Implementation method suitable for high-throughput satellite communication hybrid duplex system

Abstract

The application discloses an implementation method suitable for a high-throughput satellite communication hybrid duplex system, which acquires double-network statistical basic data containing end station round-trip propagation delay distribution and phased array antenna transmission-reception switching characteristics; constructs a dual-coupling time planning mathematical model with original network and mirror network time window start and end time as optimization variables, and solves optimal time parameters under the conditions of satisfying physical constraints such as delay reachability, antenna switching safety and main station reception zero overlap; and the main station broadcasts time slot configuration instructions to cooperatively control all network end stations to perform hybrid duplex transmission-reception. The application also introduces a closed-loop feedback mechanism based on end station local observation, estimates delay drift by using network clock reference information, and corrects model input. The application realizes zero conflict hybrid duplex communication under complex physical constraints by converting static configuration into dynamic mathematical optimization solution, and improves satellite spectrum resource utilization rate and system operation stability.

Description

Technical Field

[0001] This invention relates to the field of satellite communication technology, and in particular to a method for implementing a high-throughput satellite communication hybrid duplex system. Background Technology

[0002] High-throughput satellite (HTS) communication systems have become an important part of information infrastructure due to their wide coverage and high throughput. Among them, hybrid duplex mode, as a technology that can utilize spectrum resources and improve system capacity, enables satellite terminals to achieve efficient two-way data transmission with limited frequency resources by planning transmit and receive time slots on the time axis, thereby improving the overall performance of satellite communication systems.

[0003] Existing technologies for satellite communication time slot planning employ fixed frame structure configurations or simple request-based on-demand allocation (DAMA) strategies. The system pre-sets fixed forward and reverse window ratios and sets uniform guard intervals based on experience, determining the transmission and reception timing of terminal stations through table lookups or manual static configuration. This approach assumes a relatively stable network environment and that the hardware characteristics of terminal equipment (such as transmit / receive switching times) are consistent or highly redundant, maintaining basic link connectivity through simplified control logic.

[0004] However, existing fixed-configuration solutions struggle to balance diverse physical link constraints with dynamically changing business needs, resulting in both low resource utilization and frequent physical conflicts. Therefore, further research and innovation are needed to address these issues with existing technologies. Summary of the Invention

[0005] Purpose of the invention: In view of the above-mentioned problems in the prior art, this application provides an implementation method suitable for high-throughput satellite communication hybrid duplex systems.

[0006] Technical solution: According to one aspect of this application, a method for implementing a high-throughput satellite communication hybrid duplex system includes:

[0007] Based on the round-trip propagation delay distribution characteristics of the terminal station group and the switching time characteristics of the phased array antenna, a dual coupled time programming model is constructed with the start and end times of the time windows of the original network and the mirror network as optimization variables.

[0008] Solve the above model under the physical constraints of zero overlap between transmit and receive states and time delay reachability to generate dual-network time planning parameter data that defines the dual-network time sequence structure;

[0009] The master station generates time slot configuration instructions based on the dual-network time planning parameter data and broadcasts them, coordinating with all network endpoints to perform mixed-duplex data transmission and reception between the original network and the mirror network.

[0010] Beneficial effects: This invention transforms static configuration into dynamic mathematical optimization solutions, achieving zero-collision hybrid full-duplex communication under complex physical constraints, thus improving satellite spectrum resource utilization and system operational stability. The related technical effects will be described in detail below with reference to specific embodiments. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating an implementation method for a high-throughput satellite communication hybrid duplex system, provided as an embodiment of this application.

[0012] Figure 2 The flowchart provided in this application illustrates the construction of a dual-coupled time programming model with the start and end times of the time windows of the original network and the mirror network as optimization variables.

[0013] Figure 3 This is a flowchart illustrating the process of establishing latency reachability constraints using the round-trip propagation delay distribution characteristics of a cluster of terminal stations, as provided in an embodiment of this application.

[0014] Figure 4 This is a flowchart illustrating the process of establishing antenna switching safety constraints using the transmit / receive switching time characteristics of a phased array antenna, as provided in an embodiment of this application.

[0015] Figure 5 This is a flowchart for solving and verifying a dual-coupled time programming model, provided in an embodiment of this application. Detailed Implementation

[0016] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0017] It should be noted that the terms "first," "second," etc., in the specification and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0018] To address the aforementioned issues, the applicant conducted in-depth searches and analyses, and discovered:

[0019] Existing technologies have limitations in adapting to and matching the round-trip delay distribution of terminal stations with significant geographical differences, as well as the differences in hardware switching time for different types of phased array antennas.

[0020] Correspondingly, the fixed timing structure forces the system to reserve an excessively large global protection band to cover the worst-case scenario (such as the farthest station or the slowest switching antenna), resulting in idle and wasted spectrum resources. Next, in scenarios where phased array antennas require microsecond-level rapid switching, the lack of planning based on physical hard constraints can easily lead to overlap of transmit and receive states on the time axis, causing hardware conflicts or signal collisions when the transmitter and receiver are working simultaneously. On this basis, the lack of a real-time perception and feedback mechanism for satellite orbit drift and clock deviation makes it difficult for the system to maintain microsecond-level time slot alignment accuracy during long-term operation, which can easily lead to time slot misalignment.

[0021] To solve these problems, combined with Figures 1 to 5 The present invention will be specifically described through the following embodiments.

[0022] In some embodiments, an exemplary implementation method for master station time slot planning is provided, as follows:

[0023] A typical master station consists of a power supply board, a clock board, a switching board, and a channel board. The clock board connects to the GPS / BeiDou antenna to acquire time and synchronizes the time with the channel board via 1PPS and TOD signals. When allocating carrier resources in the network management system, the time from the clock board is used as a reference when issuing time slot plans.

[0024] Furthermore, the time slot plan is mainly divided into two parts: the mirror network and the original network. The proportion of these two network types can be flexibly adjusted to adapt to different terminal service requirements. For example, with a fixed time slice, when the mirror network has a larger number of forward time slots and the original network has a smaller number, the mirror network has stronger downlink service capabilities than the original network, but its uplink service capabilities are weaker. The protocol processor distributes the time slot plan to the FPGA via AXI-BRAM. It distributes the plan once when carrier resource allocation is complete and updates it when carrier resources change. The parameters included in the master station's time slot plan may include: time slot plan start reference time; single forward time slot time; number of forward time slots in the mirror network; mirror-to-original network forward frame interval; number of forward time slots in the original network; forward superframe end interval; original network reverse window time; original-to-mirror network reverse frame interval; mirror network reverse window length; reverse superframe end interval; and time slot plan update flag.

[0025] The initial value of the time slot plan update flag is 0. When the time slot plan parameters change and all the above parameters have been issued, the time slot plan update flag is modified to 1 for time slot plan update detection. The parameters can also be used for forward time slot transmission time calculation, which is mainly implemented on the FPGA platform. The calculation process can be carried out using the following steps:

[0026] Step S1.1: The time slot plan start reference time is used as the transmission time of the first time slot of the mirror network;

[0027] Step S1.2: When the mirror network timeslot transmission time is equal to the master station's local time (the time restored by the master station according to the clock board), a mirror network timeslot transmission flag is generated, and the mirror network timeslot count is incremented by 1.

[0028] Step S1.3: Determine whether the time slot count of the mirror network is equal to the number of forward time slots of the mirror network. If they are equal, clear the time slot count of the mirror network to zero and execute step S1.5; if they are not equal, execute step S1.4.

[0029] Step S1.4: Add the transmission time of the previous time slot of the mirror network to the transmission time of a single forward time slot to obtain the transmission time of the next time slot, and then execute step S1.2;

[0030] Step S1.5: After calculating the transmission time of the last time slot of the mirror network, the time of a single forward time slot and the forward frame interval between the mirror and original networks need to be added. The resulting time is used as the transmission time of the first time slot of the original network.

