Scheduling scheme determination method and device, medium and product

Abstract

The invention provides a scheduling scheme determination method and device, a medium and a product, relates to the technical field of cloud computing, and aims to solve the problem that a scheduling scheme determined by a general technology is lack of accuracy. The scheduling scheme determination method comprises the following steps: acquiring task information of a task to be scheduled, task processing time delay of task processing equipment and transmission time delay of a network link; and determining a plurality of to-be-selected scheduling schemes according to the task information, the task processing time delay and the transmission time delay. And determining a target scheduling scheme based on the plurality of to-be-selected scheduling schemes, a tabu search algorithm and a genetic algorithm. Wherein the task processing time delay is used for representing the duration of processing the to-be-scheduled task by the task processing equipment, and the transmission time delay is used for representing the duration of transmitting the to-be-scheduled task to the task processing equipment.

Description

Technical Field

[0001] The present application relates to the field of cloud computing technology, and in particular to a method, device, medium and product for determining a scheduling scheme. Background Art

[0002] With the development of technologies such as the Internet of Things and big data, the amount of data has exploded. Efficient processing of large amounts of data, real-time response, and resource allocation are key issues that need to be addressed. Therefore, it is necessary to coordinate the scheduling of resources across the cloud, network, edge, and end. Among them, the cloud (also known as the cloud layer) and the edge (also known as the edge layer) are used to provide computing resources and storage resources. The end (also known as the terminal layer) is used to interact with users, such as obtaining tasks to be processed and outputting task processing results. The network (also known as the network layer) is used to connect the cloud layer, edge layer, and terminal layer, and transmit data, such as tasks to be processed and task processing results.

[0003] General technologies usually focus on resource integration and scheduling of cloud, edge and end.

[0004] However, the network layer is an important link between cloud, edge, and end. The general technology ignores the impact of the network layer when determining the scheduling scheme. Therefore, in a complex network environment, the scheduling scheme determined based on general technology lacks accuracy. Summary of the invention

[0005] The embodiments of the present disclosure provide a method, device, medium and product for determining a scheduling scheme, aiming to solve the problem that the scheduling scheme determined by general technology lacks accuracy.

[0006] In order to achieve the above objectives, this application adopts the following technical solutions:

[0007] In a first aspect, a method for determining a scheduling scheme is provided, including: obtaining task information of a task to be scheduled, a task processing delay of a task processing device, and a transmission delay of a network link; the task processing delay is used to indicate the time length for a task processing device to process the task to be scheduled; the transmission delay is used to indicate the time length for transmitting the task to be scheduled to the task processing device; according to the task information, the task processing delay and the transmission delay, a plurality of candidate scheduling schemes are determined; based on the plurality of candidate scheduling schemes, a taboo search algorithm and a genetic algorithm, a target scheduling scheme is determined.

[0008] Optionally, the task information includes: the maximum tolerable delay for processing the task to be scheduled; based on the task information, task processing delay and transmission delay, determine multiple scheduling schemes to be selected, including: for each task to be scheduled and each task processing device, when the sum of the transmission delay of the task to be scheduled and the task processing delay of the task processing device is less than or equal to the maximum tolerable delay of the task to be scheduled, determine multiple scheduling schemes to be selected based on the number of tasks to be scheduled and the number of task processing devices.

[0009] Optionally, a target scheduling scheme is determined based on multiple candidate scheduling schemes, a taboo search algorithm and a genetic algorithm, including: determining the fitness of multiple candidate scheduling schemes based on the scheduling information of the candidate scheduling schemes; updating the taboo table of the taboo search algorithm based on the fitness; and determining the target scheduling scheme based on the fitness, multiple candidate scheduling schemes, the taboo table and the genetic algorithm.

[0010] Optionally, the genetic algorithm includes: selection operation, crossover operation and mutation operation; based on fitness, multiple candidate scheduling schemes, taboo table and genetic algorithm, determining the target scheduling scheme, including: performing a selection operation on multiple candidate scheduling schemes based on the taboo table and fitness to obtain at least one group of candidate scheduling schemes; each group of candidate scheduling schemes includes 2 candidate scheduling schemes, and each group of candidate scheduling schemes does not include the candidate scheduling schemes in the taboo table; performing a crossover operation on each group of candidate scheduling schemes based on the taboo table to obtain multiple candidate scheduling schemes after the crossover operation; performing a mutation operation on the multiple candidate scheduling schemes after the crossover operation based on the taboo table to obtain the candidate scheduling schemes after the mutation operation; repeating the selection operation, crossover operation and mutation operation until the convergence condition is met, and determining the candidate scheduling scheme after the mutation operation as the target scheduling scheme; the convergence condition includes: the number of schemes of the candidate scheduling schemes after the mutation operation is less than the preset number of schemes, or the number of mutation operations performed is greater than the preset number of operations.

[0011] Optionally, the taboo table includes a global taboo table and a local taboo table; the global taboo table is used to eliminate the candidate scheduling schemes in the global taboo table when performing selection operations, crossover operations and mutation operations; the local taboo table is used to eliminate the candidate scheduling schemes in the local taboo table when performing crossover operations and mutation operations.

[0012] Optionally, the scheduling scheme determination method further includes: in the process of executing the selection operation, the crossover operation and the mutation operation, updating the taboo table according to the result of each operation.

[0013] Optionally, the scheduling information includes at least one of the following: the task success rate of the scheduling scheme to be selected, the average delay of the task to be scheduled, the resource utilization of the task processing equipment, the load rate of the task processing equipment, and the bandwidth utilization of the network link. According to the scheduling information of the scheduling scheme to be scheduled, the fitness of multiple scheduling schemes to be selected is determined, including: determining the fitness as the sum of the products of each scheduling information and the weight factor of the scheduling information.

[0014] In a second aspect, a scheduling scheme determination device is provided, which includes: a communication unit and a processing unit; the communication unit is used to obtain task information of the task to be scheduled, the task processing delay of the task processing device, and the transmission delay of the network link; the task processing delay is used to indicate the time length for the task processing device to process the task to be scheduled; the transmission delay is used to indicate the time length for transmitting the task to be scheduled to the task processing device; the processing unit is used to determine multiple candidate scheduling schemes based on the task information, the task processing delay and the transmission delay; the processing unit is also used to determine the target scheduling scheme based on the multiple candidate scheduling schemes, the taboo search algorithm and the genetic algorithm.

[0015] Optionally, the task information includes: the maximum tolerable delay for processing the task to be scheduled; a processing unit, specifically used to: for each task to be scheduled and each task processing device, when the sum of the transmission delay of the task to be scheduled and the task processing delay of the task processing device is less than or equal to the maximum tolerable delay of the task to be scheduled, determine multiple scheduling schemes to be selected according to the number of tasks to be scheduled and the number of task processing devices.

[0016] Optionally, the processing unit is specifically used to: determine the fitness of multiple candidate scheduling schemes based on the scheduling information of the scheduling schemes to be scheduled; update the taboo table of the taboo search algorithm based on the fitness; determine the target scheduling scheme based on the fitness, multiple candidate scheduling schemes, the taboo table and the genetic algorithm.

[0017] Optionally, the genetic algorithm includes: selection operation, crossover operation and mutation operation; a processing unit, specifically used to: perform selection operation on multiple candidate scheduling schemes based on the taboo table and fitness to obtain at least one group of candidate scheduling schemes; each group of candidate scheduling schemes includes 2 candidate scheduling schemes, and each group of candidate scheduling schemes does not include the candidate scheduling schemes in the taboo table; perform crossover operation on each group of candidate scheduling schemes based on the taboo table to obtain multiple candidate scheduling schemes after the crossover operation; perform mutation operation on the multiple candidate scheduling schemes after the crossover operation based on the taboo table to obtain the candidate scheduling schemes after the mutation operation; repeat the selection operation, crossover operation and mutation operation until the convergence condition is met, and determine the candidate scheduling scheme after the mutation operation as the target scheduling scheme; the convergence condition includes: the number of schemes of the candidate scheduling schemes after the mutation operation is less than the preset number of schemes, or the number of times the mutation operation is performed is greater than the preset number of operations.

[0018] Optionally, the taboo table includes a global taboo table and a local taboo table; the global taboo table is used to eliminate the candidate scheduling schemes in the global taboo table when performing selection operations, crossover operations and mutation operations; the local taboo table is used to eliminate the candidate scheduling schemes in the local taboo table when performing crossover operations and mutation operations.

[0019] Optionally, the processing unit is further used to update the taboo table according to the result of each operation during the execution of the selection operation, the crossover operation and the mutation operation.

[0020] Optionally, the scheduling information includes at least one of the following: the task success rate of the selected scheduling scheme, the average delay of the task to be scheduled, the resource utilization of the task processing equipment, the load rate of the task processing equipment, the bandwidth utilization of the network link, and the processing unit is specifically used to: determine the fitness as the sum of the products of each scheduling information and the weight factor of the scheduling information.

