CN111475274A - Cloud collaborative multi-task scheduling method and device - Google Patents
Cloud collaborative multi-task scheduling method and device Download PDFInfo
- Publication number
- CN111475274A CN111475274A CN202010313381.4A CN202010313381A CN111475274A CN 111475274 A CN111475274 A CN 111475274A CN 202010313381 A CN202010313381 A CN 202010313381A CN 111475274 A CN111475274 A CN 111475274A
- Authority
- CN
- China
- Prior art keywords
- objective function
- task
- population
- preset
- cloud
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
- G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
- G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5083—Techniques for rebalancing the load in a distributed system
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The embodiment of the invention provides a cloud collaborative multi-task scheduling method and a device, wherein the method comprises the following steps: acquiring a task to be processed received by each edge node; distributing all tasks to be processed to different edge nodes for processing, using the tasks as a habitat distribution scheme, and randomly generating an initial population with a plurality of habitat distribution schemes; and acquiring an optimal solution meeting the objective function based on a preset objective function and a biophysical optimization algorithm to obtain the to-be-processed task to be distributed by each edge node. The optimal solution meeting the objective function is obtained based on a preset objective function and a biogeography optimization algorithm, the task to be processed which is to be distributed to each edge node is obtained, reasonable task distribution of all edge nodes and a central cloud node is considered, an optimal distribution scheme meeting the optimization objective can be obtained based on the preset objective function, the optimal distribution scheme is realized based on the biogeography optimization algorithm, and the calculation cost is greatly reduced.
Description
Technical Field
The invention relates to the field of edge computing, in particular to a cloud collaborative multi-task scheduling method and device.
Background
The cloud-side cooperative task scheduling is a process for unloading a computation-intensive task to a cloud server side to perform optimization by using massive intelligent terminal equipment, and is very important in a scene with high performance requirements such as system time delay and energy consumption. Under the background of everything interconnection, network edge devices are various, and numerous applications with large computing demands appear, such as face recognition, interactive games, augmented reality and the like. These applications are large in data size, high in delay requirements, and consume a large amount of computing resources and energy, so that the computing tasks of the terminal device are considered to be offloaded to remote execution.
the method comprises the steps that an edge node is closer to a user than a cloud data center, a large amount of delay caused by long-distance transmission is avoided, user experience is improved, tasks which need to be unloaded to the cloud data center to be executed are distributed to the edge cloud to be executed, the transmission burden of a Backhaul link (Backhaul L ink) of a core network is effectively relieved, and the problems of core network congestion and transmission delay of a traditional cloud computing system are solved.
Currently, a cooperative task scheduling strategy of edge computing is receiving wide attention, and an edge cloud can provide computing and storage resources for any user in a nearby area according to an open global standardization mechanism, so that edge nodes in the same area can form a shared and interconnected edge resource pool. However, the current method is limited to the cooperation among a plurality of edge nodes or the cooperation between the current edge node and the central cloud, and cannot realize the overall cooperative scheduling of the central cloud and all the edge nodes, which results in a huge amount of computation.
Disclosure of Invention
In order to solve the above problem, embodiments of the present invention provide a cloud collaborative multitask scheduling method and apparatus.
In a first aspect, an embodiment of the present invention provides a cloud collaborative multitask scheduling method, including: acquiring a task to be processed received by each edge node; distributing all tasks to be processed to different edge nodes for processing, using the tasks as a habitat distribution scheme, and randomly generating an initial population with a plurality of habitat distribution schemes; and acquiring an optimal solution meeting the objective function based on a preset objective function and a biophysical optimization algorithm to obtain the to-be-processed task to be distributed by each edge node.
Further, the obtaining an optimal solution meeting the objective function based on a preset objective function and a biophysical optimization algorithm includes: calculating the inhabitation suitability index HSI of each population, wherein according to different optimization targets, the HSI value and the objective function value form a positive correlation relationship or an inverse correlation relationship; carrying out migration operation on the population based on a preset migration rate function, and carrying out mutation operation on the population based on a preset mutation probability function; if the preset termination condition is met, outputting an optimal solution, otherwise, repeating the iterative process from the calculation of the HSI value of each population to the judgment of whether the termination condition is met.
Further, the objective function is determined according to a transmission delay and energy consumption of each task of the habitat allocation scheme.
Further, the objective function is:
wherein, theta is the weight of the time delay objective function, and theta is equal to [0,1 ] ]；Lα、Lβ、Eα、EβThe maximum and minimum values of the transmission delay and the energy consumption of the habitat in the population are respectively, and X is a habitat allocation scheme.