[0031] Step S1.6: When the original network timeslot transmission time is equal to the local time (the time recovered by the master station according to the clock board), an original network timeslot transmission flag is generated, and the original network timeslot count is incremented by 1.

[0032] Step S1.7: Determine whether the original network timeslot count is equal to the original network forward timeslot count. If they are equal, clear the original network timeslot count to zero and proceed to step S1.9; if they are not equal, proceed to step S1.8.

[0033] Step S1.8: Add the transmission time of the previous time slot of the original network to the transmission time of a single forward time slot to obtain the transmission time of the next time slot, and then execute step S1.6;

[0034] Step S1.9: After calculating the transmission time of the last time slot of the original network, the time of a single forward time slot and the forward frame interval of the original-mirror network need to be added. The resulting time is used as the transmission time of the first time slot of the mirror network. Then, proceed to step S1.2.

[0035] Example 1 provides a method for implementing a high-throughput satellite communication hybrid duplex system, specifically how to achieve zero-collision coexistence and maximized resource utilization of the original network and the mirror network on the time axis under constraints such as long latency of satellite communication links and physical time consumption during switching of phased array antennas at end stations. The specific steps include:

[0036] Step 101: Based on the round-trip propagation delay distribution characteristics of the terminal station group and the phased array antenna transmit / receive switching time characteristics, construct a dual coupled time programming model with the start and end times of the time windows of the original network and the mirror network as optimization variables.

[0037] In this step, a high-throughput satellite communication hybrid duplex system typically involves one or more high-throughput satellites (HTS) with multi-beam coverage capabilities, a ground control center (master station), and multiple user terminals (end stations) distributed across different geographical locations. An end station cluster refers to the set of all active terminals within the same beam coverage or control domain. The original network and the mirror network are two logical networks defined on the same physical channel, or two communication planes that are orthogonal in frequency (e.g., left-hand circular polarization and right-hand circular polarization) but temporally coupled. In hybrid duplex mode, an end station may need to receive forward data in the original network and then transmit reverse data in the mirror network, and vice versa.

[0038] Furthermore, constructing a dual-coupled time programming model means that the master station no longer uses a pre-fixed time slot table, such as a static TDMA frame structure, but instead establishes a mathematical optimization problem based on the physical state statistics of all end stations in the current network.

[0039] Furthermore, the round-trip propagation delay distribution characteristics reflect the signal transmission time differences caused by the different geographical distribution of the terminal station clusters. For example, the delay between near-ground terminal stations and coverage edge terminal stations may differ by tens of milliseconds. The phased array antenna transmit / receive switching time characteristics reflect the physical settling time required for the terminal station hardware to switch from receiving to transmitting (or vice versa). The master station inputs these physical characteristics as hard constraints into the model to find the optimal time parameters so that the interleaving of the two networks on the time axis does not cause physical transmit / receive conflicts at the terminal stations.

[0040] As an example, the build process can be triggered periodically, such as once every preset superframe period (e.g., every second or every minute), to adapt to changes in the terminal topology or fluctuations in service load. Furthermore, the model can be deployed in the high-performance computing unit of the main station, dynamically built using real-time collected link status data.

[0041] Step 102: Solve the above model (dual coupled time planning model) under the physical constraints of zero overlap between transmit and receive states and time delay reachability, and generate dual network time planning parameter data that defines the dual network time sequence structure.

[0042] In this step, zero overlap of transmit and receive states means that for any given end station, it cannot be in both receive and transmit states at the same time (unless it has full-duplex hardware capabilities), and sufficient hardware protection time must be maintained between two state transitions. Delay reachability means that after the forward control command or data sent by the master station arrives at the end station, the end station must have sufficient time to process it and wait for a valid reverse transmission window to open, and there cannot be a timing paradox where the transmission window has already expired when the command is received.

[0043] Specifically, solving the model involves using mathematical optimization algorithms, such as linear programming solvers or heuristic search algorithms, to find the optimal solution in a multidimensional variable space. The dual-network time programming parameter data is the output of the solution, specifically including time values ​​such as the superframe period length, the start and end times of the original network's forward window, the start and end times of the original network's backward window, and the four time parameters corresponding to the mirror network. This set of parameters outlines the skeletal structure of the dual networks over a future period.

[0044] Step 103: The master station generates a time slot configuration command based on the dual network time planning parameter data and broadcasts it, and coordinates the control of all network terminals to perform mixed duplex data transmission and reception between the original network and the mirror network.

[0045] In this step, the time slot configuration instruction converts the abstract time parameter data mentioned above into control signaling that can be parsed and executed by the end station. For example, the instruction may include a frame structure descriptor, explicitly indicating the start and end points of the next superframe and the division of each service time slot.

[0046] Accordingly, the master station sends the instruction to all end stations via a forward broadcast channel (NCR channel or dedicated control channel). Each end station receives and parses the instruction, and then, combining this with its local clock and delay compensation value, derives the precise action schedule on its local timeline. For example, based on the instruction, end station A determines that at local time t... _1 To t _2 During this period, the forward data received from the original network should be locked, at t _3 To t _4 During this period, antenna settings should be adjusted to prepare for sending reverse data in the mirror network. Through a collaborative mechanism of centralized planning at the master station and distributed execution at the end stations, zero-collision hybrid full-duplex communication was achieved across the entire network.

[0047] Example 2: Optional Implementation of the Physical Construction Process for Providing Dual-Network Statistical Foundation Data

[0048] This paper describes how to construct a dual-network statistical foundation for model input through physical sensing and data processing. This lays the data foundation for model building. Specifically, the method includes:

[0049] Step 201: Using the satellite orbital parameters in the system structure data and the geographical coordinates of the terminal stations in the terminal station dynamic status data, calculate the estimated round-trip propagation time delay of each terminal station based on the geometric propagation model and the atmospheric refraction correction model. Cluster the estimated round-trip propagation time delay according to the preset time delay interval rules to generate terminal station time delay group statistics.

[0050] In this step, the system structure data is static configuration information that the master station stores in advance or updates in real time. This data can be represented as a structure or dataset, and includes, but is not limited to, the following fields: {satellite ephemeris parameters, master station geographic coordinates, transponder processing delay, atmospheric refraction correction coefficient}.

[0051] Furthermore, the dynamic status data of the terminal station is the dynamic information reported by the terminal station through login signaling or heartbeat packets, specifically including: {unique identifier (ID) of the terminal station, GPS coordinates of the terminal station, and the moving speed vector of the terminal station}.

[0052] Specifically, the round-trip propagation delay estimate T for any terminal station i is calculated. _rtt (i) The geometric distance accumulation method is usually used. Its calculation formula can be expressed in linear text as follows:

[0053] T _rtt (i)=[Distance(GW,Sat)+Distance(Sat,Term _i )]×2 / c _light +T _atm_correction +T _hw_proc ;

[0054] Where Distance(A, B) represents the Euclidean distance between points A and B, c _light For the speed of light, T _atm_correction To add time delay to the troposphere and ionosphere calculated based on atmospheric models, T _hw_proc This provides a fixed latency for the system hardware's baseband processing. In other words, GW corresponds to the main gateway, Sat corresponds to the satellite, and Term... _i Corresponding terminal station i.

[0055] Furthermore, due to the large number of terminal stations, this directly involves tens of thousands of T... _rtt (i) The input model leads to excessive computational complexity, thus requiring clustering or grouping. A time delay step (e.g., 5ms or 10ms) can be set to group all T... _rtt(i) Mapping to the corresponding intervals. For example, terminals with latency between 240ms and 250ms are grouped into group A, and those between 250ms and 260ms are grouped into group B. For each group k, its maximum and minimum latency values ​​are calculated. The generated terminal latency group statistics are a set of group statistics, which provides a dimensionality-reduced description of the latency characteristics of all terminals in the network.