[0021] In a third aspect, a scheduling scheme determination device is provided, comprising a memory and a processor; the memory is used to store computer execution instructions, and the processor is connected to the memory through a bus; when the scheduling scheme determination device is running, the processor executes the computer execution instructions stored in the memory, so that the scheduling scheme determination device executes the scheduling scheme determination method of the first aspect.

[0022] The scheduling scheme determination device may be an electronic device, or a part of an electronic device, such as a chip system in an electronic device. The chip system is used to support the electronic device to implement the functions involved in the first aspect and any possible implementation thereof, for example, to obtain and determine the data and / or information involved in the above-mentioned scheduling scheme determination method. The chip system includes a chip, and may also include other discrete devices or circuit structures.

[0023] In a fourth aspect, a computer-readable storage medium is provided, the computer-readable storage medium comprising computer execution instructions, and when the computer execution instructions are executed on a computer, the computer executes the scheduling scheme determination method described in the first aspect.

[0024] In a fifth aspect, a computer program product is also provided, which includes a computer program or instructions. When the computer instructions are executed on a scheduling scheme determination device, the scheduling scheme determination device executes the scheduling scheme determination method as described in the first aspect above.

[0025] It should be noted that the above-mentioned computer instructions may be stored in whole or in part on a computer-readable storage medium. The computer-readable storage medium may be packaged together with the processor of the scheduling scheme determination device, or may be packaged separately from the processor of the scheduling scheme determination device, which is not limited in the embodiments of the present application.

[0026] The description of the second, third, fourth and fifth aspects of the present application can refer to the detailed description of the first aspect.

[0027] In the embodiments of the present application, the name of the above-mentioned scheduling scheme determination device does not limit the device or functional module itself. In actual implementation, these devices or functional modules may appear with other names. For example, the receiving unit may also be called a receiving module, a receiver, etc. As long as the functions of each device or functional module are similar to those of the present application, they belong to the scope of the claims of the present application and their equivalent technologies.

[0028] The technical solution provided by this application brings at least the following beneficial effects:

[0029] Based on any of the above aspects, an embodiment of the present application provides a method for determining a scheduling scheme. First, the task information of the task to be scheduled, the task processing delay of the task processing device, and the transmission delay of the network link are obtained. Then, based on the task information, the task processing delay and the transmission delay, multiple scheduling schemes to be selected are determined. Next, based on the multiple scheduling schemes to be selected, the taboo search algorithm and the genetic algorithm, the target scheduling scheme is determined. Among them, the task processing delay is used to indicate the time length for the task processing device to process the task to be scheduled, and the transmission delay is used to indicate the time length for transmitting the task to be scheduled to the task processing device.

[0030] As can be seen from the above, the present application can obtain the task information of the task to be scheduled, the task processing delay of the task processing device, and the transmission delay of the network link. And based on the task information, task processing delay and transmission delay, multiple scheduling schemes to be selected are determined. Among them, the task processing delay is used to indicate the duration of the task processing device processing the task to be scheduled, and the transmission delay is used to indicate the duration of transmitting the task to be scheduled to the task processing device, that is, the influence of the network layer. Therefore, the present application takes into account the influence of the network layer when determining the scheduling scheme, so that the determination of the scheduling scheme to be selected is more accurate. The present application can also determine the target scheduling scheme based on multiple scheduling schemes to be selected, taboo search algorithm and genetic algorithm. In this way, the scheduling schemes to be selected are further screened to obtain the final target scheduling scheme.

[0031] The beneficial effects of the first, second, third, fourth and fifth aspects of the present application can all be referred to the analysis of the above beneficial effects, which will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 A schematic diagram of an architecture of a cloud-network-edge-end architecture provided for an embodiment of the present application;

[0033] Figure 2 A schematic diagram of the structure of a scheduling scheme determination system provided in an embodiment of the present application;

[0034] Figure 3 A schematic diagram of a hardware structure of a communication device provided in an embodiment of the present application;

[0035] Figure 4 A flowchart of a method for determining a scheduling scheme provided in an embodiment of the present application;

[0036] Figure 5 A flowchart of another method for determining a scheduling scheme provided in an embodiment of the present application;

[0037] Figure 6 A flowchart of another method for determining a scheduling scheme provided in an embodiment of the present application;

[0038] Figure 7 A flowchart of another method for determining a scheduling scheme provided in an embodiment of the present application;

[0039] Figure 8 A flowchart of another method for determining a scheduling scheme provided in an embodiment of the present application;

[0040] Fig. 9 A flowchart of another method for determining a scheduling scheme provided in an embodiment of the present application;

[0041] Fig.10 A flowchart of task scheduling in a cloud-network-edge-end architecture provided in an embodiment of the present application;

[0042] Fig.11 A schematic diagram of the structure of a scheduling scheme determination device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0044] It should be noted that, in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0045] Before introducing the electric energy distribution control method provided by the present application in detail, a brief introduction is first given to the relevant elements, application scenarios, and implementation environment involved in the present application.

[0046] First, a brief introduction to the relevant elements involved in this application is given.

[0047] The cloud-network-edge architecture is a new network ecosystem that focuses on efficiently processing data and smoothly carrying application service processes.

[0048] The cloud-network-edge-end architecture is mainly composed of four key parts: cloud, network, edge, and end.

[0049] Among them, the cloud, namely cloud computing or cloud data center, is responsible for providing massive computing resources and sufficient storage resources, such as servers, storage devices, databases, network facilities and various software. The network, namely the network, is the key bridge for the interconnection between the cloud, edge and end. It is responsible for accurately transmitting the massive data and various services of the cloud data center to the edge data center and the terminal, so as to ensure the transmission of data and information in the entire architecture. The edge, namely edge computing or edge cloud data center. The edge uses a distributed computing model to process the massive data from the terminal device in the first time, respond to requests quickly, and reduce transmission delays and transmission bandwidth occupancy by pre-placing computing, storage and network services close to the data source or user. The end, namely the terminal device, includes smartphones, tablets and personal computers. These terminal devices are windows for users to directly interact with cloud services. They are not only responsible for collecting data, but also used to present the functions of various application services to users, so as to achieve seamless connection between users and cloud services.

[0050] For example, Figure 1 A schematic diagram of the architecture of a cloud-network-edge architecture is shown.

[0051] Among them, cloud data centers and edge data centers are used to provide computing resources and storage resources. Terminals are used to interact with users. The network is used to connect cloud data centers, edge cloud data centers and terminals.

[0052] Next, the application scenarios involved in this application are briefly introduced.

[0053] With the development of technologies such as the Internet of Things, big data, and artificial intelligence, the amount of data has exploded. Efficient processing of large amounts of data, real-time response, and resource allocation are key issues that need to be addressed. Therefore, it is necessary to coordinate the scheduling of resources across the cloud, network, edge, and end.

[0054] General technologies usually focus on resource integration and scheduling of cloud, edge and end.

[0055] In the cloud-network-edge integrated architecture, the user scale and network scale are huge, and the types and number of tasks are numerous, which makes the number of scheduling schemes very large (it can also be called a very large solution space). In this architecture, there will be conflicts between multiple demands such as user needs and operator needs. General technology needs to comprehensively consider the resource constraints of cloud data centers and edge data centers, as well as the latency requirements of the selected network links. General technology dynamically adjusts the scheduling scheme according to real-time information during operation by real-time monitoring and analysis of information such as task requirements and resource status to adapt to different application scenarios and load changes and reduce resource consumption. However, general technology needs to be optimized in a targeted manner to adapt to specific business scenarios, lacks cross-scenario adaptability and robustness, and is difficult to achieve general scheduling for all scenarios. In addition, general technology can usually only be optimized for a single demand, lacking overall comprehensive consideration.

[0056] Therefore, the network layer is an important link between cloud, edge and end. The general technology ignores the influence of the network layer when determining the scheduling scheme. Therefore, in a complex network environment, the scheduling scheme determined based on general technology lacks accuracy.

[0057] In response to the above problems, an embodiment of the present application provides a method for determining a scheduling scheme. First, the task information of the task to be scheduled, the task processing delay of the task processing device, and the transmission delay of the network link are obtained. Then, based on the task information, the task processing delay and the transmission delay, multiple scheduling schemes to be selected are determined. Next, based on the multiple scheduling schemes to be selected, the taboo search algorithm and the genetic algorithm, the target scheduling scheme is determined. Among them, the task processing delay is used to indicate the time length for the task processing device to process the task to be scheduled, and the transmission delay is used to indicate the time length for transmitting the task to be scheduled to the task processing device.

[0058] As can be seen from the above, the present application can obtain the task information of the task to be scheduled, the task processing delay of the task processing device, and the transmission delay of the network link. And based on the task information, task processing delay and transmission delay, multiple scheduling schemes to be selected are determined. Among them, the task processing delay is used to indicate the duration of the task processing device processing the task to be scheduled, and the transmission delay is used to indicate the duration of transmitting the task to be scheduled to the task processing device, that is, the influence of the network layer. Therefore, the present application takes into account the influence of the network layer when determining the scheduling scheme, so that the determination of the scheduling scheme to be selected is more accurate. The present application can also determine the target scheduling scheme based on multiple scheduling schemes to be selected, taboo search algorithm and genetic algorithm. In this way, the scheduling schemes to be selected are further screened to obtain the final target scheduling scheme.

