Satellite resource dynamic allocation method and system under cloud-native hybrid cloud architecture

Abstract

The invention provides a satellite resource dynamic allocation method under a cloud-native hybrid cloud architecture, and the method comprises the steps: packaging satellite load resources and cloud resources through a containerization technology, so as to form a unified virtual resource pool; constructing a deep learning model, predicting resource demands of each beam coverage area in a future time period based on satellite task historical data and real-time demands, and generating a resource matching scheme; a reinforcement learning algorithm is adopted to dynamically adjust the beam radius, channel allocation and time slot scheduling so as to match the real-time load change of cloud resources; according to the method, a satellite and cloud digital twin model is constructed, a scheduling strategy is simulated and predicted based on a resource matching scheme through the satellite and cloud digital twin model, a resource state is mirrored in real time, and a strategy library is updated according to simulation feedback data of the satellite and cloud digital twin model, so that multi-satellite task automatic recombination is realized. The invention further provides a satellite resource dynamic allocation system under the cloud and native hybrid cloud architecture and a computer storage medium. According to the method, three-dimensional cooperation of dynamic resource pooling, intelligent prediction and elastic scheduling is realized, and the problem that satellite task response time, resource utilization rate and safety cannot be achieved at the same time can be effectively solved.

Description

Technical Field

[0001] The present invention relates to the field of satellite communication technology, and in particular to a method, system and computer storage medium for dynamically allocating satellite resources under a cloud-native hybrid cloud architecture. Background Art

[0002] In the current satellite resource management environment, satellite resources (including transponders, orbits, power, etc.) are dispersed across heterogeneous cloud environments. This decentralized resource distribution leads to numerous problems. First, cross-platform scheduling is inefficient. Resource coordination and scheduling between different cloud platforms consumes a significant amount of time and effort, making it difficult to achieve fast and efficient resource allocation. Second, response latency is a serious issue. This delay not only affects the timely execution of satellite missions but can also cause delays in critical tasks, thereby impacting the operational efficiency and reliability of the entire satellite system.

[0003] Traditional satellite resource allocation strategies primarily rely on pre-allocation. This approach is particularly inadequate when faced with unexpected mission demands. The inability to dynamically adjust resource allocation based on real-time mission requirements often leads to insufficient satellite resource utilization. While existing technologies can improve resource utilization by optimizing traditional pre-allocation strategies, the improvement is limited. This inefficient resource utilization not only wastes resources but also limits the overall performance and economic benefits of satellite systems.

[0004] When it comes to data processing, sensitive data faces a dilemma. While private clouds offer a relatively secure computing environment, their limited computing power makes them inadequate for large-scale, complex data processing. On the other hand, while public clouds offer significant computing power, they carry the risk of data leakage and cannot guarantee the security of sensitive data. This conflict between security and performance makes it difficult for satellite systems to strike a balance in data processing, impacting their overall operational effectiveness and the extraction of data value.

[0005] Therefore, there is an urgent need for a method that can effectively resolve the contradiction between satellite mission response time, resource utilization and safety. Summary of the Invention

[0006] In response to the above-mentioned defects, the purpose of the present invention is to provide a method, system and computer storage medium for dynamic allocation of satellite resources under a cloud-native hybrid cloud architecture. By constructing a three-dimensional collaborative architecture of dynamic resource pooling, intelligent prediction and elastic scheduling, it effectively solves the contradiction between satellite mission response time, resource utilization and security.

[0007] In order to achieve the above technical effects, on the one hand, the present invention provides a method for dynamically allocating satellite resources under a cloud-native hybrid cloud architecture, comprising the following steps:

[0008] Using containerization technology to encapsulate satellite payload resources and cloud resources to form a unified virtual resource pool; wherein the cloud resources include computing nodes, storage capacity and network bandwidth of public and private clouds;

[0009] Constructing a deep learning model to predict resource requirements for each beam coverage area in future time periods based on historical satellite mission data and real-time demand, and generating a resource matching solution;

[0010] Using reinforcement learning algorithms to dynamically adjust beam radius, channel allocation, and time slot scheduling to match real-time load changes in the cloud resources;

[0011] A satellite and cloud digital twin model is constructed, and a scheduling strategy is simulated and predicted based on the resource matching solution through the satellite and cloud digital twin model, and resource status is mirrored in real time. The strategy library is updated according to the simulation feedback data of the satellite and cloud digital twin model to realize automatic reorganization of multi-satellite missions.

[0012] Optionally, the building of a deep learning model includes:

[0013] Construct a generator G, the input of which is a random noise sequence Z = {z1,z2,z1,…,z T}, the output is time series data Among them, θ G is the parameter of the generator G, z t is the random noise at time point t, T is the length of the sequence;

[0014] Construct a discriminator D, the input of which is the real time series data X = {x1, x2, x1, ..., x T} or the time series data The output is the probability of the authenticity of the input data; where x t is the observation value at time point t, T is the length of the sequence;

[0015] By alternately optimizing the discriminator loss function L D and the generator loss function L G Train the model, where:

[0016]

[0017] P data is the distribution of real data, P Z is the distribution of random noise;

[0018] After the model training is completed, the generator G predicts the resource requirements of each beam coverage area in the future period based on the satellite mission historical data and real-time demand, and generates a resource matching plan.

[0019] Optionally, the dynamically adjusting the beam radius, channel allocation, and time slot scheduling using a reinforcement learning algorithm includes:

[0020] Initialize the policy network parameters θ and value network parameters φ;

[0021] Predefined reinforcement learning elements, including state s t 、Action a t And the reward function r t , where state s t Contains satellite beam coverage information, channel allocation status, time slot allocation status and user service requirements; Action a t Represented as a dynamic adjustment instruction for beam radius, channel allocation or time slot scheduling; reward function r t Comprehensive calculation based on task completion rate, resource utilization and data transmission quality;

[0022] The objective function is constructed as:

[0023]

[0024] in, is the advantage function, is the state and action value function, is the value network estimated at state s t The expected return under θ (a t |s t ) means in state s t Next take action a t The probability of is the probability of the old strategy, ∈ is the parameter of the clipping range, L VF (φ) is the loss function of the value function, L ENT (θ) is the entropy regularization term, c1 and c2 are weight coefficients, is the expected value;

[0025] Update the policy network parameters θ by maximizing the objective function L(θ);

[0026] By minimizing the loss function L of the value function VF (φ) Update the value network parameter φ.