[0056] Step 202: Extract historical handover operation records of each terminal from the terminal dynamic status data, count the transmit / receive handover duration of different terminal types during actual power-on operation, construct a probability distribution model in combination with hardware nominal parameters, and generate antenna transmit / receive handover time distribution data.

[0057] In this step, the historical handover operation record is the log data of the antenna handover actions actually performed by the terminal station in past operating cycles, including: {handover type (RX-to-TX or TX-to-RX), command issuance time, hardware lock completion time, and handover success identifier}. Terminal type refers to the phased array antenna model equipped on the terminal station (such as vehicle-mounted flat panel antenna, ship-mounted parabolic antenna, etc.), and different models have different physical inertia and electronic scanning speeds.

[0058] Specifically, the main station performs statistical analysis on the historical records of each terminal type and calculates its switching time T. _switch The mean and standard deviation. Typically, the mean plus three times the standard deviation or the historical maximum value is taken as the maximum transmit / receive switching time for this type of terminal, to ensure the security of the plan.

[0059] Based on this, the generated antenna transmit / receive switching time distribution data includes the maximum transmit / receive switching time for various types of terminals and the recommended safety margin T. _margin For example, for Type I terminals, the maximum transmit / receive switching time is 500µs; for Type II terminals, the maximum transmit / receive switching time is 1200µs. These figures will directly determine the width of the protection band in subsequent time planning.

[0060] Step 203: Integrate the terminal station delay group statistics data with the antenna transmit / receive switching time distribution data in a structured manner to form dual-network statistical basic data for use by the dual-coupled time planning model.

[0061] In this step, structured integration refers to encapsulating the aforementioned scattered statistical results into a standardized input interface format (such as JSON or XML objects) for direct reading by the mathematical modeling module. This dual-network statistical foundation data serves as a bridge connecting the physical world and the mathematical model, ensuring that subsequent planning results are not only theoretically optimal but also physically feasible.

[0062] Example 3 describes possible implementation methods for the mathematical construction of a dual-coupled time programming model, particularly how to transform physical constraints into mathematical inequalities to achieve accurate control of the dual-network time-series structure. This example specifically includes:

[0063] Step 301: Read the load target data of the two networks (original network and mirror network), and define the start and end times of the forward window of the original network, the start and end times of the reverse window of the original network, the start and end times of the forward window of the mirror network, the start and end times of the reverse window of the mirror network, and the superframe period length as the set of optimization variables for the dual coupled time programming model.

[0064] In other words, the target load data of the dual networks (original network and mirror network) is read, and the start and end times of the forward window of the dual networks, the start and end times of the reverse window of the dual networks, and the length of the superframe period are defined as the set of optimization variables for the dual coupled time planning model.

[0065] In this step, the dual-network load target data includes the uplink and downlink throughput requirements of the original network and the mirror network, respectively. For example, the original network has a downlink requirement of 100Mbps and an uplink requirement of 20Mbps.

[0066] Optionally, the following non-negative continuous variables can be defined in the mathematical model:

[0067] T _cycle , that is, the superframe period length;

[0068] t _Of _s、t _Of_e That is, the start and end times of the original network forward window;

[0069] t _Or_s t _Or_e That is, the start and end times of the reverse window of the original network;

[0070] t _Mf_s t _Mf_e That is, the start and end times of the forward window of the mirror network;

[0071] t _Mr_s , t _Mr_e This refers to the start and end times of the mirror network's reverse window.

[0072] The above variables together constitute the vector space X to be solved. All time variables are defined relative to the starting point of the superframe period (t=0).

[0073] Step 302: Establish delay reachability constraints by utilizing the round-trip propagation delay distribution characteristics of the terminal station group.

[0074] Specifically, this includes: dividing the terminal station group into several terminal station delay groups, and extracting the maximum round-trip propagation delay statistics for each terminal station delay group;

[0075] For each terminal station delay group, a time interval inequality is established in the dual coupled time programming model to limit the difference between the end time of the forward window and the start time of the reverse window in the set of optimization variables.

[0076] The time interval inequality requires that the difference must be greater than the sum of the maximum round-trip propagation delay statistics of the terminal delay group and the preset processing delay, so that the terminal in the group has a sufficient physical time window to send back after receiving the forward instruction.

[0077] This step aims to address the command-response timing coordination issue in long-latency satellite links. Optionally, for any terminal delay packet k, its maximum round-trip time is Max. _Delay_k In other words, the maximum round-trip propagation delay statistic. To ensure that the endpoints within this group have sufficient time to process data received from the forward time slot and wait for the reverse time slot to arrive, the following inequality constraint must be satisfied:

[0078] t _Or_s ≥t _Of_e +Max _Delay_k +T _proc_min ;

[0079] Among them, T _proc_min This is the minimum baseband processing time required for the terminal station to process the forward data, or the minimum preset processing delay. This formula indicates that the opening time of the reverse transmission window must be later than the closing time of the forward reception window plus the signal round-trip flight time and processing time. Similarly, similar constraints need to be established for mirror networks:

[0080] t _Mr_s ≥t _Mf_e +Max _Delay_k +T _proc_min .

[0081] Furthermore, the above inequalities ensure that the coupling structure of the dual network in the time domain satisfies the laws of physical propagation.

[0082] Step 303: Establish antenna switching security constraints using the phased array antenna transmit / receive switching time characteristics.

[0083] Furthermore, the maximum transmit / receive switching time parameters for different terminal types under maximum load conditions are extracted from the antenna transmit / receive switching time distribution data;

[0084] In the dual-coupled time planning model, a state switching interval inequality is constructed for each terminal type. The state switching interval inequality limits the time difference between the end time of the receiving window and the start time of the sending window for that type of terminal, as well as the time difference between the end time of the sending window and the start time of the receiving window.

[0085] The state switching interval inequality requires that the time difference must be greater than the sum of the maximum transmit / receive switching time parameter and the preset protection safety margin to prevent transmit / receive state conflicts from occurring at the end station during the switching process between the original network and the mirror network.

[0086] This step is to prevent physical conflicts from occurring during rapid switching between transmit and receive states at the terminal station. For any terminal type m, its maximum transmit / receive switching time is T. _sw_Max (m). The model must contain a state transition interval inequality of the following form:

[0087] (t _TX_s -t _RX_e )≥T _sw_Max (m)+T _safety ;

[0088] Among them, t _TX_s Refers to the start time of the next task sent by this terminal (possibly t). _Or_s or t _Mr_s ), t _RX_e Refers to the end time of the last task received by this terminal (possibly t). _Of_e or t _Mf_e In other words, T _safety Corresponding to a preset protection margin. For example, if a certain end station is scheduled to receive forward data in the original network (ending at t) _Of_e To send reverse data on the mirror network (starting at t) _Mr_s Then the constraint is:

[0089] t _Mr_s -t _Of_e ≥T _sw_Max (m)+T _safety This constraint forces the optimization algorithm to automatically insert a sufficiently wide guard band between the time windows of the two networks.

[0090] Step 304: Combine the main station receiving channel constraints to establish a zero-overlap constraint for the main station receiving, forming a constraint set for the dual coupled time programming model.

[0091] Based on this, and taking into account the limitation on the number of demodulation channels of the master station receiver, a mutually exclusive time interval constraint is defined in the dual-coupled time planning model to prevent the original network reverse window and the mirror network reverse window from overlapping on the master station time axis.

[0092] A window interval constraint is constructed, requiring a minimum buffer time to be maintained between the end time of the original network's reverse window and the start time of the mirror network's reverse window, as well as between the end time of the mirror network's reverse window and the start time of the original network's reverse window, to meet the switching and data buffering requirements of the master station's receiving tasks. This is used to adapt to the limitations of the master station's receiving capabilities.

