Satellite networking communication method, device and equipment

Abstract

The invention provides a satellite networking communication method, device and equipment, and the method comprises the steps: carrying out the space-time window decomposition processing of a satellite constellation orbit parameter based on orbit phase synchronization, and decomposing a long-period scheduling problem into a plurality of short-period sub-problems; performing state transition conflict detection processing on the double-feed antenna system according to the space-time window decomposition result, and determining the working state of the feed antenna; parallel neighborhood search processing based on reinforcement learning is carried out according to the feed antenna resource configuration scheme, and a communication link scheduling strategy is generated through multi-thread collaborative optimization; and performing dynamic fusion processing based on state prediction according to the parallel optimization scheduling scheme, and adjusting a satellite-ground station communication link establishment time sequence through learning type template matching. Through space-time decomposition, intelligent conflict detection, reinforcement learning optimization and dynamic fusion, the problems of complexity and real-time performance of large-scale satellite networking communication scheduling are solved, and the scheduling efficiency and the communication quality are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of satellite networking, and particularly to a satellite networking communication method, device and equipment. Background Art

[0002] With the rapid development of low-orbit satellite constellation technology, large-scale satellite Internet projects have been successively deployed, and the number of satellites has increased from dozens in the traditional sense to thousands or even tens of thousands. This sharp expansion in scale has transformed the satellite-ground station communication scheduling from a small-scale task-driven mode to a large-scale network optimization problem. Traditional satellite scheduling methods mainly rely on heuristic methods such as greedy algorithms and genetic algorithms, and adopt a single optimization strategy to handle the communication link allocation between satellites and ground stations.

[0003] When dealing with large-scale satellite constellation communication scheduling, the existing technologies generally adopt simplified constraint models and single-objective optimization methods, and cannot take into account complex factors such as satellite orbit dynamics characteristics, dual-feed antenna hardware constraints, and multi-temporal window coupling at the same time. As a result, resource conflicts and low efficiency frequently occur in the generated scheduling schemes during actual deployment, seriously affecting the overall performance of large-scale satellite networking communication systems. Summary of the Invention

[0004] The main purpose of the present invention is to solve the technical problems in the existing large-scale satellite communication scheduling technology, where it is unable to effectively handle multi-constraint coupling and multi-objective optimization, resulting in low-quality and poor practicability of the scheduling scheme; In the first aspect of the present invention, a satellite networking communication method is provided, and the satellite networking communication method includes: Perform space-time window decomposition processing based on orbit phase synchronization on satellite constellation orbit parameters, decompose the long-period scheduling problem of satellite-ground station communication into multiple short-period sub-problems, and obtain the space-time window decomposition result; Perform state transition conflict detection processing on the dual-feed antenna system according to the space-time window decomposition result, determine the working states of the feed antennas in different space-time windows, and obtain the feed antenna resource allocation scheme; Perform parallel neighborhood search processing based on reinforcement learning according to the feed antenna resource allocation scheme, and generate a communication link scheduling strategy through multi-threaded collaborative optimization to obtain a parallel optimization scheduling scheme; Perform dynamic fusion processing based on state prediction according to the parallel optimization scheduling scheme, and adjust the satellite-ground station communication link establishment timing through learning-based template matching to obtain a satellite networking communication scheduling scheme.

[0005] Optionally, in the first implementation manner of the first aspect of the present invention, the processing of decomposing the satellite constellation orbit parameters based on orbit phase synchronization into spatio-temporal windows decomposes the long-period scheduling problem of satellite-ground station communication into multiple short-period sub-problems, and the spatio-temporal window decomposition result includes: Calculate the visible arc segment of each satellite relative to the ground station according to the satellite constellation orbit parameters, and determine the effective communication time boundary according to the start time and end time of the visible arc segment; Generate a visible time window between the satellite and the ground station based on the effective communication time boundary, and perform similarity analysis processing on the satellite constellation orbit parameters according to the visible time window to identify the orbit phase synchronization window; Group and classify the satellites according to the orbit phase synchronization window to obtain multiple satellite groups, and perform time-domain segmentation processing based on orbit periodicity on the long-period scheduling problems of the multiple satellite groups to obtain multiple short-period sub-problems; Associate and organize the visible time window, the orbit phase synchronization window, and the short-period sub-problems to obtain a spatio-temporal window decomposition result including the time dimension, the space dimension, and the task dimension.

[0006] Optionally, in the second implementation manner of the first aspect of the present invention, the processing of detecting state transition conflicts in the dual-feed antenna system according to the spatio-temporal window decomposition result to determine the working state of the feed antenna under different spatio-temporal windows and obtain the feed antenna resource allocation scheme includes: Perform scenario classification processing on the dual-feed antenna system of the satellite based on the link arc segment overlap relationship according to the spatio-temporal window decomposition result to identify three communication scenarios: partial overlap, complete inclusion, and complete separation; Analyze the antenna state transition rules in each communication scenario, construct a directed state graph including state nodes and transition paths, and obtain the feed antenna state transition graph; Detect the competition conflict of multiple ground stations for the same antenna and the antenna switching time overlap conflict within the same time window according to the feed antenna state transition graph, determine the conflict type and the influence range, and obtain the state transition conflict list; Accumulatively calculate the physical rotation time of the antenna, the radio frequency link establishment time, and the signal synchronization time according to the state transition conflict list, determine the minimum switching time threshold in each communication scenario, and obtain the antenna switching time constraint; Perform balanced allocation processing on the load of the dual-feed antenna system according to the antenna switching time constraint to obtain the feed antenna resource allocation scheme.

[0007] Optionally, in the third implementation manner of the first aspect of the present invention, the cumulative calculation of the physical rotation time of the antenna, the radio frequency link establishment time, and the signal synchronization time according to the state transition conflict list to determine the minimum handover time threshold in each communication scenario, and the antenna handover time constraint is obtained as follows: Analyze the delay impact of the competition conflict on the physical rotation time of the antenna and the additional overhead of the overlapping conflict on the radio frequency link establishment time according to the state transition conflict list, quantify the time penalty factors of each conflict type, and obtain the conflict time penalty coefficient; Perform a weighted cumulative operation on the physical rotation time of the antenna, the radio frequency link establishment time, and the signal synchronization time with the corresponding penalty coefficients according to the conflict time penalty coefficient to obtain the corrected handover time; Set different safety margin coefficients for the three scenarios of partial overlap, complete inclusion, and complete separation according to the handover time, and determine the minimum time interval to avoid conflicts through scenario weighted average calculation to obtain the sub-scenario handover time threshold; Perform constraint condition generation processing according to the sub-scenario handover time threshold, convert the time threshold into a hard constraint condition for the feeding antenna work scheduling, and limit the earliest available time and the latest release time of the antenna in different scenarios to obtain the antenna handover time constraint.

[0008] Optionally, in the fourth implementation manner of the first aspect of the present invention, the parallel optimization processing guided by reinforcement learning is performed according to the feeding antenna resource configuration scheme, and a communication link scheduling strategy is generated through multi-threaded collaborative optimization. The parallel optimization scheduling scheme obtained includes: Extract the communication link state information according to the feeding antenna resource configuration scheme to form a multi-dimensional state space, and select a variety of optimization operation operators to obtain the current state and the set of executable operations; Perform reinforcement learning processing based on orbit prediction according to the current state and the set of executable operations, calculate the long-term communication quality trend by predicting the position changes of the satellite in the next preset number of orbit periods, and obtain the prediction result; Integrate the prediction result into the reward function for Q-value iterative update to obtain the intelligent operation selection strategy, and perform multi-threaded parallel optimization processing on the short-period sub-problems in the spatio-temporal window decomposition result according to the intelligent operation selection strategy to obtain the multi-threaded collaborative optimization result; Integrate the communication link scheduling strategies of each thread according to the multi-threaded collaborative optimization result to obtain the parallel optimization scheduling scheme.

[0009] Optionally, in the fifth implementation manner of the first aspect of the present invention, the dynamic fusion processing based on state prediction is performed according to the parallel optimization scheduling scheme, and the satellite-ground station communication link establishment timing is adjusted through learning-based template matching to obtain the satellite networking communication scheduling scheme, including: According to the parallel optimization scheduling scheme and combining with the state information of the satellite obtained in real time, the sliding window technology is used to perform advance prediction calculations for a preset number of orbital periods on the satellite state change, and a satellite state prediction result is obtained; According to the satellite state prediction result, a constraint relaxation algorithm is performed on the adjacent sub-problem boundaries, and the scheduling conflict between time-space windows is resolved by dynamically adjusting the strictness of the boundary constraints, and a boundary conflict resolution scheme is obtained; According to the boundary conflict resolution scheme, template matching processing based on historical successful patterns is performed, and a verified learning template is selected for rapid application to obtain a learning template; According to the learning template, the timing for establishing the satellite-ground station communication link is adjusted, and the real-time state adjustment strategy and the emergency switching plan are integrated to obtain a satellite networking communication scheduling scheme.

