CN115361048B - Giant low-orbit constellation serverless edge computing task arrangement method and device - Google Patents
Giant low-orbit constellation serverless edge computing task arrangement method and device Download PDFInfo
- Publication number
- CN115361048B CN115361048B CN202210767651.8A CN202210767651A CN115361048B CN 115361048 B CN115361048 B CN 115361048B CN 202210767651 A CN202210767651 A CN 202210767651A CN 115361048 B CN115361048 B CN 115361048B
- Authority
- CN
- China
- Prior art keywords
- satellite
- task
- total cost
- function
- cost
- 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
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
-
- 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
-
- 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
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Astronomy & Astrophysics (AREA)
- Aviation & Aerospace Engineering (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Radio Relay Systems (AREA)
Abstract
The invention provides a giant low-orbit constellation serverless edge computing task arrangement method and a device, wherein the method comprises the following steps: receiving a user request by any satellite in a low orbit constellation, and constructing a virtual satellite sub-network by taking the satellite as a reference satellite, wherein the virtual satellite sub-network comprises a plurality of satellite satellites adjacent to the reference satellite; dividing a request task corresponding to a user request into a plurality of function service subtasks by a reference satellite, and respectively transmitting each function service subtask to different satellites in a virtual satellite sub-network; satellites in the virtual satellite subnetwork construct a cold start function environment or a hot start function environment based on the service type of the requested task; starting a satellite where the function service subtask is located to execute the function service subtask based on the execution sequence of the function service subtask in the request task; and outputting a task result by the satellite executing the end function service subtask, transmitting the task result to the reference satellite, and transmitting the task result to the user terminal by the reference satellite.
Description
Technical Field
The invention relates to the technical field of low-orbit satellite communication, in particular to a giant low-orbit constellation serverless edge computing task arrangement method and device.
Background
With the vigorous development of new-generation communication and network technologies such as 5G, wi-Fi6, a large number of novel user services such as Internet of vehicles, VR (Virtual Reality)/AR (Augmented Reality), 4K/8K, smart city and the like are generated, and the services greatly improve the life quality of people and promote the social development. In sharp contrast, however, telecommunications operators often do not choose to provide telecommunications capability over a full-coverage terrestrial network that is costly to build in remote mountainous areas as well as in less developed villages, making it increasingly difficult for users in these areas to be served by rapidly evolving high quality communications and networks.
In this case, satellite communication is considered as an effective technique for solving the above-mentioned challenges due to its good performance in terms of abundant radio frequency resources, large coverage, long communication distance, small ground interference, and the like. Especially in recent years, with the rapid development of the aerospace industry and automation technology, huge low-orbit constellation networks are rapidly developed. The global coverage characteristic can greatly improve satellite communication capacity, and has great potential in the aspect of communication broadband; in addition, the jumbo low rail constellation can improve the quality of service with lower signal propagation delay, which means that it is potential to process delay-sensitive and computation-intensive traffic from mobile users quickly and accurately. Therefore, the deployment and application of the huge low-orbit constellation to provide the network service processing capability has become an important development direction for further providing large-scale services in the industry.
Although the giant low-orbit constellation network has the advantages of short propagation delay, low development cost and the like, in order to provide high-quality network service, the giant low-orbit constellation still has a plurality of challenges represented by task arrangement to be solved. Firstly, in order to ensure seamless coverage on the global scale, the configuration of a huge low orbit constellation often adopts a dense mixed constellation design method, the resource deployment modes of the heterogeneous satellites are different, the single satellite has limited processing capacity for large-scale service, the traditional task arrangement mode mainly comprises the steps of uniformly transmitting a received request to a satellite controller, and distributing a certain satellite to complete a task of the request by the controller, so that the task distribution is completed by transmitting a request to the controller each time, and the processing efficiency is lower.
Disclosure of Invention
In view of the foregoing, embodiments of the present invention provide a method for arranging a huge low-orbit constellation serverless edge calculation task, which obviates or mitigates one or more of the drawbacks of the related art.
The invention provides a giant low-orbit constellation serverless edge computing task scheduling method, which comprises the following steps:
receiving a user request by any satellite in a low orbit constellation, and constructing a virtual satellite sub-network by taking the satellite as a reference satellite, wherein the virtual satellite sub-network comprises a plurality of satellite satellites adjacent to the reference satellite;
Dividing a request task corresponding to a user request into a plurality of function service subtasks by a reference satellite, and respectively transmitting each function service subtask to different satellites in a virtual satellite sub-network;
satellites in the virtual satellite subnetwork construct a cold start function environment or a hot start function environment based on the service type of the requested task;
based on the execution sequence of the function service subtasks in the request task, respectively starting the satellites where the function service subtasks are positioned to execute the function service subtasks, so that data are transmitted among a plurality of satellites in the virtual satellite subnetwork;
the satellite executing the end function service subtask outputs the task result of the request task, the task result is transmitted to the reference satellite, and the reference satellite sends the task result to the user terminal.
By adopting the scheme, on one hand, the received user request is not required to be sent to the satellite controller for task arrangement, task arrangement work can be directly completed by the received satellite, and the reference satellite can split the request task corresponding to the user request into a plurality of function service subtasks and send the function service subtasks to different satellites for execution, so that compared with the traditional mode of processing the task by only one satellite, the task processing efficiency is improved; on the other hand, the scheme can meet the requirements of different service types, such as time delay sensitive and computation intensive services, construct different function environments, provide various solutions and improve the applicability to different service types.
In some embodiments of the present invention, the step of receiving a user request by any satellite in the low orbit constellation, constructing a virtual satellite subnet using the satellite as a reference satellite comprises:
and acquiring the distance between the satellite in the low orbit constellation and the reference satellite, and taking the satellite with the distance smaller than the preset distance threshold value from the reference satellite as the satellite in the virtual satellite sub-network.
