CN110069341A - What binding function configured on demand has the dispatching method of dependence task in edge calculations - Google Patents
What binding function configured on demand has the dispatching method of dependence task in edge calculations Download PDFInfo
- Publication number
- CN110069341A CN110069341A CN201910286347.XA CN201910286347A CN110069341A CN 110069341 A CN110069341 A CN 110069341A CN 201910286347 A CN201910286347 A CN 201910286347A CN 110069341 A CN110069341 A CN 110069341A
- Authority
- CN
- China
- Prior art keywords
- task
- edge
- server
- edge server
- configuration
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
- G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
- G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5061—Partitioning or combining of resources
- G06F9/5072—Grid computing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Information Transfer Between Computers (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses the dispatching methods for having dependence task that binding function in a kind of edge calculations configures on demand, comprising: step 1, obtains the relevant parameter of network and task, chooses an initialization Edge Server；Step 2, greedy initial configuration is carried out to Edge Server with the relevant parameter of step 1 and obtains server configuration information；Step 3, the task in step 1 with dependence is indicated with directed acyclic graph, and topological sorting is carried out into topological sequences to task in directed acyclic graph；Step 4, is calculated by each task and runs the deadline terminated earliest on each Edge Server, obtains distribution and the scheduling scheme of task for the topological sequences iteration of step 3 with the server configuration information of step 2；Step 5, under the constraint of Edge Server actual capacity, each task is allocated and is dispatched according to the distribution and scheduling scheme of the task of step 4.This method can minimize the deadline of an application being made of multiple dependence task in edge calculations environment.
Description
Technical field
The present invention relates to what binding function in edge calculations field more particularly to a kind of edge calculations configured on demand dependence
The dispatching method of relational task.
Background technique
In recent years, with the fast development of cellular network and Internet of Things (IOT), high speed, the air interface of high reliability
So that the application of high complexity, high energy consumption is offloaded to long-range cloud data center operation, to make up mobile terminal computing capability
Deficiency and reduce its energy consumption.However, the propagation of long range inevitably leads to serious communication delay, this is unable to satisfy all
Such as augmented reality (VR), cognition auxiliary, car networking application program need the requirement of real-time response.In order to alleviate this problem, move
There is an important Paradigm Change in dynamic calculating field, from centralized cloud computing turn to edge calculations (Edge Computing,
Also mist is made to calculate, thin cloud calculates).The theory of edge calculations is at the edge (such as Wi-Fi access point or cellular base station) of internet
Small server, referred to as Edge Server are disposed, these servers have more powerful calculating and storage compared to mobile device
Ability, and geographically close to mobile subscriber, usual mobile subscriber can be directly connected with Edge Server by wireless network,
To greatly reduce communication delay, mobile subscriber is enable seamlessly to access cloud service in the case where low latency.
However the performance requirement and resource requirement due to mobile application are sharply increasing, edge calculations are in practical applications
Many challenges are faced with, such as:
(1) capacity limit and on demand configuration: compared with long-range cloud computing, the calculating of Edge Server and storage capacity all phases
To limited, the function of limited quantity can only be configured on Edge Server.Edge Server in order to run some task, need into
The corresponding database caches of row, the operation such as image download, installation and starting and additional environment configurations, these sequence of maneuvers can
Referred to as functional configuration, thus a task can only be run on the Edge Server with required function.If current edge service
Device does not have enough capacity that configuration is gone currently to the corresponding function of scheduler task, then decision to be needed to remove on part edge server
Configured function.The on-demand utilization rate for configuring the performance and Edge Server that will significantly affect mobile applications of function,
Therefore it is most important how to provide intelligent functional configuration strategy.
(2) task is relied on and executed parallel: mobile applications are made of multiple the task of dependence, usually with one
A directed acyclic graph (DAG) indicates.Point in figure represents different types of task, and the value on directed edge represents a task knot
Need to transmit the input that certain data volume is directed toward task as arrow after beam, so the side collection in figure also defines task execution
Successive or concurrency relation.In addition, different tasks may have different preferences, such as a Facebook to Edge Server
Video processing applications, encoding operation is computation-intensive task, is more suitable for being placed on the more powerful Edge Server of operational performance
On.In order to minimize the deadline of application as far as possible, how to design reasonable scheduling strategy is problem to be solved, packet
Include the sequence that each task in decision DAG is placed individually into task execution on which Edge Server and each Edge Server.
, there is the work of numerous studies task schedule and functional configuration problem in current mobile edge calculations field, but
Some algorithms do not account for the dependence of task in application program, and assume that applying is one independent and inseparable whole
Body.As mobile application is increasingly complicated, by wherein can parallel task be assigned on different Edge Servers operation can be effective
Optimize the performance of mobile application.But in resource-constrained edge calculations environment, how to carry out functional configuration and closed to having to rely on
The task schedule of system is urgent problem.
Summary of the invention
Based on the problems of prior art, the object of the present invention is to provide binding function in a kind of edge calculations is on-demand
The dispatching method for having dependence task of configuration, can solve task schedule in existing edge calculations and does not account in application program
The dependence of task causes in edge calculations using the problem that operational efficiency is not high.
The purpose of the present invention is what is be achieved through the following technical solutions:
Embodiment of the present invention provides the tune for having dependence task that binding function configures on demand in a kind of edge calculations
Degree method, comprising:
Step 1, edge calculations network and the relevant parameter comprising the application with dependence task are obtained, from edge meter
It calculates and chooses the initialization server that an Edge Server is output and input as the processing application in network；
Step 2, the relevant parameter of the application obtained using the step 1 takes each edge in the edge calculations network
Business device carries out greedy initial configuration and obtains server configuration information；
Step 3, the task with dependence of the application in the step 1 is indicated with directed acyclic graph, and to described
Task in directed acyclic graph carries out topological sorting, obtains the topological sequences of task；
Step 4, the topological order for the task that the server configuration information obtained using the step 2 obtains the step 3
Column iterative calculation, each task in topological sequences that calculates are placed on each Edge Server of edge calculations network and run earliest
The deadline of end simultaneously stores corresponding assigning process, each point stored according to the last one task completion time reverse search
Distribution and the scheduling scheme of all tasks are reversely rebuild with process；
Step 5, under the constraint of Edge Server actual capacity, the distribution of finally determining according to the step 4 for task and
Scheduling scheme is allocated and dispatches to each task.
As seen from the above technical solution provided by the invention, function is combined in edge calculations provided in an embodiment of the present invention
What can be configured on demand has the dispatching method of dependence task, it has the advantage that:
The method of the invention realizes the on-demand configurations of decision making function and each task to be placed individually into which edge clothes
Task executes sequence on business device and each Edge Server.Limited quantity on Edge Server can be considered in edge calculations
Functional configuration under the premise of, efficient scheduling and the on-demand configuration of function to the task with dependence improve edge service
The utilization rate of device realizes the deadline that an application being made of multiple dependence task is minimized in edge calculations environment
(deadline refers to using after Edge Server or remote cloud server unloading, obtains final operation result and is back to movement
The time of user), it reduces and applies runing time, with method (such as topcuoglu proposition used after other modifications under this scene
HEFT algorithm) compare, can be reduced 1.54~2.8 times apply the deadline, improve user experience.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, required use in being described below to embodiment
Attached drawing be briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this
For the those of ordinary skill in field, without creative efforts, it can also be obtained according to these attached drawings other
Attached drawing.
Fig. 1 is the tune for having dependence task that binding function configures on demand in edge calculations provided in an embodiment of the present invention
The flow chart of degree method；
Fig. 2 is the illustraton of model of configuration method provided in an embodiment of the present invention；
Fig. 3 is the directed acyclic graph structures schematic diagram for three kinds of applications that present invention implementation provides；
Fig. 4 is the implementation performance comparison diagram for the method that present invention implementation provides.
Specific embodiment
Below with reference to particular content of the invention, technical solution in the embodiment of the present invention is clearly and completely retouched
It states, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Based on the present invention
Embodiment, every other embodiment obtained by those of ordinary skill in the art without making creative efforts,
Belong to protection scope of the present invention.The content being not described in detail in the embodiment of the present invention belongs to professional and technical personnel in the field
The well known prior art.
As shown in Figure 1, the embodiment of the present invention provide that binding function in a kind of edge calculations configures on demand have dependence
The dispatching method of task is able to achieve the completion that an application being made of multiple dependence task is minimized in edge calculations environment
Time, and then improve in edge calculations using the efficiency of operation, promote user experience, comprising:
Step 1, edge calculations network and the relevant parameter comprising the application with dependence task are obtained, from edge meter
It calculates and chooses initialization server of the Edge Server as application in network；
Step 2, using the step 1 obtain relevant parameter to each Edge Server in the edge calculations network into
Row greed initial configuration obtains server configuration information；
Step 3, the task with dependence in the step 1 is indicated with directed acyclic graph, and to the oriented nothing
Task in ring figure carries out topological sorting, obtains the topological sequences of task；
Step 4, the topological order for the task that the server configuration information obtained using the step 2 obtains the step 3
Column iterative calculation, each task in topological sequences that calculates are placed on each Edge Server of edge calculations network and run earliest
The deadline of end simultaneously stores corresponding process, according to each process that the last one task completion time reverse search is stored come
Reversely rebuild distribution and the scheduling scheme of all tasks；
Step 5, under the constraint of Edge Server actual capacity, the distribution of finally determining according to the step 4 for task and
Scheduling scheme is allocated and dispatches to each task.
In the step 1 of the above method, edge calculations network includes:
The Edge Server of one remote cloud server and multiple isomeries, each Edge Server have limited capacity,
In, the bidirectional data transfers rate between any two Edge Server is equal.
In the step 1 of the above method, the relevant parameter for obtaining edge calculations network and the task with dependence includes:
Runing time and each Edge Server configuration of each task on each Edge Server of edge calculations network are not
The time that congenerous need to be spent.
In the step 1 of the above method, the server that greedy initial configuration starts be by all tasks on all servers
What runing time minimum value determined.This initialization server is to simulate some mobile device for one calculating task in fact
(such as application of a recognition of face) is unloaded to some Edge Server (being exactly initialization server herein), it is desirable to borrow
Help the resource of initialization server that it is helped faster to complete the calculating task, mobile device can provide required for calculating task
Input data, and final calculated result is returned to by the initialization server.
The relevant parameter of the step 2 of the above method obtained using the step 1 is to the side in the edge calculations network
Edge server carries out greedy initial configuration and obtains in server configuration information, and a corresponding number is arranged on each Edge Server
The number of group record configuration feature, the information of storage of array is the server configuration information obtained, and configuration process includes following
Step:
Step 21, under the premise of ignoring the actual capacity of Edge Server, greedy ensures each task in its operation
Good corresponding function is configured on time least Edge Server, and the number of function is recorded in the array of the Edge Server
In, and calculate under current-configuration, maximum capacity cost value is as virtual capacity in Edge Server；
Step 22, the capacity of all Edge Servers is set as virtual capacity, to the edge of non-full configuration in the step 21
Server continues to configure as follows: the runing time by all tasks in all Edge Servers sorts from small to large, successively
Judge on Edge Server corresponding to runing time that whether full configuration if full configuration skips subsequent step, if non-full configuration,
Judge whether the corresponding function of configuration task if configured skips subsequent step for the Edge Server of non-full configuration, otherwise,
It is configured and is saved in after array and jumped to next runing time again and judged, until all Edge Servers are matched
It is full.
The processing of above-mentioned steps 21 is specifically to count each Edge Server to be respectively necessary for spending under greedy allocation plan
How many capacity, maximum capacity cost value is virtual capacity (when a function occupies the capacity of a unit in Edge Server
When, maximum configured function number is equal to maximum capacity consumption number).
In above-mentioned steps 22, non-full configuration refers to that the function number of configuration is less than the virtual capacity of Edge Server；
Full configuration refers to that the function number of configuration is equal to the virtual capacity of side server.
The processing of above-mentioned steps 22 is specifically to continue functional configuration on the Edge Server of non-full configuration under step 21,
So that each Edge Server runs out of its virtual capacity, can be recorded on each Edge Server with a two-dimensional array in practice
The function of configuration.
In the step 3 of the above method, the task with dependence applied in the step 1 is indicated with directed acyclic graph
Include:
In task with dependence, a computing module in the application of each task presentation will have dependence
Task setting be a set of tasksWherein, task vjIn Edge Server skRuning time be
tjk；
With a directed acyclic graphIt indicates an application, sets directed edge e:=(vi,vj) ∈ ε expression task
vjExecution need task vi, result as input, wherein the data volume transmitted be wij。
In above-mentioned steps 4, distribution and scheduling scheme to task are obtained in dynamic programming process, dynamic programming method
It is middle can memory storage subproblem as a result, (namely the last one task) is minimized when iterating to final step
Deadline can record each task under the deadline with array from the reversed reconstruction tasks distribution of the result and scheduling scheme
Distribution and each server on task scheduling sequence, and then distribution and the scheduling scheme of finally being determined for task.
The method of the invention realizes the on-demand configurations of decision making function and each task to be placed individually into which edge clothes
Task executes sequence on business device and each Edge Server.Limited quantity on Edge Server can be considered in edge calculations
Functional configuration under the premise of, efficient scheduling and the on-demand configuration of function to the task with dependence improve edge service
The utilization rate of device realizes the deadline that an application being made of multiple dependence task is minimized in edge calculations environment
(deadline refers to using after Edge Server or remote cloud server unloading, obtains final operation result and is back to movement
The time of user), it reduces and applies runing time, improve user experience.
The embodiment of the present invention is specifically described in further detail below.
What binding function configured on demand has the dispatching method of dependence task in the edge calculations of the embodiment of the present invention, is
A kind of dispatching method for having dependence task of binding function configuration, comprising: model definition and processing step；
Wherein, network environment, each model used in (1) dispatching method are defined as follows:
(11) edge calculations network: the edge calculations network of method application of the invention is edge cloud system, wherein there is K
The Edge Server of isomery is usedIt indicates；Wherein, each Edge Server skThere is limited capacity Ck；Side
Edge server siTo sjData transmission rate be dij, set dij=dji；It include a remote cloud server s in edge cloud systemK。
(12) task dependency graph: a computing module in each application is a task, the multiple tasks of each application
With dependence, multiple set of tasks of an application are setWherein, task vjIt is taken at edge
Be engaged in device skRuning time tjkIt indicates；With a directed acyclic graph (DAG)To indicate that one is applied, and is set with
To side e:=(vi,vj) ∈ ε expression task vjExecution need task vi, result as input, wherein the data volume transmitted is
wij。
(13) server configures: task viIt can only run on the Edge Server for be configured with corresponding function, be taken at edge
Be engaged in device sjFor task viConfiguration is carried out to need to spend time rij；Set the capacity that each function accounts for a unit, Edge Server si
Last time can be to multi-configuration CiA function, default server can open an example (alternatively referred to as line for configured function
Journey) to handle corresponding task；If there is no enough capacity configuration new functions on Edge Server, decision is needed to throw away certain
A little configured functions, while terminating the corresponding example of thrown away function.
(14) directed acyclic graph simplifies: in order to simplify using expression, an empty node is added, and generate directed edge direction to have
Using initialization data amount, in addition all Ingress nodes (in-degree 0) into acyclic figure, the volume of transmitted data on these sides are
One empty node is used to collect the result of all end nodes (out-degree 0)；Empty node prevents take up any capacity, is also not carried out
Time, two empty nodes can only be placed on the same Edge Server, indicate that the initialization server of task requests is also it
Receive the Edge Server of final implementing result.The addition of this step is because the mobile device of user generally can answering calculating
With being unloaded to it apart from nearest Edge Server (being known as initialization server in the application), while the initial defeated of application
Enter data and be sent to initialization server, when scheduling, if initialization server decision (namely needs Ingress node
Original input data task as input) it has been placed on other Edge Server, that original input data to be considered that is bound to
From initialization server transport to other Edge Server bring communication delay.Otherwise last result is also serviced by initialization
Device returns to the mobile device of user, simplifies to have to utilize to directed acyclic graph and portrays this process.
The specific processing of this method includes the following steps:
Step 1, input model defines the information of middle parameters, comprising: network, directed acyclic graph, each task are each
The time that runing time and each Edge Server configuration different function on a server need to spend, and specify the application
Ingress node where server.
Step 2, the runing time letter using the information, in particular to each task of above-mentioned steps 1 on different server
Breath carries out greedy initial configuration to the Edge Server in network:
Step 21, under the premise of not considering that the actual capacity of Edge Server constrains, greedy ensures that each task exists
It is configured on the least server of its runing time, and calculates the maximum value that capacity is spent in Servers-all and (namely take
The function number of maximum configured on business device), which is referred to as virtual capacity；
Step 22, the capacity of all Edge Servers is set as virtual capacity, in above-mentioned steps 21 be not up to full configuration (
The function number exactly configured be less than virtual capacity) Edge Server continue to configure: by all tasks at all edges
The runing time of server sorts from small to large, successively judges runing time tjkCorresponding Edge Server skOn whether
Full configuration skips subsequent step if full configuration, if non-full configuration, judges Edge Server skWhether configuration task viCorresponding function
Can, if configured, subsequent step is skipped, jump to next runing time again and judged with postponing, until all sides
Edge server is matched full.The configuration also ensure do not allow on demand configure in the case where (namely server configuration function not
In the case where changing again) each task has executable Edge Server (this is the necessary condition that step 4 is run)；
Step 3, topological sorting is carried out to the directed acyclic graph inputted in above-mentioned steps 1, obtains the topological sequences of task；It is fixed
Adopted parameter fijExpression task viPlace server sjIt is upper to run the time terminated earliest.It is initialized as fij:=∞, for all
1≤i≤J and 1≤j≤K；
Step 4, using the server configuration information of above-mentioned steps 2 and the definition of step 3, according to topological sequences iteration,
Using the method for Dynamic Programming, f is obtained step by stepijValue, Until calculating the last one task fJaDeadline, and obtain distribution and the scheduling scheme of task；
Step 5, under the constraint of server actual capacity, the task of application is allocated and is adjusted according to the result of step 4
Degree was calculated using the real deadline.Consider that actual capacity is often below virtual capacity, expires under virtual capacity to certain
Foot directed acyclic graph relies on can directly running after constraint for task and also needs to spend additional queuing under actual capacity limitation
Waiting time and functional configuration time, therefore really task completion time is not the f that step 4 obtainsJa, but fJaIt adds
Certain tasks are because of queue waiting time and the time of functional configuration.Wherein, queue waiting time is when task has in satisfaction
When can be run under to the time-constrain of acyclic figure, need to wait the task available free on the server (untreated any task)
The time that example occurs.The functional configuration time refers to is configured the corresponding function of the task on demand, and replaces free instance
The time of corresponding function.
Embodiment
What binding function configured on demand has the dispatching party of dependence task in edge calculations provided in an embodiment of the present invention
Method specifically comprises the following steps:
Step 1, the directed acyclic graph of three kinds of applications of the edge calculations network and Fig. 3 of input Fig. 2 settingProvide the Edge Server s using initializationa；
Step 2, greedy initial configuration is carried out to the Edge Server in network using the information of above-mentioned steps 1, ignores side
The self-capacity of edge server, it is greedy ensure each task on the least server of its runing time with postponing, will be virtual
Capability value is set as Servers-all to spend the maximum value of capacity and set the capacity of Edge Server being the virtual capacity value.?
Under the setting of this virtual capacity, most to the non-repetitive option and installment of Edge Server greed that can be configured there are also residual capacity
The task of small runing time, until all servers are matched completely.The configuration also ensures in the case where not allowing to configure on demand
Each task has executable server；
Step 3, topological sorting is carried out to directed acyclic graph in above-mentioned steps 1, obtains the topological sequences of task.Defined parameters
fijExpression task viPlace server sjIt is upper to run the time terminated earliest.It is initialized as fij:=∞, for 1 all≤i≤
J and 1≤j≤K；
Step 4, Dynamic Programming is used according to topological sequences iteration using 2 server configuration information and 3 definition
Method, obtain f step by stepijValue,Until calculating
The last one task fJaDeadline, and obtain distribution and the scheduling scheme of task；
Step 5, it using the task distribution obtained in above-mentioned steps 4 and dispatching sequence's scheme, is provided again in above-mentioned steps 1
Actual edge calculate and calculated in network using the deadline, when task run, lacks some function and is then configured on demand,
In addition the true deadline can be calculated after configuration and queue waiting time.
In the edge calculations network of above-mentioned Fig. 2 signal, the edge of 3 Edge Servers and long-distance cloud composition is described
Cloud system, wherein some functions of Edge Server assignment configuration, one applied in Edge Server s1Initialization, by this hair
Task 1 and 3 in DAG figure is placed on s by clear and decided plan1Upper operation, and task 2 is placed on s2, task 4 is placed on s3.Due to s2On not
The corresponding function of configuration task 2, when task 2 is run, server s2It needs to download the function from cloud and to replace task 1 right
The function of answering.
Fig. 4 illustrates the implementation performance comparison diagram of the method for the present invention；Abscissa " chain query " correspondence is looked into the Fig. 4
Ask chain application, " Video Processing " corresponding video processing applications, " CDA " corresponding complex data analysis application.The present invention
The algorithm ALG-ODM of proposition reduces at least 2.8,2.28,1.54 times of deadline under these three applications.
Edge calculations network configuration applied by the present embodiment is as follows:
Those of ordinary skill in the art will appreciate that: realizing that all or part of the process in above-described embodiment method is can be with
Relevant hardware is instructed to complete by program, the program can be stored in a computer-readable storage medium, should
Program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, the storage medium can be magnetic disk, light
Disk, read-only memory (Read-Only Memory, ROM) or random access memory (Random Access Memory,
RAM) etc..
The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto,
Within the technical scope of the present disclosure, any changes or substitutions that can be easily thought of by anyone skilled in the art,
It should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be with the protection model of claims
Subject to enclosing.
Claims (6)
1. what binding function configured on demand has the dispatching method of dependence task in a kind of edge calculations, which is characterized in that packet
It includes:
Step 1, edge calculations network and the relevant parameter comprising the application with dependence task are obtained, from edge calculations net
The initialization server that an Edge Server is output and input as the processing application is chosen in network；
Step 2, the relevant parameter of the application obtained using the step 1 is to each Edge Server in the edge calculations network
It carries out greedy initial configuration and obtains server configuration information；
Step 3, the task with dependence of the application in the step 1 is indicated with directed acyclic graph, and to described oriented
Task in acyclic figure carries out topological sorting, obtains the topological sequences of task；
Step 4, the server configuration information obtained using the step 2 changes to the topological sequences for the task that the step 3 obtains
In generation, calculates, and each task in topological sequences that calculates is placed on operation on each Edge Server of edge calculations network to be terminated earliest
Deadline and store corresponding assigning process, respectively distributed according to what the last one task completion time reverse search was stored
Journey reversely rebuilds distribution and the scheduling scheme of all tasks；
Step 5, under the constraint of Edge Server actual capacity, according to the distribution and scheduling of the task that the step 4 finally determines
Scheme is allocated and dispatches to each task.
2. what binding function configured on demand has the dispatching party of dependence task in edge calculations according to claim 1
Method, which is characterized in that in the step 1 of the method, edge calculations network includes:
The Edge Server of one long-distance cloud and multiple isomeries, each Edge Server have limited capacity, wherein any two
Bidirectional data transfers rate between Edge Server is equal.
3. what binding function configured on demand has the scheduling of dependence task in edge calculations according to claim 1 or 2
Method, which is characterized in that in the step 1 of the method, obtain edge calculations network and include answering with dependence task
Relevant parameter includes:
Runing time and each Edge Server of each task on each Edge Server of edge calculations network configure different function
The time that can need to be spent.
4. what binding function configured on demand has the scheduling of dependence task in edge calculations according to claim 1 or 2
Method, which is characterized in that the relevant parameter for the application of the step 2 of the method obtained using the step 1 is to the edge
The greedy initial configuration of Edge Server progress calculated in network obtains in server configuration information, sets on each Edge Server
Set the number of a corresponding array record configuration feature, configuration process the following steps are included:
Step 21, under the premise of ignoring the actual capacity of Edge Server, greedy ensures each task in its runing time
Good corresponding function is configured on least Edge Server, and the number of function is recorded in the array of the Edge Server, and
It calculates under current-configuration, maximum capacity cost value is as virtual capacity in Edge Server；
Step 22, the capacity of all Edge Servers is set as virtual capacity, to the edge service of non-full configuration in the step 21
Device continues to configure as follows: the runing time by all tasks in all Edge Servers sorts from small to large, successively judges
Whether full configuration if full configuration skips subsequent step on Edge Server corresponding to runing time, if non-full configuration, judgement
Whether the corresponding function of configuration task if configured skips subsequent step to the Edge Server of non-full configuration, otherwise, carries out
Configure and be saved in after array and jump to next runing time again and judged, until all Edge Servers match it is full.
5. what binding function configured on demand has the dispatching party of dependence task in edge calculations according to claim 4
Method, which is characterized in that in the step 22, non-full configuration refers to that the function number of configuration is less than the virtual capacity of Edge Server；
Full configuration refers to that the function number of configuration is equal to the virtual capacity of Edge Server.
6. what binding function configured on demand has the scheduling of dependence task in edge calculations according to claim 1 or 2
Method, which is characterized in that in the step 3 of the method, indicate that having for the application in the step 1 relies on directed acyclic graph
The task of relationship includes:
In task with dependence, a computing module in the application of each task presentation, by appointing with dependence
Business is set as a set of tasks v={ v1, v2..., vJ, wherein task vjIn Edge Server skRuning time be tjk；
With a directed acyclic graphIt indicates an application, sets directed edge e:=(vi, vj) ∈ ε expression task vj's
Execution needs task vi, result as input, wherein the data volume transmitted be wij。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910286347.XA CN110069341B (en) | 2019-04-10 | 2019-04-10 | Method for scheduling tasks with dependency relationship configured according to needs by combining functions in edge computing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910286347.XA CN110069341B (en) | 2019-04-10 | 2019-04-10 | Method for scheduling tasks with dependency relationship configured according to needs by combining functions in edge computing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110069341A true CN110069341A (en) | 2019-07-30 |
| CN110069341B CN110069341B (en) | 2022-09-06 |
Family
ID=67367446
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910286347.XA Active CN110069341B (en) | 2019-04-10 | 2019-04-10 | Method for scheduling tasks with dependency relationship configured according to needs by combining functions in edge computing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110069341B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110650194A (en) * | 2019-09-23 | 2020-01-03 | 中国科学技术大学 | Task execution method based on edge calculation in computer network |
| CN110740194A (en) * | 2019-11-18 | 2020-01-31 | 南京航空航天大学 | Micro-service combination method based on cloud edge fusion and application |
| CN111756812A (en) * | 2020-05-29 | 2020-10-09 | 华南理工大学 | Energy consumption perception edge cloud cooperation dynamic unloading scheduling method |
| CN111930487A (en) * | 2020-08-28 | 2020-11-13 | 北京百度网讯科技有限公司 | Job flow scheduling method and device, electronic equipment and storage medium |
| CN113031522A (en) * | 2019-12-25 | 2021-06-25 | 沈阳高精数控智能技术股份有限公司 | Low-power-consumption scheduling method suitable for periodically dependent tasks of open type numerical control system |
| CN113986553A (en) * | 2021-11-04 | 2022-01-28 | 中国电信股份有限公司 | Model caching method and device based on mobile edge calculation, medium and equipment |
| CN115037956A (en) * | 2022-06-06 | 2022-09-09 | 天津大学 | Traffic scheduling method for cost optimization of edge server |
| WO2022236834A1 (en) * | 2021-05-14 | 2022-11-17 | Alipay (Hangzhou) Information Technology Co., Ltd. | Method and system for scheduling tasks |
| CN116880994A (en) * | 2023-09-07 | 2023-10-13 | 之江实验室 | Multiprocessor task scheduling method, device and equipment based on dynamic DAG |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018095537A1 (en) * | 2016-11-25 | 2018-05-31 | Nokia Technologies Oy | Application provisioning to mobile edge |
| CN109561148A (en) * | 2018-11-30 | 2019-04-02 | 湘潭大学 | Distributed task dispatching method in edge calculations network based on directed acyclic graph |
-
2019
- 2019-04-10 CN CN201910286347.XA patent/CN110069341B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018095537A1 (en) * | 2016-11-25 | 2018-05-31 | Nokia Technologies Oy | Application provisioning to mobile edge |
| CN109561148A (en) * | 2018-11-30 | 2019-04-02 | 湘潭大学 | Distributed task dispatching method in edge calculations network based on directed acyclic graph |
Non-Patent Citations (4)
| Title |
|---|
| H. TOPCUOGLU: "Performance-effective and low-complexity task scheduling for heterogeneous computing", 《 IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS》 * |
| HAISHENG TAN: "Online job dispatching and scheduling in edge-clouds", 《IEEE INFOCOM 2017 - IEEE CONFERENCE ON COMPUTER COMMUNICATIONS》 * |
| 赵磊: "面向移动边缘计算的边缘服务器部署及资源分配研究", 《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》 * |
| 邹云峰等: "边缘计算环境下服务质量感知的资源调度机制", 《电子技术与软件工程》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110650194A (en) * | 2019-09-23 | 2020-01-03 | 中国科学技术大学 | Task execution method based on edge calculation in computer network |
| CN110740194A (en) * | 2019-11-18 | 2020-01-31 | 南京航空航天大学 | Micro-service combination method based on cloud edge fusion and application |
| CN110740194B (en) * | 2019-11-18 | 2020-11-20 | 南京航空航天大学 | Micro-service combination method based on cloud edge fusion and application |
| CN113031522A (en) * | 2019-12-25 | 2021-06-25 | 沈阳高精数控智能技术股份有限公司 | Low-power-consumption scheduling method suitable for periodically dependent tasks of open type numerical control system |
| CN111756812B (en) * | 2020-05-29 | 2021-09-21 | 华南理工大学 | Energy consumption perception edge cloud cooperation dynamic unloading scheduling method |
| CN111756812A (en) * | 2020-05-29 | 2020-10-09 | 华南理工大学 | Energy consumption perception edge cloud cooperation dynamic unloading scheduling method |
| CN111930487A (en) * | 2020-08-28 | 2020-11-13 | 北京百度网讯科技有限公司 | Job flow scheduling method and device, electronic equipment and storage medium |
| CN111930487B (en) * | 2020-08-28 | 2024-05-24 | 北京百度网讯科技有限公司 | Job stream scheduling method and device, electronic equipment and storage medium |
| WO2022236834A1 (en) * | 2021-05-14 | 2022-11-17 | Alipay (Hangzhou) Information Technology Co., Ltd. | Method and system for scheduling tasks |
| CN113986553A (en) * | 2021-11-04 | 2022-01-28 | 中国电信股份有限公司 | Model caching method and device based on mobile edge calculation, medium and equipment |
| CN115037956A (en) * | 2022-06-06 | 2022-09-09 | 天津大学 | Traffic scheduling method for cost optimization of edge server |
| CN115037956B (en) * | 2022-06-06 | 2023-03-21 | 天津大学 | Traffic scheduling method for cost optimization of edge server |
| CN116880994A (en) * | 2023-09-07 | 2023-10-13 | 之江实验室 | Multiprocessor task scheduling method, device and equipment based on dynamic DAG |
| CN116880994B (en) * | 2023-09-07 | 2023-12-12 | 之江实验室 | Multiprocessor task scheduling method, device and equipment based on dynamic DAG |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110069341B (en) | 2022-09-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110069341A (en) | What binding function configured on demand has the dispatching method of dependence task in edge calculations | |
| CN111835827B (en) | Internet of things edge computing task unloading method and system | |
| Kaur et al. | Container-as-a-service at the edge: Trade-off between energy efficiency and service availability at fog nano data centers | |
| Téllez et al. | A tabu search method for load balancing in fog computing | |
| CN108804227B (en) | Method for computing-intensive task unloading and optimal resource allocation based on mobile cloud computing | |
| CN111541760B (en) | Complex task allocation method based on server-free mist computing system architecture | |
| CN102662764B (en) | A kind of dynamic cloud computational resource optimizing distribution method based on SMDP | |
| CN110519370B (en) | Edge computing resource allocation method based on facility site selection problem | |
| CN102857548A (en) | Mobile cloud computing resource optimal allocation method | |
| Mostafavi et al. | A stochastic approximation approach for foresighted task scheduling in cloud computing | |
| Maray et al. | Dependent task offloading with deadline-aware scheduling in mobile edge networks | |
| Wu et al. | A mobile edge computing-based applications execution framework for Internet of Vehicles | |
| CN112799823A (en) | Online dispatching and scheduling method and system for edge computing tasks | |
| JP2023526883A (en) | Scheduling methods, computing devices, and storage media for tasks | |
| Xu et al. | Multiuser computation offloading for long-term sequential tasks in mobile edge computing environments | |
| CN109408230A (en) | Docker container dispositions method and system based on energy optimization | |
| Xu et al. | Online learning algorithms for offloading augmented reality requests with uncertain demands in MECs | |
| CN117407160A (en) | Mixed deployment method for online task and offline task in edge computing scene | |
| CN118210609A (en) | Cloud computing scheduling method and system based on DQN model | |
| Durga et al. | Context-aware adaptive resource provisioning for mobile clients in intra-cloud environment | |
| Nguyen et al. | EdgePV: collaborative edge computing framework for task offloading | |
| CN108667920B (en) | Service flow acceleration system and method for fog computing environment | |
| WO2023284347A1 (en) | Task execution method and apparatus | |
| CN114595060A (en) | Task processing method and system | |
| Kortas et al. | Performance impact of the MVMM algorithm for virtual machine migration in data centres |
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 |