Joint allocation method and device for computing power and communication resources, equipment and medium

Abstract

The embodiment of the invention provides a computing power and communication resource joint allocation method and device, equipment and a medium. The method comprises the following steps: acquiring calculation requests of a plurality of users, and determining a plurality of calculation tasks according to the calculation requests; wherein each calculation task comprises a plurality of sub-tasks, and the sub-tasks are mutually independent; obtaining a satellite-ground transmission rate and an inter-satellite transmission rate of each subtask, obtaining a calculation time delay corresponding to each subtask according to the satellite-ground transmission rate and the inter-satellite transmission rate, and obtaining a target time delay value of the calculation task according to the calculation time delay; and according to the target time delay value, establishing an objective function for joint allocation of the computing power and the communication resources, and calculating the objective function to obtain a computing power and communication resource allocation strategy. The method is used for carrying out cooperative calculation in combination with multiple satellites, and the technical effect of reasonable allocation of satellite computing power resources and communication resources is achieved.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a method, apparatus, device and medium for the joint allocation of computing power and communication resources. Background Technology

[0002] With the increasing demand for satellite network services and the surge in data volume, traditional satellite-to-ground transmission and remote cloud computing models are no longer sufficient to meet real-time and bandwidth requirements. To achieve efficient data processing and transmission, the rapid development of aerospace electronics technology has led to continuous improvements in the performance of satellite-borne central processing units, making on-orbit computing and processing possible. While on-orbit computing can reduce the long response latency caused by long-distance transmission and enhance system reliability, most current satellite services employ an architecture of independent computing on a single satellite with coupled control and forwarding. This architecture is ill-suited to meet the demands of future computationally intensive services.

[0003] Research has found that by utilizing a multi-satellite collaborative mechanism, the computing resources of multiple satellites can be integrated to jointly complete a single computing task. Secondly, employing software-defined networking and network function virtualization (NFV) can abstract computing and communication resources into a unified resource pool, thereby achieving joint allocation. A computing task can be viewed as one or more virtualized network elements to be allocated. Therefore, the resource requirements of a computing task are mapped to the resource requirements of the virtualized network elements, and the network meets the operational needs of these virtualized network elements by allocating appropriate bandwidth and CPU resources. Thus, in low-Earth orbit satellite networks, the resource allocation problem for compute-intensive services can be transformed into the problem of allocating resources to virtualized network elements.

[0004] However, most current research on the allocation of virtualized network element resources focuses on terrestrial data centers and has not yet been systematically explored in conjunction with low-Earth orbit satellite networks. Therefore, how to combine multi-satellite collaborative computing to achieve a reasonable allocation of satellite computing and communication resources has become an urgent problem to be solved. Summary of the Invention

[0005] This application provides a method, apparatus, device, and medium for the joint allocation of computing power and communication resources, which is used to combine multiple satellites for collaborative computing and achieve the technical effect of rational allocation of satellite computing power and communication resources.

[0006] In a first aspect, embodiments of this application provide a method for joint allocation of computing power and communication resources. This method is applied to control equipment in a ground control center corresponding to a low-Earth orbit (LEO) satellite network system. The LEO satellite network system includes multiple LEO satellites, multiple ground users, and a ground control center. Each LEO satellite includes multiple interaction nodes and multiple computing nodes. The interaction nodes and computing nodes are connected via inter-satellite links, and the ground control center is connected to the interaction nodes via a satellite-to-ground link. The method includes:

[0007] The system acquires computing requests from multiple users and determines multiple computing tasks based on these requests. Each computing task includes multiple subtasks, and each subtask is independent of the others.

[0008] Obtain the satellite-to-ground transmission rate and inter-satellite transmission rate for each subtask. Based on the satellite-to-ground transmission rate and inter-satellite transmission rate, obtain the computational latency corresponding to each subtask. Based on the computational latency, obtain the target latency value of the computational task.

[0009] Based on the target latency value, an objective function for the joint allocation of computing power and communication resources is established, and the objective function is calculated to obtain the computing power and communication resource allocation strategy.

[0010] In one possible implementation, the satellite-to-ground transmission rate and inter-satellite transmission rate of each subtask are obtained. Based on the satellite-to-ground transmission rate and inter-satellite transmission rate, the computation delay corresponding to each subtask is obtained. Based on the computation delay, the delay value of the computation task is obtained, including:

[0011] Obtain the transmit antenna gain of the ground base station and the receive antenna gain of the interaction node in the ground control center. Based on the transmit antenna gain of the ground base station and the receive antenna gain of the interaction node in the ground control center, obtain the first signal-to-noise ratio of the satellite-to-ground link. Based on the first signal-to-noise ratio, obtain the satellite-to-ground transmission rate of the subtask.

[0012] Obtain the transmit antenna gain and receive gain of the low-Earth orbit satellite. Based on the transmit antenna gain and receive gain, obtain the second signal-to-noise ratio of the inter-satellite link. Based on the second signal-to-noise ratio, obtain the inter-satellite transmission rate of the sub-task.

[0013] Based on the satellite-to-ground transmission rate and the inter-satellite transmission rate, the computational delay corresponding to each subtask is obtained, and the planned delay of each subtask is compared to obtain the target delay value of the computation task.

[0014] In one possible implementation, an objective function for the joint allocation of computing power and communication resources is established based on the target latency value, and the objective function is calculated to obtain a computing power and communication resource allocation strategy, including:

[0015] The computing power of each computing node and the transmission bandwidth of the inter-satellite link are obtained. The computing power and transmission bandwidth are processed respectively to obtain the resource constraints corresponding to each subtask.

[0016] Based on the target latency value and the preset latency threshold, determine the latency constraints for each subtask;

[0017] Based on the target latency value, resource constraints, and latency constraints, an objective function for the joint allocation of computing power and communication resources is established.

[0018] The objective function is decoupled into a first subproblem and a second subproblem. The first subproblem indicates the location where each subtask in the computing task is offloaded to the computing node and the corresponding inter-satellite link routing subproblem. The second subproblem indicates the computing power allocation subproblem of the computing node.

[0019] The first and second subproblems are processed alternately and iteratively to obtain the computing power and communication resource allocation strategy.

[0020] In one possible implementation, the first and second subproblems are computed alternately to obtain a computing power and communication resource allocation strategy, including:

[0021] The computing power of the computing nodes is initialized, and the first subproblem is processed using a dynamic programming algorithm to obtain the location where each subtask in the computing task is unloaded to the computing node and the corresponding inter-satellite link route.

[0022] Based on the location of each subtask in the computing task unloaded to the computing node and the corresponding inter-satellite link route, the second sub-problem is processed to obtain the computing power value used to compute each subtask.

[0023] Based on the computing power value, the first and second subproblems are solved alternately again, and it is determined whether the preset convergence state is met. If so, the computing power and communication resource allocation strategy is obtained.

