CN116366133B - Unloading method and device based on low-orbit satellite edge calculation - Google Patents
Unloading method and device based on low-orbit satellite edge calculation Download PDFInfo
- Publication number
- CN116366133B CN116366133B CN202310359205.8A CN202310359205A CN116366133B CN 116366133 B CN116366133 B CN 116366133B CN 202310359205 A CN202310359205 A CN 202310359205A CN 116366133 B CN116366133 B CN 116366133B
- Authority
- CN
- China
- Prior art keywords
- satellite
- time
- time delay
- inter
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000004364 calculation method Methods 0.000 title claims abstract description 40
- 238000000034 method Methods 0.000 title claims abstract description 26
- 230000005540 biological transmission Effects 0.000 claims abstract description 29
- 238000012545 processing Methods 0.000 claims abstract description 14
- 238000012549 training Methods 0.000 claims abstract description 14
- 230000001934 delay Effects 0.000 claims abstract description 12
- 238000012360 testing method Methods 0.000 claims abstract description 7
- 239000011159 matrix material Substances 0.000 claims description 15
- 238000005457 optimization Methods 0.000 claims description 12
- 230000002123 temporal effect Effects 0.000 claims description 5
- 238000010586 diagram Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 6
- 238000004590 computer program Methods 0.000 description 5
- 230000003044 adaptive effect Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000000547 structure data Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 235000019800 disodium phosphate Nutrition 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/14—Relay systems
- H04B7/15—Active relay systems
- H04B7/185—Space-based or airborne stations; Stations for satellite systems
- H04B7/1851—Systems using a satellite or space-based relay
- H04B7/18513—Transmission in a satellite or space-based system
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/14—Relay systems
- H04B7/15—Active relay systems
- H04B7/185—Space-based or airborne stations; Stations for satellite systems
- H04B7/1851—Systems using a satellite or space-based relay
- H04B7/18519—Operations control, administration or maintenance
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W28/00—Network traffic management; Network resource management
- H04W28/02—Traffic management, e.g. flow control or congestion control
- H04W28/0231—Traffic management, e.g. flow control or congestion control based on communication conditions
- H04W28/0236—Traffic management, e.g. flow control or congestion control based on communication conditions radio quality, e.g. interference, losses or delay
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Abstract
The application discloses an unloading method and device based on low-orbit satellite edge calculation, wherein the method comprises the following steps: s1, deploying a prediction model for training and verifying on a high orbit satellite, wherein the prediction model is trained and verified by using a training set and a testing set which are divided by an input data set, the input data set is obtained by processing time delay historical data, and the time delay historical data comprises a plurality of continuous time units of inter-satellite-to-satellite topological relations, inter-satellite time delays and inter-satellite time delays; s2, predicting delay future data of the next time unit by using a prediction model, wherein the delay future data comprises inter-satellite-ground topological relation, inter-satellite time delay and inter-satellite time delay; s3, calculating and acquiring an unloading task transmission link with the lowest time delay of each terminal in the next time unit by using time delay future data to serve as a real transmission link; s4, updating the real time delay data of the next time unit to the time delay historical data; steps S2 to S4 are repeated.
Description
Technical Field
The application relates to the technical field of satellite communication, in particular to an unloading method and device based on low-orbit satellite edge calculation.
Background
Under the wide application of artificial intelligence and cloud computing, along with the development of the integration of the world and the low-orbit satellite network, the introduction of the edge computing technology into the low-orbit satellite network is one of the important points of future research. The key issues presented in 3GPP latest technical report 23.700-27 regarding how to support edge computing technology by means of on-board UPF, which illustrates that on-board edge computing technology has been taken as a future research direction by the international communication standards organization. The introduction of the spaceborne edge computing technology in the low-orbit satellite network can further improve the powerful computing power of the future 6G network.
