Remote communication integrated satellite network service scheduling method based on hypergraph and task grading

Abstract

The invention belongs to the technical field of satellite communication and network resource scheduling, and particularly relates to a remote communication fusion satellite network service scheduling method based on a hypergraph and task grading, which breaks through the limitation that the number of resource types capable of being described by a traditional time expansion graph (TEG) is limited and resource conflict modeling is difficult by constructing a multi-dimensional expansion hypergraph (MDEH). Antenna resource conflicts and transmission mode conversion can be efficiently described, and the on-satellite storage efficiency is improved; according to the priority division and preemption-delay allocation scheduling strategy (PDS) based on dynamic energy correction, the reliable transmission of high-priority (emergency) services is preferentially guaranteed, and meanwhile, the overall network service scheduling success rate is effectively improved. The problem of limitation of an existing graph model in multi-domain resource modeling is solved, and the scheduling success rate of a network in a complex service scene is improved.

Description

Technical Field

[0001] This invention belongs to the field of satellite communication and network resource scheduling technology, specifically relating to a remote communication integrated satellite network service scheduling method based on supergraph and task hierarchy. Background Technology

[0002] With further research into aerospace technologies, satellite networks are gradually becoming larger and more heterogeneous. Due to their wide-area coverage capabilities and potential advantages in user access and high capacity, satellite networks have become a major research focus. Remote Sensing and Communication Integrated Networks (RSCINs) include differentiated equipment such as remote sensing payloads and communication payloads. Their mission scheduling is affected by the time-varying networks of observation satellites (OS) and communication satellites (CS) and the differentiated downlink capacity of the satellites.

[0003] Existing methods for on-board resource scheduling and dynamic network topology representation primarily employ Snapshots Sequence Graph (SSG) and Time-Expanded Graph (TEG). However, when dealing with complex, multi-dimensional, heterogeneous satellite networks, traditional methods suffer from the following significant drawbacks: (1) Limited number of resource types that can be described: Traditional TEG can usually only model storage and communication resources in network transmission. Once it involves multidimensional heterogeneous resources such as computing resources, its limitations are exposed, and it is difficult to accurately represent the relationship between multiple heterogeneous resources in time and space.

[0004] (2) Limitations of resource conflict modeling: In heterogeneous satellite networks, data transmission is limited by radio frequency resources (such as the number of antenna ports), which inevitably leads to antenna resource conflicts. Traditional TEG can only establish multiple links to illustrate communication opportunities in the time dimension, and cannot effectively model radio frequency / energy consumption resource conflicts caused by multi-domain resource mismatch.

[0005] (3) Traditional task scheduling methods (such as heuristic algorithms) have low energy efficiency and lack reasonable on-board preemption and recovery schemes for the differences between remote sensing and communication domain services, making it difficult to guarantee the overall network service success rate during sudden emergency tasks.

[0006] Therefore, this invention provides a telecom fusion satellite network service scheduling method based on supergraph and task hierarchy, which breaks through the limitations of traditional graph models, solves the problem of multi-domain resource mismatch, and achieves efficient and highly reliable on-board service scheduling. Summary of the Invention

[0007] The purpose of this invention is to provide a telecom fusion satellite network service scheduling method based on hypergraph and task hierarchy, which solves the limitations of existing graph models in multi-domain resource modeling and improves the scheduling success rate of the network in complex service scenarios.

[0008] The specific technical solution adopted by this invention is as follows: A telecommand-based integrated satellite network service scheduling method based on hypergraph and mission hierarchy includes the following steps: Step S1: Multi-domain resource modeling of remote communication satellite network.

[0009] A typical multi-domain satellite network geometric model including remote sensing satellite constellations, communication satellite constellations and ground stations is constructed. Inter-Satellite Link (ISL) and Satellite-Ground Link (SGL) models considering space transmission loss and rain attenuation are established, as well as a mission arrival model based on the Poisson process and a dynamic energy model combining basic satellite operations, data transmission and reception and solar periodic charging. Step S2: Construct a multidimensional extended hypergraph (MDEH) to extract spatiotemporal features.

[0010] Breaking through the limitations of traditional graph models, this paper constructs composite vertices in MDEH using the concept of "resource clustering" and builds a two-layer topology using hyperedges. This hypergraph covers the dimensions of the Earth's surface grid, the transmission mode or type of communication devices, and the number of time slots. Hyperedges represent the transformation of transmission modes between communicable devices on the same grid within the same time slot, thereby reducing redundant connections while modeling resource conflicts.