[0024] In one possible implementation, a dynamic programming algorithm is used to process the first subproblem to obtain the location where each subtask in the computation task is unloaded to the computation node and the corresponding inter-satellite link route, including:

[0025] For each computation task, the inter-satellite links are processed based on pre-set path rules to determine all inter-satellite link routes that meet the target latency value;

[0026] For any inter-satellite link route, offload any subtask to any computing node location along the current inter-satellite link route;

[0027] If the time delay constraint and resource constraint are satisfied when the subtask is unloaded to the computing node location, then the computing node location that satisfies the conditions and the current inter-satellite link route are recorded as a state set.

[0028] Based on the set of states, the target state set for multiple computational tasks is obtained.

[0029] In one possible implementation, the second sub-problem is processed to obtain the computing power value for calculating each sub-task, including:

[0030] The second subproblem is standardized to obtain the standardized result.

[0031] Based on pre-set constraint factors, the standardized processing results are processed to obtain the computing power value used to calculate each subtask.

[0032] Secondly, embodiments of this application provide a device for joint allocation of computing power and communication resources, comprising:

[0033] The acquisition module is used to acquire computing requests from multiple users and determine multiple computing tasks based on the computing requests; each computing task includes multiple subtasks, and each subtask is independent of the others.

[0034] The processing module is used to obtain the satellite-to-ground transmission rate and inter-satellite transmission rate of each subtask, obtain the calculation delay corresponding to each subtask based on the satellite-to-ground transmission rate and inter-satellite transmission rate, and obtain the target delay value of the calculation task based on the calculation delay.

[0035] The calculation module is used to establish an objective function for the joint allocation of computing power and communication resources based on the target latency value, and to calculate the objective function to obtain the computing power and communication resource allocation strategy.

[0036] In one possible implementation, the processing module is further configured to:

[0037] Obtain the transmit antenna gain of the ground base station and the receive antenna gain of the interaction node in the ground control center. Based on the transmit antenna gain of the ground base station and the receive antenna gain of the interaction node in the ground control center, obtain the first signal-to-noise ratio of the satellite-to-ground link. Based on the first signal-to-noise ratio, obtain the satellite-to-ground transmission rate of the subtask.

[0038] Obtain the transmit antenna gain and receive gain of the low-Earth orbit satellite. Based on the transmit antenna gain and receive gain, obtain the second signal-to-noise ratio of the inter-satellite link. Based on the second signal-to-noise ratio, obtain the inter-satellite transmission rate of the sub-task.

[0039] Based on the satellite-to-ground transmission rate and the inter-satellite transmission rate, the computational delay corresponding to each subtask is obtained, and the planned delay of each subtask is compared to obtain the target delay value of the computation task.

[0040] In one possible implementation, the computing module is also used for:

[0041] The computing power of each computing node and the transmission bandwidth of the inter-satellite link are obtained. The computing power and transmission bandwidth are processed respectively to obtain the resource constraints corresponding to each subtask.

[0042] Based on the target latency value and the preset latency threshold, determine the latency constraints for each subtask;

[0043] Based on the target latency value, resource constraints, and latency constraints, an objective function for the joint allocation of computing power and communication resources is established.

[0044] The objective function is decoupled into a first subproblem and a second subproblem. The first subproblem indicates the location where each subtask in the computing task is offloaded to the computing node and the corresponding inter-satellite link routing subproblem. The second subproblem indicates the computing power allocation subproblem of the computing node.

[0045] The first and second subproblems are processed alternately and iteratively to obtain the computing power and communication resource allocation strategy.

[0046] In one possible implementation, the computing module is also used for:

[0047] The computing power of the computing nodes is initialized, and the first subproblem is processed using a dynamic programming algorithm to obtain the location where each subtask in the computing task is unloaded to the computing node and the corresponding inter-satellite link route.

[0048] Based on the location of each subtask in the computing task unloaded to the computing node and the corresponding inter-satellite link route, the second sub-problem is processed to obtain the computing power value used to compute each subtask.

[0049] Based on the computing power value, the first and second subproblems are solved alternately again, and it is determined whether the preset convergence state is met. If so, the computing power and communication resource allocation strategy is obtained.

[0050] In one possible implementation, the computing module is also used for:

[0051] For each computation task, the inter-satellite links are processed based on pre-set path rules to determine all inter-satellite link routes that meet the target latency value;

[0052] For any inter-satellite link route, offload any subtask to any computing node location along the current inter-satellite link route;

[0053] If the time delay constraint and resource constraint are satisfied when the subtask is unloaded to the computing node location, then the computing node location that satisfies the conditions and the current inter-satellite link route are recorded as a state set.

[0054] Based on the set of states, the target state set for multiple computational tasks is obtained.

[0055] In one possible implementation, the computing module is also used for:

[0056] The second subproblem is standardized to obtain the standardized result.

[0057] Based on pre-set constraint factors, the standardized processing results are processed to obtain the computing power value used to calculate each subtask.

[0058] Thirdly, embodiments of this application provide a device for the joint allocation of computing power and communication resources, including: a memory and a processor;

[0059] The memory stores computer-executed instructions;

[0060] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0061] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0062] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0063] This application provides a method, apparatus, device, and medium for the joint allocation of computing power and communication resources. By acquiring computing requests from multiple users and identifying multiple computing tasks, each computing task is divided into multiple independent sub-tasks, achieving fine-grained partitioning. Since the transmission rates of satellite-to-ground links and inter-satellite links in the low-Earth orbit (LEO) satellite network significantly impact the completion time of computing tasks, the computing latency of each sub-task is obtained based on the transmission rates of these links, thus yielding the target latency value for the entire computing task. By establishing an objective function for the joint allocation of computing power and communication resources based on the target latency value, the allocation of computing power and communication resources can be considered holistically, avoiding resource waste caused by optimizing computing power or communication resources individually. This achieves synergistic optimization of computing power and communication resources, which helps improve resource utilization and maximizes the performance of the LEO satellite network system while meeting the latency requirements of computing tasks. Attached Figure Description

[0064] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0065] Figure 1 An application scenario diagram of the computing power and communication resource joint allocation method provided in the embodiments of this application;

[0066] Figure 2 A flowchart illustrating a method for jointly allocating computing power and communication resources provided in this application embodiment. Figure 1 ;

[0067] Figure 3This is a schematic diagram of the structure of a computing task model provided in an embodiment of this application;

[0068] Figure 4 A flowchart illustrating a method for jointly allocating computing power and communication resources provided in this application embodiment. Figure 2 ;

[0069] Figure 5 A simulation graph showing the change in the reception rate of computing tasks as a function of the number of computing tasks, provided in an embodiment of this application.

[0070] Figure 6 A simulation diagram showing the change in response latency of a computing task as a function of the number of computing tasks, provided in an embodiment of this application.

[0071] Figure 7 A schematic diagram of the structure of the computing power and communication resource joint allocation device provided in the embodiments of this application;

[0072] Figure 8 This is a schematic diagram of the structure of the computing power and communication resource joint allocation device provided in the embodiments of this application.