[0093] Accordingly, if the master station only has one set of broadband receiving and demodulation resources, it cannot simultaneously process uplink burst signals from both the original network and the mirror network. Furthermore, the following mutually exclusive interval constraints are defined in the model:

[0094] interval [t] _Or_s , t _Or_e ] and interval [t _Mr_s , t _Mr_e The intersection of [ ] is an empty set.

[0095] In mathematical programming, this can usually be achieved by introducing 0-1 integer variables or mutually exclusive inequalities (Big-M method). For example, to find t _Mr_s ≥t _Or_e +T _guard_GW , or t _Or_s ≥t _Mr_e +T _guard_GW Among them, T _guard_GW This is the buffer time required for the main station to switch receiving channels.

[0096] Step 305: Based on the throughput requirements of different service types in the load target data of the dual networks (original network and mirror network), construct an optimization function with the objective of maximizing the effective load time ratio, and complete the construction of the dual coupled time planning model.

[0097] In this step, optimization aims to maximize the net time for transmitting valid data while satisfying all the hard constraints mentioned above. In other words, the time windows in the following formulas are all net transmission times. Accordingly, the objective function J can be expressed as:

[0098] MaximizeJ=[w _1 ×(t _Of_e -t _Of_s )+w _2 ×(t _Or_e -t _Or_s )+w _3 ×(t _Mf_e -t _Mf_s )+w _4 ×(t _Mr_e -t _Mr_s )] / T _cycle ;

[0099] Among them, w_1 to w _4 These are weighting coefficients set based on business priority. For example, if the forward and downlink demand of the original network is high in the current network, then w can be... _1 Setting it to a larger value allows the bootstrap algorithm to allocate a wider range of t. _Of In other words, allocate a wider forward window to the original network. Divide by the superframe period length T. _cycle This is for normalization, to reflect resource utilization indicators. By solving this objective function, the system can automatically find a cost-effective time window partitioning scheme.

[0100] Example 4 provides an exemplary scheme for simplified solution, verification, and parameter generation of the model, which helps in finding the optimal solution, especially when the number of model constraints is large, and how to perform engineering simplification. Accordingly, the scheme includes:

[0101] Step 401: Use a constrained optimization algorithm to iteratively solve the dual coupled time programming model to obtain a preliminary solution including time window parameters.

[0102] In this step, the choice of constraint optimization algorithm depends on the mathematical properties of the objective function and constraints. Since all time delay constraints and switching constraints are modeled as linear inequalities, and the objective function is linear, linear programming (LP) algorithms such as the simplex method or interior-point method are preferred for solving the problem. These algorithms guarantee finding the global optimum in polynomial time.

[0103] Furthermore, if the model introduces nonlinear penalty terms, such as quadratic terms to smooth window jitter, or discrete integer variables, such as 0-1 variables to represent mutually exclusive intervals, a mixed-integer linear programming (MILP) solver or heuristic search algorithm, such as a genetic algorithm or simulated annealing algorithm, can be used. Based on this, the main station can preload the solution from the previous round as the initial value for the current iteration to improve the solution speed.

[0104] For example, to reduce computational complexity, a constraint simplification strategy can be employed. The master station does not need to establish independent inequality constraints for every single terminal station in the network. Instead, the master station only establishes constraints for the boundary cases of each terminal station's latency group. For instance, for far-end groups, only the virtual terminal station with the highest latency is selected as a representative to establish constraints; for near-end groups, only the virtual terminal station with the lowest latency is selected to establish constraints. In this way, the number of constraints is reduced from being proportional to the number of terminals to being proportional to the number of groups, keeping the computational scale within a millisecond-level solvable range.

[0105] Step 402: Use the boundary grouping parameters in the terminal station delay grouping statistics to perform a physical feasibility test on the preliminary solution and determine whether there are terminal stations with specific delay groups that cannot meet the delay reachability constraints.

[0106] In this step, the preliminary solution is the set of time parameters calculated by the algorithm under simplified constraints, such as {t}. _Of_e =10ms, t _Or_s =15ms}. Physical feasibility testing is a safety verification process used to prevent the simplified model from omitting certain extreme cases.

[0107] Accordingly, the master station will substitute the initial solution into the worst-case parameters of all terminal station delay packets for verification. For example, the master station checks the maximum round-trip delay Max of the farthest distance packet. _RTT_Far Does it meet the following conditions:

[0108] t _Or_s -t _Of_e ≥Max _RTT_Far +T _proc ;

[0109] Among them, T _proc Corresponding to the preset processing delay.

[0110] If the calculation result shows t _Or_s -t _Of_e =5ms, while Max _RTT_Far +T _proc If the time is 5.2ms, the preliminary solution is determined to be unreachable for the remote packet, indicating a risk of physical conflict. In this case, a correction process must be triggered.

[0111] Step 403: When an unreachable risk is detected, the dual-coupled time planning model is reconstructed and solved again by adjusting the protection band width in the constraint set or correcting the business weight in the optimization function until the dual-network time planning parameter data that satisfies all physical constraints is output.

[0112] In this step, adjusting the constraint set typically involves tightening the boundary conditions. For example, in the case above, the main site would add stricter constraints to the model, forcing t... _Or_s -t _Of_e ≥5.3ms, or increase the global protection variable. Adjusting service weights refers to reducing the weight of certain low-priority services in the objective function, sacrificing some throughput in exchange for greater time margin.

[0113] Specifically, the solution, after multiple iterations and revisions and finally passing the feasibility test, will be formatted as dual-network time planning parameter data. This data can be represented in a compact binary format or a scalable JSON format.

[0114] For example, {SuperFrame_ID (superframe identifier): 1024,

[0115] Cycle_Length (superframe cycle length): 20000 (us),

[0116] Original_Net: {Fwd_Win (forward window): [0, 8000], Rev_Win (reverse window): [12000, 16000]}

[0117] Mirror_Net: {Fwd_Win (forward window): [8500, 11500], Rev_Win (reverse window): [16500, 19500]}

[0118] Group_Config (End Station Latency Group Configuration): [{Group_ID (Group Identifier): 1, Offset (Time Offset): 500}, {Group_ID (Group Identifier): 2, Offset (Time Offset): 800}]};

[0119] Based on this, the data structure defines the time topology of the dual network and will be sent to the broadcast module.

[0120] Example 5 describes an optional implementation of the collaborative feedback mechanism based on local observations at the terminal station, particularly how the terminal station utilizes Network Clock Reference Information (NCR) for delay drift observation, and how to construct and compress time slot quality feedback data. Furthermore, this example compensates for the limitations of the main station's panoramic view through edge sensing.

[0121] In this embodiment, each terminal station parses the time slot configuration command, generates local time slot mapping data, performs data transmission and reception based on the mapping data, observes the actual operation status of the service time slots, and generates terminal time slot quality feedback data containing time slot quality indicators, which is then reported to the main station.

[0122] In this step, the local timeslot mapping data at the terminal station is the absolute time parameter (such as t on the master station's time axis) that the terminal station will receive. _Of_s This is a product of the process of converting the data into a local counter value. Because each terminal station is at a different distance from the satellite, there is a fixed propagation time difference between its local reception time and the master station's transmission time. The terminal station must use its locally maintained delay compensation value to shift the parameters in the command to obtain the correct local operating time window.

[0123] For example, by using the time difference between the network clock reference information broadcast by the master station and the local clock of the terminal station, and combining the forward and reverse time slot relationships in the local time slot mapping data of the terminal station, the estimated value of the round-trip delay change in the current observation period is calculated.

[0124] In this step, the Network Clock Reference (NCR) is a timestamp periodically embedded in the data stream by the master station, representing the system reference time on the master station side. Accordingly, the end station compares the NCR value recovered using a phase-locked loop (PLL) with its local clock counter to calculate the delay drift Δτ(Δτ). _τ Its calculation formula can be expressed as:

[0125] Δ _τ =(T _NCR_current -T _Local_expected )-Δ _initial ;

[0126] Among them, T _NCR_current It is the currently decoded NCR value, T _Local_expected It is the expected local time of the end station in the synchronization state, Δ _initial This is the initial deviation calibrated during the last synchronization. If Δ _τ A positive and continuously increasing value indicates that the satellite is moving away from the terminal station or that atmospheric delay is increasing, meaning that the required protection zone for the terminal station must be widened. This observation index can be used to dynamically adjust the time delay grouping boundaries.

[0127] Optionally, for the service time slots of the original network and the mirror network, the number of sending acknowledgments and receiving failure modes are counted to identify abnormal packet loss caused by time slot overlap or insufficient switching, and collision probability statistics are generated.

[0128] In this step, the terminal station needs to distinguish between packet loss caused by poor channel quality and packet loss caused by time slot collisions. A decision logic based on signal characteristics can be used. That is, if the terminal station detects high-intensity radio frequency energy (high RSSI) within a certain receiving time slot but cannot demodulate valid data (CRC check failure), and this time slot is located at the edge of the time window, it is very likely because antenna switching has not yet been completed or a collision with a reverse signal has occurred. Furthermore, the terminal station counts the failure events of a specific mode, denoted as N. _collision At the same time, the total number of attempts N is counted. _total Collision probability P _coll It can be calculated as: P _coll =N _collision / N _total .

[0129] Optionally, the estimated round-trip delay variation can be integrated with the collision probability statistics to serve as the basic observation data reflecting the current physical link status of the terminal station.

[0130] Optionally, the basic observation data is subjected to dimensionality reduction processing, and the weighted average of the service slot success rate, the weighted average of the collision probability, and the statistical characteristic value of the delay offset are calculated in the original network and the mirror network respectively. The calculated weighted average and statistical characteristic values ​​are arranged in a predetermined order to construct a fixed-length slot quality evaluation vector.

[0131] In this step, to conserve bandwidth on the reverse control channel, the end station cannot upload all raw logs; vectorized compression is necessary. The time slot quality assessment vector is a fixed-length data structure, which can be defined, for example, in the following format:

[0132] V _quality =[Metric _1 Metric _2 Metric _3 Metric _4 ];

[0133] Among them, Metric _1 =Original network average success rate (weighted average, recent data has higher weight); Metric _2 =Metric Mirror Network Average Success Rate; _3 =Global collision probability factor (P) _coll (normalized value); Metric _4 =Mean round-trip delay drift (Mean) _Δ_τ ).

[0134] Optionally, the terminal station can also add a flag to the vector requesting a wider guard band, based on its own business importance level.

[0135] Optionally, the time slot quality assessment vector is packaged with the current local clock offset estimation information of the terminal station to form terminal time slot quality feedback data, which is used by the master station for low-bandwidth batch state fusion. Accordingly, this vector, together with the terminal station ID and the local clock offset value, is encapsulated into a terminal time slot quality feedback data packet and sent to the master station in the specified reverse control time slot.

[0136] Example 6 describes the specific processing procedure for feedback fusion and statistical data updating on the master station side. It explains how the master station processes feedback data from the terminal stations, uses statistical methods to update the system's basic parameters, and forms a closed adaptive loop of perception-planning-execution-feedback.

[0137] Specifically, the master station receives and integrates time slot quality feedback data from multiple end stations, corrects the round-trip propagation delay distribution characteristics and phased array antenna transmit / receive switching time characteristics in the dual-network statistical basic data, and uses the corrected dual-network statistical basic data to trigger the construction and solution of the next round of dual-coupled time planning model, thereby realizing adaptive iterative optimization of the dual-network timing structure.

[0138] Furthermore, this embodiment can be implemented by performing the following steps:

[0139] Step 601: Aggregate and detect anomalies in the multiple received terminal station time slot quality feedback data, and calculate global time slot quality statistics reflecting the current network operating status. The global time slot quality statistics include the global delay offset mean and the global collision probability.

[0140] In this step, the master station may receive feedback from thousands of end stations within a statistical period (e.g., 1 second). To prevent erroneous data reported by individual faulty end stations (such as those with GPS lockout or hardware failure) from polluting the global model, anomaly detection must be performed.

[0141] Accordingly, the master station can use the median absolute deviation (MAD) or Z-score algorithm to filter the delay drift values ​​reported by all end stations. For example, calculate the Δ for all end stations. _τ The median was used to remove outliers that deviated from the median by more than three standard deviations. The cleaned data was then used to calculate the global delay offset mean and the global collision probability P. _coll_Global .

[0142] Step 602: Use the global delay offset mean to smooth the statistical data of the terminal delay groups in the dual network statistical basic data, and update the round-trip propagation delay distribution parameters of each delay group.

[0143] In other words, the global delay offset mean is used to smoothly adjust the terminal delay group statistics and update the round-trip propagation delay distribution parameters of each delay group.

[0144] In this step, if the global delay offset mean is... _Δ_Global A positive value (e.g., +20µs) indicates that the satellite has drifted relative to the ground. The master station does not replace parameters; instead, it uses an exponentially weighted moving average (EWMA) algorithm to calculate the Max value in the terminal station delay grouping statistics. _Delay_k The following updates will be made:

[0145] New _Max_Delay =α×(Old _Max_Delay +Mean _Δ_Global )+(1-α)×Current _Measurement ;

[0146] Where α is a smoothing factor (e.g., 0.1), used to prevent drastic parameter oscillations; that is, Old _Max_Delay That is, the statistical value of the maximum round-trip propagation delay before adjustment. Current _Measurement That is, the measured value of the maximum round-trip propagation delay within the current period, New _Max_Delay The updated maximum round-trip propagation delay statistics will be used for dynamic optimization of subsequent latency reachability constraints. In this way, latency reachability constraints will automatically adjust to follow the satellite's orbital drift.

[0147] Step 603: Use the global collision probability to determine the effectiveness of the current protection interval. When the collision probability exceeds the preset threshold, increase the safety margin parameter of the antenna transmit / receive switching time distribution data in the dual-network statistical basic data.

[0148] Alternatively, the effectiveness of the current protection interval can be determined by using the global collision probability. When the collision probability exceeds a preset threshold, the safety margin parameter of the antenna transmit / receive switching time distribution data can be increased.

[0149] In this step, if the global collision probability exceeds the system-defined safety threshold (e.g., 1%), and the collisions are mainly concentrated at the edge of the time window, then the current phased array antenna switching time parameter T is inferred. _sw_Max The settings are too aggressive, or the aging of some endpoints causes the switching to be slow.

[0150] In other words, if the global collision probability exceeds the preset threshold, and the collision mainly occurs at the edge of the time window, it indicates that the maximum transmit / receive switching time is set too high, or that some end stations are aging and the switching is slowing down.

[0151] Specifically, the master station will trigger a safety margin increment mechanism, increasing the safety margin T in the antenna transmit / receive switching time distribution data. _margin Increase the step size (e.g., +50µs). This update will directly result in wider antenna switching safety constraints in the next round of model building, eliminating collisions at the physical level. It will slightly sacrifice throughput but ensure system stability.

[0152] Furthermore, the updated terminal delay group statistics and antenna transmit / receive switching time distribution data are repackaged to generate corrected dual-network statistical base data.

[0153] Example 7 provides an exemplary scheme for scheduling execution based on priority queues, describing how the master station performs specific service data filling and signal transmission after calculating time parameters. Communication services can be directly generated. In this example, the following steps can be used:

[0154] Step 701: The master station defines the physical boundaries of the forward transmission window and the reverse reception window of the original network and the mirror network on the local time axis based on the dual network time planning parameter data.

[0155] In this step, the scheduler module of the main station reads JSON-formatted parameters and sets a timer interrupt at the hardware level. For example, the timer is set to trigger forward transmission from the original network at t=0, and to trigger an interrupt at t=8000us to stop transmission and switch to the mirror network waiting state.