Further, the mutation operation on the population based on the preset mutation probability function includes:
When the individual HSI value is lower than the average HSI value of the population individuals, selecting smaller mutation probability;
When the individual HSI value is higher than the average HSI value of the population individuals, a larger mutation probability is selected.
Further, the mutation operation on the population based on the preset mutation probability function further includes: and performing mutation operation according to the fixed probability at the early stage of the iteration times, and performing self-adaptive adjustment on the mutation probability at the later stage of the iteration.
Further, the mutation probability function is:
Wherein g is the number of the iteration times, n is the sequence number of the individuals in the group, 
Is the mutation probability of the nth individual in the g iteration, g threIs the iteration number boundary of the fixed mutation operation and the self-adaptive mutation operation; cost n、Costmin、CostavgRespectively representing the HSI value of the nth individual in the previous iteration, the minimum HSI value in the population and the average HSI value in the population; k is a radical of 1、k2、k3Is a preset constant.
In a second aspect, an embodiment of the present invention provides a cloud collaborative multitask scheduling device, including: the task acquisition module is used for acquiring the to-be-processed tasks received by each edge node; the first collaborative scheduling module is used for distributing all tasks to be processed to different edge nodes for processing, and the tasks are used as a habitat distribution scheme to randomly generate an initial population with a plurality of habitat distribution schemes; and the second cooperative scheduling module is used for acquiring an optimal solution meeting the objective function based on a preset objective function and a biophysical optimization algorithm to obtain the to-be-processed task to be allocated to each edge node.
In a third aspect, an embodiment of the present invention provides an electronic device, which includes a memory, a processor, and a computer program that is stored in the memory and is executable on the processor, where the processor executes the computer program to implement the steps of the cloud collaborative multitask scheduling method according to the first aspect of the present invention.
In a fourth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps of the cloud collaborative multitask scheduling method according to the first aspect of the present invention.
According to the cloud collaborative multi-task scheduling method and device provided by the embodiment of the invention, the optimal solution meeting the objective function is obtained based on the preset objective function and the biophysical optimization algorithm, the task to be processed which is to be distributed to each edge node is obtained, the reasonable task distribution of all the edge nodes and the central cloud node is considered, the optimal distribution scheme meeting the optimization objective can be obtained based on the preset objective function, the implementation is realized based on the biophysical optimization algorithm, and the calculation overhead is greatly reduced.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
Fig. 1 is a flowchart of a cloud collaborative multitask scheduling method according to an embodiment of the present invention;
Fig. 2 is a schematic diagram of a task offloading to an edge cloud scene according to an embodiment of the present invention;
FIG. 3 is a diagram illustrating a mapping relationship between tasks and computing nodes according to an embodiment of the present invention;
FIG. 4 is a flowchart of a biophysical optimization algorithm-based embodiment of the invention;
Fig. 5 is a structural diagram of a cloud cooperative multitask scheduling device according to an embodiment of the present invention;
Fig. 6 is a schematic physical structure diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The efficient cloud-edge cooperative task scheduling method provided by the embodiment of the invention relates to an end-edge-cloud three-layer system architecture. The method comprises the following specific steps:
The "end" is the terminal device layer, which includes a number of heterogeneous devices running computationally intensive applications such as virtual reality, unmanned, etc. The computing-intensive tasks are partially or completely migrated to the cloud end with rich resources to be executed, so that the computing capability of the cloud end is expanded, and the defect of energy storage is overcome.
The 'edge' is an edge cloud, and is composed of a plurality of interconnected edge nodes, and a cloud infrastructure is deployed in an edge network close to a user to provide a nearby cloud service for the user, wherein: an edge node is a combination of a base station and a MEC server deployed thereon. The mobile terminal device is connected to the edge node through a wireless channel, and transmits the task needing to be unloaded to the edge node associated with the mobile terminal device. The edge nodes are connected with each other, and data interaction can be carried out. The main modules in the edge node include a calculation module and a transmission module. The computing power of the edge node is stronger than that of the terminal equipment, and the tasks of the terminal equipment can be effectively processed. The transmission module is used for data interaction between the edge node and other interconnected edge nodes and the center cloud.
The cloud is a central cloud at a far end, has extremely strong resources such as calculation, storage and the like, and is responsible for unified scheduling of the whole resources of the system. The central cloud mainly comprises an execution computing unit, a core network transmission module and a cooperative scheduling module, wherein: the execution computing unit means that abundant computing resources of the cloud service center can be used for executing the compute-intensive tasks, computing time delay overhead is reduced, and time delay performance of the system is optimized. The execution computing unit of the central cloud is an important composition of the computing capacity of the system, and when the system load is heavy and the resources of the edge cloud are insufficient, the computing resources of the central cloud are supplemented powerfully.