In some embodiments of the present invention, the step of splitting a request task corresponding to a user request into a plurality of function service subtasks by a reference satellite, and sending each function service subtask to a different satellite in a virtual satellite subnet, respectively, includes:
constructing a plurality of preselection schemes based on the function service subtasks and satellites of the virtual satellite subnetwork;
calculating a total cost of computing resources and a total cost of time delay for each pre-selected scheme, the total cost being calculated based on the total cost of resources and the total cost of time delay;
a pre-selected scheme with a total cost less than a pre-set cost threshold is taken as the final allocation scheme.
In some embodiments of the present invention, calculating a total cost of computing resources and a total cost of latency for each pre-selected scheme, the step of calculating a total cost based on the total cost of resources and the total cost of latency comprises:
and respectively acquiring the transmission delay cost and the processing delay cost of the function service subtasks transmitted to the satellite, and calculating the delay total cost of each preselected scheme based on the transmission delay cost and the processing delay cost.
In some embodiments of the present invention, calculating a total cost of computing resources and a total cost of latency for each pre-selected scheme, the step of calculating a total cost based on the total cost of resources and the total cost of latency comprises:
and calculating the calculation resource cost of each satellite processing function service subtask in the preselection scheme to obtain the calculation resource total cost of each preselection scheme.
In some embodiments of the present invention, the total cost of computing resources and the total cost of delay for each pre-selected scheme is calculated according to the following formula, the total cost being calculated based on the total cost of resources and the total cost of delay:
x represents the total number of function service subtasks; m represents the number of satellite satellites in the virtual satellite subnetwork; a, a ij Representing a decision parameter whether to sub-task x the function is to be serviced in a preselected scheme i Assigned to satellites M j If yes, a ij =1, otherwise a ij =0; α represents a delay cost weight, β represents a computation cost weight, α and β are both greater than 0, α+β=1; p is p i Representing function services subtask x i Bit size of (a); q i Representing function services subtask x i The size of the cpu resource required; l (L) rj Representing any satellite M in a sub-network of reference satellites and virtual satellites j Is a transmission delay cost of (a); t (T) ji Representing execution function service subtask x i Satellite M of (2) j The processing delay cost of (2); gamma ray ij Representing execution parameters, the execution of the function service subtasks needs to satisfy the execution order, if the task x is to be executed i Distribution to low-orbit satellite edge calculation nodes M j Satisfying the execution order, then γ ij =1, if not, γ ij →∞；R ji Representing a satellite M that performs this function service subtask j Is a cost of computing resources; delta represents the total cost.
In some embodiments of the present invention,for the total cost of the delay,is the total cost of the resource.
In some embodiments of the present invention, the step of constructing a cold start function environment or a hot start function environment by satellites in the virtual satellite subnetwork based on the service type of the requested task includes:
the reference satellite acquires the time delay requirement of the user request, and if the parameter of the time delay requirement is smaller than a preset time delay threshold value, the service type of the request task is determined to be time delay sensitive;
for time delay sensitive request tasks, the satellite employs a hot start function environment.
In some embodiments of the present invention, the step of constructing a cold start function environment or a hot start function environment by satellites in the virtual satellite subnetwork based on the service type of the requested task includes:
the reference satellite acquires the size of the computing resource required by the user request, and if the size of the computing resource required is larger than a preset resource threshold, the service type of the requested task is determined to be computationally intensive;
For computationally intensive request tasks, the satellite employs a cold start function environment.
In some embodiments of the present invention, in calculating the total cost of the computing resources and the total cost of the delay for each pre-selected scheme, in the formula for calculating the total cost based on the total cost of the resources and the total cost of the delay,
if the service type of the request task is time delay sensitive, the time delay cost weight is greater than the calculation cost weight;
if the service type of the requested task is computationally intensive, the computational cost weight is greater than the latency cost weight.
In another aspect, the present invention further provides a giant low-rail constellation serverless edge computing task orchestration device, which includes a computer device including a processor and a memory, where the memory stores computer instructions, and the processor is configured to execute the computer instructions stored in the memory, where the device implements the steps implemented by the method as described above when the computer instructions are executed by the processor.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description and drawings.
It will be appreciated by those skilled in the art that the objects and advantages that can be achieved with the present application are not limited to the above-described specific ones, and that the above and other objects that can be achieved with the present application will be more clearly understood from the following detailed description.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate and together with the description serve to explain the application.
FIG. 1 is a schematic diagram of an embodiment of a method for arranging edge computing tasks of a giant low-orbit constellation without server according to the present application;
FIG. 2 is a schematic diagram of another embodiment of the method for arranging edge computing tasks of a giant low-orbit constellation without server according to the present application;
FIG. 3 is a schematic diagram of the splitting and execution of function service subtasks;
FIG. 4 is a schematic diagram of a structure of a prior art 1;
FIG. 5 is a schematic diagram of a structure of prior art 2;
fig. 6 is a schematic structural view of the present application.
Detailed Description
The present application will be described in further detail with reference to the following embodiments and the accompanying drawings, in order to make the objects, technical solutions and advantages of the present application more apparent. The exemplary embodiments of the present application and the descriptions thereof are used herein to explain the present application, but are not intended to limit the application.
It should be noted here that, in order to avoid obscuring the present invention due to unnecessary details, only structures and/or processing steps closely related to the solution according to the present invention are shown in the drawings, while other details not greatly related to the present invention are omitted.
It should be emphasized that the term "comprises/comprising" when used herein is taken to specify the presence of stated features, elements, steps or components, but does not preclude the presence or addition of one or more other features, elements, steps or components.
It is also noted herein that the term "coupled" may refer to not only a direct connection, but also an indirect connection in which an intermediate is present, unless otherwise specified.
Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the drawings, the same reference numerals represent the same or similar components, or the same or similar steps.
Introduction to the prior art:
prior Art 1
The Low-orbit satellite edge computing LEC (Low-Orbit Edge computing) network taking the satellite as an edge computing network node deploys edge computing resources on the Low-orbit satellite, and data generated by the Internet of things equipment can be directly processed by the Low-orbit satellite. When an access satellite of the internet of things device (or its neighboring satellites) has the requested processing function, the end-to-end path between the internet of things device and the satellite providing the processing function does not include too many inter-satellite links. Therefore, bandwidth of these inter-satellite links can be saved. Furthermore, data processing delays will be greatly reduced because the user can obtain computing services directly from the satellite (LEC node).