[0059] The implementation environment of the above-mentioned scheduling scheme determination method can be the scheduling scheme determination system provided in the embodiment of the present application.

[0060] Figure 2FIG. 1 shows a schematic diagram of a system for determining a scheduling scheme provided by an embodiment of the present application. Figure 2 As shown, the scheduling scheme determination system includes: a scheduling scheme determination device 201, a task processing device 202, a user device 203, and a communication device 204.

[0061] The scheduling scheme determining device 201 is respectively connected to the task processing device 202 , the user device 203 , and the communication device 204 for communication.

[0062] In practical applications, the scheduling scheme determination device 201 can be connected to any number of task processing devices 202. For ease of understanding, Figure 2 An example is given for describing the connection between a scheduling scheme determination device 201 and a task processing device 202 .

[0063] In practical applications, the scheduling scheme determining device 201 can be connected to any number of user devices 203. For ease of understanding, Figure 2 An example is taken for describing that a scheduling scheme determines that a device 201 is connected to a user equipment 203 .

[0064] In practical applications, the scheduling scheme determination device 201 can be connected to any number of communication devices 204. For ease of understanding, Figure 2 An example is given for describing the connection between a scheduling scheme determination device 201 and a communication device 204.

[0065] In the embodiment of the present application, the task processing device 202 may be a cloud data center and an edge data center in the above-mentioned cloud-network-edge-end architecture. The user device 203 may be a terminal in the above-mentioned cloud-network-edge-end architecture.

[0066] In an embodiment of the present application, the task processing device 202 is used to provide the scheduling scheme determination device 201 with data for scheduling scheme determination (for example, the task processing delay of the task processing device), the user device 203 is used to provide the scheduling scheme determination device 201 with data for scheduling scheme determination (for example, task information of the task to be scheduled), and the communication device 204 is used to provide the scheduling scheme determination device 201 with data for scheduling scheme determination (for example, transmission delay), so that the scheduling scheme determination device 201 implements scheduling scheme determination according to the data sent by the task processing device 202, the user device 203, and the communication device 204.

[0067] Optionally, the entity device of the scheduling scheme determination device 201 may be a server, a terminal, or other types of electronic devices, which is not limited in the embodiments of the present application.

[0068] Optionally, the terminal may be a device that provides voice and / or data connectivity to a user, a handheld device with wireless connection function, or other processing device connected to a wireless modem. The wireless terminal may communicate with one or more core networks via a radio access network (RAN). The wireless terminal may be a mobile terminal, such as a mobile phone (or "cellular" phone) and a computer with a mobile terminal, or a portable, pocket-sized, handheld, computer-built-in or vehicle-mounted mobile device that exchanges language and / or data with a radio access network, such as a mobile phone, tablet computer, laptop computer, netbook, personal digital assistant (PDA).

[0069] Optionally, the above-mentioned server can be a server in a server cluster (consisting of multiple servers), or a chip in the server, or a system on a chip in the server, or can be implemented by a virtual machine (VM) deployed on a physical machine, which is not limited in the embodiments of the present application.

[0070] Optionally, the scheduling scheme determining device 201 and the communication device 204 may be two devices independently set up, or may be integrated into the same device. When the scheduling scheme determining device 201 and the communication device 204 are integrated into the same device, the communication device 204 may be a communication module of the scheduling scheme determining device 201.

[0071] It is easy to understand that when the scheduling scheme determining device 201 and the communication device 204 are integrated into the same device, the communication mode between the scheduling scheme determining device 201 and the communication device 204 is the communication between the internal modules of the device. In this case, the communication process between the two is the same as the communication process between the scheduling scheme determining device 201 and the communication device 204 when they are independent of each other.

[0072] For ease of understanding, the present application is described by taking the scheduling scheme determination device 201 and the communication device 204 as an example in which they are independent of each other.

[0073] The scheduling scheme determination device in the scheduling scheme determination system includes: Figure 3 The following are the components included. Figure 3 Taking the communication device shown as an example, the hardware structure of the scheduling scheme determination device is introduced.

[0074] Figure 3, which is a schematic diagram of a hardware structure of a communication device provided in an embodiment of the present application. The communication device includes a processor 31, a memory 32, a communication interface 33, and a bus 34. The processor 31, the memory 32, and the communication interface 33 can be connected via a bus 34.

[0075] The processor 31 is the control center of the communication device, which can be a processor or a general term for multiple processing elements. For example, the processor 31 can be a general-purpose central processing unit (CPU) or other general-purpose processors. Among them, the general-purpose processor can be a microprocessor or any conventional processor.

[0076] As an embodiment, the processor 31 may include one or more CPUs, such as Figure 3 CPU0 and CPU1 are shown in the figure.

[0077] The memory 32 may be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM) or other type of dynamic storage device that can store information and instructions, an electrically erasable programmable read-only memory (EEPROM), a disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.

[0078] In a possible implementation, the memory 32 may exist independently of the processor 31, and the memory 32 may be connected to the processor 31 via a bus 34 to store instructions or program codes. When the processor 31 calls and executes the instructions or program codes stored in the memory 32, the scheduling scheme determination method provided in the following embodiments of the present application can be implemented.

[0079] In the embodiment of the present application, for the communication device, the software programs stored in the memory 32 are different, so the functions implemented by the communication device are different. The functions performed by each device will be described in conjunction with the following flowchart.

[0080] In another possible implementation, the memory 32 may also be integrated with the processor 31 .

[0081] The communication interface 33 is used for the communication device to connect with other devices through a communication network, and the communication network may be Ethernet, wireless access network, wireless local area network (WLAN), etc. The communication interface 33 may include a receiving unit for receiving data and a sending unit for sending data.

[0082] The bus 34 may be an industry standard architecture (ISA) bus, a peripheral component interconnect (PCI) bus, or an extended industry standard architecture (EISA) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0083] It should be pointed out that Figure 3 The structure shown in the figure does not constitute a limitation on the communication device, except Figure 3 In addition to the components shown, the communication device may include more or fewer components than shown, or combine certain components, or arrange the components differently.

[0084] The following is a detailed introduction to the scheduling scheme determination method provided in the embodiment of the present application in conjunction with the accompanying drawings.

[0085] The scheduling scheme determination method provided in the embodiment of the present application is applied to Figure 2 The scheduling scheme determination device 201 in the scheduling scheme determination system shown in FIG. Figure 4 As shown, the scheduling scheme determination method provided in the embodiment of the present application includes:

[0086] S401, obtaining task information of a task to be scheduled, a task processing delay of a task processing device, and a transmission delay of a network link.

[0087] Among them, the task processing delay is used to indicate the time it takes for the task processing device to process the task to be scheduled, and the transmission delay is used to indicate the time it takes to transmit the task to be scheduled to the task processing device.

[0088] In the embodiment of the present application, the scheduling scheme determination device can obtain the task information of the task to be scheduled, the task processing delay of the task processing device, and the transmission delay of the network link. In this way, the scheduling scheme determination device can determine multiple scheduling schemes to be selected based on the task information, task processing delay and transmission delay.

[0089] The task information includes: the maximum tolerable delay of the task to be scheduled, that is, the time length that the task to be scheduled can accept to obtain the processing result.

[0090] It should be understood that there are multiple network links between the user equipment and the task processing device. The transmission delay is used to indicate the time length for the user equipment to transmit the task to be scheduled to the task processing device.

[0091] S402: Determine multiple scheduling schemes to be selected according to task information, task processing delay and transmission delay.

[0092] In the embodiment of the present application, the task processing delay is the time it takes for the task processing device to process the task to be scheduled, and the task transmission delay is the time it takes to transmit the task to be scheduled to the task processing device. In this way, the scheduling scheme determination device can comprehensively consider the task information of the tasks to be scheduled on the "end", the task processing delay of the task processing devices on the "cloud" and "edge", and the transmission delay of the network link of the "net" to determine multiple scheduling schemes to be selected. Therefore, when determining the scheduling scheme to be selected, the scheduling scheme determination device comprehensively considers the cloud, network, edge, and end, so that the scheduling scheme can be determined more accurately.

[0093] In some embodiments, the scheduling scheme to be selected may be regarded as a scheduling scheme individual (referred to as individual), and multiple scheduling schemes to be selected may be regarded as a scheduling scheme population (referred to as population), that is, a population includes several individuals.

[0094] Optionally, a to-be-selected scheduling scheme represents a task processing device selection scheme for multiple to-be-scheduled tasks, wherein a to-be-scheduled task can only be assigned to one task processing device.

