Low earth orbit satellite edge computing service optimization method and system based on double time scales

Abstract

The invention discloses a low earth orbit satellite edge computing service optimization method and system based on double time scales, and relates to the technical field of satellite networks and edge computing, and the method comprises the steps: obtaining a plurality of time frames and time slots, the plurality of time frames and time slots are divided based on time, and the time scales of the time frames are larger than the time scales of the time slots; dynamically adjusting satellite selection and request scheduling based on the visibility and resource state of the current satellite in each time slot; according to the method, a service deployment and migration strategy is dynamically adjusted based on long-term service requirements and resource states in each time frame, and the dynamic adjustment method is an optimization algorithm based on double time scales, so that the service quality can be improved, and the system cost can be reduced.

Description

Technical Field

[0001] The present invention relates to the technical fields of satellite networks and edge computing, and specifically to a method and system for optimizing low-earth orbit satellite edge computing services based on a dual-time scale. Background Art

[0002] With the rapid development of low-earth orbit (LEO) satellite networks, satellite edge computing has gradually become a key technology for supporting real-time data processing and low-latency services. Due to its low-orbit characteristics, LEO satellites can provide lower signal latency and are suitable for real-time data processing and edge computing services. However, due to the dynamic nature and resource constraints of satellite networks, how to optimize satellite selection, request scheduling, and service deployment in a dynamically changing network environment remains a challenging problem. Most existing technologies focus on single-time scale optimization, that is, either real-time adjustment of satellite selection and service deployment or long-term service migration and resource allocation planning. However, this single-time scale optimization method cannot effectively cope with the dynamic changes in satellite networks, easily leading to frequent service migrations and resource waste, thereby affecting the overall performance and cost of the system.

[0003] There are still deficiencies in the existing technologies for optimizing low-earth orbit satellite edge computing services based on a dual-time scale, and there is an urgent need for a method that can take into account both long-term and short-term optimization goals. Summary of the Invention

[0004] To solve the deficiencies mentioned in the above background art, the purpose of the present invention is to provide a method and system for optimizing low-earth orbit satellite edge computing services based on a dual-time scale. Through the optimization strategy of the dual-time scale, satellite selection, request scheduling, and service deployment are dynamically adjusted to improve service quality and reduce system costs.

[0005] In a first aspect, the purpose of the present invention can be achieved through the following technical solutions: A method for optimizing low-earth orbit satellite edge computing services based on a dual-time scale, the method comprising the following steps:

[0006] Obtain a plurality of time frames and time slots, wherein the plurality of time frames and time slots are divided based on time, and the time scale of the time frame is greater than the time scale of the time slot;

[0007] Dynamically adjust satellite selection and request scheduling based on the visibility and resource status of the current satellite within each time slot; dynamically adjust service deployment and migration strategies based on the long-term service requirements and resource status within each time frame, wherein the method of dynamic adjustment is an optimization algorithm based on a dual-time scale.

[0008] In combination with the first aspect, in certain implementations of the first aspect, the method further includes: the time frame t is used to process long-term decisions, including service migration and replica deployment. Each time frame contains multiple time slots, and the length of the time frame is adjusted according to the dynamics of the system and resource changes. The division of the time frame enables the system to perform resource planning within a long time range.

[0009] In combination with the first aspect, in certain implementations of the first aspect, the method further includes: the time slot k is used to process short-term decisions, including satellite selection and request scheduling. And it quickly responds to changes in user requests and satellite resources. The division of the time slot enables the system to perform real-time adjustment within a short time range.

[0010] In combination with the first aspect, in certain implementations of the first aspect, the method further includes: the selection process of the optimal satellite for dynamically adjusting satellite selection and request scheduling based on the visibility and resource status of the current satellite within each time slot:

[0011] When selecting, factors such as satellite visibility, resource status, and transmission delay need to be considered. The optimization goal of satellite selection is to minimize transmission delay and handover overhead. The communication coverage of the satellite is defined through STK simulation. If the ground station is within the coverage of the satellite, then this satellite is selected as the access satellite and the optimal satellite. Among them, the coverage of the satellite is jointly determined by the orbital altitude of the satellite, the antenna angle, and the geographical location of the ground station;

[0012] The process of request scheduling is transmitted to the service satellite through the access satellite for processing. The optimization goal of request scheduling is to minimize the processing delay and resource consumption of the request.