As shown in fig. 4, the LEC system consists of a low orbit satellite constellation, a satellite network gateway and a user side. To achieve multi-hop communication between satellites, switching and routing equipment is installed on the satellites. The satellite network gateway is deployed on the ground. It provides an interconnection between a terrestrial network and a satellite network. The low orbit satellite controller is placed in the satellite network gateway, and a local control agent is deployed on each satellite. The user terminal may obtain edge computing services from the LEC system, for example, data generated by the internet of things device may be processed by the LEC satellite nodes.
Disadvantages of Prior Art 1
1. The task orchestration decision process of the LEC scheme is too dependent on the centralized decision of the satellite network gateway, and although the user side can directly obtain the service provided by the satellite node, the task orchestration decision needs to be determined by the satellite network gateway with global topology information. This approach not only brings the gateway single point of failure problem, but also increases the transmission link length and redundant traffic. In addition, the method for making the scheduling decision by the global optimization method has no strong practical application significance in the existing huge low-orbit constellation scene, and because the limited tasks of mobility and strong interference can be generally processed cooperatively by a small number of adjacent satellites, the high-performance collaborative task scheduling decision can be realized only by solving the local optimum of the satellites close to the user.
2. The allocation and use mode of satellite resources is a pre-request-allocation mode of traditional edge calculation, and allocation and arrangement are carried out according to the resource quantity of service requests, which may cause the situation that the service does not fully use the allocated resources, and waste of satellite calculation storage resources with high cost is caused. In addition, when the service runs on the node close to full load and the actual used resources exceed the pre-allocation amount, the capacity expansion and the cooperative processing of the resources cannot be performed, and the problem of wasting resources while the task processing fails is caused.
3. It is difficult to handle computationally intensive tasks using multi-star distributed collaboration. LECs are very complex to partition and schedule subtasks based on container granularity because of the need to consider many factors, such as available computing resources and remaining energy for all satellites, link delays and capacity between all satellites. Furthermore, there may be different optimization objectives, such as minimizing task completion delays, balancing satellite loads, and minimizing communication overhead.
PRIOR ART 2
As shown in fig. 5, in the scheme of task scheduling of the giant low-orbit satellite constellation for mixing cloud computing and edge computing, the user computing task schedules the processing computing in different manners according to different requirements. For computationally intensive tasks that cannot be handled by ground users, the computational tasks may be offloaded to the giant low-orbit constellation over a wireless link or forwarded to a cloud server for processing over a low-orbit satellite. For computing tasks that the ground user has the ability to handle, it can calculate itself.
Disadvantages of Prior Art 2
1. The task arrangement mode requires that each ground user has only one calculation task to be calculated, the calculation tasks cannot be divided, the service characteristics and the requirements of the users in actual situations are not met, and in addition, the separation of the calculation intensive tasks and the collaborative arrangement processing of the calculation intensive tasks among satellites cannot be realized in a mechanism principle.
2. The scheme mainly considers the influence of the arrangement mode on task processing time and energy consumption, but the improvement of the processing efficiency of the low-orbit satellite part is realized by sharing the calculation tasks by cloud calculation and local calculation in principle, but the arrangement mode does not essentially improve the resource utilization rate of the huge low-orbit constellation, but is not applicable to remote areas with extremely weak local processing capacity because of the problem of vacant satellite calculation storage resources caused by cloud and local priority processing.
In order to solve the above problems, as shown in fig. 1, 2 and 6, the present invention provides a method for arranging a huge low-orbit constellation serverless edge computing task, which comprises the following steps:
step S100, receiving a user request by any satellite in a low orbit constellation, and constructing a virtual satellite sub-network by taking the satellite as a reference satellite, wherein the virtual satellite sub-network comprises a plurality of satellite satellites adjacent to the reference satellite;
An end user includes all individual devices or clusters of devices that use the jumbo low orbit constellation serverless edge computing network services, including not only land devices but also marine and aerial end devices. The terminal user sends a service request to the huge low orbit constellation serverless computing network, a certain low orbit satellite in the huge low orbit constellation receives the request and then serves as a serverless computing node to arrange the task, and the result is fed back to the terminal user after the processing of the huge low orbit constellation network, so that the computing task is completed.
In some embodiments of the present invention, the user request is a service request, and may be a location service or a data screening service.
All low-orbit satellites in the giant low-orbit constellation together form a complete giant low-orbit constellation distributed edge network. Each low orbit satellite is provided with a low orbit satellite serverless edge computing platform for processing serverless computing services requested by an end user, pulling and deploying functions, configuration and maintenance processing environments thereof and the like from a network. In addition, because the giant low-orbit constellation can cover the end user in the global scope, no matter the end user sends a request to any satellite in the giant low-orbit constellation network, the satellite can be used as a reference satellite, and a server-free edge computing platform on the satellite can immediately and autonomously respond to service, arrange tasks and start function operation processing. The distributed management and control network architecture and capability can ensure the dynamic autonomous capability of the huge low-orbit constellation distributed edge network, and the satellite network improves the global processing efficiency and the resource utilization rate through collaborative arrangement and common processing of function tasks.
In some embodiments of the present invention, the server-free edge computing platform on any satellite in the low orbit constellation is provided with a network controller and an orchestration controller, where the network controller is used for sensing network link topology and implementing network management and control, and includes maintenance of own network state information and routing information such as a forwarding next-hop satellite relay; the scheduling controller performs function splitting and task scheduling work of the terminal user service, cooperates with the network controller to divide the service processing virtual satellite sub-network, then performs task scheduling process according to scheduling strategy, receives and combines the server-free computing service processing result after the cooperative processing of each low orbit satellite in the sub-network, and finally feeds back the processing result to the user through the platform to complete the whole task processing process.
