Giant Constellation Operation and Maintenance and Resource Control System and Method for Computing-Networking Convergence

Abstract

The present invention discloses a giant constellation operation and maintenance and resource control system and method for computing-network convergence. The system includes: a number of MEO satellites and a number of LEO satellites. Each MEO satellite manages multiple LEO satellites in each time slot of each scheduling period. Each LEO satellite is used to obtain data of computing tasks in a preset time period, and perform data calculation according to the resource scheduling information for the time slot sent by the MEO satellite in each time slot of each scheduling period in the preset time period. Each MEO satellite is used to generate a network by using the trained resource scheduling information, generate the resource scheduling information for the time slot according to its own resource task status in the time slot and send it to each LEO satellite managed in the time slot, and perform collaborative data calculation of the managed LEO satellites according to the resource scheduling information for the time slot. The present invention has higher real-time performance, better resource control and scheduling capabilities, and a higher task completion rate.

Description

Technical Field

[0001] The present invention belongs to the technical field of satellite communication, and particularly relates to a giant constellation operation and maintenance and resource management and control system and method for computing-network convergence. Background Art

[0002] A giant constellation is a satellite network system composed of hundreds to tens of thousands of satellites, which can provide all-time and all-space continuous communication services. A hybrid multi-layer giant constellation including Low Earth Orbit (LEO) and Middle Earth Orbit (MEO) satellites, etc., has gradually become the first choice for coping with complex mission requirements due to its advantages such as wide coverage, fast response speed, and low communication latency. With the explosive growth of mission requirements, the drawbacks of the mode of satellite transmitting data back to the ground data center for processing have gradually emerged. On-orbit computing can directly perform data processing on the satellite and calculate data from sensors or other devices without transmitting the data back to the ground, reducing communication latency, improving response speed, and effectively saving bandwidth and ground processing resources. Currently, it has become a trend to actively launch and deploy high-computing-power satellites to provide powerful data processing and real-time decision-making capabilities for multiple fields.

[0003] Therefore, it is urgent to propose a new satellite operation and maintenance management and control architecture, which coordinates the computing power resources of multiple satellites through inter-satellite links to form an intelligent operation and maintenance management and control mechanism with deep integration of computing power and network. Give full play to the synergistic effect of the computing and communication resources of the giant constellation, realize more flexible and efficient task scheduling and resource allocation, support more complex space applications, and promote the development of satellite networks towards a more autonomous and intelligent direction. However, this multi-layer giant constellation structure also brings huge challenges to operation and maintenance and resource management and control. The current satellite network operation and maintenance and resource management and control methods mainly rely on satellite independent computing or only schedule and plan for static time slots. However, in the giant constellation network, the efficient allocation and scheduling of computing, communication, and energy resources between different orbit layers bring huge challenges. First, how to coordinately allocate computing and communication resources between satellites at different levels to improve task processing efficiency is an important issue faced by operation and maintenance and resource management and control; second, the tasks and resources in the satellite network are highly dynamic and uncertain, and the task load, resource availability, and communication link status change over time. Traditional static management and control strategies cannot effectively cope with these changes. Therefore, how to introduce the idea of computing-network coordination into the giant constellation network and redesign the operation and maintenance and network management and control architecture and methods has become an important development direction for future giant constellations.

[0004] Currently, there have been some studies on the operation and maintenance and resource management and control architecture and methods for giant constellations. However, these studies have the following problems:

[0005] Current architectures and scheduling methods are often limited to fixed or predictable tasks and cannot handle dynamic and complex task requirements; current methods are limited to collaborative computing by transmitting a large amount of actual task data between satellites, which will occupy a large amount of communication bandwidth and result in high latency and task response time; at the same time, existing satellite computing and communication collaborative methods also rely on the management and operation and maintenance of ground control stations; therefore, higher real-time performance and better resource control and scheduling capabilities cannot be obtained. Summary of the Invention

[0006] To solve the above problems existing in the prior art, the present invention provides a giant constellation operation and maintenance and resource control system and method for computing-network convergence.

[0007] The technical problems to be solved by the present invention are realized through the following technical solutions:

[0008] The present invention provides a giant constellation operation and maintenance and resource control system for computing-network convergence, including: a MEO satellite collaborative-level control architecture with MEO satellites and a LEO satellite edge-level control architecture with LEO satellites. The MEO satellites are connected to each other and the LEO satellites are connected to each other through communication links. Each MEO satellite manages multiple LEO satellites in each time slot of each scheduling cycle.

[0009] Each LEO satellite is used to obtain data of computing tasks for a preset time period, and within each time slot of each scheduling cycle of the preset time period, according to the resource scheduling information sent by the MEO satellite for the time slot execute the data calculation of the computing tasks.

[0010] Each MEO satellite is used to obtain its own resource task status in the time slot , adopt a trained resource scheduling information generation network, generate resource scheduling information for the time slot according to its own resource task status in the time slot and send it to each LEO satellite managed in the time slot , and execute collaborative data calculation of the computing tasks of each LEO satellite managed according to the resource scheduling information for the time slot ; the resource scheduling information for the time slot represents the proportion of the data volume calculated by the MEO satellite in the computing tasks of each LEO satellite managed in the time slot to the total data volume of the computing tasks of the LEO satellite; the value range of the resource scheduling information is from 0 to 1.

[0011] The present invention also provides a method for operation and maintenance and resource management and control of a giant constellation for computing-network convergence, which is applied to each MEO satellite in the above-mentioned system for operation and maintenance and resource management and control of a giant constellation for computing-network convergence. The method includes:

[0012] Obtain the resource task status of itself in the current time slot of the current scheduling period;

[0013] Use the trained resource scheduling information generation network to generate resource scheduling information for the current time slot according to the resource task status of itself in the current time slot. The trained resource scheduling information generation network is trained by using local reinforcement learning and federated learning methods. The resource scheduling information for the current time slot represents the proportion of the data volume calculated by the MEO satellite in the computing tasks of each LEO satellite managed in the current time slot to the total data volume of the computing tasks of the LEO satellite;

[0014] Send the resource scheduling information for the current time slot to each LEO satellite managed in the current time slot;

[0015] According to the resource scheduling information for the current time slot, determine whether to perform data calculation on the computing tasks of each LEO satellite managed in the current time slot, and the data volume that needs to be calculated when performing data calculation;

[0016] When it is necessary to perform data calculation on the computing tasks of each LEO satellite managed in the current time slot, calculate the corresponding data volume of the computing tasks of each LEO satellite managed in the current time slot according to the data volume that needs to be calculated.