At present, the main stream of satellite-borne edge computing and unloading strategies is to directly unload computing tasks to overhead satellites, but if a large amount of computing tasks are directly unloaded to overhead satellites, problems such as unbalanced use of satellite computing resources, overlarge partial satellite load, reduced cruising ability and the like are caused. In addition, the currently mainstream satellite-borne edge calculation time delay is higher, so that congestion of an unloading task queue is easy to cause, the satellite electric quantity consumption is increased, the service quality and the calculation speed of the low-orbit satellite-borne edge calculation are reduced, and the technology (automatic driving, intelligent Internet of things and the like) based on the satellite-borne edge calculation service is affected. Moreover, the mainstream spaceborne edge computing offloading strategy has no adaptive optimization capability.
Disclosure of Invention
The present application is directed to a method, apparatus, device and computer readable storage medium for unloading low-orbit satellite based edge computation, which can solve at least one of the above problems.
In order to achieve the above object, the present application provides an unloading method based on low-orbit satellite edge calculation, comprising:
s1, deploying a prediction model for training and verifying on a high orbit satellite, wherein the prediction model is trained and verified by using a training set and a testing set which are divided by an input data set, the input data set is obtained by processing time delay historical data, and the time delay historical data comprises a continuous inter-satellite-to-satellite topological relation, a satellite-to-ground time delay and an inter-satellite time delay of a plurality of time units;
s2, predicting delay future data of the next time unit by using the prediction model, wherein the delay future data comprises inter-satellite-ground topological relations, inter-satellite time delays and inter-satellite time delays;
s3, calculating and acquiring an unloading task transmission link with the lowest time delay of each terminal in the next time unit by using the time delay future data to serve as a real transmission link;
s4, updating the real time delay data of the next time unit to the time delay historical data;
steps S2 to S4 are repeated.
Optionally, the time delay history data for each of the time units is constructed as a adjacency matrix.
Optionally, the prediction model includes an input layer, a time-space domain convolution module and an output layer, and the time-space domain convolution module is responsible for performing time-space domain convolution processing on input data to obtain time features and space features.
Optionally, the prediction model includes two time-space domain convolution modules, each time-space domain convolution module includes two time-domain convolution blocks and one space-domain convolution block, the time-domain convolution blocks are used for capturing the time features, and the space-domain convolution blocks are used for capturing the space features.
Optionally, the output layer includes a time domain gating convolution block, where the time domain gating convolution block is used to merge data in a time dimension, and a full connection layer, where the full connection layer is used to output a prediction result.
In order to achieve the above object, the present application further provides an unloading policy optimization device based on low-orbit satellite edge calculation, including:
the deployment module is used for deploying a prediction model for training and verifying on a high-orbit satellite, the prediction model is trained and verified by using a training set and a testing set which are divided by an input data set, the input data set is obtained by processing time delay historical data, and the time delay historical data comprises an inter-satellite-ground topological relation, an inter-satellite time delay and an inter-satellite time delay which are acquired by a plurality of continuous time units;
the prediction module is used for predicting time delay future data of the next time unit by utilizing the prediction model, wherein the time delay future data comprises an inter-satellite-ground topological relation, an inter-satellite time delay and an inter-satellite time delay;
the acquisition module is used for calculating and acquiring an unloading task transmission link with the lowest time delay of each terminal in the next time unit by using the time delay future data to serve as a real transmission link;
and the updating module is used for updating the real time delay data of the next time unit to the time delay historical data.
Optionally, the time delay history data for each of the time units is constructed as a adjacency matrix.
Optionally, the prediction model includes an input layer, a time-space domain convolution module and an output layer, and the time-space domain convolution module is responsible for performing time-space domain convolution processing on input data to obtain time features and space features.
To achieve the above object, the present application also provides an apparatus comprising:
a processor;
a memory having stored therein executable instructions of the processor;
wherein the processor is configured to perform an offloading method based on low-earth-satellite-edge computation as described above via execution of the executable instructions.
To achieve the above object, the present application also provides a computer-readable storage medium having stored thereon a program which, when executed by a processor, implements the low-orbit satellite edge calculation-based offloading method as described above.
