CN112491742A - Method and device for scheduling edge computing terminal resources - Google Patents

Method and device for scheduling edge computing terminal resources Download PDF

Info

Publication number
CN112491742A
CN112491742A CN202011279447.9A CN202011279447A CN112491742A CN 112491742 A CN112491742 A CN 112491742A CN 202011279447 A CN202011279447 A CN 202011279447A CN 112491742 A CN112491742 A CN 112491742A
Authority
CN
China
Prior art keywords
real
scheduling
time service
service
time
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
Application number
CN202011279447.9A
Other languages
Chinese (zh)
Other versions
CN112491742B (en
Inventor
赵瑞锋
李波
李世明
卢建刚
郭文鑫
郑文杰
刘洋
陈志伟
黎浩彬
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Electric Power Dispatch Control Center of Guangdong Power Grid Co Ltd
Original Assignee
Electric Power Dispatch Control Center of Guangdong Power Grid Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Electric Power Dispatch Control Center of Guangdong Power Grid Co Ltd filed Critical Electric Power Dispatch Control Center of Guangdong Power Grid Co Ltd
Priority to CN202011279447.9A priority Critical patent/CN112491742B/en
Publication of CN112491742A publication Critical patent/CN112491742A/en
Application granted granted Critical
Publication of CN112491742B publication Critical patent/CN112491742B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/70—Admission control; Resource allocation
    • H04L47/80—Actions related to the user profile or the type of traffic
    • H04L47/801—Real time traffic
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/70—Admission control; Resource allocation
    • H04L47/80—Actions related to the user profile or the type of traffic
    • H04L47/805—QOS or priority aware
    • 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
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

The application discloses a method and a device for scheduling edge computing terminal resources, which are divided into real-time services and non-real-time services according to the types of power distribution and utilization services, optimizes the scheduling of the non-real-time services with wider service delay requirements, computing the remaining computing resources of the terminal performing the real-time service by computing the edges within one scheduling period, comparing with required computing resource required by executing non-real-time service, establishing corresponding relation between non-real-time service and scheduling period, distributing non-real-time service to multiple scheduling periods for execution, the edge computing terminal can concentrate computing resources to complete a non-real-time service in a scheduling period, meanwhile, real-time service execution can be ensured, the condition of service error caused by overload operation can be avoided, and the technical problem of low calculation reliability of the edge calculation terminal is solved.

Description

Method and device for scheduling edge computing terminal resources
Technical Field
The application relates to the technical field of power distribution, in particular to a method and a device for scheduling edge computing terminal resources.
Background
With the widespread access of distributed energy resources and the deepening of market reformation of power distribution and utilization, new equipment and new services in a power distribution and utilization system continuously emerge, and the traditional power distribution and utilization automation system cannot meet the requirements. The power distribution and utilization internet of things is a key research direction in the field of power distribution and utilization at present and is a key platform for providing support for safe and reliable operation of a power distribution and utilization system under a new background.
The edge computing terminal is a key component in a cloud, management, edge and end architecture of the power distribution and utilization internet of things, and the edge computing terminal provides guarantee for development of power distribution and utilization services at a user side and an equipment side by means of communication, storage and computing capabilities of the edge computing terminal, but resources in the edge computing terminal are limited, the constraint of computing resources is particularly obvious, and the edge computing terminal cannot provide abundant computing resources like a cloud data center. When the power distribution and utilization system service is busy, the edge computing terminal needs to process a plurality of services at the same time, and at this time, the computing resources of the edge computing terminal cannot meet the requirements of all the services, and cannot execute all the services under the specified time delay condition, so that the service error is finally caused. Therefore, how to improve the reliability of the edge computing terminal under the limited computing resources becomes a technical problem that needs to be solved urgently by those skilled in the art.
Disclosure of Invention
The application provides a resource scheduling method and device for an edge computing terminal, which are used for solving the technical problem of low computing reliability of the edge computing terminal.
First, a first aspect of the present application provides a method for scheduling resources of an edge computing terminal, including:
dividing the power distribution and utilization services in the edge computing terminal into real-time services and non-real-time services based on the types of the power distribution and utilization services;
when the total terminal computing resource of the edge computing terminal is smaller than the required computing resource of the power distribution service, computing the total real-time service computing resource and the residual computing resource in the scheduling period according to the time period corresponding to the scheduling period, wherein the residual computing resource is the difference value between the total terminal computing resource and the total real-time service computing resource;
and establishing a corresponding relation between the non-real-time service and the scheduling period based on a comparison result of the required computing resource of the non-real-time service and the residual computing resource of the scheduling period, so that the edge computing terminal executes the corresponding non-real-time service and all real-time services in the current scheduling period.
Preferably, after the edge computing terminal executes the corresponding non-real-time service and all real-time services in the scheduling period, the method further includes:
and calculating the delay and the weighted delay of each power distribution and utilization service according to the completion time of each power distribution and utilization service in the scheduling period so as to determine a resource scheduling effect quantization parameter in the scheduling period according to each power distribution and utilization service delay and weighted delay.
Preferably, the non-real-time service includes: a plurality of service blocks and service block execution order information.
Preferably, the establishing a corresponding relationship between the non-real-time service and the scheduling cycle based on a comparison result between the required computing resource of the non-real-time service and the remaining computing resource of the scheduling cycle specifically includes:
obtaining the computing resource surplus of the non-real-time service based on the comparison result of the difference value between the required computing resource of the non-real-time service and the remaining computing resource of the scheduling period;
based on the computing resource surplus and the required computing resource of the non-real-time service, respectively obtaining a scheduling influence coefficient of each non-real-time service when the non-real-time service is executed in the scheduling period through a preset first scheduling influence coefficient computing formula, wherein the first scheduling influence coefficient computing formula specifically includes:
Figure BDA0002780250720000021
in the formula (I), the compound is shown in the specification,
Figure BDA0002780250720000022
for the scheduling impact coefficient of the ith non-real-time service,
Figure BDA0002780250720000023
comparing the difference between the required computing resource of the jth non-real-time service and the remaining computing resource of the scheduling period,
Figure BDA0002780250720000024
the amount of computing resources remaining for the ith non-real-time service,
Figure BDA0002780250720000025
the task priority coefficient of the jth non-real-time service;
and establishing a corresponding relation between the non-real-time service corresponding to the minimum value of the scheduling influence coefficient and the scheduling period according to each scheduling influence coefficient.
Preferably, based on a comparison result between the required computing resource of the non-real-time service and the remaining computing resource of the scheduling period, after the establishing of the corresponding relationship between the non-real-time service and the scheduling period, the method further includes:
and calculating the ratio of the occupation amount of the calculation resources of the real-time service to the priority coefficient of the real-time service based on the scheduling period, and performing calculation resource optimization scheduling on the real-time service with the ratio larger than a preset ratio threshold value so as to schedule the calculation resources of the real-time service to the non-real-time service.
Meanwhile, a second aspect of the present application provides an edge computing terminal resource scheduling apparatus, including:
the service classification unit is used for classifying the power distribution and utilization services in the edge computing terminal into real-time services and non-real-time services based on the types of the power distribution and utilization services;
the period computing resource counting unit is used for computing the real-time service total computing resource and the residual computing resource in the scheduling period according to the time period corresponding to the scheduling period when the terminal total computing resource of the edge computing terminal is smaller than the demand computing resource of the power distribution service, wherein the residual computing resource is the difference value between the terminal total computing resource and the real-time service total computing resource;
and the service distribution unit is used for establishing the corresponding relation between the non-real-time service and the scheduling cycle based on the comparison result of the required computing resource of the non-real-time service and the residual computing resource of the scheduling cycle, so that the edge computing terminal executes the real-time service and the corresponding non-real-time service in the current scheduling cycle.
Preferably, the method further comprises the following steps:
and the resource scheduling effect evaluation unit is used for calculating the delay and the weighted delay of each power distribution and utilization service according to the completion time of each power distribution and utilization service in the scheduling period so as to determine a resource scheduling effect quantization parameter in the scheduling period according to each power distribution and utilization service delay and weighted delay.
Preferably, the non-real-time service includes: a plurality of service blocks and service block execution order information.
Preferably, the service allocation unit specifically includes:
a calculation resource surplus counting subunit, configured to obtain the calculation resource surplus of the non-real-time service based on a comparison result of a difference between the required calculation resource of the non-real-time service and the surplus calculation resource of the scheduling period;
a scheduling influence coefficient calculation unit, configured to obtain, based on the remaining amount of the computing resources of the non-real-time service and the required computing resources, a scheduling influence coefficient when each non-real-time service is executed in the scheduling period according to a preset first scheduling influence coefficient calculation formula, where the first scheduling influence coefficient calculation formula specifically is:
Figure BDA0002780250720000031
in the formula (I), the compound is shown in the specification,
Figure BDA0002780250720000041
for the scheduling impact coefficient of the ith non-real-time service,
Figure BDA0002780250720000042
comparing the difference between the required computing resource of the jth non-real-time service and the remaining computing resource of the scheduling period,
Figure BDA0002780250720000043
the amount of computing resources remaining for the ith non-real-time service,
Figure BDA0002780250720000044
the task priority coefficient of the jth non-real-time service;
and the corresponding relation constructing subunit is used for establishing a corresponding relation between the non-real-time service corresponding to the minimum value of the scheduling influence coefficients and the scheduling cycle according to each scheduling influence coefficient, so that the edge computing terminal executes the real-time service and the corresponding non-real-time service in the current scheduling cycle.
Preferably, the method further comprises the following steps:
and the real-time service resource scheduling unit is used for calculating the ratio of the occupation amount of the calculation resources of the real-time service to the priority coefficient of the real-time service based on the scheduling period, and performing calculation resource optimization scheduling on the real-time service of which the ratio is greater than a preset ratio threshold value so as to schedule the calculation resources of the real-time service to the non-real-time service.
According to the technical scheme, the method has the following advantages:
the application provides a method for scheduling edge computing terminal resources, which comprises the following steps: dividing the power distribution and utilization services in the edge computing terminal into real-time services and non-real-time services based on the types of the power distribution and utilization services; when the total terminal computing resource of the edge computing terminal is smaller than the required computing resource of the power distribution service, computing the total real-time service computing resource and the residual computing resource in the scheduling period according to the time period corresponding to the scheduling period, wherein the residual computing resource is the difference value between the total terminal computing resource and the total real-time service computing resource; and establishing a corresponding relation between the non-real-time service and the scheduling period based on a comparison result of the required computing resource of the non-real-time service and the residual computing resource of the scheduling period, so that the edge computing terminal executes the corresponding non-real-time service and all real-time services in the current scheduling period.
The method comprises the steps of dividing the service into a real-time service and a non-real-time service according to the type of the power distribution and utilization service, optimizing the scheduling of the non-real-time service with wider service delay requirement, calculating the residual calculation resources of the edge calculation terminal executing the real-time service in one scheduling period, comparing the residual calculation resources with the required calculation resources required by executing the non-real-time service, establishing the corresponding relation between the non-real-time service and the scheduling period, and distributing the non-real-time service to a plurality of scheduling periods for execution, so that the edge calculation terminal can intensively calculate the resources to complete the non-real-time service in one scheduling period, and simultaneously can ensure the real-time service execution, thereby avoiding the condition of service errors caused by overload operation, and solving the technical problem of low calculation reliability of the edge calculation terminal.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive exercise.
Fig. 1 is a schematic flowchart of a resource scheduling method for an edge computing terminal according to a first embodiment of the present disclosure;
fig. 2 is a flowchart illustrating a resource scheduling method for an edge computing terminal according to a second embodiment of the present disclosure;
fig. 3 is a schematic structural diagram of a resource scheduling apparatus for an edge computing terminal according to a first embodiment of the present disclosure.
Detailed Description
The embodiment of the application provides a resource scheduling method and device for an edge computing terminal, which are used for solving the technical problem of low computing efficiency of computing resources of the edge computing terminal.
In order to make the objects, features and advantages of the present invention more apparent and understandable, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the embodiments described below are only a part of the embodiments of the present application, and not all of the embodiments. 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 application.
Referring to fig. 1, a first embodiment of the present application provides a method for scheduling resources of an edge computing terminal, including:
step 101, based on the type of the power distribution and utilization service, dividing the power distribution and utilization service in the edge computing terminal into a real-time service and a non-real-time service.
Firstly, the power distribution service of the power distribution system edge computing terminal is classified. The distribution and utilization service establishes corresponding models according to the characteristics of different services by real-time service (RS) and non-real-time service (NS) according to the characteristics of the service, and the model of each service comprises variables representing key characteristics of the model, so that the established service model can embody the service characteristics.
And 102, when the total terminal computing resource of the edge computing terminal is smaller than the required computing resource of the power distribution service, computing the total real-time service computing resource and the residual computing resource in the scheduling period according to the time period corresponding to the scheduling period, wherein the residual computing resource is the difference value between the total terminal computing resource and the total real-time service computing resource.
It should be noted that, because different time periods and different types or calculation amounts of the real-time services executed by the edge computing terminal exist, the present application determines, according to the time period in which the scheduling cycle is located, the real-time services to be executed by the terminal in the scheduling cycle and the total calculation resources required to be occupied for executing the real-time services, and may also determine how many calculation resources, that is, the remaining calculation resources, that can be allocated to the non-real-time services, remain after the calculation resources required for executing the real-time services are ensured.
103, establishing a corresponding relation between the non-real-time service and the scheduling cycle based on a comparison result of the required computing resource of the non-real-time service and the residual computing resource of the scheduling cycle, so that the edge computing terminal executes the corresponding non-real-time service and all real-time services in the current scheduling cycle.
Next, step 102, based on the calculated remaining computing resources of the scheduling period, comparing the required computing resources required for executing the non-real-time service, and respectively allocating the non-real-time service to the scheduling period capable of supporting the service execution thereof, i.e. establishing a corresponding relationship between the non-real-time service and the scheduling period, so that the edge computing terminal only executes the real-time service that must be executed and the non-real-time service corresponding to the scheduling period in the current scheduling period.
According to the embodiment of the application, the method is divided into a real-time service and a non-real-time service according to the type of the power distribution and utilization service, and according to the characteristics that the non-real-time service has a wider requirement on service delay and is allowed to delay execution, the residual computing resources of the real-time service are executed by computing the edge computing terminal in one scheduling period, and compared with the required computing resources required by executing the non-real-time service, the corresponding relation between the non-real-time service and the scheduling period is established, the non-real-time service is distributed to a plurality of scheduling periods to be executed, so that the edge computing terminal can intensively compute the resources in one scheduling period to complete one non-real-time service, meanwhile, the real-time service execution can be ensured, the condition of service errors caused by overload operation can be avoided, and the technical problem of low computing reliability of the edge computing terminal.
The above is a detailed description of a first embodiment of a method for scheduling edge computing terminal resources provided by the present application, and the following is a detailed description of a second embodiment of the method for scheduling edge computing terminal resources provided by the present application.
Referring to fig. 2, a second embodiment of the present application provides a method for scheduling resources of an edge computing terminal based on the first embodiment, including:
step 201, based on the type of the power distribution and utilization service, the power distribution and utilization service in the edge computing terminal is divided into a real-time service and a non-real-time service.
1) For the real-time service RS:
the model of real-time service is shown in formula (1).
Figure BDA0002780250720000071
wiRepresenting the real-time computational load of real-time traffic. Specifically, the real-time computing resource size that a certain real-time service needs to occupy in the execution process may be expressed by Million Instructions Per Second (MIPS) or similar units.
tiRepresenting the time that the real-time service needs to be executed. In particular, a certain real-time service is calculated at the edge terminal according to the corresponding wiAnd executing the service, and finishing the time required by the service execution.
pi RRepresenting the priority of real-time traffic. Specifically reflecting the importance of a certain real-time service, pi RThe larger the service is, the more important the real-time service is, and the higher the sensitivity of the service to delay is.
2) For non-real-time services NS:
the non-real-time service model is shown as formula (2).
Figure BDA0002780250720000072
WiEach element in the matrix represents the calculation load size of each service block in the non-real-time service. May be specifically represented as Wi＝{wib1 wib2 wib3 wib4 wib5W thereinibnThe computation load of the nth service block representing the ith non-real-time service may be expressed in Million Instructions (MI) or the like.
Wi rEach element in the matrix represents the calculation load size of each service block which is not completed to execute in the non-real-time service. May particularly be expressed as
Figure BDA0002780250720000073
Wherein wi rbnThe computational load of the nth service block representing the ith non-real-time service, which needs to be executed, may be expressed in terms of Million Instructions (MI) or the like. If w isi rbnAnd 0, the service block is shown to have completed execution.
EiIs a matrix, each element of which represents the current execution of the non-real-time service. May be specifically represented as EiAnd { 00100 }, wherein if the matrix element is 1, it indicates that the element corresponds to the service block in the executing state. If the matrix element is 0, the corresponding service block is in a non-execution state.
Li rRepresenting the remaining computational load of the non-real-time traffic. May particularly be expressed as
Figure BDA0002780250720000081
Of fingersThe total calculation load of the service blocks in which the non-real-time service has not been executed yet represents the total calculation load of the non-real-time service.
TiThe delay requirement representing the non-real-time service specifically refers to the maximum allowable execution time of a certain non-real-time service, and if the execution time exceeds TiThe time beyond is the time delay.
Step 202, determining whether the total computing resources of the edge computing terminal are less than the required computing resources of the electricity distribution service, if so, executing step 203, and if not, executing step 208.
It should be noted that, in the current situation, whether the computing resource of the edge computing terminal can meet the requirements of all the electricity distribution services is determined. If the conditions can be met, resource scheduling is not needed. If the resource can not be met, resource scheduling is needed. The specific implementation mode of the module is shown as formula (3):
Figure BDA0002780250720000082
where DP represents the computing power of the edge computing terminal and may be expressed in millions of instructions per second or the like. If equation (3) holds for all current non-real-time services, the edge computing terminal can meet the requirements of all services in the current situation. On the contrary, if some non-real-time service does not satisfy the condition of the formula (3), the edge computing terminal cannot satisfy the requirements of all services under the current condition, and resource scheduling is required.
And 203, calculating the real-time service total calculation resource and the residual calculation resource in the scheduling period according to the time period corresponding to the scheduling period, wherein the residual calculation resource is the difference value between the terminal total calculation resource and the real-time service total calculation resource.
It can be understood that the scheduling of this embodiment is performed in one scheduling period, where the scheduling period is divided according to available scheduling bases, and the time when the resource scheduling determination module determines that resource scheduling is required or the time when the previous scheduling period ends is the starting time of the scheduling period, and the time when a new power distribution service reaches the edge computing terminal or the time when execution of a certain non-real-time service is completed is the ending time of the scheduling period.
The specific implementation mode of the scheduling period dividing module is as follows: when the starting time condition of the dispatching cycle is met, the dispatching cycle dividing module records all current power distribution service model parameters, introduces the parameters into other three sub-modules and initializes the optimization variables. And when the condition of the end time of the scheduling period is met, the scheduling period dividing module updates all current power distribution service model parameters and introduces the parameters into other three sub-modules.
Wherein the above-mentioned optimization variables are specifically denoted as si RAnd si NRespectively representing the resource scheduling results of real-time service and non-real-time service. If the value of a certain optimization variable is equal to 1, the service corresponding to the optimization variable is in an execution state in the scheduling period, and the edge computing terminal needs to schedule the computing resource to the corresponding service in the scheduling period. If the value of a certain optimized variable is equal to 0, the service corresponding to the optimized variable is in a non-execution state in the scheduling period, and the edge computing terminal does not need to schedule the computing resource to the corresponding service in the scheduling period. For example, if s3 R1, then the corresponding real-time service RS3In the execution state in the scheduling period, the edge computing terminal needs to send RS3And scheduling the computing resources.
Step 204, obtaining the computing resource surplus of the non-real-time service based on the difference comparison result of the required computing resource of the non-real-time service and the residual computing resource of the scheduling period;
step 205, based on the remaining amount of computing resources and the required computing resources of the non-real-time service, obtaining scheduling impact coefficients of each non-real-time service when executed in a scheduling period respectively through a preset first scheduling impact coefficient computing formula, where the first scheduling impact coefficient computing formula specifically is:
Figure BDA0002780250720000091
in the formula (I), the compound is shown in the specification,
Figure BDA0002780250720000092
for the scheduling impact coefficient of the ith non-real-time service,
Figure BDA0002780250720000093
the result of the difference comparison of the required computational resource for the jth non-real-time service and the remaining computational resources of the scheduling period,
Figure BDA0002780250720000094
the amount of computing resources remaining for the ith non-real-time service,
Figure BDA0002780250720000095
the task priority coefficient of the jth non-real-time service;
and step 206, establishing a corresponding relation between the non-real-time service corresponding to the minimum value of the scheduling influence coefficients and the scheduling period according to each scheduling influence coefficient.
It should be noted that, scheduling the non-real-time service actually determines the execution result of the non-real-time service in the current scheduling period, that is, selects the non-real-time service to be executed.
The following can be referred to for specific implementation details:
firstly, optimizing variables s of all real-time servicesi RIs set to 1, i.e. all real-time traffic is performed by default.
According to the real-time service model, calculating to obtain the residual calculation resource DP of the edge calculation terminal under the condition of meeting all real-time servicesr. Because the real-time service conditions in different time periods are different, DP is usedrThe method is divided into different time periods, and is specifically represented as follows:
Figure BDA0002780250720000096
different time periods DPrThe specific calculation method is as follows:
Figure BDA0002780250720000101
computing resource shortness R for non-real-time traffici LThe specific calculation method is shown in formula (6):
Figure BDA0002780250720000102
Ri Lmeans that the completion of the non-real-time service NS is executed within the specified delay requirement in the case of executing all real-time servicesiThe missing computing resources. If R isi L>0, then NS is declarediCompletion cannot be performed within the specified latency requirements. If R isi L<0, then NS is declarediCompletion can be performed within a specified latency requirement, Ri LRepresenting redundant computing resources. Calculating scheduling evaluation index Ai NAnd select Ai NAnd taking the service with the minimum value as the non-real-time service executed in the scheduling period, and assigning the optimization variable corresponding to the service to be 1. A. thei NThe specific calculation mode is as formula (7):
Figure BDA0002780250720000103
r in the formula (5)j L+Li rRepresenting selective execution of non-real-time services NSiFor non-real-time service NSjThe influence of (c). If R isj L+Li r>0, illustrates executing the service NSiWill increase NSjThe delay of (2). If R isj L+Li r<0, illustrates executing the service NSiFor service NSjWithout effect, in this case R is firstj L+Li rThe value is assigned to 0, and A is calculatedi N. Scheduling evaluation index Ai NThe overall influence of selecting to execute a certain non-real-time service on all other non-real-time services is reflected, Ai NThe smaller the delay impact of executing the non-real-time service.
And step 207, based on the ratio of the occupation amount of the computing resources of the real-time service to the priority coefficient of the real-time service in the scheduling period, performing optimal scheduling of the computing resources on the real-time service of which the ratio is greater than a preset ratio threshold value, so as to schedule the computing resources of the real-time service to the non-real-time service.
It should be noted that, all real-time services are executed by default in the non-real-time service scheduling sub-module, and a scheduling result of the non-real-time services is obtained. The real-time service scheduling provided by this embodiment reschedules the real-time service to obtain a new real-time service execution scheme as the real-time service scheduling result in the scheduling period. The specific implementation mode is as follows:
and calculating a real-time service evaluation index e, and sequencing the real-time services according to the size of e.
The specific calculation of e is shown in equation (8):
Figure BDA0002780250720000104
e represents the ratio of the size of the calculation resource occupied by the real-time service to the priority of the real-time service, the larger e is, the more resources occupied by the real-time service are occupied, the priority is low, and the better effect is achieved by scheduling the resources to support other services from the larger e is.
Further, the scheduling evaluation index A can be calculated according to the sequence of e from large to smallj R。
If the scheduling evaluation index Aj R>If 0, the optimization variable of the corresponding real-time service is assigned to be 0, namely the real-time service is not executed, and then the calculation is continuedA of a real-time servicej R. If the scheduling evaluation index Aj RIf the value is less than or equal to 0, the real-time service is continuously executed, and A of other real-time services is not calculated any morej R. The specific calculation mode of the scheduling evaluation index is shown as formula (9):
Figure BDA0002780250720000111
wherein TeiTo RSjTo the NSiIn the case of (2), NSiThe time required for the execution to complete. Te (Te)i' to not use RSjTo the NSiIn the case of (2), NSiThe time required for the execution to complete. NS (server)iRepresenting selectively performed non-real-time traffic, TeiAnd TeiThe value of' may utilize DPrAnd (4) calculating.
And step 208, calculating the delay and the weighted delay of each power distribution and utilization service according to the completion time of each power distribution and utilization service in the scheduling period so as to determine a resource scheduling effect quantization parameter in the scheduling period according to the delay and the weighted delay of each power distribution and utilization service.
It should be noted that, after the non-real-time service scheduling and the resource optimization scheduling of the real-time service are performed, the time when each service is performed can be recorded, the delay and the weighted delay of each service are calculated, the total delay and the total weighted delay are obtained, and the effect of the resource scheduling is evaluated.
The specific calculation mode of the total delay and the total weighted delay is shown in the formulas (10) and (11):
Figure BDA0002780250720000112
Figure BDA0002780250720000113
wherein
Figure BDA0002780250720000114
teiRepresenting the time of completion of real-time service execution, TeiRepresenting the time when the non-real-time traffic execution is completed.
The above is a detailed description of a second embodiment of the method for scheduling edge computing terminal resources provided by the present application, and the following is a detailed description of a first embodiment of the apparatus for scheduling edge computing terminal resources provided by the present application.
Referring to fig. 3, a third embodiment of the present application provides an edge computing terminal resource scheduling apparatus, including:
the service classification unit 301 is configured to classify the power distribution and utilization services in the edge computing terminal into real-time services and non-real-time services based on the types of the power distribution and utilization services;
a period computing resource counting unit 302, configured to, when a total computing resource of a terminal of an edge computing terminal is smaller than a required computing resource of the power distribution service, compute a total computing resource of the real-time service and a remaining computing resource in a scheduling period according to a time period corresponding to the scheduling period, where the remaining computing resource is a difference between the total computing resource of the terminal and the total computing resource of the real-time service;
the service allocating unit 303 is configured to establish a corresponding relationship between the non-real-time service and the scheduling period based on a comparison result between the required computing resource of the non-real-time service and the remaining computing resource of the scheduling period, so that the edge computing terminal executes the real-time service and the corresponding non-real-time service in the current scheduling period.
Further, still include:
and the resource scheduling effect evaluating unit 305 is configured to calculate a delay and a weighted delay of each power distribution service according to a completion time of each power distribution service in the scheduling period, so as to determine a resource scheduling effect quantization parameter in the scheduling period according to each power distribution service delay and weighted delay.
Further, the non-real-time service includes: a plurality of service blocks and service block execution order information.
Further, the service allocating unit 303 specifically includes:
a calculation resource residual amount statistic subunit 3031, configured to obtain the calculation resource residual amount of the non-real-time service based on a comparison result of a difference between the required calculation resource of the non-real-time service and the residual calculation resource of the scheduling period;
a scheduling impact coefficient calculation subunit 3032, configured to obtain scheduling impact coefficients of each non-real-time service when executed in a scheduling period respectively through a preset first scheduling impact coefficient calculation formula based on the remaining amount of computing resources and required computing resources of the non-real-time service, where the first scheduling impact coefficient calculation formula specifically is:
Figure BDA0002780250720000121
in the formula (I), the compound is shown in the specification,
Figure BDA0002780250720000122
for the scheduling impact coefficient of the ith non-real-time service,
Figure BDA0002780250720000123
the result of the difference comparison of the required computational resource for the jth non-real-time service and the remaining computational resources of the scheduling period,
Figure BDA0002780250720000124
the amount of computing resources remaining for the ith non-real-time service,
Figure BDA0002780250720000125
the task priority coefficient of the jth non-real-time service;
the corresponding relationship establishing subunit 3033 is configured to establish a corresponding relationship between the non-real-time service corresponding to the minimum value of the scheduling impact coefficients and the scheduling period according to each scheduling impact coefficient, so that the edge computing terminal executes the real-time service and the corresponding non-real-time service in the current scheduling period.
Further, still include:
and the real-time service resource scheduling unit 304 is configured to calculate a ratio between the occupation amount of the calculation resources of the real-time service and a priority coefficient of the real-time service based on the scheduling period, and perform calculation resource optimization scheduling on the real-time service of which the ratio is greater than a preset ratio threshold, so as to schedule the calculation resources of the real-time service to the non-real-time service.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The terms "first," "second," "third," "fourth," and the like in the description of the application and the above-described figures, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the application described herein are, for example, capable of operation in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
The units described as separate parts may or may not be physically separate, and 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 units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. 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 method according to the 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.
The above embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should 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 in the embodiments of the present application.

Claims (10)

1. A method for scheduling resources of an edge computing terminal is characterized by comprising the following steps:
dividing the power distribution and utilization services in the edge computing terminal into real-time services and non-real-time services based on the types of the power distribution and utilization services;
when the total terminal computing resource of the edge computing terminal is smaller than the required computing resource of the power distribution service, computing the total real-time service computing resource and the residual computing resource in the scheduling period according to the time period corresponding to the scheduling period, wherein the residual computing resource is the difference value between the total terminal computing resource and the total real-time service computing resource;
and establishing a corresponding relation between the non-real-time service and the scheduling period based on a comparison result of the required computing resource of the non-real-time service and the residual computing resource of the scheduling period, so that the edge computing terminal executes the corresponding non-real-time service and all real-time services in the current scheduling period.
2. The method according to claim 1, wherein the edge computing terminal further includes, after executing the corresponding non-real-time service and all real-time services in the scheduling period:
and calculating the delay and the weighted delay of each power distribution and utilization service according to the completion time of each power distribution and utilization service in the scheduling period so as to determine a resource scheduling effect quantization parameter in the scheduling period according to each power distribution and utilization service delay and weighted delay.
3. The method according to claim 1, wherein the non-real-time traffic includes: a plurality of service blocks and service block execution order information.
4. The method according to claim 3, wherein the establishing the corresponding relationship between the non-real-time service and the scheduling period based on the comparison result between the required computing resource of the non-real-time service and the remaining computing resource of the scheduling period specifically comprises:
obtaining the computing resource surplus of the non-real-time service based on the comparison result of the difference value between the required computing resource of the non-real-time service and the remaining computing resource of the scheduling period;
based on the computing resource surplus and the required computing resource of the non-real-time service, respectively obtaining a scheduling influence coefficient of each non-real-time service when the non-real-time service is executed in the scheduling period through a preset first scheduling influence coefficient computing formula, wherein the first scheduling influence coefficient computing formula specifically includes:
Figure FDA0002780250710000011
in the formula (I), the compound is shown in the specification,
Figure FDA0002780250710000012
for the scheduling impact coefficient of the ith non-real-time service,
Figure FDA0002780250710000013
comparing the difference between the required computing resource of the jth non-real-time service and the remaining computing resource of the scheduling period,
Figure FDA0002780250710000021
the amount of computing resources remaining for the ith non-real-time service,
Figure FDA0002780250710000022
the task priority coefficient of the jth non-real-time service;
and establishing a corresponding relation between the non-real-time service corresponding to the minimum value of the scheduling influence coefficient and the scheduling period according to each scheduling influence coefficient.
5. The method according to claim 1, wherein after the establishing of the corresponding relationship between the non-real-time service and the scheduling period based on the comparison result between the required computing resource of the non-real-time service and the remaining computing resource of the scheduling period, the method further comprises:
and calculating the ratio of the occupation amount of the calculation resources of the real-time service to the priority coefficient of the real-time service based on the scheduling period, and performing calculation resource optimization scheduling on the real-time service with the ratio larger than a preset ratio threshold value so as to schedule the calculation resources of the real-time service to the non-real-time service.
6. An edge computing terminal resource scheduling apparatus, comprising:
the service classification unit is used for classifying the power distribution and utilization services in the edge computing terminal into real-time services and non-real-time services based on the types of the power distribution and utilization services;
the period computing resource counting unit is used for computing the real-time service total computing resource and the residual computing resource in the scheduling period according to the time period corresponding to the scheduling period when the terminal total computing resource of the edge computing terminal is smaller than the demand computing resource of the power distribution service, wherein the residual computing resource is the difference value between the terminal total computing resource and the real-time service total computing resource;
and the service distribution unit is used for establishing the corresponding relation between the non-real-time service and the scheduling cycle based on the comparison result of the required computing resource of the non-real-time service and the residual computing resource of the scheduling cycle, so that the edge computing terminal executes the real-time service and the corresponding non-real-time service in the current scheduling cycle.
7. The apparatus for scheduling resources for an edge computing terminal according to claim 6, further comprising:
and the resource scheduling effect evaluation unit is used for calculating the delay and the weighted delay of each power distribution and utilization service according to the completion time of each power distribution and utilization service in the scheduling period so as to determine a resource scheduling effect quantization parameter in the scheduling period according to each power distribution and utilization service delay and weighted delay.
8. The apparatus for scheduling resources for an edge computing terminal according to claim 6, wherein the non-real-time traffic includes: a plurality of service blocks and service block execution order information.
9. The apparatus for scheduling resource of an edge computing terminal according to claim 8, wherein the service allocating unit specifically includes:
a calculation resource surplus counting subunit, configured to obtain the calculation resource surplus of the non-real-time service based on a comparison result of a difference between the required calculation resource of the non-real-time service and the surplus calculation resource of the scheduling period;
a scheduling influence coefficient calculation unit, configured to obtain, based on the remaining amount of the computing resources of the non-real-time service and the required computing resources, a scheduling influence coefficient when each non-real-time service is executed in the scheduling period according to a preset first scheduling influence coefficient calculation formula, where the first scheduling influence coefficient calculation formula specifically is:
Figure FDA0002780250710000031
in the formula (I), the compound is shown in the specification,
Figure FDA0002780250710000032
for the scheduling impact coefficient of the ith non-real-time service,
Figure FDA0002780250710000033
computing resources and for j' th non-real-time traffic demandThe difference comparison of the remaining computational resources of the scheduling period,
Figure FDA0002780250710000034
the amount of computing resources remaining for the ith non-real-time service,
Figure FDA0002780250710000035
the task priority coefficient of the jth non-real-time service;
and the corresponding relation constructing subunit is used for establishing a corresponding relation between the non-real-time service corresponding to the minimum value of the scheduling influence coefficients and the scheduling cycle according to each scheduling influence coefficient, so that the edge computing terminal executes the real-time service and the corresponding non-real-time service in the current scheduling cycle.
10. The apparatus for scheduling resources for an edge computing terminal according to claim 6, further comprising:
and the real-time service resource scheduling unit is used for calculating the ratio of the occupation amount of the calculation resources of the real-time service to the priority coefficient of the real-time service based on the scheduling period, and performing calculation resource optimization scheduling on the real-time service of which the ratio is greater than a preset ratio threshold value so as to schedule the calculation resources of the real-time service to the non-real-time service.
CN202011279447.9A 2020-11-16 2020-11-16 Method and device for scheduling edge computing terminal resources Active CN112491742B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011279447.9A CN112491742B (en) 2020-11-16 2020-11-16 Method and device for scheduling edge computing terminal resources

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011279447.9A CN112491742B (en) 2020-11-16 2020-11-16 Method and device for scheduling edge computing terminal resources

Publications (2)

Publication Number Publication Date
CN112491742A true CN112491742A (en) 2021-03-12
CN112491742B CN112491742B (en) 2022-08-30

Family

ID=74931115

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011279447.9A Active CN112491742B (en) 2020-11-16 2020-11-16 Method and device for scheduling edge computing terminal resources

Country Status (1)

Country Link
CN (1) CN112491742B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113886094A (en) * 2021-12-07 2022-01-04 浙江大云物联科技有限公司 Resource scheduling method and device based on edge calculation
CN116302456A (en) * 2023-05-24 2023-06-23 浙江毫微米科技有限公司 Meta universe computing resource scheduling system

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1889778A (en) * 2006-07-31 2007-01-03 华为技术有限公司 High-speed down block business attachment managing method, base station and system
CN101414958A (en) * 2007-10-18 2009-04-22 华为技术有限公司 Method and apparatus for scheduling business
CN105871745A (en) * 2016-05-31 2016-08-17 浪潮集团有限公司 Mixed-service-based dispatching method in SDN network
CN106793133A (en) * 2017-01-06 2017-05-31 国网江苏省电力公司信息通信分公司 The dispatching method of multi-service QoS is ensured in a kind of electric power wireless communication system
CN110121213A (en) * 2019-03-22 2019-08-13 全球能源互联网研究院有限公司 A kind of multi-service resource dispatching method and device
CN110602747A (en) * 2019-08-05 2019-12-20 国网浙江省电力有限公司绍兴供电公司 Method for scheduling wide-band and narrow-band mixed service channel resources of power wireless communication system
CN111511031A (en) * 2020-04-27 2020-08-07 全球能源互联网研究院有限公司 Frequency allocation method and system based on power business characteristics

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1889778A (en) * 2006-07-31 2007-01-03 华为技术有限公司 High-speed down block business attachment managing method, base station and system
CN101414958A (en) * 2007-10-18 2009-04-22 华为技术有限公司 Method and apparatus for scheduling business
CN105871745A (en) * 2016-05-31 2016-08-17 浪潮集团有限公司 Mixed-service-based dispatching method in SDN network
CN106793133A (en) * 2017-01-06 2017-05-31 国网江苏省电力公司信息通信分公司 The dispatching method of multi-service QoS is ensured in a kind of electric power wireless communication system
CN110121213A (en) * 2019-03-22 2019-08-13 全球能源互联网研究院有限公司 A kind of multi-service resource dispatching method and device
CN110602747A (en) * 2019-08-05 2019-12-20 国网浙江省电力有限公司绍兴供电公司 Method for scheduling wide-band and narrow-band mixed service channel resources of power wireless communication system
CN111511031A (en) * 2020-04-27 2020-08-07 全球能源互联网研究院有限公司 Frequency allocation method and system based on power business characteristics

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113886094A (en) * 2021-12-07 2022-01-04 浙江大云物联科技有限公司 Resource scheduling method and device based on edge calculation
CN113886094B (en) * 2021-12-07 2022-04-26 浙江大云物联科技有限公司 Resource scheduling method and device based on edge calculation
CN116302456A (en) * 2023-05-24 2023-06-23 浙江毫微米科技有限公司 Meta universe computing resource scheduling system
CN116302456B (en) * 2023-05-24 2023-08-29 浙江毫微米科技有限公司 Meta universe computing resource scheduling system

Also Published As

Publication number Publication date
CN112491742B (en) 2022-08-30

Similar Documents

Publication Publication Date Title
CN112491742B (en) Method and device for scheduling edge computing terminal resources
CN104796422A (en) Online customer service staff equilibrium assignment method and online customer service staff equilibrium assignment device
CN109861850B (en) SLA-based stateless cloud workflow load balancing scheduling method
CN102662764A (en) Dynamic cloud computing resource optimization allocation method based on semi-Markov decision process (SMDP)
CN115421930B (en) Task processing method, system, device, equipment and computer readable storage medium
CN105022668B (en) Job scheduling method and system
CN111176840A (en) Distributed task allocation optimization method and device, storage medium and electronic device
CN109327844A (en) A kind of cell capacity-enlarging method and device
CN112764920A (en) Edge application deployment method, device, equipment and storage medium
CN117290090A (en) Node cooperative resource allocation method, device, equipment and storage medium
CN114780244A (en) Container cloud resource elastic allocation method and device, computer equipment and medium
CN117827456A (en) Computing power resource scheduling method and device, electronic equipment and storage medium
CN114546646A (en) Processing method and processing apparatus
CN117707763A (en) Hierarchical calculation scheduling method, system, equipment and storage medium
CN116302453B (en) Task scheduling method and device for quantum electronic hybrid platform
CN111475251A (en) Cluster container scheduling method, system, terminal and storage medium
CN115840649A (en) Method and device for allocating partitioned capacity block type virtual resources, storage medium and terminal
Ambati et al. Modeling and analyzing waiting policies for cloud-enabled schedulers
CN112003900B (en) Method and system for realizing high service availability under high-load scene in distributed system
CN114035940A (en) Resource allocation method and device
CN112714037A (en) Method, device and equipment for evaluating guarantee performance of online service quality
Pucher et al. EXFed: Efficient Cross-Federation with Availability SLAs on Preemptible IaaS Instances
CN117707797B (en) Task scheduling method and device based on distributed cloud platform and related equipment
CN118796380A (en) Resource scheduling method and device of virtual machine
Vaezpour et al. Mobile telecom cloud brokerage with orchestrated multi-tier resource pooling

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