[0010] Optionally, in the sixth implementation manner of the first aspect of the present invention, the adjusting the timing for establishing the satellite-ground station communication link according to the learning template, integrating the real-time state adjustment strategy and the emergency switching plan, and obtaining the satellite networking communication scheduling scheme includes: Mapping the successful scheduling pattern in the learning template to the current satellite orbit configuration, and adjusting the connection time points of the satellite-ground station through time offset and priority reordering to obtain an optimized link timing; According to the optimized link timing and the satellite state information, preset dynamic adjustment rules are matched to obtain a real-time state adjustment strategy; According to the real-time state adjustment strategy, standby communication links and switching times are pre-set for emergencies, and a multi-level emergency response mechanism and a resource reallocation plan are formulated to obtain an emergency switching plan; According to the optimized link timing, the real-time state adjustment strategy and the emergency switching plan, scheme integration processing is performed to obtain a satellite networking communication scheduling scheme.

[0011] The second aspect of the present invention provides a satellite networking communication device, and the satellite networking communication device includes: A space-time decomposition module, configured to perform space-time window decomposition processing based on orbital phase synchronization on the orbital parameters of the satellite constellation, decompose the long-period scheduling problem of satellite-ground station communication into multiple short-period sub-problems, and obtain a space-time window decomposition result; A conflict detection module, configured to perform state transition conflict detection processing on the dual-feed antenna system according to the space-time window decomposition result, determine the working state of the feed antenna in different space-time windows, and obtain a feed antenna resource configuration plan; An intelligent optimization module for performing parallel neighborhood search processing based on reinforcement learning according to the feeder antenna resource configuration scheme, generating a communication link scheduling strategy through multi-threaded collaborative optimization, and obtaining a parallel optimization scheduling scheme; A dynamic fusion module for performing dynamic fusion processing based on state prediction according to the parallel optimization scheduling scheme, adjusting the satellite-ground station communication link establishment timing through learning-based template matching, and obtaining a satellite networking communication scheduling scheme.

[0012] A third aspect of the present invention provides a satellite networking communication device, including: a memory and at least one processor, instructions are stored in the memory, and the memory and the at least one processor are interconnected by a line; the at least one processor calls the instructions in the memory to cause the satellite networking communication device to execute the steps of the above-mentioned satellite networking communication method.

[0013] The above-mentioned satellite networking communication method, device and equipment decompose the long-term scheduling problem into multiple short-term sub-problems by performing spatio-temporal window decomposition processing based on orbital phase synchronization on the satellite constellation orbit parameters; perform state transition conflict detection processing on the dual-feeder antenna system according to the spatio-temporal window decomposition result to determine the working state of the feeder antenna; perform parallel neighborhood search processing based on reinforcement learning according to the feeder antenna resource configuration scheme, and generate a communication link scheduling strategy through multi-threaded collaborative optimization; perform dynamic fusion processing based on state prediction according to the parallel optimization scheduling scheme, and adjust the satellite-ground station communication link establishment timing through learning-based template matching. The present invention solves the complexity and real-time problems of large-scale satellite networking communication scheduling through spatio-temporal decomposition, intelligent conflict detection, reinforcement learning optimization and dynamic fusion, and significantly improves the scheduling efficiency and communication quality.

[0014] Other features and advantages of the present invention will be described in the following specification, and, in part, will be obvious from the specification, or will be understood by implementing the present invention. The objectives and other advantages of the present invention are achieved and obtained by the structures specifically pointed out in the specification, claims and drawings.

[0015] To make the above objectives, features and advantages of the present invention more obvious and understandable, the following specific preferred embodiments are given, and in conjunction with the accompanying drawings, the detailed description is as follows. Description of the Drawings

[0016] Figure 1 It is a schematic diagram of the first embodiment of the satellite networking communication method in the embodiments of the present invention; Figure 2 It is a schematic diagram of an embodiment of the satellite networking communication device in the embodiments of the present invention; Figure 3 It is a schematic diagram of an embodiment of the satellite networking communication equipment in the embodiments of the present invention. Specific Embodiments

[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0018] The terms "including" and "having" and any variations thereof mentioned in the embodiments of the present invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device including a series of steps or units is not limited to the listed steps or units, but optionally further includes other steps or units not listed, or optionally further includes other steps or units inherent to these processes, methods, products, or devices.

[0019] To facilitate the understanding of this embodiment, a satellite networking communication method disclosed in the embodiments of the present invention will be introduced in detail first. As Figure 1 shown, this method includes the following steps: 101. Perform space-time window decomposition processing on the satellite constellation orbit parameters based on orbit phase synchronization, decompose the long-period scheduling problem of satellite-ground station communication into multiple short-period sub-problems, and obtain the space-time window decomposition result; In an embodiment of the present invention, the performing space-time window decomposition processing on the satellite constellation orbit parameters based on orbit phase synchronization, decomposing the long-period scheduling problem of satellite-ground station communication into multiple short-period sub-problems, and obtaining the space-time window decomposition result includes: calculating the visible arc segment of each satellite relative to the ground station according to the satellite constellation orbit parameters, and determining the effective communication time boundary according to the start time and end time of the visible arc segment; generating a visible time window between the satellite and the ground station based on the effective communication time boundary, and performing similarity analysis processing on the satellite constellation orbit parameters according to the visible time window to identify the orbit phase synchronization window; grouping and classifying the satellites according to the orbit phase synchronization window to obtain multiple satellite groups, and performing time-domain segmentation processing based on orbit periodicity on the long-period scheduling problems of the multiple satellite groups to obtain multiple short-period sub-problems; associating and organizing the visible time window, the orbit phase synchronization window, and the short-period sub-problems to obtain a space-time window decomposition result including the time dimension, the space dimension, and the task dimension.

[0020] Specifically, for the spatio-temporal window decomposition process, precise geometric calculations are first performed on the orbital parameters of the satellite constellation. The system extracts the orbital elements of each satellite, including the semi-major axis, eccentricity, orbital inclination, right ascension of the ascending node, argument of perigee, and mean anomaly, and calculates the three-dimensional position coordinates of the satellite at any moment through the Keplerian orbit model. The Earth ellipsoid model uses the WGS-84 coordinate system, and the system determines the position vector of the ground station in the geocentric coordinate system based on its longitude, latitude coordinates, and altitude. The calculation of the visible arc segment is based on geometric line-of-sight analysis. The system establishes the line-of-sight vector from the ground station to the satellite and determines whether this vector intersects the Earth's surface. In the specific calculation process, the system constructs a spherical equation with the Earth's center as the origin, and the intersection of the line-of-sight vector and the sphere is judged using the quadratic equation solution method. When the line-of-sight vector has no intersection with the Earth's surface, the satellite is visible to the ground station. The system traverses the discrete time points on the satellite orbit to identify the conversion moments of the visibility state, that is, the rising moment when the satellite changes from invisible to visible and the setting moment when the satellite changes from visible to invisible. These two moments form the start and end boundaries of the visible arc segment.

[0021] Specifically, for the treatment of the atmospheric refraction effect, the standard atmospheric model is adopted. The system calculates the refraction angle correction value according to the atmospheric density distribution on the signal propagation path. In the specific implementation, the system divides the atmosphere into multiple concentric spherical shells, and the refractive index within each spherical shell is calculated according to the exponential decay model. The signal path tracing adopts the ray tracing algorithm. Starting from the ground station emission point, the refraction direction is calculated at each atmospheric interface according to Snell's law until the satellite position is reached. The bending of the line-of-sight path caused by the refraction effect causes the apparent position of the satellite to deviate from the geometric position. The system determines the actual visible time boundary considering atmospheric refraction through iterative calculations. After obtaining the geometric visible arc segment, the system further deducts the satellite antenna pointing adjustment time and the radio frequency signal establishment time. The antenna adjustment time is calculated according to the angular velocity limit of the satellite attitude control system, and the signal establishment time includes the cumulative duration of carrier acquisition, symbol synchronization, and frame synchronization. After these corrections, the system obtains the effective communication time boundary corresponding to each satellite-ground station.