[0027] On the other hand, based on the same inventive structure, the present invention also provides a satellite resource dynamic allocation system under a cloud-native hybrid cloud architecture, including a resource virtualization module, a resource prediction and scheduling module, a cloud-native management module, and a hybrid cloud interface module; wherein:

[0028] The resource virtualization module is configured to encapsulate satellite payload resources and cloud resources through containerization technology to form a unified virtual resource pool; wherein the cloud resources include computing nodes, storage capacity and network bandwidth of public and private clouds;

[0029] The resource prediction and scheduling module is configured to use a deep learning model to predict resource requirements for each beam coverage area in the future based on historical satellite mission data and real-time demand, generate a resource matching plan, and dynamically adjust beam radius, channel allocation, and time slot scheduling using a reinforcement learning algorithm to match real-time load changes of the cloud resources.

[0030] The cloud-native management module is configured to simulate and predict the scheduling strategy based on the resource matching solution through the satellite and cloud digital twin model and mirror the resource status in real time, and update the strategy library according to the simulation feedback data of the satellite and cloud digital twin model to realize the automatic reorganization of multi-satellite missions;

[0031] The hybrid cloud interface module is configured to achieve seamless connection between satellite resources and multiple cloud platforms.

[0032] Optionally, the resource prediction and scheduling module specifically includes:

[0033] The spatiotemporal prediction submodule is configured to predict user needs based on historical task data and / or historical user needs using optimization algorithms or deep learning models;

[0034] The scheduling submodule is configured to dynamically adjust the resource allocation plan based on the prediction results of the spatiotemporal prediction submodule, user needs, satellite resource status and other information.

[0035] Optionally, the spatiotemporal prediction submodule is specifically used to:

[0036] Construct a generator G, the input of which is a random noise sequence Z = {z1,z2,z1,…,z T}, the output is time series data Among them, θ G is the parameter of the generator G, z t is the random noise at time point t, T is the length of the sequence;

[0037] Construct a discriminator D, the input of which is the real time series data X = {x1, x2, x1, ..., x T} or the time series data The output is the probability of the authenticity of the input data; where x t is the observation value at time point t, T is the length of the sequence;

[0038] By alternately optimizing the discriminator loss function LD and the generator loss function L G Train the model, where:

[0039]

[0040] P data is the distribution of real data, P Z is the distribution of random noise;

[0041] After the model training is completed, the generator G predicts the resource requirements of each beam coverage area in the future period based on the satellite mission historical data and real-time demand, and generates a resource matching plan.

[0042] Optionally, the scheduling submodule is specifically used to:

[0043] Initialize the policy network parameters θ and value network parameters φ;

[0044] Predefined reinforcement learning elements, including state s t 、Action a t And the reward function r t , where state s t Contains satellite beam coverage information, channel allocation status, time slot allocation status and user service requirements; Action a t Represented as a dynamic adjustment instruction for beam radius, channel allocation or time slot scheduling; reward function r t Comprehensive calculation based on task completion rate, resource utilization and data transmission quality;

[0045] The objective function is constructed as:

[0046]

[0047] in, is the advantage function, is the state and action value function, is the value network estimated at state s t The expected return under θ (a t |s t ) means in state s t Next take action a t The probability of is the probability of the old strategy, ∈ is the parameter of the clipping range, L VF (φ) is the loss function of the value function, L ENT (θ) is the entropy regularization term, c1 and c2 are weight coefficients, is the expected value;

[0048] Update the policy network parameters θ by maximizing the objective function L(θ);

[0049] By minimizing the loss function L of the value function VF (φ) Update the value network parameter φ.

[0050] Optionally, the cloud native management module specifically includes:

[0051] The data acquisition submodule is configured to connect sensors and devices in the satellite constellation and hybrid cloud through IoT technology. It collects data on satellite location, status, mission execution, and cloud resource usage in real time, calculates the constellation topology, and obtains global satellite status and global resource usage, which are then sent to the multi-source data fusion submodule.

[0052] a communication monitoring submodule configured to collect bandwidth and delay of mutual access between the entire constellation topology, generate a delay matrix between satellites, and send it to the multi-source data fusion submodule;

[0053] The multi-source data fusion submodule is configured to collect information from the data acquisition submodule and the communication monitoring submodule, and generate a dynamic panoramic resource view of the satellite constellation and hybrid cloud based on the global resource usage, the global satellite status, and the delay matrix;

[0054] The simulation submodule is configured to simulate the operation of satellite constellations and hybrid clouds in different scenarios through simulation technology;

[0055] A real-time feedback submodule is configured to compare the simulation results of the simulation submodule with actual operation data and adjust satellite missions and cloud resource configurations in real time;

[0056] The strategy preview submodule is configured to allow users to test and verify different scheduling strategies in a virtual environment.

[0057] Optionally, the hybrid cloud interface module specifically includes:

[0058] The cloud platform docking submodule is responsible for docking with various mainstream cloud platforms to achieve resource integration and management;

[0059] The elastic expansion submodule is configured to dynamically adjust the scale of cloud computing resources according to resource demand to meet the real-time needs of tasks;

[0060] The security isolation submodule is configured to ensure the security of resources and data between the satellite system and the cloud platform, and prevent data leakage and malicious attacks.

[0061] The present invention also provides a computer storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for dynamic allocation of satellite resources under a cloud-native hybrid cloud architecture as described in any one of the above items.

[0062] Compared with existing technologies, this invention offers at least the following advantages: It comprehensively organizes and virtualizes satellite resources to create a unified resource view. It also employs a machine learning-based prediction model, combining historical satellite mission data with current mission requirements, to estimate resource demand trends for different missions and rationally schedule resources for the satellite communication system. By monitoring resource utilization efficiency in real time and providing feedback to adjust allocation strategies, it achieves dynamic and optimized allocation of satellite resources within a cloud-native hybrid cloud architecture, improving resource utilization and mission adaptability, and ensuring stable satellite service operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 A flowchart of the steps of the method for dynamic allocation of satellite resources under the cloud-native hybrid cloud architecture provided by the first embodiment of the present invention;

[0064] Figure 2 A structural block diagram of the satellite resource dynamic allocation system under the cloud-native hybrid cloud architecture provided by the second embodiment of the present invention;

[0065] Figure 3 A structural block diagram of the resource prediction and scheduling module of the satellite resource dynamic allocation system under the cloud-native hybrid cloud architecture provided by the second embodiment of the present invention;

[0066] Figure 4 A structural block diagram of the cloud-native management module of the satellite resource dynamic allocation system under the cloud-native hybrid cloud architecture provided by the second embodiment of the present invention;

[0067] Figure 5 This is a structural block diagram of the hybrid cloud interface module of the satellite resource dynamic allocation system under the cloud-native hybrid cloud architecture provided by the second embodiment of the present invention. DETAILED DESCRIPTION

[0068] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0069] It should be noted that references to "one embodiment," "an embodiment," "an example embodiment," etc., in this specification indicate that the described embodiment may include specific features, structures, or characteristics, but not every embodiment must include these specific features, structures, or characteristics. Furthermore, such references do not necessarily refer to the same embodiment. Furthermore, when specific features, structures, or characteristics are described in conjunction with an embodiment, whether or not explicitly described, it is understood that incorporating such features, structures, or characteristics into other embodiments is within the knowledge of those skilled in the art.