[0013] In combination with the first aspect, in certain implementations of the first aspect, the method further includes: the service deployment process of dynamically adjusting service deployment and migration strategies based on the long-term service demand and resource status within each time frame: The deployment strategy of service replicas needs to consider the computing, storage, and communication resources of the satellite to optimize the deployment location of service replicas;

[0014] The migration strategy is carried out when the satellite has insufficient resources or user requirements change. The optimization goal is to minimize migration overhead and service interruption time.

[0015] In combination with the first aspect, in certain implementations of the first aspect, the method further includes: the optimization process of the optimization algorithm based on a dual time scale is processed through a pre-constructed polynomial model. Among them, the pre-constructed polynomial model is a model for minimizing social cost, and the expression of the model for minimizing social cost is as follows:

[0016]

[0017] Among them indicates whether the ground station is connected to the satellite at time tk indicates whether the satellite is a serving satellite indicates the user request transmitted by the ground station indicates the transmission delay between the ground station and the satellite indicates the satellite service cost and respectively represent the handover cost and service migration cost for accessing the satellite indicates the delay in the satellite network, D i and C m respectively represent the receiving capacity and service capacity of the satellite; Constraint (1a) ensures that each ground station selects only one satellite as the access satellite in each time slot, Constraint (1b) ensures that the access capacity of each satellite is respected in each time slot, Constraint (1c) ensures that the user requests received by each access satellite are scheduled in each time slot, Constraint (1d) ensures that the service capacity of each satellite is respected in each time slot, Constraint (1e) ensures that the total number of service replicas existing in the satellite network is always consistent with the specified one in each time frame, and Constraint (1f) enforces the domain of all decision variables.

[0018] Combined with the first aspect, in some implementations of the first aspect, the method further includes: The adopted algorithm for dynamically adjusting satellite selection and request scheduling is as follows:

[0019] Gradually increase the dual variables through the primal-dual method until each dual constraint is satisfied, select the satellite selection scheme that minimizes the transmission delay and satisfies the capacity constraint, and output the satellite selection and request scheduling decisions within each time slot;

[0020] And first check whether the current solution still satisfies the constraint conditions. If not, roll back to the previous solution and re-solve the sub-problem to update the solution, and delay switching to access the position of the satellite until the current cache position causes the request scheduling to be infeasible or the accumulated non-replacement cost exceeds a predetermined threshold.

[0021] Combined with the first aspect, in some implementations of the first aspect, the method further includes: The adopted algorithm for dynamically adjusting service deployment and migration strategies is as follows:

[0022] First initialize the dual variables, and then increase the dual variables until each dual constraint is satisfied. During this process, select the service replica placement scheme that minimizes the cost and satisfies the constraint conditions, and finally output the service replica placement decisions within each time frame;

[0023] And delay changing the serving satellite group.

[0024] Second aspect, to achieve the above object, the present invention discloses an optimization system for low-earth orbit satellite edge computing services based on a dual time scale, including:

[0025] A time division module, configured to obtain a plurality of time frames and time slots, wherein the plurality of time frames and time slots are divided based on time, and the time scale of the time frame is greater than the time scale of the time slot;

[0026] A dynamic adjustment module, configured to dynamically adjust satellite selection and request scheduling based on the visibility and resource status of the current satellite within each time slot; and dynamically adjust service deployment and migration strategies based on the long-term service demand and resource status within each time frame, wherein the method of dynamic adjustment is an optimization algorithm based on a dual time scale.

[0027] In yet another aspect of the present invention, to achieve the above object, a terminal device is disclosed, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. The memory stores a computer program capable of running on the processor. When the processor loads and executes the computer program, the optimization method for low-earth orbit satellite edge computing services based on the dual time scale as described above is adopted.

