A service scheduling and deployment method for resources of an internet of things edge computing node

Abstract

The application relates to a service scheduling and deployment method for Internet of Things edge computing node resources, which is realized based on an Internet of Things edge computing system. The system comprises a cloud data center, a convergence layer core network and a plurality of access layer switch networks connected in sequence. Each access layer switch network is in communication connection with a plurality of edge computing nodes, and each edge computing node is in communication connection with a plurality of data acquisition terminals. The application comprehensively considers the current running state of the edge computing system, network delay problems and running results after deployment, and performs service scheduling and deployment of Internet of Things edge computing node resources on each real-time data stream computing application. The application is suitable for the decision environment of the real-time computing task load and bandwidth dynamic change of the Internet of Things edge computing node, realizes global optimization from the system level, takes into account the single node load capacity, and realizes the optimal scheme and application planning and deployment decision suggestion of the long-period stable operation and real-time task flexible scheduling strategy.

Description

Technical Field

[0001] This invention relates to the field of distribution network Internet of Things (IoT) technology, and specifically to a service scheduling and deployment method for IoT edge computing node resources. Background Technology

[0002] In recent years, with the rise of the Internet of Things (IoT), researchers have begun to use intelligent edge computing frameworks to offload real-time computing tasks. The basic idea is to deploy computing devices at the network edge, close to the data source, forming a "data acquisition terminal device - edge computing node - cloud data center" architecture. Based on predictable or monitored task load information of the computing nodes, applications are dynamically deployed on edge computing nodes or in the cloud data center. This fully utilizes the stronger computing capabilities of edge computing nodes compared to data acquisition terminal devices and effectively avoids the high latency caused by data transmission in the core network, thereby significantly improving the service quality of real-time services. In IoT intelligent edge computing, limited by the bandwidth, computing power, and other hardware resources of edge computing nodes, a crucial issue is how to rationally utilize the predictable or monitored real-time task technology and network load to schedule resource allocation tasks for deployable edge computing nodes, thereby reducing application latency and improving the efficient utilization of edge computing node resources.

[0003] The long distances between data acquisition terminals, edge computing nodes, and cloud data center servers result in significant communication costs, data transmission latency, and remote computing energy consumption, negatively impacting real-time applications. Therefore, migrating some of the computing and storage capabilities of remote cloud devices to edge IoT gateways and deploying a three-tier architecture for IoT edge computing gateways is becoming the preferred technical solution for most IoT enterprises. However, at the same time, IoT edge networks exhibit randomness and dynamism; some real-time business applications are highly sensitive to latency and energy consumption, and the prolonged operation of these applications can also lead to high energy consumption.

[0004] In IoT edge computing, edge computing nodes need to decide when and where to deploy applications to achieve the optimal real-time service deployment scheme for the nearest computing edge. Current methods employ heuristic algorithms for global optimization, considering the link conditions of the fronthaul and backhaul networks. They optimize task deployment decisions while ensuring latency and determine whether to migrate buffered tasks to other nearby edge computing gateways within each time slot, achieving dynamic load balancing at the edge. However, current methods only consider fixed-pattern edge computing gateway service deployment schemes, neglecting the resource availability efficiency and optimization paths of adjacent edge computing gateways. They all rely on heuristic learning techniques for resource allocation and management, depending on past workload states while ignoring the current running state. Therefore, they are not suitable for optimization and improvement schemes for real-time service program deployment in dynamic environments. Summary of the Invention

[0005] The purpose of this invention is to propose a service scheduling and deployment method for IoT edge computing node resources, so as to perform service scheduling and deployment of IoT edge computing node resources.

[0006] This invention proposes a service scheduling and deployment method for IoT edge computing node resources, which is implemented based on an IoT edge computing system. The system includes a cloud data center, an aggregation layer core network, and multiple access layer switch networks. The cloud data center is communicatively connected to the aggregation layer core network, the multiple access layer switch networks are communicatively connected to the aggregation layer core network, each access layer switch network is communicatively connected to multiple edge computing nodes, and each edge computing node is communicatively connected to multiple data acquisition terminals.

[0007] The method includes the following steps:

[0008] The cloud data center obtains the real-time data streaming computing application that needs to be deployed, analyzes the computing resource requirements involved in the real-time data streaming computing application that needs to be deployed, and sends it to all edge computing nodes.