[0070] In addition, certain words are used in the specification and subsequent claims to refer to specific components or parts. It should be understood by those with ordinary knowledge in the relevant field that manufacturers may use different nouns or terms to refer to the same component or part. This specification and subsequent claims do not use differences in names as a way to distinguish components or parts, but rather use differences in the functions of components or parts as the criteria for distinction. The words "including" and "comprising" mentioned throughout the specification and subsequent claims are open-ended terms and should be interpreted as "including but not limited to". In addition, the word "connect" here includes any direct and indirect electrical connection means. Indirect electrical connection means include connection through other devices.

[0071] Before describing the embodiments of the present application in detail, the technical concept of the present application is briefly described:

[0072] The present invention realizes dynamic three-dimensional collaborative optimization of satellite resources through a cloud-native hybrid cloud architecture, specifically including: 1. Resource virtualization: abstracting satellite payloads (computing / storage / communication) and cloud resources (public cloud / private cloud) into a unified resource pool through containerization technology, breaking the boundary of satellite and ground resources. 2. Intelligent prediction: using generative adversarial networks (GANs) to analyze historical mission data, predict resource requirements in future beam areas, and generate allocation plans. 3. Elastic scheduling: using proximal strategy optimization algorithms to adjust beam radius, channel allocation and time slot scheduling in real time to match cloud load changes. 4. Digital twin closed-loop construction of a satellite-cloud digital twin model, through the communication delay matrix and panoramic resource view simulation strategy, combined with real-time feedback dynamic optimization task reorganization. The three-dimensional collaborative mechanism (virtualization + prediction + scheduling) of the present invention effectively solves the contradiction between resource dispersion, response delay and low utilization; using integrated scheduling of satellite and ground resources, sensitive data is stored in the private cloud, and computing-intensive tasks are transferred to the public cloud, taking into account both security and performance.

[0073] The following describes the specific principles of the satellite resource dynamic allocation method under the cloud-native hybrid cloud architecture of the present application in conjunction with specific embodiments.

[0074] Example 1

[0075] Figure 1A method for dynamically allocating satellite resources in a cloud-native hybrid cloud architecture provided by an embodiment of the present invention is shown, including the following steps:

[0076] S101: Encapsulating satellite payload resources and cloud resources through containerization technology to form a unified virtual resource pool; wherein the cloud resources include computing nodes, storage capacity and network bandwidth of public clouds and private clouds.

[0077] The satellite payload resources include hardware resources such as computing, storage and communication.

[0078] Specifically, containerization technology is used to abstract the hardware resources and cloud resources of satellite payloads into flexibly allocated virtual resources; at the same time, computing nodes, storage capacity, network bandwidth, etc. in public or private clouds are included in the scope of unified management; this virtualization and integration method breaks the boundary between satellite resources and cloud computing resources, forming a unified resource pool.

[0079] S102: Build a deep learning model. This model uses historical satellite mission data and real-time demand to predict resource requirements for each beam coverage area in the future and generate a resource matching solution. Specifically, a deep learning model is built to predict resource requirements for each beam coverage area in the future and generate an optimal matching solution.

[0080] In an optional embodiment, the constructing of the deep learning model includes:

[0081] Construct a generator G, the input of which is a random noise sequence Z = {z1,z2,z1,…,z T}, the output is time series data Among them, θ G is the parameter of the generator G, z t is the random noise at time point t, T is the length of the sequence;

[0082] Construct a discriminator D, the input of which is the real time series data X = {x1, x2, x1, ..., x T} or the time series data The output is the probability of the authenticity of the input data; where x t is the observation value at time point t, T is the length of the sequence;

[0083] By alternately optimizing the discriminator loss function L D and the generator loss function L G Train the model, where:

[0084]

[0085] P datais the distribution of real data, P Z is the distribution of random noise;

[0086] After the model training is completed, the generator G predicts the resource requirements of each beam coverage area in the future period based on the satellite mission historical data and real-time demand, and generates a resource matching plan.

[0087] The generator G generates time series data, takes random noise Z as input, and outputs a generated time series

[0088] The discriminator D is used to distinguish between real data X and generated data It takes time series data as input and outputs a probability value, which indicates the probability that the input data is real data. Mathematically, the discriminator D can be expressed as: p = D(X; θ D )or Among them, θ D are the parameters of the discriminator D.

[0089] The loss function L of the discriminator D D Aims to maximize the ability to distinguish between real data and generated data; the loss function L of the generator G G The goal is to minimize the discriminator D’s ability to identify generated data, that is, the generated data should be identified as real data as much as possible.

[0090] Alternately optimize the discriminator loss function L D and the generator loss function L G That is, alternately training the discriminator D and the generator G, specifically including:

[0091] Fixed θ G , update θ by gradient descent D To minimize L D ;

[0092] Fixed θ D , update θ by gradient descent G To minimize L G .

[0093] After training is complete, use the generator G to predict future time series data. Given a random noise sequence Z, the generator G will output a predicted time series In this way, user demand time series data can be effectively predicted and generated to achieve forward-looking and scientific resource allocation.

[0094] S103: Using a reinforcement learning algorithm to dynamically adjust beam radius, channel allocation, and time slot scheduling to match the real-time load changes of the cloud resources.

[0095] In an optional implementation, step S103 includes:

[0096] Initialize the policy network parameters θ and value network parameters φ;

[0097] Predefined reinforcement learning elements, including state s t 、Action a t And the reward function r t , where state s t Contains satellite beam coverage information, channel allocation status, time slot allocation status and user service requirements; Action a t Represented as a dynamic adjustment instruction for beam radius, channel allocation or time slot scheduling; reward function r t Comprehensive calculation based on task completion rate, resource utilization and data transmission quality;

[0098] Under the current strategy, interact with the environment and sample a series of states, actions and rewards {s t , a t , r t}.

