Edge computing node management method and device, equipment and storage medium

Abstract

The embodiment of the invention provides an edge computing node management method and device, equipment and a storage medium, and the method comprises the steps: collecting first target information of a first edge node in a target region, analyzing the first target information through a preset genetic algorithm, and generating an optimal distribution deployment scheme about the first edge node, based on an optimized distribution deployment scheme, screening a second edge node from the first edge nodes, determining a first resource allocation amount, a to-be-executed first task and a first task scheduling amount in the second edge node by using a preset optimization strategy, and obtaining second target information of the second edge node under the condition of determining that the execution of the first task is completed, and comparing the second target information with the first target information, and adjusting according to a comparison result to obtain a second resource allocation amount and a second task scheduling amount of the second edge node, and the embodiment of the invention improves the computing efficiency and the resource utilization rate of the system.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to an edge computing node management method, device, equipment and storage medium. Background Art

[0002] With the development of technologies such as the Internet of Things, smart cities, and autonomous driving, edge computing has been widely used in many fields. For example, in the Internet of Things environment, a large number of data generated by sensors and devices need to be processed and analyzed in real time. Edge computing reduces data transmission delays and bandwidth consumption by processing near the data source, and improves the response speed and reliability of the system. Applications such as intelligent transportation systems, smart grids, and environmental monitoring require real-time processing of a large amount of sensor data. Edge computing can perform calculations close to the data source, provide fast decision support, optimize the use of urban resources, and improve the efficiency of urban operations. Self-driving cars need to process a large amount of data from cameras, lidars, and other sensors in real time. Edge computing can manage edge computing nodes inside or near the vehicle, provide real-time driving decisions, reduce dependence on the cloud, and improve driving safety and reliability. In the intelligent manufacturing environment, the operating data generated by machines and equipment needs to be monitored and analyzed in real time. Edge computing can manage edge computing nodes inside the factory, provide timely equipment maintenance and production optimization, and improve production efficiency and quality.

[0003] In order to make up for the shortcomings of insufficient computing power, limited energy supply and small coverage of a single edge computing node, multiple edge computing nodes build an edge intelligent network through multi-machine networking, resource sharing, collaborative services, etc., which can provide more computing resources and a larger coverage. In order to improve the service quality of edge intelligent networks, current technologies have set up a variety of edge computing node deployment schemes, such as: through receiving request tasks, notifying service function (SF) information, matching game between request tasks and SF, and unloading request tasks, etc., to achieve a stable match between request tasks, service functions SF and edge computing nodes; or based on the distribution of user data hotspots, using the user's received signal strength to determine the division area where it is located, considering factors such as communication scenarios and connection costs when deploying edge computing nodes, and deploying nodes through simulated annealing, to achieve reasonable edge computing node management.

[0004] However, among the above solutions, the former does not involve the overall distribution optimization of edge computing nodes, and has deficiencies in node resource utilization and overall system efficiency. The latter focuses on communication scenarios and connection costs, and does not consider other influencing factors. In addition, both solutions do not consider optimizing the resource allocation on edge nodes after deployment, and there is still room for improvement. Summary of the invention

[0005] The purpose of the embodiments of the present invention is to provide an edge computing node management method, device, equipment and storage medium to solve the problem that the current edge computing node deployment solution does not involve the overall distribution optimization of edge computing nodes, has deficiencies in node resource utilization and overall system efficiency, considers relatively single influencing factors, and does not consider the optimization of resource allocation on edge nodes, and there is still room for improvement. The specific technical solution is as follows:

[0006] In a first aspect of the implementation of the present invention, a method for managing an edge computing node is first provided, the method comprising:

[0007] Collecting first target information of a plurality of first edge nodes in a target area;

[0008] Analyzing the first target information by using a preset genetic algorithm to generate an optimized distribution deployment plan for a plurality of the first edge nodes;

[0009] Based on the optimized distributed deployment solution, selecting a second edge node from a plurality of the first edge nodes;

[0010] Determine the first resource allocation amount, the first task to be executed and the first task scheduling amount in the second edge node by using a preset optimization strategy;

[0011] When it is determined that the first tasks are all completed, obtaining second target information of the second edge node, wherein the target information includes at least one of target task request information, target node status information, target network status information and target environment information;

[0012] The second target information is compared with the first target information, and adjustments are made according to the comparison result to obtain a second resource allocation amount and a second task scheduling amount of the second edge node.

[0013] Optionally, the analyzing the first target information by using a preset genetic algorithm to generate an optimized distribution deployment scheme for a plurality of the first edge nodes includes:

[0014] Obtaining target parameters set for the initialization population in the genetic algorithm, the target parameters including population size, crossover probability and mutation probability, the population size refers to the number of individuals specified to be included in the initialization population;

[0015] Analyzing the first target information by using the target parameter to determine an initialization population of the first edge node, wherein the initialization population includes a plurality of first distribution deployment schemes for the first edge node, and each first distribution deployment scheme corresponds to an individual in the initialization population;

[0016] Obtaining a first fitness value of each first distribution deployment scheme through a fitness function preset in the genetic algorithm;

[0017] Based on the first fitness value, determine the probability of selection of each first distribution deployment scheme using a target selection probability formula;

[0018] Screening out an excellent second distribution deployment scheme from the first distribution deployment schemes according to the selection probability;

[0019] Performing a crossover operation and a mutation operation on the second distribution deployment scheme by using a crossover device and a mutator in the genetic algorithm to obtain a third distribution deployment scheme;

[0020] Obtaining a second fitness value of each third distribution deployment scheme through the preset fitness function;

[0021] When it is determined that the second fitness value satisfies a preset condition, a third distribution deployment scheme corresponding to the second fitness value that satisfies the preset condition is determined as an optimized distribution deployment scheme for the first edge nodes.

