Edge heterogeneous computing power equipment scheduling method and system, storage medium and equipment

Abstract

The invention provides an edge heterogeneous computing power equipment scheduling method and system, a storage medium and equipment, and the method comprises the steps: carrying out the scanning of edge heterogeneous computing power assets, recognizing and registering all edge computing power equipment and the computing power resources thereof, and forming a computing power asset list; based on the computing power asset list, analyzing a service scene and a service demand, and predicting a service demand change of the edge heterogeneous computing power; collecting edge heterogeneous computing power state parameters, and assigning the collected state parameters based on business demand changes to obtain a comprehensive evaluation value; and cooperatively scheduling each edge heterogeneous computing power device by adopting an RSDL scheduling algorithm and a red-black tree structure according to the obtained comprehensive evaluation value. According to the real-time business requirements of the tasks and the load condition of the edge computing power equipment, the computing resources are dynamically allocated, the tasks are allocated to the nodes with low loads for calculation, and the overall energy consumption is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of heterogeneous computing, and particularly relates to a method, system, storage medium and device for scheduling edge heterogeneous computing power devices. Background Art

[0002] With the booming rise of edge computing, we have witnessed the emergence of edge computing power in various forms, such as GPU servers, embedded AI boxes, and FPGAs. These diverse computing power structures have brought unprecedented flexibility and efficiency to data processing and computing tasks. However, this increase in heterogeneity has also brought huge challenges in resource management and scheduling, making efficient computing power allocation and utilization an urgent problem to be solved.

[0003] In the prior art, the existence of heterogeneous computing power resources makes traditional resource management and scheduling methods no longer applicable. Each heterogeneous computing power resource has its unique performance and characteristics. How to quickly and accurately select the most suitable computing power resource according to application requirements has become a technical problem. In addition, since additional mechanisms are required to support communication and collaborative work between heterogeneous resources, this also increases the complexity of management and scheduling.

[0004] Secondly, the demands of edge computing power users show a trend of diversification and rapid growth. From simple data processing to complex AI computing, from real-time video monitoring to remote industrial control, the computing power requirements of various application scenarios are different. However, the current heterogeneous computing power resources at the edge and on the device side are still in an "island" state, lacking collaborative management, making it difficult to effectively integrate and utilize these resources, and thus unable to meet the requirements of complex application scenarios. Summary of the Invention

[0005] In view of the deficiencies of the prior art, the present invention provides a method for scheduling edge heterogeneous computing power devices, which includes: Performing a scan of edge heterogeneous computing power assets, identifying and registering each edge computing power device and its computing power resources, and forming a computing power asset list; Based on the computing power asset list, analyzing the business scenario and business requirements, and predicting the change of business requirements for edge heterogeneous computing power; Collecting the state parameters of edge heterogeneous computing power, and assigning values to the collected state parameters based on the change of business requirements to obtain a comprehensive evaluation value; According to the obtained comprehensive evaluation value, and using the RSDL scheduling algorithm and the red-black tree structure, coordinately scheduling each edge heterogeneous computing power device.

[0006] Furthermore, the collecting of the state parameters of edge heterogeneous computing power includes: collecting by using an edge computing power gateway and collecting by using an edge AI computing power server.

[0007] Furthermore, assigning values to the collected state parameters to obtain a comprehensive evaluation value includes: for the collected state parameters, using the entropy weight assignment method to calculate the entropy weight value of the state parameters and obtain the comprehensive evaluation value.

[0008] Furthermore, using the entropy weight assignment method to calculate the entropy weight value of the state parameters includes: Processing the collected state parameters using the range standardization method and calculating the state parameter proportion; Calculating the entropy value of the state parameters according to the calculated parameter proportion; Calculating the coefficient of variation according to the calculated entropy value, and calculating the entropy weight value of the state parameters according to the coefficient of variation; Calculating and obtaining the comprehensive evaluation value of the edge heterogeneous computing power device according to the calculated entropy weight value and the state parameter proportion.