Fig. 1 is a flowchart of a cloud collaborative multitask scheduling method according to an embodiment of the present invention, and as shown in fig. 1, the embodiment of the present invention provides a cloud collaborative multitask scheduling method, including:
101. And acquiring the to-be-processed task received by each edge node.
Fig. 2 is a schematic view of a scene in which a task is offloaded to an edge cloud according to an embodiment of the present invention, where the scene is composed of a mobile terminal and an edge cloud service infrastructure. The edge computing can fully compute or preprocess the tasks of the terminal equipment at the edge node, and the problems of congestion and delay of a core network are relieved. End-to-end delay is a key optimization target of network application, and directly influences system performance and user experience. The edge cloud service infrastructure mainly comprises an edge service access point AP, an edge computing server and a wired network for connecting each edge node.
Mobile devices in the system have been associated with a particular base station through some policy, and the mobile devices are connected to the base station via a wireless channel. Consider the situation in which, in a multi-cell mobile cellular network scenario, devices share radio channel resources in a Time Division Multiple Access (TDMA) manner. In a short period of time (with a negligible length), each terminal in the system runs a real-time delay sensitive application to generate a calculation task, and the calculation capability of each terminal is limited to be unloaded and executed.
The computing nodes are collectively referred to as nodes in the edge service infrastructure that can independently perform computations. The computing nodes may be computing resources of an edge server deployed on the base station side, or may be servers or deployed virtual machines of other edge data centers.
Tasks are transmitted over a wireless channel to an associated edge node, provided that each task can be executed locally at its associated edge node or transmitted over a wired network to other computing nodes in the system. The transmission of tasks between compute nodes consumes bandwidth resources at the input and output nodes and creates transmission delays. The unloading execution of the data processing task generated by the mobile terminal device mainly comprises the following steps:
(1) The mobile terminal equipment transmits task data to be unloaded to the edge node corresponding to the position of the mobile terminal equipment through a wireless channel;
(2) The scheduling program specifies an execution place for the task, and can be left in local execution or scheduled to other interconnected edge cloud nodes according to the situation. If the task data are scheduled to be executed by other nodes, the task data are transmitted to the nodes through a wired network;
(3) The calculation results are sent back from the edge node to the mobile device.
In many practical scenes, the data volume of the calculation result of the data calculation task is very small, such as face recognition, video analysis and the like, the time for transmitting the calculation result in the step (3) back to the terminal equipment can be ignored, and the end-to-end time delay mainly takes the transmission time delay and the calculation time delay of a wireless channel into consideration.
In 101, the execution subject is a central cloud server or a central cloud related device. The mobile terminal sends the tasks to be processed to the edge nodes, the central cloud acquires the condition of each task to be processed through the edge nodes, and the acquired parameters can be related to the tasks, such as the calculation amount, the data amount and the like, so that after the tasks are distributed to the edge nodes, the transmission delay overhead, the calculation amount overhead, the energy consumption overhead and the like of each task when the distribution scheme is executed are calculated according to the parameters.
102. Distributing all tasks to be processed to different edge nodes for processing, and randomly generating an initial population with a plurality of habitat distribution schemes as a habitat distribution scheme.
The invention is realized by adopting a biophysical optimization algorithm (BBO), which is a novel evolutionary algorithm and is based on a biophysical theory. The basic idea is derived from the geographical distribution and species migration law of the natural organisms. The BBO algorithm has strong utilization capability on beneficial information in a group, good global search capability and high convergence speed.
Fig. 3 is a schematic diagram of a mapping relationship between tasks and computing nodes according to an embodiment of the present invention. The scheduling process of the central cloud requires that a unique computing node is assigned to all tasks in the system, and the unique computing node may be an edge node or a central cloud node. It is specified that tasks in edge cloud systems are indivisible, i.e. the same task specifies that unique nodes are all executing. Each computing node may be assigned multiple tasks that share the computing resources of the node, and the cloud service infrastructure may assign the computing resources to the tasks through virtualization techniques. 102, one allocation scheme of all nodes corresponds to one habitat of the BBO algorithm, the initial population is composed of a plurality of allocation schemes, the initial allocation result may be generated randomly, the task is allocated to the edge node, and the unallocated one is completed by the central cloud node.
103. And acquiring an optimal solution meeting the objective function based on a preset objective function and a biophysical optimization algorithm to obtain the to-be-processed task to be distributed by each edge node.