[0011] Step S3: Dynamic priority allocation of services based on real-time energy status.

[0012] The real-time energy status of satellite nodes (the sum of remaining energy and expected collected energy) is extracted as a weighting factor. By setting an energy status threshold and adjustment factor, the energy status is mapped to a priority correction value, and the original static service priority is dynamically updated, so that the priority of high-energy-consuming services or low-power nodes is adaptively adjusted.

[0013] Step S4: Calculate the queuing delay for multi-priority services.

[0014] Based on the preemption rules (high-priority services can preempt service resources at any time, and interrupted services can be resumed from the point of interruption), using Little's law and the compound Poisson process, we derive the expected queuing delays of the highest priority services and non-highest priority services after considering the preemption amplification effect.

[0015] Step S5: Execute the on-board service queue preemption-delay scheduling strategy (PDS). Arriving regular and emergency tasks are merged and sorted according to arrival time and dynamic priority. Then, a three-level allocation strategy is executed: priority is given to allocating idle antennas; if there are no idle resources, high-priority tasks are interrupted after comparing energy status to perform preemption of low-priority tasks; if the interrupted regular service has not exceeded the maximum allowable waiting time, it enters the delay allocation queue to wait for rearrangement and recovery.

[0016] The technical effects achieved by this invention are as follows: This invention overcomes the limitations of traditional Time-Spreading Graphs (TEGs), which restrict the number of resource types they can describe and struggle with resource conflict modeling, by constructing a Multidimensional Extended Hypergraph (MDEH). This MDEH efficiently describes antenna resource conflicts and transmission mode transitions, improving on-board storage efficiency. The proposed Priority Partitioning and Preemption-Delay Allocation Scheduling Strategy (PDS) based on Dynamic Energy Correction effectively improves the overall network service scheduling success rate while prioritizing reliable transmission of high-priority (urgent) services. Simulations show that the overall scheduling success rate remains above 91% even with regular tasks lasting up to 1200 seconds and sudden emergency tasks; for emergency tasks lasting 300 seconds, the success rate reaches over 95%, representing an overall performance improvement of approximately 20% compared to the traditional First-Come, First-Served (FCFS) conservative strategy. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the satellite network model architecture designed in this invention; Figure 2 This is a schematic diagram of the MDEH architecture designed in this invention; Figure 3 This is a diagram of the MDEH multidimensional extended architecture designed in this invention; Figure 4 This is a graph showing the impact of emergency tasks of different durations on the success rate of business scheduling. Figure 5 This is a graph showing the impact of emergency missions of different durations on the disruption of routine missions; Figure 6 This is a performance graph of the overall service scheduling interruption rate for different durations; Figure 7 This is a performance graph showing the overall business scheduling success rate for different durations; Figure 8 This is a statistical chart showing the success rate of subtasks for tasks of different priorities; Figure 9 This is a chart showing the success rate of parent tasks for tasks of different priorities. Figure 10 It is an end-to-end latency statistics chart for tasks of different priorities; Figure 11It is a statistical chart of the success rate and latency of subtasks and parent tasks under a certain load. Detailed Implementation

[0018] To make the objectives and advantages of this invention clearer, the invention will be specifically described below with reference to embodiments. It should be understood that the following text is merely used to describe one or more specific embodiments of the invention and does not strictly limit the scope of protection specifically claimed by the invention.