[0073] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0074] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0075] With the continuous growth in demand for satellite network services, the amount of data generated by these services has increased dramatically. The traditional model of transmitting data via satellite to cloud computing centers is no longer sufficient to fully meet user needs. Specifically, this manifests in several ways: long transmission delays exist between satellite and ground stations, as well as between satellites and remote cloud centers, directly leading to significantly extended service response times and marked deficiencies in real-time performance; satellite backhaul links already bear a heavy burden, and transmitting computing service data further exacerbates bandwidth constraints; long-distance transmission reduces service reliability and security to some extent, posing potential risks to the entire satellite network service.

[0076] Against this backdrop, satellite networks should not merely act as relays, but also undertake a significant amount of data computation. Thanks to the rapid development of aerospace electronics technology, the performance of central processing units (CPUs) on satellites has continuously improved, making on-orbit computing and onboard processing technically possible. In particular, on-orbit computing can effectively reduce the response latency caused by long-distance transmission, thereby improving system reliability.

[0077] However, the scarcity of onboard computing resources remains a pressing issue that needs to be addressed. Satellite central processing units (CPUs) need to handle numerous tasks, including sensing, control, and maintenance, resulting in very limited computing resources available for computation. Therefore, a reasonable collaborative mechanism, an open architecture, and effective resource allocation methods are crucial. However, most current satellite services employ a single-satellite architecture that couples independent computing with control and forwarding. This architecture will struggle to meet the demands of future computationally intensive services.

[0078] Research has found that multi-satellite collaboration mechanisms can integrate the computing resources of multiple satellites to jointly complete computing tasks. Secondly, the introduction of software-defined networking (SDN) and network function virtualization (NFV) paradigms allows for the construction of a more flexible and open architecture. SDN and NFV paradigms can abstract computing and communication resources into a unified resource pool for allocation. A computing task can be viewed as one or more virtualized network elements to be allocated, the number of which is determined by the service model. Allocating bandwidth and CPU resources to a computing task is equivalent to allocating resources to NFV. Therefore, in low-Earth orbit (LEO) satellite networks, the resource allocation problem for computationally intensive services can be transformed into the problem of allocating resources to virtualized network elements.

[0079] However, most current research on the allocation of virtualized network element resources focuses on terrestrial data centers and has not yet been systematically explored in conjunction with low-Earth orbit satellite networks. Therefore, how to combine multi-satellite collaborative computing to achieve a reasonable allocation of satellite computing and communication resources has become an urgent problem to be solved.

[0080] To address the aforementioned issues, this application provides a method for jointly allocating computing and communication resources. This method acquires computing requests from multiple users and identifies multiple computing tasks, dividing each task into several independent sub-tasks for fine-grained management. Since the transmission rates of satellite-to-ground and inter-satellite links in low-Earth orbit (LEO) satellite networks significantly impact the completion time of computing tasks, the computational latency of each sub-task is determined based on these transmission rates, leading to the target latency value for the entire computing task. By establishing an objective function for the joint allocation of computing and communication resources based on the target latency value, the allocation of computing and communication resources can be considered holistically, avoiding resource waste caused by optimizing computing or communication resources in isolation. This achieves synergistic optimization of computing and communication resources, improving resource utilization and maximizing the performance of the LEO satellite network system while meeting the latency requirements of computing tasks.

[0081] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0082] Figure 1 This is an application scenario diagram of the computing power and communication resource joint allocation method provided in the embodiments of this application, such as... Figure 1As shown, the system corresponding to the joint allocation method of computing power and communication resources in this embodiment is a low-Earth orbit (LEO) satellite network system. This LEO satellite network system includes space-based and ground-based components. The ground-based component includes a ground control center and multiple ground users. The ground control center includes control equipment used to receive computing requests from ground users and the usage status of LEO satellite resources to determine computing power and communication resource allocation strategies, thereby deciding on the offloading strategy for computing tasks. The space-based component consists of multiple LEO satellites in the same or different orbits, which can be divided into multiple interaction nodes and multiple computing nodes according to their functions. It should be noted that the interaction nodes and computing nodes are connected via inter-satellite links, and the ground control center is connected to the interaction nodes via a satellite-to-ground link. The interaction nodes are equipped with hardware devices such as sensors, central processing units, and storage devices, enabling them to sense the surrounding computing nodes and link load. They are also responsible for signaling interaction with the ground control center, sending the collected computing node load information to the ground control center, and forwarding the ground control center's signaling to the computing nodes within its management range. The computing nodes are equipped with advanced computing devices and chips, capable of performing calculations on user computing requests. In addition, the computing nodes are also equipped with software-defined network switches, which can also forward data. Optionally, the space-based system may also include forwarding nodes, which are equipped with software-defined network switches but not with high-performance computing equipment; they are only responsible for forwarding data and not for processing it.

[0083] For example, when a ground user is located in a remote area, unable to directly connect to a terrestrial mobile network, and has limited computing power to handle the large amounts of data generated by computing tasks, one possible implementation is as follows: the ground user can send the data generated by the computing task to the ground control center via user equipment or IoT devices. The control equipment in the ground control center determines the computing power and communication resource allocation strategy based on the ground user's computing request and the availability of low-Earth orbit satellite resources. The computing task is then sent to an interaction node, which in turn sends it to the corresponding computing node for computation. After the computing node completes the computation, it returns the result to the ground user's user equipment or IoT device via the interaction node and the ground control center. Another possible implementation involves the control equipment in the ground control center determining the computing power and communication resource allocation strategy based on the ground user's computing request and the availability of low-Earth orbit satellite resources, and sending this strategy to the ground user's user equipment or IoT device. The ground user then directly sends the computing task to the interaction node, which in turn sends it to the corresponding computing node for computation. After the computing node completes the computation, it returns the result to the ground user's user equipment or IoT device via the interaction node.

[0084] Figure 2 A flowchart illustrating a method for jointly allocating computing power and communication resources provided in this application embodiment. Figure 1 ,like Figure 2 As shown, this method is applied to control equipment in the ground control center corresponding to a low-Earth orbit satellite network system, including:

[0085] S201. Obtain computing requests from multiple users and determine multiple computing tasks based on the computing requests.

[0086] In this embodiment, the control equipment in the ground control center receives computing requests from multiple users and classifies them into different computing tasks based on the type of the computing request. Each computing task includes multiple subtasks, and each subtask is independent of the others. For example, subtasks include data processing tasks, communication tasks, orbit calculation tasks, etc.

[0087] Optional, such as Figure 3 As shown, for computationally intensive computing tasks requiring significant computational and processing power, such as parallel or distributed computing, a large computing task is decomposed into multiple independently executable subtasks. These subtasks have no data or sequential dependencies. Specifically, let K represent the set of computing tasks, where... If we represent a single computational task, then we have ,in Represents a single computational task Subtasks. Using sets This represents the set of data for the computation task. Subtasks Given the amount of data, the amount of data for each subtask in this computational task can be expressed as: Once all subtasks have been computed, the results can be combined. This typically involves integrating the output data from the subtasks to obtain the final computational result.