[0017] Compared with the prior art, the beneficial effects of the present invention:

[0018] The system proposed by the present invention includes two - level architectures: the MEO satellite collaboration level and the LEO satellite edge level. Each MEO satellite in the MEO satellite collaboration - level architecture can obtain its own real - time resource - task status, generate real - time resource - scheduling information based on its own real - time resource - task status, and according to the real - time resource - scheduling information, perform collaborative data calculations for the computing tasks of each managed LEO satellite, and enable each managed LEO satellite to perform data calculations for computing tasks according to the real - time resource - scheduling information. Therefore, through the MEO satellite collaboration - level architecture, real - time allocation of computing tasks for the overall satellite network and comprehensive utilization of inter - layer resources can be carried out. Through the LEO satellite edge - level architecture, real - time computing of edge nodes and real - time collaborative execution of communication tasks can be carried out. Thus, without relying on the control of ground controllers, dynamic scheduling and efficient utilization of resources can be carried out according to real - time task requirements, not limited to fixed tasks or predictable tasks. Moreover, there is no problem of high latency and task response time caused by transmitting a large amount of task data between different MEOs for collaborative computing between MEO satellites. It can achieve efficient computing scheduling for dynamic and complex task requirements, improve the task completion volume of the overall satellite network, and not only has higher real - time performance and better resource control and scheduling capabilities, but also has a higher task completion rate.

[0019] The following will further elaborate on the present invention in conjunction with the accompanying drawings and specific embodiments. Brief Description of the Drawings

[0020] Figure 1 is a schematic diagram of an architecture of a giant constellation operation and maintenance and resource control system for computing - network convergence provided by an embodiment of the present invention;

[0021] Figure 2 is another schematic diagram of an architecture of a giant constellation operation and maintenance and resource control system for computing - network convergence provided by an embodiment of the present invention;

[0022] Figure 3 is a schematic flowchart of a method for giant constellation operation and maintenance and resource control for computing - network convergence provided by an embodiment of the present invention;

[0023] Figure 4 is a schematic diagram of task completion volume obtained through simulation experiments provided by an embodiment of the present invention;

[0024] Figure 5 is a graph of energy consumption results obtained through simulation experiments provided by an embodiment of the present invention. Detailed Embodiments

[0025] The following further describes the present invention in detail with specific embodiments, but the embodiments of the present invention are not limited thereto.

[0026] The present invention provides a giant constellation operation and maintenance and resource management and control system for computing-network convergence. The system includes: an MEO satellite collaborative-level management and control architecture with MEO satellites and a LEO satellite edge-level management and control architecture with LEO satellites, and both being positive integers greater than 1. The MEO satellites are connected to each other and the LEO satellites are connected to each other through communication links. Each MEO satellite manages multiple LEO satellites within each time slot of each scheduling period. The LEO satellites managed by each MEO satellite within different time slots may be different or the same. The number of LEO satellites managed by different MEO satellites within different time slots may be the same or different.

[0027] Each LEO satellite is used to obtain data of a computing task for a preset time period and, within each time slot of each scheduling period of the preset time period, perform data calculation of the computing task according to the resource scheduling information sent by the MEO satellite for the time slot . Each MEO satellite is used to obtain its own resource task status for the time slot , generate a network using the trained resource scheduling information, generate resource scheduling information for the time slot according to its own resource task status for the time slot and send it to each LEO satellite managed within the time slot , and perform collaborative data calculation of the computing tasks of each LEO satellite managed according to the resource scheduling information for the time slot ; the resource scheduling information for the time slot represents the proportion of the data volume calculated by the MEO satellite in the total data volume of the computing tasks of each LEO satellite managed within the time slot . Here, each LEO satellite collects data from the outside according to the computing task to obtain data of the corresponding computing task. For example, by performing ecological monitoring on the surrounding environment, data of the ecological monitoring task is obtained; by performing meteorological observation on the surrounding environment, data of the meteorological monitoring task is obtained; and by acquiring remote sensing images of the surrounding environment, data of the computing task related to the remote sensing image is obtained, and so on. Each computing task corresponds to a preset time period to indicate that the computing of the computing task needs to be performed within this time period and the computing of the computing task is completed. Exemplarily, Figure 1 is a schematic diagram of the architecture of a giant constellation operation and maintenance and resource management and control system for computing-network convergence provided by an embodiment of the present invention, as shown in Figure 3As shown, the system includes three MEO satellites. There are communication links between the MEO satellites, and communication links between the MEO satellites and the LEO satellites. The area formed by the dotted lines below each MEO satellite represents the signal radiation range of that MEO satellite. Figure 3 The semi - circle in Figure 3 represents the Earth, and the arrow pointing from the Earth to the LEO represents obtaining ground data from the LEO. Each MEO satellite manages multiple LEO satellites within each time slot. Each LEO satellite can obtain data for computing tasks from the outside world. For example, it can obtain task data by receiving data sent from devices such as ships and cars on the ground.

[0028] Here, the value range of the resource scheduling information is from 0 to 1. Thus, according to the resource scheduling information of each time slot, the MEO satellite and the LEO satellite can determine whether to perform data calculation for the computing task in that time slot, and when data calculation for the computing task is required, the amount of data to be calculated; among them, when the value range of the resource scheduling information of a certain time slot is 0, it means that in that time slot, the MEO satellite does not need to undertake any data calculation for all the data amounts required to be calculated in the computing tasks of each LEO satellite it manages, while each LEO satellite managed by the MEO satellite needs to undertake data calculation for all the data amounts required to be calculated in its own computing task in that time slot; when the value range of the resource scheduling information is 1, it means that the MEO satellite needs to perform data calculation for all the data amounts required to be calculated in the computing tasks of each LEO satellite it manages in that time slot, while each LEO satellite managed by the MEO satellite does not need to undertake any data calculation for all the data amounts required to be calculated in its own computing task in that time slot; when the value of the resource scheduling information of a certain time slot is 0.4, it means that in that time slot, the MEO satellite needs to undertake 40% of the data calculation for all the data amounts required to be calculated in the computing tasks of each LEO satellite it manages, while each LEO satellite managed by the MEO satellite needs to undertake 60% of the data calculation for all the data amounts required to be calculated in its own computing task in that time slot. Exemplarily, the value of the resource scheduling information can be one of {0, 0.2, 0.4, 0.6, 0.8, 1}. Based on this, each LEO satellite is specifically used to determine whether to perform data calculation for the computing task in the time slot according to the resource scheduling information sent by the MEO satellite for the time slot and the amount of data to be calculated when performing data calculation, and calculate the corresponding amount of data for the computing task in the time slot according to the amount of data to be calculated. Each MEO satellite is specifically used to determine in the time slot according to the resource scheduling information for the time slot Whether data calculations for the computing tasks of each managed LEO satellite are performed, and the amount of data to be calculated when performing data calculations. According to the amount of data to be calculated, in a time slot the corresponding data volume of the computing tasks of each managed LEO satellite is calculated.