[0009] All edge computing nodes perform a self-assessment of computing resources based on the computing resource requirements involved in the real-time data streaming computing application to be deployed. If the self-assessment of computing resources passes, the edge computing node applies to the cloud data center to be added to the service list of optional deployment nodes and sends the self-assessment result of computing resources to the cloud data center. If the self-assessment of computing resources passes, the edge computing node does not perform any action.

[0010] The cloud data center generates a service list of optional deployment nodes based on the application of the edge computing nodes, and obtains the network latency information of multiple edge computing nodes in the service list of optional deployment nodes;

[0011] The cloud data center performs calculations based on the network latency information of multiple edge computing nodes in the service list of the optional deployment node and the self-assessment results of computing resources. It selects the optimal edge computing node in the service list as the deployment node and runs the real-time data streaming computing task of the real-time data streaming computing application that needs to be deployed. It evaluates whether the deployment result meets the expected business indicators. If it does, the deployment ends successfully. If it does not, it selects the next edge computing node in the service list of the optional deployment node for iterative deployment until the deployment ends successfully or fails.

[0012] Preferably, the method further includes the following steps:

[0013] The cloud data center obtains a set of real-time data streaming computing applications, which includes multiple real-time data streaming computing applications, and extracts one from the set of real-time data streaming computing applications without replacement as the real-time data streaming computing application that needs to be deployed at the moment.

[0014] When the deployment of the real-time data streaming computing application to be deployed is completed successfully or fails, the cloud data center determines whether the number of applications in the current real-time data streaming computing application set is 0. If not, it selects one application from the real-time data streaming computing application set without replacement as the real-time data streaming computing application to be deployed. If yes, it terminates the deployment of all applications in the real-time data streaming computing application set.

[0015] Preferably, the computing resource requirements include the data acquisition terminals and their data types involved in real-time data streaming computing applications.

[0016] Preferably, all edge computing nodes perform a self-assessment of computing resources based on the computing resource requirements involved in the currently deployed real-time data streaming computing application, including:

[0017] All edge computing nodes acquire real-time CPU resources, memory resources, and hardware storage resources, and acquire terminal information and data types of multiple data acquisition terminals connected to them.

[0018] All edge computing nodes analyze whether they can meet the computing resource requirements based on the real-time CPU resources, memory resources, and hardware storage resources, as well as the terminal information and the data types being processed. If they meet the requirements, the computing resource self-assessment passes and a computing resource self-assessment result is generated; otherwise, the computing resource self-assessment fails.

[0019] Preferably, the self-assessment results of computing resources include the real-time CPU resources, memory resources, and hardware storage resources of the edge computing node, as well as the terminal information and data types of the multiple data acquisition terminals connected to it.

[0020] Preferably, obtaining the network latency information of multiple edge computing nodes in the service list of the optional deployment node includes:

[0021] The cloud data center sends a network latency query to each of the multiple edge computing nodes in the service list of the optional deployment node; the network latency query includes the sending time information and the network latency query content.

[0022] After receiving network latency query information, multiple edge computing nodes in the service list of the optional deployment nodes extract the sending time information based on the network latency query content, obtain the receiving time information of the received network latency query information, calculate their network latency information based on the sending time information and the receiving time information, and then feed their network latency information back to the cloud data center.

[0023] The embodiments of the present invention have at least the following beneficial effects:

[0024] Taking into account the current operating status of the edge computing system, network latency issues, and post-deployment results, this paper proposes a business scheduling and deployment approach for IoT edge computing node resources for various real-time data streaming computing applications. This approach is suitable for decision-making environments where the real-time computing task load and bandwidth of IoT edge computing nodes change dynamically. From a system-level perspective, it provides global optimization, takes into account the load capacity of individual nodes, and achieves the optimal solution and application planning and deployment decision recommendations for long-term stable operation and flexible real-time task scheduling strategies.

[0025] Other features and advantages of embodiments of the present invention will be set forth in the following description. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 This is a flowchart of a service scheduling and deployment method for IoT edge computing node resources in an embodiment of the present invention.

[0028] Figure 2 This is a structural diagram of an Internet of Things (IoT) edge computing system according to an embodiment of the present invention.

[0029] Figure 3 This is a flowchart illustrating a service scheduling and deployment method for IoT edge computing node resources in an embodiment of the present invention. Detailed Implementation

[0030] The various exemplary embodiments, features, and aspects of this disclosure will now be described in detail with reference to the accompanying drawings. Furthermore, numerous specific details are set forth in the following detailed embodiments to better illustrate the invention. Those skilled in the art will understand that the invention can be practiced without certain specific details. In some instances, means well known to those skilled in the art have not been described in detail in order to highlight the spirit of the invention.