[0088] Optionally, after obtaining the subtasks based on the computation task, a six-tuple can be used. This is used to describe the user's computational task, in order to mathematically model the computational task. Among them, and These are two binary constants, used to represent the computational tasks. The starting point and the ending point. Represents computational task Is the starting point? If so, then ; Represents computational task Is the endpoint the If so, then . Represents computational task Number of subtasks , Represents computational task The The amount of data for each subtask express The Minimum computational requirements for each subtask express The longest service response time requirement, i.e., the computing task The required latency cannot exceed .

[0089] S202. Obtain the satellite-to-ground transmission rate and inter-satellite transmission rate for each subtask. Based on the satellite-to-ground transmission rate and inter-satellite transmission rate, obtain the computation delay corresponding to each subtask. Based on the computation delay, obtain the target delay value of the computation task.

[0090] It is understandable that the satellite-to-ground transmission rate refers to the data transmission rate between the ground control center and the interaction node, while the inter-satellite transmission rate refers to the data transmission rate between the interaction node and the computing node. The satellite-to-ground and inter-satellite transmission rates can be obtained from pre-tested system data or data obtained from actual on-orbit monitoring. The satellite-to-ground transmission rate is affected by various factors, such as the transmit antenna gain and power of the ground base station in the ground control center, and the receive antenna gain of the interaction node. The inter-satellite transmission rate is also affected by factors such as the distance between the interaction node and the computing node, the communication frequency band, transmit power, and receive sensitivity.

[0091] Optionally, the transmit antenna gain of the ground base station and the receive antenna gain of the interaction node in the ground control center are obtained. Based on these gains, the first signal-to-noise ratio (SNR) of the satellite-to-ground link is calculated, and the satellite-to-ground transmission rate of the sub-task is obtained. The transmit antenna gain and receive antenna gain of the low-Earth orbit satellite are obtained. Based on these gains, the second SNR of the inter-satellite link is calculated, and the inter-satellite transmission rate of the sub-task is obtained. Based on the satellite-to-ground transmission rate and the inter-satellite transmission rate, the computational delay corresponding to each sub-task is obtained, and the calculated delays of each sub-task are compared to obtain the target delay value for the computational task.

[0092] In this embodiment, the R&D user can pre-build a low-Earth orbit (LEO) satellite network topology model based on a pre-constructed LEO satellite network system. Specifically, the entire LEO satellite network system deploying a terrestrial mobile communication network can be abstracted as an undirected graph. ,in, Used to represent all nodes and links in an undirected graph. Used to represent all nodes and links in an undirected graph. Represents a set of low-Earth orbit satellite nodes. Represents the set of ground users. Represents the set of inter-satellite links. This represents the set of satellite-to-ground links.

[0093] Based on this, the transmit antenna gain of the ground base station and the receive antenna gain of the interaction node in the ground control center can be directly obtained from the equipment specifications or measured using dedicated testing equipment. (Based on the transmit antenna gain of the ground base station in the ground control center...) and the receiving gain of the interactive node antenna The first signal-to-noise ratio of the satellite-to-ground link is obtained, and the specific formula is as follows:

[0094] (1)

[0095] in, This refers to the transmit power of the ground base station's transmitting antenna in the ground control center. It is random noise. Other noise interference items, among which, Indicates the source of interference The channel gain coefficient, Indicates the time-varying channel state. This represents the transmission loss of the satellite-to-ground link. Transmission loss can be expressed as free path loss:

[0096] (2)

[0097] d(g,v) represents the computational task. From the initiating user to the interaction node The distance between them; The center frequency represents the carrier frequency used for satellite-to-ground transmission.

[0098] Based on the first signal-to-noise ratio, the satellite-to-ground transmission rate of the sub-task is obtained as follows:

[0099] (3)

[0100] in, This refers to the transmission bandwidth.

[0101] When considering inter-satellite links, obtain the transmit antenna gain of low-Earth orbit satellites. and receive gain Based on the transmit antenna gain and receive gain, the second signal-to-noise ratio of the inter-satellite link is obtained:

[0102] (4)

[0103] in, This refers to the antenna transmit power of the low-Earth orbit satellite. It should be noted that when a subtask is sent to the computing node for computation, it is sent from the interaction node to the computing node. This refers to the antenna transmit power of the interaction node; after the subtask calculation is completed, it is sent from the computing node to the interaction node. To calculate the antenna transmit power of the node; For noise power, For interactive nodes and computing nodes The distance between them. Furthermore, the noise power can be expressed as:

[0104] (5)

[0105] Where, k B B is the Boltzmann constant. c The effective noise bandwidth is T, where T is the thermodynamic ambient temperature at which the noise is generated.

[0106] And order:

[0107] (6)

[0108] in The center frequency represents the carrier frequency used for inter-satellite transmission; It is the speed of light. Therefore, the second signal-to-noise ratio of the inter-satellite link is expressed as:

[0109] (7)

[0110] Based on the second signal-to-noise ratio, the rate of the subtask is obtained:

[0111] (8)

[0112] in, This refers to the transmission bandwidth for inter-satellite transmission.

[0113] Optional, for each subtask The computation latency depends not only on the amount of data to be computed, but also on the computation density and the computing power of the central processing unit on the computing node, which can be specifically expressed as:

[0114] (9)

[0115] in, Represents computational density, indicating the number of CPU cores required to compute each bit, expressed in bits per core. This is a continuous variable, representing the number of CPU cores allocated to this subtask. Furthermore, each subtask has a minimum computational requirement, also expressed in terms of the number of CPU cores.

[0116] Optional, for each subtask The computational latency includes both uplink and downlink latency. Uplink latency includes transmission and propagation delays between the ground control center and the interaction node, and between the interaction node and the computing node. Downlink latency includes transmission and propagation delays between the computing node and the interaction node, and between the interaction node and the ground control center. Therefore, the computational latency of the subtask can be expressed as:

[0117] (10)

[0118] in, For satellite-to-ground propagation delay, and These are two binary variables; the first represents the inter-satellite link. Was it used for transmission? The Data from each subtask is transmitted to the computing node, which represents the inter-satellite link. Was it used for transmission? The The data from each subtask is returned to the interaction node. Indicates the inter-satellite link between the interaction node and the computing node. The propagation delay between them. For ease of calculation, consider the scenario where the interaction nodes in the uplink and downlink overlap. Furthermore, the computation delay of the subtask can be expressed as:

[0119] (11)

[0120] For computational tasks After each subtask is assigned to a computing node, the subtasks are processed synchronously, thus the computing tasks... The target latency value is equal to the computation latency of its longest-running subtask, which is the computation latency of the subtask assigned to the computation node farthest from the interaction node:

[0121] (12)

[0122] S203. Based on the target latency value, establish an objective function for the joint allocation of computing power and communication resources, and calculate the objective function to obtain the computing power and communication resource allocation strategy.