[0022] Optionally, the target task request information includes at least one of task type, computing requirements, and time constraints; the target node status information includes at least one of the central processing unit utilization, memory utilization, storage space, and operating status of each edge computing node; the target network status information includes at least one of the network bandwidth, network delay, and packet loss rate of the network where the edge computing node is located; and the target environment information includes at least one of the temperature, humidity, and power supply status of the environment where each edge computing node is located.

[0023] Optionally, the fitness function formula is:

[0024]

[0025] Among them, α, β, γ, δ are weight parameters, It refers to the sum of n tasks in the first distribution deployment scheme, D i refers to the computational requirements of the ith task, x i It means that the i-th task is assigned to the first edge node of the target. It refers to the CPU usage of the target first edge node. Refers to the memory usage of the target first edge node, MEM total It refers to the total memory usage of the first edge node in the first distributed deployment scheme. Refers to the storage space of the target first edge node, STOR total It refers to the total storage space of the first edge node in the first distributed deployment solution. It refers to the network delay of the network where the target first edge node is located. It refers to the packet loss rate of the network where the target first edge node is located, f(x) refers to the first fitness value, and the target first edge node is the first edge node in the first distributed deployment scheme.

[0026] Optionally, the determining the first resource allocation amount, the first task to be executed, and the first task scheduling amount in the second edge node by using a preset optimization strategy includes:

[0027] Acquire first node state information and first network state information about the second edge node from the first target information;

[0028] The first resource allocation amount, the first task to be executed and the first task scheduling amount in the second edge node are determined by using the preset optimization strategy, the first node state information and the first network state information.

[0029] Optionally, after determining the first resource allocation amount, the first task to be executed, and the first task scheduling amount in the second edge node by using a preset optimization strategy, the method further includes:

[0030] When it is determined that the second edge node starts to execute the first task, obtaining the task execution status of the second edge node through a node status monitor;

[0031] When it is determined that the task execution status is a state where no task is executed, obtaining an unexecuted second task from the first task, and continuing to execute the second task;

[0032] When it is determined that the execution quantity of the first task reaches the first task scheduling quantity and the task execution status is a state where no task is executed, a task execution completion notification is sent.

[0033] Optionally, comparing the second target information with the first target information and adjusting according to the comparison result to obtain a second resource allocation amount and a second task scheduling amount of the second edge node includes:

[0034] Comparing the second target information with the first target information to determine target difference information between the second target information and the first target information;

[0035] Generate a report log using the target difference information, and transmit the report log to a target page;

[0036] In response to the user's adjustment operation on the target page, a second resource allocation amount and a second task scheduling amount in the second edge node are obtained.

[0037] In a second aspect of the implementation of the present invention, there is also provided an edge computing node management device, including:

[0038] A first acquisition module, used to acquire first target information of a plurality of first edge nodes in a target area;

[0039] A first generating module, configured to analyze the first target information by using a preset genetic algorithm to generate an optimized distribution deployment scheme for a plurality of the first edge nodes;

[0040] A first screening module, configured to screen a second edge node from a plurality of first edge nodes based on the optimized distributed deployment solution;

[0041] A first determination module, configured to determine a first resource allocation amount, a first task to be executed, and a first task scheduling amount in the second edge node by using a preset optimization strategy;

[0042] A first acquisition module is used to acquire second target information of the second edge node when it is determined that the first tasks are all completed, wherein the target information includes at least one of target task request information, target node status information, target network status information and target environment information;

[0043] The first adjustment module is used to compare the second target information with the first target information, and make adjustments according to the comparison result to obtain a second resource allocation amount and a second task scheduling amount of the second edge node.

[0044] In a third aspect of the implementation of the present invention, there is also provided a communication device, comprising: a transceiver, a memory, a processor, and a program stored in the memory and executable on the processor;

[0045] The processor is used to read the program in the memory to implement and execute any of the above-mentioned edge computing node management methods.

[0046] In a fourth aspect of the implementation of the present invention, a computer-readable storage medium is also provided, wherein instructions are stored in the computer-readable storage medium, and when the computer-readable storage medium is run on a computer, the computer executes any of the above-mentioned edge computing node management methods.

[0047] The edge computing node management method provided by the embodiment of the present invention collects first target information of several first edge nodes in a target area; analyzes the first target information through a preset genetic algorithm to generate an optimized distribution deployment plan for several first edge nodes; sets a distribution deployment plan for edge nodes by comprehensively considering task requests, node status, network status and environmental information, thereby ensuring the rationality and effectiveness of node deployment; and selects a second edge node from several first edge nodes based on the optimized distribution deployment plan; determines the first resource allocation amount, the first task to be executed and the first task scheduling amount in the second edge node by using a preset optimization strategy; and obtains the second edge node when it is determined that all first tasks have been executed. The second target information is obtained, where the target information includes at least one of target task request information, target node status information, target network status information and target environment information; the second target information is compared with the first target information, and adjusted according to the comparison result to obtain a second resource allocation amount and a second task scheduling amount of the second edge node, and the resource and task allocation are dynamically adjusted and optimized through the comparison result, thereby improving the computing efficiency and resource utilization of the system. In the embodiment of the present invention, the distributed deployment scheme of the edge nodes is set by comprehensively considering the task request, node status, network status and environment information, and the resource allocation and task scheduling of the edge nodes are optimized and adjusted at the same time, thereby improving the computing efficiency and resource utilization of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below.

[0049] Figure 1 A flowchart of a method for managing edge computing nodes provided by an embodiment of the present invention;

[0050] Figure 2 for Figure 1 The flowchart of step 106 in the edge computing node management method provided by the embodiment of the present invention is shown;

[0051] Figure 3 for Figure 1 A schematic diagram of module connections in an edge computing node management method provided by an embodiment of the present invention is shown;

[0052] Figure 4 for Figure 1 A schematic diagram of data flows of different modules in an edge computing node management method provided by an embodiment of the present invention is shown;

[0053] Figure 5 is a structural schematic diagram of a fault handling device provided by an embodiment of the present invention;

[0054] Figure 6 It is a structural diagram of a communication device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0055] To make the purpose, technical scheme and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings. However, it will be appreciated by those skilled in the art that in the embodiments of the present invention, many technical details are proposed in order to enable the reader to better understand the present application. However, even without these technical details and various changes and modifications based on the following embodiments, the technical scheme claimed in the present application can be implemented. The division of the following embodiments is for the convenience of description, and should not constitute any limitation on the specific implementation of the present invention. The various embodiments can be combined and referenced with each other without contradiction.