The present application also provides a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the electronic device reads the computer instructions from the computer readable storage medium and executes the computer instructions to cause the electronic device to perform the low-orbit satellite edge calculation based offloading method as described above.
In the application, an inter-satellite-ground topological relation, an inter-satellite time delay and an inter-satellite time delay which are acquired by a plurality of continuous time units are used as time delay historical data to train a prediction model which is deployed on a high-orbit satellite, the inter-satellite-ground topological relation, the inter-satellite time delay and the inter-satellite time delay of the next time unit are predicted by using the prediction model to calculate an unloading task transmission link with the lowest time delay of each terminal in the next time unit to be used as a real transmission link, and real time delay data of the next time unit is updated to the time delay historical data to continuously predict time delay future data and calculate the next real transmission link. By repeating the steps, the self-adaptive optimization of the low-orbit satellite edge calculation unloading strategy can be realized, an unloading task transmission link with the lowest time delay is always used as a real transmission link, congestion of an unloading task queue is not easy to cause, the satellite electric quantity consumption is reduced, the service quality and the calculation speed of the low-orbit satellite-borne edge calculation are improved, and meanwhile, the problems of unbalanced use of satellite-borne calculation resources, overlarge partial satellite load, insufficient cruising ability and the like can be effectively relieved.
Drawings
FIG. 1 is a schematic block diagram of an unloading strategy adaptive optimization system based on space-borne edge computation according to an embodiment of the application.
FIG. 2 is a flow chart of an offloading method based on low-orbit satellite edge calculation according to an embodiment of the application.
FIG. 3 is a schematic diagram of an adjacency matrix constructed from time delay history data of one time unit according to an embodiment of the present application.
FIG. 4 is a schematic diagram of an undirected graph formed from data of one time unit according to an embodiment of the present application.
Fig. 5 is a schematic diagram of a space-time diagram constructed from the M Zhang Moxiang diagram in accordance with an embodiment of the application.
FIG. 6 is a schematic diagram of a predictive model according to an embodiment of the application.
Fig. 7 is a schematic block diagram of an unloading device based on low-orbit satellite edge calculation according to an embodiment of the application.
Fig. 8 is a schematic block diagram of an apparatus according to an embodiment of the present application.
Detailed Description
In order to describe the technical content, construction features, achieved objects and effects of the present application in detail, the following description is made with reference to the accompanying drawings.
Example 1
Referring to fig. 1, the application discloses an unloading strategy self-adaptive optimization system based on satellite-borne edge calculation, comprising: a plurality of ground terminals (UEs), a plurality of low-Orbit Satellites (LEOs) in which servers of an edge calculation function are disposed, and a high-orbit satellite (GEO) 30 in which delay data between inter-satellite links in the respective low-orbit satellites, calculation resource information of the respective low-orbit satellites, and the like are stored. The edge computing server architecture of the low orbit satellite comprises a satellite-borne infrastructure layer, a resource virtualization layer, an edge computing service layer and an edge computing management layer, wherein the satellite-borne infrastructure layer mainly comprises computing function facilities, network function facilities and storage function facilities in a satellite physical layer. The resource virtualization layer provides resource virtualization based on the existing physical layer resources, and virtualizes a single satellite-borne infrastructure resource into a plurality of satellite-borne infrastructure logic resources through a virtualization technology, wherein each logic resource can bear different computing services, and the computing services can run in mutually independent spaces without mutual influence. The edge computing service layer provides the service of space-borne edge computing. The edge computing management layer provides an unloading strategy function, a scheduling algorithm function, a resource management function and a task management function, and the unloading strategy function is mainly responsible for executing a planet carrier edge computing task unloading strategy and realizing adaptive optimization of the unloading strategy based on a prediction model deployed on a high orbit satellite.
