Computing resource scheduling method based on network cloud collaboration

Abstract

The invention discloses a computing resource scheduling method based on network cloud collaboration, which comprises a central cloud node, an edge cloud node and a mobile node, and the mobile node applies for an intelligent computing resource scheduling service. The central cloud node intelligently operates four computing resource scheduling modes of multi-machine redundancy, window switching, grading processing and continuous service to schedule network cloud resources, so that on-demand centralized optimal use of the computing resources based on the network cloud system is realized, and the capabilities such as reliability, safety, real-time performance and the like of using the cloud computing resources by the mobile node are effectively improved; and a solid guarantee is provided for information acquisition, data processing, analysis decision, communication service and the like of the user.

Description

Technical Field

[0001] The present invention relates to computing resource scheduling technology, and in particular to a computing resource scheduling method based on network cloud collaboration. Background Art

[0002] Distributed cloud computing resources are an important innovation in the field of modern information technology. By distributing computing resources, storage resources, and network resources in data centers across various locations, it achieves elastic expansion and efficient utilization of computing power. As business needs grow, users can easily obtain computing resources from cloud services without having to worry about the limitations of physical hardware and high maintenance costs and operating expenses. At the same time, distributed deployment also enhances the system's fault tolerance, anti-destruction, and reliability.

[0003] Distributed wide area network, this patent mainly studies the distributed wide area network with wireless characteristics, using advanced wireless transmission technology (such as 4G / 5G / 6G) to achieve cross-regional and large-scale wireless connections. This network architecture overcomes the physical limitations and high costs of traditional limited networks. At present, relying on three-dimensional means of space and air, it can achieve wireless coverage communication in large areas and has high-speed and stable network service capabilities, which is also a necessary means to achieve mobile communication.

[0004] Network-cloud collaboration is the product of the deep integration of network information technology and cloud computing. It closely combines network resources with cloud computing capabilities to achieve efficient data transmission, storage and processing, as well as flexible resource allocation and sharing. In this process, the network, as the circulation blood vessel of information, provides infrastructure support for cloud computing services, while cloud computing provides power for various applications with its powerful computing power and flexible expansion capabilities.

[0005] At present, in the network-cloud collaboration scenario, users are mainly classified into two categories according to their mobility: fixed users and mobile users. Mobile users need to access the wireless communication devices of various communication nodes at will during their mobility to realize data communication transmission. In this process, the access communication nodes can change dynamically to ensure that users can communicate anytime and anywhere.

[0006] To sum up, after the integration of distributed wireless wide area network and cloud computing, the capabilities of the network and the cloud support and coordinate with each other. In this process, users who want to obtain the best cloud computing resources, especially mobile users, need to choose appropriate cloud resource scheduling methods according to business characteristics. In this process, the cloud resource scheduling method needs to be coordinated with network access and network changes. This coordination method will need to solve coordination problems in various scenarios, which is relatively complex and difficult. Therefore, it is urgently needed to realize efficient scheduling of computing resources for network-cloud collaboration based on network-cloud characteristics and business characteristics. Summary of the invention

[0007] In view of the problems raised by the background technology, the present invention discloses a computing resource scheduling method based on network cloud collaboration, which includes a central cloud node, an edge cloud node, and a mobile node. Among them, the edge cloud node and the central cloud node communicate data through a wireless link, the edge cloud node and the mobile node communicate data through a wireless link, the edge cloud node and the edge cloud node communicate data through a wireless link, the edge cloud node is composed of an edge cloud facility and an edge communication device, the central cloud node is composed of a central cloud facility and a central communication device, and the mobile node is composed of an application terminal and a mobile communication device; the application terminal needs cloud computing resources during work, which are provided by the edge cloud facility and the central cloud facility; the central cloud facility perceives the network status of the entire network and the status of the entire cloud computing resources in real time, and the central cloud facility uniformly schedules the network cloud resources, and defaults to providing the mobile node with the scheduling service of the nearest resources. After the mobile node applies for the intelligent computing resource scheduling service, it will intelligently run the four computing resource scheduling modes of multi-machine redundancy, window switching, upgrade processing, and continued service according to the authority range according to the system capability opening strategy and the mobile node identity authority to schedule the network cloud resources.