[0056] Reference Figure 1 , shows a flowchart of the steps of an edge computing node management method provided by an embodiment of the present invention, the method may include:

[0057] Step 101: collect first target information of a plurality of first edge nodes in a target area.

[0058] In the embodiment of the present invention, the target area can be based on the user demand analysis, including the user's geographical distribution, the frequency and type of service requests, the peak traffic period, etc., to select the area with dense users or high service demand, so as to ensure that the edge node can effectively serve the target user group. The area can also be selected based on the quality and coverage of the network infrastructure to ensure that the edge node can provide low-latency and high-bandwidth services. Cost-effectiveness can also be considered to select the area that can provide the best service quality within the budget, etc. The present invention is not specifically limited here.

[0059] In the embodiment of the present invention, the first target information includes at least one of target task request information, target node status information, target network status information and target environment information. Among them, the target task request information can be obtained by a task request receiver set in the data processing center, and the task request receiver receives and parses the target task request information through API, Web service and other interfaces to obtain at least one of task type, computing requirements and time constraints. The target node status information can be obtained by a node status monitor set on each first edge node, and the target node status information includes at least one of the CPU usage rate, memory usage rate, storage space and running status of each edge computing node. The target network status information can be obtained by a network status monitor on a key node where the network connected to the first edge node is located, and the target network status information includes at least one of the network bandwidth, network delay and packet loss rate of the network where the edge computing node is located. The target environment information can be obtained by an environment information collector located at an edge node that is sensitive to environmental conditions, and the target environment information includes at least one of the temperature, humidity and power supply status of the environment where each edge computing node is located.

[0060] Step 102: Analyze the first target information by using a preset genetic algorithm to generate an optimized distribution deployment plan for a plurality of first edge nodes.

[0061] The preset genetic algorithm in the embodiment of the present invention includes operations such as initializing the population, selection, crossover, mutation and fitness evaluation. The population generator is responsible for initializing the population according to the input data, the selector selects excellent individuals according to the set fitness function, the crossover and mutator generate new individuals through crossover and mutation operations, and the fitness evaluator performs fitness evaluation on each individual, and the evaluation results are used to select the next generation of population. Through continuous iteration, the genetic algorithm module finds the optimal edge computing node distribution scheme, that is, the optimized distribution deployment scheme for several first edge nodes, wherein initializing the population involves creating a set of random solutions, which are called "individuals", and the entire set is called a "population", and an individual can represent the deployment scheme of an edge node. This may include the location, quantity, configuration parameters, etc. of the node. When generating the initial population, it should be ensured that the individuals in the population have a certain diversity to cover different areas in the problem space and increase the chance of finding the global optimal solution. The specific steps include:

[0062] Obtain the target parameters set for the initialization population in the genetic algorithm. The target parameters include population size, crossover probability and mutation probability. The population size refers to the number of individuals specified in the initialization population.

[0063] Analyzing the first target information through the target parameter to determine an initialization population of the first edge node, wherein the initialization population includes a plurality of first distribution deployment schemes for the first edge node, and each first distribution deployment scheme corresponds to an individual in the initialization population;

[0064] Obtaining a first fitness value of each first distribution deployment scheme through a fitness function preset in the genetic algorithm;

[0065] Based on the first fitness value, determining the probability of each first distribution deployment scheme being selected using a target selection probability formula;

[0066] Selecting an excellent second distribution deployment scheme from the first distribution deployment scheme by the selection probability;

[0067] Performing crossover and mutation operations on the second distribution deployment plan through a crossover and a mutator in a genetic algorithm to obtain a third distribution deployment plan;

[0068] Obtaining a second fitness value of each third distribution deployment scheme through a preset fitness function;

[0069] When it is determined that the second fitness value satisfies the preset condition, the third distribution deployment scheme corresponding to the second fitness value that satisfies the preset condition is determined as the optimized distribution deployment scheme for the plurality of first edge nodes.

[0070] Through the above steps, the genetic algorithm can gradually optimize the distribution deployment plan of edge nodes and eventually find one or more optimal solutions to meet specific performance and cost goals.

[0071] It should be noted that after obtaining the second fitness value of each third distribution deployment scheme through the preset fitness function, if it is determined that the second fitness value does not meet the preset condition, the process of selection, crossover, mutation and fitness evaluation will continue to be repeated. One process is called an iteration operation. When the second fitness value meets the preset condition, the iteration will be stopped, or when the number of iteration operations is greater than the target threshold, the iteration will also be stopped. The specific steps include:

[0072] The complete process of realizing a selection operation, a crossover operation, a mutation operation and a fitness evaluation operation in a genetic algorithm is defined as an iterative operation;

[0073] When it is determined that the second fitness value does not meet the preset condition, performing an iterative operation, and starting a target counter to record the number of iterative operations;

[0074] When it is determined that the second fitness value does not meet the preset condition and the number of iterative operations is greater than the target threshold, the fourth distributed deployment scheme obtained by the last iterative operation is determined as the optimized distributed deployment scheme for the first edge nodes.

[0075] Furthermore, the fitness function formula is:

[0076]

[0077] Among them, α, β, γ, δ are weight parameters, It refers to the sum of n tasks in the first distribution deployment scheme, D i refers to the computational requirements of the ith task, x i It means that the i-th task is assigned to the first edge node of the target. It refers to the CPU usage of the target first edge node. Refers to the memory usage of the target first edge node, MEM total It refers to the total memory usage of the first edge node in the first distributed deployment scheme. Refers to the storage space of the target first edge node, STOR total It refers to the total storage space of the first edge node in the first distributed deployment solution. It refers to the network delay of the network where the target first edge node is located. It refers to the packet loss rate of the network where the target first edge node is located, f(x) refers to the first fitness value, and the target first edge node is the first edge node in the first distributed deployment scheme.