Step S200, dividing a request task corresponding to a user request into a plurality of function service subtasks by a reference satellite, and respectively transmitting each function service subtask to different satellites in a virtual satellite subnet;
in some embodiments of the present invention, the request task may be a model training task, and the plurality of function service subtasks corresponding to the model training task may include sequentially executed data collection, data screening, model training, and the like, and the function service subtasks may be respectively sent to different satellites in the virtual satellite subnetwork for execution, and further, if the data collection includes a plurality of steps, the data collection may also be divided into a plurality of function service subtasks again.
In some embodiments of the present invention, if two data screening methods are required at the same time, two different screening methods may be respectively allocated to different satellites as function service subtasks.
In some embodiments of the present invention, the dividing manner of the function service subtasks may be preset by a technician according to the requirements.
Step S300, satellites in the virtual satellite subnetwork construct a cold start function environment or a hot start function environment based on the service type of the requested task;
in some embodiments of the present invention, if the service type of the requested task is a time delay sensitive type, such as a positioning service in automatic driving, the time delay requirement is higher, a hot start function environment is adopted, if the hot start function environment is constructed, the satellite starts to pull function images required by respective processing function operations from the network and configures an operation environment thereof, such as a download map in the positioning service, when receiving the task to be completed; if the cold start function environment is constructed, only when the request task is executed to the function service subtask responsible for the satellite, the function mirror images required by the respective processing function operation are pulled from the network and the running environment is configured.
By adopting the scheme, the timeliness of processing the time delay sensitive type request task can be ensured, and the processing speed is improved; however, for the request task with lower time delay requirement, the environment does not need to be built in advance, and the workload is increased by building the environment in advance.
Step S400, based on the execution sequence of the function service subtasks in the request task, respectively starting the satellites where the function service subtasks are located to execute the function service subtasks, so that data are transferred among a plurality of satellites in the virtual satellite sub-network;
in some embodiments of the present invention, if the request task is a model training task, the multiple function service subtasks corresponding to the model training task may include sequentially performed data collection, data screening and model training, and the data collection, data screening and model training are respectively sent to three satellites a, b and c, then the data collected by a is sent to b, the data is sent to c after b screening, and c performs training based on the data, and the trained model is fed back to the client.
The method comprises the steps of splitting a function service without a server according to functions required to be realized by a request task to obtain a function service subtask, wherein the function service subtask is in the meaning of splitting a complete service into a group of function operations, each low-orbit satellite edge computing node can configure a part or all of function execution required running environments to process related functions, and finally, all the processed function operations are combined into a complete service result to be provided for a terminal user. The splitting of function services is mathematically a directed acyclic graph (Directed acyclic graph, DAG) problem, as shown in fig. 3, splitting f into several function services subtasks { x } 1 ,x 2 ,…,x i …, it can be seen that there may be a call order between these function service subtasks, i.e. the output of one function service subtask may be the input of another function service subtask, the directed path set d= { D in the figure 1 ,d 2 ,…,d i ,…,d n Then the dependency relationship between function service subtasks is represented, such as directed path d 1 Representing function services subtask x 1 Serving subtask x as a function of output 2 Thus the function serves subtask x 2 Must serve subtask x at function 1 And triggering after the operation is finished.
Step S500, the satellite executing the end function service subtask outputs the task result of the request task, the task result is transmitted to the reference satellite, and the reference satellite sends the task result to the user side.
In some embodiments of the present invention, the function service subtask at the end is the last function service subtask in the execution order of the function service subtask.
By adopting the scheme, on one hand, the received user request is not required to be sent to the satellite controller for task arrangement, task arrangement work can be directly completed by the received satellite, and the reference satellite can split the request task corresponding to the user request into a plurality of function service subtasks and send the function service subtasks to different auxiliary satellites for execution, so that compared with the traditional mode of processing the task by only one satellite, the task processing efficiency is improved; on the other hand, the scheme can meet the requirements of different service types, such as time delay sensitive and computation intensive services, construct different function environments, provide various solutions and improve the applicability of different service types.
In some embodiments of the present invention, the step of receiving a user request by any satellite in the low orbit constellation, constructing a virtual satellite subnet using the satellite as a reference satellite comprises:
and acquiring the distance between the satellite in the low orbit constellation and the reference satellite, and taking the satellite with the distance smaller than the preset distance threshold value from the reference satellite as the satellite in the virtual satellite sub-network.
The relative motion between low-orbit satellites can cause unstable links, so that dynamic network topology and frequent route update influence the stability of inter-satellite task collaborative arrangement, and the traditional global-range task arrangement method is excessively high in complexity and difficult to adapt to the large-scale time-varying state of the constellation network;
the giant low orbit constellation has the characteristics of numerous low orbit satellites, distributed network structure, rapid dynamic change, frequent inter-satellite interaction and the like. In the practical application scene, besides the independent processing of the local satellite, the low-orbit satellite can only select other few satellites with relatively close peripheral distances to cooperatively process the calculation service, but generally does not select the satellite with relatively far link distances through repeated forwarding, so that network jitter and data loss caused by relatively strong inter-satellite channel interference and relatively long-distance satellite position movement can be avoided.
In some embodiments of the present invention, the step of splitting a request task corresponding to a user request into a plurality of function service subtasks by a reference satellite, and sending each function service subtask to a different satellite in a virtual satellite subnet, respectively, includes:
constructing a plurality of preselection schemes based on the function service subtasks and satellites of the virtual satellite subnetwork;
calculating a total cost of computing resources and a total cost of time delay for each pre-selected scheme, the total cost being calculated based on the total cost of resources and the total cost of time delay;
a pre-selected scheme with a total cost less than a pre-set cost threshold is taken as the final allocation scheme.
In some embodiments of the present invention, in the step of constructing the preselection schemes, the satellites in each of the preselection schemes are capable of meeting the latency requirements and computational resource requirements required by the function service subtasks in matching of the function service subtasks.