[0099] Policy function π θ (a t |s t ) means in state s t Next take action a t The probability of , which is output by the policy network in the algorithm.

[0100] According to the sampled data, the advantage function A(s) of each state-action pair is calculated. t , a t ).

[0101] The objective function is constructed as:

[0102]

[0103] in, is the advantage function, is the state and action value function, is the value network estimated at state s t The expected return under θ (a t |s t ) means in state s t Next take action a t The probability of is the probability of the old strategy, ∈ is the parameter of the clipping range, L VF (φ) is the loss function of the value function, L ENT (θ) is the entropy regularization term, c1 and c2 are weight coefficients, is the expected value;

[0104] The objective function combines policy optimization, value function update, and entropy regularization.

[0105] Update the policy network parameters θ by maximizing the objective function L(θ);

[0106] By minimizing the loss function L of the value function VF (φ) Update the value network parameter φ.

[0107] Furthermore, sampling, advantage function calculation, strategy update and value function update are repeated until the strategy converges, and corresponding actions are taken under the current state of the satellite communication system to achieve reasonable scheduling of the resources of the satellite communication system.

[0108] S104: Build a satellite-cloud digital twin model. Use the satellite-cloud digital twin model to simulate and predict scheduling strategies based on the resource matching solution and mirror resource status in real time. Update the strategy library based on simulation feedback from the satellite-cloud digital twin model to achieve automatic reconfiguration of multi-satellite missions. Specifically, by building the satellite-cloud digital twin model, rehearsing scheduling strategies and mirroring resource status in real time, and combining online feedback to update the strategy library, automatic reconfiguration of multi-satellite missions is achieved.

[0109] Example 2

[0110] Figure 2 Another embodiment of the present invention provides a satellite resource dynamic allocation system 100 under a cloud-native hybrid cloud architecture, including a resource virtualization module 10, a resource prediction and scheduling module 20, a cloud-native management module 30, and a hybrid cloud interface module 40; wherein:

[0111] The resource virtualization module 10 is configured to encapsulate satellite payload resources and cloud resources through containerization technology to form a unified virtual resource pool; wherein, the cloud resources include computing nodes, storage capacity and network bandwidth of public and private clouds; the resource prediction and scheduling module 20 is configured to predict the resource requirements of each beam coverage area in the future period based on the historical data and real-time demand of satellite missions through a deep learning model, and generate a resource matching plan; and adopt a reinforcement learning algorithm to dynamically adjust the beam radius, channel allocation and time slot scheduling to match the real-time load changes of the cloud resources; the cloud native management module 30 is configured to simulate and predict the scheduling strategy based on the resource matching plan through the satellite and cloud digital twin model and mirror the resource status in real time, and update the strategy library according to the simulation feedback data of the satellite and cloud digital twin model to realize automatic reorganization of multiple satellite missions; the hybrid cloud interface module 40 is configured to realize seamless docking of satellite resources with multiple cloud platforms.

[0112] Specifically, the resource virtualization module 10 uses containerization technology to abstract the hardware resources of the onboard payload (such as sensors, communication equipment, processing units, etc. on the satellite) into flexibly allocated virtual resources. At the same time, it is closely connected with the hybrid cloud interface, and the resources in the cloud platform (such as computing nodes, storage capacity, network bandwidth, etc. in the public cloud or private cloud) are also included in the scope of unified management. This virtualization and integration method breaks the boundary between satellite resources and cloud computing resources, forming a unified resource pool; the resource prediction and scheduling module 20 is used to use optimization algorithms or machine learning models based on information such as satellite mission history data, satellite resource status and user needs to predict user needs and generate resource allocation plans; the cloud native management module 30 is used to synchronize the satellite constellation with the hybrid cloud status in real time, and is responsible for the overall management of the resources of the entire constellation; the hybrid cloud interface module 40 is configured to achieve seamless docking of satellite resources with multiple cloud platforms, while ensuring efficient resource utilization, security and scalability.

[0113] The advantage of the resource virtualization module 10 of this embodiment lies in its flexibility and integration. Through containerization technology, satellite hardware resources are abstracted into dynamically allocated virtual resources, avoiding resource waste. At the same time, combined with the hybrid cloud interface, it breaks the boundaries between satellite and cloud platform resources, forms a unified resource pool, improves resource allocation capabilities and adaptability, provides a basis for system optimization, reduces management complexity, and enhances system reliability and stability.

[0114] In one application example, a constellation of low-Earth orbit satellites provides broadband communications services to users worldwide. These satellites are equipped with a variety of payloads, including communication antennas, processing units, and storage devices. The system also accesses public and private cloud resources to process satellite data and perform compute-intensive tasks. Users are distributed across diverse geographic regions, and their communication needs vary significantly across time and space. Dynamic resource allocation is required to meet user needs and improve resource utilization.

[0115] Optionally, this embodiment is developed based on the resource management framework Kubernetes. Kubernetes is an open-source container orchestration platform for automating the deployment, scaling, and management of containerized applications. kubectl is the official command-line tool for Kubernetes, providing a simple yet powerful interface for communicating with, managing, and monitoring Kubernetes clusters.

[0116] This embodiment preferably encapsulates the hardware resources of the satellite payload (such as sensors, communication equipment, processing units on the satellite, etc.) and the resources in the cloud platform (such as computing nodes, storage capacity, network bandwidth, etc. in the public cloud or private cloud) through the containerization technology Kubernetes. First, define a container template to abstract the hardware resources of the satellite payload into dynamically allocated virtual resource units; for example, the communication antenna bandwidth of the satellite is divided into multiple virtual channels, each channel has a fixed bandwidth capacity; the computing power of the processing unit is divided into multiple virtual computing nodes, each node has a specific processor and memory configuration; the storage device is divided into multiple virtual storage volumes, each volume has a specific storage capacity. Then use the container orchestration tool Kubernetes to manage the encapsulated resources to form a unified resource pool. Through the Kubernetes API interface, the creation, deletion, update and query operations of resources are realized to ensure flexible allocation and efficient utilization of resources.

[0117] like Figure 3 As shown, in an optional embodiment, the resource prediction and scheduling module 20 specifically includes a spatiotemporal prediction submodule 201 and a scheduling submodule 202, wherein:

[0118] The spatiotemporal prediction submodule 201 is configured to predict user needs based on historical mission data and / or historical user needs using an optimization algorithm or a deep learning model; the scheduling submodule 202 is configured to dynamically adjust the resource allocation plan based on the prediction results of the spatiotemporal prediction submodule and information such as user needs and satellite resource status to maximize resource utilization and improve the efficiency of mission execution.