[0008] The system capability opening strategy refers to the strategy for scheduling and allocating computing resources for the capacity configuration of the central cloud facility. By establishing a mapping relationship from the node type and node identity authority dimensions, the scheduling of computing resources with different capabilities can be realized, including the management of the scheduling and execution of the four strategies of multi-machine redundancy, window switching, upgraded processing, and continued service. For example, from the perspective of node type, some nodes are only allowed to use multi-machine redundancy and window switching strategies, and some nodes are allowed to use upgraded processing strategies; from the perspective of identity, some nodes with certain identities are only allowed to continue service, and some nodes with certain identities can allow four types of services.

[0009] The network cloud collaborative system is characterized by being distributed, large-scale, and heterogeneous in communication means. When affected by bad weather, electromagnetic interference, terrain shielding, etc., the communication link will be unstable; and when the mobile nodes of the network cloud collaborative system move quickly, the edge cloud nodes that the mobile nodes directly communicate with will change; or a certain edge cloud node will be directly disconnected and inaccessible due to communication link reasons. Under the above conditions, the cloud computing resources of the mobile node run on a certain edge cloud node, and the anti-destruction capability and robustness are difficult to guarantee.

[0010] This patent designs the multi-machine redundant computing resource scheduling service capability. The central cloud facility perceives the network situation of the entire network and the situation of the entire cloud computing resource in real time. The central cloud facility calculates and generates a multi-machine redundant computing resource scheduling strategy, and provides two sets of cloud computing resources, primary and backup, for mobile nodes that need redundant computing resource scheduling services. Under normal circumstances, the primary computing resource handles the computing tasks of the mobile node, and the backup computing resource will synchronize the parameters and status information of the primary computing resource in real time. When the primary computing resource cannot provide services, the central cloud facility will promote the backup computing resource to the primary computing resource, and the mobile node will quickly obtain the computing service of the backup computing resource. At the same time, the central cloud facility schedules and distributes the backup computing resources in other edge cloud nodes.

[0011] All cloud computing resources refer to all cloud computing resources in the network cloud collaborative system. An overlay network is established between the central cloud facilities and each edge cloud node. The basic connectivity of the physical network is used to create a network cloud collaborative control logic network on it to ensure the rapid scheduling of the central cloud node to each edge cloud node, the reliability of path optimization, and the programmability.

[0012] The process of the central cloud facility computing and generating a multi-machine redundant computing resource scheduling strategy is as follows: S101, the application terminal of the mobile node sends an information request to the central cloud facility, requesting the central cloud facility to provide it with a multi-machine redundant computing resource scheduling service; S102, the central cloud facility determines whether to provide the service according to the system capability opening policy. If not, a service denial message is sent and the process ends; if provided, a service confirmation message is sent and the process proceeds to the next step; S103. The central cloud facility perceives the status of the entire network and the status of all cloud computing resources in real time, allocates two sets of cloud computing resources, primary and backup, to the application terminals, and dynamically keeps the primary computing resources mainly allocated to the edge cloud node closest to the mobile node, and the backup computing resources mainly allocated to the edge cloud node with the best signal evaluation with the mobile node at the next moment.

[0013] Mobile nodes have the characteristic of high-speed mobility. In the process of movement, in order to ensure the fast and efficient transmission of information, cloud computing resources can quickly serve mobile nodes and realize time-sensitive network cloud collaboration. It is necessary to schedule and allocate cloud computing resources to the best location so that computing resources can provide services nearby. When mobile nodes need to provide fast and agile services, they will initiate a window switching computing resource scheduling request, and the central cloud facility will calculate and generate a window switching computing resource scheduling strategy.

[0014] The process of switching computing resource scheduling strategy of the central cloud facility computing generation window is as follows: S201, the application terminal of the mobile node sends an information request to the central cloud facility, requesting the central cloud facility to provide it with a window switching computing resource scheduling service; S202, the central cloud facility determines whether to provide the service according to the system capability opening policy. If not, a service denial message is sent and the process ends; if provided, a service confirmation message is sent and the process proceeds to the next step; S203, the central cloud facility perceives the network situation of the entire network and the situation of the entire cloud computing resources in real time, allocates cloud computing resources to application terminals, and allocates computing resources to edge cloud nodes with the best signal evaluation with mobile nodes; S204: When the edge cloud node with the best signal evaluation with the mobile node changes, dynamically migrate the cloud computing resources to the new edge cloud node based on the migration change strategy.