In some embodiments of the present invention, in the step of constructing a plurality of pre-selected scenarios based on the function service subtasks and satellites of the virtual satellite subnetwork, a plurality of pre-selected scenarios may be obtained in a permutation and combination manner, and each pre-selected scenario may be represented as a decision matrix a= [ a ] ij ] X×M ，a ij Representing a decision parameter whether to sub-task x the function is to be serviced in a preselected scheme i Assigned to satellite m j If yes, a ij =1, otherwise a ij =0, x denotes the total number of function service subtasks, M denotes the number of satellite satellites in the virtual satellite subnetwork, i starting from 1, i.e. the function service subtasks include (x 1 、x 2 … …), j is calculated from 0, i.e. the satellites of the virtual satellite subnetwork comprise (m 0 、m 1 ……)，m 0 A reference satellite may be represented.
a 10 =1 means that the function is served by subtask x 1 Assigned to reference satellite m 0 The decision matrix may ensure that each function is assigned a subtask and only to a unique low-orbit satellite edge compute node.
The reference satellite may also be assigned a function service subtask in this scenario.
In some embodiments of the present invention, calculating a total cost of computing resources and a total cost of latency for each pre-selected scheme, the step of calculating a total cost based on the total cost of resources and the total cost of latency comprises:
and respectively acquiring the transmission delay cost and the processing delay cost of the function service subtasks transmitted to the satellite, and calculating the delay total cost of each preselected scheme based on the transmission delay cost and the processing delay cost.
In some embodiments of the present invention, the propagation delay cost may be estimated by a reference satellite based on a distance, and the processing delay cost of the satellite may be estimated according to the size of the function service subtasks and the cpu parameters of the satellite.
In some embodiments of the present invention, calculating a total cost of computing resources and a total cost of latency for each pre-selected scheme, the step of calculating a total cost based on the total cost of resources and the total cost of latency comprises:
and calculating the calculation resource cost of each satellite processing function service subtask in the preselection scheme to obtain the calculation resource total cost of each preselection scheme.
In some embodiments of the invention, the computational resource cost of each satellite may be the size of cpu required to serve the subtasks for the satellite processing functions.
In some embodiments of the present invention, the total cost of computing resources and the total cost of delay for each pre-selected scheme is calculated according to the following formula, the total cost being calculated based on the total cost of resources and the total cost of delay:
x represents the total number of function service subtasks; m represents the number of satellite satellites in the virtual satellite subnetwork; a, a ij Representing a decision parameter whether to sub-task x the function is to be serviced in a preselected scheme i Assigned to satellites M j If yes, a ij =1, otherwise a ij =0; α represents a delay cost weight, β represents a computation cost weight, α and β are both greater than 0, α+β=1; p is p i Representing function services subtask x i Bit size of (a); q i Representing function services subtask x i The size of the cpu resource required; l (L) rj Representing reference satellite and any satellite in a virtual satellite sub-networkM j Is a transmission delay cost of (a); t (T) ji Representing execution function service subtask x i Satellite M of (2) j The processing delay cost of (2); gamma ray ij Representing execution parameters, the execution of the function service subtasks needs to satisfy the execution order, if the task x is to be executed i Distribution to low-orbit satellite edge calculation nodes M j Satisfying the execution order, then γ ij =1, if not, γ ij →∞；R ji Representing a satellite M that performs this function service subtask j Is a cost of computing resources; delta represents the total cost.
In some embodiments of the present invention,for the total cost of the delay,is the total cost of the resource.
In some embodiments of the invention, the above formula may also be expressed as Where E represents the expectation of the random variable.
With the above scheme, the optimization problem proposed by the method is an NP-hard problem, so that a near-optimal final allocation scheme can be obtained from a plurality of pre-selected schemes through some solving algorithms such as a dynamic integer programming algorithm, a branch delimitation algorithm, a greedy algorithm and a heuristic algorithm.
In some embodiments of the present invention, the step of constructing a cold start function environment or a hot start function environment by satellites in the virtual satellite subnetwork based on the service type of the requested task includes:
The reference satellite acquires the time delay requirement of the user request, and if the parameter of the time delay requirement is smaller than a preset time delay threshold value, the service type of the request task is determined to be time delay sensitive;
for time delay sensitive request tasks, the satellite employs a hot start function environment.
In some embodiments of the present invention, the step of constructing a cold start function environment or a hot start function environment by satellites in the virtual satellite subnetwork based on the service type of the requested task includes:
the reference satellite acquires the size of the computing resource required by the user request, and if the size of the computing resource required is larger than a preset resource threshold, the service type of the requested task is determined to be computationally intensive;
for computationally intensive request tasks, the satellite employs a cold start function environment.
By adopting the scheme, after the final distribution scheme is issued in the virtual satellite sub-network, each satellite needs to pull the function mirror image required by each processing function service sub-task from the network and configure the running environment of the function mirror image, if the satellite node stores the related function mirror image, the function service sub-task is not required to be pulled again, the function service sub-task is started in a hot start mode and a cold start mode, and the function operation starting mode depends on whether the service type of the request task belongs to a time delay sensitive mode or a computation intensive mode. For time delay sensitive service, all environments required by function operation are pre-configured immediately by each low-orbit satellite edge node after a final allocation scheme is received, a container example is started in advance according to mirror images, service can be immediately processed when input parameters of a function service subtask come, and the hot start mode can accelerate the processing process to ensure the stability of time delay; and for the computation-intensive business, a cold start mode is adopted, when parameters required by the function service subtask are transmitted to the edge node of the low-orbit satellite, the container instance is started again to process the user service, and under the condition of meeting the non-harsh time delay requirement, the function service subtask of the computation-intensive business is processed in a short time in a concentrated manner so as to ensure the computation-intensive high-quality resource requirement.
In some embodiments of the present invention, if the requested task is not delay sensitive, a cold start function environment is employed.
In some embodiments of the present invention, in calculating the total cost of the computing resources and the total cost of the delay for each pre-selected scheme, in the formula for calculating the total cost based on the total cost of the resources and the total cost of the delay,
if the service type of the request task is time delay sensitive, the time delay cost weight is greater than the calculation cost weight;
if the service type of the requested task is computationally intensive, the computational cost weight is greater than the latency cost weight.