[0119] Therefore, the resource prediction and scheduling module 20 uses optimization algorithms or machine learning models based on satellite mission historical data, resource status, user needs and other information to accurately predict user needs and generate resource allocation plans, thereby improving the scientific nature and foresight of resource allocation.

[0120] In one example, the spatiotemporal prediction submodule 201 uses a deep learning module to predict the resource requirements of each beam coverage area over a period of time in the future. The spatiotemporal prediction submodule 201 is specifically used to:

[0121] Construct a generator G, the input of which is a random noise sequence Z = {z1,z2,z1,…,z T}, the output is time series data Among them, θ G is the parameter of the generator G, z t is the random noise at time point t, T is the length of the sequence;

[0122] Construct a discriminator D, the input of which is the real time series data X = {x1, x2, x1, ..., x T} or the time series data The output is the probability of the authenticity of the input data; where x t is the observation value at time point t, T is the length of the sequence;

[0123] By alternatingly optimizing the discriminator loss function L D and the generator loss function L G Train the model, where:

[0124]

[0125] P data is the distribution of real data, P Z is the distribution of random noise;

[0126] After the model training is completed, the generator G predicts the resource requirements of each beam coverage area in the future period based on the satellite mission historical data and real-time demand, and generates a resource matching plan.

[0127] Before model training, historical data of satellite missions are collected, including information such as mission type, mission time, mission location, resource usage, etc., and the data is preprocessed, such as data cleaning and normalization.

[0128] In a specific implementation, X may be resource demand data of each beam coverage area in the past period of time.

[0129] The random noise sequence Z is a randomly generated numerical sequence used to provide input to the generator G, enabling it to generate diverse data.

[0130] Optionally, the generator G uses LSTM (Long Short-Term Memory) to generate time series data. It takes random noise Z as input and outputs a generated time series Mathematically, the generator G can be expressed as The goal of the generator G is to generate resource demand forecast data for each beam coverage area in the future based on the input random noise and historical data.

[0131] LSTM (Long Short-Term Memory) is a special recurrent neural network architecture primarily used for processing sequence data. By introducing mechanisms such as input, forget, and output gates, it effectively addresses the vanishing and exploding gradient problems that plague traditional recurrent neural networks when processing long sequences. This allows it to better capture long-term dependencies within sequences and excels in fields such as natural language processing and time series prediction. LSTMs are used for tasks such as text generation, machine translation, and stock price prediction, providing a powerful tool for sequence modeling.

[0132] The discriminator D is used to distinguish between real data X and generated data It takes time series data as input and outputs a probability value indicating the probability that the input data is real data. The task of the discriminator D is to distinguish whether the input data is real data or synthetic data generated by the generator G.

[0133] Mathematically, the discriminator D can be expressed as: p = D(X; θ D )or Among them, θ D are the parameters of the discriminator D.

[0134] Discriminator loss function L D The first term in the formula Represents the recognition loss of the discriminator D for the real data. It is hoped that the output probability D(X) of the discriminator D for the real data is close to 1; the second term It represents the recognition loss of the discriminator D for the generated data. It is hoped that the output probability D(G(Z)) of the discriminator D for the generated data is close to 0.

[0135] The loss function L of the generator G G Minimize the discriminator D's ability to identify generated data, that is, the generated data is identified as real data as much as possible:

[0136] Alternately train the discriminator D and generator G: fix θ G , update θ by gradient descent D To minimize L D ; Fixed θ D , update θ by gradient descent G To minimize L G This alternating training method allows the generator and the discriminator to compete with each other during the training process. The generator continuously generates more realistic data, while the discriminator D continuously improves its ability to distinguish.

[0137] After training is completed, the generator G is used to predict future time series data. That is, by giving a random noise sequence Z, the generator G will output a predicted time series In this way, user demand time series data can be effectively predicted and generated to achieve forward-looking and scientific resource allocation.

[0138] In one example, the scheduling submodule 202 uses a reinforcement learning algorithm to dynamically adjust beam radius, channel allocation, and time slot scheduling to match the real-time load of cloud resources. The scheduling submodule 202 is specifically used to:

[0139] Initialize the policy network parameters θ and value network parameters φ;

[0140] Predefined reinforcement learning elements, including state s t 、Action a t And the reward function r t , where state s t Contains satellite beam coverage information, channel allocation status, time slot allocation status and user service requirements; Action a t Represented as a dynamic adjustment instruction for beam radius, channel allocation or time slot scheduling; reward function r t Comprehensive calculation based on task completion rate, resource utilization and data transmission quality;

[0141] Under the current strategy, interact with the environment and sample a series of states, actions and rewards {s t , a t , r t}.

[0142] In the specific implementation, the state s t It can be expressed as a vector, including satellite beam coverage information, frequency allocation, time slot allocation and user service requirements, etc. Action a t Represents adjustments to beam resources, frequency resources, and time slot resources. Reward function r t It can be designed based on the system's performance indicators, such as throughput, user satisfaction, interference level, etc.

[0143] Specifically, a reward is a scalar value used to evaluate the quality of an action in a certain state, thereby influencing future decisions. In the satellite resource scheduling scenario of this embodiment, the reward can be a comprehensive evaluation of various factors, such as task completion, resource utilization, data transmission quality, task priority satisfaction, energy consumption, and system stability. Specifically, if the satellite can efficiently complete the scheduled observation or data transmission tasks while rationally utilizing resources, ensuring high data transmission quality, meeting the requirements of high-priority tasks, reducing energy consumption, and maintaining stable system operation, then these factors combined can form a positive reward signal, incentivizing the intelligent agent to make better scheduling decisions in the future.

[0144] Policy function π θ (at |s t ) means in state s t Next take action a t The probability of , which is output by the policy network in the algorithm.

[0145] Optionally, the policy network is implemented using an actor-critic network. Actor-critic networks are a widely used reinforcement learning architecture that combines the strengths of policy gradients (actors) and value function estimation (critics). Actors are responsible for selecting actions based on the current policy, obtaining rewards by interacting with the environment, and updating the policy to improve the expected return of actions. The critic evaluates the value of the actor's actions, providing feedback on the quality of the policy, and updating the value function to more accurately estimate future rewards. The two work together: the actor uses the critic's evaluation to optimize the policy, and the critic uses the actor's experience to improve the value estimate, resulting in a more efficient learning and decision-making process.

[0146] According to the sampled data, the advantage function A(s) of each state-action pair is calculated. t , a t ), the advantage function measures the t Next take action a t How good or bad it is relative to the average case (value function).