[0019] like Figure 1 As shown, a remote communication integrated satellite network service scheduling method based on hypergraph and task hierarchy includes the following steps: Step S1: Multi-domain resource modeling of remote communication satellite network; A typical multi-domain satellite network geometric model including remote sensing satellite constellations, communication satellite constellations and ground stations is constructed. Inter-Satellite Link (ISL) and Satellite-Ground Link (SGL) models considering space transmission loss and rain attenuation are established, as well as a mission arrival model based on the Poisson process and a dynamic energy model combining basic satellite operations, data transmission and reception and solar periodic charging. Step S2: Construct a multidimensional extended hypergraph (MDEH) to extract spatiotemporal features; Breaking through the limitations of traditional graph models, this paper constructs composite vertices in MDEH based on the concept of "resource clustering" and builds a two-layer topology with hyperedges. The hypergraph covers the dimensions of the Earth's surface grid, the transmission mode or type of communication equipment, and the number of time slots. The hyperedges represent the conversion of transmission modes between communicable devices on the same grid within the same time slot, thereby reducing redundant connections while modeling resource conflicts. Step S3: Dynamically prioritize services based on real-time energy status; The real-time energy status of satellite nodes is extracted as a weighting factor. The real-time energy status is the sum of the remaining energy and the expected collected energy. By setting an energy status threshold and adjustment factor, the energy status is mapped to a priority correction value, and the original static service priority is dynamically updated, so that the priority of high-energy-consuming services or low-power nodes is adaptively adjusted. Step S4: Calculate the queuing delay for multi-priority services; Based on the preemption rules (high-priority services can preempt service resources at any time, and interrupted services can be resumed from the point of interruption), the expected queuing delays of the highest priority services and non-highest priority services after considering the preemption amplification effect are derived using Little's law and the compound Poisson process. The preemption rules are as follows: high-priority services can preempt service resources at any time, and interrupted services can be resumed from the point of interruption. Step S5: Execute the on-board service queue preemption-delay scheduling strategy (PDS); Arriving routine and emergency tasks are merged and sorted according to arrival time and dynamic priority. Then, a three-level allocation strategy is executed: priority is given to allocating idle antennas; if there are no idle resources, high-priority tasks are interrupted and low-priority tasks are preempted after comparing energy status; if the interrupted routine service has not exceeded the maximum allowed waiting time, it enters the delayed allocation queue to wait for rearrangement and recovery.

[0020] Preferably, in step S1, a typical multi-domain satellite network model is first established, which consists of a remote sensing satellite constellation, a communication satellite constellation, and a set of ground stations. A schematic diagram of the network model architecture is shown below. Figure 1 As shown; based on this, the following sub-models are established: Link Model: Derive the achievable rates of Inter-Satellite Link (ISL) and Satellite-to-Ground Link (SGL); for Inter-Satellite Link (ISL), its time slot... The transmission performance is affected by parameters such as transmit antenna gain, receive antenna gain, constant transmit power, Boltzmann constant, noise temperature in the overall system, and free space transmission loss. ; in, and These represent the gains of the transmitting and receiving antennas, respectively. This represents a constant transmission power; furthermore, Represents the Boltzmann constant and The noise temperature represents the overall system noise level. It is the required signal-to-noise ratio. Represents the preset margin for inter-satellite links; Represents free space transmission loss. At the speed of light, Represents the center frequency of the frequency band. Represents time slot The slant distance between the two satellites; For inter-satellite links, the achievable rate is not only related to the satellite's transmission power and path loss, but also needs to take into account the propagation loss caused by external factors such as rain attenuation. ; in (Unit: W) represents the satellite in the time slot The transmission power relative to the corresponding ground station at that time; Indicates path loss; This represents propagation loss and is mainly used to describe the loss caused by external factors such as rain attenuation. Represents noise; Task arrival model: For file transfer services in the communication domain, the traffic arrival process is modeled as a Poisson process. The probability distribution of the number of tasks arriving within a time period satisfies a specific arrival rate; ; in Indicates data arrival rate. Represents a communication satellite within a time slot Number of tasks received.

[0021] Preferably, in step S2, a three-dimensional extended hypergraph MDEH is constructed; the three dimensions of the three-dimensional extended hypergraph MDEH are defined as follows: The number of grid cells on the Earth's surface represents the location mapping; The transmission mode or type of communication equipment; Number of time slots: Divide the task scheduling time into time slots of equal length; By constructing this two-layer topology, the hyperedge can effectively represent the transmission mode conversion between communicable devices on the same grid within the same time slot, thereby accurately characterizing the conflict situation of radio frequency resources such as antennas. by Represents a three-dimensional MDEH; where It represents the set of composite vertices. It represents the set of superedges. It is a finite-dimensional set. These represent various types of satellites and ground stations, as well as ISL and SGL; number of dimensions This represents the order of the MDEH; Define a composite vector It is by A set consisting of Cartesian products of MDEH of order; each composite vertex depends on elements of different orders; with This represents the dimensional list of MDEH. Wei; then: ; The corresponding architecture diagram is as follows Figure 2 As shown in the figure, Loc represents the location, that is, the current location of the satellite mapped to the grid position of the ground network; after the Earth's surface is divided into different grids, satellites in different domains may cover the same area at a certain moment with periodic movement; through the transmitted beams between satellites, the superedge of the mutual communication equipment of the beam range is formed. Figure 3The typical MDEH representation of the remote communication integrated satellite network is shown; the various dimensions can be explained as follows: (1) The first dimension is defined as the number of grids on the Earth's surface; (2) The second dimension can be represented as the transmission mode or type of communication equipment; (3) The third dimension can be represented as the number of time slots; It can be seen that different forms of hyperedges are composed of multiple dimensions; Assuming that the trajectory range of the satellite for a period of time is divided into three grids, marked as 1, 2, and 3; The eight double-arrow yellow solid lines in the figure represent the conversion of transmission modes between communication equipment on the same grid in the same time slot; Assuming that the task scheduling time is divided into equal-length time slots with a time slot interval of; The dashed arrows represent the possible transmissions between adjacent time slot grids.