[0029] Figure 2 is another schematic diagram of the architecture of the giant constellation operation and maintenance and resource management and control system for computing-network convergence provided by the present invention. As Figure 2 shown, each MEO satellite includes: a resource task status monitoring module, a data calculation and processing module, and a computing-network convergence scheduling and optimization module. The resource task status monitoring module is used to obtain the resource usage and computing task execution status of the MEO satellite itself and each managed LEO satellite in each time slot of each scheduling cycle, and obtain the resource task status of the MEO satellite in the time slot The resource usage includes the remaining power and communication resources; the computing task execution status includes: whether the execution of the computing task is in progress, and the remaining amount of data to be calculated, etc. In the present invention, the resource task status of each MEO satellite in the time slot includes: the energy status of the MEO satellite in the time slot the link status and the on-board task data status of the MEO satellite in the time slot ; among them, the energy status of the MEO satellite in the time slot represents: the remaining power of the MEO satellite in the time slot ; the link status of the MEO satellite in the time slot represents: the total link capacity on the communication link between the MEO satellite and the managed LEO satellites in the time slot ; the on-board task data status of the MEO satellite in the time slot represents: the total amount of data of the computing tasks of the LEO satellites managed by the MEO satellite in the time slot ; the on-board task data status of the MEO satellite in the time slot represents: the total amount of data of the computing tasks of the LEO satellites managed by the MEO satellite in the time slot

[0030] The data calculation and processing module is used to receive the data of the computing tasks of each managed LEO satellite when data calculations for the computing tasks of each managed LEO satellite need to be performed in the time slot and, according to the amount of data to be calculated, in the time slot Calculate the corresponding data volume of the computing tasks of each LEO satellite managed therein. It should be noted that when the MEO satellite needs to calculate the data of the computing tasks of each LEO satellite managed in a certain time slot, it obtains the corresponding data volume from each LEO satellite managed and performs the calculation. The computing-network fusion scheduling optimization module is used to train a resource scheduling information generation network by using local reinforcement learning and federated learning methods, obtain a trained resource scheduling information generation network, and share and aggregate the network parameters of the resource scheduling information generation network with the remaining MEO satellites during the process of training the resource scheduling information generation network, and, by using the trained resource scheduling information generation network, generate resource scheduling information for the time slot according to the resource task status of the MEO satellite itself in the time slot and send the resource scheduling information for the time slot to each LEO satellite managed in the time slot . In some embodiments, continuing to refer to the above Figure 2 , the computing-network fusion scheduling optimization module may include an intelligent learning sub-module and a cooperative sharing sub-module. The intelligent learning sub-module is used to train a resource scheduling information generation network by using local reinforcement learning and federated learning methods, obtain a trained resource scheduling information generation network, and, by using the trained resource scheduling information generation network, generate resource scheduling information for the time slot according to the resource task status of the MEO satellite in the time slot . The cooperative sharing sub-module is used to share and aggregate the network parameters of the resource scheduling information generation network with the remaining MEO satellites during the process of training the resource scheduling information generation network, and, send the resource scheduling information for the time slot to each LEO satellite managed in the time slot .

[0031] In the present invention, as Figure 2 shown, each LEO satellite includes: a data acquisition module, a resource task status monitoring module, and a data calculation and processing module. The data acquisition module is used to acquire task data in the environment where the LEO satellite itself is located. The resource task status monitoring module is used to obtain the computing resources, communication resources, and computing task execution status of the LEO satellite itself in real time. The data calculation and processing module is used to process the acquired data and execute edge computing tasks. On this basis, in some embodiments, when each LEO satellite among multiple LEO satellites managed by the same MEO satellite performs corresponding data calculations according to real-time resource scheduling information, it can perform collaborative calculations with the remaining LEO satellites among the multiple LEO satellites managed by the same MEO satellite by using existing algorithms. Based on this, as Figure 2As shown, each LEO satellite further includes a performance optimization module, which is used to obtain the operating parameters of the LEO satellite itself and the resource scheduling strategy for the remaining LEO satellites by using the existing LEO satellite cooperative scheduling method according to the status of the computing resources and communication resources of the LEO satellite itself and the status of computing task execution, so as to improve the system performance. It should be noted that the operating parameters of the LEO satellite itself and the resource scheduling strategy for the remaining LEO satellites need to be determined according to the specific existing algorithms adopted by the LEO satellite, and the present invention does not limit this. Based on this, on the basis of introducing the computing and communication cooperation mechanism and the strategy parameter cooperation mechanism between MEO satellites, the present invention can further improve the task scheduling ability of the entire satellite network by introducing the cooperative scheduling and control mechanism of inter-layer satellites.

[0032] In the present invention, the trained resource scheduling information generation network is trained by local reinforcement learning and federated learning methods under the condition that the optimization objective is to minimize the energy consumption of the system in a scheduling period, and the constraint conditions of the optimization objective are the energy constraint conditions of each LEO satellite, the energy constraint conditions of each MEO satellite, the computing constraint conditions of each LEO satellite, the computing constraint conditions of each MEO satellite, and the communication constraint conditions of each LEO satellite. Exemplarily, the resource scheduling information generation network is an Actor network.