[0147] The state-action value function can be estimated using a deep Q-network Q(s t , a t ;θ)=f(s t , a t ;θ), where θ is the parameter of the neural network. f(s t , a t ; θ) is the output of the neural network, indicating that in state s t Next take action a t The expected cumulative reward.

[0148] The objective function is constructed as:

[0149]

[0150] in, is the advantage function, is the state and action value function, is the value network estimated at state s t The expected return under θ (a t |s t ) means in state s t Next take action a t The probability of is the probability of the old strategy, ∈ is the parameter of the clipping range, L VF (φ) is the loss function of the value function, L ENT (θ) is the entropy regularization term, c1 and c2 are weight coefficients, is the expected value;

[0151] The above objective function combines policy optimization, value function update and entropy regularization.

[0152] Loss function of value function Entropy regularization term L ENT (θ)=H(π θ (s t )). Where H(π θ (s t ))=-∑ a π θ (a|s t )logπ θ (a|s t ) represents the entropy of the policy and the uncertainty of the policy distribution.

[0153] The working principle of the objective function is to limit the scope of policy updates. If the policy update ratio exceeds the preset range (i.e., greater than 1+∈ or less than 1-∈), the update will be clipped to prevent drastic changes in the policy. At the same time, combined with the mean square error loss To optimize the value function; by adding the entropy regularization term L ENT (θ), encourages strategic exploration and prevents premature convergence to local optimality.

[0154] Increasing entropy can prevent the policy from converging to a local optimum too early.

[0155] Update the policy network parameters θ by maximizing the objective function L(θ);

[0156] By minimizing the loss function L of the value function VF (φ) Update the value network parameter φ.

[0157] Furthermore, sampling, calculating the advantage function, updating the strategy and updating the value function are repeated until the strategy converges, and corresponding actions are taken under the current state of the satellite communication system to achieve reasonable scheduling of the resources of the satellite communication system.

[0158] Through multiple iterations of training, the policy network converges and obtains the optimal resource scheduling strategy. In actual operation, resource allocation is dynamically adjusted according to real-time task requirements and resource status, improving resource utilization and task execution efficiency.

[0159] See also Figure 4Optionally, the cloud native management module 30 specifically includes a data acquisition submodule 301, a communication monitoring submodule 302, a multi-source data fusion submodule 303, a simulation submodule 304, a real-time feedback submodule 305, and a strategy preview submodule 306, wherein:

[0160] The data acquisition submodule 301 is configured to connect the sensors and devices in the satellite constellation and hybrid cloud through the Internet of Things technology, collect data on the satellite's position, status, mission execution and cloud resource usage in real time, calculate the constellation topology, obtain the global satellite status and global resource usage and send them to the multi-source data fusion submodule 303; the communication monitoring submodule 302 is configured to collect the bandwidth and delay of mutual access between the entire constellation topology, generate the delay matrix between satellites, and send it to the multi-source data fusion submodule 303; the multi-source data fusion submodule 303 is configured to collect the data acquisition submodule 301 The information transmitted by the communication monitoring submodule 302 is used to generate a dynamic panoramic resource view of the satellite constellation and hybrid cloud according to the global resource usage, global satellite status, and delay matrix; the simulation submodule 304 is configured to simulate the operation of the satellite constellation and hybrid cloud in different scenarios through simulation technology; the real-time feedback submodule 305 is configured to compare the simulation results of the simulation submodule 304 with the actual operation data, and adjust the satellite mission and cloud resource configuration in real time; the strategy preview submodule 306 is configured to allow users to test and verify different scheduling strategies in a virtual environment, so as to discover potential problems in advance and optimize strategies.

[0161] In an optional embodiment, the data acquisition submodule 301 is used to connect sensors and devices in the satellite constellation with resources in the hybrid cloud through Internet of Things technology, such as using a satellite communication link to transmit the satellite's telemetry data to a ground station and then upload it to the cloud platform; collect satellite location information, task execution status, resource usage and other data in real time, and collect resource usage in public and private clouds, such as CPU and memory utilization of computing nodes; and clean, filter and format the collected data to remove invalid data and noise data, thereby improving data accuracy and consistency.

[0162] In an optional embodiment, the communication monitoring submodule 302 is used to monitor the communication bandwidth and delay within the satellite constellation in real time through the communication link between satellites, such as using a response mechanism to periodically send test signals and measure the round-trip time; monitor the communication link between the satellite and the ground station and the network communication between the ground station and the cloud platform, use network monitoring tools to measure the communication bandwidth and delay, and generate communication performance data; generate a delay matrix based on the monitored communication delay data, which is a two-dimensional array representing the communication delay between nodes. For example, for a constellation of n satellites, the delay matrix size is n×n; since the communication delay will change in real time, the delay matrix needs to be updated regularly, for example, once every 5 minutes.

[0163] In an optional embodiment, the multi-source data fusion submodule 303 is used to fuse data from different sources, eliminate redundancy and contradictions, and extract valuable information, such as analyzing communication performance by combining satellite position information and communication delay information, and evaluating resource demand and utilization by combining cloud resource usage and satellite mission execution; generating a resource view based on the fused data, including information such as the distribution, usage rate, and communication performance of satellite and cloud resources, such as displaying the satellite constellation topology and cloud platform resource pool through a graphical interface; since the resource status will change in real time, the resource view needs to be updated in real time, for example, once every 1 minute.

[0164] In an optional embodiment, the simulation submodule 304 is used to build a virtual simulation environment based on the topological structure of the satellite constellation, satellite performance parameters, cloud platform resource configuration and mission requirements, and simulate processes such as satellite movement, mission execution, resource allocation and communication interaction; the strategy preview submodule 306 is used to define different simulation scenarios, such as normal operation scenarios, emergency mission scenarios, resource failure scenarios, etc., to observe the dynamic adjustment capabilities of the system under different scenarios; run different scheduling strategies in the simulation environment, and record data such as resource usage, task execution efficiency, and communication performance; analyze the simulation results, evaluate the performance of the scheduling strategy, discover potential problems and optimize the strategy, such as adjusting strategy parameters or introducing new mechanisms; allow users to preview different scheduling strategies in a virtual environment, discover problems in advance and optimize them.

[0165] In an optional embodiment, the real-time feedback submodule 305 is used to compare the simulation results with the actual operation data, analyze the differences, such as comparing resource utilization and task response time, and evaluate the accuracy of the simulation results and the rationality of the scheduling strategy; generate feedback information based on the comparison results and feed it back to the scheduling submodule 202, such as prompting to adjust the resource allocation strategy or optimize communication performance; the scheduling submodule 202 adjusts the satellite mission and cloud resource configuration in real time according to the feedback information, such as adjusting the task priority or communication link.