[0015] Edge cloud nodes are divided into large and small nodes according to their scale and capabilities. Large nodes are mainly composed of platforms such as satellites, airships, airplanes, ships, and fixed base stations, and small nodes are composed of low-weight platforms such as portable stations, small drones, and small unmanned vehicles. The processing capabilities of some edge cloud nodes are limited. In some cases, mobile nodes will quickly cross multiple edge cloud nodes. The rapid switching of computing resources of edge cloud nodes will result in large management overhead and low response speed. This patent designs an upgraded processing mode in the computing resource scheduling mode, in which cloud computing resources will rise to the upper-level edge cloud node or central cloud node.

[0016] The process of computing resource scheduling strategy for central cloud facility computing generation and upgrading is as follows: S301, the application terminal of the mobile node sends an information request to the central cloud facility, requesting the central cloud facility to provide it with upgraded processing computing resource scheduling services; S302, the central cloud facility determines whether to provide the service according to the system capability opening policy. If not, a service denial message is sent and the process ends; if provided, a service confirmation message is sent and the process proceeds to the next step; S303. The central cloud facility dispatches and allocates the cloud computing resources serving the mobile node to the central cloud node, or dispatches and allocates them to the edge cloud node between the central cloud node and the currently serving edge cloud node.

[0017] The computing resource scheduling strategy is designed to continue the service strategy. This strategy is used to meet the requirements of key data and key operation calculations. The computing resources are relatively fixed to ensure the consistency of data and computing resources, so as to improve the reliability and security of computing services. The process of generating the computing resource scheduling strategy for continued service by the central cloud facility computing is as follows: S401, the application terminal of the mobile node sends an information request to the central cloud facility, requesting the central cloud facility to provide it with a continuous service computing resource scheduling service; S402, the central cloud facility determines whether to provide the service according to the system capability opening policy. If not, a service denial message is sent and the process ends; if provided, a service confirmation message is sent and the process proceeds to the next step; S403, the central cloud facility perceives the network status of the entire network and the status of all cloud computing resources in real time, and allocates the computing resources of the nearest edge cloud node to the mobile node for the first time. When the edge cloud node that the mobile node is connected to for nearby communication changes, no computing resource scheduling is performed, but only the spare computing capacity of the edge cloud node that the mobile node is connected to is sent to the mobile node. The computing resource capacity includes idle computing CPU, idle memory, idle storage resources, network delay, and network bandwidth information.

[0018] The beneficial technical effect of the present invention is: the mobile node of the present invention accesses the network cloud collaborative system at any time through wireless communication means, applies for four types of computing resource scheduling strategies such as multi-machine redundancy, window switching, upgraded processing, and continued service on demand, perceives the global situation through cloud center facilities, intelligently schedules and allocates computing resources, and realizes the on-demand and optimized use of computing resources based on the network cloud system, effectively improving the reliability, security, real-time and other capabilities of mobile nodes in using cloud computing resources, and providing solid guarantees for user information acquisition, data processing, analysis and decision-making, and communication services. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings described herein are used to provide further understanding of the present invention and constitute a part of this application. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0020] Attached Figure 1 , one A system networking diagram of a computing resource scheduling method based on network-cloud collaboration; Attached Figure 2 , one A multi-machine redundancy strategy step diagram of a computing resource scheduling method based on network cloud collaboration; Attached Figure 3 , one A window switching strategy step diagram of a computing resource scheduling method based on network-cloud collaboration; Attached Figure 4 , one A step diagram of an upgraded processing strategy for a computing resource scheduling method based on network-cloud collaboration; Attached Figure 5 , one A diagram of the steps of a continuous service strategy for a computing resource scheduling method based on network-cloud collaboration. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] See attached Figure 1 As shown, this embodiment provides a system networking diagram of a computing resource scheduling method based on network cloud collaboration. In the figure, the edge cloud node is composed of edge cloud facilities and edge communication equipment, the central cloud node is composed of central cloud facilities and central communication equipment, and the mobile node is composed of application terminals and mobile communication equipment; wherein the central cloud node, the edge cloud node, and the mobile cloud node are interconnected by wireless communication means to realize data communication and transmission. In the figure, the central cloud node and the edge cloud node can be fixed infrastructure, including towers, houses, or mobile platforms, including flying platforms, aviation platforms, vehicle platforms, and ship platforms. The mobile node uses a vehicle as an example, which also includes other embodiments such as aircraft. In the figure, mobile node D is interconnected with edge cloud node B at time T0 and with edge cloud node C at time T1. In this process, mobile node D is connected to the communication equipment of edge cloud node C from edge cloud node B. At the same time, the access capability of mobile node D to edge cloud computing resources has changed. The data transmission bandwidth of accessing the cloud computing resources of edge cloud node B is reduced and the delay is increased. The data transmission bandwidth of accessing the cloud computing resources of edge cloud node C is increased and the delay is reduced.

[0023] See attached Figure 2 As shown in FIG. 1 , this embodiment provides a multi-machine redundancy strategy step diagram of a computing resource scheduling method based on network cloud collaboration. Figure 1 For example, the steps are as follows: Step S101, mobile node D is connected to edge cloud node B at time T0, and initiates a multi-machine redundant computing resource scheduling service application to the central cloud facility of central cloud node A; Step S102: The central cloud facility of the central cloud node A determines whether to provide multi-machine redundant computing resource scheduling service according to the node type and identity authority information of the mobile node D; Step S103, the central cloud facility of the central cloud node A perceives the network situation of the entire network and the situation of the entire cloud computing resources, allocates two sets of cloud computing resources, primary and backup, to the application terminal, and dynamically keeps the primary computing resources mainly allocated to the edge cloud node closest to the mobile node, and the backup computing resources are mainly allocated to the edge cloud node with the best signal evaluation with the mobile node at the next moment. At time T0, edge cloud node B is the primary and edge cloud node C is the backup. At time T1, edge cloud node C is the primary and edge cloud node B is the backup. See attached Figure 3 As shown in FIG. 1 , this embodiment provides a window switching strategy step diagram of a computing resource scheduling method based on network-cloud collaboration. Figure 1 For example, the steps are as follows: S201, mobile node D is connected to edge cloud node B at time T0, and initiates a window switching computing resource scheduling service application to the central cloud facility of central cloud node A; S202, the central cloud facility determines whether to provide the service according to the system capability opening policy. If not, a service denial message is sent and the process ends; if provided, a service confirmation message is sent and the process proceeds to the next step; S203, the central cloud facility perceives the network situation and the situation of the entire cloud computing resources in real time, allocates cloud computing resources to the application terminal of the mobile node D, and allocates computing resources to the edge cloud node B with the best signal evaluation with the mobile node at time T0; S204, when the edge cloud node with the best signal evaluation with the mobile node changes, dynamically migrate the cloud computing resources to the new edge cloud node based on the migration change strategy, and allocate computing resources to the edge cloud node C with the best signal evaluation with the mobile node at time T1; See attached Figure 4 As shown in FIG. 1 , this embodiment provides a step diagram of a processing strategy for upgrading a computing resource scheduling method based on network-cloud collaboration. Figure 1 For example, the steps are as follows: S301, mobile node D is connected to edge cloud node B at time T0, and initiates a processing computing resource scheduling service to the central cloud facility of central cloud node A; S302, the central cloud facility determines whether to provide the service according to the system capability opening policy. If not, a service denial message is sent and the process ends; if provided, a service confirmation message is sent and the process proceeds to the next step; S303: The central cloud facility dispatches and allocates the cloud computing resources serving the mobile nodes to the central cloud node A.

[0024] See attached Figure 5 As shown in FIG. 1 , this embodiment provides a continuous service strategy step diagram of a computing resource scheduling method based on network cloud collaboration. Figure 1 For example, the steps are as follows: S401, mobile node D is interconnected with edge cloud node B at time T0, and initiates a continued service computing resource scheduling service to the central cloud facility of central cloud node A; S402, the central cloud facility determines whether to provide the service according to the system capability opening policy. If not, a service denial message is sent and the process ends; if provided, a service confirmation message is sent and the process proceeds to the next step; S403, the central cloud facility perceives the network status of the entire network and the status of all cloud computing resources in real time, and allocates the computing resources of the nearest edge cloud node B to the mobile node D for the first time. When the edge cloud node for the nearest communication access of the mobile node D changes to the edge cloud node C, no computing resource scheduling is performed, but only the spare computing capacity of the edge cloud node C for the nearest communication access of the mobile node D is sent to the mobile node D. The computing resource capacity includes idle computing CPU, idle memory, idle storage resources, network delay, and network bandwidth information.

Claims

1. A computing resource scheduling method based on network-cloud collaboration, characterized in that: It includes central cloud nodes, edge cloud nodes, and mobile nodes. The edge cloud nodes communicate with the central cloud nodes through wireless links, the edge cloud nodes communicate with the mobile nodes through wireless links, and the edge cloud nodes communicate with each other through wireless links. The edge cloud nodes are composed of edge cloud facilities and edge communication equipment, the central cloud nodes are composed of central cloud facilities and central communication equipment, and the mobile nodes are composed of application terminals and mobile communication equipment. Application terminals need cloud computing resources during work, which are provided by edge cloud facilities and central cloud facilities. The central cloud facility perceives the network status of the entire network and the status of all cloud computing resources in real time. The central cloud facility uniformly dispatches network cloud resources and provides scheduling services for nearby resources for mobile nodes by default. After the mobile node applies for intelligent computing resource scheduling services, it will intelligently run the four computing resource scheduling modes of multi-machine redundancy, window switching, upgraded processing, and continued service to schedule network cloud resources according to the system capability opening strategy and mobile node identity authority within the authority range.

2. A computing resource scheduling method based on network-cloud collaboration according to claim 1, characterized in that: The process of central cloud facility computing to generate multi-machine redundant computing resource scheduling strategy is as follows: S101, the application terminal of the mobile node sends an information request to the central cloud facility, requesting the central cloud facility to provide it with a multi-machine redundant computing resource scheduling service; S102, the central cloud facility determines whether to provide the service according to the system capability opening policy. If not, a service denial message is sent and the process ends; If provided, a confirmation service message will be sent and the next step will be entered; S103. The central cloud facility perceives the status of the entire network and the status of all cloud computing resources in real time, allocates two sets of cloud computing resources, primary and backup, to the application terminals, and dynamically keeps the primary computing resources mainly allocated to the edge cloud node closest to the mobile node, and the backup computing resources mainly allocated to the edge cloud node with the best signal evaluation with the mobile node at the next moment.

3. The computing resource scheduling method based on network-cloud collaboration according to claim 1, characterized in that: The process of switching computing resource scheduling strategies in the central cloud facility computing generation window is as follows: S201, the application terminal of the mobile node sends an information request to the central cloud facility, requesting the central cloud facility to provide it with a window switching computing resource scheduling service; S202, the central cloud facility determines whether to provide the service according to the system capability opening policy. If not, a service denial message is sent and the process ends; If provided, a confirmation service message will be sent and the next step will be entered; S203, the central cloud facility perceives the network situation of the entire network and the situation of the entire cloud computing resources in real time, allocates cloud computing resources to application terminals, and allocates computing resources to edge cloud nodes with the best signal evaluation with mobile nodes; S204: When the edge cloud node with the best signal evaluation with the mobile node changes, dynamically migrate the cloud computing resources to the new edge cloud node based on the migration change strategy.

4. The computing resource scheduling method based on network-cloud collaboration according to claim 1, characterized in that: The process of computing resource scheduling strategy for central cloud facility computing generation and upgrading is as follows: S301, the application terminal of the mobile node sends an information request to the central cloud facility, requesting the central cloud facility to provide it with upgraded processing computing resource scheduling services; S302, the central cloud facility determines whether to provide the service according to the system capability opening policy. If not, a service denial message is sent and the process ends; If provided, a confirmation service message will be sent and the next step will be entered; S303: The central cloud facility dispatches and allocates the cloud computing resources serving the mobile node to the central cloud node, or dispatches and allocates them to the edge node between the central cloud node and the current service edge node.

5. The computing resource scheduling method based on network-cloud collaboration according to claim 1, characterized in that: The process of central cloud facility computing to generate continuous service computing resource scheduling strategy is as follows: S401, the application terminal of the mobile node sends an information request to the central cloud facility, requesting the central cloud facility to provide it with a continuous service computing resource scheduling service; S402, the central cloud facility determines whether to provide the service according to the system capability opening policy. If not, a service denial message is sent and the process ends; If provided, a confirmation service message will be sent and the next step will be entered; S403, the central cloud facility perceives the network status of the entire network and the status of all cloud computing resources in real time, and allocates the computing resources of the nearest edge cloud node to the mobile node for the first time. When the edge cloud node that the mobile node is connected to for nearby communication changes, no computing resource scheduling is performed, but only the spare computing capacity of the edge cloud node that the mobile node is connected to is sent to the mobile node. The computing resource capacity includes idle computing CPU, idle memory, idle storage resources, network delay, and network bandwidth information.

Images