[0022] Specifically, based on the effective communication time boundary, the system constructs a standardized visible time window data structure. Each time window record contains a satellite identifier, a ground station identifier, the Julian date of the window start time, the Julian date of the window end time, the window duration, and an estimated communication quality parameter. The system then performs a clustering analysis on the orbital parameters of all satellites, calculating the Euclidean distance between orbital elements as a similarity metric. The similarity of orbital altitude is evaluated by the difference in semi-major axis, the similarity of inclination is calculated by the difference in the cosine value of the orbital inclination, and the similarity of period is directly compared by the numerical difference in orbital period. The system sets a similarity threshold and groups the satellites that meet the threshold conditions into the same similarity group. The identification of the orbital phase synchronization window is based on the calculation of the phase difference. The system selects a reference time to calculate the true anomaly of each satellite and identifies the periodic pattern of the phase relationship by analyzing the variation law of the true anomaly over time. When the phase differences of multiple satellites remain relatively stable within a certain time window, the system marks this time period as the orbital phase synchronization window. The judgment of phase stability uses sliding window variance analysis, and when the variance of the phase difference is lower than the set threshold, the phase relationship is considered stable.

[0023] Specifically, after the orbital phase synchronization window is determined, the system groups and classifies the satellites according to the phase relationship. The grouping algorithm uses the connected component detection method in graph theory, regarding the satellites with a stable phase relationship as the connected nodes in the graph, and finally obtaining several independent satellite groups. For each satellite group, the system performs time-domain segmentation processing, and the segmentation strategy is based on the greatest common divisor of the orbital periods of the satellites within the group. The system first calculates the greatest common divisor of the orbital periods of all satellites within the group and uses integer multiples of this common divisor as the basic unit for time-domain segmentation. The long-period scheduling problem is segmented according to this basic unit, and the scheduling problem within each time segment constitutes an independent short-period sub-problem. During the time-domain segmentation process, the system ensures that the continuity constraint at the boundary of adjacent time segments is maintained, and a smooth transition is achieved by setting an overlapping buffer near the time segment boundary.

[0024] Specifically, the system finally organizes the visible time window, the orbital phase synchronization window, and the short-period sub-problems in a structured manner. The organization in the time dimension uses a time index tree structure, arranging all time windows in chronological order and establishing a fast query index. The organization in the space dimension is based on the geographical distribution of satellite groups and ground stations, constructing a two-dimensional spatial grid index system. The organization in the task dimension sorts the short-period sub-problems according to priority and dependency relationships, establishing a hierarchical structure for task execution. The three dimensions are associated through a cross-reference table, and each data element contains pointers to related data in other dimensions. The entire data structure adopts a multi-level index design, supporting fast data retrieval based on various query conditions such as time, space, and task.

[0025] 102. Perform state transition conflict detection and processing on the dual-feed antenna system according to the spatio-temporal window decomposition result, determine the operating states of the feed antennas under different spatio-temporal windows, and obtain the resource allocation scheme for the feed antennas. In an embodiment of the present invention, the performing state transition conflict detection and processing on the dual-feed antenna system according to the spatio-temporal window decomposition result, determining the operating states of the feed antennas under different spatio-temporal windows, and obtaining the resource allocation scheme for the feed antennas includes: performing scenario classification processing based on the link arc segment overlap relationship on the dual-feed antenna system of the satellite according to the spatio-temporal window decomposition result, and identifying three communication scenarios of partial overlap, complete inclusion, and complete separation; analyzing the antenna state transition rules in each communication scenario, constructing a directed state graph including state nodes and transition paths, and obtaining the feed antenna state transition graph; detecting the competition conflict of multiple ground stations for the same antenna and the antenna switching time overlap conflict within the same time window according to the feed antenna state transition graph, determining the conflict type and the influence range, and obtaining the state transition conflict list; performing cumulative calculation on the physical rotation time of the antenna, the radio frequency link establishment time, and the signal synchronization time according to the state transition conflict list, determining the minimum switching time threshold in each communication scenario, and obtaining the antenna switching time constraint; performing load balancing allocation processing on the load of the dual-feed antenna system according to the antenna switching time constraint, and obtaining the resource allocation scheme for the feed antennas.

[0026] Specifically, the state transition conflict detection and processing of the dual-feed antenna system first performs an overlap relationship analysis on the link arc segments of each satellite based on the spatio-temporal window decomposition result. The system extracts the communication time windows between each satellite and different ground stations, and judges the overlap situation between any two time windows through the time interval comparison algorithm. In the specific calculation process, the system represents each time window as a time interval of the start time and the end time, and determines the overlap degree through the interval intersection operation. When there is an intersection but neither of the two time windows contains the other, the system classifies it as a partial overlap scenario. In this case, the satellite needs to serve two ground stations simultaneously during the overlapping period. The complete inclusion scenario occurs when one time window is completely inside another time window. At this time, the shorter communication task is completely covered by the longer communication task. The complete separation scenario corresponds to the situation where the two time windows do not intersect on the time axis at all. In this scenario, the satellite can serve different ground stations in sequence without concurrent processing. The system traverses all pairs of communication time windows of each satellite to establish a scenario classification matrix including scenario type identification, overlap duration, and associated ground station information.

[0027] Specifically, the dual-feed antenna system includes two independent feed antenna units, and each antenna unit can independently establish and maintain a communication link with the ground station. In partially overlapping scenarios, the system analyzes the specific mode of antenna state transition. When the satellite is about to enter the communication window of the second ground station, the idle antenna unit completes the pointing adjustment in advance and establishes a new communication link, while the original antenna unit continues to maintain the connection with the first ground station. After the overlapping period ends, the antenna unit that has completed the task releases the connection and enters the standby state. The state transition strategy in the fully contained scenario needs to balance communication efficiency and resource utilization. The system preferentially assigns communication tasks with longer durations to the main antenna unit, and shorter tasks to the auxiliary antenna unit. The processing in the fully separated scenario is relatively simple. The system sequentially allocates antenna resources in chronological order to ensure that there is enough transition time to complete antenna reorientation after the previous communication task ends. Based on these transition modes, the system constructs an antenna state transition diagram. The nodes in the diagram represent the working states of the antenna, including the idle state, single-antenna working state, dual-antenna working state, and transition state. The directed edges between states represent possible transition paths, and the weights of the edges correspond to the time overhead required for the transition.

[0028] Specifically, after the antenna state transition diagram is constructed, the system executes a conflict detection algorithm to identify resource competition and time conflict problems. The competition conflict detection is achieved by analyzing the service requests of multiple ground stations for the same antenna unit within the same time period. The system maintains a time occupancy table for each antenna unit, recording the allocated communication tasks and their time ranges. When a new communication request overlaps with the existing occupied time, the system marks it as a competition conflict and records the identification of the conflicting ground station, the conflict time period, and the severity of the conflict. The detection of antenna switching time overlap conflicts involves more complex timing analysis. The system calculates the complete time required for each antenna state transition, including the end processing time of the current task, the physical rotation time of the antenna, and the initialization time of the new task. When the time interval between two adjacent communication tasks is less than the complete transition time, the system identifies it as a switching time overlap conflict. The assessment of the conflict impact range is achieved by analyzing the conflict propagation path. A single conflict event affects subsequent scheduling arrangements through task dependencies. The system uses a graph traversal algorithm to trace the propagation range of the conflict and quantify the impact degree.

[0029] Specifically, the system then precisely calculates various time parameters based on the state transition conflict list. The calculation of the physical rotation time of the antenna is based on the mechanical characteristics of the satellite antenna system. The system calculates the required angular change based on the current antenna pointing and the target pointing, and combines the angular velocity limit of the antenna rotation to calculate the rotation time. The radio frequency link establishment time includes the cumulative duration of multiple stages such as carrier signal capture, symbol clock recovery, frame synchronization establishment, and channel equalization convergence. The system determines the standard time for each stage according to the communication protocol specifications and the complexity of the signal processing algorithm. The signal synchronization time involves clock synchronization and Doppler frequency offset compensation between the satellite and the ground station. The system calculates the Doppler frequency shift based on the satellite orbit altitude and the relative motion speed, and combines the convergence characteristics of the synchronization algorithm to determine the synchronization establishment time. During the cumulative calculation process, the system considers the change characteristics of each time component in different communication scenarios. The switching time in partially overlapping scenarios is relatively short because the antenna units can work in parallel, while the switching time in completely separated scenarios is relatively long because a complete antenna reorientation process is required. The system calculates the minimum switching time threshold for each communication scenario through the scenario weighted average method to ensure normal state transition under the worst conditions.