[0166] The hybrid cloud interface module 40 is responsible for connecting with various mainstream cloud platforms (including public and private clouds) to achieve resource integration and management. Figure 5 Optionally, the hybrid cloud interface module 40 specifically includes a cloud platform docking submodule 401, an elastic expansion submodule 402, and a security isolation submodule 403, wherein:

[0167] The cloud platform docking submodule 401 is configured to be responsible for docking with multiple mainstream cloud platforms to achieve resource integration and management; the elastic expansion submodule 402 is configured to dynamically adjust the scale of cloud computing resources according to resource requirements to meet the real-time needs of the task; the security isolation submodule 403 is configured to ensure the security of resources and data between the satellite system and the cloud platform to prevent data leakage and malicious attacks.

[0168] In an optional embodiment, the cloud platform connection submodule 401 is configured as follows: It selects a supported cloud platform, such as public clouds like AWS, Azure, and Google Cloud, and private clouds like OpenStack and VMware vSphere, ensuring they have good interface support and security; configures interfaces to programmatically manage cloud resources; registers satellite resources and cloud platform resources to form a unified resource pool, and regularly synchronizes resource status information to ensure the system can monitor resource usage in real time. For example, it obtains information such as cloud platform resource utilization and virtual machine operating status through the interface and integrates this information with satellite resource status information.

[0169] In an optional embodiment, the elastic expansion submodule 402 is specifically used to combine the prediction results of the resource prediction and scheduling module 20 to analyze the resource demand gap. When it is predicted that the resource demand exceeds the existing capacity, the elastic expansion mechanism is automatically triggered to dynamically increase resources such as computing nodes, storage capacity or network bandwidth, and allocate the newly added resources to the corresponding tasks, optimize the resource allocation plan, and dynamically adjust the resource allocation according to the task execution status and resource utilization rate, and recycle excess resources in a timely manner.

[0170] In an optional embodiment, the security isolation submodule 403 is specifically used to formulate strict security policies covering aspects such as access control, data encryption, and network security, such as setting role-based access control policies to limit user access rights to resources, and encrypting the storage and transmission of sensitive data; achieving network isolation between the satellite system and the cloud platform through virtual private networks (VPN) and firewall technologies, and configuring firewall rules to limit illegal access and data leakage; encrypting data in the satellite and cloud platform to ensure the security of data during transmission and storage, and limiting access rights to data through access control mechanisms; conducting security audits regularly to check the implementation of security policies, monitoring the system security status in real time through security monitoring tools, and promptly detecting and handling security threats.

[0171] The present invention also provides a computer storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for dynamically allocating satellite resources under the cloud-native hybrid cloud architecture as described in the above embodiment is implemented.

[0172] The advantages of the computer storage medium over the prior art are the same as those of the satellite resource dynamic allocation system, which will not be described in detail here.

[0173] In summary, the method and system for dynamic allocation of satellite resources under the cloud-native hybrid cloud architecture described in the present invention achieves the three-dimensional collaboration of dynamic resource pooling, intelligent prediction, and elastic scheduling by adopting the cloud-native hybrid cloud architecture. Dynamic resource pooling can integrate scattered satellite resources to form a unified resource pool, improving the flexibility and utilization of resource allocation; intelligent prediction can predict resource requirements in advance based on historical data and real-time task requirements, providing a basis for resource scheduling; elastic scheduling can dynamically adjust resource allocation based on prediction results and real-time task requirements, ensuring timely response and efficient execution of satellite missions. Through this innovative three-dimensional collaborative architecture, the present invention can provide a new idea and solution for satellite resource management, which has important innovative value and practical application significance.

[0174] It should be noted that the present invention can be implemented in software and / or a combination of software and hardware, for example, can be implemented using an application specific integrated circuit (ASIC), a general purpose computer or any other similar hardware device. In one embodiment, the software program of the present invention can be executed by a processor to implement the above steps or functions. Similarly, the software program of the present invention (including related data structures) can be stored in a computer-readable recording medium, for example, a RAM memory, a magnetic or optical drive or a floppy disk and similar devices. In addition, some steps or functions of the present invention can be implemented using hardware, for example, as a circuit that cooperates with a processor to perform each step or function.

[0175] It should be noted that, in this article, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present invention is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in the opposite order according to the functions involved. For example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0176] Of course, the present invention may have many other embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art may make various corresponding changes and modifications based on the present invention, but these corresponding changes and modifications should all fall within the scope of protection of the claims attached to the present invention.

Claims

1. A method for dynamically allocating satellite resources under a cloud-native hybrid cloud architecture, characterized in that: Including steps: Using containerization technology to encapsulate satellite payload resources and cloud resources to form a unified virtual resource pool; wherein the cloud resources include computing nodes, storage capacity and network bandwidth of public and private clouds; Constructing a deep learning model to predict resource requirements for each beam coverage area in future time periods based on historical satellite mission data and real-time demand, and generating a resource matching solution; Using reinforcement learning algorithms to dynamically adjust beam radius, channel allocation, and time slot scheduling to match real-time load changes in the cloud resources; A satellite and cloud digital twin model is constructed, and a scheduling strategy is simulated and predicted based on the resource matching solution through the satellite and cloud digital twin model, and resource status is mirrored in real time. The strategy library is updated according to the simulation feedback data of the satellite and cloud digital twin model to realize automatic reorganization of multi-satellite missions.

2. The method for dynamically allocating satellite resources under the cloud-native hybrid cloud architecture according to claim 1, characterized in that: The construction of the deep learning model includes: Construct a generator G, the input of which is a random noise sequence Z = {z1,z2,z1,…,z T }, the output is time series data Among them, θ G is the parameter of the generator G, z t is the random noise at time point t, T is the length of the sequence; Construct a discriminator D, the input of which is the real time series data X = {x1, x2, x1, ..., x T } or the time series data The output is the probability of the authenticity of the input data; where x t is the observation value at time point t, T is the length of the sequence; By alternatingly optimizing the discriminator loss function L D and the generator loss function L G Train the model, where: P data is the distribution of real data, P Z is the distribution of random noise; After the model training is completed, the generator G predicts the resource requirements of each beam coverage area in the future period based on the satellite mission historical data and real-time demand, and generates a resource matching plan.