[0078] For example, the above genetic algorithm process is represented by coding. First, the task request Ti in the edge node is determined, and the node Nj assigned to it is determined. The specific task in the task request Ti is xi, so when initializing the population, each individual X is represented by a vector of length n, and each element xi in the vector represents the task T i Assigned nodes N j . Coding format: X = [x 1 ,x 2 ,…,x n ], where xi∈{N 1 ,N 2 ,…,N m}, randomly generate P individuals to form an initial population P = {X 1 ,X 2 ,…,X P For example, for a task set T = {T 1 ,T 2 ,T 3} and node set N = {N 1 ,N2 ,N 3}, the mathematical expression of an individual is X 1 =[N 1 ,N 2 ,N 3 ], the target selection probability formula is the probability formula set in the roulette selection method:

[0079]

[0080] Among them, f(X i ) is the fitness value of individual i, which is used to measure the performance of the individual. It is the sum of the reciprocals of all individual fitnesses and is used for normalization to ensure that the sum of the selection probabilities of all individuals is 1. Single-point crossover or multi-point crossover is used during the crossover operation to select two parent individuals X i and X j , randomly select a crossover point k and generate a new individual: During the mutation operation, the individuals in the population are mutated and individuals X are randomly selected. i A task T i , assign it to another random node N i . The individual after mutation: Then, the fitness of the newly generated individuals is evaluated and their fitness value f(X) is calculated. Individuals with higher fitness values ​​are selected to enter the next generation, and the maximum number of iterations max_itermax or the fitness value change is less than the threshold ∈ is set as the termination condition. When the termination condition is reached, the optimal individual is output as the final optimization solution, and the genetic algorithm module outputs the optimal individual X opt , represents the optimal node distribution solution. Optimal solution: X opt =[N 1 ,N 2 ,N 1 ,…,N 3 ].

[0082] Step 103: based on the optimized distribution deployment solution, select a second edge node from a plurality of first edge nodes.

[0083] The optimized distributed deployment scheme in the embodiment of the present invention specifies the deployment location and number of edge nodes. Because the optimized distributed deployment scheme is a genetic algorithm that optimizes the deployment of nodes by simulating natural selection and genetic mechanisms, its goal is to find the optimal or nearly optimal node subset while satisfying specific constraints (such as cost, performance, coverage, etc.). Therefore, the optimized distributed deployment scheme is to screen several first edge nodes in the target area and select the second edge nodes with suitable locations and numbers.

[0084] Step 104: Determine the first resource allocation amount, the first task to be executed, and the first task scheduling amount in the second edge node by using a preset optimization strategy.

[0085] The preset optimization strategy in the embodiment of the present invention is set in the strategy selector. The strategy selector selects the most suitable optimization strategy according to the current state of the edge node, that is, determines the amount of various resources to be allocated to each second edge node and the amount of task scheduling to be executed. Among them, the current state of the edge node is monitored by the node status monitor, and the monitored content is the operating state (such as node state and network state) and resource usage of all second edge nodes. The resource allocator adjusts the allocation of computing resources and storage resources of the node according to the optimization strategy to ensure the reasonable distribution and efficient execution of tasks among nodes. The load balancer is responsible for balancing the load of each node to ensure that the system runs efficiently while avoiding single point overload. The task scheduler allocates and schedules tasks according to the optimization strategy and real-time status. The specific steps include:

[0086] Acquire first node state information and first network state information about the second edge node from the first target information;

[0087] The first resource allocation amount, the first task to be executed and the first task scheduling amount in the second edge node are determined by using the preset optimization strategy, the first node state information and the first network state information.

[0088] The above steps ensure that the second edge node can reasonably allocate resources and schedule tasks according to the current state information and optimization strategy to achieve the best balance between system performance and resource utilization.

[0089] In addition, after determining that the second edge node starts to execute the first task, the task execution status of the second edge node will be obtained in real time through the node status monitor. If it is found that no task is currently being executed (that is, the task execution status is "no task execution status"), it is necessary to obtain the unexecuted second task from the first task and continue to execute it until all the first tasks are completed. A task execution completion notification is sent to inform the staff that all scheduled tasks have been completed as planned. The specific steps include:

[0090] When it is determined that the second edge node starts to execute the first task, obtaining the task execution status of the second edge node through the node status monitor;

[0091] When it is determined that the task execution status is a non-executing task status, obtaining the unexecuted second task from the first task, and continuing to execute the second task;

[0092] When it is determined that the execution quantity of the first task reaches the first task scheduling quantity and the task execution status is a state where no task is executed, a task execution completion notification is sent.

[0093] The above steps ensure that the second edge node can effectively execute the task, and can promptly notify the relevant system or user after the task is completed to modify the task execution process, thereby avoiding errors caused by failure to obtain new tasks after the first task is completed.

[0094] Step 105: When it is determined that all first tasks have been completed, obtain second target information of the second edge node.

[0095] The second target information in the embodiment of the present invention includes at least one of target task request information, target node status information, target network status information and target environment information.

[0096] Step 106: compare the second target information with the first target information, and make adjustments based on the comparison result to obtain a second resource allocation amount and a second task scheduling amount of the second edge node.

[0097] In the embodiment of the present invention, the second target information is compared with the first target information to determine the difference between the optimized result and the original result. The difference information is generated into a report log and transmitted to the target page. The user adjusts the first resource allocation amount and the first task scheduling amount in the second edge node according to the content in the log on the target page.

[0098] Further, step 106, such as Figure 2 As shown:

[0099] Step 1061, comparing the second target information with the first target information to determine target difference information between the second target information and the first target information.

[0100] Step 1062, generating a report log using the target difference information, and transmitting the report log to the target page.

[0101] Step 1063 , in response to the user's adjustment operation on the target page, obtain a second resource allocation amount and a second task scheduling amount in the second edge node.

[0102] The above steps ensure that the system can reflect the state changes of edge nodes in real time and provide them to users through the user interface for viewing and intervention. Users can make necessary adjustments to resource allocation and task scheduling based on the difference information in the report log to optimize the performance and efficiency of the system.

[0103] It can be seen that in the embodiment of the present invention, by collecting and analyzing task requests, node status, network status and environmental information, the distribution of edge computing nodes is optimized using a genetic algorithm to improve computing efficiency and resource utilization. Multiple modules are used, including an input module, a genetic algorithm module, an optimization strategy module, an edge computing node management module, a communication module, a data storage module and an output module. The connections between the modules are as follows: Figure 3As shown, the connection between the input module (the input module consists of a task request receiver, a node status monitor, a network status monitor and an environmental information collector) and the genetic algorithm module: the input module transmits the task request, node status, network status and environmental information to the genetic algorithm module through a high-performance data bus. The task request receiver parses the task request, obtains the task description data, the CPU usage rate, memory usage rate and other information collected by the node status monitor, the network performance data monitored by the network status monitor, and the environmental data such as temperature and humidity collected by the environmental information collector, and transmits them to the population generator in the genetic algorithm module through the data bus for initializing the population. This connection ensures the real-time and accuracy of the data, and the population generation is adjusted in real time with the change of the input data. The connection between the genetic algorithm module and the optimization strategy module: the genetic algorithm module transmits the optimized distribution deployment scheme after fitness evaluation to the optimization strategy module through the internal high-speed data exchange interface. The population generated by the population generator forms a new population through selection, crossover and mutation operations. After the fitness evaluator evaluates the new population, the evaluation results and the optimal individuals are transmitted to the optimization strategy module. The strategy selector of the optimization strategy module determines the first resource allocation amount, the first task to be executed and the first task scheduling amount (resource task allocation plan) in the second edge node according to the optimization distribution deployment plan and the preset optimization strategy. This connection ensures that the optimization plan output by the genetic algorithm can be efficiently transmitted to the optimization strategy module to achieve rapid optimization. Connection between the optimization strategy module and the edge computing node management module: The optimization strategy module exchanges data with the edge computing node management module through a local area network connection. The resource task allocation plan generated by the strategy selector is transmitted to the edge computing node management module through the network interface. The resource allocator adjusts the computing resource allocation of the node according to the received resource task allocation plan, and the task scheduler schedules the task execution according to the new resource allocation plan. This connection ensures that the optimization plan generated by the optimization strategy module can be quickly applied to node management to achieve optimal resource allocation and efficient scheduling of tasks. Connection between the edge computing node management module and the communication module: The edge computing node management module is connected to the communication module through a standard Ethernet interface to realize data communication between nodes and between nodes and the central server. The node operation status and resource usage monitored by the node monitor are transmitted to the communication module through the network interface. The communication interface is responsible for establishing a connection with each node, and the data transmission protocol ensures the stability and reliability of data transmission. This connection ensures that the node management module can communicate with each edge node in real time to maintain data synchronization and task collaboration. Connection between the communication module and the data storage module: The communication module transmits the collected data to the data storage module through a high-speed data transmission channel. The data transmission protocol ensures the integrity and security of the data. The network controller manages the data traffic and transmits the node status data, network performance data, and environmental data to the data storage module.The database management system of the data storage module classifies and stores the received data, and the cache manager caches the real-time data. This connection ensures that the data collected by the communication module can be stored efficiently and safely, which is convenient for subsequent query and analysis. Connection between the data storage module and the genetic algorithm module: The data storage module is connected to the genetic algorithm module through the internal storage bus to provide historical data and intermediate results for the genetic algorithm module. The historical data stored in the database management system and the real-time data cached by the cache manager are transmitted to the genetic algorithm module through the storage bus for optimizing the algorithm parameters and improving the optimization effect. The intermediate results and final optimization scheme generated by the genetic algorithm module are also stored in the data storage module to ensure the traceability and checkability of the data. This connection ensures that the data provided by the data storage module can timely and accurately support the optimization process of the genetic algorithm. Connection between the data storage module and the output module: The data storage module transmits the optimization results and the generated report log to the output module through the data interface. The optimization results and report log data stored in the database management system are transmitted to the result generator of the output module through the data interface. The result generator generates the final optimization results, the report generator generates the system operation report and the optimization report, and the interface controller transmits the control instructions to the corresponding node or system administrator. This connection ensures that the data in the data storage module can be efficiently transferred to the output module, and the optimization effect and operation status can be intuitively displayed. Connection between the output module and the edge computing node management module: The output module transmits the control instructions to the edge computing node management module through the control bus. The control instructions generated by the interface controller are transmitted to the edge computing node management module through the control bus, and the edge computing node management module adjusts the resource allocation and task scheduling of the node according to the control instructions. This connection ensures that the control instructions generated by the output module can be transmitted to the node management module in a timely and accurate manner, realizing dynamic adjustment and optimization of the system.

[0104] Through the above connection relationship, the input and output data flows of different modules can be obtained, such as Figure 4As shown, the input module obtains task requests, node status, network status, and environmental information, and outputs the processed input data after parsing, namely, task type, computing requirements, at least one of time constraints, the CPU usage rate of each edge computing node, memory usage rate, storage space, and running status, the network bandwidth of the network where the edge computing node is located, at least one of network delay and packet loss rate, and at least one of the temperature, humidity, and power supply status of the environment where each edge computing node is located. The parsed data is input into the genetic algorithm module. After the genetic algorithm module receives the data transmitted by the input module, the population generator outputs the initialization population according to these data. Each individual represents a distribution scheme of an edge computing node. The number of individuals in the initialization population is determined by the system parameters. The population generator ensures the diversity of the population so as to improve the optimization ability of the algorithm in subsequent iterations. The initialization population is iterated by the algorithm to obtain an optimized distribution deployment scheme, and the optimized distribution deployment scheme is input into the optimization strategy module. Through the preset optimization strategy and the edge computing node management module, the node resource allocation and task scheduling scheme are obtained, and the node resource allocation and task scheduling scheme are input into the communication module to obtain real-time communication data, and the communication module transmits the collected data to the data storage module through a high-speed data transmission channel. The data transmission protocol ensures the integrity and security of the data. The network controller manages data traffic and transmits node status data, network performance data and environmental data to the data storage module. The optimization results and report data are obtained based on the stored data and transmitted to the output module. After the output module outputs, the user adjusts and generates control instructions, which are transmitted to the edge computing node management module for adjustment. The adjusted node resource allocation and task scheduling are obtained to achieve dynamic adjustment and optimization of the system.

[0105] The edge computing node management method provided by the embodiment of the present invention collects first target information of several first edge nodes in a target area; analyzes the first target information through a preset genetic algorithm to generate an optimized distribution deployment plan for several first edge nodes; sets a distribution deployment plan for edge nodes by comprehensively considering task requests, node status, network status and environmental information, thereby ensuring the rationality and effectiveness of node deployment; and selects a second edge node from several first edge nodes based on the optimized distribution deployment plan; determines the first resource allocation amount, the first task to be executed and the first task scheduling amount in the second edge node by using a preset optimization strategy; and obtains the second edge node when it is determined that all first tasks have been executed. The second target information is obtained, where the target information includes at least one of target task request information, target node status information, target network status information and target environment information; the second target information is compared with the first target information, and adjusted according to the comparison result to obtain a second resource allocation amount and a second task scheduling amount of the second edge node, and the resource and task allocation are dynamically adjusted and optimized through the comparison result, thereby improving the computing efficiency and resource utilization of the system. In the embodiment of the present invention, the distributed deployment scheme of the edge nodes is set by comprehensively considering the task request, node status, network status and environment information, and the resource allocation and task scheduling of the edge nodes are optimized and adjusted at the same time, thereby improving the computing efficiency and resource utilization of the system.

[0106] Reference Figure 5 , which shows a schematic diagram of the structure of an edge computing node management device provided by an embodiment of the present invention, such as Figure 5 As shown, the device may include:

[0107] The first collection module 201 is used to collect first target information of a plurality of first edge nodes in a target area;

[0108] A first generating module 202, configured to analyze the first target information by using a preset genetic algorithm to generate an optimized distribution deployment scheme for a plurality of the first edge nodes;

[0109] A first screening module 203, configured to screen a second edge node from a plurality of first edge nodes based on the optimized distribution deployment solution;

[0110] A first determination module 204 is used to determine the first resource allocation amount, the first task to be executed and the first task scheduling amount in the second edge node by using a preset optimization strategy;

[0111] A first acquisition module 205 is used to acquire second target information of the second edge node when it is determined that the first tasks are all completed, wherein the target information includes at least one of target task request information, target node status information, target network status information and target environment information;

[0112] The first adjustment module 206 is used to compare the second target information with the first target information, and make adjustments according to the comparison result to obtain a second resource allocation amount and a second task scheduling amount of the second edge node.

[0113] Optionally, the first generating module 202 specifically includes:

[0114] A first acquisition submodule is used to acquire target parameters set for the initialization population in the genetic algorithm, wherein the target parameters include population size, crossover probability and mutation probability, and the population size refers to the number of individuals specified to be included in the initialization population;

[0115] A first determination submodule is used to analyze the first target information through the target parameter to determine an initialization population of the first edge node, wherein the initialization population includes a plurality of first distribution deployment schemes for the first edge node, and each first distribution deployment scheme corresponds to an individual in the initialization population;

[0116] A second acquisition submodule is used to obtain a first fitness value of each first distribution deployment scheme through a fitness function preset in the genetic algorithm;

[0117] A second determination submodule is used to determine the selection probability of each first distribution deployment scheme based on the first fitness value using a target selection probability formula;

[0118] A first screening submodule, configured to screen out excellent second distribution deployment solutions from the first distribution deployment solutions according to the selection probability;

[0119] A first operation submodule, configured to perform a crossover operation and a mutation operation on the second distribution deployment scheme by using a crossover device and a mutator in the genetic algorithm to obtain a third distribution deployment scheme;

[0120] A third acquisition submodule, used to obtain a second fitness value of each third distribution deployment scheme through the preset fitness function;

[0121] The third determination submodule is used to determine the third distribution deployment scheme corresponding to the second fitness value that meets the preset condition as the optimized distribution deployment scheme for the first edge nodes when it is determined that the second fitness value meets the preset condition.

[0122] Optionally, the target task request information includes at least one of the task type, computing requirements, and time constraints; the target node status information includes at least one of the central processing unit utilization, memory utilization, storage space, and operating status of each edge computing node; the target network status information includes at least one of the network bandwidth, network delay, and packet loss rate of the network where the edge computing node is located; and the target environment information includes at least one of the temperature, humidity, and power supply status of the environment where each edge computing node is located.

[0123] Optionally, the fitness function formula is:

[0124]

[0125] Among them, α, β, γ, δ are weight parameters, It refers to the sum of n tasks in the first distribution deployment scheme, D i refers to the computational requirements of the ith task, x i It means that the i-th task is assigned to the first edge node of the target. It refers to the CPU usage of the target first edge node. Refers to the memory usage of the target first edge node, MEM total It refers to the total memory usage of the first edge node in the first distributed deployment scheme. Refers to the storage space of the target first edge node, STOR total It refers to the total storage space of the first edge node in the first distributed deployment solution. It refers to the network delay of the network where the target first edge node is located. It refers to the packet loss rate of the network where the target first edge node is located, f(x) refers to the first fitness value, and the target first edge node is the first edge node in the first distributed deployment scheme.