[0028] Advantages of the present invention:

[0029] Through the optimization strategy of the dual time scale, the present invention dynamically adjusts satellite selection, request scheduling, and service deployment, and can effectively balance long-term and short-term optimization goals, improve system performance, and reduce system costs. Description of the drawings

[0030] To more clearly illustrate the technical solutions in the embodiments of the present invention or in the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings;

[0031] Figure 1 It is a schematic flowchart of the method of the present invention;

[0032] Figure 2 It is a flowchart of the core algorithm of the present invention;

[0033] Figure 3 It is a system scenario diagram of the present invention;

[0034] Figure 4 It is a dual time scale concept diagram of the present invention;

[0035] Figure 5 It is a schematic diagram of the system structure of the present invention.

[0036] Figure 6 It is a diagram showing the experimental results of the comparison of the normalized total cost of the present invention under different constellation scales and different algorithms;

[0037] Figure 7 It is a diagram showing the experimental results of the comparison of the normalized costs of different algorithms in each time slot under the same constellation scale and the entire time range of the present invention. Detailed implementation manners

[0038] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts belong to the scope of protection of the present invention.

[0039] Embodiment 1:

[0040] As Figure 1 shown, a method for optimizing low-earth orbit satellite edge computing services based on a dual-time scale, the method includes the following steps:

[0041] S101: Obtain a plurality of time frames and time slots, wherein the plurality of time frames and time slots are divided based on time, and the time scale of the time frame is greater than the time scale of the time slot;

[0042] Specifically, time is divided into a plurality of time frames and time slots. The time frame belongs to a larger time scale and is used for long-term decisions (such as service migration and replica deployment). The time slot belongs to a smaller time scale and is used for short-term decisions (such as satellite selection and request scheduling). Through the division of the dual-time scale, the optimization goals of both the long term and the short term can be taken into account simultaneously.

[0043] Time Frame: The time frame t is used to process long-term decisions, such as service migration and replica deployment. Each time frame contains a plurality of time slots, and the length of the time frame can be adjusted according to the dynamics of the system and the changes in resources. The division of the time frame enables the system to perform resource planning within a relatively long time range, avoiding frequent service migrations and resource waste.

[0044] Time Slot: The time slot k is used to process short-term decisions, such as satellite selection and request scheduling. The length of each time slot is short, enabling rapid response to user requests and changes in satellite resources. The division of the time slot enables the system to perform real-time adjustments within a relatively short time range, ensuring the continuity and efficiency of services.

[0045] Through the division of the dual-time scale, the system can make dynamic adjustments between the long term and the short term, avoiding frequent service migrations and resource waste.

[0046] S102: Dynamically adjust satellite selection and request scheduling based on the visibility and resource status of the current satellite within each time slot; dynamically adjust service deployment and migration strategies based on long-term service requirements and resource status within each time frame, where the method of dynamic adjustment is an optimization algorithm based on a dual time scale.

[0047] Within each time slot, based on the visibility and resource status of the current satellite, select the optimal satellite as the access satellite and schedule user requests to the service satellite for processing. By dynamically adjusting satellite selection and request scheduling, the dynamic changes of the satellite network can be effectively addressed.

[0048] Satellite Selection: Within each time slot, the ground station needs to select one satellite from multiple visible satellites as the access satellite. When making the selection, factors such as satellite visibility, resource status, and transmission delay need to be considered. Specifically, the optimization goal of satellite selection is to minimize transmission delay and handover overhead. We define the communication coverage range of the satellite through the simulation of STK (Satellite ToolKit). If the ground station is within the coverage range of the satellite, then this satellite can be selected as the access satellite. The coverage range of the satellite is jointly determined by the satellite's orbital altitude, antenna angle, and the geographical location of the ground station. Through STK simulation, we can accurately calculate the visibility between the ground station and the satellite within each time slot, thus providing a basis for satellite selection.

[0049] Request Scheduling: User requests are first transmitted to the service satellite through the access satellite for processing. The optimization goal of request scheduling is to minimize the processing delay and resource consumption of the requests. By dynamically adjusting satellite selection and request scheduling, the system can respond to changes in user requirements within a short period of time, ensuring service continuity and efficiency.

[0050] By dynamically adjusting satellite selection and request scheduling, the system can respond to changes in user requirements within a short period of time, ensuring service continuity and efficiency.