[0030] Specifically, the final stage involves performing load balancing allocation on the dual-feed antenna system based on the antenna switching time constraint. The system establishes an antenna load evaluation model, comprehensively considering multiple dimensions such as antenna usage frequency, cumulative working time, conversion times, and task complexity. The load balancing algorithm adopts a dynamic adjustment strategy. The system monitors the load status of the two antenna units in real time and triggers the reallocation mechanism when the load difference exceeds the preset threshold. During the reallocation process, the system preferentially transfers lightweight tasks on the high-load antenna to the low-load antenna, and the transfer operation needs to meet the time constraint and communication quality requirements. The load evaluation also considers the performance differences and aging degrees of the antenna units. The system adjusts the task allocation weights of each antenna unit according to the historical performance data to ensure the reliability and service life of the overall system. The feed antenna resource configuration plan includes a detailed task allocation table, expected load level, switching time arrangement, and backup plan settings for each antenna unit. The configuration plan also includes emergency handling strategies and resource reallocation rules in case of anomalies. The entire configuration plan is stored in a structured data format, supporting real-time query and dynamic update operations.

[0031] Further, the physical rotation time of the antenna, the radio frequency link establishment time, and the signal synchronization time are cumulatively calculated according to the state transition conflict list to determine the minimum handover time threshold for each communication scenario, and the antenna handover time constraint is obtained as follows: Analyze the delay impact of the competition conflict on the physical rotation time of the antenna and the additional overhead of the overlapping conflict on the radio frequency link establishment time according to the state transition conflict list, quantify the time penalty factors of each conflict type, and obtain the conflict time penalty coefficient; Perform a weighted cumulative operation on the physical rotation time of the antenna, the radio frequency link establishment time, and the signal synchronization time with the corresponding penalty coefficients according to the conflict time penalty coefficient to obtain the corrected handover time; Set different safety margin coefficients for the three scenarios of partial overlap, complete inclusion, and complete separation according to the handover time, and determine the minimum time interval to avoid conflicts through scenario weighted average calculation to obtain the handover time threshold for each scenario; Perform constraint condition generation processing according to the handover time threshold for each scenario, convert the time threshold into a hard constraint condition for the feed antenna operation scheduling, and limit the earliest available time and the latest release time of the antenna in different scenarios to obtain the antenna handover time constraint.

[0032] Specifically, the calculation process of the antenna handover time constraint first conducts an in-depth time impact analysis on various conflicts in the state transition conflict list. The system extracts the competition conflict events recorded in the conflict list, and determines the delay impact of the competition conflict on the physical rotation time of the antenna by analyzing the specific situation of multiple ground stations simultaneously requesting the same antenna resource. The delay caused by the competition conflict mainly stems from the need for the antenna to select and switch between multiple target directions. The system calculates the azimuth and elevation angle differences of each competing ground station. When the angle difference is large, the antenna needs to perform a larger rotation action. The quantification of the delay impact adopts a piecewise function calculation method based on the angle change amount. Small angle changes correspond to smaller delay coefficients, while large angle changes correspond to significantly increased delay coefficients. The system also analyzes the additional overhead of the overlapping conflict on the radio frequency link establishment time. When the overlapping conflict occurs, the antenna system needs to maintain the state information of multiple radio frequency links simultaneously, increasing the computational burden and memory occupancy of the signal processor. The calculation of the additional overhead is based on the number of concurrent links and the signal processing complexity. The system determines the processing overhead of each additional link according to parameters such as the modulation method, coding rate, and data transmission rate. Due to the resource limitations of the signal processor, the link establishment time increases non-linearly with the number of concurrent links. The system uses an exponential growth model to quantify this overhead relationship.

[0033] Specifically, based on the conflict impact analysis results, the system calculates the time penalty factors corresponding to each conflict type. The penalty factor calculation for competition conflicts takes into account two dimensions: conflict intensity and duration. The conflict intensity is measured by the number of ground stations competing for antenna resources simultaneously, and the duration corresponds to the time span of the conflict event. The system establishes a mapping relationship between the conflict intensity and the penalty factor. When the number of competing ground stations is 2, a basic penalty factor is set, and the penalty factor increases step by step as the number of competitors increases. The impact of duration is achieved through a time weight function. Short-duration conflicts correspond to smaller weight coefficients, while long-duration conflicts significantly increase the penalty level. The penalty factor calculation for overlapping conflicts is based on the time window overlap degree and the frequency resource occupancy rate. The overlap degree is calculated by the ratio of the overlapping time to the total communication time, and the frequency occupancy rate reflects the consumption degree of the system spectrum resources by the simultaneously operating links. The system determines the basic penalty factor for overlapping conflicts by synthesizing these two factors and adjusts it according to the specific configuration parameters of the satellite communication system. The final determination of the conflict time penalty coefficient uses a weighted average method. The system statistically analyzes the occurrence frequency of various conflicts based on historical conflict data and calculates the comprehensive penalty coefficient using the frequency as the weight.

[0034] Specifically, the calculation process of the corrected handover time involves mathematical operation combinations of basic time parameters and penalty coefficients. The correction of the antenna physical rotation time is achieved through multiplication. The system multiplies the standard rotation time by the competition conflict penalty coefficient to obtain the actual rotation time considering the conflict impact. The correction of the RF link establishment time uses addition. The system adds the basic link establishment time and the additional overhead time generated by overlapping conflicts. The correction of the signal synchronization time needs to consider the combined impact of multiple factors. The system first calculates the product of the basic synchronization time and the frequency conflict penalty coefficient, and then adds the additional synchronization time caused by the increased processor load. During the weighted accumulation operation process, the system assigns corresponding weight coefficients to different time components. The weight assignment is based on the importance and uncertainty level of each time component in the overall handover process. The antenna physical rotation time usually occupies the main part of the total handover time, so a higher weight is assigned. The weights of the RF link establishment time and the signal synchronization time are relatively lower but still need to be precisely considered. The system calculates the corrected handover time for each conflict scenario through weighted summation. This time value reflects the actual duration required to complete the antenna state conversion under conflict conditions.

[0035] Specifically, the process of determining the time threshold for scene switching requires setting differential safety margin coefficients for three different communication scenarios. The safety margin coefficient for the partially overlapping scenario is set relatively low because the dual-feed antenna system can process two communication tasks in parallel and the time pressure is relatively small. The system dynamically adjusts the safety margin according to the degree of overlap and task complexity, using a smaller safety coefficient when the overlap is low or the task is relatively simple, and appropriately increasing the safety margin when the overlap is high or the task is complex. The safety margin coefficient for the fully contained scenario needs to consider the complexity of resource scheduling. Since one task is completely covered by another task, the system needs to reserve additional time to handle task priorities and resource reallocation issues. The safety margin coefficient for the fully separated scenario is relatively high because the antenna requires a complete reorientation process, and any time estimation error may cause delays in subsequent tasks. The system uses a scene weighted average method to calculate the minimum time interval, and the weight allocation is based on the occurrence frequency and impact degree of each scenario in actual operation. Statistical analysis shows that the partially overlapping scenario has the highest occurrence frequency, so the largest weight is assigned. The fully contained scenario has a lower occurrence frequency but a greater impact degree, and the fully separated scenario is between the two. Through weighted average calculation, the system obtains the time threshold for scene switching that can cover all scenario requirements.

[0036] Specifically, the constraint condition generation process converts the calculated time threshold into a constraint rule that can be directly used by the scheduling algorithm. The system establishes a time availability matrix for each antenna unit, where the rows of the matrix correspond to time periods and the columns correspond to different working states. The expression of the constraint condition uses an inequality form, specifying that the start time of any new task must be later than the sum of the end time of the previous task and the switching time threshold. The calculation of the earliest available time is based on the expected end time of the current task and the switching time threshold of the corresponding scenario. The system determines the earliest moment when the antenna unit can accept a new task through time addition operations. The determination of the latest release time needs to consider the deadline requirements of the task and the overall scheduling goal of the system. The system calculates the latest moment when each task must be completed according to task priorities and time window constraints. The antenna switching time constraints are stored in the form of a data structure, including information such as constraint type identification, time limit value, applicable scenario, and constraint strength. The constraint strength is divided into two categories: hard constraints and soft constraints. Hard constraints correspond to physical limitations that cannot be violated, while soft constraints correspond to performance requirements that can be appropriately adjusted. The entire constraint system is implemented through hierarchical organization. Global constraints apply to all scheduling decisions, and local constraints only take effect under specific conditions. This design ensures the feasibility and efficiency of the scheduling algorithm under complex constraint conditions.