In the embodiment of the invention, the objective function is determined according to the optimization target, for example, if the optimization target is that the weighted average time delay of the task in the system is minimum, the objective function is determined according to the weighted average time delay of the task in the system. It is considered that the tasks in the system have different sensitivities to the delay, or the users have different requirements on the delay performance of the system. A weight can be set for each task to reflect the priority of the task or the requirement degree of the time delay performance so as to optimize the whole time delay performance in the system. And then based on BBO algorithm optimization, selecting an optimal solution meeting the objective function, wherein the optimal solution corresponds to a distribution scheme of each task to a corresponding computing node, and can ensure that the objective function is met, namely the optimization objective is achieved. It should be noted that tasks that are not assigned to the edge nodes are processed at the central cloud node.
The cloud collaborative multi-task scheduling method provided by the embodiment of the invention is based on a preset objective function and a biogeography optimization algorithm, obtains an optimal solution meeting the objective function, obtains a task to be processed which is to be distributed by each edge node, considers the reasonable task distribution of all the edge nodes and the central cloud node, can obtain an optimal distribution scheme meeting an optimization objective based on the preset objective function, and is realized based on the biogeography optimization algorithm, so that the calculation overhead is greatly reduced.
Based on the content of the foregoing embodiment, as an optional embodiment, the obtaining an optimal solution satisfying an objective function based on a preset objective function and a biophysical optimization algorithm includes: calculating the inhabitation suitability index HSI of each population, wherein according to different optimization targets, the HSI value and the objective function value form a positive correlation relationship or an inverse correlation relationship; the method comprises the steps of arranging populations in a descending order according to HSI, carrying out migration operation on the populations based on a preset migration rate function, and carrying out mutation operation on the populations based on a preset mutation probability function; if the preset termination condition is met, outputting an optimal solution, otherwise, repeating the iterative process from the calculation of the HSI value of each population to the judgment of whether the termination condition is met.
Fig. 4 is a flowchart of a biophysical optimization-based algorithm according to an embodiment of the invention, and as shown in fig. 4, the core of the algorithm is migration and mutation. Wherein Smax is the maximum population number, I is the maximum mobility, E is the maximum mobility, and Mmax represents the upper limit of the mutation rate.
The migration process refers to information sharing between habitats and determines local optimizing capacity. The migration process is determined by the migration probability, and a conventional mobility model, such as an exponential mobility model, a cosine mobility model, etc., may be used. The mutation imitates the change caused by an emergency, the global searching capability is influenced, the operation is also carried out according to the probability, and the corresponding mutation function can be set to realize. According to different optimization targets, the objective function reflects the HSI value in the BBO algorithm, if the optimization target is that the total transmission delay of the system is minimum, the HSI value of the population individual with small total transmission delay is large, and the HSI value of the population individual with large total transmission delay is small, so that an inverse correlation relationship is formed. If the optimization target is larger, the optimization target is better, and the HSI value and the objective function value form a positive correlation relationship.
Based on the contents of the above embodiments, as an alternative embodiment, the objective function is determined according to the time delay and the energy consumption of each task of the habitat solution.
Besides time delay, energy consumption of task execution is also an important index to be considered by cloud service, energy consumption overhead is a main cost of cloud service infrastructure operation, and energy consumption is generated by computing resources consumed by task execution and data transmission among nodes. The time delay and the energy consumption are jointly optimized in the edge computing cooperative task scheduling problem, the time delay and the energy consumption are jointly used as optimization targets, the optimization targets are mutually balanced, and the optimization direction can be adjusted through a weight factor.
The embodiment of the invention considers the optimization of energy consumption while improving the time delay performance of the system, and can lower the cost of cloud service. The embodiment of the invention combines the time delay cost and the energy consumption cost of the system to define the system overhead, solves the multi-objective optimization problem of time delay and energy consumption, and is favorable for task allocation to obtain the optimal result.
Based on the content of the foregoing embodiment, as an alternative embodiment, the objective function is:
wherein, theta is the weight of the time delay objective function, and theta is equal to [0,1 ] ]；Lα、Lβ、Eα、EβThe maximum value and the minimum value of the transmission delay and the maximum value and the minimum value of the energy consumption of the habitat in the population are respectively, and X is a habitat allocation scheme.
And adopting an aggregation function method in the multi-objective optimization problem of cloud-edge cooperative task scheduling to combine the optimization of time delay and energy consumption into one objective optimization function. The aggregation function reduces the numerical difference caused by different target types by respectively carrying out standardization processing on the two target functions; the value of theta quantifies how heavily the system focuses on the optimization objective, and single-objective optimization of latency is performed when theta is 1.