Referring to fig. 1 to 6, the application discloses an unloading method based on low-orbit satellite edge calculation, which is applied to the system and comprises the following steps:
s1, deploying a prediction model for training and verifying on a high orbit satellite, wherein the prediction model is trained and verified by using a training set and a testing set which are divided by an input data set, the input data set is obtained by processing time delay historical data, and the time delay historical data comprises inter-satellite-ground topological relations, inter-satellite time delays and inter-satellite time delays which are acquired by a plurality of continuous time units. The inter-satellite-ground topological relation is the topological relation between each terminal and each low-orbit satellite and between each low-orbit satellite, the inter-satellite time delay is the time delay data between each terminal and each low-orbit satellite, and the inter-satellite time delay is the time delay data between each low-orbit satellite. The time unit may be defined according to practical situations, for example, may be defined as 5 minutes.
Referring to fig. 3 and 4, specifically, delay history data of each time unit is constructed as an adjacency matrix. In this example, S1 to S4 are low-orbit satellites, UE1 and UE2 are terminals, a value of 0 in the matrix indicates that there is no connection between nodes, and if it is not 0, it indicates inter-node link delay.
In order to obtain the time delay history data, data acquisition can be performed every 30 seconds, data of every 5 minutes (a time unit) are aggregated into an undirected graph, and the undirected graph is used for constructing an adjacency matrix.
Referring to fig. 5, further, the M Zhang Moxiang chart forms a piece of data, specifically configured as a temporal undirected graph (space-time diagram), to be input into the prediction model for prediction. Correspondingly, the data output by the prediction model is a graph of the next time unit.
In particular, the time delay history data for each time unit may be directly provided by a third party.
Referring to fig. 6, specifically, the prediction model includes an input layer, a time-space domain convolution module, and an output layer, where the time-space domain convolution module is responsible for performing time-space domain convolution processing on input data to obtain time features and space features.
More specifically, the prediction model includes two time-space domain convolution modules, each including two time-space domain convolution blocks and one space-space domain convolution block, the time-space domain convolution blocks being used to capture the temporal features and the space-space domain convolution blocks being used to capture the spatial features.
In a specific example, spatial features are spatially extracted by directly using the graph structure data to perform higher order feature extraction, K is the graph convolution kernel size, T k Is a polynomial expansion approximation of the Laplace matrix, θ k Is a polynomial coefficient:
the final graph convolution of data with C channels (channels) is expressed as:
C i and C o Is the size of the feature map of the input and output, the input is a satellite link traffic map comprising consecutive M time units, each time unit of the satellite link traffic map can be represented by a matrix, the dimension n x C, n representing n samples, C representing the feature dimension, e.g. C of column i i Dimension represents sample i C i Dimension characteristics.
For time features, a width K can be used t Is convolved to extract K t-1 Is a time characteristic of (a).
More specifically, the output layer includes a time domain gating convolution block for merging data of a time dimension and a full connection layer for outputting a prediction result.
S2, predicting delay future data of the next time unit by using a prediction model, wherein the delay future data comprises inter-satellite-ground topological relations, inter-satellite time delays and inter-satellite time delays.
And S3, calculating and acquiring an unloading task transmission link with the lowest time delay of each terminal in the next time unit by using time delay future data to serve as a real transmission link.
S4, updating the real time delay data (inter-satellite-ground topological relation, inter-satellite time delay and inter-satellite time delay) of the next time unit to time delay history data (namely, the time delay data of the earliest time unit is removed, and the real time delay data of the next time unit is complemented to the time delay data of the latest time unit), so that the updated time delay history data can be used for predicting the next time delay future data continuously.
Steps S2 to S4 are repeated.
In the application, an inter-satellite-ground topological relation, an inter-satellite time delay and an inter-satellite time delay which are acquired by a plurality of continuous time units are used as time delay historical data to train a prediction model which is deployed on a high-orbit satellite, the inter-satellite-ground topological relation, the inter-satellite time delay and the inter-satellite time delay of the next time unit are predicted by using the prediction model to calculate an unloading task transmission link with the lowest time delay of each terminal in the next time unit to be used as a real transmission link, and real time delay data of the next time unit is updated to the time delay historical data to continuously predict time delay future data and calculate the next real transmission link. By repeating the steps, the self-adaptive optimization of the low-orbit satellite edge calculation unloading strategy can be realized, an unloading task transmission link with the lowest time delay is always used as a real transmission link, congestion of an unloading task queue is not easy to cause, the satellite electric quantity consumption is reduced, the service quality and the calculation speed of the low-orbit satellite-borne edge calculation are improved, and meanwhile, the problems of unbalanced use of satellite-borne calculation resources, overlarge partial satellite load, insufficient cruising ability and the like can be effectively relieved.