[0051] Service Deployment and Migration

[0052] Within each time frame, optimize the deployment and migration strategies of service replicas based on long-term service requirements and resource status. By dynamically adjusting service deployment and migration, resource waste and service interruption can be effectively avoided.

[0053] Service Deployment: The deployment strategy of service replicas needs to consider the computing, storage, and communication resources of the satellite. By optimizing the deployment location of service replicas, resource utilization can be maximized and service latency can be reduced.

[0054] Service Migration: When satellite resources are insufficient or user requirements change, the system needs to dynamically migrate service replicas. The optimization goal of service migration is to minimize migration overhead and service interruption time.

[0055] By dynamically adjusting service deployment and migration, the system can maintain efficient resource utilization in the long term, avoiding resource waste and service interruption.

[0056] Design an optimization algorithm based on a dual time scale. Through online optimization, dynamically adjust satellite selection, request scheduling, and service deployment. Through the dual time scale optimization algorithm, it is possible to effectively balance long-term and short-term optimization goals, improve system performance, and reduce system costs.

[0057] Figure 2 Shows the overall framework of the algorithm design. We designed multiple algorithms to handle decision-making problems at different time scales. Specifically, Algorithm 1 and Algorithm 2 are used to handle the service replica placement problem at a large time scale and the satellite selection and request scheduling problem at a small time scale, respectively.

[0058] Algorithm 1 (Large Time Scale): This algorithm solves the service replica placement problem through the primal-dual method. The algorithm first initializes the dual variables and then gradually increases the dual variables until each dual constraint is satisfied. During this process, the algorithm selects a service replica placement scheme that minimizes the cost and satisfies the constraint conditions. Finally, the algorithm outputs the service replica placement decision for each time frame.

[0059] Algorithm 2 (Small Time Scale): This algorithm is used to handle the satellite selection and request scheduling problem. The algorithm gradually increases the dual variables through the primal-dual method until each dual constraint is satisfied. During this process, the algorithm selects a satellite selection scheme that minimizes the transmission delay and satisfies the capacity constraint. Finally, the algorithm outputs the satellite selection and request scheduling decision for each time slot.

[0060] In terms of subsequent service replica placement and migration, we designed Algorithm 3 and Algorithm 4 to balance decisions at different time scales. Algorithm 4 dynamically adjusts satellite selection and request scheduling at a small time scale, while Algorithm 3 adjusts the placement and migration of service replicas at a large time scale. They respectively need to use the feasible solutions output by the primal-dual algorithms of Algorithm 2 and Algorithm 1 for calculation.

[0061] Algorithm 4 (Small Time Scale): This algorithm dynamically adjusts satellite selection and request scheduling within each time slot. The algorithm first checks whether the current solution still satisfies the constraint conditions. If not, it reverts to the previous feasible solution and re-solves the sub-problem to update the solution. The algorithm tries to minimize the handover overhead by delaying the handover to access the satellite's position until the current cache position causes the request scheduling to be infeasible or the accumulated non-replacement cost exceeds a predetermined threshold.

[0062] Algorithm 3 (Large time scale): This algorithm adjusts the placement and migration of service replicas within each time frame. By delaying the change of the service satellite group, the algorithm avoids unnecessary service replica migration costs. The algorithm uses the output of Algorithm 4 as input to ensure the consistency of short-term decisions with long-term goals.

[0063] The optimization process of the proposed dual-time-scale optimization algorithm is processed through a pre-constructed polynomial model. The pre-constructed polynomial model is a model for minimizing social cost, and the expression of the model for minimizing social cost is as follows: First, the ground station collects requests from ground users. In this project, it is defined that the ground station is within the coverage area of the satellite and the satellite is within the visible range of the ground station. Then, at the current moment, this satellite is the satellite flying over the ground station. The ground station selects one of these flying satellites as the access satellite to transmit the service request through the satellite-ground link.