[0033] Exemplarily, the expressions of the energy constraint conditions of each LEO satellite and the energy constraint conditions of each MEO satellite are as follows:

[0034] ;

[0035] ;

[0036] Among them, , , , is the LEO satellite set composed of LEO satellites in the system, is the th LEO satellite in the LEO satellite set, takes values of , is each time slot in a scheduling period , is the MEO satellite set composed of MEO satellites in the system, is the th MEO satellite in the MEO satellite set, takes values of , is the The maximum battery capacity of a LEO satellite, which is the percentage of the maximum depth of discharge of the LEO satellite. The percentage of the maximum depth of discharge describes the lowest voltage level to which the battery can be safely discharged during use. That is to say, when the maximum depth of discharge is exceeded, it may lead to a decrease in battery capacity or performance degradation, is the amount of electricity already available for the th LEO satellite in each time slot, is the amount of electricity charged for the th LEO satellite in each time slot, is the amount of transmission energy consumption of the th LEO satellite in each time slot, is the amount of computing energy consumption of the th LEO satellite in each time slot; is the amount of receiving energy consumption of the The calculated energy consumption of each MEO satellite in each time slot. It should be noted that the amount of electricity charged to each satellite in each time slot in this system, as well as the link capacity between satellites in this system, need to be obtained in advance. Specifically, it can be obtained by pre-obtaining the satellite link capacity matrix and satellite energy acquisition matrix of this system during the required time period. The satellite link capacity matrix includes the link capacity matrix between all LEO satellites and MEO satellites in each time slot of each scheduling period within this system. Moreover, each element in the link capacity matrix represents the link capacity between a certain LEO and a certain MEO in this time slot, or the link capacity between two certain MEOs. A value of 0 in the link capacity matrix indicates that the corresponding satellites are not visible to each other, while other values greater than 0 in the link capacity matrix represent the magnitude of the link capacity between the corresponding satellites. The satellite energy acquisition matrix includes: LEO satellite energy acquisition matrix and MEO satellite energy acquisition matrix. The LEO satellite energy acquisition matrix includes the dynamic energy acquisition matrix of LEO satellites in each time slot of each scheduling period within this system. The MEO satellite energy acquisition matrix includes the dynamic energy acquisition matrix of MEO satellites in each time slot of each scheduling period within this system. Each element in the dynamic energy acquisition matrix of LEO satellites represents the amount of charge filled by a corresponding LEO satellite in this time slot. When a certain element is 0, it means that the corresponding LEO satellite is covered by the earth's shadow in this time slot and cannot obtain solar energy. The same applies to the dynamic energy acquisition matrix of MEO satellites.

[0037] Exemplarily, the expression of the calculation constraint condition for each MEO satellite is as follows:

[0038] ;

[0039] Wherein, is the amount of data processed by the th MEO satellite in each time slot , represents the maximum calculation limit of the MEO satellite.

[0040] Exemplarily, the expression of the calculation constraint condition for each LEO satellite is as follows:

[0041] ;

[0042] Wherein, is the amount of data processed by the th LEO satellite in each time slot , represents the maximum calculation limit of the LEO satellite.

[0043] Exemplarily, the expression of the communication constraint condition for each LEO satellite is as follows:

[0044] ;

[0045] Among them, is the th LEO satellite in the time slot managed by a MEO satellite, is the time slot from to link capacity on the link, is the time slot from to amount of data transmitted.

[0046] Exemplarily, the expression of the energy consumption of the system is as follows:

[0047] ;

[0048] ;

[0049] ;

[0050] ;

[0051] ;

[0052] Among them, is the set of scheduling strategies obtained by all MEO satellites through the resource scheduling information generation network, where the scheduling strategy is the network parameter of the resource scheduling information generation network, is the total energy consumption of the system in a scheduling period, represents the minimum total energy consumption of the system in a scheduling period, is the resource task status of the satellite in each time slot , represents the set of scheduling strategies of all MEO satellites in the state, is the th transmission power of the LEO satellite, is the time slot length of the time slot , is the th hardware calculation coefficient of the LEO satellite, is the th calculation rate of the LEO satellite in each time slot ; is the th receiving power of the MEO satellite, is in the One MEO satellite in a time slot Managed and sent to the th MEO satellite A LEO satellite that transmits data, Is the time slot From To The link capacity on the link, Is in the time slot From To The amount of data transmitted, For the th MEO satellite The hardware calculation coefficient, For the th MEO satellite In each time slot The calculation rate.

[0053] The present invention takes into account the dynamic fluctuations in energy acquisition caused by the periodic switching of satellites between the earth's shadow area and the direct sunlight area. According to these actual fluctuation characteristics, the dynamic consumption and acquisition of energy corresponding to satellites in the process of computing-network collaboration are modeled. Based on the modeling of the dynamic consumption and acquisition of energy corresponding to satellites in the process of computing-network collaboration, the optimization objective and the constraint conditions of the optimization objective for generating the training resource scheduling information network are constructed, making the present invention more efficient and more in line with the real application environment, so that the high-energy-consuming computing tasks in the actual satellite network scenario can be efficiently completed, avoiding the situation of performance degradation or service interruption when performing on-board computing, which is a high-energy-consuming service.

[0054] In some embodiments, the trained resource scheduling information generation network deployed on each MEO satellite is obtained by training using the methods described in S101~S106 below. The following will take MEO satellite as an example to illustrate the steps of this training method:

[0055] S101. At the th time slot in the th round of training, obtain the resource task status of MEO satellite at the th time slot; wherein, each round of training corresponds to a scheduling period, each scheduling period has time slots, and training is performed every preset update step time slots; when the resource task status of MEO satellite at the th time slot is the preset status. ​

[0056] It should be noted that within each scheduling period, training is performed once every preset update step number of time slots. is a positive integer, and the value of , is a preset integer. At the first time slot in the first round of training, the obtained MEO satellite The resource task status at the first time slot is a preset status, and the preset status can be set according to actual needs, which is not limited in the present invention.

[0057] S102. Generate a network based on the resource scheduling information obtained in the th time slot and the resource task status of the th time slot , and determine the resource scheduling information of the th time slot ; where, when , the network generated from the resource scheduling information obtained in the th time slot is the network generated from the resource scheduling information obtained in the last training in the nd round of training.

[0058] Specifically, input the resource task status of the th time slot into the network generated from the resource scheduling information of the th time slot, and the network generated from the resource scheduling information of the th time slot outputs the resource scheduling information of the th time slot .

[0059] S103. Based on the resource scheduling information of the th time slot , determine the reward of the MEO satellite at the th time slot and the resource task status of the th time slot .