Example two
Referring to fig. 7, the application discloses an unloading strategy optimization device based on low-orbit satellite edge calculation, which comprises:
the deployment module 100 is configured to deploy a prediction model for training and verifying on a high orbit satellite, where the prediction model is trained and verified by using a training set and a testing set divided by an input data set, the input data set is obtained by processing delay history data, and the delay history data includes inter-satellite-to-satellite topological relations, inter-satellite time delays and inter-satellite time delays collected by a plurality of continuous time units.
The prediction module 200 is configured to predict delay future data of a next time unit by using a prediction model, where the delay future data includes an inter-satellite-to-satellite topological relationship, an inter-satellite delay, and an inter-satellite delay.
And the acquisition module 300 is used for calculating and acquiring the task-off transmission link with the lowest time delay of each terminal in the next time unit by using the time delay future data as a real transmission link.
The updating module 400 updates the real time delay data of the next time unit to the time delay history data.
In the application, an inter-satellite-ground topological relation, an inter-satellite time delay and an inter-satellite time delay which are acquired by a plurality of continuous time units are used as time delay historical data to train a prediction model which is deployed on a high-orbit satellite, the inter-satellite-ground topological relation, the inter-satellite time delay and the inter-satellite time delay of the next time unit are predicted by using the prediction model to calculate an unloading task transmission link with the lowest time delay of each terminal in the next time unit to be used as a real transmission link, and real time delay data of the next time unit is updated to the time delay historical data to continuously predict time delay future data and calculate the next real transmission link. By repeating the steps, the self-adaptive optimization of the low-orbit satellite edge calculation unloading strategy can be realized, an unloading task transmission link with the lowest time delay is always used as a real transmission link, congestion of an unloading task queue is not easy to cause, the satellite electric quantity consumption is reduced, the service quality and the calculation speed of the low-orbit satellite-borne edge calculation are improved, and meanwhile, the problems of unbalanced use of satellite-borne calculation resources, overlarge partial satellite load, insufficient cruising ability and the like can be effectively relieved.
Referring to fig. 3 and 4, specifically, delay history data of each time unit is constructed as an adjacency matrix. In this example, S1 to S4 are low-orbit satellites, UE1 and UE2 are terminals, a value of 0 in the matrix indicates that there is no connection between nodes, and if it is not 0, it indicates inter-node link delay.
In order to obtain the time delay history data, data acquisition can be performed every 30 seconds, data of every 5 minutes (a time unit) are aggregated into an undirected graph, and the undirected graph is used for constructing an adjacency matrix.
Referring to fig. 5, further, the M Zhang Moxiang chart forms a piece of data, specifically configured as a temporal undirected graph (space-time diagram), to be input into the prediction model for prediction. Correspondingly, the data output by the prediction model is a graph of the next time unit.
In particular, the time delay history data for each time unit may be directly provided by a third party.
Referring to fig. 6, specifically, the prediction model includes an input layer, a time-space domain convolution module, and an output layer, where the time-space domain convolution module is responsible for performing time-space domain convolution processing on input data to obtain time features and space features.
More specifically, the prediction model includes two time-space domain convolution modules, each including two time-space domain convolution blocks and one space-space domain convolution block, the time-space domain convolution blocks being used to capture the temporal features and the space-space domain convolution blocks being used to capture the spatial features.