[0064] Second, the user request first arrives at the access satellite i of ground station j and then is sent to the satellite with service replicas m , that is, the service satellite for processing. Since the service capacity of the service satellite may be limited, an access satellite will distribute the service to multiple service satellites for processing. Here, the neighbor satellites in the same orbit and adjacent orbits of the access satellite are preferentially selected. In this project, the time range is divided into time frame t and time slot k, where the time frame contains several time slots. By proposing this multi-time-scale dynamic control solution, this project aims to provide an efficient and flexible management method for the task coordination service of multi-satellite cooperation. This project introduces a polynomial online multi-satellite cooperation scheduling algorithm based on primal-dual, comprehensively considering the transmission delay of the satellite-ground and satellite-satellite links, the task cost of the access satellite and the service satellite, the migration cost of the task and the replica in the satellite network, and the handover cost of the ground station accessing the satellite, to form a model for minimizing social cost.

[0065]

[0066] where represents whether the ground station is connected to the satellite at time tk, represents whether the satellite is a service satellite, represents the user request transmitted by the ground station, represents the transmission delay between the ground station and the satellite, represents the satellite service cost, and respectively represent the handover cost of the access satellite and the service migration cost, represents the delay in the satellite network, D i and C mrespectively represent the receiving capacity and service capacity of the satellite; Constraint (1a) ensures that each ground station selects only one satellite as the access satellite in each time slot, Constraint (1b) ensures that the access capacity of each satellite is respected in each time slot, Constraint (1c) ensures that the user requests received by each access satellite are scheduled in each time slot, Constraint (1d) ensures that the service capacity of each satellite is respected in each time slot, Constraint (1e) ensures that the total number of service replicas existing in the satellite network is always consistent with the specified one in each time frame, and Constraint (1f) enforces the domain of all decision variables.

[0067] The optimization problem of the system is described by constructing a polynomial model, and a polynomial online algorithm based on the primal-dual algorithm is designed to solve the above model. To avoid the handover cost caused by the frequent handover of the access satellite across time slots, we introduce a non-greedy algorithm for optimization to achieve the co-evolution of the tasks of multi-satellite cooperation in satellite selection, request scheduling, and service provision under two time scales.

[0068] Polynomial model: We construct a model for minimizing the social cost, with the goal of minimizing transmission delay, handover cost, service migration cost, and resource consumption. By means of constraint conditions, it is ensured that each ground station selects only one satellite as the access satellite in each time slot and respects the access capacity and service capacity of each satellite.

[0069] Primal-dual algorithm: We design a polynomial online algorithm based on the primal-dual algorithm, by gradually increasing the dual variables until each dual constraint is satisfied. During this process, the algorithm selects the service replica placement scheme that minimizes the cost and satisfies the constraint conditions. Specifically, Algorithms 1 and 2 constructed on two time scales respectively

[0070] Non-greedy algorithm: To avoid the handover cost caused by the frequent handover of the access satellite across time slots, we introduce a non-greedy algorithm for optimization. This algorithm delays the position of the access satellite for handover and tries to minimize the handover overhead until the current cache position leads to the infeasibility of request scheduling or the accumulated non-replacement cost exceeds a predetermined threshold. Specifically, Algorithms 3 and 4 constructed on two time scales respectively.

[0071] Specifically, the solution of the present invention is further elaborated below through embodiments:

[0072] Figure 3It shows a system scenario where a ground station is covered by multiple satellites. The ground station selects the most suitable satellite for communication to ensure the reliability and efficiency of communication. Due to the orbital dynamics of satellites, the ground station needs to select different satellites for connection within different time windows. This scenario of multi-satellite coverage is very common in LEO satellite communication systems, especially when the satellite density increases and the communication network requirements become complex. Since the resources of a single satellite are limited, the ground station may need to distribute user requests to multiple service satellites for processing. The selection of service satellites usually gives priority to neighbor satellites in the same orbit or adjacent orbits as the access satellite to reduce the transmission delay of inter-satellite links. Due to the high-speed movement of satellites, the connection time between the ground station and satellites is limited, and satellite resources (such as computing, storage, and bandwidth) are dynamically changing. Therefore, the system needs to adjust satellite selection and request scheduling in real time to cope with these dynamic changes.