[0009] Furthermore, according to the obtained comprehensive evaluation value, and using the RSDL scheduling algorithm and the red-black tree structure to coordinately schedule each edge heterogeneous computing power device includes: Configuring static priority parameters for each edge heterogeneous computing power device, where the size of the static priority parameters is determined by the comprehensive evaluation value; Configuring virtual running time parameters for each edge heterogeneous computing power device, where the virtual running time parameters are determined by the static priority parameters; Using a red-black tree as the queue structure setting, and inserting each edge heterogeneous computing power device into the edge heterogeneous computing power device ready queue according to the virtual running time parameters; Using the RSDL scheduling algorithm to maintain the edge heterogeneous computing power device ready queue and coordinately schedule each edge heterogeneous computing power device.

[0010] An edge heterogeneous computing power device scheduling system, the system includes: a computing power device scanning module, a computing power demand prediction module, a computing power device state monitoring module, and a computing power device scheduling module; The computing power device scanning module is used to perform edge heterogeneous computing power asset scanning, identify and register each edge computing power device and its computing power resources, and form a computing power asset list; The computing power demand prediction module is used to analyze the business scenario and business demand based on the computing power asset list, and predict the change of the business demand for edge heterogeneous computing power; The computing power device state monitoring module is used to collect the edge heterogeneous computing power state parameters, and assign values to the collected state parameters based on the change of the business demand to obtain a comprehensive evaluation value; The computing power device scheduling module is used to coordinately schedule each edge heterogeneous computing power device according to the obtained comprehensive evaluation value, and using the RSDL scheduling algorithm and the red-black tree structure.

[0011] Further, the computing power device status monitoring module is specifically configured to collect data using an edge computing power gateway and an edge AI computing power server.

[0012] Further, the computing power device status monitoring module is specifically configured to calculate the entropy weight value of the status parameters using the entropy weight assignment method for the collected status parameters, and obtain a comprehensive evaluation value.

[0013] Further, the computing power device status monitoring module is specifically configured to Process the collected status parameters using the range standardization method and calculate the proportion of the status parameters; Calculate the entropy value of the status parameters according to the calculated parameter proportion; Calculate the coefficient of variation according to the calculated entropy value, and calculate the entropy weight value of the status parameters according to the coefficient of variation; Calculate and obtain the comprehensive evaluation value of the edge heterogeneous computing power device according to the calculated entropy weight value and the status parameter proportion.

[0014] Further, the computing power device scheduling module is specifically configured to Configure static priority parameters for each edge heterogeneous computing power device, where the size of the static priority parameters is determined by the comprehensive evaluation value; Configure virtual running time parameters for each edge heterogeneous computing power device, where the virtual running time parameters are determined by the static priority parameters; Use a red-black tree as the queue structure setting, and insert each edge heterogeneous computing power device into the edge heterogeneous computing power device ready queue according to the virtual running time parameter sorting; Use the RSDL scheduling algorithm to maintain the edge heterogeneous computing power device ready queue and coordinate the scheduling of each edge heterogeneous computing power device.

[0015] A computer-readable storage medium stores a computer program therein, and when the computer program is executed by a processor, the steps of any of the above-mentioned edge heterogeneous computing power device scheduling methods are implemented.

[0016] An electronic device includes a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete communication with each other through the communication bus; The memory is used to store a computer program; The processor is configured to implement the steps of any of the above-mentioned edge heterogeneous computing power device scheduling methods when executing the program stored in the memory.

[0017] A computer program product includes a computer program / instructions, and when the computer program / instructions are executed by a processor, the steps of any of the above-mentioned edge heterogeneous computing power device scheduling methods are implemented.

[0018] Compared with the prior art, the present invention has the following advantages: 1. The present invention proposes a method for scheduling edge heterogeneous computing power devices. By dynamically allocating computing resources according to the real-time service requirements of tasks and the load conditions of edge computing power devices, tasks are assigned to nodes with lower loads for calculation, reducing overall energy consumption. In addition, energy optimization methods such as task migration, sleep wake-up, and dynamic voltage and frequency adjustment can be adopted to further reduce energy consumption.

[0019] 2. Through methods such as task partitioning, task scheduling, and resource management, efficient video intelligent analysis task calculation can be achieved on edge nodes, significantly reducing the overall calculation time and latency.

[0020] 3. Dynamic scheduling can balance the load of nodes, avoiding system performance degradation caused by overloading of certain nodes. By optimizing resource allocation and task scheduling, the throughput of the system can be improved to handle more tasks.

[0021] Other features and advantages of the present invention will be described in the following specification, and part of them will become obvious from the specification or be understood by implementing the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the structures pointed out in the specification, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0023] Figure 1 Shows an edge computing power network related to the present invention.

[0024] Figure 2 Shows a flowchart of a method for scheduling edge heterogeneous computing power devices according to the present invention.

[0025] Figure 3 Shows a schematic diagram of the modules of a system for scheduling edge heterogeneous computing power devices according to the present invention.

[0026] Figure 4 Shows a red-black tree queue structure set in an embodiment of the present invention.

[0027] Figure 5 Shows a relationship table between vruntime parameters and static_prio parameters of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0029] An edge computing power network related to the present invention, as Figure 1 shown, its structure includes an edge heterogeneous computing power collaborative scheduling system, an edge computing power gateway, and edge computing power facilities. Among them, the edge computing power gateway is an intelligent device integrating functions such as high-performance computing, network communication, and data storage. It is located at the edge of the network, that is, near the source of data generation, aiming to improve data processing efficiency and response speed. This box has powerful computing capabilities, can execute local data analysis and processing tasks, can perform real-time processing and analysis on data collected by devices such as sensors and cameras, and provides accurate data support for fields such as intelligent security, industrial manufacturing, and environmental monitoring; The edge computing power facility is an AI computing power server device privately deployed at the edge. By integrating high-performance graphics processing units (GPUs), it provides powerful parallel computing capabilities, can quickly process and analyze large-scale data sets at the edge of the network, especially for complex computing tasks such as deep learning, computer vision, and image processing, and realizes low-latency and high-efficiency intelligent decision-making and data processing; The edge heterogeneous computing power collaborative scheduling system is used to uniformly manage and efficiently schedule edge computing resources such as edge gateway boxes and edge computing power servers. Through intelligent computing power allocation and task scheduling strategies, it optimizes resource utilization, improves data processing speed and system response capabilities, and ensures that various edge intelligent applications can run stably and efficiently. This platform can monitor the operating status of edge devices in real time and dynamically adjust computing power resources to meet the requirements of computing power, latency, and bandwidth in different application scenarios. Through the edge computing power collaborative management platform, users can achieve centralized management, flexible configuration, and intelligent scheduling of edge computing resources, improve business operation efficiency, reduce operation and maintenance costs, and provide strong support for digital transformation and intelligent upgrading.

[0030] In an embodiment of the present invention, as Figure 2 shown, a method for scheduling edge heterogeneous computing power devices includes the following steps: S1. Perform a scan of edge heterogeneous computing power assets, identify and register each edge computing power device and its computing power resources, and form a computing power asset list.

[0031] S2. Analyze the business scenarios and requirements based on the computing power asset list, predict the changes in the business requirements for edge heterogeneous computing power, and ensure the reasonable allocation and efficient utilization of computing power resources.

[0032] S3. Collect the status parameters of edge heterogeneous computing power, assign values to the collected status parameters based on the changes in business requirements, obtain a comprehensive evaluation value, and use this to monitor and record edge heterogeneous computing power problems.

[0033] S3.1. Collect the status parameters of edge computing power.

[0034] Optionally, the status parameters collected by the edge AI computing power server include: device performance parameters, network status parameters, task execution parameters, and energy consumption parameters. The status parameters collected by the edge computing power gateway include: device connection parameters, data traffic parameters, protocol conversion parameters, and local processing parameters.

[0035] Optionally, the device performance parameters include CPU usage rate, memory occupancy rate, disk space usage, etc., which reflect the overall performance status of the server. The network status parameters include network bandwidth usage, network latency, and packet loss rate, etc., which help to understand the network connection status of the server. The task execution parameters include the number of running tasks, task execution progress, and task resource consumption, etc., which are used to monitor and manage the tasks on the server. The energy consumption parameters, including the energy consumption of the server, are also important objects to be collected, which helps to achieve the operation and maintenance goal of green energy conservation. The device connection parameters include the connection status, device type, device identification, etc. of various devices (such as sensors, PLCs, etc.) connected to the gateway. The data traffic parameters include the number of sent and received data packets, data transmission rate, etc., which helps to understand the data processing ability of the gateway. The protocol conversion parameters include that the gateway usually supports the conversion of multiple communication protocols, so it is necessary to collect the detailed information of protocol conversion to ensure the correct transmission of data. The local processing parameters include the information such as the processing result, processing time, and resource consumption when the gateway processes data locally, to evaluate the efficiency and effect of local processing.

[0036] S3.2. Assign values to the collected status parameters to obtain a comprehensive evaluation value, and use this to monitor edge computing power devices, discover and record edge heterogeneous computing power problems.

[0037] In another embodiment of the present invention, for the collected edge computing power status parameters, the entropy weight assignment method is used to calculate the entropy weight value of the status parameters and obtain a comprehensive evaluation value. The steps include: S3.2.1. Assume n state parameters of edge computing power for m - time collection in a cycle , and its matrix X is expressed as:

[0038] S3.2.2. Process the state parameters using the inspection standardization method, and calculate the parameter proportion. For example, calculate the parameter proportion P of the j - th parameter state of the i - th edge heterogeneous computing power device ij , and its formula is expressed as:

[0039] S3.2.3. According to the calculated parameter proportion, calculate the entropy value of the j - th state parameter. Among them, the formula for calculating the entropy value e j is expressed as,

[0040] where is the normalization constant.

[0041] S3.2.4. According to the calculated entropy value, calculate the coefficient of variation d j , and its formula is expressed as,

[0042] S3.2.5. According to the coefficient of variation, calculate the entropy weight w j , and its formula is expressed as,

[0043] S3.2.6. According to the calculated entropy weight and parameter proportion, calculate the comprehensive evaluation value Z of the i - th edge computing power device (server facility or edge gateway) i , and its formula is expressed as,

[0044] S4. According to the obtained comprehensive evaluation value, and using the RSDL scheduling algorithm and the red - black tree structure, coordinately schedule each edge heterogeneous computing power device.

[0045] Optionally, the core of the RSDL algorithm lies in its unique priority processing mechanism and the combination of "small polling" and "large rotation".

[0046] In another embodiment of the present invention, first, configure a static priority parameter (static_prio parameter) for each edge heterogeneous computing power device, and the scheduling platform allows it to execute with this priority. Among them, the size of the static_prio parameter is determined by the comprehensive evaluation value Z of the state parameter entropy weight.

[0047] Then, configure virtual runtime parameters (vruntime parameters) for each edge heterogeneous computing device to record the cumulative call duration of the node.

[0048] Finally, set up an edge heterogeneous computing device ready queue (MECQ) in the global system. Use a red - black tree as the data structure. The red - black tree sacrifices the superior condition of strict height balance. It only requires partial balance, reduces the requirement for rotations, and thus improves performance. The red - black tree can perform search, insertion, and deletion operations with a time complexity of O(log2n).

[0049] Optionally, the edge heterogeneous computing device with the smallest vruntime parameter value will be the next to run, so as to ensure that the cumulative running duration of each task is evenly distributed as much as possible.

[0050] In another embodiment of the present invention, a typical round - robin algorithm is used to rotate among the edge heterogeneous computing devices with the highest priority. When an edge heterogeneous computing device uses up its vruntime parameter quota at a given priority level, it will be given a new quota. For example, the quota given to the virtual running time of the i - th edge heterogeneous computing device next time is expressed by the formula, vruntime_i = vruntime_i×(weight = 1024) / static_prio_i where the quota of the vruntime parameter is determined by the static priority static_prio parameter, as Figure 5 shown. The entropy - weight comprehensive evaluation value Z of the edge heterogeneous computing device with a high priority is relatively high, the growth of the vruntime parameter is relatively slower, the single - run time is shorter, and the call frequency is higher; the entropy - weight comprehensive evaluation value Z of the edge heterogeneous computing device with a low priority is relatively low, the growth of the vruntime parameter is relatively faster, the single - run time is longer, and the call frequency is lower, so that even the edge heterogeneous computing device with the lowest priority can finally be utilized.

[0051] In another embodiment of the present invention, as Figure 4 shown, first maintain a red - black tree sorted according to the vruntime parameter value through RSDL. All runnable scheduling entities are inserted into the red - black tree sorted by the vruntime parameter value. The scheduling edge heterogeneous computing device with a smaller vruntime parameter value for the execution time is arranged on the left side of the red - black tree. At the next task scheduling, select the scheduling entity with less virtual running time to run.

[0052] In another embodiment of the present invention, as Figure 3As shown, an edge heterogeneous computing power equipment scheduling system includes: a computing power equipment scanning module, a computing power demand prediction module, a computing power equipment status monitoring module and a computing power equipment scheduling module.

[0053] Optionally, the computing power scanning module is used to scan edge heterogeneous computing power assets, identify and register each edge computing power device and its computing power resources, and form a computing power asset list; The computing power demand prediction module is used to analyze business scenarios and business needs based on the computing power asset list, and predict changes in business demand for edge heterogeneous computing power; The computing power equipment detection module is used to collect edge heterogeneous computing power status parameters, and assign values ​​to the collected status parameters based on changes in business needs to obtain a comprehensive evaluation value, so as to monitor and record edge heterogeneous computing power problems; The computing power equipment scheduling module is used to coordinate the scheduling of various edge heterogeneous computing power devices based on the comprehensive evaluation value obtained, using the RSDL scheduling algorithm and the red-black tree structure.

[0054] Based on the above disclosed content, the present invention also provides an electronic device accordingly. The electronic device of an embodiment of the present invention includes at least one electrically connected processor and at least one storage medium, the storage medium is electrically connected to the processor, wherein the storage medium stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the method as described above.

[0055] Based on the same inventive concept, the present invention also provides a storage medium, which stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method as described above.

[0056] The above description and the accompanying drawings fully illustrate the embodiments of the present invention so that those skilled in the art can practice them. Other embodiments may include structural and other changes. The embodiments represent only possible variations. Unless explicitly required, individual components and functions are optional, and the order of operations may vary. Parts and features of some embodiments may be included in or replace parts and features of other embodiments. The embodiments of the present invention are not limited to the structures described above and shown in the accompanying drawings, and various modifications and changes may be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A method for scheduling edge heterogeneous computing devices, characterized in that: The method comprises, Scan edge heterogeneous computing assets, identify and register each edge computing device and its computing resources, and form a computing asset list; Based on the computing power asset list, analyze business scenarios and business needs, and predict changes in business needs for edge heterogeneous computing power; Collect edge heterogeneous computing power status parameters, and assign values ​​to the collected status parameters based on changes in business needs to obtain a comprehensive evaluation value; Based on the comprehensive evaluation value obtained, the RSDL scheduling algorithm and red-black tree structure are used to coordinately schedule various edge heterogeneous computing devices.

2. The method for scheduling edge heterogeneous computing devices according to claim 1, characterized in that: The collecting of edge heterogeneous computing power status parameters includes: using an edge computing power gateway to collect edge heterogeneous computing power status parameters and using an edge AI computing power server to collect edge heterogeneous computing power status parameters.

3. The method for scheduling edge heterogeneous computing devices according to claim 1, characterized in that: The step of assigning values ​​to the collected state parameters to obtain a comprehensive evaluation value includes: using an entropy weight assignment method to calculate the entropy weight of the state parameters to obtain a comprehensive evaluation value.

4. The method for scheduling edge heterogeneous computing devices according to claim 3 is characterized in that: The entropy weight assignment method is used to calculate the entropy weight of the state parameter, including: The collected state parameters are processed by the range standardization method to calculate the proportion of state parameters; According to the calculated parameter proportion, the entropy value of the state parameter is calculated; According to the calculated entropy value, the coefficient of variation is calculated, and according to the coefficient of variation, the entropy weight of the state parameter is calculated; According to the calculated entropy weight and state parameter proportion, the comprehensive evaluation value of the edge heterogeneous computing power equipment is calculated.

5. The method for scheduling edge heterogeneous computing devices according to claim 1, characterized in that: According to the obtained comprehensive evaluation value, the RSDL scheduling algorithm and the red-black tree structure are used to coordinately schedule various edge heterogeneous computing devices, including: Configure a static priority parameter for each edge heterogeneous computing device, wherein the size of the static priority parameter is determined by the comprehensive evaluation value; Configuring virtual runtime parameters for each edge heterogeneous computing device, wherein the virtual runtime parameters are determined by static priority parameters; A red-black tree is used as the queue structure setting, and each edge heterogeneous computing device is sorted according to the virtual running time parameter and inserted into the edge heterogeneous computing device ready queue; The RSDL scheduling algorithm is used to maintain the ready queue of edge heterogeneous computing devices and coordinate the scheduling of various edge heterogeneous computing devices.

6. An edge heterogeneous computing device scheduling system, characterized in that: The system includes: a computing power equipment scanning module, a computing power demand prediction module, a computing power equipment status monitoring module and a computing power equipment scheduling module; The computing power device scanning module is used to scan edge heterogeneous computing power assets, identify and register each edge computing power device and its computing power resources, and form a computing power asset list; The computing power demand prediction module is used to analyze business scenarios and business needs based on the computing power asset list, and predict changes in business demand for edge heterogeneous computing power; The computing power equipment status monitoring module is used to collect edge heterogeneous computing power status parameters, and assign values ​​to the collected status parameters based on changes in business requirements to obtain a comprehensive evaluation value; The computing power device scheduling module is used to coordinately schedule various edge heterogeneous computing power devices based on the obtained comprehensive evaluation value, and adopts the RSDL scheduling algorithm and the red-black tree structure.

7. The edge heterogeneous computing device scheduling system according to claim 6, characterized in that: The computing power equipment status monitoring module is specifically used to collect edge heterogeneous computing power status parameters using an edge computing power gateway and to collect edge heterogeneous computing power status parameters using an edge AI computing power server.

8. The edge heterogeneous computing device scheduling system according to claim 6, characterized in that: The computing power equipment status monitoring module is specifically used to calculate the entropy weight of the collected status parameters using the entropy weight assignment method to obtain a comprehensive evaluation value.

9. The edge heterogeneous computing device scheduling system according to claim 8, characterized in that: The computing power equipment status monitoring module is specifically used to The collected state parameters are processed by the range standardization method to calculate the proportion of state parameters; According to the calculated parameter proportion, the entropy value of the state parameter is calculated; According to the calculated entropy value, the coefficient of variation is calculated, and according to the coefficient of variation, the entropy weight of the state parameter is calculated; According to the calculated entropy weight and state parameter proportion, the comprehensive evaluation value of the edge heterogeneous computing power equipment is calculated.

10. The edge heterogeneous computing device scheduling system according to claim 6, characterized in that: The computing power equipment scheduling module is specifically used to: Configure a static priority parameter for each edge heterogeneous computing device, wherein the size of the static priority parameter is determined by the comprehensive evaluation value; Configuring virtual runtime parameters for each edge heterogeneous computing device, wherein the virtual runtime parameters are determined by static priority parameters; A red-black tree is used as the queue structure setting, and each edge heterogeneous computing device is sorted according to the virtual running time parameter and inserted into the edge heterogeneous computing device ready queue; The RSDL scheduling algorithm is used to maintain the ready queue of edge heterogeneous computing devices and coordinate the scheduling of various edge heterogeneous computing devices.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the edge heterogeneous computing device scheduling method described in any one of claims 1-5 are implemented.

12. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory, used to store computer programs; The processor is used to implement the steps of the edge heterogeneous computing device scheduling method described in any one of claims 1-5 when executing the program stored in the memory.

13. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the steps of the edge heterogeneous computing device scheduling method as described in any one of claims 1-5 are implemented.

Images