[0037] 103. Perform parallel neighborhood search processing based on reinforcement learning according to the feed antenna resource configuration scheme, generate a communication link scheduling strategy through multi-threaded collaborative optimization, and obtain a parallel optimized scheduling scheme; In one embodiment of the present invention, the parallel optimization processing guided by reinforcement learning is performed according to the feeder antenna resource configuration scheme, and a communication link scheduling strategy is generated through multi-threaded collaborative optimization. The obtained parallel optimization scheduling scheme includes: extracting communication link state information according to the feeder antenna resource configuration scheme to form a multi-dimensional state space, selecting a variety of optimization operation operators to obtain the current state and the set of executable operations; performing reinforcement learning processing based on orbit prediction according to the current state and the set of executable operations, calculating the long-term communication quality trend by predicting the position changes of the satellite in the next preset number of orbit periods to obtain a prediction result; integrating the prediction result into the reward function to perform Q-value iterative update to obtain an intelligent operation selection strategy, and performing multi-threaded parallel optimization processing on the short-period sub-problems in the spatio-temporal window decomposition result according to the intelligent operation selection strategy to obtain a multi-threaded collaborative optimization result; integrating the communication link scheduling strategies of each thread according to the multi-threaded collaborative optimization result to obtain a parallel optimization scheduling scheme.

[0038] Specifically, the parallel optimization processing guided by reinforcement learning first extracts detailed communication link state information from the feeder antenna resource configuration scheme. The system analyzes the task allocation table of each antenna unit in the configuration scheme and extracts key parameters including link establishment status, antenna pointing angle, signal strength, data transmission rate, remaining communication time, and load level. These parameters form the basis vectors of the multi-dimensional state space. Among them, the link establishment status is represented by binary coding to indicate whether the antenna is currently performing a communication task. The antenna pointing angle is represented by a pair of azimuth and elevation values to indicate the current spatial pointing of the antenna. The signal strength reflects the quality level of the current communication link. The data transmission rate describes the real-time data traffic situation. The remaining communication time represents the estimated completion time of the current task, and the load level comprehensively reflects the working intensity of the antenna unit. The system normalizes these discrete and continuous parameters to construct a unified state vector representation. The dimension number of the multi-dimensional state space is dynamically determined according to the number of satellites and ground stations. The dual-feed antenna system of each satellite corresponds to a specific sub-region in the state space. The selection of optimization operation operators is based on the actual requirements of satellite communication scheduling. The system designs eight basic operation types such as link establishment operator, link release operator, antenna switching operator, load adjustment operator, priority rearrangement operator, time window adjustment operator, frequency allocation operator, and emergency handling operator. Each operator defines specific operation objects, execution conditions, and expected effects. The link establishment operator is responsible for establishing a new communication connection between the satellite and the ground station, and the link release operator is responsible for terminating the existing communication link and releasing resources.

[0039] Specifically, the core of the reinforcement learning process lies in the decision-making mechanism based on orbit prediction. The system uses a satellite orbit dynamics model to predict the precise position changes of each satellite within a specific future orbit period. The orbit prediction calculation uses numerical integration methods to solve the satellite's motion equations, considering the combined effects of the Earth's gravitational field, atmospheric drag, solar radiation pressure, and other perturbing forces. The prediction period is set based on the satellite orbit characteristics and the time scale of communication scheduling. For low-Earth orbit satellites, 3-5 orbit periods are usually selected as the prediction window, corresponding to a time span of 6-15 hours. The position prediction results include the three-dimensional coordinates and velocity vectors of the satellite in the geocentric coordinate system. The system further calculates the changes in the geometric relationship of the satellite relative to each ground station, including distance changes, line-of-sight angle changes, and Doppler shift changes. The assessment of the long-term communication quality trend is based on signal propagation models and link budget calculations. The system calculates propagation loss factors such as free space path loss, atmospheric attenuation, and multipath effects based on the predicted geometric relationship. Communication quality metrics include parameters such as signal-to-noise ratio, bit error rate, data transmission rate, and link availability. The system identifies the change trends and periodic patterns of these metrics through time series analysis methods. The prediction results are stored in the form of a numerical vector, containing the communication quality evaluation values and uncertainty quantification metrics for each prediction time point.

[0040] Specifically, the Q-value iterative update process integrates the orbit prediction results into the design of the reinforcement learning reward function. The construction of the reward function comprehensively considers two dimensions: immediate reward and long-term reward. The immediate reward reflects the direct impact of the current operation on the system state, while the long-term reward reflects the impact of the future communication quality trend on the decision-making value. The system calculates the long-term reward component based on the predicted communication quality changes. When the prediction shows that a certain scheduling decision can maintain a high communication quality in the future period, this decision receives a positive long-term reward. The reward function also considers factors such as resource utilization efficiency, load balancing degree, and system stability, and forms a comprehensive reward signal through weighted combination. The Q-value update uses the temporal difference learning algorithm. The system maintains a Q-value table to record the value estimates of each state-action pair. During the update process, the system calculates the target Q-value based on the current state, the executed action, the obtained reward, and the transferred new state, and then uses the learning rate parameter to adjust the existing Q-value to approach the target value. The intelligent operation selection strategy is implemented based on the ε-greedy strategy. The system selects the action with the highest Q-value most of the time, while retaining a small probability of random exploration mechanism to prevent falling into local optima. The strategy also introduces a dynamic adjustment mechanism to adaptively adjust the exploration probability according to the learning progress and environmental changes.

[0041] Specifically, the multi-threaded parallel optimization process assigns short-cycle sub-problems in the spatio-temporal window decomposition results to different computing threads for independent processing. Each thread is responsible for the scheduling optimization tasks of a specific time period and satellite group, and the threads exchange status information and coordinate decisions through a shared memory area. The thread allocation strategy is based on the principle of computational load balancing, and the system dynamically adjusts the number of threads and task allocation according to the scale and complexity of the sub-problems. An independent reinforcement learning agent runs inside each thread, and the agent makes local optimization decisions according to the assigned sub-problems. The inter-thread coordination mechanism is realized through message passing and semaphore synchronization. When a certain thread discovers a high-quality scheduling scheme, the relevant information is transmitted to other threads for reference through a broadcast mechanism. During the collaborative optimization process, the system also implements a dynamic load reallocation mechanism, and when a certain thread finishes its task ahead of schedule, it automatically takes over part of the workload of other threads. The multi-threaded collaborative optimization result includes the local scheduling schemes generated by each thread, the scheme quality evaluation indicators, and the inter-thread coordination information.

[0042] Specifically, the generation of the parallel optimization scheduling scheme requires global integration of the local scheduling strategies of each thread. The integration process first checks the consistency and compatibility between the local schemes and identifies possible resource conflicts and time conflicts. The system adopts a priority-based conflict resolution strategy, and assigns priorities to conflicting schemes according to factors such as scheme quality, time urgency, and resource importance. The high-priority schemes remain unchanged, and the low-priority schemes are adjusted accordingly to eliminate conflicts. The integration algorithm also considers the global optimization effect of the schemes, and finds a scheme combination that can maximize the overall system performance through linear programming or heuristic search methods. The final parallel optimization scheduling scheme includes a complete satellite-ground station communication schedule, an antenna resource allocation scheme, a backup strategy setting, and performance expectation indicators. The scheme also includes a real-time adjustment interface to support dynamic modification and optimization according to the actual situation during execution.

[0043] 104. Perform dynamic fusion processing based on state prediction according to the parallel optimization scheduling scheme, and adjust the satellite-ground station communication link establishment timing through learning-based template matching to obtain a satellite networking communication scheduling scheme.

[0044] In one embodiment of the present invention, the dynamic fusion processing based on state prediction according to the parallel optimization scheduling scheme adjusts the timing of satellite-ground station communication link establishment through learning-based template matching, and the satellite networking communication scheduling scheme obtained includes: combining the parallel optimization scheduling scheme with the state information of the satellite obtained in real time, and using the sliding window technique to perform advance prediction calculation for a preset number of orbital periods on the satellite state change to obtain a satellite state prediction result; performing a constraint relaxation algorithm processing on the adjacent sub-problem boundaries according to the satellite state prediction result, and resolving the scheduling conflict between space-time windows by dynamically adjusting the strictness of the boundary constraints to obtain a boundary conflict resolution scheme; performing template matching processing based on historical successful patterns according to the boundary conflict resolution scheme, and selecting a verified learning-based template for rapid application to obtain a learning-based template; adjusting the timing of satellite-ground station communication link establishment according to the learning-based template, and integrating the real-time state adjustment strategy and the emergency switching plan to obtain a satellite networking communication scheduling scheme.