[0022] Preferably, in step 3, the business priorities are initially divided into... Several levels; since satellites are dynamic nodes, the real-time energy state of the satellite is introduced as a weighting factor for dynamic adjustment: First, the satellite's energy state in the current time slot is calculated. relative energy state : ; in, This represents the satellite's remaining energy at the start of the time slot. The energy expected to be collected within this time slot based on orbit prediction. This represents the maximum capacity of the satellite's battery. Secondly, define the mapping factor. Map the energy state to a priority correction value: ; in, This represents the energy state threshold; if the value is below this threshold, the satellite enters "energy-saving mode". All belong to adjustment factors; when When the value is relatively high, the correction value is greater than 1, which has a positive gain on priority; when When the value is below the threshold, the correction value is less than 1, which has a reverse effect on the priority; the original static business priority will be adjusted. Change to dynamic priority : ; in, Energy consumption weights are inherent to the business itself; for high-energy-consuming businesses, the system adaptively adjusts their priority.

[0023] Preferably, in step S4, a preemption mechanism is adopted, allowing high-priority services to preempt service resources at any time, and interrupted services to resume from the point of interruption upon recovery; preemptive task distribution is performed on the satellite, and this process needs to ensure that: Mission arrival: Regarding the satellite mission arrival process, each priority level... Service arrival parameters The Poisson distribution process; the mission service time on the satellite has any distribution form and is not limited to the exponential distribution; Task priority: in the priority set Among these, the smaller the value, the higher the priority; and high-priority services can preempt service resources at any time; interrupted tasks can resume their original business from the point of interruption when they are restored. The queuing latency for tasks of different priorities is mainly divided into two categories: highest priority and non-highest priority, and the derivations are performed separately for each category: At the highest priority: Assuming the remaining service time is The satellite's mission queue contains There are 1 task; therefore, the waiting time is satisfied: ; Since the total task latency includes service time, the queuing latency for the highest priority task can be expressed as: ; Non-highest priority queuing delay For tasks that are not of the highest priority, define cumulative strength: ; In the process of satellite service delivery, the arrival of high-priority services constitutes a composite Poisson process; let... Priority for arrival during service period Given the number of services, the expected total service time is: ; Because tasks may be preempted, the actual service time will be amplified. Taking the expected value of the total service time yields: ; Therefore, the effective service and the excess waiting time are combined to obtain the final queuing delay for non-highest priority services: .

[0024] Preferably, in step S5, preemptive and delayed scheduling strategies (PDS) are performed based on the available resources on the satellite, including merging, sorting, and dynamic priority adjustment of arriving tasks, and allocating resources according to the existing available resources on the satellite. During the allocation process, resources are preempted based on priority and arrival time. Preempted services are considered interrupted services, and tasks arriving later will be allocated with delay. The PDS strategy, through a dynamic priority adjustment mechanism and changes in the rate of high-priority tasks, combined with a three-level allocation strategy—idle allocation, preemptive allocation, and delayed allocation—ensures that the overall service success rate is maximized when high-priority services arrive. Specifically, the process is as follows: Task merging and sorting: Calculate dynamic priorities for routine and emergency tasks, incorporate energy factors, and sort them in ascending order of arrival time; Resource allocation strategy: Prioritize allocating resources to idle antennas; when antenna resources are insufficient, perform energy verification, taking into account the energy status of the satellite where the preempted task is located. If the currently arriving task has a higher priority, the preemption is completed; if the current satellite is below a certain energy threshold, the priority is dynamically divided using the method in step 3, prioritizing the QoS performance of higher priority tasks. Delayed allocation: If the preempted service has not exceeded the maximum allowed waiting time, the queued task will be allocated later; the delay length and delay number are related to the actual visible window of the satellite; each task is set with a corresponding maximum delay number based on its priority, and only tasks exceeding the number will be added to the interruption queue and participate in the interruption count; the main purpose of this strategy is to ensure the excellent QoS service indicators of high-priority tasks while preventing other preempted tasks from experiencing a significant decline in QoS indicators due to task interruption and preemption.