3. The method for dynamically allocating satellite resources under a cloud-native hybrid cloud architecture according to claim 1, wherein: The use of a reinforcement learning algorithm to dynamically adjust beam radius, channel allocation, and time slot scheduling includes: Initialize the policy network parameters θ and value network parameters φ; Predefined reinforcement learning elements, including state s t 、Action a t And the reward function r t , where state s t Contains satellite beam coverage information, channel allocation status, time slot allocation status and user service requirements; Action a t Represented as a dynamic adjustment instruction for beam radius, channel allocation or time slot scheduling; reward function r t Comprehensive calculation based on task completion rate, resource utilization and data transmission quality; The objective function is constructed as: in, is the advantage function, is the state and action value function, is the value network estimated at state s t The expected return under θ (a t |s t ) means in state s t Next take action a t The probability of is the probability of the old strategy, ∈ is the parameter of the clipping range, L VF (φ) is the loss function of the value function, L ENT (θ) is the entropy regularization term, c1 and c2 are weight coefficients, is the expected value; Update the policy network parameters θ by maximizing the objective function L(θ); By minimizing the loss function L of the value function VF (φ) Update the value network parameter φ.

4. A satellite resource dynamic allocation system under a cloud-native hybrid cloud architecture, characterized in that: It includes resource virtualization module, resource prediction and scheduling module, cloud native management module and hybrid cloud interface module; among them: The resource virtualization module is configured to encapsulate satellite payload resources and cloud resources through containerization technology to form a unified virtual resource pool; wherein the cloud resources include computing nodes, storage capacity and network bandwidth of public and private clouds; The resource prediction and scheduling module is configured to use a deep learning model to predict resource requirements for each beam coverage area in the future based on historical satellite mission data and real-time demand, generate a resource matching plan, and dynamically adjust beam radius, channel allocation, and time slot scheduling using a reinforcement learning algorithm to match real-time load changes of the cloud resources. The cloud-native management module is configured to simulate and predict the scheduling strategy based on the resource matching solution through the satellite and cloud digital twin model and mirror the resource status in real time, and update the strategy library according to the simulation feedback data of the satellite and cloud digital twin model to realize the automatic reorganization of multi-satellite missions; The hybrid cloud interface module is configured to achieve seamless connection between satellite resources and multiple cloud platforms.

5. The satellite resource dynamic allocation system under the cloud native hybrid cloud architecture according to claim 4 is characterized in that: The resource prediction and scheduling module specifically includes: The spatiotemporal prediction submodule is configured to predict user needs based on historical task data and / or historical user needs using optimization algorithms or deep learning models; The scheduling submodule is configured to dynamically adjust the resource allocation plan based on the prediction results of the spatiotemporal prediction submodule, user needs, satellite resource status and other information.

6. The satellite resource dynamic allocation system under the cloud native hybrid cloud architecture according to claim 5 is characterized in that: The spatiotemporal prediction submodule is specifically used for: Construct a generator G, the input of which is a random noise sequence Z = {z1,z2,z1,…,z T }, the output is time series data Among them, θ G is the parameter of the generator G, z t is the random noise at time point t, T is the length of the sequence; Construct a discriminator D, the input of which is the real time series data X = {x1, x2, x1, ..., x T } or the time series data The output is the probability of the authenticity of the input data; where x t is the observation value at time point t, T is the length of the sequence; By alternatingly optimizing the discriminator loss function L D and the generator loss function L G Train the model, where: P data is the distribution of real data, P Z is the distribution of random noise; After the model training is completed, the generator G predicts the resource requirements of each beam coverage area in the future period based on the satellite mission historical data and real-time demand, and generates a resource matching plan.

7. The satellite resource dynamic allocation system under the cloud native hybrid cloud architecture according to claim 5 is characterized in that: The scheduling submodule is specifically used for: Initialize the policy network parameters θ and value network parameters φ; Predefined reinforcement learning elements, including state s t 、Action a t And the reward function r t , where state s t Contains satellite beam coverage information, channel allocation status, time slot allocation status and user service requirements; Action a t Represented as a dynamic adjustment instruction for beam radius, channel allocation or time slot scheduling; reward function r t Comprehensive calculation based on task completion rate, resource utilization and data transmission quality; The objective function is constructed as: in, is the advantage function, is the state and action value function, is the value network estimated at state s t The expected return under θ (a t |s t ) means in state s t Next take action a t The probability of is the probability of the old strategy, ∈ is the parameter of the clipping range, L VF (φ) is the loss function of the value function, L ENT (θ) is the entropy regularization term, c1 and c2 are weight coefficients, is the expected value; Update the policy network parameters θ by maximizing the objective function L(θ); By minimizing the loss function L of the value function VF (φ) Update the value network parameter φ.

8. The satellite resource dynamic allocation system under the cloud native hybrid cloud architecture according to claim 4 is characterized in that: The cloud native management module specifically includes: The data acquisition submodule is configured to connect sensors and devices in the satellite constellation and hybrid cloud through IoT technology. It collects data on satellite location, status, mission execution, and cloud resource usage in real time, calculates the constellation topology, and obtains global satellite status and global resource usage, which are then sent to the multi-source data fusion submodule. a communication monitoring submodule configured to collect bandwidth and delay of mutual access between the entire constellation topology, generate a delay matrix between satellites, and send it to the multi-source data fusion submodule; The multi-source data fusion submodule is configured to collect information from the data acquisition submodule and the communication monitoring submodule, and generate a dynamic panoramic resource view of the satellite constellation and hybrid cloud based on the global resource usage, the global satellite status, and the delay matrix; The simulation submodule is configured to simulate the operation of satellite constellations and hybrid clouds in different scenarios through simulation technology; A real-time feedback submodule is configured to compare the simulation results of the simulation submodule with actual operation data and adjust satellite missions and cloud resource configurations in real time; The strategy preview submodule is configured to allow users to test and verify different scheduling strategies in a virtual environment.

9. The satellite resource dynamic allocation system under the cloud native hybrid cloud architecture according to claim 4, characterized in that: The hybrid cloud interface module specifically includes: The cloud platform docking submodule is responsible for docking with various mainstream cloud platforms to achieve resource integration and management; The elastic expansion submodule is configured to dynamically adjust the scale of cloud computing resources according to resource demand to meet the real-time needs of tasks; The security isolation submodule is configured to ensure the security of resources and data between the satellite system and the cloud platform, and prevent data leakage and malicious attacks.

10. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for dynamically allocating satellite resources under a cloud-native hybrid cloud architecture according to any one of claims 1 to 3 is implemented.

Images