The system average delay overhead can be expressed as:
Wherein J represents the number of edge nodes, J is the total number, I is the number of tasks to be distributed, and I jIs the total number of tasks of the corresponding node. T is j,ifor transmission delay of the corresponding task, alpha j,iIs the delay weight of the corresponding task.
The system average energy consumption overhead can be expressed as:
Wherein E is j,iRepresents the energy consumption overhead of the corresponding task, and I is the total task number.
According to the embodiment of the invention, the two target functions are respectively subjected to standardization processing through the aggregation function, the numerical difference caused by different target types is reduced, and the targeted single-target optimization can be realized according to the value of theta of 0 or 1.
Based on the content of the foregoing embodiment, as an optional embodiment, performing mutation operation on the population based on a preset mutation probability function includes: when the individual HSI value is lower than the average HSI value of all individuals in the population, selecting smaller mutation probability; when the individual HSI value is higher than the average HSI value of all individuals in the population, a larger mutation probability is selected.
For simple BBO, mutation parameters are confirmed in advance, and cannot be flexibly adjusted according to actual conditions in the evolution process, so that precocity and local convergence easily occur. The embodiment of the invention provides a hybrid adaptive BBO (HABBO) algorithm by adaptively adjusting mutation probability according to iteration times and individual quality in the iterative optimization process of the algorithm, thereby remarkably improving the search efficiency.
In the implementation of the mutation module of the BBO algorithm, a random number between 0 and 1 is generated, and if rand (0,1) is less than the mutation probability of an individual, the individual is mutated. In the mutation probability adaptive BBO proposed in the embodiment of the present invention, the mutation probability of an individual is affected by the relationship between the HSI value of the individual and the average value of the whole HSI in the last iteration. And replacing the HSI value with the system overhead calculated according to the objective function of the specific problem. When the individual overhead value is lower than the average overhead value of all individuals in the population, the solution quality is better, and the excellent individual is protected by selecting smaller mutation probability; and when the individual cost is higher than the average cost value of all individuals in the population, the quality of the solution is poor, and a larger mutation probability is selected for exploration.
According to the embodiment of the invention, when the HSI value of an individual is lower than the average HSI value of population individuals, a smaller mutation probability is selected; when the individual HSI value is higher than the average HSI value of the population individuals, a larger mutation probability is selected, so that the premature and local convergence of the algorithm can be avoided.
Based on the content of the foregoing embodiment, as an optional embodiment, the mutation operation is performed on the population based on a preset mutation probability function, and the method further includes: and performing mutation operation according to the fixed probability at the early stage of the iteration times, and performing self-adaptive adjustment on the mutation probability at the later stage of the iteration.
In the early stage of evolution, excellent individuals in a population also need to enhance the optimization efficiency of the algorithm through mutation operation with a certain probability. Therefore, the evolution process is divided into an early stage and a later stage, mixed self-adaptive mutation operation is carried out, mutation operation with fixed probability is carried out in the early stage, and mutation probability is changed in a self-adaptive mode in the later stage. Therefore, the global optimization capability of the algorithm in the middle and later stages can be improved, premature convergence is avoided, and the efficiency of the whole algorithm is improved.
Based on the contents of the above embodiments, as an alternative embodiment, the mutation probability function is:
Wherein g is the number of the iteration times, n is the sequence number of the individuals in the group, 
Is the mutation probability of the nth individual in the g iteration, g threIs the iteration number boundary of the fixed mutation operation and the self-adaptive mutation operation; cost n、Costmin、CostavgRespectively representing the HSI value of the nth individual in the previous iteration, the minimum HSI value in the population and the average HSI value in the population; k is a radical of 1、k2、k3Is a preset constant.
Fig. 5 is a structural diagram of a cloud cooperative multi-task scheduling device according to an embodiment of the present invention, and as shown in fig. 5, the cloud cooperative multi-task scheduling device includes: a task obtaining module 501, a first cooperative scheduling module 502 and a second cooperative scheduling module 503. The task obtaining module 501 is configured to obtain a to-be-processed task received by each edge node; the first cooperative scheduling module 502 is configured to allocate all tasks to be processed to different edge nodes for processing, serve as a habitat allocation scheme, and randomly generate an initial population having multiple habitat allocation schemes; the second cooperative scheduling module 503 is configured to obtain an optimal solution meeting the objective function based on a preset objective function and a biophysical optimization algorithm, and obtain a to-be-processed task to be allocated to each edge node.
The device embodiment provided in the embodiments of the present invention is for implementing the above method embodiments, and for details of the process and the details, reference is made to the above method embodiments, which are not described herein again.