In a specific example, spatial features are spatially extracted by directly using the graph structure data to perform higher order feature extraction, K is the graph convolution kernel size, T k Is a polynomial expansion approximation of the Laplace matrix, θ k Is a polynomial coefficient:
the final graph convolution of data with C channels (channels) is expressed as:
C i and C o Is the size of the feature map of the input and output, the input is a satellite link traffic map comprising consecutive M time units, each time unit of the satellite link traffic map can be represented by a matrix, the dimension n x C, n representing n samples, C representing the feature dimension, e.g. C of column i i Dimension represents sample i C i Dimension characteristics.
For time features, a width K can be used t Is convolved to extract K t-1 Is a time characteristic of (a).
More specifically, the output layer includes a time domain gating convolution block for merging data of a time dimension and a full connection layer for outputting a prediction result.
Example III
Referring to fig. 8, the present application discloses an apparatus comprising:
a processor 30;
a memory 40 having stored therein executable instructions of the processor 30;
wherein the processor 30 is configured to perform the low-orbit satellite edge calculation based offloading method as described in embodiment one, via execution of executable instructions.
Example IV
The application discloses a computer readable storage medium, wherein a program is stored, and when the program is executed by a processor, the unloading method based on the low-orbit satellite edge calculation is realized.
Example five
Embodiments of the present application disclose a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the electronic device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions to cause the electronic device to perform the low-orbit satellite edge calculation based offloading method as described in embodiment one.
It should be appreciated that in embodiments of the present application, the processor may be a central processing module (CentralProcessing Unit, CPU), which may also be other general purpose processors, digital signal processors (DigitalSignal Processor, DSPs), application specific integrated circuits (Application SpecificIntegrated Circuit, ASICs), off-the-shelf programmable gate arrays (Field-Programmable Gate Array, FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
Those skilled in the art will appreciate that the processes implementing all or part of the methods of the above embodiments may be implemented by hardware associated with computer program instructions, and the program may be stored in a computer readable storage medium, where the program when executed may include processes of embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-only memory (ROM), a Random access memory (Random AccessMemory, RAM), or the like.
In the foregoing embodiments, the descriptions of the embodiments are emphasized, and for parts of one embodiment that are not described in detail, reference may be made to the related descriptions of other embodiments.
The foregoing disclosure is only illustrative of the preferred embodiments of the present application and is not to be construed as limiting the scope of the application, which is defined by the appended claims.
Claims (10)
1. An unloading method based on low-orbit satellite edge calculation, comprising:
s1, deploying a prediction model for training and verifying on a high orbit satellite, wherein the prediction model is trained and verified by using a training set and a testing set which are divided by an input data set, the input data set is obtained by processing time delay historical data, and the time delay historical data comprises a continuous inter-satellite-to-satellite topological relation, a satellite-to-ground time delay and an inter-satellite time delay of a plurality of time units;
s2, predicting delay future data of the next time unit by using the prediction model, wherein the delay future data comprises inter-satellite-ground topological relations, inter-satellite time delays and inter-satellite time delays;
s3, calculating and acquiring an unloading task transmission link with the lowest time delay of each terminal in the next time unit by using the time delay future data to serve as a real transmission link;
s4, updating the real time delay data of the next time unit to the time delay historical data;
steps S2 to S4 are repeated.
2. The low-orbit satellite edge based offloading method of claim 1, wherein the time delay history data for each of the time units is constructed as a adjacency matrix.
3. The unloading method based on low-orbit satellite edge calculation according to claim 1, wherein the prediction model comprises an input layer, a time-space domain convolution module and an output layer, and the time-space domain convolution module is responsible for performing time-space domain convolution processing on input data to obtain time features and space features.
4. The low orbit satellite edge based offloading method of claim 3, wherein the prediction model comprises two time-space domain convolution modules, each comprising two time-space domain convolution blocks and one space-domain convolution block, the time-domain convolution blocks being used to capture the temporal features and the space-domain convolution blocks being used to capture the spatial features.
5. The method of unloading based on low-orbit satellite edge computation according to claim 3, wherein the output layer comprises a time-domain-gating convolution block for merging data in a time dimension and a full-connection layer for outputting a prediction result.