[0025] In actual operation, this invention: Simulation verification was performed according to the method described in the specific implementation plan: Simulation conditions: The total simulation duration is set to 24 hours (86,400 seconds), the maximum preemption waiting time is 300 seconds, the number of low-priority (regular business) services is 1,500, and the duration of emergency services fluctuates between 60 and 1,200 seconds. Detailed parameters are shown in Table 1.

[0026]

[0027] The satellite network parameters adopt a network architecture of 3 observation satellites in solar orbit and 72 low-Earth orbit (LEO) satellites following the Walker constellation deployment. The orbital parameters are shown in Table 2.

[0028]

[0029] Figure 4 and Figure 5 This simulation analysis focuses on the basic scheduling concept of high-priority tasks preempting low-priority tasks. The total simulation duration is set to 24 hours, and the total number of regular tasks is 2000. To more realistically simulate the suddenness of emergency tasks, simulation experiments and results are designed considering the varying duration of emergency tasks (60-1200 seconds) and different numbers of emergency tasks.

[0030] Simulation curves show that the scheduling success rate remains above 91% even for regular tasks lasting up to 1200 seconds, while for emergency tasks lasting 300 seconds, the success rate is above 95%. The simulation further considers the impact of different emergency task packet lengths on service scheduling and regular task interruptions. Under these conditions, with exceptionally long and high emergency task arrival rates, the system's interruption rate for regular services is around 20%. This simulation only describes the number of regular services interrupted by emergency tasks; adding recovery and rescheduling mechanisms for interrupted services would further improve the service scheduling success rate performance.

[0031] Figures 6 to 7 This approach takes into account that in real-world scenarios, services generated on the OS not only have different priorities but also different times. Therefore, it's necessary to maintain a sorted task queue within the client-server (CS) that receives the services, and to allocate resources to each service when resources are sufficient on the CS. This task sorting optimization method is a two-dimensional sorting approach, designed to ensure the priority execution of urgent tasks and the rapid turnaround of short tasks. Simulation results are used to obtain... Figure 6 The simulation results show that the scheduling strategy used for comparison employs a First Come First Served (FCFS) delayed allocation strategy, strictly processing tasks according to their arrival order. This strategy provides a two-level allocation mechanism for tasks: the first level prioritizes allocating idle antennas, while the second level delays task allocation until an idle antenna or transmission resource becomes available. Under this strategy, the presence of priority attributes for tasks is largely irrelevant, as the CS execution logic itself is FCFS, and allocated tasks will not be interrupted.

[0032] It is clear that regarding the interruption rate of routine tasks, because the priority preemption strategy itself interrupts and delays the allocation of routine services when satellite resources are insufficient, it will inevitably cause the current system to delay or lose some routine services, thus causing the routine service interruption rate to rise along with the increase in the arrival rate of emergency tasks. However, analysis of the curves shows that the impact of traditional scheduling strategies and preemptive strategies on the interruption rate of routine tasks fluctuates by about 5%, and the increase in the interruption rate caused by the preemptive strategy is not unacceptable. More strongly, in terms of the overall service scheduling success rate, the priority strategy is superior to the conservative scheduling strategy as the emergency task arrival rate increases. Furthermore, the longer the task duration and the higher the emergency task arrival rate, the greater the difference in performance curves between the two strategies. The maximum performance difference is about 20%. Under short service conditions, there is no significant difference between the two in terms of either the routine service interruption rate or the overall service scheduling rate. This further demonstrates the feasibility and advancement of PDS, or priority preemption strategy.

[0033] To further illustrate the performance advantages and lower end-to-end latency of the PDS strategy for end-to-end transmission of tasks with different priorities, the same satellite orbital parameters as shown in Table 2 were used. A similar satellite network was constructed based on hypergraph theory, and the PDS strategy was applied to the task distribution process of different priorities within it, serving as the primary basis for task preemption, waiting, and forwarding.