The cloud collaborative multi-task scheduling device provided by the embodiment of the invention is based on a preset objective function and a biogeography optimization algorithm, obtains an optimal solution meeting the objective function, obtains a task to be processed to be allocated to each edge node, considers the reasonable task allocation of all the edge nodes and the central cloud node, can obtain an optimal allocation scheme meeting an optimization objective based on the preset objective function, and is realized based on the biogeography optimization algorithm, thereby greatly reducing the calculation overhead.
Fig. 6 is a schematic entity structure diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 6, the electronic device may include: a processor 601, a communication Interface 602, a memory 603 and a bus 604, wherein the processor 601, the communication Interface 602 and the memory 603 complete communication with each other through the bus 604. The communication interface 602 may be used for information transfer of an electronic device. The processor 601 may call logic instructions in the memory 603 to perform a method comprising: acquiring a task to be processed received by each edge node; distributing all tasks to be processed to different edge nodes for processing, using the tasks as a habitat distribution scheme, and randomly generating an initial population with a plurality of habitat distribution schemes; and acquiring an optimal solution meeting the objective function based on a preset objective function and a biophysical optimization algorithm to obtain the to-be-processed task to be distributed by each edge node.
In addition, the logic instructions in the memory 603 may be implemented in the form of software functional units and stored in a computer readable storage medium when the logic instructions are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the above-described method embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In another aspect, an embodiment of the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program is implemented to perform the transmission method provided in the foregoing embodiments when executed by a processor, and for example, the method includes: acquiring a task to be processed received by each edge node; distributing all tasks to be processed to different edge nodes for processing, using the tasks as a habitat distribution scheme, and randomly generating an initial population with a plurality of habitat distribution schemes; and acquiring an optimal solution meeting the objective function based on a preset objective function and a biophysical optimization algorithm to obtain the to-be-processed task to be distributed by each edge node.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods of the various embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A cloud collaborative multitask scheduling method is characterized by comprising the following steps:
Acquiring a task to be processed received by each edge node;
Distributing all tasks to be processed to different edge nodes for processing, using the tasks as a habitat distribution scheme, and randomly generating an initial population with a plurality of habitat distribution schemes;
And acquiring an optimal solution meeting the objective function based on a preset objective function and a biophysical optimization algorithm to obtain the to-be-processed task to be distributed by each edge node.
2. The cloud collaborative multitask scheduling method according to claim 1, wherein the obtaining of an optimal solution satisfying a preset objective function based on the preset objective function and a biophysical optimization algorithm includes:
Calculating the inhabitation suitability index HSI of each population, wherein according to different optimization targets, the HSI value and the objective function value form a positive correlation relationship or an inverse correlation relationship;
Carrying out migration operation on the population based on a preset migration rate function, and carrying out mutation operation on the population based on a preset mutation probability function;
If the preset termination condition is met, outputting an optimal solution, otherwise, repeating the iterative process from the calculation of the HSI value of each population to the judgment of whether the termination condition is met.
3. The cloud collaborative multitasking scheduling method according to claim 1, characterized in that the objective function is determined according to a transmission delay and an energy consumption of each task of a habitat allocation plan.
4. The cloud collaborative multitasking scheduling method according to claim 3, characterized in that the objective function is
wherein, theta is the weight of the time delay objective function, and theta is equal to [0,1 ] ]；Lα、Lβ、Eα、EβThe maximum value and the minimum value of the transmission delay and the maximum value and the minimum value of the energy consumption of the habitat in the population are respectively, and X is a habitat allocation scheme.
5. The cloud collaborative multitask scheduling method according to claim 1, wherein the mutation operation on the population based on a preset mutation probability function includes:
When the individual HSI value is lower than the average HSI value of the population individuals, selecting smaller mutation probability;
When the individual HSI value is higher than the average HSI value of the population individuals, a larger mutation probability is selected.
6. The cloud collaborative multitask scheduling method according to claim 5, wherein the mutation operation is performed on the population based on a preset mutation probability function, and further comprising:
And performing mutation operation according to the fixed probability at the early stage of the iteration times, and performing self-adaptive adjustment on the mutation probability at the later stage of the iteration.
7. The cloud collaborative multitask scheduling method according to claim 6, characterized in that the mutation probability function is:
Wherein g is the number of the iteration times, n is the sequence number of the individuals in the group, 
Is the mutation probability of the nth individual in the g iteration, g threIs the iteration number boundary of the fixed mutation operation and the self-adaptive mutation operation; cost n、Costmin、CostavgRespectively representing the HSI value of the nth individual in the previous iteration, the minimum HSI value in the population and the average HSI value in the population; k is a radical of 1、k2、k3Is a preset constant.