For each low-orbit satellite edge computing node in the virtual satellite sub-network, after all function operations born by the node are processed completely and the processing result is sent to the low-orbit satellite processed by the next step, the low-orbit satellite can release resources and clear the configured container environment without waiting for the resource environment to be cleared after all function operations are processed completely. The purpose of this is to free up as soon as possible the computational storage resources of the low-orbit satellites that do not need to be processed for use by other user service processes. And finally, after all the function operations are finished, the reference satellite sends a notification of the disassembly of the virtual satellite subnetwork to all the low-orbit satellites in the virtual satellite subnetwork, and all the low-orbit satellites are restored to the original normal running state.
Aiming at the problems of low task scheduling resource utilization efficiency and poor cooperative processing effect of the current giant low-orbit constellation, the proposal refers to the ideas of non-server computing (Serverless computing) and mobile edge computing (Mobile Edge Computing, MEC) in a ground network, introduces a non-server edge computing technology in the giant low-orbit constellation to solve the problems, and provides a solution idea for meeting the demands of time delay sensitive and computation intensive services. The core idea is that on one hand, a function in server-free computing, namely a service (Functions as a Service, faaS) method is utilized to disassemble user service into a function-independent stateless function, and the function-independent stateless function is dynamically processed in an event-triggered mode, so that the on-demand allocation and deployment of resources are realized, and the utilization rate of the resources is improved. On the other hand, the network scheduling function is sunk to the heterogeneous low-orbit satellite network edge by utilizing edge calculation, the calculation processing request of the user is responded quickly, flexible task scheduling between adjacent satellites is carried out, and the service experience quality (QoE, quality ofExperience) of the user is improved.
At present, the application of a server-free computing technology in a giant low-orbit constellation edge network is still in a starting stage, most of the existing schemes focus on the aspects of independently constructing an edge computing system or a cloud-edge combined platform architecture design, mobility management and scheduling in the giant low-orbit constellation, and how to further invent a server-free edge computing task scheduling scheme of the giant low-orbit constellation so as to realize the efficient utilization of constellation resources is still blank. Therefore, the scheme focuses on the huge low orbit constellation scene, designs a server-free edge computing task arrangement network and a method, and realizes flexible deployment of server-free computing service in the huge low orbit constellation and efficient utilization of heterogeneous satellite computing resources.
The beneficial effects of the scheme are that
1. The scheme provides the distributed management and decision capability of realizing the network on each satellite in the huge low orbit constellation, and solves the problems of complex service flow and large redundant flow caused by frequent interaction between an end user and the low orbit satellite and the centralized control arrangement controller in the traditional centralized control constellation edge network.
2. The invention provides a scheme for deploying a low-orbit satellite serverless edge computing platform with network control, arrangement control and serverless computing service processing functions on all low-orbit satellites in a giant low-orbit constellation, so as to combine edge computing and serverless computing in the giant low-orbit constellation to rapidly and efficiently process end user services.
3. The invention splits the request task of the terminal user into a group of function operations corresponding to the function service subtasks, considers the time delay and the resource requirement of the service to divide and process the virtual satellite subnetwork of the service, and arranges the function operations to a plurality of low orbit satellites in the virtual subnetwork for cooperative processing, thereby realizing the improvement of service performance and the efficient utilization of resources.
4. The invention provides the whole architecture of the huge low-orbit constellation serverless edge computing distributed network structure, the functional modules thereof and the communication flow and organization relation between the modules.
5. The invention provides a complete full-flow mechanism for processing a calculation task without a server edge in a huge low-orbit constellation network, which comprises the following steps: the method comprises the steps of initiating a service request by an end user, dividing the service to process a virtual satellite subnet, splitting a function service, distributing an arrangement decision, configuring a function operation starting environment, cooperating with a function operation and combining processing results, releasing resources and dismissing the virtual subnet.
6. The invention considers whether the service requirement of the terminal user belongs to time delay sensitivity or computation intensive, flexibly selects the starting mode of the function environment, the weight ratio of the time delay cost and the computation resource cost, and needs to meet the matching of the function operation dependency relationship, so as to obtain the approximate solution of the optimization task arrangement model to make the calculation task arrangement decision without a server edge, thereby meeting the high-performance requirement of the service delay and the resource of the terminal user.
The embodiment of the invention also provides a giant low-orbit constellation serverless edge computing task arrangement device, which comprises computer equipment, wherein the computer equipment comprises a processor and a memory, the memory stores computer instructions, the processor is used for executing the computer instructions stored in the memory, and the device realizes the steps realized by the method when the computer instructions are executed by the processor.
The embodiment of the invention also provides a computer readable storage medium, on which a computer program is stored, which when being executed by a processor, is used for realizing the steps realized by the method for arranging the edge calculation tasks of the huge low-orbit constellation without a server. The computer readable storage medium may be a tangible storage medium such as Random Access Memory (RAM), memory, read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, floppy disks, hard disk, a removable memory disk, a CD-ROM, or any other form of storage medium known in the art.
Those of ordinary skill in the art will appreciate that the various illustrative components, systems, and methods described in connection with the embodiments disclosed herein can be implemented as hardware, software, or a combination of both. The particular implementation is hardware or software dependent on the specific application of the solution and the design constraints. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention. When implemented in hardware, it may be, for example, an electronic circuit, an Application Specific Integrated Circuit (ASIC), suitable firmware, a plug-in, a function card, or the like. When implemented in software, the elements of the invention are the programs or code segments used to perform the required tasks. The program or code segments may be stored in a machine readable medium or transmitted over transmission media or communication links by a data signal carried in a carrier wave.
It should be understood that the invention is not limited to the particular arrangements and instrumentality described above and shown in the drawings. For the sake of brevity, a detailed description of known methods is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method processes of the present invention are not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between steps, after appreciating the spirit of the present invention.