[0156] Step 702: Within the forward transmission window, the master station performs queue management and priority sorting of data packets to be sent for different service types according to the service priority constraints in the dual network load target data, and generates master station hybrid duplex scheduling execution data.

[0157] In this step, the main station maintains multiple logical queues, which correspond to different service QoS levels such as voice (high priority), video (medium priority), and file download (low priority).

[0158] Specifically, when the raw network forward window opens, the scheduler checks the high-priority queue. If the queue is not empty, it retrieves a packet from the head of the queue for encapsulation; if the queue is empty, it checks the next-level queue. During the encapsulation process, the scheduler must calculate the remaining time of the current window in real time.

[0159] For example, if the transmission time corresponding to the length of the currently retrieved data packet exceeds the remaining time of the window, the master station can implement a fragmentation strategy (splitting the data packet, sending one part in the current window, and sending the remaining part in the next window) or a postponement strategy (putting the packet back to the head of the queue, and filling the current window with idle frames). This logic strictly limits data transmission to the planned time t. _Of_s and t _Of_e There will be no overflow that would interfere with the subsequent reverse receiving window.

[0160] Step 703: The master station performs signal modulation and transmission in the corresponding original network forward time slot or mirror network forward time slot according to the master station's hybrid duplex scheduling execution data, and opens the receiving channel in the corresponding reverse receiving window to centrally receive uplink burst data from the end station.

[0161] In this step, the mixed-duplex scheduling execution data is a baseband waveform stream that has undergone physical layer encoding (such as LDPC encoding) and modulation (such as QPSK / 8PSK). The master station's RF front-end, according to the scheduling instructions, activates the power amplifier (PA) to transmit signals or the low-noise amplifier (LNA) to receive signals at precise timing. Through microsecond-level control, the master station achieves high-speed switching between the original network and the mirror network, exhibiting a mixed-duplex communication effect.

[0162] In summary, this invention constructs a closed-loop system encompassing physical sensing, mathematical modeling, optimization, and execution feedback. This system not only theoretically solves the duplex conflict problem in high-throughput satellite communication but also fully considers computational complexity, feedback bandwidth overhead, and hardware physical limitations in its engineering implementation.

[0163] Example 8 describes an exemplary scheme for hybrid duplex transmission and reception execution and hardware state control on the terminal side. This scheme involves collaborative execution and further illustrates how the terminal station utilizes local mapping data to drive the underlying hardware to complete hybrid duplex operations. Accordingly, this example can be implemented by performing the following steps:

[0164] Step 801: Based on the local time slot mapping data, the terminal station starts and stops the working state of the receiver and transmitter on the local time axis, and generates the terminal station hybrid duplex transceiver execution data by combining the transceiver switching control strategy of the shared phased array antenna.

[0165] In this step, the local timeslot mapping data at the terminal station is a timetable that the terminal station has parsed and performed local latency correction on. The terminal station's central controller (CPU or FPGA) reads this timetable and converts it into a series of timestamped hardware control instruction sequences, which constitute the terminal station's mixed-duplex transmit and receive execution data.

[0166] Accordingly, this process involves precise timing control at the microsecond level. The end station controller continuously scans the local clock, and when the current time is close to the start time of a forward receive time slot of a certain original network, the controller will issue an antenna pointing adjustment and low noise amplifier (LNA) activation command in advance; when it is close to the start time of the reverse transmit time slot of the mirror network, the controller will issue a power amplifier (PA) warm-up and modulator turn-on command in advance.

[0167] Step 802: During the transmit / receive state switching process, the terminal station performs pre-switching control based on hardware characteristics to ensure that the antenna is in a stable state at the start of the service time slot. This is used to solve the physical delay problem of phased array antenna beam reconfiguration and RF channel switching.

[0168] An alternative implementation is for the terminal station to allocate a lead time before the time slot boundary specified in the mapped data, based on its own hardware's transmit / receive switching control command response time. For example, if the next task is a reverse transmission, and the start time specified in the mapped data is T... _start The physical switching time of the antenna is T. _sw Then the terminal station will be at time (T) _start -T _sw -T _margin A switching command is issued. This pre-switching mechanism ensures that when time T reaches... _startAt that time, the antenna has completed the beamforming weight update and impedance matching from the receive mode to the transmit mode, avoiding bit loss at the beginning of the time slot.

[0169] It should be understood that, through hardware state machine control, the terminal station can achieve high-frequency, seamless switching between receiving data from the original network and sending data from the mirror network.

[0170] Example 9 describes an exemplary scheme for a rapid system reconfiguration and triggering mechanism based on real-time monitoring, particularly the verification phase, and also describes the exception handling logic during system operation. Further, this example is based on event-triggered updates (fast loop) driven by real-time faults, which includes:

[0171] Step 901: The master station performs joint analysis on the master station's mixed-duplex scheduling execution data and the end station's mixed-duplex transmit and receive execution data, and monitors the data transmission integrity, bit error rate, and transmit / receive conflict in each time slot in real time.

[0172] In this step, the master station acts as both the scheduler and the supervisor. It calculates the real-time bit error rate (BER) and packet loss rate (PER) by comparing the planned data to the actual received data.

[0173] Furthermore, the master station will pay special attention to service packets located at the edge of the time window (i.e., the area most prone to conflict due to improper time slot planning). If the master station finds consecutive checksum errors in the frame header or tail of uplink data packets from a specific end station, or if the master station detects an abnormal energy burst within the guard band that should be idle, it indicates that the current dual-network time planning parameters are no longer applicable (for example, a sudden change in latency due to satellite maneuvering exceeds the tolerance of the original guard band).

[0174] Step 902: When signs of violating zero-conflict constraints are detected or resource utilization deviates significantly from expectations, a new round of rapid reconfiguration is triggered by adjusting the configuration of dual-network load target data or shortening the update cycle of the next time planning. This step is part of the emergency response mechanism. When monitored indicators exceed safety thresholds, the master station immediately abandons the planning parameters for the current cycle and forcibly triggers a full-process reconfiguration.

[0175] The implementation can be achieved using the following strategies: First, a degradation strategy, in which the master station temporarily modifies the dual network load target data input into the model to artificially reduce the service throughput requirement in exchange for a wider time protection band and prioritize the stability of the connection; Second, an acceleration strategy, in which the master station shortens the period of model calculation and parameter broadcasting from the original (for example) 1 second to 100 milliseconds to track channel changes at a faster frequency.

[0176] It should be understood that this mechanism enables the system to recover to a stable state with zero conflicts through replanning when it encounters sudden disturbances, demonstrating the high availability of the system.

[0177] According to one aspect of this application, the data for constructing a dual-network time planning model may also include:

[0178] In this step, the master station reads the basic statistical data and target load data of both networks. Based on the superframe structure designed by the system, it determines the set of time planning parameters that need to be optimized. The master station defines the superframe period length variable on the time axis and defines the start time, end time, start time, and end time of the original network forward window, the mirror network forward window, the mirror network forward window, the mirror network reverse window, and the mirror network reverse window as variables to be optimized.

[0179] Building upon this, the main station introduces auxiliary variables such as the forward window length, the reverse window length, and the intervals between forward and reverse windows, forming a set of time planning variables. The main station then organizes these variables in a structured manner according to the logical division of the original and mirror networks, obtaining time planning variable data that serves as the independent variables of the model. This time planning variable data is denoted as the dual-network time planning variable data. It is used to construct various constraints and objective functions.

[0180] Furthermore, the master station reads the terminal station latency group statistics data and dual network time planning variable data from the dual network statistical basic data, and constructs reachability constraints between forward data arrival and reverse data return for each group based on the maximum round-trip propagation latency and processing latency margin within different latency groups.