[0095] For example, there are three tasks to be scheduled (task 1, task 2, and task 3), and two task processing devices (device A and device B). 3 There are eight scheduling schemes to be selected. Among them, the scheduling scheme a: Task 1, Task 2, and Task 3 are all assigned to device A. The scheduling scheme b: Task 1 is assigned to device A, and Task 2 and Task 3 are all assigned to device B. The scheduling scheme c: Task 1 and Task 2 are all assigned to device A, and Task 3 is assigned to device B. The scheduling scheme d: Task 1 and Task 3 are all assigned to device A, and Task 2 is assigned to device B. The scheduling scheme e: Task 1, Task 2, and Task 3 are all assigned to device B. The scheduling scheme f: Task 1 is assigned to device B, and Task 2 and Task 3 are all assigned to device A. The scheduling scheme g: Task 1 and Task 2 are all assigned to device B, and Task 3 is assigned to device A. The scheduling scheme h: Task 1 and Task 3 are all assigned to device B, and Task 2 is assigned to device A.

[0096] S403: Determine a target scheduling solution based on multiple candidate scheduling solutions, a taboo search algorithm and a genetic algorithm.

[0097] In the embodiment of the present application, the scheduling scheme determination device is based on a genetic algorithm, and can operate (selection operation, cross operation and mutation operation) on multiple candidate scheduling schemes at the same time, and after iteration, determine the optimal target scheduling scheme or a near-optimal target scheduling scheme. The scheduling scheme determination device is based on a taboo search algorithm, which can avoid repeated operations in the process of determining the target scheduling scheme, so that the obtained target scheduling scheme is the optimal scheduling scheme among all scheduling schemes (also called the global optimal solution), rather than the optimal scheduling scheme among some scheduling schemes (also called the local optimal solution). In this way, the scheduling scheme determination device can be based on multiple candidate scheduling schemes, taboo search algorithms and genetic algorithms to obtain the optimal target scheduling scheme.

[0098] The taboo search algorithm is a heuristic search algorithm based on neighborhood search. The algorithm can select a series of specific search directions, and further select the search direction that can achieve the most changes in the objective function value to avoid falling into the local optimal solution. The taboo search algorithm can also record and select the optimization process that has been carried out to guide the direction of the next search.

[0099] Genetic algorithm is a method to search for the optimal solution by simulating the natural evolution process. Through a series of selection operations, crossover operations and mutation operations, the algorithm can quickly obtain better optimization results.

[0100] The scheduling scheme determination device can combine the taboo search algorithm and the genetic algorithm to determine the optimized target scheduling scheme based on multiple candidate scheduling schemes.

[0101] Specifically, first, the scheduling scheme determination device can perform a global search for multiple scheduling schemes based on a genetic algorithm. Then, the scheduling scheme determination device improves the operation process involved in the genetic algorithm through a taboo search algorithm (which can also be a neighborhood search). In this way, the scheduling scheme determination device can obtain a target scheduling scheme in the process of iterating the above operations.

[0102] In some embodiments, the task information includes: the maximum tolerable delay for processing the task to be scheduled, combined with Figure 4 ,like Figure 5 As shown, in S402, multiple scheduling schemes to be selected are determined according to the task information, task processing delay and transmission delay, including:

[0103] S501. For each task to be scheduled and each task processing device, when the sum of the transmission delay of the task to be scheduled and the task processing delay of the task processing device is less than or equal to the maximum tolerable delay of the task to be scheduled, determine multiple scheduling schemes to be selected according to the number of tasks to be scheduled and the number of task processing devices.

[0104] In the embodiment of the present application, the task to be scheduled can be successfully scheduled only when the sum of the transmission delay of the task to be scheduled and the task processing delay of the task processing device is less than or equal to the maximum tolerable delay of the task to be scheduled. Therefore, in the process of determining multiple scheduling schemes to be selected based on the task information, task processing delay and transmission delay, the scheduling scheme determination device should ensure that each scheduling scheme to be selected meets the scheduling conditions, that is, the sum of the transmission delay of the task to be scheduled and the task processing delay of the task processing device is less than or equal to the maximum tolerable delay of the task to be scheduled. The number of scheduling schemes to be selected is closely related to the number of tasks to be scheduled and the number of task processing devices. Therefore, the scheduling scheme determination device can determine multiple scheduling schemes to be selected based on the number of tasks to be scheduled and the number of task processing devices.

[0105] Since the scheduling scheme determination device can determine multiple candidate scheduling schemes based on task information, task processing delay and transmission delay. Moreover, the fitness of multiple candidate scheduling schemes (also known as the quality of the candidate scheduling schemes), the number of candidate scheduling schemes (also known as the size of the population) and the diversity of the candidate scheduling schemes will affect the efficiency of the scheduling scheme determination device in determining the target scheduling scheme (also known as the convergence speed performance from multiple candidate scheduling schemes to the target scheduling scheme). Therefore, for each task to be scheduled and each task processing device, when the sum of the transmission delay of the task to be scheduled and the task processing delay of the task processing device is less than or equal to the maximum tolerable delay of the task to be scheduled, the scheduling scheme determination device can determine multiple candidate scheduling schemes based on the number of tasks to be scheduled and the number of task processing devices.

[0106] In one achievable manner, the scheduling scheme determination device selects the network link for transmitting the task to be scheduled and the task processing device for processing the task to be scheduled according to the maximum tolerable delay of the task to be scheduled. Moreover, the sum of the transmission delay of the task to be scheduled and the delay of the task processing device should be less than or equal to the maximum tolerable delay of the task to be scheduled. In this way, when the scheduling scheme determination device determines the network link of one of the tasks to be scheduled, it can screen out all the network links that meet the above scheduling conditions, and select one of them with the minimum delay or the most sufficient bandwidth. The scheduling scheme determination device can also combine the network link of a task to be scheduled with the task processing device, and obtain scheduling sub-schemes for multiple tasks to be scheduled, so as to obtain a complete scheduling scheme to be selected.

[0107] In another achievable manner, when the task to be scheduled is a delay-sensitive task, the scheduling scheme determination device determines that the task processing device that processes the scheduled task is a task processing device that is relatively close to the task to be scheduled. When the task to be scheduled is a task with high delay tolerance, the scheduling scheme determination device determines that the task processing device that processes the scheduled task is a task processing device that is relatively far from the task to be scheduled.

[0108] By repeating the above process (also called heuristic initialization (scheduling scheme) population), a scheduling scheme population containing multiple scheduling schemes to be selected can be obtained. All scheduling schemes in the scheduling scheme population meet the above scheduling conditions (also called delay constraints). In this way, the scheduling scheme determination device can reduce the amount of calculation for determining the target scheduling scheme and the time for determining the target scheduling scheme (also called convergence time) based on this.

[0109] From the above, it can be seen that the present application adopts the above-mentioned scheme to determine multiple candidate scheduling schemes, which can be generated according to delay constraints or resource constraints, so as to reduce the proportion of invalid scheduling schemes, improve the quality of candidate scheduling schemes, and accelerate the determination of the target scheduling scheme.

[0110] In some embodiments, the task information includes: the maximum tolerable delay for processing the task to be scheduled, combined with Figure 5 ,like Figure 6 As shown, in S403, based on multiple candidate scheduling schemes, taboo search algorithm and genetic algorithm, a target scheduling scheme is determined, specifically including:

[0111] S601: Determine the fitness of multiple scheduling schemes to be selected according to the scheduling information of the scheduling scheme to be scheduled.

[0112] In the embodiment of the present application, the fitness is used to characterize the effectiveness of the candidate scheduling scheme, and further, the fitness can be used to characterize the scheduling efficiency of the processing task under the candidate scheduling scheme. The scheduling scheme determination device can determine the fitness of multiple candidate scheduling schemes based on the scheduling information of the candidate scheduling schemes (for example, the task success rate of the candidate scheduling schemes, the average delay of the tasks to be scheduled).

[0113] In one achievable manner, the scheduling information includes: the task success rate of the selected scheduling scheme, the average delay of the scheduled tasks, the resource utilization rate of the task processing equipment, the load rate of the task processing equipment, and the bandwidth utilization rate of the network link. The scheduling scheme determination device can directly determine the fitness of the selected scheduling scheme based on the above scheduling information.

[0114] In another achievable manner, the scheduling information includes: whether the i-th task to be scheduled is successfully assigned to the j-th task processing device, the total number of tasks to be scheduled, the resource consumption of the i-th task to be scheduled in the j-th task processing device, the total resource amount of the j-th task processing device, the network link delay selected by the i-th task to be scheduled, the resource amount required by the i-th task to be scheduled, the calculation speed of the i-th task to be scheduled in the task processing device, the network link bandwidth occupied by the j-th task processing device of the i-th task to be scheduled, and the sum of all network link bandwidths that can be transmitted to the j-th task processing device. The scheduling scheme determination device can determine the task success rate of the selected scheduling scheme, the average delay of the tasks to be scheduled, the resource utilization rate of the task processing device, the load rate of the task processing device, and the bandwidth utilization rate of the network link based on the above scheduling information. And further determine the adaptability of the selected scheduling scheme based on the task success rate of the selected scheduling scheme, the average delay of the tasks to be scheduled, the resource utilization rate of the task processing device, the load rate of the task processing device, the bandwidth utilization rate of the network link, and other information.

