorithm is used to solve the problem of selecting elements from the cost matrix that minimizes the total cost, thus obtaining the optimal scheduling result.

8. A satellite-to-ground laser link scheduling system, characterized in that, include: The random scenario generation module is used to obtain the link availability prediction data of each candidate link and generate possible available random scenarios for each candidate link based on the link availability prediction data; the available random scenarios include availability realization values ​​and scenario occurrence probabilities. The weighted expected data volume calculation module is used to calculate the weighted expected data volume of the corresponding scenario based on the availability realization value and the probability of scenario occurrence of the random scenario. The model building module is used to construct a stochastic programming model with the goal of maximizing the sum of weighted expected data of the scenario within the scheduling period, and with constraints such as the number of satellite-to-ground laser links established by the satellite, the number of satellite-to-ground laser links accessed by the ground station, and whether the satellite and the ground station have established satellite-to-ground laser links. The result generation module is used to solve the stochastic programming model, obtain the optimal scheduling result, and schedule the satellite-to-ground laser link based on the optimal scheduling result.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images