[0060] Specifically, the MEO satellite performs corresponding data calculations according to the resource scheduling information of the th time slot , and then obtains the reward of the MEO satellite at the th time slot , as well as the resource task status of the MEO satellite at the th time slot , where , is a MEO satellite in the th time slot and the total energy consumption of the local system composed of LEO satellites managed in the th time slot, is the receiving energy consumption of the MEO satellite in the th time slot, is the computing energy consumption of the MEO satellite in the th time slot, is the set of LEO satellites managed by the MEO satellite in the th time slot, is the transmitting energy consumption of each LEO satellite managed by the MEO satellite in the th time slot in the th time slot, is the computing energy consumption of each LEO satellite managed by the MEO satellite in the th time slot in the th time slot. After determining the reward for the th time slot and the resource task status for the th time slot of the MEO satellite based on the resource scheduling information for the th time slot, it further includes: using the resource task status for the th time slot , the resource task status for the th time slot , the resource scheduling information for the th time slot and the reward for the th time slot as an experience sample < > and storing it in the experience buffer of the MEO satellite , , , >.

[0061] S104. Based on , a preset update step size and the resource scheduling information obtained for the th time slot, generate a network, determine whether to update the network parameters of the current resource scheduling information generation network, and obtain the resource scheduling information generation network for the th time slot.

[0062] Specifically, for and the preset update step size​ Perform a remainder operation to obtain a second operation result; when the second operation result is 0, use the loss function and the experience samples in the experience buffer to update the network parameters of the current Actor network, and obtain the Actor network obtained in the th time slot. After that, use the loss function and the experience samples in the experience buffer to update the network parameters of the current Critic network, and obtain the Critic network obtained in the th time slot; when the second operation result is not 0, do not update the network parameters, and directly use the Actor network obtained in the th time slot as the Actor network in the th time slot.

[0063] Specifically, when updating the network parameters of the Actor network, obtain some experience samples from the experience buffer of the MEO satellite , calculate the loss according to the obtained experience samples and the loss function , and use the gradient descent method to obtain the network parameters of the Actor network in the th time slot , and using the network parameters of the Actor network in the th time slot, the Actor network obtained in the th time slot can be obtained.

[0064] Exemplarily, The expression of is as follows:

[0065] ;

[0066] ;

[0067] Where is the advantage function under and . The advantage function measures the goodness or badness of choosing a specific action (i.e., specific resource scheduling information) compared to the average of all possible actions in a given state. That is, the advantage function measures whether the reward obtained by taking under is higher or lower than the average level. is the action value function of taking under . It represents the expected value of all possible future cumulative rewards after taking under . is the state value function under . It represents the expected value of all possible future cumulative rewards under . is the probability ratio of the policy, where the policy represents the network parameters of the Actor network, and the probability ratio of the policy refers to the new policy and the old policy under the same state and action. At the current th time slot, the new policy refers to the network parameters of the th time slot of the Actor network, and the old policy refers to the network parameters of the th time slot of the Actor network. is to clip the probability ratio of the policy. The clipping threshold is a hyperparameter used to control the step size of policy update. Through the clipping operation, the probability ratio of the policy is restricted within to prevent the policy from being updated too much. represents the mathematical expectation.

[0068] Exemplarily, the expression of

[0069] ;

[0070] where represents the network parameters of the th time slot of the Actor network, is the learning rate, which is used to control the step size of parameter update. represents taking the derivative of .

[0071] Specifically, when updating the network parameters of the Critic network, some experience samples are obtained from the experience buffer of this MEO satellite, and the loss is calculated according to the obtained experience samples and the loss function , and the gradient descent method is used to obtain the network parameters of the Critic network at the th time slot , and using the network parameters of the Critic network at the th time slot , the Critic network at the th time slot can be obtained.

[0072] Exemplarily, the expression of

[0073] ;

[0074] ;

[0075] where is the estimated value of the expected cumulative reward after taking a certain action in the current state, which is a target value.R Represents the current reward, which is the estimated value of the reward by the Critic network for the next state and represents the expected value of all possible cumulative rewards that can be obtained in the future starting from the next state. It is the estimate of the state output by the Critic network, is the discount factor, used to measure the importance. The closer the discount factor is to 1, the greater the proportion of in the target value, indicating the estimated value output by the Critic network and the target value The mean squared error between them. During the training process, the Critic network adjusts its parameters by minimizing the mean squared error between the target value and the estimated value to make the estimated value it outputs closer to the target value .

[0076] Exemplarily, the expression of

[0077] ;

[0078] wherein, is the network parameter of the Critic network at the th time slot, is the learning rate, used to control the step size of parameter update, is the derivative of with respect to

[0079] S105. Based on , the preset aggregation step size and the resource scheduling information generation network obtained at the th time slot, determine whether to perform network parameter aggregation with the remaining MEO satellites and obtain the resource scheduling information generation network obtained at the latest th time slot.

[0080] Specifically, perform a modulo operation on and the preset aggregation step size to obtain a first operation result; when the first operation result is 0, obtain the Actor network of the remaining MEO satellites at the th time slot, and use the preset aggregation update matrix to aggregate the MEO satellite with the remaining​ The network parameters of the Actor network obtained in the th time slot of the th MEO satellite are aggregated to obtain the network parameters of the th MEO satellite after aggregation in the th time slot. The Actor network obtained in the th time slot of the MEO satellite is updated using the network parameters of the th MEO satellite after aggregation in the th time slot to obtain the latest Actor network obtained in the th time slot of the MEO satellite. When the first operation result is not 0, the Actor network obtained in the th time slot of the MEO satellite is used as the latest Actor network obtained in the th time slot of the MEO satellite. Exemplarily, the expression of the network parameters after aggregation in the th time slot is as follows: ;

[0081] ;

[0082] ;

[0083] Among them, is the network parameter after aggregation in the th time slot, is the set of network parameters of the Actor network obtained in the th time slot of the th MEO satellite, is the preset aggregation update matrix, is the federated aggregation factor. The federated aggregation factor represents the weight of the elements on the diagonal, which usually represents the degree of retention of each MEO satellite's own network parameters. is the non - diagonal element,

[0084] which represents the influence of the other S106. Generate a network based on the resource task status of the th time slot of the MEO satellite and the resource scheduling information obtained in the latest th time slot of the MEO satellite, and continue to determine the latest resource scheduling information generation network in the th round of training in the th time slot until reaching the​​​​ At the th time slot of the round of training, the round of training is completed, and the round of training continues until the iteration termination condition is reached, and the trained resource scheduling information generation network of the MEO satellite is obtained.

[0085] For example, when the number of training rounds reaches the preset round threshold, it can be considered that the iteration termination condition is reached, and thus the Actor network obtained in the last training of the last round can be used as the trained Actor network of the MEO satellite.