[0045] Specifically, the dynamic fusion processing based on state prediction first deeply combines the parallel optimization scheduling scheme with the real-time satellite state information. The system extracts the current working state of each satellite from the telemetry data stream, including key parameters such as power supply voltage, current consumption, attitude angle, angular velocity, payload working state, memory usage rate, and communication subsystem state. The power state information reflects the energy supply capacity and remaining battery level of the satellite, the attitude information describes the current spatial orientation and stability status of the satellite, and the payload state indicates the working condition and health status of the equipment carried by the satellite. The system compares and analyzes these real-time state data with the expected state in the parallel optimization scheduling scheme to identify the deviation between the actual state and the planned state. The implementation of the sliding window technique uses a fixed-length time window to continuously sample and analyze historical state data. The window length is set to the preset number of orbital periods, usually corresponding to 3-5 complete orbital periods. The sliding window moves forward at a fixed time interval. Each time it moves, the oldest data point is discarded and the latest observed data is added. The state prediction calculation identifies the trend and periodic pattern of satellite state change based on the historical data within the window. The system uses time series analysis methods to establish a mathematical model of state change. The prediction model considers the deterministic characteristics of satellite orbital motion and the influence of random disturbance factors, and captures the main patterns of state change through polynomial fitting and harmonic analysis methods.

[0046] Specifically, the satellite state prediction results include the state estimation values and corresponding uncertainty intervals at each prediction time point. During the prediction calculation process, the system pays special attention to the key state parameters that affect communication performance, such as the decay trend of power supply, the change of attitude control accuracy, and the aging degree of communication equipment. The power supply prediction is based on the historical data of the power generation efficiency of solar panels and the load power consumption, considering the periodic change of the sun illumination angle and the long-term decay of the battery capacity. The attitude prediction comprehensively analyzes the actuator performance of the attitude control system and the influence of external disturbance torques, including the effects of atmospheric drag, solar radiation pressure, and the earth's magnetic field. The communication equipment state prediction is based on the historical change trends of indicators such as the operating temperature, power output, and signal quality of the equipment. The system establishes a prediction model for the performance decay of the equipment through regression analysis methods. The prediction results are represented in the form of a probability distribution, providing uncertainty quantification information for the subsequent decision-making process. The state prediction also considers the occurrence probability of abnormal events. The system statistically analyzes the occurrence frequency and impact degree of various abnormal events based on historical failure data and includes the handling plans for abnormal situations in the prediction results.

[0047] Specifically, the generation of the boundary conflict resolution scheme is based on the dynamic adjustment mechanism of the constraint relaxation algorithm. The conflicts at the boundaries of adjacent sub-problems mainly stem from the simplifying assumptions in the space-time window decomposition process and the dynamic changes in actual operation. These conflicts are manifested as overlaps in resource requirements, conflicts in time arrangements, and contradictions in performance requirements. The constraint relaxation algorithm searches for feasible solutions by gradually relaxing the strictness of the constraint conditions. The system first identifies the specific constraint conditions that cause conflicts and then evaluates the impact of relaxing these constraints on the overall system performance. The adjustment of the constraint strictness adopts a hierarchical strategy. The system classifies the constraint conditions according to their importance and influence scope, giving priority to keeping the hard constraints related to safety and reliability unchanged and appropriately relaxing the soft constraints related to performance optimization. During the dynamic adjustment process, the system evaluates the feasibility and risk level of relaxing the constraints based on the satellite state prediction results. When the prediction shows that a certain satellite is in good condition and has a margin, the system appropriately increases the task load of this satellite to relieve the pressure on other satellites. The adjustment of the boundary constraints also considers the influence of time factors. The system dynamically adjusts the strictness of the time constraints according to the urgency and time flexibility of the tasks. Urgent tasks are given higher priorities and stricter time guarantees, while non-urgent tasks are allowed appropriate time delays to make way for urgent tasks.

[0048] Specifically, the learning-based template matching process is based on the recognition and reuse mechanism of historical successful scheduling patterns. The system maintains a template library containing historical successful scheduling cases, and each template records the scheduling decisions, execution results, and performance evaluation data in specific scenarios. The template matching process first analyzes the characteristic parameters of the current scheduling scenario, including key elements such as the number of satellites, ground station distribution, communication demand patterns, time constraints, and resource limitations. Feature extraction uses a multi-dimensional vector representation method, and the system maps the scenario features to specific points in a high-dimensional feature space. The template matching algorithm identifies the most matching scheduling pattern by calculating the similarity between the current scenario and historical templates. The similarity calculation is based on the Euclidean distance and weighted cosine similarity of feature vectors. The system also considers the impact of time context and gives higher matching weights to historical cases that are closer in time. The verified learning-based templates contain detailed descriptions of scheduling strategies, expected performance indicators, and applicable conditions. The verification process of the templates is based on the statistical analysis of historical execution results, and only scheduling patterns that meet the preset performance thresholds are included in the available template library. The application of learning-based templates uses a parameterized adjustment method, and the system adaptively adjusts the template parameters according to the differences between the current scenario and historical templates to ensure that the template strategy can adapt to the specific situation.

[0049] Specifically, the final generation of the satellite constellation communication scheduling plan requires the comprehensive integration of the scheduling strategy of the learning-based template with the real-time status adjustment mechanism and the emergency switching plan. The learning-based template provides a basic scheduling framework and decision-making logic, and the system adjusts the establishment time and release time of the satellite-ground station communication link according to the timing arrangement in the template. The timing adjustment process considers the impact of satellite state prediction results. When the prediction shows that a certain satellite is in a poor state during a specific period, the system actively transfers the communication tasks during that period to other satellites or delays their execution. The real-time status adjustment strategy establishes the mapping relationship between status monitoring triggers and corresponding adjustment actions, and automatically triggers corresponding adjustment measures when the satellite status deviates from the expected range. The adjustment measures include various means such as load redistribution, task priority adjustment, activation of backup resources, and optimization of communication parameters. The emergency switching plan designs a rapid response mechanism for various emergencies, including abnormal scenarios such as satellite failures, ground station outages, bad weather, and cyber attacks. Each emergency plan contains detailed fault detection methods, impact assessment procedures, resource redistribution strategies, and recovery operation processes. The final satellite constellation communication scheduling plan is organized in a structured data format, including a complete schedule, resource allocation table, performance expectation indicators, monitoring trigger conditions, and emergency handling processes. The plan also provides a real-time update interface to support dynamic adjustment and optimization operations.

[0050] Further, adjusting the satellite-ground station communication link establishment timing according to the learning template, integrating the real-time status adjustment strategy and the emergency switching plan, the satellite networking communication scheduling plan obtained includes: mapping the successful scheduling mode in the learning template to the current satellite orbit configuration, adjusting the connection time points of the satellite-ground station through time offset and priority reordering to obtain the optimized link timing; matching the preset dynamic adjustment rules according to the optimized link timing and satellite status information to obtain the real-time status adjustment strategy; preparing a standby communication link and switching timing for emergencies according to the real-time status adjustment strategy, formulating a multi-level emergency response mechanism and resource reallocation plan to obtain the emergency switching plan; performing plan integration processing according to the optimized link timing, real-time status adjustment strategy and emergency switching plan to obtain the satellite networking communication scheduling plan.

[0051] Specifically, in the process of mapping the scheduling mode of the learning template, the geometric relationship differences between the current satellite orbit configuration and historical successful cases are first analyzed. The system extracts the orbital elements and spatial position information of each satellite in the current constellation and conducts a comparative analysis with the historical orbit configurations recorded in the learning template. The mapping algorithm uses the coordinate transformation method to transform the satellite position relationship in the historical case into the current orbit reference system. The transformation process takes into account the spatial orientation differences of the orbital planes, the changes in satellite phase distribution, and the adjustment of ground station coverage. The system realizes the effective migration of the scheduling strategy by establishing the correspondence relationship between satellite identifiers, and the correspondence relationship is determined based on the orbital feature similarity and functional equivalence of the satellites. When a certain satellite in the historical case has similar orbital altitude, inclination, and payload configuration to a specific satellite in the current constellation, the system establishes a mapping relationship between the two. The time offset adjustment is calculated based on the orbital period difference and phase offset. The system calculates the time offset according to the orbital phase difference between the current satellite and the corresponding satellite in the historical case to ensure that the scheduling timing can adapt to the current orbit configuration. The priority reordering process comprehensively considers multiple factors such as task importance, time urgency, and resource availability. The system re-evaluates and adjusts the task priorities in the historical template according to the current communication requirements and constraints. The reordering algorithm uses the multi-objective optimization method to balance the requirements of different objectives such as communication efficiency, resource utilization rate, and system stability.