[0115] Optionally, the scheduling scheme determination device can abstract scheduling information such as the task success rate of the selected scheduling scheme, the average delay of the tasks to be scheduled, the resource utilization of the task processing equipment, the load rate of the task processing equipment, and the bandwidth utilization of the network link into an objective function that needs to be optimized.

[0116] Maximizing the task success rate of the candidate scheduling solution (also known as maximizing the overall task success rate) satisfies the following formula:

[0117]

[0118] Among them, x i,j The decision variable used to indicate whether the i-th task to be scheduled is successfully assigned to the j-th task processing device, and N is used to indicate the total number of tasks to be scheduled.

[0119] In one achievable way, That is, a task to be scheduled can only be assigned to one task processing device.

[0120] When the i-th task to be scheduled is successfully assigned to the j-th task processing device, the decision variable x i,j The value of is 1. When the i-th task to be scheduled is not successfully assigned to the j-th task processing device, the decision variable x i,j The value of is 0.

[0121] Whether the task to be scheduled is successfully allocated depends on whether the scheduling scheme determines whether the device allocates the required task processing equipment (also called computing resources, storage resources), network links (also called network resources) for the task to be scheduled, and whether the time taken for the task to be scheduled from initiation to completion is less than its maximum tolerable delay.

[0122] When the scheduling scheme determines that the device allocates the required task processing equipment and network links to the task to be scheduled, and the time taken from initiation to completion of the task to be scheduled is less than its maximum tolerable delay, the scheduling scheme determines that the task to be scheduled is successfully allocated to the task processing equipment.

[0123] Correspondingly, when the scheduling scheme determines that the device does not allocate the required task processing equipment and network link for the task to be scheduled, or the time taken for the task to be scheduled to be completed from initiation to completion is greater than its maximum tolerated delay, the scheduling scheme determines that the task to be scheduled is not successfully allocated to the task processing equipment.

[0124] Maximizing the resource utilization of task processing equipment (also known as maximizing the resource utilization of cloud data centers) satisfies the following formula:

[0125]

[0126] Among them, r i,j It is used to represent the resource consumption of the i-th task to be scheduled on the j-th task processing device, R j Used to represent the total amount of resources of the j-th task processing device.

[0127] In one achievable way, That is, for any task to be scheduled and any task processing device, the resource consumption on a task processing device cannot exceed the total resource consumption of the task processing device.

[0128] Minimizing the average delay of the scheduled tasks (also known as minimizing the average delay on the basis of ensuring the success rate of the tasks) satisfies the following formula:

[0129]

[0130] Among them, t link(i) It is used to represent the network link delay selected by the i-th task to be scheduled, C i It is used to represent the amount of resources required by the i-th task to be scheduled, C pi Used to indicate the computing speed of the i-th task to be scheduled on the task processing device.

[0131] Maximizing the bandwidth utilization of a network link (also known as maximizing bandwidth utilization) satisfies the following formula:

[0132]

[0133] Among them, b i,j It is used to represent the network link bandwidth occupied by the jth task processing device of the i-th task to be scheduled, B jIt is used to represent the sum of all network link bandwidths that can be transmitted to the j-th task processing device.

[0134] In one achievable way, That is, for any task processing device, the bandwidth occupied on the network link does not exceed the sum of the bandwidths of all network links that can be transmitted to the j-th task processing device.

[0135] Minimizing the load rate of the task processing equipment (also known as minimizing the cloud data center load) satisfies the following formula:

[0136]

[0137] Among them, x i,j The decision variable used to indicate whether the i-th scheduled task is successfully assigned to the j-th task processing device, r i,j It is used to represent the resource consumption of the i-th task to be scheduled on the j-th task processing device, R j Used to indicate the total amount of resources of the j-th task processing device.

[0138] S602: Update the taboo table of the taboo search algorithm according to the fitness.

[0139] Based on the above embodiments, it can be seen that fitness can be used to characterize the effectiveness of the candidate scheduling scheme. In the embodiment of the present application, the taboo table is used to store (part of) the candidate scheduling scheme. It should be understood that when the target scheduling scheme is subsequently determined based on multiple candidate scheduling schemes, the scheduling scheme determination device can eliminate part of the candidate scheduling schemes from the multiple candidate scheduling schemes according to the taboo table. Therefore, the scheduling scheme determination device can update the taboo table of the taboo search algorithm according to the fitness.

[0140] The taboo table includes the global taboo table and the local taboo table.

[0141] In one achievable method, the scheduling scheme determination device can store the candidate scheduling scheme with the highest fitness (also called the global optimal solution) and the candidate scheduling scheme with the lowest fitness (also called the global worst solution) among multiple candidate scheduling schemes into a global taboo table to update the global taboo table of the taboo search algorithm.

[0142] S603: Determine a target scheduling solution based on fitness, multiple candidate scheduling solutions, a taboo table and a genetic algorithm.

[0143] Based on the above embodiments, it can be seen that the scheduling scheme determination device is based on a genetic algorithm and can operate on multiple candidate scheduling schemes at the same time, and after iteration, determine the optimal target scheduling scheme or a target scheduling scheme close to the optimal one. The scheduling scheme determination device is based on a taboo table, which can avoid repeated operations in the process of determining the target scheduling scheme, so that the scheduling scheme determination device can obtain the final target scheduling scheme based on multiple candidate scheduling schemes.

[0144] In some embodiments, the genetic algorithm includes: a selection operation, a crossover operation, and a mutation operation, combined with Figure 6 ,like Figure 7 As shown, in S603, based on fitness, multiple candidate scheduling solutions, taboo table and genetic algorithm, a target scheduling solution is determined, including:

[0145] S701. Perform a selection operation on multiple candidate scheduling schemes based on a taboo table and fitness to obtain at least one group of candidate scheduling schemes.

[0146] Each group of candidate scheduling schemes includes 2 candidate scheduling schemes, and each group of candidate scheduling schemes does not include the candidate scheduling schemes in the taboo table.

[0147] In the embodiment of the present application, the selection operation in the genetic algorithm can be understood as selecting a better scheduling scheme from multiple scheduling schemes to be selected. Therefore, the present application can eliminate some scheduling schemes in the current scheduling schemes to be selected according to the taboo table, and select the scheduling schemes to be selected with higher fitness according to the fitness, so as to obtain at least one group of scheduling schemes to be selected.

[0148] The selection operation is used to select some better candidate scheduling schemes from multiple candidate scheduling schemes, so as to save and disseminate the better candidate scheduling schemes, so that the scheduling scheme determination device gradually determines the target scheduling scheme.

[0149] In one practicable manner, the scheduling scheme determining device may apply a roulette wheel selection method to perform a selection operation on a plurality of candidate scheduling schemes to obtain at least one group of candidate scheduling schemes.

[0150] Among them, the probability of the mth candidate scheduling scheme being selected satisfies the following formula:

[0151]

[0152] Among them, M represents the number of scheduling options to be selected, Fitness Cm Used to indicate the probability of the mth candidate scheduling scheme being selected.

[0153] S702: Perform a crossover operation on each group of candidate scheduling solutions based on the taboo table to obtain multiple candidate scheduling solutions after the crossover operation.

[0154] In the embodiment of the present application, the crossover operation in the genetic algorithm can be understood as exchanging and combining some sub-schedules in two candidate scheduling schemes to generate a new candidate scheduling scheme. Therefore, the present application can perform a crossover operation on each group of candidate scheduling schemes, and based on the taboo table, eliminate individual schemes in the new candidate scheduling schemes (i.e., the same schemes as the candidate scheduling schemes in the taboo table). In this way, the scheduling scheme determination device can increase the diversity of the candidate scheduling schemes and avoid repeated operations on the candidate scheduling schemes in the taboo table.

[0155] The crossover operation is used to exchange and combine the scheduling schemes of individual tasks to be scheduled in a set of candidate scheduling schemes to obtain a new candidate scheduling scheme.

[0156] Specifically, first, the scheduling scheme determination device performs a crossover operation (for example, single-point crossover, two-point crossover, uniform crossover) according to a traditional genetic algorithm to obtain a plurality of candidate scheduling schemes after the crossover operation (also referred to as offspring of each group of scheduling schemes). Then, the scheduling scheme determination device can check whether there are candidate scheduling methods in the local taboo table among the plurality of candidate scheduling schemes after the crossover operation based on the taboo search algorithm. If so, the scheduling scheme determination device re-executes the crossover operation on each group of candidate scheduling schemes. If not, the scheduling scheme determination device can add the plurality of candidate scheduling schemes after the crossover operation to the local promotion table.

[0157] Optionally, the scheduling scheme determination device may also perform a neighborhood search on multiple candidate scheduling schemes after the cross operation processing.

[0158] Specifically, the scheduling scheme determination device can adjust the network link for transmitting the task to be scheduled, change the task processing device for processing the task to be scheduled, and migrate the task to be scheduled. At the same time, the scheduling scheme determination device needs to ensure that the scheduling scheme after the above operations meets the scheduling requirements, as well as the allocation of task processing device resources and network link resources. In this way, the scheduling scheme determination device can obtain a better candidate scheduling scheme than the current candidate scheduling scheme, so as to obtain a candidate scheduling scheme that is more inclined to the global optimality.