[0034] Figures 8 to 9The paper presents the interruption rate and success rate curves of end-to-end transmission of main tasks arriving under different loads. These curves are derived by dividing the main task into multiple subtasks, assigning different priorities to each subtask, distributing them across the network, tracking the transmission process of these subtasks, and statistically analyzing the quantity and quality of received subtasks at the corresponding ground stations. It can be seen that when the network load is light, inter-satellite link bandwidth, satellite antennas, and buffer resources are relatively sufficient. At this time, task queuing is less prevalent (consistent with the performance of the PRM / G / 1 queuing theory model under low arrival rates). Therefore, all three algorithms have high success rates. However, as the task volume surges, severe resource conflicts occur in hotspots in the network (such as specific relay communication satellites or satellite nodes above ground stations). Different tasks on different transmission paths eventually collide and preempt each other based on their respective priorities. As the load increases, high-priority tasks maintain a low interruption rate and a high success rate. Meanwhile, preempted medium-priority and low-priority tasks wait in the queue due to insufficient inter-satellite relay resources, eventually leading to timeouts or excessive preemption attempts and forced interruptions. However, even low-priority tasks maintain a success rate of over 80%, demonstrating that the PDS strategy, while ensuring the transmission of high-priority tasks, also prevents preempted medium- and low-priority tasks from being completely abandoned by the network, exhibiting good robustness.

[0035] Figures 10 to 11 Under the same simulation framework and parameters, all subtask queues were counted at the ground station, combined into a parent task, and their completion rate and end-to-end latency curve were calculated. It can be seen that the end-to-end performance of the parent task remains at a high level despite increasing load, with high-priority tasks performing best, and low-priority tasks also within an acceptable range. Emergency tasks arrive in about 0.5 seconds, while tasks with high network load and low priority arrive in about 1.5 seconds, meeting the requirements for rapid task return in dense LEO (Low Orbit) communication resource scenarios. The simulation results in the figure also demonstrate that the PDS (Programmable Targeting System) strategy avoids deep queuing on a single link in terms of time. Although multipath transmission may result in more hops for some subtask paths, under high load, the queuing time saved by avoiding congestion far outweighs the propagation latency caused by detours. Therefore, the maximum latency of emergency remote sensing tasks, represented by the highest priority, remains relatively stable. Meanwhile, other secondary priority parent tasks can guarantee a basic completion rate and relatively fast end-to-end latency. The verification method can ensure reliable and fast data transmission of emergency remote sensing tasks, while preventing other tasks from experiencing large-scale data stagnation due to resource sacrifice, which would lead to excessive degradation of system performance.

[0036] The above description is merely a preferred embodiment of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention. Structures, devices, and operating methods not specifically described or explained in this invention are implemented according to conventional methods in the art unless otherwise specified or limited.

Claims

1. A telecommand-based integrated satellite network service scheduling method based on hypergraph and task hierarchy, characterized in that: Includes the following steps: Step S1: Multi-domain resource modeling of remote communication satellite network; A typical multi-domain satellite network geometric model including remote sensing satellite constellations, communication satellite constellations and ground stations is constructed. Inter-satellite link and satellite-to-ground link models considering space transmission loss and rain attenuation, a mission arrival model based on Poisson process, and a dynamic energy model combining basic satellite operations, data transmission and reception, and periodic solar charging are established. Step S2: Construct a multidimensional extended hypergraph to extract spatiotemporal features; The concept of "resource clustering" is used to construct composite vertices in MDEH, and a two-layer topology is constructed with hyperedges. This hypergraph covers the grid dimension of the Earth's surface, the transmission mode or type dimension of communication devices, and the number of time slots. The hyperedges represent the conversion of transmission modes between communicable devices on the same grid within the same time slot. Step S3: Dynamically prioritize services based on real-time energy status; The real-time energy status of satellite nodes is extracted as a weighting factor, and the real-time energy status is the sum of the remaining energy and the expected collected energy. By setting energy state thresholds and adjustment factors, energy state is mapped to priority correction values, and the original static service priorities are dynamically updated, so that the priorities of high-energy-consuming services or low-power nodes are adaptively adjusted. Step S4: Calculate the queuing delay for multi-priority services; Based on the preemption rules, using Little's law and the compound Poisson process, we derive the expected queuing delays of the highest priority and non-highest priority services after considering the preemption amplification effect. The preemption rules are as follows: high-priority services can preempt service resources at any time, and interrupted services can be resumed from the point of interruption. Step S5: Execute the on-board service queue preemption-delay scheduling strategy; Arriving routine and emergency tasks are merged and sorted according to arrival time and dynamic priority. Then, a three-level allocation strategy is executed: priority is given to allocating idle antennas; if there are no idle resources, high-priority tasks are interrupted and low-priority tasks are preempted after comparing energy status; if the interrupted routine service has not exceeded the maximum allowed waiting time, it enters the delayed allocation queue to wait for rearrangement and recovery.