[0031] See Figure 1 The embodiments of the present invention propose a service scheduling and deployment method for IoT edge computing node resources, based on, for example, Figure 2 The diagram illustrates an IoT edge computing system implementation. The system includes a cloud data center, an aggregation layer core network, and multiple access layer switch networks. The cloud data center is communicatively connected to the aggregation layer core network, and the multiple access layer switch networks are communicatively connected to the aggregation layer core network. Each access layer switch network is communicatively connected to multiple edge computing nodes, and each edge computing node is communicatively connected to multiple data acquisition terminals.

[0032] See Figure 1 The method of this invention includes the following steps:

[0033] Step S101: The cloud data center obtains the real-time data streaming computing application that needs to be deployed, analyzes the computing resource requirements involved in the real-time data streaming computing application that needs to be deployed, and sends it to all edge computing nodes.

[0034] Step S102: All edge computing nodes perform a self-assessment of computing resources based on the computing resource requirements involved in the real-time data streaming computing application to be deployed. If the self-assessment of computing resources passes, the edge computing node applies to the cloud data center to be added to the service list of optional deployment nodes and sends the self-assessment result of computing resources to the cloud data center. If the self-assessment of computing resources passes, the edge computing node does not perform any action.

[0035] Step S103: The cloud data center generates a service list of optional deployment nodes based on the application of the edge computing nodes, and obtains the network latency information of multiple edge computing nodes in the service list of optional deployment nodes;

[0036] Step S104: The cloud data center performs calculations based on the network latency information of multiple edge computing nodes in the service list of the optional deployment node and the self-assessment results of computing resources. It selects the optimal edge computing node in the service list as the deployment node and runs the real-time data streaming computing task of the real-time data streaming computing application that needs to be deployed. It evaluates whether the deployment result meets the expected business indicators. If it does, the deployment ends successfully. If it does not, it selects the next edge computing node in the service list of the optional deployment node for iterative deployment until the deployment ends successfully or fails.

[0037] Specifically, in the embodiments of the present invention, see [reference]. Figure 3 The method further includes the following steps:

[0038] Step S100: The cloud data center obtains a set of real-time data streaming computing applications, which includes multiple real-time data streaming computing applications, and extracts one from the set of real-time data streaming computing applications without replacement as the real-time data streaming computing application that needs to be deployed at the moment.

[0039] Step S105: When the deployment of the real-time data streaming computing application to be deployed is completed successfully or fails, the cloud data center determines whether the number of applications in the current real-time data streaming computing application set is 0. If not, one application is extracted from the real-time data streaming computing application set without replacement as the real-time data streaming computing application to be deployed. If yes, the deployment of all applications in the real-time data streaming computing application set is terminated.

[0040] Specifically, in this embodiment of the invention, the computing resource requirements include the data acquisition terminals and their data types involved in real-time data streaming computing applications.

[0041] Specifically, in this embodiment of the invention, all edge computing nodes perform a self-assessment of computing resources based on the computing resource requirements involved in the currently deployed real-time data streaming computing application, including:

[0042] All edge computing nodes acquire real-time CPU resources, memory resources, and hardware storage resources, and acquire terminal information and data types of multiple data acquisition terminals connected to them.

[0043] All edge computing nodes analyze whether their own real-time CPU resources, memory resources, and hardware storage resources, as well as the terminal information and the data types being processed, can simultaneously meet the computing resource requirements based on the real-time CPU resources, memory resources, and hardware storage resources, as well as the terminal information and the data types being processed. If they meet the requirements, the computing resource self-assessment passes and a computing resource self-assessment result is generated; otherwise, the computing resource self-assessment fails.

[0044] Specifically, in this embodiment of the invention, the self-evaluation result of computing resources includes not only the result of whether the self-evaluation of computing resources passes, but also the real-time CPU resources, memory resources and hardware storage resources of the edge computing node, as well as the terminal information and data types of the multiple data acquisition terminals connected to it.

[0045] Specifically, in this embodiment of the invention, obtaining the network latency information of multiple edge computing nodes in the service list of the optional deployment node includes:

[0046] The cloud data center sends a network latency query to each of the multiple edge computing nodes in the service list of the optional deployment node; the network latency query includes the sending time information and the network latency query content.

[0047] After receiving network latency query information, multiple edge computing nodes in the service list of the optional deployment nodes extract the sending time information based on the network latency query content, obtain the receiving time information of the received network latency query information, calculate their network latency information based on the sending time information and the receiving time information, and then feed their network latency information back to the cloud data center.

[0048] In this embodiment of the invention, the original fixed-mode manual pre-assessment planning and deployment is changed to real-time dynamic planning by the computer system. This enables the output of a system operation assessment report and a business deployment strategy suggestion report every hour, allowing business operators to select business deployment nodes in a timely and efficient manner and providing a basis for real-time decision-making for dynamically adjusting business deployment plans. This effectively supports the lean operation and maintenance level of the system and improves the overall intelligent service level of the system.

[0049] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or technical improvements to the embodiments in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for service scheduling deployment of resources of an Internet of Things edge computing node, characterized in that, The application is based on an Internet of Things edge computing system, which comprises a cloud data center, a convergence layer core network, a plurality of access layer switch networks, the cloud data center being in communication connection with the convergence layer core network, the plurality of access layer switch networks being in communication connection with the convergence layer core network, each access layer switch network being in communication connection with a plurality of edge computing nodes, and each edge computing node being in communication connection with a plurality of data collection terminals. The method comprises the following steps: The cloud data center acquires a real-time data stream computing application to be currently deployed, analyzes computing resource requirements involved in the real-time data stream computing application to be currently deployed, and sends the computing resource requirements to all edge computing nodes; the computing resource requirements comprise data collection terminals and data types involved in the real-time data stream computing application; All edge computing nodes perform computing resource self-evaluation according to the computing resource requirements involved in the real-time data stream computing application to be currently deployed, if the computing resource self-evaluation is passed, the edge computing node applies to join a service list of optional deployment nodes to the cloud data center, and sends a computing resource self-evaluation result to the cloud data center, if the computing resource self-evaluation is passed, the edge computing node does not perform any action; The cloud data center generates a service list of optional deployment nodes according to the application of the edge computing node, and acquires network delay information of a plurality of edge computing nodes in the service list of the optional deployment nodes; The cloud data center performs operation according to the network delay information of the plurality of edge computing nodes in the service list of the optional deployment nodes and the computing resource self-evaluation result, selects an optimal edge computing node in the service list as a deployment node, and runs a real-time data stream computing task of the real-time data stream computing application to be currently deployed, evaluates whether a deployment running result meets an expected business index, if yes, the deployment is successfully ended, if not, the next edge computing node in the service list of the optional deployment nodes is selected for iterative deployment until the deployment is successfully ended or the deployment is failed to end. 2.The method of claim 1, wherein, The method further comprises the following steps: The cloud data center acquires a real-time data stream computing application set, the real-time data stream computing application set comprising a plurality of real-time data stream computing applications, and extracts one from the real-time data stream computing application set as a real-time data stream computing application to be currently deployed; When the real-time data stream computing application to be currently deployed is successfully ended or failed to end, the cloud data center judges whether the number of applications in the current real-time data stream computing application set is 0, if not, extracts one from the real-time data stream computing application set as a real-time data stream computing application to be currently deployed; if yes, ends the deployment of all applications in the real-time data stream computing application set. 3.The method of Claim 2, wherein, The computing resource self-evaluation of all edge computing nodes according to the computing resource requirements involved in the real-time data stream computing application to be currently deployed comprises: The all edge computing nodes acquire real-time CPU resources, memory resources and hardware storage resources, and acquire terminal information of a plurality of data acquisition terminals connected thereto and data types processed thereby; The all edge computing nodes analyze whether the computing resource demand can be met according to the real-time CPU resources, memory resources and hardware storage resources, the terminal information and the data types processed thereby, and if yes, the computing resource self-evaluation is passed, and a computing resource self-evaluation result is generated, and if not, the computing resource self-evaluation is not passed.

4. The method of claim 3, wherein, The computing resource self-evaluation result includes real-time CPU resources, memory resources and hardware storage resources of the edge computing nodes, and terminal information of a plurality of data acquisition terminals connected thereto and data types processed thereby.

5. The method of claim 1, wherein, The network delay information of the plurality of edge computing nodes in the service list of the optional deployment node includes: The cloud data center respectively sends a network delay query information to the plurality of edge computing nodes in the service list of the optional deployment node; the network delay query information contains sending time information and network delay query content; After receiving the network delay query information, the plurality of edge computing nodes in the service list of the optional deployment node extracts the sending time information according to the network delay query content, acquires receiving time information of the received network delay query information, calculates network delay information according to the sending time information and the receiving time information, and then feeds back the network delay information to the cloud data center.

Images