[0159] Optionally, the scheduling scheme determination device can prioritize based on historical information or prior knowledge, and select network links that can generate high-quality candidate scheduling schemes to a certain extent. That is, the scheduling scheme determination device can prioritize adjusting network links with smaller transmission delays, or prioritize optimizing the scheduled tasks with the smallest maximum tolerable delay or delay-sensitive scheduled tasks.

[0160] Among them, the scheduling scheme determination device can adjust the network link for transmitting the task to be scheduled, that is, the scheduling scheme determination device can change the network link for transmitting the task to be scheduled.

[0161] The scheduling scheme determining device can change the task processing device that processes the task to be scheduled, that is, the scheduling scheme determining device can replace the task processing device that processes any task to be scheduled in the selected scheduling scheme with another task processing device.

[0162] The scheduling scheme determines that the device can migrate the tasks to be scheduled. The scheduling scheme determines that the device can migrate part of the tasks to be scheduled from the task processing device with overloaded resources to another idle task processing device.

[0163] Optionally, the scheduling scheme determining device may dynamically adjust the size of the neighborhood according to the quality of the scheduling scheme to be selected.

[0164] In one practicable manner, in the initial stage, the scheduling solution determination device may select a larger neighborhood for searching, and as the genetic algorithm and the taboo search algorithm proceed, the scheduling solution determination device may gradually narrow the neighborhood to focus on searching for candidate scheduling solutions in a larger area.

[0165] In another possible implementation, the scheduling scheme determines that when the device is performing a neighborhood search, it can first adjust the network link that transmits the task to be scheduled (i.e., perform a smaller local operation), and then change the task processing device that processes the task to be scheduled, or migrate the task to be scheduled (i.e., perform a larger change operation).

[0166] S703 , performing a mutation operation on the multiple candidate scheduling solutions after the crossover operation based on the taboo table to obtain the candidate scheduling solutions after the mutation operation.

[0167] In the embodiment of the present application, the mutation operation in the genetic algorithm can be understood as fine-tuning the candidate scheduling scheme to generate a new candidate scheduling scheme. Therefore, the present application can perform a mutation operation on multiple candidate scheduling schemes after the crossover operation, and based on the taboo table, eliminate individual schemes in the new candidate scheduling scheme (i.e., the same scheme as the candidate scheduling scheme in the taboo table). In this way, the scheduling scheme determination device can avoid the similarity of multiple candidate scheduling schemes, increase the diversity of the candidate scheduling schemes, and avoid repeated operations on the candidate scheduling schemes in the taboo table.

[0168] The mutation operation is used to change some of the candidate scheduling schemes, that is, to change the scheduling schemes of individual tasks to be scheduled in the candidate scheduling schemes.

[0169] In one achievable manner, the scheduling scheme determination device may pre-acquire the mutation rates of multiple candidate scheduling schemes in this round, and perform a mutation operation after the scheduling scheme determination device performs a crossover operation on the multiple candidate scheduling schemes to obtain the candidate scheduling schemes after the mutation operation.

[0170] The scheduling scheme determination device checks the candidate scheduling scheme obtained after the mutation operation based on the taboo table. If the candidate scheduling scheme obtained after the mutation operation is present in the taboo table, the mutation operation is re-executed. If the candidate scheduling scheme obtained after the mutation operation is not present in the taboo table, the scheduling scheme determination device may store the candidate scheduling scheme obtained after the mutation operation into the local taboo table to update the local taboo table.

[0171] S704 , repeat the selection operation, the crossover operation, and the mutation operation until the convergence condition is met, and determine the candidate scheduling solution after the mutation operation as the target scheduling solution.

[0172] The convergence conditions include: the number of candidate scheduling solutions after the mutation operation is less than the preset number of solutions, or the number of mutation operations performed is greater than the preset number of operations.

[0173] In the embodiment of the present application, the scheduling solution determination device may repeat the above operations until the convergence condition is met and then stop the repetitive operations. In this way, the scheduling solution determination device can obtain the final target scheduling solution.

[0174] In some embodiments, the taboo table includes a global taboo table and a local taboo table. The global taboo table is used to eliminate the candidate scheduling solutions in the global taboo table when performing selection operations, crossover operations, and mutation operations. The local taboo table is used to eliminate the candidate scheduling solutions in the local taboo table when performing crossover operations and mutation operations.

[0175] In some embodiments, the scheduling scheme determination method further includes: in the process of executing the selection operation, the crossover operation and the mutation operation, updating the taboo table according to the result of each operation.

[0176] In one practicable manner, after the scheduling scheme determining device performs a selection operation on multiple candidate scheduling schemes, the scheduling scheme determining device may clear the local taboo table and store the candidate scheduling schemes obtained by this operation into the local taboo table to update the local taboo table.

[0177] In another achievable manner, after the scheduling scheme determination device performs a selection operation on multiple candidate scheduling schemes, the scheduling scheme determination device may store the candidate scheduling scheme with the highest fitness and the candidate scheduling scheme with the lowest fitness obtained by this operation into a global taboo table to update the global taboo table.

[0178] In another achievable manner, after the scheduling scheme determination device performs a selection operation on multiple candidate scheduling schemes, the scheduling scheme determination device may store the candidate scheduling schemes obtained by this operation into a local taboo table, and delete the earliest record according to the capacity of the global taboo table to maintain the size of the table.

[0179] In one practicable manner, after the scheduling scheme determining device performs a cross operation on multiple candidate scheduling schemes, the scheduling scheme determining device may clear the local taboo table and store the candidate scheduling schemes obtained by this operation into the local taboo table to update the local taboo table.

[0180] In another achievable manner, after the scheduling scheme determination device performs a cross operation on multiple candidate scheduling schemes, the scheduling scheme determination device may store the candidate scheduling scheme with the highest fitness and the candidate scheduling scheme with the lowest fitness obtained by this operation into a global taboo table to update the global taboo table.

[0181] In one achievable manner, after the scheduling scheme determination device performs a mutation operation on multiple candidate scheduling schemes, the scheduling scheme determination device may clear the local taboo table and store the candidate scheduling schemes obtained by this operation into the local taboo table to update the local taboo table.

[0182] In another achievable manner, after the scheduling scheme determination device performs a mutation operation on multiple candidate scheduling schemes, the scheduling scheme determination device may store the candidate scheduling scheme with the highest fitness and the candidate scheduling scheme with the lowest fitness obtained by this operation into a global taboo table to update the global taboo table.

[0183] As can be seen from the above, the present application applies a global taboo table to store the candidate scheduling scheme with the highest fitness in the global scope (also known as the global optimal solution) and the candidate scheduling scheme with the lowest fitness (also known as the global worst solution) to ensure that the scheduling scheme determination device avoids known optimal solutions and inferior solutions during the search process based on the taboo search algorithm to search for candidate scheduling schemes in a larger range. The local taboo table stores all candidate scheduling schemes in each round of crossover and mutation operations to limit the generation of invalid solutions and duplicate solutions during the crossover and mutation operations, improve local search efficiency and ensure the diversity of candidate scheduling schemes.

[0184] In some embodiments, in combination Figure 7 ,like Figure 8 As shown, in S601, the fitness of multiple scheduling schemes to be selected is determined according to the scheduling information of the scheduling scheme to be scheduled, which specifically includes:

[0185] S801: Determine the fitness as the sum of the products of each piece of scheduling information and the weight factor of the scheduling information.

[0186] The scheduling information includes at least one of the following: the task success rate of the selected scheduling scheme, the average delay of the scheduled tasks, the resource utilization rate of the task processing equipment, the load rate of the task processing equipment, and the bandwidth utilization rate of the network link.

[0187] In the embodiment of the present application, the weight factor is used to indicate the importance of the scheduling information. Since there may be conflicts between the various scheduling information, it is necessary to combine multiple scheduling information according to their importance, that is, the scheduling scheme determination device can adjust the weight factor of the scheduling information according to user needs, and determine the sum of the products of each scheduling information and the weight factor of the scheduling information as the fitness. In order to achieve a comprehensive balance of multiple scheduling information. And determine the target scheduling scheme based on data such as fitness. In this way, the target scheduling scheme obtained by the scheduling scheme determination device is more reasonable.

[0188] When the scheduling information includes: the task success rate of the selected scheduling scheme, the average delay of the scheduled tasks, the resource utilization of the task processing equipment, the load rate of the task processing equipment, and the bandwidth utilization of the network link, the scheduling scheme determination device can determine the fitness as the sum of the products of each scheduling information and the weight factor of the scheduling information.

[0189] The scheduling scheme determination device can abstract scheduling information such as the task success rate of the selected scheduling scheme, the average delay of the tasks to be scheduled, the resource utilization of the task processing equipment, the load rate of the task processing equipment, the bandwidth utilization of the network link, etc. into an objective function that needs to be optimized, and determine the sum of the product of its objective function and the weight factor of the objective function as the fitness.

