A resource allocation method for multi-access edge computing

Abstract

The present invention proposes a resource allocation method for multi-access edge computing, which belongs to the field of artificial intelligence technology. An edge controller acts as an agent for the edge node to establish a long connection with multiple terminal devices, so that computing tasks can be transmitted simultaneously on the same spectrum resources, and edge computing nodes are used to allocate computing tasks, thereby improving spectrum utilization and computing resource utilization. The edge controller monitors resource metadata corresponding to the edge node in multiple terminal devices through the long connection; when the resource metadata changes, the edge controller notifies the edge node of the change data of the resource metadata, thereby enabling the edge node to perform corresponding calculation processing according to the variable data, and jointly optimizing the resource allocation plan.

Description

Technical Field

[0001] The present invention belongs to the field of artificial intelligence technology, and specifically relates to a resource allocation method for multi-access edge computing. Background Art

[0002] Edge computing technology is a new technology that is currently emerging. By placing computing, storage, bandwidth, and application resources on the edge of the network, it can reduce transmission delays and bandwidth consumption, and better support applications in areas such as industrial control, autonomous driving, distribution network differential protection, and cloud gaming.

[0003] Although the application of edge computing technology can provide ultra-low latency application services, it increases the complexity of resource allocation and scheduling for computing, network, etc. For example, when deploying multiple mobile edge computing applications on an edge computing server, it is necessary to comprehensively consider the allocation of computing and network resources to provide the best service experience for multiple mobile edge computing applications.

[0004] At present, the resource allocation methods for multi-access edge computing nodes usually use optimization theory to determine the allocation of network resources and computing resources. In addition, machine learning, game theory, and contract theory are also applied. However, the implementation is very complicated.

[0005] The required computing resources and transmission delay are calculated based on the existing computing and network resources, without considering the replacement of computing and network resources. For example, when network resources are scarce, the computing delay can be reduced by allocating more computing resources, and the overall processing delay is less than the total service delay requirement. Summary of the invention

[0006] The present invention aims at the problems in the prior art. The technical solution adopted in the present application is: a resource allocation method for multi-access edge computing, comprising: obtaining operating parameters of a terminal device to be allocated resources; constructing a multi-access edge computing network model based on the operating parameters; The edge controller acts as an agent for the edge node to establish long connections with multiple terminal devices and use edge computing nodes to distribute computing tasks; The classification tree generator is used to classify the terminal devices to obtain the centroid coordinates of each type. The deployment plan generator uses the particle swarm optimization algorithm to solve the centroid coordinates, builds the terminal device task deployment set, and inputs it into the deployment plan generator for allocation; Utilize statistical data to estimate the demand for terminal devices in the next cycle, and pre-allocate the types and quantities of services to edge servers based on the estimated data.

[0007] Furthermore, the multi-access edge computing network model has minimum resource requirements, and the terminal devices are set to allocate resources in a corresponding number of intervals in order to operate normally. This can effectively prevent terminal devices with larger services from being allocated to servers with smaller computing resources, thereby shortening the application latency.

[0008] Furthermore, a classification tree generator is used to classify the terminal devices to obtain the centroid coordinates of each category. The deployment plan generator constructs a terminal device task deployment set based on the centroid coordinates through a particle swarm optimization algorithm. The task deployment set is input into the deployment plan generator, and the optimal cloud computing service system is selected according to the set deployment criteria. The terminal devices are then allocated to the corresponding optimal cloud computing service system to achieve reasonable allocation of resources.

[0009] Furthermore, the edge controller acts as an agent for the edge node to establish a long connection with multiple terminal devices, transmits computing tasks simultaneously on the same spectrum resources, and uses the edge computing nodes to distribute computing tasks. The edge controller monitors the resource metadata corresponding to the edge node in multiple terminal devices through the long connection; when the resource metadata changes, the edge controller notifies the edge node of the change data of the resource metadata.

[0010] Furthermore, the edge computing controller adopts the ARMCortex-A7 core board, and is connected to the RS485 interface, Ethernet interface and network interface. It combines a high-performance edge computing processor with an embedded file system to achieve rapid scheduling and concurrent processing of multi-business application systems.

[0011] Compared with the prior art, the advantages and positive effects of the present invention are: The present invention provides a resource allocation method for multi-access edge computing, comprising: an edge controller acts as an agent for the edge node to establish a long connection with multiple terminal devices, computing tasks can be transmitted simultaneously on the same spectrum resources, and edge computing nodes are used to allocate computing tasks to improve spectrum utilization and computing resource utilization. The edge controller monitors resource metadata corresponding to the edge node in multiple terminal devices through the long connection; when the resource metadata changes, the edge controller notifies the edge node of the change data of the resource metadata, so that the edge node performs corresponding computing processing according to the variable data, and jointly optimizes the resource allocation plan. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0013] Figure 1 A schematic diagram of a process flow of a resource allocation method for multi-access edge computing of the present invention; Figure 2 It is a schematic diagram of the structure of the edge controller of the present invention; DETAILED DESCRIPTION In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described below in conjunction with the accompanying drawings and embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0014] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways than those described herein. Therefore, the present invention is not limited to the specific embodiments of the following disclosure.

[0015] Embodiment 1, as Figure 1-Figure 2 As shown, the present application provides a resource allocation method for multi-access edge computing, including: obtaining operating parameters of a terminal device to be allocated resources; constructing a multi-access edge computing network model based on the operating parameters; The edge controller acts as an agent for the edge node to establish long connections with multiple terminal devices and use edge computing nodes to distribute computing tasks; The edge controller includes a classification tree generator and a solution generator. The classification tree generator is used to classify the terminal devices to obtain the centroid coordinates of each type. The deployment solution generator uses the particle swarm optimization algorithm to solve the centroid coordinates and construct a terminal device task deployment set. Input the task deployment set into the deployment plan generator, select the optimal cloud computing service system according to the set deployment criteria, and allocate the terminal device to the corresponding optimal cloud computing service system to achieve reasonable allocation of resources and improve resource utilization; Statistical data is used to estimate the demand for terminal devices in the next cycle, and the types and quantities of services are pre-allocated to edge servers based on the estimated data, allowing for more efficient resource allocation, improved resource utilization, and reduced application latency.

[0016] The minimum resource requirement sets the resource allocation of the corresponding number interval to the terminal device in order to operate normally. This can effectively prevent the terminal devices with larger services from being allocated to servers with smaller computing resources, thereby shortening the application latency.

[0017] Specifically, the edge controller acts as an agent for the edge node to establish a long connection with multiple terminal devices, so that computing tasks can be transmitted simultaneously on the same spectrum resources, and computing tasks are distributed using edge computing nodes to improve spectrum utilization and computing resource utilization. The edge controller monitors the resource metadata corresponding to the edge node in multiple terminal devices through the long connection; when the resource metadata changes, the edge controller notifies the edge node of the change data of the resource metadata, so that the edge node can perform corresponding calculation processing according to the variable data, and jointly optimize the resource allocation plan.

[0018] By adopting edge computing to serially transmit computing tasks, the communication of computing nodes can be controlled. The speed of transmission tasks can be analyzed by node communication calculation, and the transmission method can be adjusted accordingly. The node transmission communication time is obtained by using the node calculation formula. The transmission communication time of threads in each computing node is processed to obtain the optimized transmission communication time:

[0019] in: is the optimized transmission communication time between the nth node and the hth thread of the mth CPU, is the inherent time of the communication protocol, is the transmission inherent delay of the mth CPU of the nth computing node, is the amount of data transmitted by the hth thread of the mth CPU of the nth node, The total bandwidth of the compute node is the total thread bandwidth of each CPU in the compute node.

[0020] The edge node performs corresponding calculations based on the variable data. The specific steps include: converting the variable data into a unified UML format, modeling the unified data, defining model attributes, and establishing a knowledge graph to build internal data associations. It uses name matching and content sampling matching combined with NLP algorithms for automatic benchmarking, thereby classifying and grading the variable data, and allocating resources based on the graded data.

[0021] The edge computing controller uses the ARMCortex-A7 core board and is connected to the RS485 interface, Ethernet interface and network interface. It combines a high-performance edge computing processor with an embedded file system to achieve rapid scheduling and concurrent processing of multi-business application systems, and uses multi-interface long-distance collection. Figure 2 As shown, a corresponding computing program can be configured internally to execute corresponding instruction operations.

[0022] The above description is only a preferred embodiment of the present invention and does not limit the present invention in other forms. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A resource allocation method for multi-access edge computing, characterized in that: include: Obtaining operating parameters of the terminal device to be allocated resources; Build a multi-access edge computing network model based on operating parameters; The edge controller acts as an agent for the edge node to establish long connections with multiple terminal devices and use edge computing nodes to distribute computing tasks; The classification tree generator is used to classify the terminal devices to obtain the centroid coordinates of each type. The deployment plan generator uses the particle swarm optimization algorithm to solve the centroid coordinates, builds the terminal device task deployment set, and inputs it into the deployment plan generator for allocation; Utilize statistical data to estimate the demand for terminal devices in the next cycle, and pre-allocate the types and quantities of services to edge servers based on the estimated data.

2. A resource allocation method for multi-access edge computing according to claim 1, characterized in that: The multi-access edge computing network model has minimum resource requirements, and the terminal device is set to allocate resources in a corresponding number of intervals in order to operate normally. It can effectively prevent terminal devices with larger services from being allocated to servers with smaller computing resources, thereby shortening the application delay.

3. A resource allocation method for multi-access edge computing according to claim 1, characterized in that: The classification tree generator is used to classify the terminal devices to obtain the centroid coordinates of each category. The deployment plan generator constructs a terminal device task deployment set based on the centroid coordinates through a particle swarm optimization algorithm. The task deployment set is input into the deployment plan generator, and the optimal cloud computing service system is selected according to the set deployment criteria. The terminal devices are allocated to the corresponding optimal cloud computing service system to achieve reasonable allocation of resources.

4. The resource allocation method for multi-access edge computing according to claim 1, characterized in that: The edge controller acts as an agent for the edge node to establish a long connection with multiple terminal devices, transmits computing tasks simultaneously on the same spectrum resources, and uses the edge computing nodes to distribute computing tasks. The edge controller monitors the resource metadata corresponding to the edge node in multiple terminal devices through the long connection; when the resource metadata changes, the edge controller notifies the edge node of the change data of the resource metadata.

5. The resource allocation method for multi-access edge computing according to claim 1, characterized in that: The edge computing controller adopts the ARMCortex-A7 core board and is connected to the RS485 interface, Ethernet interface and network interface. It combines a high-performance edge computing processor with an embedded file system to achieve rapid scheduling and concurrent processing of multi-business application systems.

Images