In this disclosure, features that are described and/or illustrated with respect to one embodiment may be used in the same way or in a similar way in one or more other embodiments and/or in combination with or instead of the features of the other embodiments.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, and various modifications and variations can be made to the embodiments of the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. The method for arranging the edge computing tasks of the huge low-orbit constellation without the server is characterized by comprising the following steps:
receiving a user request by any satellite in a low orbit constellation, and constructing a virtual satellite sub-network by taking the satellite as a reference satellite, wherein the virtual satellite sub-network comprises a plurality of satellite satellites adjacent to the reference satellite;
dividing a request task corresponding to a user request into a plurality of function service subtasks by a reference satellite, and respectively transmitting each function service subtask to different satellites in a virtual satellite sub-network;
satellites in the virtual satellite subnetwork construct a cold start function environment or a hot start function environment based on the service type of the requested task;
based on the execution sequence of the function service subtasks in the request task, respectively starting the satellites where the function service subtasks are positioned to execute the function service subtasks, so that data are transmitted among a plurality of satellites in the virtual satellite subnetwork;
the satellite executing the end function service subtask outputs the task result of the request task, the task result is transmitted to the reference satellite, and the reference satellite sends the task result to the user terminal.
2. The server-less edge computing task orchestration method for a giant low-orbit constellation according to claim 1, wherein the step of receiving a user request by any satellite in the low-orbit constellation and constructing a virtual satellite subnetwork with the satellite as a reference satellite comprises:
And acquiring the distance between the satellite in the low orbit constellation and the reference satellite, and taking the satellite with the distance smaller than the preset distance threshold value from the reference satellite as the satellite in the virtual satellite sub-network.
3. The method for arranging the edge computing tasks of the huge low orbit constellation without server according to claim 1, wherein the step of dividing the request task corresponding to the user request into a plurality of function service sub-tasks by the reference satellite and sending each function service sub-task to different satellites in the virtual satellite sub-network respectively comprises the following steps:
constructing a plurality of preselection schemes based on the function service subtasks and satellites of the virtual satellite subnetwork;
calculating a total cost of computing resources and a total cost of time delay for each pre-selected scheme, the total cost being calculated based on the total cost of resources and the total cost of time delay;
a pre-selected scheme with a total cost less than a pre-set cost threshold is taken as the final allocation scheme.
4. A method of server-less edge computing task orchestration for a giant low-orbit constellation according to claim 3, wherein the step of computing a total cost of computing resources and a total cost of delay for each pre-selected scheme, the total cost being computed based on the total cost of resources and the total cost of delay comprises:
and respectively acquiring the transmission delay cost and the processing delay cost of the function service subtasks transmitted to the satellite, and calculating the delay total cost of each preselected scheme based on the transmission delay cost and the processing delay cost.
5. The method of server-less edge computing task orchestration for a giant low-orbit constellation according to claim 4, wherein the step of computing a total cost of computing resources and a total cost of latency for each pre-selected scheme, the step of computing a total cost based on the total cost of resources and the total cost of latency comprises:
and calculating the calculation resource cost of each satellite processing function service subtask in the preselection scheme to obtain the calculation resource total cost of each preselection scheme.
6. The method of server-less edge computing task orchestration for a giant low-orbit constellation according to claim 5, wherein the total cost of computing resources and the total cost of delay for each pre-selected scheme is calculated according to the following formula, and the total cost is calculated based on the total cost of resources and the total cost of delay:
x represents a function service subtaskA total number; m represents the number of satellite satellites in the virtual satellite subnetwork; a, a ij Representing a decision parameter whether to sub-task x the function is to be serviced in a preselected scheme i Assigned to satellites M j If yes, a ij =1, otherwise a ij =0; α represents a delay cost weight, β represents a computation cost weight, α and β are both greater than 0, α+β=1; p is p i Representing function services subtask x i Bit size of (a); q i Representing function services subtask x i The size of the cpu resource required; l (L) rj Representing any satellite M in a sub-network of reference satellites and virtual satellites j Is a transmission delay cost of (a); t (T) ji Representing execution function service subtask x i Satellite M of (2) j The processing delay cost of (2); gamma ray ij Representing execution parameters, the execution of the function service subtasks needs to satisfy the execution order, if the task x is to be executed i Distribution to low-orbit satellite edge calculation nodes M j Satisfying the execution order, then γ ij =1, if not, γ ij →∞；R ji Representing a satellite M that performs this function service subtask j Is a cost of computing resources; delta represents the total cost.
7. The method for arranging the edge computing tasks of the giant low-orbit constellation serverless according to any one of claims 1 to 6, wherein the step of constructing a cold start function environment or a hot start function environment by satellites in the virtual satellite subnetwork based on the service type of the requested task comprises:
the reference satellite acquires the time delay requirement of the user request, and if the parameter of the time delay requirement is smaller than a preset time delay threshold value, the service type of the request task is determined to be time delay sensitive;
for time delay sensitive request tasks, the satellite employs a hot start function environment.
8. The method for arranging the edge computing tasks of the huge low orbit constellation without server according to claim 1, wherein the step of constructing a cold start function environment or a hot start function environment by satellites in the virtual satellite subnetwork based on the service type of the requested task comprises:
The reference satellite acquires the size of the computing resource required by the user request, and if the size of the computing resource required is larger than a preset resource threshold, the service type of the requested task is determined to be computationally intensive;
for computationally intensive request tasks, the satellite employs a cold start function environment.
9. The server-less edge computing task orchestration method for the giant low-orbit constellations of claim 6, wherein in computing the total cost of computing resources and the total cost of delay for each pre-selected scheme, in a formula for computing the total cost of resources and the total cost of delay based on the total cost of resources and the total cost of delay,
if the service type of the request task is time delay sensitive, the time delay cost weight is greater than the calculation cost weight;
if the service type of the requested task is computationally intensive, the computational cost weight is greater than the latency cost weight.
10. A giant low-rail constellation serverless edge computing task orchestration device, comprising a computer device comprising a processor and a memory, the memory having stored therein computer instructions for executing the computer instructions stored in the memory, the device implementing the steps implemented by the method according to any one of claims 1-9 when the computer instructions are executed by the processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210767651.8A CN115361048B (en) | 2022-07-01 | 2022-07-01 | Giant low-orbit constellation serverless edge computing task arrangement method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210767651.8A CN115361048B (en) | 2022-07-01 | 2022-07-01 | Giant low-orbit constellation serverless edge computing task arrangement method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115361048A CN115361048A (en) | 2022-11-18 |
| CN115361048B true CN115361048B (en) | 2023-08-15 |
Family
ID=84030501
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210767651.8A Active CN115361048B (en) | 2022-07-01 | 2022-07-01 | Giant low-orbit constellation serverless edge computing task arrangement method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115361048B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116722912A (en) * | 2023-08-09 | 2023-09-08 | 银河航天(北京)网络技术有限公司 | Block chain-based satellite distributed operation method, device and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110647391A (en) * | 2019-09-27 | 2020-01-03 | 北京邮电大学 | Edge computing method and system for satellite-ground cooperative network |
| CN112799784A (en) * | 2021-02-01 | 2021-05-14 | 军事科学院系统工程研究院系统总体研究所 | Low-orbit satellite network optimal task allocation method based on decentralized computing |
| WO2021221736A2 (en) * | 2020-05-01 | 2021-11-04 | Intel Corporation | Edge computing in satellite connectivity environments |
| CN113794494A (en) * | 2021-07-30 | 2021-12-14 | 湖北微源卓越科技有限公司 | Edge computing architecture and computing unloading optimization method for low-earth-orbit satellite network |
| CN113839704A (en) * | 2021-09-30 | 2021-12-24 | 东南大学 | Mobile edge calculation method for integration of dense low-earth orbit satellite and land |
| CN114268357A (en) * | 2021-11-28 | 2022-04-01 | 西安电子科技大学 | Low-earth-orbit-satellite-based edge computing task unloading method, system, equipment and application |
| WO2022135478A1 (en) * | 2020-12-25 | 2022-06-30 | 第四范式（北京）技术有限公司 | Task orchestration method and orchestration apparatus |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20220004950A1 (en) * | 2020-07-02 | 2022-01-06 | Raytheon Company | Distributed orchestration for satellite groups and other groups |
-
2022
- 2022-07-01 CN CN202210767651.8A patent/CN115361048B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110647391A (en) * | 2019-09-27 | 2020-01-03 | 北京邮电大学 | Edge computing method and system for satellite-ground cooperative network |
| WO2021221736A2 (en) * | 2020-05-01 | 2021-11-04 | Intel Corporation | Edge computing in satellite connectivity environments |
| WO2022135478A1 (en) * | 2020-12-25 | 2022-06-30 | 第四范式（北京）技术有限公司 | Task orchestration method and orchestration apparatus |
| CN112799784A (en) * | 2021-02-01 | 2021-05-14 | 军事科学院系统工程研究院系统总体研究所 | Low-orbit satellite network optimal task allocation method based on decentralized computing |
| CN113794494A (en) * | 2021-07-30 | 2021-12-14 | 湖北微源卓越科技有限公司 | Edge computing architecture and computing unloading optimization method for low-earth-orbit satellite network |
| CN113839704A (en) * | 2021-09-30 | 2021-12-24 | 东南大学 | Mobile edge calculation method for integration of dense low-earth orbit satellite and land |
| CN114268357A (en) * | 2021-11-28 | 2022-04-01 | 西安电子科技大学 | Low-earth-orbit-satellite-based edge computing task unloading method, system, equipment and application |
Non-Patent Citations (1)
| Title |
|---|
| Integrating Edge Computing into Low Earth Orbit Satellite Networks: Architecture and Prototype;Chengcheng Li;《IEEE Access 》;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115361048A (en) | 2022-11-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110647391B (en) | Edge computing method and system for satellite-ground cooperative network | |
| CN109714219B (en) | Virtual network function rapid mapping method based on satellite network | |
| Zhang et al. | Satellite mobile edge computing: Improving QoS of high-speed satellite-terrestrial networks using edge computing techniques | |
| Hu et al. | Ready player one: UAV-clustering-based multi-task offloading for vehicular VR/AR gaming | |
| CN109257091B (en) | Global load balancing satellite-ground cooperative network networking device and method | |
| Garg et al. | MobQoS: Mobility-aware and QoS-driven SDN framework for autonomous vehicles | |
| Sahni et al. | Multihop offloading of multiple DAG tasks in collaborative edge computing | |
| Cui et al. | Latency optimization for hybrid GEO–LEO satellite-assisted IoT networks | |
| CN109947574A (en) | A kind of vehicle big data calculating discharging method based on mist network | |
| Tang et al. | Federated learning for intelligent transmission with space-air-ground integrated network toward 6G | |
| CN115361048B (en) | Giant low-orbit constellation serverless edge computing task arrangement method and device | |
| CN116547648A (en) | Method and apparatus for supporting application mobility in a multiple access edge computing platform architecture | |
| Pourghasemian et al. | AI-based mobility-aware energy efficient resource allocation and trajectory design for NFV enabled aerial networks | |
| CN112020001A (en) | Time slot resource allocation method for multi-station multi-machine system of unmanned aerial vehicle | |
| CN113114335B (en) | Software-defined space-based network networking architecture based on artificial intelligence | |
| Nguyen et al. | Multi-agent DRL-based task offloading in hierarchical HAP-LAP networks | |
| Shukla et al. | Software-defined network based resource allocation in distributed servers for unmanned aerial vehicles | |
| CN109547345B (en) | Software-defined airborne network system and content-driven routing method | |
| CN115459828B (en) | Autonomous migration method of service for low-orbit constellation edge micro cloud | |
| CN114710200B (en) | Satellite network resource arrangement method and system based on reinforcement learning | |
| Liu et al. | Research on Dynamic Spectrum Allocation of Space-Air-Ground Integration | |
| Lin et al. | State machine with tracking tree and traffic allocation scheme based on cumulative entropy for satellite network | |
| CN115664487A (en) | Intelligent air-ground integrated network based on brain neuron aggregation architecture | |
| Murugan et al. | Efficient Space Communication and Management (SCOaM) Using Cognitive Radio Networks Based on Deep Learning Techniques: Cognitive Radio in Space Communication | |
| Ling et al. | An optimized forward scheduling algorithm in a software-defined satellite network |
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 |