[0181] Furthermore, for each type of delay packet, the master station calculates the minimum time interval that needs to be retained between the end time of the forward window of the original network and the start time of the available reverse time slot for that type of end station, and represents this interval using a time planning variable. The master station organizes the reachability inequality constraints of all delay packets to form a set of delay reachability constraints in the original network direction; a similar set of delay reachability constraints in the mirror network direction is constructed to ensure that the forward and reverse windows in the mirror network also meet the delay requirements of various types of end stations.

[0182] Furthermore, the main site will uniformly encode the constraint set based on time planning variables and delay grouping statistical parameters into delay reachability constraint data, which can be used to build time planning models.

[0183] Furthermore, the master station reads the antenna transmit / receive switching time distribution data from the dual-network statistical basic data, and combines the dual-network time planning variable data and latency reachability constraint data to construct transmit / receive switching safety constraints for the phased array antennas on the master station side and the phased array antennas on each terminal station side.

[0184] Furthermore, for each terminal type, the master station extracts the maximum transmit / receive switching time of that type of terminal during high-load periods and the transmit / receive switching time of the master station's antenna under maximum load conditions from the antenna transmit / receive switching time distribution data. Based on the maximum switching time, the master station introduces a protection time for transmit / receive state switching into the time planning variables, requiring that the time interval between the end of the transmit window and the beginning of the receive window, and between the end of the receive window and the beginning of the transmit window, is not less than the maximum transmit / receive switching time of the corresponding antenna type plus a safety margin.

[0185] Furthermore, through comprehensive analysis of all terminal types and the master station antennas, the master station constructs a handover safety constraint inequality with time planning variables as independent variables and antenna transmit / receive switching time distribution as parameters, forming antenna handover safety constraint data. This data, together with delay reachability constraint data, defines the safety boundary of dual-network time slot planning in terms of transmit / receive switching.

[0186] Optionally, the master station reads the dual-network time planning variable data and, in conjunction with the system structure data regarding the number of master station receiving channels and demodulation capabilities, constructs a zero-overlap constraint on the original network reverse window and the mirror network reverse window on the master station's time axis.

[0187] Optionally, the master station represents the start and end times of the original network reverse window and the mirror network reverse window as time planning variables. By sorting the intervals on the time axis, the master station is required to allow only one type of reverse window to be in the receiving state at any given time, or to maintain the necessary guard interval before and after the receiving state. In order to take into account the hardware receiver switching and buffering requirements, the master station further adds a minimum interval time condition in addition to the zero overlap condition, so that there is sufficient buffer time during the switching of the two types of network reverse windows.

[0188] Optionally, the master station will uniformly organize the constraint expressions into zero-overlap constraint data received by the master station, which can be combined with the time delay reachability constraint data and the antenna switching security constraint data to form a constraint set for the time planning problem.

[0189] For example, the master station reads the dual-network load target data, and combines it with the dual-network time planning variable data, latency reachability constraint data, antenna switching security constraint data, and master station reception zero overlap constraint data to construct the objective function of the time planning problem.

[0190] Accordingly, based on the parameters such as uplink traffic volume, downlink traffic volume, and priority weight of the original and mirrored networks given in the dual-network load target data, the master station multiplies the total available time slot duration of each type of service within the corresponding window by the weight of that service, and adds the contributions of each type of service to obtain a weighted sum of the effective payload time. The master station then divides the weighted sum by the superframe period length variable to obtain the objective function expression reflecting resource utilization.

[0191] In the objective function, the master station can set different weights for the time slot contributions of different networks and directions based on the latency sensitivity and reliability requirements of different service types, so as to reflect the different usage strategies between the original network and the mirror network. The master station organizes this objective function together with the aforementioned constraints into dual-network time planning optimization objective data. This dual-network time planning optimization objective data will be encapsulated into the time planning model along with various constraint data.

[0192] In this step, the master station encapsulates the dual-network time planning variable data, latency reachability constraint data, antenna switching security constraint data, master station reception zero overlap constraint data, and dual-network time planning optimization objective data into a unified mathematical model description. This model description includes variable definitions, constraint sets, and objective function expressions, and is encoded using a unified data structure for easy invocation by subsequent time planning algorithm solvers.

[0193] In this step, the master station will attach the model version number and applicable statistical time window information during encapsulation to identify the dual-network statistical foundation data on which the model is based. This encapsulated model data is the dual-network time programming model data, which can be used to perform the solution process of zero-conflict adaptive time programming.

[0194] According to another aspect of this application, the following approach can also be used to solve the dual-network time programming model:

[0195] In this step, the main station reads the data from the dual-network time programming model and analyzes the size of the constraints, the number of variables, and the structural characteristics of the objective function. Based on whether the linear programming conditions are met or whether there are obvious nonlinear terms, the main station selects an appropriate solution algorithm type, including linear programming, mixed-integer programming, or hierarchical heuristic search.

[0196] In this step, the master station initializes the solution parameters of the selected algorithm, such as the maximum number of iterations, convergence threshold, and initial solution strategy. If necessary, the dual-network time planning parameter data obtained from the previous round of time planning can be used as the initial solution for the current solution process to accelerate convergence. The master station records the selected solution algorithm type and the initialized algorithm parameters together to form the time planning solution configuration data.

[0197] Optionally, the master station reads the dual-network time planning model data and time planning solution configuration data, and combines the end station grouping information in the end station delay grouping statistics to load and simplify the constraints in the model by group.

[0198] Optionally, the master station generates representative constraints for different sets of endpoints based on latency groups, instead of defining constraints for each endpoint individually, significantly reducing the number of constraints. The master station abstracts the upper and lower bounds of latency for each group, replacing the constraints of individual endpoints within a group with group-level interval constraints.

[0199] Optionally, for antenna transmit / receive handover security constraints, the master station also groups terminals according to terminal type, abstracting the handover time distribution of terminals of the same type into a unified constraint expression. While retaining the boundary conditions of key terminals, the master station filters out redundant constraints, obtaining a constraint set of manageable size that still covers the worst-case scenario, forming the simplified time planning model data. This simplified time planning model data will be used as the model input during actual solution processing.

[0200] Furthermore, based on the simplified time planning model data and time planning solution configuration data, the master station initiates the time planning algorithm to iteratively optimize the time planning variables for the dual-network network. During the solution process, the master station monitors constraint satisfaction and objective function convergence. After the algorithm provides a preliminary solution, the master station uses the original latency reachability constraint data and antenna switching safety constraint data to perform a feasibility test on the solution. If it is found that certain packets may face risks of transmission unreachability or insufficient handover under boundary conditions, the master station will slightly tighten the constraints or adjust the objective function weights by increasing the corresponding packet protection time, shortening some window lengths, or reducing the weights of certain services. After adjustment, the master station reruns the solution algorithm until a solution that meets the feasibility requirements under all key constraints is obtained, forming verified time planning result data, which is denoted as the dual-network time planning solution result data.

[0201] Furthermore, the master station maps the time planning variable values ​​in the dual-network time planning solution data back to specific physical time parameters, including the superframe period length, the start and end times of the original network forward window, the start and end times of the original network reverse window, the start and end times of the mirror network forward window, the start and end times of the mirror network reverse window, and the interval time between each window. Based on the variable values, the master station calculates the length of each type of window and the length of the protection interval between windows. Simultaneously, combining the terminal station delay group statistics, it records the protection band length corresponding to each delay group. The master station organizes the solved and verified time parameters into a unified data structure, forming dual-network time planning parameter data, which can be converted into dual-network time slot configuration command data and broadcast by the master station to each terminal station.