[0073] Figure 4 It shows a two-stage decision-making framework that divides time into a large time scale (time frame) and a small time scale (time slot). This division of the two time scales enables the system to take into account both long-term and short-term optimization goals and effectively cope with the dynamic changes of the satellite network. On the large time scale (time frame), the system processes long-term decisions, such as the deployment and migration of service replicas. Each time frame contains multiple time slots, and the length of the time frame can be adjusted according to the dynamicity and resource changes of the system. The decisions of the time frame mainly focus on the long-term deployment strategy of service replicas to maximize resource utilization and reduce service latency. Within each time frame, the system optimizes the deployment location of service replicas according to the long-term service demand and resource status. The deployment strategy of service replicas needs to consider the computing, storage, and communication resources of satellites, and the goal is to minimize service latency and resource consumption. When satellite resources are insufficient or user demands change, the system needs to dynamically migrate service replicas, and the optimization goal of service migration is to minimize migration overhead and service interruption time. On the small time scale (time slot), the system processes short-term decisions, such as satellite selection and request scheduling. The length of each time slot is short, which can quickly respond to changes in user requests and satellite resources. The decisions of the time slot mainly focus on real-time satellite selection and request scheduling to ensure service continuity and efficiency. Within each time slot, the ground station needs to select one of the multiple visible satellites as the access satellite. When making the selection, factors such as satellite visibility, resource status, and transmission delay need to be considered. The optimization goal of satellite selection is to minimize transmission delay and handover overhead. User requests are first transmitted to service satellites through the access satellite for processing, and the optimization goal of request scheduling is to minimize the processing delay and resource consumption of requests.

[0074] Embodiment 2: Second aspect, as Figure 5 shown, to achieve the above object, the present invention discloses a low-Earth orbit satellite edge computing service optimization system based on a dual time scale, including:

[0075] A time division module 11, configured to obtain a plurality of time frames and time slots, wherein the plurality of time frames and time slots are divided based on time, and the time scale of the time frame is greater than the time scale of the time slot;

[0076] A dynamic adjustment module 12, configured to dynamically adjust satellite selection and request scheduling based on the visibility and resource status of the current satellite within each time slot; and dynamically adjust service deployment and migration strategies based on long-term service requirements and resource status within each time frame, wherein the dynamic adjustment method is an optimization algorithm based on a dual time scale.

[0077] To verify the technical effects of the present invention, multi-dimensional experimental verifications are carried out based on a simulated Walker-Delta low-Earth orbit satellite constellation system (including 16 orbital planes, 24 satellites per plane, an orbital altitude of 1000 kilometers, and an inclination angle of 53°) and 20 ground stations within China. The ground stations are dynamically connected to the satellites through a communication coverage range with a 30° cone angle. The experimental time span is 10 consecutive time frames (each frame contains 12 time slots), simulating dynamic user requests and resource change scenarios. By comparing with traditional greedy algorithms, random algorithms, and single time scale methods, the dual time scale optimization strategy of the present invention shows significant advantages in terms of total cost, delay, and handover overhead. As Figure 6 (Total cost comparison chart) shows, under a typical constellation scale, the total cost of the present invention is significantly reduced compared with the single time scale method, with an optimization amplitude of 20%, and is reduced by 33% and 45% respectively compared with the greedy algorithm and the random algorithm. The curves in the figure clearly show that due to the lack of a coordination mechanism for long-term resource planning and short-term dynamic adjustment in traditional methods, the total cost shows a fluctuating upward trend as the experimental time progresses. However, through the joint optimization of the dual time scale framework of the present invention, the total cost curve converges smoothly, verifying the coordination effect of long-term service deployment strategies (such as dynamic migration of service replicas) and short-term request scheduling (such as satellite selection and task allocation). Further combined with Figure 7 (Single time slot cost change chart) analysis, in the critical time slots of satellite handover (such as time slot 6 and time slot 12), the traditional method has a sharp increase in cost peaks due to frequent handovers. However, the delay handover strategy of the present invention smooths the transition through a dynamic caching mechanism, suppressing the peak fluctuations, and the cost is reduced by 5% - 50% compared with other methods. It can be seen from the figure that the cost curves of traditional algorithms show significant spikes in handover time slots. The present invention reduces the additional overhead caused by connection interruptions by predicting satellite coverage changes and caching tasks in advance, and at the same time reduces transmission delay and resource competition conflicts by optimizing the task allocation of service satellites.

[0078] In addition, by comparing the long-term cost convergence ( Figure 6 ) and short-term stability ( Figure 7) The dual-time-scale framework of the present invention demonstrates outstanding dynamic adaptability. Traditional single-time-scale methods only focus on real-time response and neglect long-term resource planning, resulting in frequent service migrations and cumulative costs. In contrast, greedy and random algorithms lack a global optimization perspective and cannot balance switching overhead and resource utilization. In comparison, the present invention realizes a double improvement in resource utilization and service quality through a hierarchical long-term and short-term decision-making mechanism, preallocating service replicas and optimizing migration paths on the time-frame scale, and dynamically adjusting access satellites and task scheduling on the time-slot scale. Experimental data further shows that the transmission delay of the present invention is reduced by an average of 30%, and the switching frequency is reduced by 40%, verifying its robustness and efficiency in a dynamic satellite network environment.

[0079] The above experimental results fully prove that the dual-time-scale optimization method of the present invention effectively solves key problems such as frequent switching, resource competition, and cost accumulation in the low-earth orbit satellite edge computing scenario by coordinating long-term resource planning and short-term dynamic response, providing a practical technical solution for service optimization in a complex space network environment.

[0080] Based on the same inventive concept, the present invention also provides a computer device, which includes: one or more processors, and a memory for storing one or more computer programs; the program includes program instructions, and the processor is used to execute the program instructions stored in the memory. The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is used to implement one or more instructions, specifically used to load and execute one or more instructions in the computer storage medium to implement the above method.

[0081] It should be further noted that, based on the same inventive concept, the present invention also provides a computer storage medium, on which a computer program is stored. When the computer program is run by a processor, it executes the above-mentioned method. The storage medium may be any combination of one or more computer-readable media. The computer-readable media may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electro-magnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (a non-exhaustive list) of the computer-readable storage medium include: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, the computer-readable storage medium may be any tangible medium that contains or stores a program, and the program may be used by or in combination with an instruction execution system, apparatus, or device.

[0082] In the description of this specification, the description with reference to terms such as "an embodiment", "example", "specific example", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present disclosure. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0083] The above shows and describes the basic principles, main features, and advantages of the present disclosure. Those skilled in the art should understand that the present disclosure is not limited by the above embodiments. The above embodiments and the descriptions in the specification only illustrate the principles of the present disclosure. Without departing from the spirit and scope of the present disclosure, the present disclosure will have various changes and improvements, and these changes and improvements all fall within the scope of the present disclosure claimed.

Claims

1. A low-orbit satellite edge computing service optimization method based on dual time scales, characterized in that: The method comprises the following steps: Acquire a plurality of time frames and time slots, wherein the plurality of time frames and time slots are divided based on time, and a time scale of the time frame is greater than a time scale of the time slot; Satellite selection and request scheduling are dynamically adjusted based on the visibility and resource status of the current satellite in each time slot; service deployment and migration strategies are dynamically adjusted based on the long-term service demand and resource status in each time frame, wherein the dynamic adjustment method is an optimization algorithm based on a dual time scale.

2. The dual-time-scale low-orbit satellite edge computing service optimization method according to claim 1 is characterized in that: The time frame t is used to process long-term decisions, including service migration and replica deployment. Each time frame contains multiple time slots. The length of the time frame is adjusted according to the dynamics of the system and resource changes. The division of time frames enables the system to plan resources over a long period of time.

3. The dual-time-scale low-orbit satellite edge computing service optimization method according to claim 1 is characterized in that: The time slot k is used to process short-term decisions, including satellite selection and request scheduling, and quickly respond to user requests and changes in satellite resources. The division of time slots enables the system to make real-time adjustments within a short time frame.