[0086] In the present invention, is the deployment number of the MEO satellite that minimizes the overall scheduling complexity of the system when there are LEO satellites in the system; the expression of the overall scheduling complexity of the system is as follows:

[0087] ;

[0088] ;

[0089] Among them, is the overall scheduling complexity of the system, represents the scheduling complexity when each MEO satellite manages multiple LEO satellites, is the dimension of the input data of the trained resource scheduling information generation network, is the number of neurons in the hidden layer of the trained resource scheduling information generation network, is the dimension of the output data of the trained resource scheduling information generation network. For example, is 6. In the expression of , the four terms other than constitute the scheduling complexity when the multiple LEO satellites managed by each MEO satellite perform collaborative calculations. " " and " " are both multiplication signs. Exemplarily, Table 1 below gives the optimal MEO deployment cases determined by using the above expression of the overall scheduling complexity of the system under different LEO numbers, where

[0090] Table 1 Optimal MEO Deployment Numbers Corresponding to Different LEO Numbers

[0091]

[0092] ​​Existing methods ignore the relationship between the scheduling complexity in the actual operation of the operation and maintenance and resource management and control architecture and methods of giant constellations and the scale of the number of satellites in the network. Based on the proposed operation and maintenance and resource management and control system for giant constellations oriented to computing-network convergence, the present invention proposes a calculation model for the overall scheduling complexity of the system, enabling the operation and maintenance and resource management and control system for giant constellations oriented to computing-network convergence proposed by the present invention to obtain the optimal number of MEO satellite deployments with low scheduling complexity under different numbers of LEO satellite deployments, so that the present invention can provide guidance for the construction and deployment of actual satellite constellations.

[0093] The present invention also provides an operation and maintenance and resource management and control method for giant constellations oriented to computing-network convergence. This method is applied to each MEO satellite in the above system, as Figure 3 shown. This method includes:

[0094] S201. Obtain the resource task status of itself in the current time slot of the current scheduling period.

[0095] S202. Use the trained resource scheduling information generation network to generate resource scheduling information for the current time slot according to the resource task status of itself in the current time slot; the trained resource scheduling information generation network is trained by using local reinforcement learning and federated learning methods; the resource scheduling information for the current time slot represents the proportion of the data volume executed by the MEO satellite in the computing tasks of each LEO satellite managed in the current time slot to the total data volume of the computing tasks of the LEO satellite.

[0096] S203. Send the resource scheduling information for the current time slot to each LEO satellite managed in the current time slot.

[0097] S204. Determine whether to perform data calculation on the computing tasks of each LEO satellite managed in the current time slot according to the resource scheduling information for the current time slot, and the data volume that needs to be calculated when performing data calculation.

[0098] S205. When it is necessary to perform data calculation on the computing tasks of each LEO satellite managed in the current time slot, perform the corresponding data volume calculation on the computing tasks of each LEO satellite managed in the current time slot according to the data volume that needs to be calculated.

[0099] It should be noted that the specific principles of each step in the operation and maintenance and resource management and control method for giant constellations oriented to computing-network convergence have been described in detail when introducing the operation and maintenance and resource management and control system for giant constellations oriented to computing-network convergence above, and will not be elaborated here.

[0100] Compared with the prior art, the present invention has the following advantages:

[0101] (1) In view of the multi-layer complex structure and computing control of the giant constellation, the present invention proposes an operation and maintenance and resource control system and method for the giant constellation oriented to the integration of computing and networking. The system includes two-level control units, namely the MEO satellite coordination level and the LEO satellite edge level. By introducing a computing and communication coordination mechanism, an inter-layer satellite coordination scheduling and control mechanism, and a policy parameter coordination mechanism among MEO satellites, dynamic scheduling and efficient utilization of resources are achieved. At the MEO coordination level, it is mainly responsible for the global operation and maintenance and resource control of the entire satellite network, including the allocation of computing tasks and the comprehensive utilization of inter-layer resources; at the LEO edge level, it focuses on the collaborative execution of real-time computing and communication tasks of edge nodes. This feature enables the present invention to perform efficient computing scheduling in practical applications, without relying on the control of ground controllers, and has higher real-time performance and better resource control and scheduling capabilities.

[0102] (2) In view of the actual multi-layer satellite network scenario of the giant constellation, considering problems such as dynamically uncertain computing tasks, the dynamic fluctuation process of satellite energy acquisition and consumption, and resource limitations in reality. In order to ensure the efficient completion of high-energy-consuming computing tasks in the actual satellite network scenario, the present invention focuses on the dynamic fluctuation of energy acquisition caused by the periodic switching of satellites between the earth's shadow area and the direct sunlight area, and the arrival of dynamically uncertain computing tasks, and constructs the computing and networking coordination process of computing and communication resources in the multi-layer giant constellation and its corresponding dynamic consumption and acquisition process of energy according to these actual fluctuation characteristics. This feature makes the present invention more efficient and more in line with the real application environment.

[0103] (3) Different from the existing planning technologies, the present invention decides the proportion of task data of LEO satellites assisted by MEO satellites for collaborative computing in each time slot through continuous learning. At the same time, considering the long-term optimal resource control decision under the arrival of dynamic tasks, the coordination between the LEO satellite and the MEO satellite layer is realized. The present invention conducts computing and networking coordination among MEO satellites through the sharing, interaction, and aggregation of policy parameters (instead of the relay transmission of task data) among MEO satellites, which can reduce the occupancy of communication bandwidth and delay, and achieve the computing and networking coordination of the entire satellite network. Through efficient LEO-MEO inter-layer computing and networking coordination and the aggregation coordination of policy parameters among MEO satellites, the present invention realizes the efficient computing and networking integration control of the giant constellation, further improving the task completion volume of the entire satellite network. This feature enables the present invention to have better overall network dynamic decision-making capabilities and higher task completion rates.

[0104] (4) Existing methods ignore the scheduling complexity in the actual operation of the operation and maintenance and resource control architecture and methods for megaconstellations and the relationship with the scale of the number of satellites in the network. In view of the multi-layer complex structure and computing control problems of megaconstellations, the present invention proposes an operation and maintenance and resource control system and method for megaconstellations oriented to the integration of computing and networking. According to this system and method, the optimal number of MEO satellite deployments with low scheduling complexity can be obtained under different numbers of LEO satellite deployments. This feature enables the present invention to provide guidance for the construction and deployment of actual satellite constellations.

[0105] The following verifies some technical effects of the present invention through simulation experiments.