[0190] The fitness satisfies the following formula:

[0191] f(x)=max(αf1(x)+βf2(x)-γf′3(x)+δf4(x)-εf5(x));

[0192] Among them, f1(x) is used to represent the task success rate of the maximization of the candidate scheduling scheme, f2(x) is used to represent the resource utilization of the maximization task processing equipment, f′3(x) is used to represent the normalized objective function of minimizing the average delay of the scheduled tasks, f4(x) is used to represent the bandwidth utilization of the network link, and f5(x) is used to represent the load rate of the task processing equipment. α is used to represent the weight factor for maximizing the task success rate of the candidate scheduling scheme, β is used to represent the weight factor for maximizing the resource utilization of the task processing equipment, γ is used to represent the weight factor for minimizing the average delay of the scheduled tasks, δ is used to represent the weight factor for maximizing the bandwidth utilization of the network link, and ε is used to represent the weight factor for minimizing the load rate of the task processing equipment. The sum of the absolute values ​​of each weight factor is 1, α+β+γ+δ+ε=1.

[0193] Optionally, since the unit for minimizing the average delay of the task to be scheduled is inconsistent with the unit of other scheduling information, the scheduling scheme determination device needs to perform normalization processing on f3(x). In addition, the values ​​of other scheduling information are all in the form of ratios, that is, the values ​​of the scheduling information are all greater than or equal to 0 and less than or equal to 1. Therefore, the scheduling scheme determination device can map the f3(x) value to an interval greater than or equal to 0 and less than or equal to 1 to obtain f′3(x), and f′3(x) satisfies the following formula:

[0194]

[0195] Among them, f3(x) is used to represent the minimization of the average delay of the scheduled tasks, f3(x) min It is used to represent the theoretical minimum delay of the tasks to be scheduled, that is, the sum of the minimum delays of all feasible network links of the tasks to be scheduled, f3(x) max It is used to indicate the theoretical maximum delay of tasks to be scheduled, that is, the sum of the maximum delays of all feasible network links of tasks to be scheduled.

[0196] In summary, the scheduling scheme determination device performs a global search through a genetic algorithm to improve the diversity of candidate scheduling schemes. The scheduling scheme determination device improves the crossover operation process through a neighborhood search, and applies the local optimization capability of the taboo search algorithm to fine-tune the candidate scheduling scheme after the mutation operation, taking into account the capabilities of global search and local optimization to obtain the optimal target scheduling scheme. Among them, when performing global search and local optimization, the scheduling scheme determination device ensures the exploration of candidate scheduling schemes in a global scope based on the global taboo table and the local taboo table, avoids generating repeated or explored candidate scheduling schemes, and prevents the algorithm from falling into a local optimum. Without sacrificing the quality of the candidate scheduling scheme, the rate of determining the candidate scheduling scheme is improved.

[0197] In summary, Fig. 9 A flow chart of yet another method for determining a scheduling solution is shown.

[0198] First, the scheduling scheme determination device can initialize the global taboo table and the local taboo table. That is, the scheduling scheme determination device can clear the global taboo table and the local taboo table. So that the global taboo table is used to store the candidate scheduling scheme with the highest global fitness and the candidate scheduling scheme with the lowest global fitness. So that the local taboo table is used to store all the candidate scheduling schemes obtained during the crossover operation and the mutation operation.

[0199] Then, the scheduling scheme determination device can initialize multiple scheduling schemes to be selected. That is, the scheduling scheme determination device can determine multiple scheduling schemes to be selected according to the maximum tolerable delay of the task to be scheduled, the task processing delay of the task processing device and the transmission delay of the network link.

[0200] Next, the scheduling solution determination device may determine the fitness of multiple candidate scheduling solutions, and update the global taboo table and the local taboo table based on the fitness of the multiple candidate scheduling solutions.

[0201] The global taboo table records the candidate scheduling scheme with the highest fitness and the candidate scheduling scheme with the lowest fitness among multiple candidate scheduling schemes. The local taboo table records all the current candidate scheduling schemes.

[0202] Next, the scheduling scheme determination device can repeatedly iterate the above operations and determine whether the convergence conditions are met, that is, determine whether there is a candidate scheduling scheme that meets the threshold among the current multiple candidate scheduling schemes. If so, the scheduling scheme determination device outputs the optimal target scheduling scheme.

[0203] If not, the scheduling scheme determines whether the device has reached the preset number of iterations.

[0204] If the preset number of iterations has been reached, the scheduling scheme determination device outputs the current optimal target scheduling scheme.

[0205] If the preset number of iterations is not reached, the scheduling scheme determination device performs a selection operation. Then, the scheduling scheme determination device performs a crossover operation based on the taboo table and updates the local taboo table. Then, the scheduling scheme determination device performs a mutation operation based on the taboo table and updates the local taboo table. Then, the scheduling scheme determination device locally optimizes and updates the local taboo table based on the taboo search algorithm.

[0206] Next, the scheduling solution determination device may proceed to determine the fitness of multiple candidate scheduling solutions, and repeat the operation until the scheduling solution determination device outputs the optimal target scheduling solution or outputs the current optimal target scheduling solution.

[0207] For example, Fig.10 A flow chart of task scheduling in a cloud-network-edge architecture is shown.

[0208] First, the terminal initiates tasks, which can form a task cluster.

[0209] Then, the task scheduling system in the cloud-network-edge architecture can be classified according to task types.

[0210] Among them, tasks include: long-term resource occupation type tasks and service type tasks. Long-term resource occupation type tasks (for example, applying for a virtual machine) require long-term occupation of network resources (for example, cloud storage cameras). Service type tasks have high real-time requirements, such as cloud games, augmented reality, and virtual reality.

[0211] For tasks that occupy resources for a long time, a priority algorithm can be applied to process such tasks. That is, according to the priority scheduling algorithm, priorities are set for cloud data centers and edge cloud data centers, and such tasks are assigned to cloud data centers. In addition, the load conditions of cloud data centers and edge cloud data centers and the task latency of tasks that occupy resources for a long time are comprehensively considered, and the most appropriate cloud data center or edge cloud data center is selected to assign tasks that occupy resources for a long time.

[0212] Since long-term resource-occupying tasks will occupy resources for a long time, the resource inventory of the allocated cloud data center, edge cloud data center (also called task processing equipment) and network link resource inventory are deducted to obtain the scheduling result of long-term resource-occupying tasks.

[0213] For service type tasks, it is necessary to update the task processing equipment resource inventory and network resource inventory according to the scheduling results of long-term resource occupation type tasks. And according to the target scheduling scheme determined by the scheduling scheme device, the scheduling is performed (also known as the intelligent scheduling algorithm). At the same time, the network equipment connected to the task terminal, the assigned task processing equipment, the selected network link, and the result of whether the task is successfully scheduled are obtained.

[0214] Determine whether all tasks are successfully scheduled. If the scheduling fails, obtain the scheduling failure set and reprocess the failed tasks. That is, first update the inventory of task processing equipment resources and network link resources that have been released by the successful tasks, and then schedule again the target scheduling plan determined by the scheduling plan determination device until all tasks are successfully scheduled.

[0215] Optionally, record the number of task scheduling times.

[0216] If the scheduling is successful, the service type task scheduling result is output.

[0217] The scheduling results of long-term resource-occupying type tasks and service type tasks are integrated to obtain the final scheduling results.

[0218] As can be seen from the above, the above task scheduling is based on the cloud-network-edge-end integrated architecture. The terminal obtains each accessed network device, access delay, and the task information table initiated by it (including the computing resources, storage resources, and maximum tolerable delay required for each task). The network obtains the network topology and each network link and its transmission delay. The cloud data center and edge cloud data center obtain the computing resources and storage resource inventory of the current cloud data center and edge cloud data center, as well as their computing power performance indicators.

[0219] The above mainly introduces the solution provided by the embodiment of the present application from the perspective of the method. In order to realize the above functions, it includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0220] The embodiment of the present application can divide the functional modules of the scheduling scheme determination device according to the above method example. For example, each functional module can be divided corresponding to each function, or two or more functions can be integrated into one processing module. The above integrated module can be implemented in the form of hardware or in the form of software functional modules. Optionally, the division of modules in the embodiment of the present application is schematic and is only a logical function division. There may be other division methods in actual implementation.

[0221] Fig.11 FIG. 1 is a schematic diagram showing the structure of a scheduling scheme determination device provided in an embodiment of the present application. Fig.11 As shown, the scheduling scheme determination device includes: a communication unit 1101 and a processing unit 1102.

[0222] The communication unit 1101 is used to obtain the task information of the task to be scheduled, the task processing delay of the task processing device, and the transmission delay of the network link; the task processing delay is used to indicate the time length for the task processing device to process the task to be scheduled; the transmission delay is used to indicate the time length for transmitting the task to be scheduled to the task processing device; the processing unit 1102 is used to determine a plurality of candidate scheduling schemes according to the task information, the task processing delay and the transmission delay; the processing unit 1102 is also used to determine a target scheduling scheme based on a plurality of candidate scheduling schemes, a taboo search algorithm and a genetic algorithm.