4. The dual-time-scale low-orbit satellite edge computing service optimization method according to claim 1, characterized in that: The process of selecting the optimal satellite for dynamically adjusting satellite selection and requesting scheduling based on the visibility and resource status of the current satellite in each time slot: When selecting, factors such as satellite visibility, resource status, and transmission delay need to be considered. The optimization goal of satellite selection is to minimize transmission delay and switching overhead. The communication coverage of the satellite is defined through STK simulation. If the ground station is within the coverage of the satellite, this satellite is selected as the access satellite and the optimal satellite. The coverage of the satellite is determined by the satellite's orbital altitude, antenna angle, and the geographical location of the ground station. The request scheduling process is transmitted from the access satellite to the service satellite for processing. The optimization goal of request scheduling is to minimize the processing delay and resource consumption of the request.

5. The dual-time-scale low-orbit satellite edge computing service optimization method according to claim 1, characterized in that: The service deployment process of dynamically adjusting service deployment and migration strategies based on long-term service demand and resource status in each time frame: the deployment strategy of service replicas needs to consider the computing, storage and communication resources of the satellite to optimize the deployment location of service replicas; The migration strategy is implemented when satellite resources are insufficient or user needs change, and the optimization goal is to minimize the migration overhead and service interruption time.

6. The dual-time-scale low-orbit satellite edge computing service optimization method according to claim 1, characterized in that: The optimization process of the dual-time-scale optimization algorithm is processed by a pre-constructed polynomial model, wherein the pre-constructed polynomial model is a social cost minimization model, and the expression of the social cost minimization model is as follows: in Indicates whether the ground station is connected to the satellite at time tk. Indicates whether the satellite is a service satellite. represents the user request transmitted by the ground station, represents the transmission delay between the ground station and the satellite, represents the satellite service cost, and They represent the switching cost of accessing the satellite and the service migration cost respectively, represents the delay in the satellite network, D i With C m They represent the receiving capability and service capability of the satellite respectively; constraint (1a) ensures that each ground station selects only one satellite as the access satellite in each time slot, constraint (1b) ensures that the access capability of each satellite is respected in each time slot, constraint (1c) ensures that the user requests received by each access satellite are scheduled in each time slot, constraint (1d) ensures that the service capability of each satellite is respected in each time slot, constraint (1e) ensures that the total number of service replicas existing in the satellite network is always consistent with the specified one in each time frame, and constraint (1f) is the domain that enforces all decision variables.

7. The dual-time-scale low-orbit satellite edge computing service optimization method according to claim 4, characterized in that: The algorithm used to dynamically adjust satellite selection and request scheduling is as follows: The dual variables are gradually increased through the primal-dual method until each dual constraint is satisfied, and the satellite selection scheme that minimizes the transmission delay and satisfies the capacity constraint is selected, and the satellite selection and request scheduling decision in each time slot is output; and first checking whether the current solution still satisfies the constraints. If not, it falls back to the previous solution and re-solves the subproblem to update the solution by delaying the switching of the visited satellite positions until the current cache position causes the request scheduling to be infeasible or the accumulated non-replacement cost exceeds a predetermined threshold.

8. The dual-time-scale low-orbit satellite edge computing service optimization method according to claim 5, characterized in that: The algorithm used to dynamically adjust the service deployment and migration strategy is as follows: First, initialize the dual variable, then increase the dual variable until each dual constraint is satisfied. In this process, select the service replica placement scheme that minimizes the cost and satisfies the constraints, and finally output the service replica placement decision within each time frame; and changing the serving satellite set by delay.

9. A low-orbit satellite edge computing service optimization system based on dual time scales, characterized in that: include: A time division module, used to obtain a plurality of time frames and time slots, wherein the plurality of time frames and time slots are divided based on time, and the time scale of the time frame is greater than the time scale of the time slot; A dynamic adjustment module is used to dynamically adjust satellite selection and request scheduling based on the visibility and resource status of the current satellite in each time slot; and dynamically adjust service deployment and migration strategies based on the long-term service demand and resource status in each time frame, wherein the dynamic adjustment method is an optimization algorithm based on a dual time scale.

10. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that: The memory stores a computer program that can be run on the processor. When the processor loads and executes the computer program, the low-orbit satellite edge computing service optimization method based on dual time scales described in any one of claims 1 to 8 is adopted.

Images