6. An unloading strategy optimization device based on low-orbit satellite edge calculation, which is characterized by comprising:
the deployment module is used for deploying a prediction model for training and verifying on a high-orbit satellite, the prediction model is trained and verified by using a training set and a testing set which are divided by an input data set, the input data set is obtained by processing time delay historical data, and the time delay historical data comprises an inter-satellite-ground topological relation, an inter-satellite time delay and an inter-satellite time delay which are acquired by a plurality of continuous time units;
the prediction module is used for predicting time delay future data of the next time unit by utilizing the prediction model, wherein the time delay future data comprises an inter-satellite-ground topological relation, an inter-satellite time delay and an inter-satellite time delay;
the acquisition module is used for calculating and acquiring an unloading task transmission link with the lowest time delay of each terminal in the next time unit by using the time delay future data to serve as a real transmission link;
and the updating module is used for updating the real time delay data of the next time unit to the time delay historical data.
7. The low-orbit satellite edge based offloading policy optimization device of claim 6, wherein the latency history data for each of the time units is constructed as a adjacency matrix.
8. The unloading strategy optimization device based on low-orbit satellite edge calculation according to claim 6, wherein the prediction model comprises an input layer, a time-space domain convolution module and an output layer, and the time-space domain convolution module is responsible for performing time-space domain convolution processing on input data to obtain time features and space features.
9. An apparatus, comprising:
a processor;
a memory having stored therein executable instructions of the processor;
wherein the processor is configured to perform the low-orbit satellite edge calculation based offloading method of any one of claims 1 to 5 via execution of the executable instructions.
10. A computer readable storage medium having stored thereon a program, wherein the program when executed by a processor implements the low-orbit satellite edge calculation based offloading method of any one of claims 1 to 5.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310359205.8A CN116366133B (en) | 2023-04-06 | 2023-04-06 | Unloading method and device based on low-orbit satellite edge calculation |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310359205.8A CN116366133B (en) | 2023-04-06 | 2023-04-06 | Unloading method and device based on low-orbit satellite edge calculation |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116366133A CN116366133A (en) | 2023-06-30 |
CN116366133B true CN116366133B (en) | 2023-10-27 |
Family
ID=86941753
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310359205.8A Active CN116366133B (en) | 2023-04-06 | 2023-04-06 | Unloading method and device based on low-orbit satellite edge calculation |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116366133B (en) |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108809839A (en) * | 2018-07-17 | 2018-11-13 | 湖南理工学院 | Wireless Mesh backbone network network flow control methods and device |
CN109714263A (en) * | 2019-01-18 | 2019-05-03 | 北京邮电大学 | A kind of routing resource and device in satellite communication network |
CN114866133A (en) * | 2022-05-12 | 2022-08-05 | 重庆邮电大学 | Computing unloading method for satellite cloud edge collaborative computing |
KR102484004B1 (en) * | 2022-08-23 | 2022-12-30 | 재단법인대구경북과학기술원 | Method and apparatus for performing task offloading between terminal and satellite in mec network |
CN115622608A (en) * | 2022-09-29 | 2023-01-17 | 广州爱浦路网络技术有限公司 | Method, system and medium for optimization of offloading strategies based on low-earth-orbit satellite edge calculation |
CN115633332A (en) * | 2022-10-26 | 2023-01-20 | 南京熊猫电子股份有限公司 | Vehicle-mounted satellite communication terminal system and communication method under edge unloading |
CN115720226A (en) * | 2022-09-08 | 2023-02-28 | 杭州电子科技大学 | Time delay perception resource reservation method based on space-time task demand prediction |