[0126] Optionally, the first determining module 204 specifically includes:

[0127] a fourth acquisition submodule, configured to acquire first node state information and first network state information about the second edge node from the first target information;

[0128] The fourth determination submodule is used to determine the first resource allocation amount, the first task to be executed and the first task scheduling amount in the second edge node by using a preset optimization strategy, the first node status information and the first network status information.

[0129] Optionally, the edge computing node management device further includes:

[0130] A second acquisition module is used to acquire the task execution status of the second edge node through a node status monitor when it is determined that the second edge node starts to execute the first task;

[0131] A third acquisition module is used to acquire an unexecuted second task from the first task when it is determined that the task execution status is a non-executed task status, and continue to execute the second task;

[0132] The sending module is used to send a task execution completion notification when it is determined that the execution quantity of the first task reaches the first task scheduling quantity and the task execution status is a state where no task is executed.

[0133] Optionally, the first adjustment module 206 specifically includes:

[0134] a fifth determination submodule, configured to compare the second target information with the first target information to determine target difference information between the second target information and the first target information;

[0135] A transmission submodule, used for generating a report log according to the target difference information, and transmitting the report log to a target page;

[0136] The adjustment submodule is used to obtain a second resource allocation amount and a second task scheduling amount in the second edge node in response to the user's adjustment operation on the target page.

[0137] The edge computing node management method provided by the embodiment of the present invention collects first target information of several first edge nodes in a target area; analyzes the first target information through a preset genetic algorithm to generate an optimized distribution deployment plan for several first edge nodes; sets a distribution deployment plan for edge nodes by comprehensively considering task requests, node status, network status and environmental information, thereby ensuring the rationality and effectiveness of node deployment; and selects a second edge node from several first edge nodes based on the optimized distribution deployment plan; determines the first resource allocation amount, the first task to be executed and the first task scheduling amount in the second edge node by using a preset optimization strategy; and obtains the second edge node when it is determined that all first tasks have been executed. The second target information is obtained, where the target information includes at least one of target task request information, target node status information, target network status information and target environment information; the second target information is compared with the first target information, and adjusted according to the comparison result to obtain a second resource allocation amount and a second task scheduling amount of the second edge node, and the resource and task allocation are dynamically adjusted and optimized through the comparison result, thereby improving the computing efficiency and resource utilization of the system. In the embodiment of the present invention, the distributed deployment scheme of the edge nodes is set by comprehensively considering the task request, node status, network status and environment information, and the resource allocation and task scheduling of the edge nodes are optimized and adjusted at the same time, thereby improving the computing efficiency and resource utilization of the system.

[0138] The embodiment of the present invention also provides a communication device, such as Figure 6 As shown, it includes a processor 301, a communication interface 302, a memory 303 and a communication bus 304, wherein the processor 301, the communication interface 302, and the memory 303 communicate with each other through the communication bus 304.

[0139] Memory 303, used for storing computer programs;

[0140] The processor 301 is used to execute the program stored in the memory 303 to implement the following steps:

[0141] Collecting first target information of a plurality of first edge nodes in a target area;

[0142] Analyzing the first target information by using a preset genetic algorithm to generate an optimized distribution deployment plan for a plurality of the first edge nodes;

[0143] Based on the optimized distributed deployment solution, selecting a second edge node from a plurality of the first edge nodes;

[0144] Determine the first resource allocation amount, the first task to be executed and the first task scheduling amount in the second edge node by using a preset optimization strategy;

[0145] When it is determined that the first tasks are all completed, obtaining second target information of the second edge node, wherein the target information includes at least one of target task request information, target node status information, target network status information and target environment information;

[0146] The second target information is compared with the first target information, and adjustments are made according to the comparison result to obtain a second resource allocation amount and a second task scheduling amount of the second edge node.

[0147] The communication bus mentioned in the above terminal can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into a physical address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0148] The communication interface is used for communication between the above terminal and other devices.

[0149] The memory may include a random access memory (RAM) or a non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0150] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.

[0151] The present invention also provides a readable storage medium, when the instructions in the storage medium are executed by a processor of an electronic device, the electronic device can execute the edge computing node management method of the aforementioned embodiment.

[0152] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0153] The algorithm and display provided herein are not inherently related to any particular computer, virtual device or other equipment. According to the above description, it is obvious that the structure required for constructing this type of device is. In addition, the present invention is not directed to any specific programming language. It should be understood that various programming languages ​​can be utilized to realize the content of the present invention described herein, and the description of the above specific language is to disclose the best mode of the present invention.

[0154] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0155] Similarly, it should be understood that in order to streamline the present invention and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, various features of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed method should not be interpreted as reflecting the following intention: that the claimed invention requires more features than those expressly recited in each claim. Rather, as reflected in the claims below, inventive aspects lie in less than all of the features of the individual embodiments disclosed above. Therefore, the claims that follow the detailed description are hereby expressly incorporated into the detailed description, with each claim itself serving as a separate embodiment of the present invention.

[0156] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and in addition they may be divided into a plurality of submodules or subunits or subcomponents. Except that at least some of such features and / or processes or units are mutually exclusive, all features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed in this manner may be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0157] The various component embodiments of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It should be understood by those skilled in the art that a microprocessor or digital signal processor (DSP) may be used in practice to implement some or all of the functions of some or all of the components in the sorting device according to the present invention. The present invention may also be implemented as a device or apparatus program for executing part or all of the methods described herein. Such a program for implementing the present invention may be stored on a computer-readable medium, or may be in the form of one or more signals. Such a signal may be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0158] It should be noted that the above embodiments illustrate the present invention rather than limit it, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets shall not be construed as a limitation on the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "one" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising a number of different elements and by means of a suitably programmed computer. In a unit claim enumerating a number of devices, several of these devices may be embodied by the same hardware item. The use of the words first, second, and third, etc., does not indicate any order. These words may be interpreted as names.

[0159] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices, apparatuses and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0160] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0161] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0162] It should be noted that the various data-related processes in the embodiments of the present application are all carried out in compliance with the corresponding data protection laws and policies of the country where the device is located, and with the authorization given by the owner of the corresponding device.