[0123] In this embodiment, an objective function for the joint allocation of computing power and communication resources is established based on the target latency value. The purpose of this objective function is to optimize the allocation of computing power and communication resources while meeting the target latency value of the computing task. For example, the objective function includes minimizing computing power resources, minimizing the use of communication resources (such as reducing the bandwidth consumption of satellite-to-ground and inter-satellite links), and ensuring that all subtasks can be completed within the target latency value. The objective function includes variables representing computing power resources (such as the performance of the central processing unit on the computing node) and communication resources (satellite-to-ground and inter-satellite transmission rates, bandwidth, etc.), and their relationship with the computing latency. The established objective function can be calculated using optimization algorithms. Through calculation, the optimal computing power and communication resource configuration strategy is obtained. This strategy explicitly specifies how to allocate computing power and communication resources among different subtasks and different computing nodes, and how to allocate satellite-to-ground and inter-satellite communication bandwidth to meet the needs of each subtask, thereby ensuring that the entire computing task can be completed efficiently within the target latency value.

[0124] The computing power and communication resource joint allocation method provided in this application allows the ground control center to accurately grasp the computing needs of the entire low-Earth orbit satellite network system by acquiring computing requests from multiple users and identifying multiple computing tasks. This facilitates targeted resource allocation for specific sub-tasks, improving the accuracy of resource allocation. Dividing each computing task into multiple independent sub-tasks facilitates fine-grained management of complex computing tasks. Calculating the corresponding computing latency by acquiring the satellite-to-ground transmission rate and inter-satellite transmission rate for each sub-task fully considers the communication link characteristics of the low-Earth orbit satellite network system. Since the transmission rates of satellite-to-ground and inter-satellite links in the low-Earth orbit satellite network significantly impact the completion time of computing tasks, this method allows for accurate estimation of the computing latency of each sub-task, thereby obtaining the target latency value for the entire computing task. By establishing an objective function for the joint allocation of computing power and communication resources based on the target latency value, the allocation of computing power and communication resources can be considered holistically. This avoids the waste of resources caused by optimizing computing power or communication resources separately, and achieves synergistic optimization of computing power and communication resources. This helps to improve resource utilization and maximize the performance of the low-Earth orbit satellite network system while meeting the latency requirements of computing tasks. For example, it can handle more computing tasks or improve the processing speed of computing tasks with the same total amount of resources.

[0125] Figure 4 A flowchart illustrating a method for jointly allocating computing power and communication resources provided in this application embodiment. Figure 2 ,like Figure 4 As shown, in this embodiment... Figure 2Based on the embodiments, a detailed explanation is provided regarding the establishment of an objective function for the joint allocation of computing power and communication resources according to the target latency value, and the calculation of the objective function to obtain the computing power and communication resource allocation strategy. This method includes:

[0126] S401. Obtain the computing power of each computing node and the transmission bandwidth of the inter-satellite link, process the computing power and transmission bandwidth respectively, and obtain the resource constraints corresponding to each subtask.

[0127] In this embodiment, the computing power of a computing node refers to its ability to execute subtasks, which depends on the node's hardware configuration, such as the model and number of cores of its central processing unit (CPU). The computing power of a computing node can be directly obtained from the model and performance parameters of the central server mounted on it, or by using benchmarking tools to measure the node's performance. The transmission bandwidth of the inter-satellite link refers to the frequency bandwidth resources provided by the communication link between the interacting node and the computing node. It depends on the design of the low-Earth orbit satellite communication system, including the operating frequency band and bandwidth settings of the communication equipment on the satellite. During the planning and design phase of the satellite communication system, the transmission bandwidth of the inter-satellite link is determined based on mission requirements (such as data transmission volume and transmission rate requirements). These parameters can be obtained from the overall design document of the satellite communication system. Simultaneously, during satellite operation in orbit, the actual available transmission bandwidth can also be obtained by monitoring the inter-satellite link. For each subtask, the computing power of the computing node and the transmission bandwidth resource constraints of the inter-satellite link are comprehensively considered.

[0128] Optionally, the allocation of inter-satellite computing tasks is constrained by computing power resources and the transmission bandwidth resources of inter-satellite links. The computing power consumed by the computing tasks undertaken by a computing node should not exceed its own tolerance, specifically:

[0129] (13)

[0130] in, It is a binary variable representing a computational task. The Are individual subtasks deployed on compute nodes? superior.

[0131] When the total bandwidth occupied by transmission tasks assigned to a link cannot exceed the maximum bandwidth of that link, we have:

[0132] (14)

[0133] in, express subtasks Required transmission bandwidth Interstellar links The capacity.

[0134] S402. Determine the latency constraints for each subtask based on the target latency value and the preset latency threshold.

[0135] In this embodiment, the preset latency threshold is a hard time limit. If this threshold is exceeded, the execution of the subtask will cause the entire computing task to fail to meet the requirements.

[0136] Optional, computational tasks The target latency value should not exceed its maximum tolerable limit. ,have:

[0137] (15)

[0138] Optionally, in addition to the resource and latency constraints mentioned above, subtasks are also subject to deployment constraints. These deployment constraints are as follows: First, since subtasks do not support further splitting, each subtask can and can only select one computing node for computation, specifically as follows:

[0139] (16)

[0140] in, It is a binary variable representing a computational task. The Are individual subtasks deployed on compute nodes? Up. Using binary variables. Represents computational task The deployment is considered successful if and only if all subtasks are deployed successfully. Only when the deployment is considered successful are the following conditions met:

[0141] (17)

[0142] During transmission, the data streams related to the subtasks need to be forwarded to the computing nodes via the interaction nodes. Indicates link Was it used for computational tasks? The Data transmission starting point for each subtask Then we have the formula:

[0143] (18)

[0144] Subtask The data needs to be routed from the interaction node to the selected computing node for computation, where Indicated by The set of links within one hop from which the origin is located. Indicated by This is the set of links within one hop of the destination. After the computing node completes the computation, it needs to return the result to the user. Indicates link Was it used to transfer computing tasks? The For each sub-task, the formula is:

[0145] (19)

[0146] For ease of calculation, and They actually overlap.

[0147] S403. Based on the target latency value, resource constraints, and latency constraints, establish an objective function for the joint allocation of computing power and communication resources.

[0148] In this embodiment, the objective function is designed to optimize the joint allocation of computing power and communication resources to achieve optimal system performance. This joint allocation method can adapt more flexibly to changes when facing tasks of different scales and complexities. When a new task is added or the task requirements change, the allocation of computing power and communication resources can be readjusted according to new target latency values, resource constraints, etc.