8. A cloud collaborative multitask scheduling device, comprising:
The task acquisition module is used for acquiring the to-be-processed tasks received by each edge node;
The first collaborative scheduling module is used for distributing all tasks to be processed to different edge nodes for processing, and the tasks are used as a habitat distribution scheme to randomly generate an initial population with a plurality of habitat distribution schemes;
And the second cooperative scheduling module is used for acquiring an optimal solution meeting the objective function based on a preset objective function and a biophysical optimization algorithm to obtain the to-be-processed task to be allocated to each edge node.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the cloud collaborative multitasking scheduling method according to any one of claims 1 to 7 when executing the program.
10. A non-transitory computer readable storage medium, having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the cloud collaborative multitasking scheduling method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010313381.4A CN111475274B (en) | 2020-04-20 | 2020-04-20 | Cloud collaborative multi-task scheduling method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010313381.4A CN111475274B (en) | 2020-04-20 | 2020-04-20 | Cloud collaborative multi-task scheduling method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111475274A true CN111475274A (en) | 2020-07-31 |
| CN111475274B CN111475274B (en) | 2023-04-18 |
Family
ID=71754294
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010313381.4A Active CN111475274B (en) | 2020-04-20 | 2020-04-20 | Cloud collaborative multi-task scheduling method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111475274B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112202866A (en) * | 2020-09-25 | 2021-01-08 | 大拓无限(重庆)智能科技有限公司 | Method, device and equipment for task scheduling |
| CN112287609A (en) * | 2020-12-28 | 2021-01-29 | 之江实验室 | End, edge and cloud collaborative computing device for robot task division |
| CN112650585A (en) * | 2020-12-24 | 2021-04-13 | 山东大学 | Novel edge-cloud collaborative edge computing platform, method and storage medium |
| CN112702401A (en) * | 2020-12-15 | 2021-04-23 | 北京邮电大学 | Multi-task cooperative allocation method and device for power Internet of things |
| CN112910698A (en) * | 2021-01-27 | 2021-06-04 | 网宿科技股份有限公司 | CDN coverage scheme adjusting method, device and equipment |
| CN113015216A (en) * | 2021-02-05 | 2021-06-22 | 浙江大学 | Burst task unloading and scheduling method facing edge service network |
| CN113342504A (en) * | 2021-07-02 | 2021-09-03 | 西安邮电大学 | Intelligent manufacturing edge calculation task scheduling method and system based on cache |
| WO2022057811A1 (en) * | 2020-09-17 | 2022-03-24 | 浙江大学 | Edge server-oriented network burst load evacuation method |
| CN115599529A (en) * | 2022-11-15 | 2023-01-13 | 阿里巴巴（中国）有限公司（Cn） | Edge cloud function computing system and method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109684075A (en) * | 2018-11-28 | 2019-04-26 | 深圳供电局有限公司 | A method of calculating task unloading is carried out based on edge calculations and cloud computing collaboration |
| CN110751293A (en) * | 2019-09-29 | 2020-02-04 | 浙江财经大学 | Cloud manufacturing multi-task scheduling optimization method based on game theory |
| CN110851277A (en) * | 2019-11-08 | 2020-02-28 | 中国石油大学(华东) | Task scheduling strategy based on edge cloud cooperation in augmented reality scene |
-
2020
- 2020-04-20 CN CN202010313381.4A patent/CN111475274B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109684075A (en) * | 2018-11-28 | 2019-04-26 | 深圳供电局有限公司 | A method of calculating task unloading is carried out based on edge calculations and cloud computing collaboration |
| CN110751293A (en) * | 2019-09-29 | 2020-02-04 | 浙江财经大学 | Cloud manufacturing multi-task scheduling optimization method based on game theory |
| CN110851277A (en) * | 2019-11-08 | 2020-02-28 | 中国石油大学(华东) | Task scheduling strategy based on edge cloud cooperation in augmented reality scene |
Non-Patent Citations (2)
| Title |
|---|
| XIAOXIA LUO等: "Joint Makespan-aware and Load Balance-aware Optimization of Task Scheduling in Cloud", 《2019 IEEE 10TH INTERNATIONAL CONFERENCE ON SOFTWARE ENGINEERING AND SERVICE SCIENCE (ICSESS)》 * |
| 专祥涛著: "《最优化方法基础》", 30 April 2018 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022057811A1 (en) * | 2020-09-17 | 2022-03-24 | 浙江大学 | Edge server-oriented network burst load evacuation method |
| US11784931B2 (en) | 2020-09-17 | 2023-10-10 | Zhejiang University | Network burst load evacuation method for edge servers |