2. The telecommand scheduling method for remote communication integrated satellite networks based on supergraph and task hierarchy as described in claim 1, characterized in that: In step S1, a typical multi-domain satellite network model is first established, which consists of a remote sensing satellite constellation, a communication satellite constellation, and a set of ground stations. Based on this, the following sub-models are established: Link Model: Deriving the achievable rates of inter-satellite links and satellite-to-ground links; for inter-satellite links, their time slots... The transmission performance is affected by parameters such as transmit antenna gain, receive antenna gain, constant transmit power, Boltzmann constant, noise temperature in the overall system, and free space transmission loss. ； in, and These represent the gains of the transmitting and receiving antennas, respectively. This represents a constant transmission power; furthermore, Represents the Boltzmann constant and The noise temperature represents the overall system noise level. It is the required signal-to-noise ratio. Represents the preset margin for inter-satellite links; Represents free space transmission loss. At the speed of light, Represents the center frequency of the frequency band. Represents time slot The slant distance between the two satellites; For inter-satellite links, the achievable rate is not only related to the satellite's transmission power and path loss, but also needs to take into account the propagation loss caused by external factors such as rain attenuation. ； in (Unit: W) represents the satellite in the time slot The transmission power relative to the corresponding ground station at that time; Indicates path loss; This represents propagation loss and is mainly used to describe the loss caused by external factors such as rain attenuation. Represents noise; Task arrival model: For file transfer services in the communication domain, the traffic arrival process is modeled as a Poisson process. The probability distribution of the number of tasks arriving within a time period satisfies a specific arrival rate; ； in Indicates data arrival rate. Represents a communication satellite within a time slot Number of tasks received.

3. The telecommand scheduling method for remote communication fusion satellite networks based on supergraph and task hierarchy as described in claim 2, characterized in that: In step S2, a three-dimensional extended hypergraph MDEH is constructed; the three dimensions of the three-dimensional extended hypergraph MDEH are defined as follows: The number of grid cells on the Earth's surface represents the location mapping; The transmission mode or type of communication equipment; Number of time slots: Divide the task scheduling time into time slots of equal length; by Represents a three-dimensional MDEH; where It represents the set of composite vertices. It represents the set of superedges. It is a finite-dimensional set. These represent various types of satellites and ground stations, as well as ISL and SGL; number of dimensions This represents the order of the MDEH; Define a composite vector It is by A set consisting of Cartesian products of MDEH of order; each composite vertex depends on elements of different orders; with This represents the dimensional list of MDEH. Wei; then: ； After dividing the Earth's surface into different grids, satellites from different domains may periodically cover the same area at a certain moment; the transmitted beams between satellites form the super-edge of mutual communication equipment within the beam range.

4. The telecommand scheduling method for remote communication fusion satellite networks based on supergraph and task hierarchy as described in claim 3, characterized in that: In step 3, the business priorities are initially divided into: Several levels; since satellites are dynamic nodes, the real-time energy state of the satellite is introduced as a weighting factor for dynamic adjustment: First, the satellite's energy state in the current time slot is calculated. relative energy state : ； in, This represents the satellite's remaining energy at the start of the time slot. The energy expected to be collected within this time slot based on orbit prediction. This represents the maximum capacity of the satellite's battery. Secondly, define the mapping factor. Map the energy state to a priority correction value: ； in, This represents the energy state threshold; if the value is below this threshold, the satellite enters "energy-saving mode". All belong to adjustment factors; when When the value is relatively high, the correction value is greater than 1, which has a positive gain on priority; when When the value is below the threshold, the correction value is less than 1, which has a reverse effect on the priority; the original static business priority will be adjusted. Change to dynamic priority : ； in, Energy consumption weights are inherent to the business itself; for high-energy-consuming businesses, the system adaptively adjusts their priority.