[0149] Optionally, based on the target latency value, resource constraints, and latency constraints, an objective function for the joint allocation of computing and communication resources is established, which maximizes the system's revenue, as follows:

[0150]

[0151] in, The overhead weighting factor is the target latency value. The weighting factor for the benefits brought by business success. The objective function aims to minimize the target latency value while maximizing the number of acceptable computational tasks.

[0152] S404. Decouple the objective function into a first subproblem and a second subproblem.

[0153] In this embodiment, since the objective function is a non-convex optimization problem, inter-satellite service allocation, routing, and intra-satellite computing power allocation are coupled, and communication and computing resources require joint decision-making, making it difficult to find the optimal solution for the objective function in polynomial time. Therefore, to simplify the calculation process, the objective function can be divided into a first subproblem and a second subproblem. The first subproblem indicates the location where each subtask in the computing task is offloaded to the computing node and the corresponding inter-satellite link routing subproblem, while the second subproblem indicates the computing power allocation subproblem of the computing node.

[0154] S405. Initialize the computing power value of the computing nodes. For each computing task, process the inter-satellite links based on the pre-set path rules to determine all inter-satellite link routes that meet the target latency value.

[0155] In this embodiment, the pre-set path rules can be formulated based on the low-Earth orbit satellite network topology model. For example, if the low-Earth orbit satellite network topology model is a mesh topology, it can be stipulated that the path selection from the source node to the target node should follow the shortest path principle (measured by link hop count or transmission delay).

[0156] For example, for each computational task To simplify the search space, we can first use a weighted shortest path algorithm to obtain all inter-satellite link routes that satisfy the target latency value. All subtasks of this computational task Will be uninstalled first The same route On the router, offloading should only be considered if resources are scarce on that router. On another route.

[0157] S406. For any inter-satellite link route, offload any subtask to any computing node location along the current inter-satellite link route. If the time delay constraint and resource constraint are satisfied when the subtask is offloaded to the computing node location, then the computing node locations that satisfy the conditions and the current inter-satellite link route are recorded as a state set.

[0158] In this embodiment, for any inter-satellite link route, there are multiple computing nodes in the inter-satellite link route. Any subtask is unloaded to all computing node positions along the current inter-satellite link route, and it is determined at all computing node positions whether the subtask meets the latency constraint and resource constraint conditions. If it does, the corresponding computing node position and the current inter-satellite link route are recorded as a state set.

[0159] For example, for the selected Unload subtasks one by one to the computing nodes along their path. Above. For any If the subtask is assigned to If the allocation strategy can satisfy its latency and resource constraints, then this allocation strategy is denoted as a state. ; Traversal By considering all computational nodes, all state sets can be obtained. In one possible implementation, for ease of computation, only the M states with the smallest inter-satellite link routing hop counts can be retained.

[0160] S407. Based on the state set, obtain the target state set for multiple computational tasks.

[0161] In this embodiment, when a subtask is unloaded, the unloading process of the previous subtask is repeated when unloading the next subtask to obtain the current subtask's state set, until the state set of the entire computing request is obtained, that is, the target state set of multiple computing tasks. This target state set is used to represent the location of each subtask in the computing task unloaded to the computing node and the corresponding inter-satellite link route.

[0162] S408. Based on the location of each subtask in the computing task unloaded to the computing node and the corresponding inter-satellite link route, the second sub-problem is standardized to obtain the standardized processing result. Based on the pre-set constraint factor, the standardized processing result is further processed to obtain the computing power value used to calculate each subtask.

[0163] In this embodiment, the objective function is first equivalently represented as:

[0164]

[0165] When the location of each subtask in the computing task is unloaded to the computing node and the corresponding inter-satellite link route is determined, the above problem is transformed into:

[0166]

[0167] Next, the above issues will be standardized:

[0168]

[0169] Based on pre-set constraint factors, the standardized results are processed:

[0170]

[0171] Among them, variables and These are pre-set constraint factors, i.e., non-negative Lagrange multipliers.

[0172] For computational tasks The The optimal allocation of computing power for each subtask. And the optimal Lagrange multiplier and The Karush-Kuhn-Tucker (KKT) condition must be met, that is:

[0173]

[0174] Based on the above formula, the computing power value used to calculate each subtask is obtained. size.

[0175] S409. Based on the computing power value, solve the first subproblem and the second subproblem alternately.

[0176] S410. Determine whether the preset convergence state is met.

[0177] S411. If yes, obtain the computing power and communication resource allocation strategy; if no, repeat step S405.

[0178] In steps S409-S411, the preset convergence state can be defined based on the change in the objective function value. For example, if after multiple alternating solutions, the change in the solutions to the first and second subproblems is less than a preset threshold, then the convergence state has been reached; or after multiple alternating solutions, the solutions to the first and second subproblems remain essentially unchanged or change very little, then the convergence state has been reached. Once the convergence state is reached, the corresponding computing power and communication resource allocation strategy is obtained. This strategy indicates the location where each subtask is offloaded to a computing node, the inter-satellite link route, and the computing power allocated to that subtask by the computing node.

[0179] The computing power and communication resource joint allocation method provided in this application obtains the computing power of each computing node and the transmission bandwidth of the inter-satellite link, and processes them to obtain the resource constraints of subtasks. This helps to clarify the limitations of each subtask in terms of computing power and communication bandwidth, avoiding resource waste. The latency constraints of each subtask are determined based on the target latency value and a preset latency threshold, enabling the system to flexibly respond to the different latency requirements of different subtasks. An objective function for the joint allocation of computing power and communication resources is established based on the target latency value, resource constraints, and latency constraints. By decoupling the objective function into a first subproblem and a second subproblem, the complex joint allocation problem is decomposed into relatively simpler subproblems. This reduces the complexity of the problem and allows for more targeted algorithms and strategies to process the two subproblems separately. The computing power of the computing nodes is initialized, and inter-satellite link routes that meet the target latency value are determined based on preset path rules. By unloading subtasks to computing node locations and checking whether they meet latency and resource constraints, a set of states is selected that meets the requirements for computing node locations and inter-satellite link routes. This allows for the accurate identification of resource allocation and task unloading combinations that meet system requirements, providing foundational data for determining the overall optimal state of multiple computing tasks. The second subproblem is standardized, and the computing power value used to calculate each subtask is obtained based on constraint factors, making the computing power calculation more standardized and accurate. This helps to accurately determine the computing power required for each subtask based on specific network conditions and task requirements, providing an accurate basis for the rational allocation of resources. The optimal resource allocation scheme is gradually approximated by alternately solving the first and second subproblems. This iterative solution process can adapt to dynamic changes in the system and continuously adjust the resource allocation strategy to achieve better system performance.

[0180] Optionally, the computing power and communication resource joint allocation method proposed in the embodiments of this application is applied to a real-world scenario simulation. The specific simulation experiments and corresponding results are as follows:

[0181] In one possible implementation, the simulation parameters for the experiment are set as follows: A satellite topology containing 6 Iridium satellites is generated using the satellite toolkit. This topology includes 2 orbital planes, with 3 satellites distributed on each plane. The orbital plane inclination is 45°, the altitude is 780 km, and the constellation type is Walker. The inter-satellite link bandwidth is 100 Mbps. The number of subtasks in each computational task is randomly generated from [1,2], requiring a transmission bandwidth of [3,10] Mbps. The end-to-end latency requirement is uniformly distributed between 60 ms and 120 ms. Each subtask has a minimum computing power requirement of 2 cores and a maximum of 8 cores.

[0182] Figure 5The graphs showing the service acceptance rate as a function of the number of computed services are presented for five algorithms, including the one proposed in this invention. Figure 5 As shown in the figure, there are 5 curves, representing the following scenarios for the proposed algorithm (joint allocation algorithm of computing power and communication resources): fixed computing power allocation of 4, no allocation with fixed computing power allocation of 4, fixed computing power allocation of 8, random allocation, and allocation with fixed computing power allocation of 4 based on the proportion of computing power. Figure 5 It can be seen that the algorithm with a fixed computing power of 4 has the highest service acceptance rate, while the proposed algorithm has the second highest. This is because the smaller the fixed computing power, the more tasks it can accept while keeping the capacity constant. The proposed algorithm (the algorithm for joint allocation of computing power and communication resources) and the algorithm for random allocation of computing power have similar acceptance rates, while the algorithm with a fixed computing power of 8 has the lowest acceptance rate, further verifying the conclusion that the acceptance rate is affected by the lowest computing power.

[0183] Figure 6 The graphs showing the service response latency of five algorithms, including the one proposed in this invention, as a function of the number of computing services are presented. Figure 6 As shown, the curve with a fixed computing power of 4 has the longest response latency, exceeding that of the curve with the lowest response latency. This is because the computing power allocated to each subtask is relatively small, resulting in relatively large processing and transmission latency. Random allocation is highly random, making it impossible to flexibly allocate computing power based on satellite resource latency conditions and latency requirements, making it difficult to find a balance between service acceptance rate and response latency. The fixed computing power of 8 and the proposed algorithm have relatively low response latency because the computing power value of the fixed computing power algorithm is relatively large. However, the higher computing power value limits the number of satellite services, reducing the overall service acceptance rate. After fixing satellite capacity and minimum computing power requirements, there is a trade-off between service response latency and service acceptance rate. If a lower response latency is required, the number of services that the satellite can serve will inevitably be limited, leading to a decrease in the system's service acceptance rate. Conversely, improving the acceptance rate involves allocating as little computing power as possible to each task, allowing a single satellite to deploy multiple tasks. However, this significantly increases the computational latency of tasks, resulting in slow service response.

[0184] Figure 7 This is a schematic diagram of the structure of the computing power and communication resource joint allocation device provided in the embodiments of this application, as shown below. Figure 7 As shown, the computing power and communication resource joint allocation device 700 provided in this embodiment includes:

[0185] The acquisition module 701 is used to acquire computing requests from multiple users and determine multiple computing tasks based on the computing requests; wherein each computing task includes multiple subtasks, and each subtask is independent of the others.

[0186] The processing module 702 is used to obtain the satellite-to-ground transmission rate and inter-satellite transmission rate of each subtask, obtain the calculation delay corresponding to each subtask based on the satellite-to-ground transmission rate and inter-satellite transmission rate, and obtain the target delay value of the calculation task based on the calculation delay.

[0187] The calculation module 703 is used to establish an objective function for the joint allocation of computing power and communication resources based on the target latency value, and to calculate the objective function to obtain the computing power and communication resource allocation strategy.

[0188] In one possible implementation, the processing module 702 is further configured to:

[0189] Obtain the transmit antenna gain of the ground base station and the receive antenna gain of the interaction node in the ground control center. Based on the transmit antenna gain of the ground base station and the receive antenna gain of the interaction node in the ground control center, obtain the first signal-to-noise ratio of the satellite-to-ground link. Based on the first signal-to-noise ratio, obtain the satellite-to-ground transmission rate of the subtask.

[0190] Obtain the transmit antenna gain and receive gain of the low-Earth orbit satellite. Based on the transmit antenna gain and receive gain, obtain the second signal-to-noise ratio of the inter-satellite link. Based on the second signal-to-noise ratio, obtain the inter-satellite transmission rate of the sub-task.

[0191] Based on the satellite-to-ground transmission rate and the inter-satellite transmission rate, the computational delay corresponding to each subtask is obtained, and the planned delay of each subtask is compared to obtain the target delay value of the computation task.

[0192] In one possible implementation, the computing module 703 is further configured to:

[0193] The computing power of each computing node and the transmission bandwidth of the inter-satellite link are obtained. The computing power and transmission bandwidth are processed respectively to obtain the resource constraints corresponding to each subtask.

[0194] Based on the target latency value and the preset latency threshold, determine the latency constraints for each subtask;

[0195] Based on the target latency value, resource constraints, and latency constraints, an objective function for the joint allocation of computing power and communication resources is established.

[0196] The objective function is decoupled into a first subproblem and a second subproblem. The first subproblem indicates the location where each subtask in the computing task is offloaded to the computing node and the corresponding inter-satellite link routing subproblem. The second subproblem indicates the computing power allocation subproblem of the computing node.

[0197] The first and second subproblems are processed alternately and iteratively to obtain the computing power and communication resource allocation strategy.

[0198] In one possible implementation, the computing module 703 is further configured to:

[0199] The computing power of the computing nodes is initialized, and the first subproblem is processed using a dynamic programming algorithm to obtain the location where each subtask in the computing task is unloaded to the computing node and the corresponding inter-satellite link route.

[0200] Based on the location of each subtask in the computing task unloaded to the computing node and the corresponding inter-satellite link route, the second sub-problem is processed to obtain the computing power value used to compute each subtask.

[0201] Based on the computing power value, the first and second subproblems are solved alternately again, and it is determined whether the preset convergence state is met. If so, the computing power and communication resource allocation strategy is obtained.

[0202] In one possible implementation, the computing module 703 is further configured to:

[0203] For each computation task, the inter-satellite links are processed based on pre-set path rules to determine all inter-satellite link routes that meet the target latency value;

[0204] For any inter-satellite link route, offload any subtask to any computing node location along the current inter-satellite link route;

[0205] If the time delay constraint and resource constraint are satisfied when the subtask is unloaded to the computing node location, then the computing node location that satisfies the conditions and the current inter-satellite link route are recorded as a state set.

[0206] Based on the set of states, the target state set for multiple computational tasks is obtained.

[0207] In one possible implementation, the computing module 703 is further configured to:

[0208] The second subproblem is standardized to obtain the standardized result.

[0209] Based on pre-set constraint factors, the standardized processing results are processed to obtain the computing power value used to calculate each subtask.