[0202] To address the dilemma of guard band setting caused by the inability of existing fixed configuration schemes to adapt to differences in geographical latency and hardware switching time at terminal stations (too large wastes resources, too small causes conflicts), this solution employs a dual-coupled time programming mathematical model. The round-trip propagation delay distribution characteristics of the terminal station group and the phased array antenna transmit / receive switching time characteristics are transformed into latency reachability inequality constraints and state switching interval constraints in the mathematical model, iteratively solving the problem with the objective of maximizing the effective payload time proportion. This approach digitizes physical constraints, and the automatically generated dynamic timing structure can match the current physical link state. While ensuring zero physical layer conflicts (solving hardware conflict issues), the guard band is compressed to its theoretical minimum, improving resource utilization (solving resource waste issues).

[0203] Building upon this, this solution employs a terminal station collaborative feedback and adaptive update mechanism to address the lack of a real-time sensing mechanism in the system, which leads to a decrease in alignment accuracy due to satellite drift or clock skew during long-term operation. The embodiment describes how the terminal station uses Network Clock Reference (NCR) information to calculate delay drift, distinguishes between channel noise and signal collisions, and generates a quantified quality assessment vector that is fed back to the master station. The master station then dynamically corrects the basic statistical data and triggers model reconstruction based on this vector. This closed-loop control method allows time planning parameters to adaptively adjust with environmental changes, resolving the timing synchronization problem during long-term operation.

[0204] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all fall within the protection scope of the present invention.

Claims

1. A method of implementation suitable for high throughput satellite communication hybrid duplex system, characterized in that, The method comprises the following steps: Based on the round-trip propagation delay distribution characteristics of the terminal group and the transmission-reception switching time characteristics of the phased array antenna, a dual-coupled time planning model is constructed, taking the start and end time of the time window of the original network and the mirror network as the optimization variables; Under the physical constraints of meeting the zero overlap of transmission-reception state and the delay reachability, the above model is solved to generate dual-network time planning parameter data defining the time sequence structure of the dual network; The master station generates time slot configuration instructions according to the dual-network time planning parameter data and broadcasts them, and cooperatively controls the full-network terminals to perform mixed duplex data transmission between the original network and the mirror network; Wherein, the dual-coupled time planning model taking the start and end time of the time window of the original network and the mirror network as the optimization variables comprises: reading the dual-network load target data, defining the dual-network forward window start and end time, the dual-network reverse window start and end time, and the superframe cycle length as the optimization variable set of the dual-coupled time planning model; the delay reachability constraint is established by using the round-trip propagation delay distribution characteristics of the terminal group, the antenna switching safety constraint is established by using the transmission-reception switching time characteristics of the phased array antenna, and the master station reception zero overlap constraint is established by combining the master station reception channel limit to form the constraint set of the dual-coupled time planning model; based on the throughput demand of different service types in the dual-network load target data, an optimization function is constructed to maximize the effective payload time ratio as the target, and the dual-coupled time planning model is completed; Wherein, the delay reachability constraint is established by using the round-trip propagation delay distribution characteristics of the terminal group, which comprises: dividing the terminal group into several terminal delay groups, and extracting the maximum round-trip propagation delay statistical value of each terminal delay group; for each terminal delay group, a time interval inequality is established in the dual-coupled time planning model to limit the difference between the forward window end time and the reverse window start time in the optimization variable set; the time interval inequality requires that the difference must be greater than the sum of the maximum round-trip propagation delay statistical value of the terminal delay group and the preset processing delay, so that the terminals in the group have enough physical time window to perform reverse transmission after receiving the forward instruction; Wherein, the master station reception zero overlap constraint is established by combining the master station reception channel limit, which comprises: based on the demodulation channel number limit of the master station receiver, a mutual time interval constraint is defined in the dual-coupled time planning model to prohibit the overlap of the original network reverse window and the mirror network reverse window on the master station time axis; a window interval constraint is constructed to require that the end time of the original network reverse window and the start time of the mirror network reverse window, and the end time of the mirror network reverse window and the start time of the original network reverse window, must be kept apart by a minimum buffer time.

2. The method according to claim 1, characterized in that, The antenna switching safety constraint is established by using the transmission-reception switching time characteristics of the phased array antenna, which comprises: Extract the maximum transmission-reception switching time parameter of different terminal types under the maximum load condition from the antenna transmission-reception switching time distribution data; In the dual coupled time planning model, a state switching interval inequality is constructed for each terminal type, which limits the time difference between the end time of the receiving window and the start time of the sending window of the terminal of this type, and the time difference between the end time of the sending window and the start time of the receiving window; The state switching interval inequality requires that the time difference must be greater than the sum of the maximum transceiver switching time parameter and the preset protection safety margin, to prevent the transceiver state conflict of the terminal during the switching process of the original network and the mirror network.

3. The method according to claim 1, characterized in that, The method further includes solving and verifying the dual coupled time planning model, specifically: A constraint optimization algorithm is used to iteratively solve the dual coupled time planning model to obtain a preliminary solution containing time window parameters; The boundary packet parameters in the terminal delay packet statistical data are used to perform physical feasibility detection on the preliminary solution to determine whether the terminal of a specific delay packet cannot meet the delay reachability constraint; When the unreachable risk is detected, the protection bandwidth in the constraint set or the service weight in the optimization function is adjusted to trigger the reconstruction and re-solution of the dual coupled time planning model until the dual-network time planning parameter data that meets all physical constraints is output.

4. The method according to claim 1, characterized by, The method further includes the coordination time slot quality feedback of the terminal and the main station, specifically: Each terminal analyzes the time slot configuration instruction, generates local time slot mapping data of the terminal, executes data transmission and reception according to the mapping data, observes the running state of the actual service time slot, and generates terminal time slot quality feedback data containing time slot quality indicators and reports them to the main station; The main station receives and fuses the terminal time slot quality feedback data from multiple terminals, and modifies the round-trip propagation delay distribution characteristics and phased array antenna transceiver switching time characteristics in the dual-network statistical basic data; The modified dual-network statistical basic data is used to trigger the next round of dual coupled time planning model construction and solution.

5. The method according to claim 4, characterized in that, The terminal time slot quality feedback data is generated, including: The time difference between the network clock reference information broadcast by the main station and the local clock of the terminal is used to calculate the round-trip delay change estimate value in the current observation period, combined with the forward and reverse time slot relationship in the local time slot mapping data of the terminal; For the service time slots of the original network and the mirror network, the number of sending confirmations and the receiving failure mode are counted to identify abnormal packet loss caused by time slot overlap or insufficient switching, and a collision probability statistical indicator is generated; The round-trip delay change estimate value and the collision probability statistical indicator are integrated as basic observation data reflecting the current physical link state of the terminal.

6. The method according to claim 5, characterized in that, The terminal time slot quality feedback data is further generated, including: The basic observation data is processed by dimension reduction, and the weighted average value of the service time slot success rate, the weighted average value of the collision probability, and the statistical characteristic value of the delay offset in the original network and the mirror network are calculated respectively; The calculated weighted average values and statistical characteristic values are arranged in a predetermined order to construct a fixed-length time slot quality evaluation vector; The time slot quality evaluation vector and the current local clock offset estimation information of the terminal are packaged to form the terminal time slot quality feedback data.

7. The method according to claim 4, characterized in that, The main station modifies the round-trip propagation delay distribution characteristics and phased array antenna transceiver switching time characteristics in the dual-network statistical basic data, including: The received multiple end station time slot quality feedback data is aggregated and abnormity detected, global time slot quality statistical data reflecting current network running state is calculated, and the global time slot quality statistical data includes global time delay offset mean value and global collision probability; The global time delay offset mean value is used to perform smooth adjustment on the end station time delay group statistical data, and the round trip propagation time delay distribution parameters of each time delay group are updated; The global collision probability is used to judge the effectiveness of the current protection interval, and when the collision probability exceeds a preset threshold, the safety margin parameter of the antenna transmission-reception switching time distribution data is increased; The updated end station time delay group statistical data and the antenna transmission-reception switching time distribution data are re-encapsulated to generate corrected double network statistical basic data.

Images