[0052] Specifically, the adjustment of satellite-ground station connection time points involves fine-tuning the timing in the historical success patterns. The system analyzes the start time, end time, and duration of each communication event in historical cases, and calculates the optimal adjusted connection time in combination with the visibility window under the current orbital configuration. The adjustment process takes into account the changes in the geometric relationship between the satellite and the ground station, including key time points such as the moment of maximum elevation angle, the moment of signal strength peak, and the moment of minimum Doppler shift. The system uses an optimization algorithm to find the connection time combination that can maximize communication quality and efficiency, and the optimization objectives include indicators such as signal strength, communication duration, data transmission rate, and link stability. The adjustment of time points also needs to meet various constraints, including antenna switching time limit, power budget constraint, frequency resource constraint, and ground station processing capacity constraint. The optimized link timing is organized in the form of a schedule, recording the precise timing of each communication event, the participating satellites and ground stations, the expected communication quality, and resource consumption and other information. The schedule also contains time tolerance information, reflecting the adjustment flexibility of each time point and the tolerance to time deviation.

[0053] Specifically, the formulation of the real-time state adjustment strategy is based on the matching analysis of the optimized link timing and the current satellite state information. The system has established a rule base containing a variety of dynamic adjustment rules, and each rule defines the adjustment actions and execution logic under specific state conditions. The matching process of dynamic adjustment rules uses pattern recognition methods. The system compares the current satellite state vector with the state patterns in the rule base to identify the adjustment rule that best suits the current situation. The rule matching considers the numerical range, change trend, and combined characteristics of state parameters. When the power level of the satellite is lower than the preset threshold, the power consumption optimization rule is triggered. When the attitude control accuracy decreases, the stability enhancement rule is activated. When the temperature of the communication device is too high, the heat dissipation protection rule is started. The real-time state adjustment strategy consists of components such as monitoring triggers, execution conditions, adjustment actions, and recovery mechanisms. The monitoring trigger defines the state parameters that need to be continuously monitored and the corresponding threshold settings. The execution condition describes the specific condition combinations required to trigger the adjustment action. The adjustment actions cover various operation types such as load reduction, task delay, resource reallocation, and parameter optimization, and each operation has clear execution steps and expected effects. The recovery mechanism ensures that the system can return to the normal working state after the adjustment action is executed, including processes such as state monitoring, effect evaluation, and gradual recovery.

[0054] Specifically, the formulation of the emergency switching plan designs a hierarchical response mechanism for various emergencies. The system classifies emergency situations into four levels: minor anomalies, general failures, serious accidents, and catastrophic failures according to the severity and impact scope of the emergencies. Each level corresponds to different response intensities and handling strategies. Minor anomalies mainly include situations such as slight decline in equipment performance, signal quality fluctuations, and temporary overload of the load. The handling strategy mainly focuses on parameter adjustment and load optimization. General failures involve problems such as single device failure, communication interruption, and attitude deviation, and it is necessary to activate backup resources and reallocate tasks. Serious accidents include situations such as simultaneous failure of multiple key devices, satellite out-of-control, and long-term interruption of the ground station, which require large-scale resource reorganization and task rearrangement. Catastrophic failures correspond to extreme situations such as complete satellite failure, serious damage to the ground station, and large-scale network paralysis, and it is necessary to activate the highest-level emergency response mechanism. The preset of the backup communication link is based on network topology analysis and reliability assessment. The system identifies key communication paths and configures multiple backup links for each path. The selection of the backup link takes into account principles such as geographical dispersion, device independence, and performance equivalence to ensure that communication services can be quickly switched to the backup link when the main link fails. The determination of the switching timing is based on the fault detection algorithm and the results of predictive analysis. The system anticipates possible fault moments through real-time monitoring and trend analysis, and prepares the switching operation in advance to reduce the service interruption time.

[0055] Specifically, the final integration and processing of the satellite networking communication scheduling plan organizes the optimized link timing, real-time state adjustment strategy, and emergency switching plan into a complete execution plan. The integration process first checks the consistency and compatibility between the various components to ensure that the timing arrangement, adjustment strategy, and emergency plan are logically coordinated and unified. The consistency check includes content such as time conflict detection, resource requirement verification, and constraint condition matching. The system identifies and solves potential conflict problems through the constraint solving algorithm. The plan integration also establishes the interface relationship and data flow between the various components to ensure that different strategies can be effectively coordinated and switched during the execution process. The integrated scheduling plan is organized in a hierarchical architecture, with the top layer being the overall scheduling plan, the middle layer being the specific execution strategy, and the bottom layer being the detailed operation instructions. The plan also includes a complete monitoring system and feedback mechanism to support real-time monitoring, effect evaluation, and dynamic adjustment during the execution process. The monitoring system covers all key performance indicators and state parameters, and the feedback mechanism ensures that abnormal situations can be detected and processed in a timely manner.

[0056] In this embodiment, by performing space-time window decomposition processing based on orbital phase synchronization on the satellite constellation orbit parameters, the long-term scheduling problem is decomposed into multiple short-term sub-problems; according to the space-time window decomposition result, conflict detection processing for the state transition of the dual-feed antenna system is performed to determine the operating state of the feed antenna; according to the feed antenna resource allocation scheme, parallel neighborhood search processing based on reinforcement learning is performed, and a communication link scheduling strategy is generated through multi-threaded collaborative optimization; according to the parallel optimization scheduling scheme, dynamic fusion processing based on state prediction is performed, and the timing for establishing the satellite-ground station communication link is adjusted through learning-based template matching. Through space-time decomposition, intelligent conflict detection, reinforcement learning optimization, and dynamic fusion, the present invention solves the complexity and real-time problems of large-scale satellite network communication scheduling, and significantly improves the scheduling efficiency and communication quality.

[0057] The satellite network communication method in the embodiment of the present invention is described above. Next, the satellite network communication device in the embodiment of the present invention will be described. The satellite network communication device is shown in Figure 2 , and an embodiment of the satellite network communication device in the embodiment of the present invention includes: A space-time decomposition module 201, configured to perform space-time window decomposition processing based on orbital phase synchronization on the satellite constellation orbit parameters, decompose the long-term scheduling problem of satellite-ground station communication into multiple short-term sub-problems, and obtain a space-time window decomposition result; A conflict detection module 202, configured to perform state transition conflict detection processing on the dual-feed antenna system according to the space-time window decomposition result, determine the operating state of the feed antenna in different space-time windows, and obtain a feed antenna resource allocation scheme; An intelligent optimization module 203, configured to perform parallel neighborhood search processing based on reinforcement learning according to the feed antenna resource allocation scheme, generate a communication link scheduling strategy through multi-threaded collaborative optimization, and obtain a parallel optimization scheduling scheme; A dynamic fusion module 204, configured to perform dynamic fusion processing based on state prediction according to the parallel optimization scheduling scheme, adjust the timing for establishing the satellite-ground station communication link through learning-based template matching, and obtain a satellite network communication scheduling scheme.

[0058] In the embodiments of the present invention, the satellite networking communication device runs the above satellite networking communication method. The satellite networking communication device performs space-time window decomposition processing based on orbital phase synchronization on the satellite constellation orbit parameters, decomposes the long-period scheduling problem into multiple short-period sub-problems; performs state transition conflict detection processing on the dual-feed antenna system according to the space-time window decomposition result to determine the working state of the feed antenna; performs parallel neighborhood search processing based on reinforcement learning according to the feed antenna resource allocation scheme, and generates a communication link scheduling strategy through multi-threaded collaborative optimization; performs dynamic fusion processing based on state prediction according to the parallel optimization scheduling scheme, and adjusts the timing of establishing the satellite-ground station communication link through learning-based template matching. The present invention solves the complexity and real-time problems of large-scale satellite networking communication scheduling through space-time decomposition, intelligent conflict detection, reinforcement learning optimization, and dynamic fusion, and significantly improves the scheduling efficiency and communication quality.

[0059] Above Figure 2 The satellite networking communication device in the embodiments of the present invention is described in detail from the perspective of modular functional entities. Next, the satellite networking communication device in the embodiments of the present invention is described in detail from the perspective of hardware processing.

[0060] Figure 3 FIG. is a schematic structural diagram of a satellite networking communication device provided by an embodiment of the present invention. The satellite networking communication device 300 may vary greatly due to configuration or performance differences, and may include one or more processors (central processing units, CPU) 310 (for example, one or more processors) and a memory 320, and one or more storage media 330 for storing application programs 333 or data 332 (for example, one or more mass storage device terminals). Among them, the memory 320 and the storage media 330 may be transient storage or persistent storage. The program stored in the storage media 330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations on the satellite networking communication device 300. Further, the processor 310 may be set to communicate with the storage media 330 and execute a series of instruction operations in the storage media 330 on the satellite networking communication device 300 to implement the steps of the above satellite networking communication method.

[0061] The satellite networking communication device 300 may further include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating systems 331, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, and so on. Those skilled in the art can understand, Figure 3The structure of the satellite networking communication device shown does not constitute a limitation on the satellite networking communication device provided by the present invention, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0062] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described system, device, or unit can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0063] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0064] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A satellite networking communication method, characterized in that, The satellite networking communication method includes: Performing a spatio-temporal window decomposition process on the satellite constellation orbit parameters based on orbital phase synchronization, decomposing the long-period scheduling problem of satellite-ground station communication into multiple short-period sub-problems, and obtaining the spatio-temporal window decomposition result; Performing a state transition conflict detection process on the dual-feed antenna system according to the spatio-temporal window decomposition result, determining the working states of the feed antennas in different spatio-temporal windows, and obtaining the feed antenna resource allocation scheme; Performing a parallel neighborhood search process based on reinforcement learning according to the feed antenna resource allocation scheme, generating a communication link scheduling strategy through multi-threaded collaborative optimization, and obtaining the parallel optimization scheduling scheme; Performing a dynamic fusion process based on state prediction according to the parallel optimization scheduling scheme, adjusting the satellite-ground station communication link establishment timing through learning-based template matching, and obtaining the satellite networking communication scheduling scheme.

2. The satellite networking communication method according to claim 1, wherein The performing a spatio-temporal window decomposition process on the satellite constellation orbit parameters based on orbital phase synchronization, decomposing the long-period scheduling problem of satellite-ground station communication into multiple short-period sub-problems, and obtaining the spatio-temporal window decomposition result includes: Calculating the visible arc segment of each satellite relative to the ground station according to the satellite constellation orbit parameters, and determining the effective communication time boundary according to the start time and end time of the visible arc segment; Generating a visible time window between the satellite and the ground station based on the effective communication time boundary, and performing a similarity analysis process on the satellite constellation orbit parameters according to the visible time window to identify the orbital phase synchronization window; Grouping and classifying the satellites according to the orbital phase synchronization window, obtaining multiple satellite groups, and performing a time-domain segmentation process based on orbital periodicity on the long-period scheduling problems of the multiple satellite groups to obtain multiple short-period sub-problems; Associating and organizing the visible time window, the orbital phase synchronization window, and the short-period sub-problems to obtain a spatio-temporal window decomposition result including the time dimension, the space dimension, and the task dimension.

3. The satellite networking communication method according to claim 1, wherein The performing a state transition conflict detection process on the dual-feed antenna system according to the spatio-temporal window decomposition result, determining the working states of the feed antennas in different spatio-temporal windows, and obtaining the feed antenna resource allocation scheme includes: Performing a scenario classification process on the dual-feed antenna system of the satellite based on the link arc segment overlap relationship according to the spatio-temporal window decomposition result, and identifying three communication scenarios: partial overlap, complete inclusion, and complete separation; Analyzing the antenna state transition rules in each communication scenario, constructing a directed state graph including state nodes and transition paths, and obtaining the feed antenna state transition graph; Detecting the competition conflict of multiple ground stations for the same antenna and the antenna switching time overlap conflict within the same time window according to the feed antenna state transition graph, determining the conflict type and the influence range, and obtaining the state transition conflict list; Accumulatively calculating the antenna physical rotation time, the RF link establishment time, and the signal synchronization time according to the state transition conflict list, determining the minimum switching time threshold in each communication scenario, and obtaining the antenna switching time constraint; Performing an equal load distribution process on the load of the dual-feed antenna system according to the antenna switching time constraint, and obtaining the feed antenna resource allocation scheme.

4. The satellite networking communication method according to claim 3, wherein Accumulatively calculate the physical rotation time of the antenna, the radio frequency link establishment time, and the signal synchronization time according to the state transition conflict list, determine the minimum handover time threshold under each communication scenario, and obtain the antenna handover time constraint, including: Analyze the delay effect of the competition conflict on the physical rotation time of the antenna and the additional overhead of the overlapping conflict on the radio frequency link establishment time according to the state transition conflict list, quantify the time penalty factors of each conflict type, and obtain the conflict time penalty coefficient; Perform a weighted accumulation operation on the physical rotation time of the antenna, the radio frequency link establishment time, and the signal synchronization time with the corresponding penalty coefficients according to the conflict time penalty coefficient to obtain the corrected handover time; Set different safety margin coefficients for the three scenarios of partial overlap, complete inclusion, and complete separation according to the handover time, and determine the minimum time interval to avoid conflicts through scenario weighted average calculation to obtain the sub-scenario handover time threshold; Perform constraint condition generation processing according to the sub-scenario handover time threshold, convert the time threshold into a hard constraint condition for the feeder antenna operation scheduling, and limit the earliest available time and the latest release time of the antenna under different scenarios to obtain the antenna handover time constraint.

5. The satellite networking communication method according to claim 1, wherein, Perform parallel optimization processing guided by reinforcement learning according to the feeder antenna resource configuration plan, and generate a communication link scheduling strategy through multi-threaded collaborative optimization. The obtained parallel optimization scheduling plan includes: Extract the communication link state information according to the feeder antenna resource configuration plan to form a multi-dimensional state space, and select a variety of optimization operation operators to obtain the current state and the set of executable operations; Perform reinforcement learning processing based on orbit prediction according to the current state and the set of executable operations, calculate the long-term communication quality trend by predicting the position change of the satellite in the next preset number of orbital periods, and obtain the prediction result; Integrate the prediction result into the reward function for Q-value iterative update to obtain the intelligent operation selection strategy, and perform multi-threaded parallel optimization processing on the short-period sub-problems in the spatio-temporal window decomposition result according to the intelligent operation selection strategy to obtain the multi-threaded collaborative optimization result; Integrate the communication link scheduling strategies of each thread according to the multi-threaded collaborative optimization result to obtain the parallel optimization scheduling plan.

6. The satellite networking communication method according to claim 1, wherein Perform dynamic fusion processing based on state prediction according to the parallel optimization scheduling plan, and adjust the satellite-ground station communication link establishment timing through learning-based template matching to obtain the satellite networking communication scheduling plan, including: Combine the parallel optimization scheduling plan with the real-time obtained satellite state information, and use the sliding window technology to perform advance prediction calculation on the satellite state change for the next preset number of orbital periods to obtain the satellite state prediction result; Perform a constraint relaxation algorithm processing on the adjacent sub-problem boundaries according to the satellite state prediction result, and solve the scheduling conflict between spatio-temporal windows by dynamically adjusting the strictness of the boundary constraints to obtain the boundary conflict resolution plan; Perform template matching processing based on historical successful patterns according to the boundary conflict resolution plan, and select a verified learning-based template for rapid application to obtain the learning-based template; Adjust the satellite-ground station communication link establishment timing according to the learning template, integrate the real-time status adjustment strategy and the emergency switching plan, and obtain the satellite networking communication scheduling plan.

7. The satellite networking communication method according to claim 6, wherein The adjusting the satellite-ground station communication link establishment timing according to the learning template, integrating the real-time status adjustment strategy and the emergency switching plan, and obtaining the satellite networking communication scheduling plan includes: Map the successful scheduling mode in the learning template to the current satellite orbit configuration, and adjust the connection time points of the satellite-ground station through time offset and priority reordering to obtain the optimized link timing. Match the preset dynamic adjustment rules according to the optimized link timing and satellite status information to obtain the real-time status adjustment strategy. Prepare the standby communication link and switching timing for emergencies according to the real-time status adjustment strategy, formulate a multi-level emergency response mechanism and resource reallocation plan to obtain the emergency switching plan. Perform scheme integration processing according to the optimized link timing, real-time status adjustment strategy and emergency switching plan to obtain the satellite networking communication scheduling plan.

8. A satellite networking communication device, characterized in that, The satellite networking communication device includes: A space-time decomposition module, configured to perform space-time window decomposition processing based on orbital phase synchronization on the satellite constellation orbit parameters, decompose the long-period scheduling problem of satellite-ground station communication into multiple short-period sub-problems, and obtain the space-time window decomposition result. A conflict detection module, configured to perform state transition conflict detection processing on the dual-feed antenna system according to the space-time window decomposition result, determine the working state of the feed antenna in different space-time windows, and obtain the feed antenna resource allocation plan. An intelligent optimization module, configured to perform parallel neighborhood search processing based on reinforcement learning according to the feed antenna resource allocation plan, and generate a communication link scheduling strategy through multi-threaded collaborative optimization to obtain a parallel optimization scheduling plan. A dynamic fusion module, configured to perform dynamic fusion processing based on state prediction according to the parallel optimization scheduling plan, and adjust the satellite-ground station communication link establishment timing through learning template matching to obtain the satellite networking communication scheduling plan.

9. A satellite networking communication device, characterized in that, The satellite networking communication device includes: a memory and at least one processor, and instructions are stored in the memory. The at least one processor calls the instructions in the memory so that the satellite networking communication device executes the steps of the satellite networking communication method according to any one of claims 1-7.

Images