[0223] Optionally, the task information includes: the maximum tolerable delay for processing the task to be scheduled; the processing unit 1102 is specifically used to: for each task to be scheduled and each task processing device, when the sum of the transmission delay of the task to be scheduled and the task processing delay of the task processing device is less than or equal to the maximum tolerable delay of the task to be scheduled, determine multiple scheduling schemes to be selected according to the number of tasks to be scheduled and the number of task processing devices.

[0224] Optionally, the processing unit 1102 is specifically used to: determine the fitness of multiple candidate scheduling schemes based on the scheduling information of the scheduling schemes to be scheduled; update the taboo table of the taboo search algorithm based on the fitness; and determine the target scheduling scheme based on the fitness, multiple candidate scheduling schemes, the taboo table and the genetic algorithm.

[0225] Optionally, the genetic algorithm includes: selection operation, crossover operation and mutation operation; processing unit 1102 is specifically used to: perform selection operation on multiple candidate scheduling schemes based on the taboo table and fitness to obtain at least one group of candidate scheduling schemes; each group of candidate scheduling schemes includes 2 candidate scheduling schemes, and each group of candidate scheduling schemes does not include the candidate scheduling schemes in the taboo table; perform crossover operation on each group of candidate scheduling schemes based on the taboo table to obtain multiple candidate scheduling schemes after the crossover operation; perform mutation operation on the multiple candidate scheduling schemes after the crossover operation based on the taboo table to obtain the candidate scheduling schemes after the mutation operation; repeat the selection operation, crossover operation and mutation operation until the convergence condition is met, and determine the candidate scheduling scheme after the mutation operation as the target scheduling scheme; the convergence condition includes: the number of schemes of the candidate scheduling schemes after the mutation operation is less than the preset number of schemes, or the number of mutation operations performed is greater than the preset number of operations.

[0226] Optionally, the taboo table includes a global taboo table and a local taboo table; the global taboo table is used to eliminate the candidate scheduling schemes in the global taboo table when performing selection operations, crossover operations and mutation operations; the local taboo table is used to eliminate the candidate scheduling schemes in the local taboo table when performing crossover operations and mutation operations.

[0227] Optionally, the processing unit 1102 is further configured to update the taboo table according to the result of each operation during the execution of the selection operation, the crossover operation and the mutation operation.

[0228] Optionally, the scheduling information includes at least one of the following: the task success rate of the selected scheduling scheme, the average delay of the task to be scheduled, the resource utilization of the task processing equipment, the load rate of the task processing equipment, and the bandwidth utilization of the network link. The processing unit 1102 is specifically used to determine the fitness as the sum of the products of each scheduling information and the weight factor of the scheduling information.

[0229] An embodiment of the present application also provides a computer-readable storage medium, which includes computer execution instructions. When the computer execution instructions are executed on a computer, the computer executes the scheduling scheme determination method provided in the above embodiment.

[0230] The embodiment of the present application also provides a computer program, which can be directly loaded into a memory and contains software code. After being loaded and executed by a computer, the computer program can implement the scheduling scheme determination method provided in the above embodiment.

[0231] Those skilled in the art should be aware that in one or more of the above examples, the functions described in this application can be implemented using hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer-readable storage media and communication media, wherein communication media include any media that facilitates the transmission of computer programs from one place to another. Storage media can be any available media that can be accessed by general or special-purpose computers.

[0232] Through the description of the above implementation methods, technical personnel in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0233] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules or units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the coupling or direct coupling or communication connection between each other shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms. The unit described as a separate component may or may not be physically separated, and the component displayed as a unit may be a physical unit or multiple physical units, that is, it may be located in one place, or it may be distributed in multiple different places. Some or all of the units can be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0234] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on such an understanding, the technical solution of the embodiment of the present application is essentially or the part that contributes to the general technology or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium, including several instructions to enable a device (which can be a single-chip microcomputer, chip, etc.) or a processor (processor) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a U disk, a mobile hard disk, a ROM, a RAM, a disk, or an optical disk.

[0235] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for determining a scheduling scheme, characterized in that: The method comprises: Obtaining task information of the task to be scheduled, task processing delay of the task processing device, and transmission delay of the network link; the task processing delay is used to indicate the duration of the task processing device processing the task to be scheduled; the transmission delay is used to indicate the duration of transmitting the task to be scheduled to the task processing device; Determining a plurality of scheduling schemes to be selected according to the task information, the task processing delay and the transmission delay; Based on the multiple candidate scheduling solutions, the taboo search algorithm and the genetic algorithm, a target scheduling solution is determined.

2. The method according to claim 1, characterized in that: The task information includes: a maximum tolerable delay for processing the task to be scheduled; and determining a plurality of scheduling schemes to be selected according to the task information, the task processing delay and the transmission delay, including: For each of the tasks to be scheduled and each task processing device, when the sum of the transmission delay of the task to be scheduled and the task processing delay of the task processing device is less than or equal to the maximum tolerable delay of the task to be scheduled, the multiple scheduling schemes to be selected are determined according to the number of the tasks to be scheduled and the number of task processing devices.

3. The method according to claim 1, characterized in that The step of determining a target scheduling scheme based on the plurality of candidate scheduling schemes, the taboo search algorithm and the genetic algorithm comprises: Determining the fitness of the plurality of the to-be-selected scheduling schemes according to the scheduling information of the to-be-scheduled schemes; Updating a taboo table of the taboo search algorithm according to the fitness; The target scheduling solution is determined based on the fitness, the multiple candidate scheduling solutions, the taboo table and the genetic algorithm.

4. The method according to claim 3, characterized in that The genetic algorithm includes: a selection operation, a crossover operation and a mutation operation; the step of determining the target scheduling scheme based on the fitness, the multiple candidate scheduling schemes, the taboo table and the genetic algorithm includes: Performing the selection operation on the multiple candidate scheduling schemes based on the taboo table and the fitness to obtain at least one group of candidate scheduling schemes; each group of candidate scheduling schemes includes 2 of the candidate scheduling schemes, and each group of candidate scheduling schemes does not include a candidate scheduling scheme in the taboo table; Performing the crossover operation on each group of candidate scheduling solutions based on the taboo table to obtain a plurality of candidate scheduling solutions after the crossover operation; Based on the taboo table, the mutation operation is performed on the multiple candidate scheduling schemes after the crossover operation is processed to obtain the candidate scheduling schemes after the mutation operation is processed; Repeat the selection operation, the crossover operation and the mutation operation until a convergence condition is met, and determine the candidate scheduling scheme after the mutation operation as the target scheduling scheme; the convergence condition includes: the number of the candidate scheduling schemes after the mutation operation is less than the preset number of schemes, or the number of times the mutation operation is performed is greater than the preset number of operations.

5. The method according to claim 4, characterized in that The taboo table includes a global taboo table and a local taboo table; the global taboo table is used to eliminate the candidate scheduling schemes in the global taboo table when performing the selection operation, the crossover operation and the mutation operation; The local taboo table is used to eliminate the candidate scheduling schemes in the local taboo table when executing the crossover operation and the mutation operation.

6. The method according to claim 4, characterized in that The method further comprises: In the process of executing the selection operation, the crossover operation and the mutation operation, the taboo table is updated according to the result of each operation.

7. The method according to claim 3, characterized in that The scheduling information includes at least one of the following: the task success rate of the to-be-scheduled scheduling scheme, the average delay of the to-be-scheduled task, the resource utilization rate of the task processing device, the load rate of the task processing device, and the bandwidth utilization rate of the network link. The determining the fitness of the plurality of the to-be-scheduled scheduling schemes according to the scheduling information of the to-be-scheduled scheduling schemes includes: The sum of the products of each of the scheduling information and the weight factors of the scheduling information is determined as the fitness.

8. A scheduling scheme determination device, characterized in that: The device comprises: a communication unit and a processing unit; The communication unit is used to obtain task information of the task to be scheduled, the task processing delay of the task processing device, and the transmission delay of the network link; the task processing delay is used to indicate the time length for the task processing device to process the task to be scheduled; the transmission delay is used to indicate the time length for transmitting the task to be scheduled to the task processing device; The processing unit is used to determine a plurality of scheduling schemes to be selected according to the task information, the task processing delay and the transmission delay; The processing unit is further used to determine a target scheduling scheme based on the multiple candidate scheduling schemes, the taboo search algorithm and the genetic algorithm.

9. A scheduling scheme determination device, characterized in that: include: A processor and a memory; wherein the memory is used to store one or more programs, and the one or more programs include computer-executable instructions. When the device is running, the processor executes the computer-executable instructions stored in the memory to enable the device to perform the method described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: When the computer-executable instructions stored in the computer-readable storage medium are executed by a processor of a scheduling solution determination device, the scheduling solution determination device can perform the method according to any one of claims 1 to 7.

11. A computer program product, characterized in that The computer program product comprises: a computer program or instructions, and when the computer program or instructions are run on a computer, the computer is caused to perform the method according to any one of claims 1 to 7.

Images