Claims

1. A method for managing edge computing nodes, characterized in that: The method comprises: Collecting first target information of a plurality of first edge nodes in a target area; Analyzing the first target information by using a preset genetic algorithm to generate an optimized distribution deployment plan for a plurality of the first edge nodes; Based on the optimized distributed deployment solution, selecting a second edge node from a plurality of the first edge nodes; Determine the first resource allocation amount, the first task to be executed and the first task scheduling amount in the second edge node by using a preset optimization strategy; When it is determined that the first tasks are all completed, obtaining second target information of the second edge node, wherein the target information includes at least one of target task request information, target node status information, target network status information and target environment information; The second target information is compared with the first target information, and adjustments are made according to the comparison result to obtain a second resource allocation amount and a second task scheduling amount of the second edge node.

2. The method according to claim 1, characterized in that The analyzing the first target information by a preset genetic algorithm to generate an optimized distribution deployment scheme for a plurality of the first edge nodes includes: Obtaining target parameters set for the initialization population in the genetic algorithm, the target parameters including population size, crossover probability and mutation probability, the population size refers to the number of individuals specified to be included in the initialization population; Analyzing the first target information by using the target parameter to determine an initialization population of the first edge node, wherein the initialization population includes a plurality of first distribution deployment schemes for the first edge node, and each first distribution deployment scheme corresponds to an individual in the initialization population; Obtaining a first fitness value of each first distribution deployment scheme through a fitness function preset in the genetic algorithm; Based on the first fitness value, determine the probability of selection of each first distribution deployment scheme using a target selection probability formula; Screening out an excellent second distribution deployment scheme from the first distribution deployment schemes according to the selection probability; Performing a crossover operation and a mutation operation on the second distribution deployment scheme by using a crossover device and a mutator in the genetic algorithm to obtain a third distribution deployment scheme; Obtaining a second fitness value of each third distribution deployment scheme through the preset fitness function; When it is determined that the second fitness value satisfies a preset condition, a third distribution deployment scheme corresponding to the second fitness value that satisfies the preset condition is determined as an optimized distribution deployment scheme for the first edge nodes.

3. The method according to claim 2, characterized in that The target task request information includes at least one of the task type, computing requirements, and time constraints; the target node status information includes at least one of the central processing unit utilization, memory utilization, storage space, and operating status of each edge computing node; the target network status information includes at least one of the network bandwidth, network delay, and packet loss rate of the network where the edge computing node is located; and the target environment information includes at least one of the temperature, humidity, and power supply status of the environment where each edge computing node is located.

4. The method according to claim 3, characterized in that The fitness function formula is: Among them, α, β, γ, δ are weight parameters, It refers to the sum of n tasks in the first distribution deployment scheme, D i refers to the computational requirements of the ith task, x i It means that the i-th task is assigned to the first edge node of the target. It refers to the CPU usage of the target first edge node. Refers to the memory usage of the target first edge node, MEM total It refers to the total memory usage of the first edge node in the first distributed deployment scheme. Refers to the storage space of the target first edge node, STOR total It refers to the total storage space of the first edge node in the first distributed deployment solution. It refers to the network delay of the network where the target first edge node is located. It refers to the packet loss rate of the network where the target first edge node is located, f(x) refers to the first fitness value, and the target first edge node is the first edge node in the first distributed deployment scheme.

5. The method according to claim 1, characterized in that The using a preset optimization strategy to determine the first resource allocation amount, the first task to be executed, and the first task scheduling amount in the second edge node includes: Acquire first node state information and first network state information about the second edge node from the first target information; The first resource allocation amount, the first task to be executed and the first task scheduling amount in the second edge node are determined by using the preset optimization strategy, the first node state information and the first network state information.

6. The method according to claim 1, characterized in that After determining the first resource allocation amount, the first task to be executed and the first task scheduling amount in the second edge node by using the preset optimization strategy, the method further includes: When it is determined that the second edge node starts to execute the first task, obtaining the task execution status of the second edge node through a node status monitor; When it is determined that the task execution status is a state where no task is executed, obtaining an unexecuted second task from the first task, and continuing to execute the second task; When it is determined that the execution quantity of the first task reaches the first task scheduling quantity and the task execution status is a state where no task is executed, a task execution completion notification is sent.

7. The method according to claim 1, characterized in that The comparing the second target information with the first target information and adjusting according to the comparison result to obtain the second resource allocation amount and the second task scheduling amount of the second edge node includes: Comparing the second target information with the first target information to determine target difference information between the second target information and the first target information; Generate a report log using the target difference information, and transmit the report log to a target page; In response to the user's adjustment operation on the target page, a second resource allocation amount and a second task scheduling amount in the second edge node are obtained.

8. An edge computing node management device, characterized in that: include: A first acquisition module, used to acquire first target information of a plurality of first edge nodes in a target area; A first generating module, configured to analyze the first target information by using a preset genetic algorithm to generate an optimized distribution deployment scheme for a plurality of the first edge nodes; A first screening module, configured to screen a second edge node from a plurality of first edge nodes based on the optimized distributed deployment solution; A first determination module, configured to determine a first resource allocation amount, a first task to be executed, and a first task scheduling amount in the second edge node by using a preset optimization strategy; A first acquisition module is used to acquire second target information of the second edge node when it is determined that the first tasks are all completed, wherein the target information includes at least one of target task request information, target node status information, target network status information and target environment information; The first adjustment module is used to compare the second target information with the first target information, and make adjustments according to the comparison result to obtain a second resource allocation amount and a second task scheduling amount of the second edge node.

9. A communication device, characterized in that: include: A transceiver, a memory, a processor, and a program stored on the memory and executable on the processor; The processor is used to read the program in the memory to implement the steps in the edge computing node management method as described in any one of claims 1-7.

10. A readable storage medium for storing a program, characterized in that: When the stored program is executed by the processor, the steps in the edge computing node management method as described in any one of claims 1 to 7 are implemented.

Images