CN115733542A (en) * | 2022-11-24 | 2023-03-03 | 北京信息科技大学 | Inter-satellite link path planning method, device and medium |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11395308B2 (en) * | 2019-04-30 | 2022-07-19 | Fujitsu Limited | Monitoring-based edge computing service with delay assurance |
-
2023
- 2023-04-06 CN CN202310359205.8A patent/CN116366133B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108809839A (en) * | 2018-07-17 | 2018-11-13 | 湖南理工学院 | Wireless Mesh backbone network network flow control methods and device |
CN109714263A (en) * | 2019-01-18 | 2019-05-03 | 北京邮电大学 | A kind of routing resource and device in satellite communication network |
CN114866133A (en) * | 2022-05-12 | 2022-08-05 | 重庆邮电大学 | Computing unloading method for satellite cloud edge collaborative computing |
KR102484004B1 (en) * | 2022-08-23 | 2022-12-30 | 재단법인대구경북과학기술원 | Method and apparatus for performing task offloading between terminal and satellite in mec network |
CN115720226A (en) * | 2022-09-08 | 2023-02-28 | 杭州电子科技大学 | Time delay perception resource reservation method based on space-time task demand prediction |
CN115622608A (en) * | 2022-09-29 | 2023-01-17 | 广州爱浦路网络技术有限公司 | Method, system and medium for optimization of offloading strategies based on low-earth-orbit satellite edge calculation |
CN115633332A (en) * | 2022-10-26 | 2023-01-20 | 南京熊猫电子股份有限公司 | Vehicle-mounted satellite communication terminal system and communication method under edge unloading |
CN115733542A (en) * | 2022-11-24 | 2023-03-03 | 北京信息科技大学 | Inter-satellite link path planning method, device and medium |
Non-Patent Citations (1)
Title |
---|
基于端边云协同的边缘计算任务卸载研究;刘黎明;《中国优秀硕士学位论文全文数据库 (信息科技辑)》;第3章 * |
Also Published As
Publication number | Publication date |
---|---|
CN116366133A (en) | 2023-06-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Shabayek et al. | A simulation model for the Kwai Chung container terminals in Hong Kong | |
Pérez et al. | A resilient and distributed near real-time traffic forecasting application for Fog computing environments | |
CN112188442A (en) | Vehicle networking data-driven task unloading system and method based on mobile edge calculation | |
US10417558B1 (en) | Methods and systems for artificial neural network optimistic event processing | |
CN108923983B (en) | Method and device for predicting opportunistic network link and readable storage medium | |
US10853720B1 (en) | Traffic condition forecasting using matrix compression and deep neural networks | |
CN112884086B (en) | Model training method, device, equipment, storage medium and program product | |
CN114265631A (en) | Mobile edge calculation intelligent unloading method and device based on federal meta-learning | |
CN114626503A (en) | Model training method, target detection method, device, electronic device and medium | |
CN112612553A (en) | Container technology-based edge computing task unloading method | |
CN112887905A (en) | Task unloading method based on periodic resource scheduling in Internet of vehicles | |
Mohammed et al. | Smart edge computing for 5 g/6 g satellite iot for reducing inter transmission delay | |
Strelkovskaya et al. | Different extrapolation methods in Problems of Forecasting | |
CN116366133B (en) | Unloading method and device based on low-orbit satellite edge calculation | |
CN112819258A (en) | Bank branch to store customer quantity prediction method and device | |
CN112534399A (en) | Semantic-based Internet of things equipment data processing related application program installation method and device | |
CN112860402A (en) | Dynamic batch processing task scheduling method and system for deep learning inference service | |
CN115907136B (en) | Electric automobile dispatching method, device, equipment and computer readable medium | |
CN116663717A (en) | Traffic flow prediction model training and traffic flow prediction method, device and equipment | |
CN112287178A (en) | Data cleaning method and device, electronic equipment and storage medium | |
CN108770014B (en) | Calculation evaluation method, system and device of network server and readable storage medium | |
White et al. | Smart cities with deep edges | |
Sirojan et al. | Enabling deep learning on embedded systems for iot sensor data analytics: Opportunities and challenges | |
CN115220131A (en) | Meteorological data quality inspection method and system | |
US11755302B2 (en) | Application placement device and application placement program |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |