CN111352731A - Method, system, apparatus and medium for distributing tasks in edge computing network - Google Patents
Method, system, apparatus and medium for distributing tasks in edge computing network Download PDFInfo
- Publication number
- CN111352731A CN111352731A CN202010102713.4A CN202010102713A CN111352731A CN 111352731 A CN111352731 A CN 111352731A CN 202010102713 A CN202010102713 A CN 202010102713A CN 111352731 A CN111352731 A CN 111352731A
- Authority
- CN
- China
- Prior art keywords
- edge
- task
- tasks
- computing network
- edge computing
- 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.)
- Pending
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
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
-
- 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]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
- G06F9/505—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering the load
Abstract
The invention discloses a method, a system, a device and a storage medium for distributing tasks in an edge computing network. The method for distributing tasks in the edge computing network enables the number of successfully distributed tasks in the edge computing network to be the largest on the premise of meeting constraint conditions, and distributes the tasks to the edge nodes which are most likely to be completed, so that all the edge nodes can meet task requirements and improve the efficiency of the edge computing network. The invention is widely applied to the technical field of the Internet of things.
Description
Technical Field
The invention relates to the technical field of internet of things, in particular to a method, a system, a device and a storage medium for distributing tasks in an edge computing network.
Background
The internet of things (iot), with its ability to connect a large variety of intelligent devices across a wide geographic area, has become part of the infrastructure for many advanced applications that will lead to the emergence of smart cities and other interconnected communities. This trend has driven the development of various smart device interconnect applications. However, despite the rapid advances in internet of things related technology, the development of internet of things applications is limited by limited computing resources, including CPU, storage, etc., in each internet of things device. This situation is further exacerbated by the proliferation of various internet of things applications involving resource intensive operations.
Due to the severe limitations on computing resources of internet of things devices, tasks of various applications requiring a large amount of computing resources are often offloaded to a computing system with sufficient computing resources, such as a server, a cloud system, or a data center for processing. The task unloading improves the performance of the application program of the Internet of things and reduces the energy consumption of the equipment of the Internet of things. However, using an offloading method typically results in additional overhead through data transmission over the wide area network, which increases latency and network congestion for internet of things applications, particularly those involving resource-intensive operations.
The advent of edge computing reduces the negative impact of offloading tasks to computing systems such as servers. However, edge computing also has its own drawbacks, such as limited computing resources of some edge computing devices, and load imbalance among these devices. In order to effectively mine the supporting potential of the edge computing for the application of the internet of things, it is necessary to effectively manage tasks in the edge computing network.
The currently proposed methods for task distribution in the edge computing network aim to minimize the average task completion time or minimize the system operation cost, but do not consider the resource limitation of the edge computing network and the influence of the network bandwidth sharing and security constraint on the task distribution performance.
Disclosure of Invention
In view of at least one of the above technical problems, it is an object of the present invention to provide a method, system, apparatus and storage medium for distributing tasks in an edge computing network.
The technical scheme adopted by the invention is as follows: in one aspect, an embodiment of the present invention includes a method for distributing tasks in an edge computing network, including:
generating all feasible allocation schemes according to the constraint conditions of the task allocation problem;
traversing all feasible distribution schemes to obtain an optimal distribution scheme, wherein the optimal distribution scheme is a scheme with the smallest sum of all task flows in the edge computing network;
and according to the optimal distribution scheme, distributing tasks in the edge computing network.
Further, the constraints of the task allocation problem include at least one of:
each task can be distributed to one edge node only and executed and completed at the edge node;
each edge node has a storage capacity not lower than the total data capacity of all tasks distributed to the edge node;
all virtual machines on each edge node need to meet the safety condition of the task distributed to the edge node;
the total number of tasks distributed to the edge node does not exceed the total number of virtual machines on the edge node;
the completion time of the task needs to be earlier than the deadline of the task;
the difference between the traffic bandwidth of a task flowing into an edge node and the traffic bandwidth of a task flowing out of the edge node is equal to the traffic bandwidth from an access node of the task to the edge node;
and calculating the bandwidth capacity of each edge in the network, wherein the total bandwidth of all the traffic passing through each edge does not exceed the bandwidth capacity of the edge.
Further, the optimal allocation scheme is to maximize the number of successfully distributed tasks in the edge computing network on the premise that the constraint condition is satisfied, and may be described as:
in the formula (I), the compound is shown in the specification,
for the total number of tasks, m is the total number of nodes in the edge calculation network, xk,iIndicating whether the k-th task is distributed to the ith node for execution, when xk,iIf the value is 1, the result is yes, and if the value is 0, the result is no; s.t shows the constraint condition is satisfied
The maximum value is taken.
Further, the step of performing task distribution in the edge computing network according to the optimal allocation scheme specifically includes:
calculating a probability value of a target task which is allocated to all edge nodes in an edge computing network to complete, wherein the target task is any one task in tasks needing to be allocated;
performing descending order arrangement on all edge nodes according to the probability values;
and distributing the target tasks to corresponding edge nodes to be executed according to the arrangement sequence.
Further, the step of calculating the probability value that the target task is assigned to all edge nodes in the edge computing network to complete specifically includes:
to pair
Performing linear scaling processing to make the definition domains of the formulas continuous;
and sequentially calculating the probability value of the target task distributed to all edge nodes in the edge calculation network to complete according to the formula.
Furthermore, the calculated probability value needs to be verified, if the probability value is an invalid probability value, the invalid probability value is abandoned, and then all edge nodes are arranged in a descending order again.
Further, the probability is verified using the following formula:
in the formula (I), the compound is shown in the specification,
for the total number of tasks, n is the total number of edges in the edge calculation network, fk,jCalculating the flow of the kth task passing through the jth edge of the edge calculation network;
if the formula has a feasible solution, the probability value is represented as an effective probability value, and if the formula has no feasible solution, the probability value is represented as an invalid probability value.
In another aspect, an embodiment of the present invention further includes a system for distributing tasks in an edge computing network, including:
the first processing module is used for generating all feasible distribution schemes according to the constraint conditions of the task distribution problem;
the second processing module is used for traversing all feasible distribution schemes to obtain an optimal distribution scheme, wherein the optimal distribution scheme is a scheme with the smallest sum of all task flows in the edge computing network;
and the third processing module is used for distributing tasks in the edge computing network according to the optimal distribution scheme.
In another aspect, an embodiment of the present invention further includes an apparatus for distributing tasks in an edge computing network, including a memory and a processor, where the memory is used to store at least one program, and the processor is used to load the at least one program to perform the method for distributing tasks in an edge computing network.
In another aspect, embodiments of the present invention also include a storage medium having stored therein processor-executable instructions, which when executed by a processor, are configured to perform the method of distributing tasks in an edge computing network.
The invention has the beneficial effects that: the invention provides a method for distributing tasks in an edge computing network, which maximizes the number of successfully distributed tasks in the edge computing network on the premise of meeting constraint conditions, distributes the tasks to edge nodes which are most likely to be completed, ensures that all the edge nodes can meet the task requirements and improve the efficiency of the edge computing network, fully considers the precondition of resource limitation in the edge computing network, meets the requirement of quality of service (QoS) of the tasks completed in the edge computing network, can effectively improve the number of the tasks to be completed in the edge computing network, balances the load of the edge nodes and effectively utilizes the resources in the network.
Drawings
FIG. 1 is a flowchart illustrating steps of a method for distributing tasks in an edge computing network according to an embodiment;
FIG. 2 is a pseudo-code diagram of an algorithmic implementation according to an embodiment;
FIG. 3 is a diagram illustrating a comparison of the number of tasks performed by the method, the random method, and the local method according to the embodiment of the present invention, when the number of input tasks is changed in the simulation experiment;
FIG. 4 is a diagram illustrating a comparison of the number of tasks performed by the method, the random method, and the local method according to the embodiment of the present invention, when the size of data of an input task is changed in the simulation experiment;
fig. 5 is a comparison graph of the number of tasks completed by the method, the random method, and the local method according to the embodiment of the present invention using edge computing networks with different connection degrees in the simulation experiment.
Detailed Description
Referring to fig. 1, the present embodiment includes a method of distributing tasks in an edge computing network, the method comprising the steps of:
s1, generating all feasible allocation schemes according to constraint conditions of task allocation problems;
s2, traversing all feasible distribution schemes to obtain an optimal distribution scheme, wherein the optimal distribution scheme is a scheme with the smallest sum of all task flows in the edge computing network;
and S3, distributing the tasks in the edge computing network according to the optimal distribution scheme.
The application of the internet of things needs an edge computing network to provide computing power for the edge computing network, but a certain task is allocated to a node in the edge computing network to be executed, so that the task requirement can be met, the efficiency of the edge computing network can be improved, and a task allocation scheme is needed; traversing all feasible distribution schemes to obtain an optimal distribution scheme, wherein the optimal distribution scheme is a scheme with the smallest task flow in the edge computing network; the optimal allocation scheme is to maximize the number of tasks successfully distributed in the edge computing network on the premise of satisfying the constraint condition, and can be described by the following formula:
in the formula (I), the compound is shown in the specification,
for the total number of tasks, m is the total number of nodes in the edge calculation network, xk,iIndicating whether the k-th task is distributed to the ith node for execution, when xk,iIf the value is 1, the result is yes, and if the value is 0, the result is no; s.t mean "make", i.e. make under the precondition that the constraint condition is satisfied
The maximum value is taken.
In this embodiment, the implementation of formula 1 further needs to satisfy the following 3 preconditions:
(1) all processed tasks have the same priority;
(2) all processed tasks are independent of each other;
(3) the operating systems of all virtual machines in the edge computing network are the same, as long as the virtual machines have resources to handle the incoming tasks.
In the process of obtaining the optimal allocation scheme through the formula 1, the task problem to be allocated is formalized, wherein the problem formalization symbol definition is shown in the table 1,
TABLE 1 problem formalization symbol definitions
In this embodiment, step S3, that is, the step of performing task distribution in the edge computing network according to the optimal allocation scheme, is performed by the following steps:
s301, calculating a probability value of a target task which is allocated to all edge nodes in an edge computing network to complete, wherein the target task is any one of tasks needing to be allocated;
s302, performing descending order arrangement on all edge nodes according to the probability values;
and S303, distributing the target tasks to corresponding edge nodes to execute according to the arrangement sequence.
In this embodiment, the probability value that the target task is allocated to all edge nodes in the edge computing network to complete is calculated by formula 1, before this, a task set T needs to be established first, the task set T is initialized, and the task set T that is successfully allocated is obtainedAccAnd an unallocated task set TFailSetting as an empty set, and setting the binary variable x in the formula (1)k,i∈ {0,1} are replaced by
The objective function, namely the domain of formula 1, is continuous, which is convenient for solving, that is, linear scaling processing is performed on formula (1), and then formula 1 is solved to obtain
The value of (a) is,
the value of (b) represents the probability value of the k-th task being assigned to the ith node in the network for completion, so that the probability value of the target task being assigned to all edge nodes in the edge computing network for completion can be calculated, that is, the task k is kept unchanged and is sequentially calculated
I.e. the probability value of task k assigned to all edge nodes (m) to complete is calculated in turn, e.g. the task to be assigned is known from task set T
Total sum of all
Each task sequentially calculates tasks t0The probability values assigned to all m edge nodes for completion are respectively:
then, according to the probability values obtained by calculation, all edge nodes are arranged in a descending order, and according to the arrangement order, the target task is distributed to the corresponding edge nodes to be executed, wherein the process is to carry out the task t in the task set0The most likely completed node is assigned to perform to improve the efficiency of the edge computing network. Likewise, all of the task set may be computed
Probability values that tasks are assigned to all m edge nodes to complete; then, according to the calculated probability values, all edge nodes are arranged in a descending order, and according to the arrangement order, the target tasks are distributed to the corresponding edge nodes to be executed, so that each task in the task set is distributed to the node which is most possibly completed to be executed, and further, the task set is provided with the edge nodes which are most possibly completed to be executedThe efficiency of the high edge computing network. In this embodiment, according to the arrangement order of all edge nodes, the value of the maximum probability value arranged at the head of the queue is set to 1, and the other probability values are set to 0, and the probability value obtained by calculation is further verified by a formula, which is as follows:
in the formula (I), the compound is shown in the specification,
for the total number of tasks, n is the total number of edges in the edge calculation network, fk,jCalculating the flow of the kth task passing through the jth edge of the edge calculation network;
in the process, data routing in the edge computing network is considered, if the formula 2 has a feasible solution, the probability value is represented as a valid probability value, and if the formula 2 has no feasible solution, the probability value is represented as an invalid probability value. In this embodiment, if equation 2 has a feasible solution, its corresponding probability value is obtained
Effectively, here probability value
Distributing the task k to the corresponding node i for executing the maximum probability value arranged at the head of the queue, and simultaneously distributing the task tkMoving from task set T to successfully distributed task set TAccIn represents a task tkHas been successfully allocated; if formula 2 has a feasible solution, then its corresponding probability value is obtained
Invalid, at which point the invalid probability values are discarded
The rank in the ranked order of probability values is verified in the second, i.e., the second largest, probability value substitution equation 2, and if there is a feasible solution,then task k is assigned to the node corresponding to the second highest probability value and so on until the k task is assigned to the node most likely to complete. Such as task t0The probability values assigned to all m edge nodes for completion are respectively:
wherein the probability maximum is
At this time, will
Substituting in formula 2, and judging whether formula 2 has a feasible solution, if so, then executing task t0Assigned to the 1 st node in the node set, and if there is no feasible solution, the second approximate value in the probability value ranking order is taken, such as
Substituting in formula 2, and judging whether formula 2 has a feasible solution, if so, then executing task t0Assigning to the 2 nd node in the node set, and if there is no feasible solution, taking the third rough probability value in the probability value ranking order, such as
Substituting in formula 2, and judging whether formula 2 has a feasible solution, if so, then executing task t0Assign to the 2 nd node in the node set to execute, and so on until the task t is executed0Assigned to the most likely node to complete.
In this embodiment, the constraint conditions that the task allocation problem needs to satisfy are as follows:
c-1: task allocation constraints
Each task can only be distributed to one edge node, and executed and completed at that node,
the formula is expressed as:
c-2: node storage constraints
Each edge node viThe storage capacity possessed must be equal to or greater than the total data capacity of all the tasks distributed to the node,
the formula is expressed as:
c-3: safety conditions
At edge node viAll virtual machines on must be secure enough to satisfy the security conditions of the task distributed to the node,
the formula is expressed as: x is the number ofk,ip′k≤pi(formula 5);
c-4: node sharing constraints
Distribution to edge nodes viHas a total number of tasks of at most an edge node viNumber of virtual machines h oni，
The formula is expressed as:
c-5: task completion time constraints
Completion time of task τkMust be earlier than the task's deadline deltakThe task completion time can be divided into the transmission time of data between the internet of things device and the access node, the transmission time of data between the access node and the execution node and the task completion time of the execution node,
the formula is expressed as:
c-6.1: link bandwidth conservation constraints
For executing task tkEdge node v ofiIn other words, the ingress edge node viAbout task tkTraffic bandwidth minus egress edgeNode viAbout task tkIs equal to the slave task tkAccess node a ofkTo the executing node viThe bandwidth of the traffic of (a) is,
the formula is expressed as:
c-6.2: link bandwidth constraints
All passing edges ejMust not exceed edge ejThe bandwidth capacity of (a) of (b),
the formula is expressed as:
among the above conditions, C-1, C-3, C-4, C-5, C-6.1 and C-6.2 are objective constraint conditions that must be satisfied for completing task allocation, and C-2 embodies the satisfaction of the service quality requirement of the task.
In this embodiment, the number of tasks distributed to the edge computing network needs to be maximized on the premise that the constraint conditions of formula 3, formula 4, formula 5, formula 6, formula 7, formula 8, and formula 9 are satisfied
The maximum value is taken.
In this embodiment, it is assumed that all tasks are completed at their deadlines, which has no impact on task distribution. Given executing node viOn the premise of (7):
solving equation 10 yields:
wherein the content of the first and second substances,
if it is not
Or
Then node viCan not be taken as task tkBecause the constraint of equation 7 cannot be satisfied;
if it is not
Then formula 7 is in fact equivalent to
In order for equation 13 to hold, i.e., all tasks are completed at their deadlines, this has no impact on the optimal allocation scheme; variables in this hypothesis in equation 8 below
Can be composed of constant
And (4) replacing. That is, the linearization of the constraint includes two parts: (1) for each task tkIs removed so that
Or
At this time node viCan not be taken as task tkThe execution node of (1); (2) the constraints of equations 7 and 8 are replaced by the following equations:
in this case, it is necessary to satisfy the constraint conditions of formula 3, formula 4, formula 5, formula 6, formula 9, and formula 15
The maximum value is taken to maximize the number of tasks distributed to the edge computing network.
Referring to fig. 2, equation 1, namely the implementation process of the task allocation algorithm, may encode the algorithm, namely equation 1, and the computer background replaces the algorithm, and the specific pseudo code of the algorithm is shown in fig. 2.
In summary, the method for allocating tasks in an edge computing network according to the embodiment of the present invention has the following advantages:
the embodiment of the invention provides a method for distributing tasks in an edge computing network, which maximizes the number of successfully distributed tasks in the edge computing network on the premise of meeting constraint conditions, distributes the tasks to edge nodes which are most likely to be completed, ensures that all the edge nodes can meet the task requirements and improve the efficiency of the edge computing network, simultaneously fully considers the precondition of resource limitation in the edge computing network, meets the requirement of quality of service (QoS) of the tasks completed in the edge computing network, can effectively improve the number of the tasks to be completed in the edge computing network, balances the load of the edge nodes and effectively utilizes the resources in the network.
In another aspect, the present embodiment further includes a system for distributing tasks in an edge computing network, including:
the first processing module is used for generating all feasible distribution schemes according to the constraint conditions of the task distribution problem;
the second processing module is used for traversing all feasible distribution schemes to obtain an optimal distribution scheme, wherein the optimal distribution scheme is a scheme with the smallest sum of all task flows in the edge computing network;
and the third processing module is used for distributing tasks in the edge computing network according to the optimal distribution scheme.
The system can be a server or a personal computer and the like, and the method for distributing tasks in the edge computing network can be written into a computer program and written into the server or the personal computer, so that the system for distributing tasks in the edge computing network can be obtained, and the system can be operated to realize the same technical effect as the method for distributing tasks in the edge computing network.
In this embodiment, the apparatus for distributing tasks in an edge computing network includes a memory for storing at least one program and a processor for loading the at least one program to perform the method of the embodiment.
The memory may also be separately produced and used to store a computer program corresponding to the method of allocating tasks in an edge computing network. When the memory is connected to the processor, the stored computer program is read out by the processor and executed, so as to implement the method for allocating tasks in the edge computing network, thereby achieving the technical effects described in the embodiments.
It should be noted that, unless otherwise specified, when a feature is referred to as being "fixed" or "connected" to another feature, it may be directly fixed or connected to the other feature or indirectly fixed or connected to the other feature. Furthermore, the descriptions of upper, lower, left, right, etc. used in the present disclosure are only relative to the mutual positional relationship of the constituent parts of the present disclosure in the drawings. As used in this disclosure, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. In addition, unless defined otherwise, all technical and scientific terms used in this example have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in the description of the embodiments herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this embodiment, the term "and/or" includes any combination of one or more of the associated listed items.
It will be understood that, although the terms first, second, third, etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element of the same type from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of the present disclosure. The use of any and all examples, or exemplary language ("e.g.," such as "or the like") provided with this embodiment is intended merely to better illuminate embodiments of the invention and does not pose a limitation on the scope of the invention unless otherwise claimed.
It should be recognized that embodiments of the present invention can be realized and implemented by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer readable memory. The methods may be implemented in a computer program using standard programming techniques, including a non-transitory computer-readable storage medium configured with the computer program, where the storage medium so configured causes a computer to operate in a specific and predefined manner, according to the methods and figures described in the detailed description. Each program may be implemented in a high level procedural or object oriented programming language to communicate with a computer system. However, the program(s) can be implemented in assembly or machine language, if desired. In any case, the language may be a compiled or interpreted language. Furthermore, the program can be run on a programmed application specific integrated circuit for this purpose.
Further, operations of processes described in this embodiment can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The processes described in this embodiment (or variations and/or combinations thereof) may be performed under the control of one or more computer systems configured with executable instructions, and may be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) collectively executed on one or more processors, by hardware, or combinations thereof. The computer program includes a plurality of instructions executable by one or more processors.
Further, the method may be implemented in any type of computing platform operatively connected to a suitable interface, including but not limited to a personal computer, mini computer, mainframe, workstation, networked or distributed computing environment, separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, and the like. Aspects of the invention may be embodied in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, optically read and/or write storage medium, RAM, ROM, or the like, such that it may be read by a programmable computer, which when read by the storage medium or device, is operative to configure and operate the computer to perform the procedures described herein. Further, the machine-readable code, or portions thereof, may be transmitted over a wired or wireless network. The invention described in this embodiment includes these and other different types of non-transitory computer-readable storage media when such media include instructions or programs that implement the steps described above in conjunction with a microprocessor or other data processor. The invention also includes the computer itself when programmed according to the methods and techniques described herein.
A computer program can be applied to input data to perform the functions described in the present embodiment to convert the input data to generate output data that is stored to a non-volatile memory. The output information may also be applied to one or more output devices, such as a display. In a preferred embodiment of the invention, the transformed data represents physical and tangible objects, including particular visual depictions of physical and tangible objects produced on a display.
The above description is only a preferred embodiment of the present invention, and the present invention is not limited to the above embodiment, and any modifications, equivalent substitutions, improvements, etc. within the spirit and principle of the present invention should be included in the protection scope of the present invention as long as the technical effects of the present invention are achieved by the same means. The invention is capable of other modifications and variations in its technical solution and/or its implementation, within the scope of protection of the invention.
In addition, the method described in the examples is compared with the local method and the random method. The local method refers to that the application of the Internet of things is handed to the access node of the Internet of things for processing, and does not depend on an edge computing network. The random method refers to randomly assigning a node capable of meeting task requirements as an execution node of the application of the Internet of things. To fully demonstrate the superiority of the method described in the examples, three sets of simulation experiments were performed. The first group uses different task numbers to perform simulation, the second group uses tasks with different data sizes to perform simulation, and the third group uses different edge computing network connectivity to perform simulation.
The topology of an edge computing network is generated using a Waxman model, which defines the probability of a connection between any two nodes u and v in a network, determined by four parameters, the distance d (u, v) between u and v, the distance L of the two furthest nodes in the network, and the two artificially defined parameters α, β and β, with a value range of (0, 1), the larger the β the higher the connection density of the network, the smaller the α the higher the density of short connections than long connections in the network.
In the first and second experiments, the number of nodes of the network was set to 15 and the number of connections to 80, α and β were both set to 0.8. in the third simulation experiment, the number of nodes to 15, α and β were set to 0.1 to 1 and the step size was 0.1, generating 28, 28, 36, 38, 38, 54, 66, 80, 96, 116 connections, respectively.
The three simulation experiments are operated 20 times each to obtain the average number of completed tasks, and the results are more stable through 20 repeated experiments, thereby being beneficial to comparing the three methods. The experimental results are shown in fig. 3, 4 and 5. In the simulation experiments, the number of tasks completed was used to evaluate the performance of the different methods.
(1) Simulation experiments using different numbers of tasks
In the simulation experiment, the number of the input tasks is changed from 50 to 400, and the tasks are increased by 50 each time. Referring to fig. 3, fig. 3 shows the results of the simulation, which indicates that the method described in the example accomplishes 54% more tasks than the local method and 100% more tasks than the random method. It is also found that the method according to the embodiments performs better than both of the two basic methods when the number of tasks increases, because the method according to the embodiments balances the load of the nodes and makes more efficient use of the resources in the network. The overall performance of the stochastic method is worse than that of the local method because the stochastic method delays data transmission in the edge computing network.
(2) Simulation experiments using tasks of different data sizes
In the present simulation experiment, the number of input tasks was set to 150, and the data size of the tasks was changed from 20MB to 200MB, with each increment of 20 MB. Referring to fig. 4, fig. 4 shows simulation results. The results show that the method described in the examples performed 52% more tasks than the local method and 86% more tasks than the random method, on average. The method described in the embodiments still performs well when the data size of the task increases, while the random method performs poorly because the method described in the embodiments effectively balances the load, reducing it to some extent, although the transmission load of the data increases as the data size increases.
(3) Simulation experiment using edge computing networks of different degrees of connectivity
Referring to fig. 5, fig. 5 shows simulation results, which indicate that the method of the embodiment completes 34% more tasks than the local method and 205% more tasks than the random method, and as the network connection degree increases, the performance of the method of the embodiment also increases, because the resources are more fully utilized due to the higher-degree connection, the load balancing is more effective, and the task distribution does not exist in the local method, so the performance of the network is not affected by the change of the connection degree.
Summary of the invention
Through comparison of three simulation experiments, the method in the embodiment is better than two basic methods, and meanwhile, the method in the embodiment is more superior in performance when the number of input tasks is large, the data size of the tasks is large, and the connection degree of the edge network is high.
Claims (10)
1. A method for distributing tasks in an edge computing network, comprising:
generating all feasible allocation schemes according to the constraint conditions of the task allocation problem;
traversing all feasible distribution schemes to obtain an optimal distribution scheme, wherein the optimal distribution scheme is a scheme with the smallest sum of all task flows in the edge computing network;
and according to the optimal distribution scheme, distributing tasks in the edge computing network.
2. The method of claim 1, wherein the constraints of the task assignment problem include at least one of:
each task is specified to be distributed to an edge node, and executed and completed at the edge node;
each edge node has a storage capacity not lower than the total data capacity of all tasks distributed to the edge node;
all virtual machines on each edge node need to meet the safety condition of the task distributed to the edge node;
the total number of tasks distributed to the edge node does not exceed the total number of virtual machines on the edge node;
the completion time of the task needs to be earlier than the deadline of the task;
the difference between the traffic bandwidth of a task flowing into an edge node and the traffic bandwidth of a task flowing out of the edge node is equal to the traffic bandwidth from an access node of the task to the edge node;
and calculating the bandwidth capacity of each edge in the network, wherein the total bandwidth of all the traffic passing through each edge does not exceed the bandwidth capacity of the edge.
3. The method for distributing tasks in the edge computing network according to claim 2, wherein the optimal distribution scheme is to maximize the number of tasks successfully distributed in the edge computing network on the premise that the constraint condition is satisfied, and can be described as:
s.t constraint condition
In the formula (I), the compound is shown in the specification,
for the total number of tasks, m is the total number of nodes in the edge calculation network, xk,iIndicating whether the k-th task is distributed to the ith node for execution, when xk,iIf the value is 1, the result is yes, and if the value is 0, the result is no; s.t shows the constraint condition is satisfied
The maximum value is taken.
4. The method according to claim 3, wherein the step of performing task distribution in the edge computing network according to the optimal distribution scheme specifically includes:
calculating a probability value of a target task which is allocated to all edge nodes in an edge computing network to complete, wherein the target task is any one task in tasks needing to be allocated;
performing descending order arrangement on all edge nodes according to the probability values;
and distributing the target tasks to corresponding edge nodes to be executed according to the arrangement sequence.
5. The method according to claim 4, wherein the step of calculating the probability value that the target task is assigned to all edge nodes in the edge computing network to complete specifically comprises:
to pair
Performing linear scaling processing to make the definition domains of the formulas continuous;
s.t constraint condition
And sequentially calculating the probability value of the target task distributed to all edge nodes in the edge calculation network to complete according to the formula.
6. The method as claimed in claim 4, wherein the computed probability value is further verified, and if the computed probability value is an invalid probability value, the invalid probability value is discarded, and all edge nodes are sorted again in descending order.
7. The method of claim 6, wherein the probability is verified using the following equation:
s.t constraint condition
In the formula (I), the compound is shown in the specification,
for the total number of tasks, n is the total number of edges in the edge calculation network, fk,jCalculating the flow of the kth task passing through the jth edge of the edge calculation network;
if the formula has a feasible solution, the probability value is represented as an effective probability value, and if the formula has no feasible solution, the probability value is represented as an invalid probability value.
8. A system for distributing tasks in an edge computing network, comprising:
the first processing module is used for generating all feasible distribution schemes according to the constraint conditions of the task distribution problem;
the second processing module is used for traversing all feasible distribution schemes to obtain an optimal distribution scheme, wherein the optimal distribution scheme is a scheme with the smallest sum of all task flows in the edge computing network;
and the third processing module is used for distributing tasks in the edge computing network according to the optimal distribution scheme.
9. An apparatus for distributing tasks in an edge computing network, comprising a memory for storing at least one program and a processor for loading the at least one program to perform the method for distributing tasks in an edge computing network as claimed in any one of claims 1 to 7.
10. A storage medium having stored therein processor-executable instructions for performing a method of distributing tasks in an edge computing network as claimed in any one of claims 1 to 7 when executed by a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010102713.4A CN111352731A (en) | 2020-02-19 | 2020-02-19 | Method, system, apparatus and medium for distributing tasks in edge computing network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010102713.4A CN111352731A (en) | 2020-02-19 | 2020-02-19 | Method, system, apparatus and medium for distributing tasks in edge computing network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111352731A true CN111352731A (en) | 2020-06-30 |
Family
ID=71192286
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010102713.4A Pending CN111352731A (en) | 2020-02-19 | 2020-02-19 | Method, system, apparatus and medium for distributing tasks in edge computing network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111352731A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113011009A (en) * | 2021-03-01 | 2021-06-22 | 澳门科技大学 | Parameter optimization method and device based on MoreData mechanism and storage medium |
| CN113569673A (en) * | 2021-07-16 | 2021-10-29 | 国家石油天然气管网集团有限公司西气东输分公司 | Low-power-consumption video monitoring mechanism for construction detection of engineering machinery |
| CN113590335A (en) * | 2021-08-11 | 2021-11-02 | 南京大学 | Task load balancing method based on grouping and delay estimation in tree edge network |
| CN113660508A (en) * | 2021-07-16 | 2021-11-16 | 国家石油天然气管网集团有限公司西气东输分公司 | Multi-edge computing device cooperation task allocation algorithm for intelligent video identification |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108319502A (en) * | 2018-02-06 | 2018-07-24 | 广东工业大学 | A kind of method and device of the D2D tasks distribution based on mobile edge calculations |
| CN109800072A (en) * | 2019-01-22 | 2019-05-24 | 深圳市简智联信息科技有限公司 | Task schedule optimization method and device based on edge calculations |
| CN110798858A (en) * | 2019-11-07 | 2020-02-14 | 华北电力大学（保定） | Distributed task unloading method based on cost efficiency |
-
2020
- 2020-02-19 CN CN202010102713.4A patent/CN111352731A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108319502A (en) * | 2018-02-06 | 2018-07-24 | 广东工业大学 | A kind of method and device of the D2D tasks distribution based on mobile edge calculations |
| CN109800072A (en) * | 2019-01-22 | 2019-05-24 | 深圳市简智联信息科技有限公司 | Task schedule optimization method and device based on edge calculations |
| CN110798858A (en) * | 2019-11-07 | 2020-02-14 | 华北电力大学（保定） | Distributed task unloading method based on cost efficiency |
Non-Patent Citations (1)
| Title |
|---|
| YAOZHONG SONG ET.AL: "An Approach to QoS-based Task Distribution in Edge Computing Networks for IoT Applications" * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113011009A (en) * | 2021-03-01 | 2021-06-22 | 澳门科技大学 | Parameter optimization method and device based on MoreData mechanism and storage medium |
| CN113011009B (en) * | 2021-03-01 | 2024-01-30 | 澳门科技大学 | Parameter optimization method and device based on MoreData mechanism and storage medium |
| CN113569673A (en) * | 2021-07-16 | 2021-10-29 | 国家石油天然气管网集团有限公司西气东输分公司 | Low-power-consumption video monitoring mechanism for construction detection of engineering machinery |
| CN113660508A (en) * | 2021-07-16 | 2021-11-16 | 国家石油天然气管网集团有限公司西气东输分公司 | Multi-edge computing device cooperation task allocation algorithm for intelligent video identification |
| CN113590335A (en) * | 2021-08-11 | 2021-11-02 | 南京大学 | Task load balancing method based on grouping and delay estimation in tree edge network |
| CN113590335B (en) * | 2021-08-11 | 2023-11-17 | 南京大学 | Task load balancing method based on grouping and delay estimation in tree edge network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111352731A (en) | Method, system, apparatus and medium for distributing tasks in edge computing network | |
| Xu et al. | A method based on the combination of laxity and ant colony system for cloud-fog task scheduling | |
| CN109491790B (en) | Container-based industrial Internet of things edge computing resource allocation method and system | |
| Yang et al. | A framework for partitioning and execution of data stream applications in mobile cloud computing | |
| CN109582448B (en) | Criticality and timeliness oriented edge calculation task scheduling method | |
| CN108270805B (en) | Resource allocation method and device for data processing | |
| CN111176840B (en) | Distribution optimization method and device for distributed tasks, storage medium and electronic device | |
| CN110502321A (en) | A kind of resource regulating method and system | |
| Wadhwa et al. | Optimized task scheduling and preemption for distributed resource management in fog-assisted IoT environment | |
| Armant et al. | Semi-online task assignment policies for workload consolidation in cloud computing systems | |
| CN111159859B (en) | Cloud container cluster deployment method and system | |
| CN112181613A (en) | Heterogeneous resource distributed computing platform batch task scheduling method and storage medium | |
| Madhumala et al. | Virtual machine placement using energy efficient particle swarm optimization in cloud datacenter | |
| CN112465371B (en) | Resource data distribution method, device and equipment | |
| CN113535381A (en) | Service function chain mapping method, device and equipment based on improved genetic algorithm | |
| Yassir et al. | Graph-based model and algorithm for minimising big data movement in a cloud environment | |
| CN115756772A (en) | Dynamic arrangement and task scheduling method and system for edge cloud mixed operation | |
| CN109165729A (en) | The dispatching method and system of neural network | |
| Reshmi et al. | Profit and resource availability-constrained optimal handling of high-performance scientific computing tasks | |
| Moreira et al. | Evolutionary acyclic graph partitioning | |
| Yao et al. | A Data-driven method for adaptive resource requirement allocation via probabilistic solar load and market forecasting utilizing digital twin | |
| CN115639762A (en) | Intelligent robot scheduling method and device, computing equipment and computer storage medium | |
| CN113703936A (en) | Method for creating computing power container, computing power platform, electronic device and storage medium | |
| WO2020221799A1 (en) | Apparatus and method to dynamically optimize parallel computations | |
| CN111784029A (en) | Fog node resource allocation method |
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 | ||
| CB02 | Change of applicant information |
Address after: 510630 room 1101, building 1, No.16 Keyun Road, Tianhe District, Guangzhou City, Guangdong Province (office use only) Applicant after: Guangdong Yitong Lianyun Intelligent Information Co.,Ltd. Address before: 510630 building 1101, No.16 Keyun Road, Tianhe District, Guangzhou City, Guangdong Province Applicant before: YITONG CENTURY INTERNET OF THINGS RESEARCH INSTITUTE (GUANGZHOU) Co.,Ltd. |
|
| CB02 | Change of applicant information |