[0106] Simulation scenario

[0107] The simulation scenario includes 144 LEO satellites with an orbital altitude of 780 km and 6 MEO satellites with an orbital altitude of 8000 km. Each satellite is equipped with edge computing and communication capabilities. The hardware computing coefficients of LEO satellites and MEO satellites 、 are 7×10 -26 J / cycle and 7×10 -24 J / cycle respectively. J is joule and cycle is revolution. The hardware computing coefficient of a satellite represents the amount of energy consumed in one computer clock cycle. In this scenario, each LEO satellite can collect, transmit, and process ecological monitoring data. The link capacity between satellites is 300 Mbps. The receiving and transmitting powers of LEO satellites are 8W and 10W respectively. The learning rate , the number of time slots (i.e., the scheduling period) , the discount factor , the total number of training iterations , the federated aggregation factor is 0.5, is 2, is 4, and the clipping threshold .

[0108] Simulation content and results

[0109] The present invention is compared with the independent control and scheduling algorithm and the non-interlayer cooperation algorithm. As the computing requirement of task data increases, the task completion amounts of the present invention, the independent control and scheduling algorithm, and the non-interlayer cooperation algorithm are as Figure 4 shown, and the energy consumption results are as Figure 5 shown. Among them, Figure 4 and Figure 5 's "proposed cooperation algorithm" represents the present invention, and the computing requirement The unit of [the relevant quantity] is cycles / bit (cycles per bit), the unit of the task data volume is Gbit, and the unit of the energy consumption is KJ (kilojoule). Specifically, Figure 4 shows the amount of task data completed by the present invention, the independent control and scheduling algorithm, and the non-inter-layer cooperation algorithm as the computing requirement of the task data increases. It can be clearly seen from Figure 4 that, since the present invention can adapt to the dynamic environment and cooperate through the MEO inter-satellite federation, it can complete the most tasks under different burst task volumes. Compared with other methods, it can significantly increase the amount of task data completed. To further explore and explain the source of the advantages of the present invention, the present invention further studied the energy consumption situation, as shown in Figure 5 . Combining Figure 4 and Figure 5 it can be seen that when the computing requirement of the task data is relatively small, the task completion situations of different methods under limited resources are similar because the resources are relatively sufficient at this time. When the computing requirement of the task data is relatively large, the present invention can cooperate with the available communication and computing resources on different satellites through intelligent federation and complete more task data with lower resource consumption. Therefore, the present invention completes more task data through the intelligent cooperation of communication and computing resources under the realistic condition of limited resources.

[0110] It should be noted that the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more features. In the description of the present invention, "a plurality" means two or more unless otherwise specifically defined.

[0111] In the description of this specification, the description referring to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine the different embodiments or examples described in this specification.

[0112] In the specification, the word "including" does not exclude other components or steps, and "a" or "one" does not exclude the case of a plurality. Certain measures are described in different embodiments, but this does not mean that these measures cannot be combined to produce good effects.

[0113] The above content is a further detailed description of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is only limited to these descriptions. For those of ordinary skill in the technical field to which the present invention pertains, without departing from the concept of the present invention, several simple deductions or substitutions can still be made, and all should be regarded as belonging to the protection scope of the present invention.

Claims

1. A giant constellation operation and maintenance and resource management system for computing and network integration, characterized in that: include: A MEO satellite collaborative-level control architecture with K MEO satellites and a LEO satellite edge-level control architecture with N LEO satellites. MEO satellites and LEO satellites are connected via communication links. Each MEO satellite manages multiple LEO satellites in each time slot of each scheduling cycle. Each LEO satellite is used to obtain data of a computing task in a preset time period, and in each time slot t of each scheduling period of the preset time period, determine whether to perform data calculation of the computing task in the time slot t and the amount of data to be calculated when performing data calculation according to the resource scheduling information for the time slot t sent by the MEO satellite, and calculate the corresponding amount of data of the computing task in the time slot t according to the amount of data to be calculated; Each MEO satellite is used to obtain its own resource task status in the time slot t, use the trained resource scheduling information generation network, generate resource scheduling information for the time slot t according to its own resource task status in the time slot t and send it to each LEO satellite managed in the time slot t, and determine whether to perform data calculation of the computing task of each LEO satellite managed in the time slot t and the amount of data required to be calculated when performing data calculation according to the resource scheduling information for the time slot t, and calculate the corresponding data amount of the computing task of each LEO satellite managed in the time slot t according to the amount of data required to be calculated; the resource scheduling information of the time slot t represents the proportion of the data amount calculated by the MEO satellite in the computing task of each LEO satellite managed in the time slot t to the total data amount of the computing task of the LEO satellite; the value range of the resource scheduling information is 0 to 1.

2. The giant constellation operation and maintenance and resource management system for computing-network integration according to claim 1 is characterized in that: The resource mission status of each MEO satellite at the time slot t includes: the energy status, link status and on-board mission data status of the MEO satellite at the time slot t; wherein the energy status of the MEO satellite at the time slot t represents: the remaining power of the MEO satellite in the time slot t; the link status of the MEO satellite at the time slot t represents: the sum of the link capacities on the communication links between the MEO satellite and the managed LEO satellites in the time slot t; the on-board mission data status of the MEO satellite at the time slot t represents: the sum of the data amounts of the computing tasks of the LEO satellites managed by the MEO satellite in the time slot t.

3. The giant constellation operation and maintenance and resource management system for computing-network integration according to claim 1 is characterized in that: K is the number of MEO satellites deployed to minimize the overall scheduling complexity of the system when there are N LEO satellites in the system; the expression of the overall scheduling complexity of the system is as follows: Ψ=ψ·K; Among them, Ψ is the overall scheduling complexity of the system, ψ represents the scheduling complexity when each MEO satellite manages multiple LEO satellites, Lin is the dimension of the input data of the trained resource scheduling information generation network, Lhidden is the number of neurons in the hidden layer of the trained resource scheduling information generation network, and Lout is the dimension of the output data of the trained resource scheduling information generation network.

4. The giant constellation operation and maintenance and resource management system for computing-network integration according to claim 1 is characterized in that: The trained resource scheduling information generation network is trained by local reinforcement learning and federated learning methods with the optimization goal of minimizing the energy consumption of the system in a scheduling cycle, and with the energy constraints of each LEO satellite, the energy constraints of each MEO satellite, the computing constraints of each LEO satellite, the computing constraints of each MEO satellite, and the communication constraints of each LEO satellite as constraints of the optimization goal.

5. The giant constellation operation and maintenance and resource management system for computing-network integration according to claim 4 is characterized in that: The expressions of the energy constraints of each LEO satellite and each MEO satellite are as follows: in, t∈{1,...T}, is a LEO satellite set consisting of the N LEO satellites in the system, ln is the nth LEO satellite in the LEO satellite set, n is 1, 2, ..., N, t is each time slot in a scheduling period T, is a MEO satellite set composed of the K MEO satellites in the system, mk is the mth MEO satellite in the MEO satellite set, and the value of m is 1, 2, ..., K, is the maximum battery capacity of the nth LEO satellite, ΓL is the percentage of the maximum discharge depth of the LEO satellite, is the amount of power that the nth LEO satellite has in each time slot, is the amount of electricity charged by the nth LEO satellite in each time slot, is the transmission energy consumption of the nth LEO satellite in each time slot, is the calculated energy consumption of the nth LEO satellite in each time slot; is the maximum battery capacity of the m-th MEO satellite, ΓM is the percentage of the maximum discharge depth of the MEO satellite, is the amount of power available for the m-th MEO satellite in each time slot, is the amount of electricity charged by the m-th MEO satellite in each time slot, is the receiving energy consumption of the m-th MEO satellite in each time slot, is the calculated energy consumption of the mth MEO satellite in each time slot.

6. The giant constellation operation and maintenance and resource management system for computing-network integration according to claim 1 is characterized in that: Each MEO satellite includes: The resource task status monitoring module is used to obtain the resource usage and computing task execution status of the MEO satellite itself and each LEO satellite managed by it in each time slot t of each scheduling cycle, and obtain the resource task status of the MEO satellite in the time slot t; a data calculation processing module, configured to receive data of the calculation task of each LEO satellite managed when data calculation of the calculation task of each LEO satellite managed is required in the time slot t, and calculate the corresponding data amount of the calculation task of each LEO satellite managed in the time slot t according to the data amount to be calculated; The computing-network fusion scheduling optimization module is used to train the resource scheduling information generation network by using local reinforcement learning and federated learning methods, obtain the trained resource scheduling information generation network, and share and aggregate the network parameters of the resource scheduling information generation network with the remaining K-1 MEO satellites during the training of the resource scheduling information generation network, and use the trained resource scheduling information generation network to generate resource scheduling information for the time slot t according to the resource task status of the MEO satellite itself in the time slot t, and send the resource scheduling information for the time slot t to each LEO satellite managed in the time slot t.

7. A method for giant constellation operation and maintenance and resource management for computing-network integration, characterized in that: Applied to each MEO satellite in the system of any one of claims 1 to 6 above, the method comprises: Get the resource task status of the current time slot in the current scheduling cycle; The resource scheduling information generation network is trained to generate resource scheduling information for the current time slot according to its own resource task status in the current time slot; the trained resource scheduling information generation network is trained by using local reinforcement learning and federated learning methods; the resource scheduling information of the current time slot represents the proportion of the amount of data calculated by the MEO satellite to the total amount of data of the LEO satellite's computing tasks in the computing tasks of each LEO satellite managed in the current time slot; Sending resource scheduling information for the current time slot to each LEO satellite managed in the current time slot; Determine, according to resource scheduling information for the current time slot, whether to perform data calculation of the computing task of each managed LEO satellite in the current time slot, and the amount of data that needs to be calculated when performing the data calculation; When data calculation of the computing task of each managed LEO satellite needs to be performed in the current time slot, the corresponding data amount of the computing task of each managed LEO satellite is calculated in the current time slot according to the data amount that needs to be calculated.

8. The method for giant constellation operation and maintenance and resource management for computing-network integration according to claim 7 is characterized in that: The training method of the trained resource scheduling information generation network of each MEO satellite includes: At the cth time slot in the kth round of training, the resource task state of the MEO satellite in the cth time slot is obtained; wherein each round of training corresponds to a scheduling cycle, each scheduling cycle has T time slots, and training is performed every preset update step time slots; when c=1, the resource task state of the MEO satellite in the cth time slot is a preset state; Determine the resource scheduling information of the c-th time slot according to the resource scheduling information generation network obtained in the c-1th time slot and the resource task status of the c-th time slot; wherein, when c=1, the resource scheduling information generation network obtained in the c-1th time slot is the resource scheduling information generation network obtained in the last training in the k-1th round of training; Determining a reward for the c-th time slot and a resource task status for the c+1-th time slot of the MEO satellite based on the resource scheduling information for the c-th time slot; Based on c, the preset update step size and the resource scheduling information generation network obtained in the c-1th time slot, determine whether to update the network parameters of the current resource scheduling information generation network, and obtain the resource scheduling information generation network obtained in the cth time slot; Based on c, the preset aggregation step size and the resource scheduling information generation network obtained in the c-th time slot, determine whether to perform network parameter aggregation with the remaining K-1 MEO satellites, and obtain the latest resource scheduling information generation network obtained in the c-th time slot; According to the resource task status of the c+1th time slot of the MEO satellite and the resource scheduling information generation network obtained in the latest cth time slot, continue to determine the latest resource scheduling information generation network in the c+1th time slot in the kth round of training, until the Tth time slot of the kth round of training is reached, the kth round of training is completed, and the k+1th round of training is continued until the iteration termination condition is reached, and the trained resource scheduling information generation network of the MEO satellite is obtained.

9. The method for giant constellation operation and maintenance and resource management for computing-network integration according to claim 8, characterized in that: The generating network based on c, the preset aggregation step size and the resource scheduling information obtained in the c-th time slot, determining whether to perform network parameter aggregation with the remaining K-1 MEO satellites, and obtaining the latest resource scheduling information generated network obtained in the c-th time slot, includes: Performing a modulo operation on c and a preset aggregation step length to obtain a first operation result; When the first operation result is 0, the resource scheduling information generation network obtained in the c-th time slot of the remaining K-1 MEO satellites is obtained, and the network parameters of the resource scheduling information generation network obtained in the c-th time slot of the MEO satellite and the remaining K-1 MEO satellites are aggregated by using a preset aggregation update matrix to obtain the aggregated network parameters of the c-th time slot, and the resource scheduling information generation network obtained in the c-th time slot of the MEO satellite is updated by using the aggregated network parameters of the c-th time slot to obtain the latest resource scheduling information generation network obtained in the c-th time slot; When the first operation result is not 0, the resource scheduling information generation network obtained in the c-th time slot of the MEO satellite is used as the resource scheduling information generation network obtained in the latest c-th time slot.

Images