[0210] The computing power and communication resource allocation device provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0211] Figure 8 This is a schematic diagram of the structure of a device for jointly allocating computing and communication resources as provided in an embodiment of this application. Figure 8As shown, the computing power and communication resource joint allocation device 800 provided in this embodiment includes at least one processor 801 and a memory 802. Optionally, the device 800 further includes a communication component 803. The processor 801, memory 802, and communication component 803 are connected via a bus 804.

[0212] In a specific implementation, at least one processor 801 executes computer execution instructions stored in memory 802, causing at least one processor 801 to perform the above-described method.

[0213] The specific implementation process of processor 801 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0214] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0215] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0216] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0217] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0218] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0219] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0220] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0221] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0222] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0223] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0224] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0225] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0226] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with the relevant laws, regulations, and standards of the relevant countries and regions, have taken necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation access points for users to choose to authorize or refuse.

[0227] It should be noted that the terms "first," "second," etc., in the claims, specification, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, products, or apparatus.

[0228] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for jointly allocating computing power and communication resources, characterized in that, The method is applied to control equipment in a ground control center corresponding to a low-Earth orbit satellite network system. The low-Earth orbit satellite network system includes multiple low-Earth orbit satellites, multiple ground users, and a ground control center. Each low-Earth orbit satellite includes multiple interactive nodes and multiple computing nodes. The interactive nodes and computing nodes are connected via inter-satellite links, and the ground control center is connected to the interactive nodes via a satellite-to-ground link. The method includes: The system acquires multiple computing requests from the users and determines multiple computing tasks based on the computing requests; wherein each computing task includes multiple subtasks, and each subtask is independent of the others. Obtain the satellite-to-ground transmission rate and inter-satellite transmission rate for each subtask; based on the satellite-to-ground transmission rate and the inter-satellite transmission rate, obtain the computational delay corresponding to each subtask; and based on the computational delay, obtain the target delay value of the computational task. Based on the target latency value, an objective function for the joint allocation of computing power and communication resources is established, and the objective function is calculated to obtain a computing power and communication resource allocation strategy.

2. The method according to claim 1, characterized in that, Obtain the satellite-to-ground transmission rate and inter-satellite transmission rate of the sub-task; based on the satellite-to-ground transmission rate and the inter-satellite transmission rate, obtain the computational delay corresponding to each sub-task; and based on the computational delay, obtain the delay value of the computational task, including: The transmit antenna gain of the ground base station in the ground control center and the receive antenna gain of the interaction node are obtained. Based on the transmit antenna gain of the ground base station in the ground control center and the receive antenna gain of the interaction node, the first signal-to-noise ratio of the satellite-to-ground link is obtained. Based on the first signal-to-noise ratio, the satellite-to-ground transmission rate of the subtask is obtained. The transmit antenna gain and receive gain of the low-orbit satellite are obtained. Based on the transmit antenna gain and receive gain, the second signal-to-noise ratio of the inter-satellite link is obtained. Based on the second signal-to-noise ratio, the inter-satellite transmission rate of the sub-task is obtained. Based on the satellite-to-ground transmission rate and the inter-satellite transmission rate, the computational delay corresponding to each sub-task is obtained, and the planned delay of each sub-task is compared to obtain the target delay value of the computational task.

3. The method according to claim 1, characterized in that, Based on the target latency value, an objective function for the joint allocation of computing power and communication resources is established, and the objective function is calculated to obtain a computing power and communication resource allocation strategy, including: The computing power of each computing node and the transmission bandwidth of the inter-satellite link are obtained, and the computing power and transmission bandwidth are processed respectively to obtain the resource constraints corresponding to each subtask. Based on the target latency value and the preset latency threshold, determine the latency constraints corresponding to each subtask; Based on the target latency value, the resource constraints, and the latency constraints, establish the objective function for the joint allocation of computing power and communication resources; The objective function is decoupled into a first subproblem and a second subproblem, wherein the first subproblem is used to indicate the location where each subtask in the computing task is offloaded to the computing node and the corresponding inter-satellite link routing subproblem, and the second subproblem is used to indicate the computing power allocation subproblem of the computing node; The first subproblem and the second subproblem are processed alternately and iteratively to obtain the computing power and communication resource allocation strategy.

4. The method according to claim 3, characterized in that, The computing power and communication resource allocation strategy is obtained by alternately calculating and processing the first subproblem and the second subproblem, including: The computing power value of the computing node is initialized, and the first sub-problem is processed by a dynamic programming algorithm to obtain the location where each sub-task in the computing task is unloaded to the computing node and the corresponding inter-satellite link route. Based on the location of each subtask in the computing task unloaded to the computing node and the corresponding inter-satellite link route, the second sub-problem is processed to obtain the computing power value used to calculate each subtask. Based on the computing power value, the first subproblem and the second subproblem are solved alternately again, and it is determined whether the preset convergence state is met. If so, the computing power and communication resource configuration strategy is obtained.

5. The method according to claim 4, characterized in that, The first subproblem is processed using a dynamic programming algorithm to obtain the location where each subtask in the computation task is offloaded to the computation node and the corresponding inter-satellite link route, including: For each computational task, the inter-satellite links are processed based on pre-set path rules to determine all inter-satellite link routes that meet the target latency value; For any inter-satellite link route, offload any of the aforementioned subtasks to all computing node locations along the current inter-satellite link route; If the subtask satisfies the latency constraint and resource constraint when it is unloaded to the computing node location, then the computing node location that satisfies the conditions and the current inter-satellite link route are recorded as a state set. Based on the set of states, a set of target states for multiple computational tasks is obtained.

6. The method according to claim 4, characterized in that, The second sub-problem is processed to obtain the computing power value used to calculate each sub-task, including: The second subproblem is standardized to obtain the standardized result. Based on pre-set constraint factors, the standardized processing results are processed to obtain the computing power value used to calculate each subtask.

7. A device for jointly allocating computing power and communication resources, characterized in that, The device includes: The acquisition module is used to acquire multiple computing requests from the users and determine multiple computing tasks based on the computing requests; wherein each computing task includes multiple subtasks, and each subtask is independent of the others; The processing module is used to obtain the satellite-to-ground transmission rate and inter-satellite transmission rate of each sub-task, obtain the calculation delay corresponding to each sub-task based on the satellite-to-ground transmission rate and the inter-satellite transmission rate, and obtain the target delay value of the calculation task based on the calculation delay. The calculation module is used to establish an objective function for the joint allocation of computing power and communication resources based on the target latency value, and to calculate the objective function to obtain a computing power and communication resource allocation strategy.

8. A device for jointly allocating computing power and communication resources, characterized in that, The device includes: a memory and a processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the computing power and communication resource joint allocation method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the computing power and communication resource joint allocation method as described in any one of claims 1-6.

10. A computer program product comprising a computer program that, when executed by a processor, implements the method for jointly allocating computing power and communication resources as described in any one of claims 1-6.

Images