5. The telecommand scheduling method for remote communication fusion satellite networks based on supergraph and task hierarchy as described in claim 4, characterized in that: In step S4, a preemptive mechanism is adopted, allowing high-priority services to preempt service resources at any time, and interrupted services can resume from the point of interruption upon recovery. Preemptive task distribution is performed on the satellite, and this process needs to ensure that: Mission arrival: Regarding the satellite mission arrival process, each priority level... Service arrival parameters The Poisson distribution process; the mission service time on the satellite has any distribution form and is not limited to the exponential distribution; Task priority: in the priority set Among these, the smaller the value, the higher the priority; and high-priority services can preempt service resources at any time; interrupted tasks can resume their original business from the point of interruption when they are restored. The queuing latency for tasks of different priorities is mainly divided into two categories: highest priority and non-highest priority, and the derivations are performed separately for each category: At the highest priority: Assuming the remaining service time is The satellite's mission queue contains There are 1 task; therefore, the waiting time is satisfied: ； in, Indicates the expected waiting time for tasks in the queue; This indicates that the first [item] in the queue... The random service time of each task, that is, the year in which Satellite World completes its first task. The specific time required to send each data packet; Since the total task latency includes service time, the queuing latency for the highest priority task can be expressed as: ； in, This represents the expected total dwell time of the highest priority task in the system. It refers to the average service rate of the system, that is, the throughput capacity of the satellite nodes, which is the number of tasks that the system can process per unit of time. Specifically refers to satellites For the highest priority service rate; This reflects the busy level or resource utilization of the satellite communication link or processor. Similarly... Representative satellite Load status for the highest priority tasks; The second moment representing the service time of the highest priority task; Non-highest priority queuing delay For tasks that are not of the highest priority, define cumulative strength: ； In the process of satellite service delivery, the arrival of high-priority services constitutes a composite Poisson process; let... Priority for arrival during service period Given the number of services, the expected total service time is: ； in and All values ​​represent priority indexes; the smaller the value, the higher the priority. This represents the priority of the current assessment target task. Represents higher than Other priorities; This represents the cumulative business intensity, reflecting all services in the system that are at or above the current level. The proportion of the total bandwidth or total computing resources of the satellite system used for the mission; For the first The ideal service time for a mission of this level; For the first A higher priority arrival The service time for the task; Because tasks may be preempted, the actual service time will be amplified. Taking the expected value of the total service time yields: ； in and Indicates the first The effective (actual) service time and expected value of a mission; due to the preemption mechanism, regular missions will be frequently interrupted; the effective service time refers to the time consumed by the regular mission itself plus the total time that is forced to stop due to being preempted by emergency missions, that is, the actual time consumed by the mission on the star after being amplified due to interruption. Therefore, the effective service and the excess waiting time are combined to obtain the final queuing delay for non-highest priority services: ； This represents the first The expected total queuing delay for priority tasks.

6. The telecommand scheduling method for remote communication fusion satellite networks based on supergraph and task hierarchy as described in claim 5, characterized in that: In step S5, preemption-delay allocation scheduling is performed based on the available resources on the satellite, including merging, sorting, and dynamic priority adjustment of arriving tasks, and allocation is carried out according to the existing available resources on the satellite. During the allocation process, resources are preempted based on priority and arrival time. Preempted services are considered interrupted services, and tasks arriving later will be allocated with delay. The PDS strategy, through a dynamic priority adjustment mechanism and changes in the high-priority task rate, combined with a three-level allocation strategy—idle allocation, preemption allocation, and delayed allocation—ensures that the overall service success rate is maximized when high-priority services arrive. Specifically, the process is as follows: Task merging and sorting: Calculate dynamic priorities for routine and emergency tasks, incorporate energy factors, and sort them in ascending order of arrival time; Resource allocation strategy: Prioritize allocating resources to idle antennas; when antenna resources are insufficient, perform energy verification, taking into account the energy status of the satellite where the preempted task is located. If the currently arriving task has a higher priority, the preemption is completed; if the current satellite is below a certain energy threshold, the priority is dynamically divided using the method in step 3, prioritizing the QoS performance of higher priority tasks. Delayed allocation: If the preempted service has not exceeded the maximum allowed waiting time, the queued task will be allocated later; the delay length and delay count are related to the actual visible window of the satellite; each task is set with a corresponding maximum delay count based on its priority, and only tasks that exceed the maximum delay count will be added to the interrupt queue and participate in the interrupt count.

Images