| CN112202866B (en) * | 2020-09-25 | 2022-08-23 | 大拓无限(重庆)智能科技有限公司 | Method, device and equipment for task scheduling |
| CN112202866A (en) * | 2020-09-25 | 2021-01-08 | 大拓无限(重庆)智能科技有限公司 | Method, device and equipment for task scheduling |
| CN112702401A (en) * | 2020-12-15 | 2021-04-23 | 北京邮电大学 | Multi-task cooperative allocation method and device for power Internet of things |
| CN112650585A (en) * | 2020-12-24 | 2021-04-13 | 山东大学 | Novel edge-cloud collaborative edge computing platform, method and storage medium |
| CN112287609B (en) * | 2020-12-28 | 2021-03-30 | 之江实验室 | End, edge and cloud collaborative computing device for robot task division |
| CN112287609A (en) * | 2020-12-28 | 2021-01-29 | 之江实验室 | End, edge and cloud collaborative computing device for robot task division |
| CN112910698A (en) * | 2021-01-27 | 2021-06-04 | 网宿科技股份有限公司 | CDN coverage scheme adjusting method, device and equipment |
| CN112910698B (en) * | 2021-01-27 | 2023-08-22 | 网宿科技股份有限公司 | CDN coverage scheme adjusting method, device and equipment |
| CN113015216A (en) * | 2021-02-05 | 2021-06-22 | 浙江大学 | Burst task unloading and scheduling method facing edge service network |
| CN113342504A (en) * | 2021-07-02 | 2021-09-03 | 西安邮电大学 | Intelligent manufacturing edge calculation task scheduling method and system based on cache |
| CN113342504B (en) * | 2021-07-02 | 2023-04-21 | 西安邮电大学 | Intelligent manufacturing edge computing task scheduling method and system based on cache |
| CN115599529A (en) * | 2022-11-15 | 2023-01-13 | 阿里巴巴（中国）有限公司（Cn） | Edge cloud function computing system and method |
| CN115599529B (en) * | 2022-11-15 | 2023-03-10 | 阿里巴巴（中国）有限公司 | Edge cloud function computing system and method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111475274B (en) | 2023-04-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111475274B (en) | Cloud collaborative multi-task scheduling method and device | |
| CN111930436B (en) | Random task queuing unloading optimization method based on edge calculation | |
| Nishio et al. | Client selection for federated learning with heterogeneous resources in mobile edge | |
| CN109684075B (en) | Method for unloading computing tasks based on edge computing and cloud computing cooperation | |
| CN108809695B (en) | Distributed uplink unloading strategy facing mobile edge calculation | |
| CN109857546B (en) | Multi-server mobile edge computing unloading method and device based on Lyapunov optimization | |
| CN110493360B (en) | Mobile edge computing unloading method for reducing system energy consumption under multiple servers | |
| CN111445111B (en) | Electric power Internet of things task allocation method based on edge cooperation | |
| CN111132235B (en) | Mobile offload migration algorithm based on improved HRRN algorithm and multi-attribute decision | |
| He et al. | QoE-based cooperative task offloading with deep reinforcement learning in mobile edge networks | |
| WO2023040022A1 (en) | Computing and network collaboration-based distributed computation offloading method in random network | |
| CN110096362A (en) | A kind of multitask discharging method based on Edge Server cooperation | |
| Li et al. | Distributed task offloading strategy to low load base stations in mobile edge computing environment | |
| CN112650581A (en) | Cloud-side cooperative task scheduling method for intelligent building | |
| CN114567895A (en) | Method for realizing intelligent cooperation strategy of MEC server cluster | |
| CN113993218A (en) | Multi-agent DRL-based cooperative unloading and resource allocation method under MEC architecture | |
| CN113918240A (en) | Task unloading method and device | |
| Li et al. | Dynamic computation offloading based on graph partitioning in mobile edge computing | |
| Wu et al. | A mobile edge computing-based applications execution framework for Internet of Vehicles | |
| CN113992677A (en) | MEC calculation unloading method for delay and energy consumption joint optimization | |
| CN113590279A (en) | Task scheduling and resource allocation method for multi-core edge computing server | |
| Shu et al. | Joint offloading strategy based on quantum particle swarm optimization for MEC-enabled vehicular networks | |
| Nan et al. | A New Task Scheduling Scheme Based on Genetic Algorithm for Edge Computing. | |
| CN116489708B (en) | Meta universe oriented cloud edge end collaborative mobile edge computing task unloading method | |
| CN114615705B (en) | Single-user resource allocation strategy method based on 5G network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |