CN110413392A - The method of single task migration strategy is formulated under a kind of mobile edge calculations scene - Google Patents

The method of single task migration strategy is formulated under a kind of mobile edge calculations scene Download PDF

Info

Publication number
CN110413392A
CN110413392A CN201910675567.1A CN201910675567A CN110413392A CN 110413392 A CN110413392 A CN 110413392A CN 201910675567 A CN201910675567 A CN 201910675567A CN 110413392 A CN110413392 A CN 110413392A
Authority
CN
China
Prior art keywords
task
energy consumption
ant
subtask
migration
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
CN201910675567.1A
Other languages
Chinese (zh)
Other versions
CN110413392B (en
Inventor
方娟
徐玮豪
陈勇
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.)
Beijing University of Technology
Original Assignee
Beijing University of Technology
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 Beijing University of Technology filed Critical Beijing University of Technology
Priority to CN201910675567.1A priority Critical patent/CN110413392B/en
Publication of CN110413392A publication Critical patent/CN110413392A/en
Priority to PCT/CN2019/124423 priority patent/WO2021012584A1/en
Priority to JP2020533281A priority patent/JP6942397B2/en
Application granted granted Critical
Publication of CN110413392B publication Critical patent/CN110413392B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
    • G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Biomedical Technology (AREA)
  • Artificial Intelligence (AREA)
  • Mathematical Physics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

The invention proposes a kind of method for formulating single task migration strategy under mobile edge calculations scene, under the scene for solving mobile edge, the problem of losing interaction capabilities using bulk migration scheme bring, increase bandwidth of base station pressure.The specific implementation steps are as follows: firstly, the carrying out migration calculating of the task is divided into the different subtasks with relation of interdependence, and guaranteeing that each subtask can individually carry out calculation processing, while cannot migrate subtask node location in determining figure.Secondly, generating a cum rights directed acyclic graph, the calculation amount of each node on behalf data in figure according to the dependence between each subtask, each edge represents the traffic of data between different components.Then each specific execution position in transportable subtask is iterated to calculate out using ant group algorithm, i.e. confirmation migrates to edge calculations server and still locally completes operation.It finally obtains for the purpose of reducing energy consumption of mobile equipment, the suboptimal solution of the single task migration strategy based on ant group algorithm.

Description

The method of single task migration strategy is formulated under a kind of mobile edge calculations scene
Technical field
The invention belongs to mobile edge calculations fields, for the purpose of reducing mobile device task immigration energy consumption, the base of design In the single task migration strategy of ant group algorithm.
Background technique
The mobile terminals such as smart phone or tablet computer are popularized, very far-reaching to mobile and wireless network influence, and And the revolution of global mobile network is thus caused.This kind of mobile device is faced with low memory capacity, high energy consumption, low bandwidth and height The network environment of delay.Mobile cloud computing (MCC) is integrated as cloud computing and mobile computing, and it is considerable to be that mobile device is brought Ability, and by centralized cloud provide storage, calculate and the energy.However, with the appearance of a large amount of mobile devices, the front MCC Face more stern challenge, such as high latency, security breaches and the low network coverage.In next generation mobile networks (such as 5G) Under scene, these problems may become more prominent.According to the report that Cisco's visual web index is issued recently, the year two thousand twenty is arrived, The whole world will have 11,600,000,000 mobile connection equipment to be used.In order to solve growing network demand, mobile edge calculations (MEC) concept is born.The main purpose of MEC is to solve the challenge from MCC system.MEC is by (such as depositing cloud resource Storage and computing capability) edge of wireless access network is deployed to enhance the ability of MCC.This provides quick and strong for terminal user Big computing capability, energy efficiency, memory capacity, mobility and environment sensing is supported.Before this, the referred to as interconnection of cloudlet Network edge technology has been introduced into deployment mobile cloud service.However, still not due to its limited WiFi coverage area cloudlet It can solve existing challenge.So the research field in MEC still has a large amount of problems to need to be solved.
In academia, Rudenko A et al., which proposes task immigration earliest and can be effectively reduced the energy consumption of mobile device, to be prolonged The long working time.The experiment that the research is executed by the program that laptop uploads calculation amount complexity to distal end desktop computer, is tested Above-mentioned guess is demonstrate,proved.Wen Y and Zhang W et al. proposes a kind of task immigration scheme of combined optimization, when mobile application exists When mobile device locally executes, calculating energy consumption is minimized by the clock frequency of most preferably schedule mobile device, when movement is answered When executing in clone's cloud, transmission energy consumption is minimized by configuring the transimission power of wireless channel.Since operation task is moved The computing resource and energy that algorithm itself is mobile device to be consumed are moved, so Huerta-Canepa G et al. puts forward one Execution history and the current application state of foundation mobile application are planted to carry out the scheme of task immigration.The program sets task and moves It moves the adaptive adjustment of decision: when mobile device itself computing resource and battery capacity abundance, using dynamic decision scheme, mentioning High task execution performance；When mobile device own resource deficiency, according to historic state, migration is just carried out before task execution and is determined Plan, overhead caused by reducing because of dynamic decision.In task immigration, wireless network also influences whether the energy that task is completed Consumption, Huang D and Wang P et al. propose a kind of method for changing task immigration strategy according to wireless network environment dynamic, And mobile application is configured to the directed acyclic graph model of multiple subtasks, by distributing the execution position of each subtask, most The execution energy consumption of smallization mobile application.
In addition, some research simultaneously optimizes task execution time and energy consumption.WuHuaming et al. is mentioned Go out a kind of task immigration scheme for executing in shortening and being weighed between time and saving energy consumption, realizes cloud calculating The elasticity of resource, distribution according to need.Li Tianze et al. proposes a kind of consumption of complex energy, time delay and server and executes The prioritization scheme of cost, for the task immigration under MEC environment.The complexity for the task immigration algorithm that multiple targets coexist is often Excessively high, in order to reduce the time complexity of migration algorithm, Wang J et al. proposes a kind of based on Liapunov (Lyapunov) The low complex degree task immigration algorithm of optimum theory, and the energy consumption for executing time and mobile device can be reduced simultaneously.
Relative to the mobile cloud computing environment of tradition, under new mobile edge calculations environment, user and MEC server away from From closer, communication overhead of the task immigration in data transmission is greatly reduced in this.Traditional task immigration strategy is being formulated When, as soon as task is all regarded into entirety, if task is all given MEC server process by migration, not task if migration It is performed locally.Such migration strategy is apparently not optimal for the mobile device for frequently carrying out data communication with server 's.The task immigration mode proposed by the present invention split single task and formulate migration decision, can be improved task execution performance, Task execution expense is reduced, calculating task is subjected to more fine-grained division.The concrete property of task (task topology is utilized at this time The size etc. of transmitted data amount between structure, task computation amount, task) a set of task immigration algorithm of design also just seems especially heavy It wants.
Summary of the invention
It when carrying out task immigration decision, is integrally moved using a task as unit under existing technical solution Move decision.It can need in real scene there are many tasks and the interaction of mobile device frequent progress, and these operations must be It locally executes.Therefore, current technology scheme is that have the part for not being suitable for true scene.Under the scene at mobile edge, The mobile device of user can frequently be interacted with the base station of operator, if losing interaction using the scheme of bulk migration Also it will increase the bandwidth pressure of base station while ability, this is clearly not meet practical application.
The present invention designs one and is based on ant group algorithm task immigration strategy in single user MEC system, is turned by that will apply It is changed to the digraph comprising multiple subtasks, is introduced into the concept of pheromones in ant group algorithm to calculate current subtask migration decision Probability, optimization aim is minimised as with energy consumption of mobile equipment and generates a group scheduling strategy, and algorithm is subjected to successive ignition, no It is disconnected the scheduling strategy of generation to be optimized to approach optimal task scheduling strategy.The present invention fully considers each subtask The case where, to formulate whole migration strategy, it ensure that each not transportable subtask can be performed locally to meet The needs of user's interaction, are more applicable for real scene, improve optimization efficiency.
In order to achieve the above object, the present invention devises following scheme comprising the steps of:
Step 1, it is first temporarily stored in buffer queue when Random Task reaches, when the system call task execution, by task Multiple subtasks that can be independently executed are divided into, set V={ v is represented by1,v2,…vi,…vj,…,vN, N is expressed as son The total number of task.V is divided into two groups of different set V simultaneouslyloc、Wherein VlocRepresentative must be performed locally Can not migration component, VoffIndicate can migrate to the assembly set of MEC server for decision-making.And each There are a unique entrance affairs and outlet affairs in task-set V, wherein predecessor transaction is not present in the entrance affairs, out Subsequent transactions are not present in mouth affairs.The present invention will also define a binary variable Eij∈ { 0,1 } is to indicate between each task Dependence:
ForThere is an eij, for indicating the volume of transmitted data between task i and task j.Finally lead to It crosses subtask collection V and transaction dependency set of relations E and forms a directed acyclic graph G=(V, E).
Step 2, the parameters for determining and initializing mobile edge calculations model, establish energy consumption model.
The purpose of task immigration strategy proposed by the present invention be in order to optimize mobile device end execute energy consumption, need thus to appoint The position (locally execute or the end MEC executes) for being engaged in executing is defined, and uses set A={ A1,A2,……,ANIndicate each The execution position of task, and
The present invention indicates task computation amount using ω (CPU cycles), and f indicates that the CPU of equipment executes rate, and T indicates to appoint The execution time of business.If task is performed locally, the time can will be locally executed are as follows: Tl=ωifl -1.If task is in calculating speed For fcMEC server end execute when, task complete needed for time are as follows: Tc=ωifc -1。
Assuming that power unit when P is CPU execution task is (W), then mobile device is performed locally the energy consumption of task CPU It may be expressed as: El=PlTl.If task is executed in MEC server end, end equipment is moved at this time and does not need carry out task operation, but There is still a need for consumption basal energies to maintain equipment operation, energy consumption Eb=PbTcIt indicates.Wherein Pb(W) mobile device CPU is indicated Power when idle, Tc(s) mobile device standby time is indicated.Due to PbMuch smaller than Pl, so task immigration strategy just can be Mobile device saves energy consumption.
In data transmission consumption, R is usedsAnd RrRespectively indicate channel speed (mobile terminal to the end MEC) sum number of data upload It is (bit/s), P according to download channels rate (end MEC to mobile terminal) unitsAnd PrWhen respectively indicating data transmission and data receiver Power of communications, unit be (W).
When task j is executed in MEC server end, and its previous task i is executed in mobile device end, when the transmission of task Between are as follows:The energy of consumption are as follows:
When task j mobile terminal execute and its previous task MEC server end execute, the transmission time of task are as follows:The energy of consumption are as follows:According to the energy consumption model constructed above, entire mobile device executes completion The total energy consumption individually applied can indicate are as follows:
N indicates the total number of subtask, and equation rightward second portion is indicated from first subtask to height second from the bottom The energy consumption summation of task, wherein [El(1-Ai)+EbAi] indicate mobile device CPU energy consumption.Formula (3)Part indicates total transmission energy consumption of task, | Ai-Aj| for judging task i and its postposition Whether task j, if carrying out operation in MEC server in mobile device or, will not generate in same position progress operation Transmit energy consumption.It is performed locally because the last one subtask determines and follow-up work is not present, by the energy consumption of its task It is added in the front end of energy consumption calculation model.
So far task creation is the model for minimizing mobile device total energy consumption E (A) by we, due to each transportable task There are two kinds of selections, migrate or do not migrate, then the gross migration decision of N number of task will have 2NA solution.If using enumerating Method calculates optimal energy consumption solution of going out on missions, and time complexity is excessively high, is not particularly suited for actual production.So the present invention is calculated using ant colony Method solves this complex task model.
Step 3, the pheromone concentration in each paths, duty cycle number t and ant number m are initialized.
In order to calculate the specific execution energy consumption gone out on missions, the actual calculating position in each subtask just must determine, also It is set A={ A1,A2,……,ANValue.Task computation amount and data traffic needed for each subtask are not quite similar, In order to reduce task energy consumption as far as possible, we are more prone to for the subtask for calculating dependent form to be transferred to the progress of edge calculations node Operation, and the subtask that data traffic is high but calculation amount is low is given mobile device and is locally handled.Ant group algorithm is in foundation The concentration calculation of pheromones is gone out on missions migration probability on different paths, to obtain task immigration strategy.The present invention uses τc(0) ={ τ1c(0)、……、τNc(0)}、τl(0)={ τ1l(0)、……、τNl(0) } respectively indicate ant group algorithm start execute when, respectively Pheromone concentration of the subtask on migration path and not migration path, and forThere is τic(0)=τil(0)= δ, (δ ∈ (0,1)).ForThere is τil(0)=+ ∞.During initialization task cycle-index and every wheel recycle simultaneously The number m of ant.
Step 4, path selected by every ant is obtained using ant group algorithm, the task total energy consumption model designed according to step 2 From path selected by all m ants, least energy consumption E is selectedmin(A) corresponding path is as preferred under this duty cycle Task immigration strategy continues to execute step 5 after m ants all under this duty cycle complete task
Wherein, path selected by ant calculates probability by migration and determines, and under t subtask circulation, every ant appoints son The probability P that business i migration calculatesic(t) calculation formula is as follows:
The meaning of each symbol is as follows in above formula:
T indicates duty cycle number, also illustrates that the moment；
·τic(t) indicate that t moment migrates task i to the concentration of pheromones on this paths of MEC server, til(t) table Show the concentration of pheromones on this paths of t moment task i local computing；
α indicates pheromones heuristic greedy method (α ∈ [0,5]), it reflects effect of the pheromones to ant Path selection；
·It is a heuristic function, indicates the expected degree that task i needs to migrate, value is in the present inventionIt can be seen that eijIt is smallerIt is bigger, that is, task i migration desired value it is higher；
β indicate the heuristic function factor (β ∈ [0,5]), reflect heuristic function instruct ant colony search in it is relatively heavy Want degree；
After the execution position for obtaining each subtask by formula 4, ant number k is reset, every ant is appointed according to epicycle Migration strategy execution task of being engaged in calculates the energy consumption in path selected by every ant, and more according to the energy consumption model that step 2 designs New lowest energy consumption Emin(A).Step 5 is continued to execute after all m ants of epicycle complete task.
Step 5, if not up to preset duty cycle number, updates pheromone concentration, and return step 4 continues to seek Look for more preferably task immigration strategy；If reaching preset duty cycle number, step 6 is continued to execute.
The pheromone concentration more new formula is as follows:
τic(t+1)=(1- ρ) * τic(t)+Δτic(t,t+1) (5)
Wherein, ρ is pheromones volatilization factor (ρ ∈ [0.1,0.99]), and 1- ρ indicates remaining information prime factor, Δ τic(t, T+1 it) is expressed as increment of the pheromones after a wheel task iteration, is calculated by formula 6:
M is the total number of ant in one cycle,Indicate kth ant at task i task immigration this The pheromones left on paths, the pheromones that every ant is left on migration path at the task i then indicate by formula 7, Middle Q is a normal number (Q ∈ [1,10]), for controlling the quantity for the pheromones that every ant leaves.
Step 6, the preferred task immigration strategy that last time duty cycle obtains is OPTIMAL TASK migration strategy, according to OPTIMAL TASK migration strategy carries out task distribution, executes edge calculations.
Compared with prior art, the invention has the characteristics that:
Present invention design is in view of in real MEC scene, there are the applications that many needs and mobile subscriber frequently interact, These application program bulk migration operations can undoubtedly be increased significantly into communication overhead, lead to higher energy consumption of mobile equipment.This hair It is bright first to convert the digraph comprising multiple subtasks for application to be processed, it is then repeatedly traversed using ant group algorithm to be processed Task image finally obtains the task immigration strategy suboptimal solution using energy consumption as optimization aim.Guaranteeing task compared to other algorithms On the basis of execution efficiency, the time complexity of task execution is reduced, while fine granularity splits the mode of task, to greatest extent Energy consumption of mobile equipment is reduced, the service quality of whole MEC system is improved.
Detailed description of the invention
To make the purpose of the present invention, scheme is more easy-to-understand, and below in conjunction with attached drawing, the present invention is further described.
Fig. 1 is that finegrained tasks of the present invention divide figure；
Fig. 2 is task execution flow chart；
Specific embodiment
Step 1, as Figure 1 shows is the finegrained tasks division figure applied, and the present invention will be more using being divided into A subtask independently executed, and indicated with a digraph G=(V, E).What the node ν ∈ V expression in Fig. 1 was split Subtask, the side e in Fig. 1ijTransmission data between ∈ E expression task, such as: eijAfter the completion of expression task i is executed, it can pass Defeated eijData give task j, and data that task j is only transmitted after the task i of receiving has been executed could start to hold Row.Subtask in figure is segmented into 2 liang of classes: one kind be must locally execute task (such as the audio-video collection of user with The interaction etc. of mobile terminal), the solid task 1,4,6 being expressed as in Fig. 1 is expressed asAnother kind of is transportable Task, hollow task 2,3,5 as shown in figure 1, is expressed as
Invention defines a binary variable Eij∈ { 0,1 } is to indicate the dependence between each task.Formula 1 E is worked as in expressionijWhen=1, task j can just start to execute after the completion of task i is executed, E in the case of otherij=0.Such as in Fig. 1 E12=1, E45=1, E63=0, if also, there are two a tasks or when more than two previous tasks, it is necessary to it waits until all preposition Task just can be performed after completing.
Step 2, the energy consumption model of mobile edge calculations, initialization task parameter are established.Energy consumption model is divided by the present invention Two parts, local computing energy consumption and migration calculate energy consumption.Assuming that the calculation amount of each subtask is respectively ωi(CPU Cycles), it is f that CPU, which executes rate,l, executing power when calculating is Pl, then task, which locally executes energy consumption, may be expressed as: El=Plωifl -1.It is calculated if task needs to migrate, E can be used in standby energy consumption of the mobile device during task immigrationb=Pbωifc -1Table Show, simultaneously as task immigration will generate communication energy consumption, so the present invention uses respectivelyWithIndicate the energy consumption that data are uploaded and downloaded.The purpose of task immigration strategy proposed by the present invention is to optimize and move Dynamic equipment end executes energy consumption, needs to be defined the position (locally execute or the end MEC executes) of task execution thus, use Ai∈ { 0,1 } indicates the execution position of some task.Obviously, all must being performed locally for tasks can only by mobile device into Row operation, so forThe total energy consumption of task can then be gone out by 3 table of formula.
Step 3, the specific energy consumption for wanting to obtain task execution on the basis of step 2 need to only determine set A={ A1, A2,……,ANValue.First according to the pheromone concentration τ of initializationc(0)、τl(0) and formula 4, first run task is calculated Migration probability Pic(0).The k ant of the subsequent first round passes through Pic(0) respective first time task immigration strategy A can be acquired respectively ={ A1,A2,……,ANValue, (method for solving is as follows: assuming that P1c(0)=λ generates one by completely random method at this time Digital μ ∈ (0,1), the A if 0 < λ≤μ1=1, the A if 1 > λ > μ1=0.) finally using you can get it the task total energy consumption of formula 3, And minimum value is chosen in this k energy consumption and is recorded as Emin(A)。
Step 4, according to least energy consumption Emin(A) and formula 5-7, the pheromone concentration on different paths is updated.Using new Pheromone concentration and formula 4 calculate the task allocation probability P of next roundic(t), pass through Pic(t) new appoint can be obtained again Be engaged in migration strategy A, then updates optimal energy consumption Emin(A)。
Step 5, if not up to preset duty cycle number t, return step 4 continually look for more preferably task immigration Strategy；If reaching preset duty cycle number t, step 6 is continued to execute.
Step 6, the preferred task immigration strategy that last time duty cycle obtains is OPTIMAL TASK migration strategy, according to OPTIMAL TASK migration strategy carries out task distribution, executes edge calculations.Specific execution flow chart is as shown in Figure 2.

Claims (3)

1. formulating the method for single task migration strategy under a kind of mobile edge calculations scene, it is characterised in that comprise the steps of:
Step 1, it is first temporarily stored in buffer queue when Random Task reaches, when the system call task execution, task is divided For multiple subtasks that can be independently executed, it is represented by set V={ v1, v2... vi... vj..., vN, N is expressed as son The total number of task, while V is divided into two groups of different setWherein VlocRepresenting must hold locally Capable not transportable subtask, VoffIndicate can migrate to the subtask of MEC server set for decision-making, and There are a unique entrance affairs and outlet affairs by each task-set V, wherein there is no leading for the entrance affairs Subsequent transactions are not present in affairs, outlet affairs, and the present invention will also define a binary variable Eij∈ { 0,1 } is to indicate each Dependence between business:
ForThere is an eij, for indicating the volume of transmitted data between task i and task j, finally by appoint Business collection V and transaction dependency set of relations E forms a directed acyclic graph G=(V, E)；
Step 2, task total energy consumption model is established in conjunction with task computation and multiplexed transport energy consumption model, and initializes total energy consumption model Each parameter；
Step 3, the pheromone concentration in each paths is initialized, ant number m in duty cycle number t and every wheel circulation；
Step 4, path selected by every ant is obtained using ant group algorithm, the task total energy consumption model designed according to step 2 is from institute Have in path selected by m ant, selects least energy consumption Emin(A) corresponding path is as the preferred task under this duty cycle Migration strategy continues to execute step 5 after m ants all under this duty cycle complete task,
Wherein, path selected by ant calculates probability by migration and determines, and under t subtask circulation, every ant moves subtask i Move the probability P calculatedic(t) calculation formula is as follows:
The meaning of each symbol is as follows in above formula:
T indicates duty cycle number；
·τic(t) indicate that t moment migrates task i to the concentration of pheromones on this paths of MEC server, τil(t) when indicating t The concentration of pheromones on this paths of quarter task i local computing；
α indicates pheromones heuristic greedy method, it reflects effect of the pheromones to ant Path selection；
·It is a heuristic function, indicates the expected degree that task i needs to migrate, value is in the present inventionIt can be seen that eijIt is smallerIt is bigger, that is, task i migration desired value it is higher；
β indicates the heuristic function factor, reflects heuristic function and is instructing the relative importance in ant colony search；
Step 5, if not up to preset duty cycle number, updates pheromone concentration, and return step 4 continually looks for more Excellent task immigration strategy；If reaching preset duty cycle number, step 6 is continued to execute,
The pheromone concentration more new formula is as follows:
τic(t+1)=(1- ρ) * τic(t)+Δτic(t, t+1) (5)
Wherein, ρ is pheromones volatilization factor, and 1- ρ indicates remaining information prime factor, Δ τic(t, t+1) is expressed as pheromones warp Increment after crossing a wheel task iteration, is calculated by formula 6:
M is the total number of ant in one cycle,Indicate kth ant this road of task immigration at task i The pheromones left on diameter, the pheromones that every ant is left on migration path at the task i then indicate that wherein Q is by formula 7 One normal number, for controlling the quantity for the pheromones that every ant leaves,
E hereinmin(A) it represents t subtask and recycles corresponding least energy consumption；
Step 6, the preferred task immigration strategy that last time duty cycle obtains is OPTIMAL TASK migration strategy, according to optimal Task immigration strategy carries out task distribution, executes edge calculations.
2. the method for formulating single task migration strategy under a kind of mobile edge calculations scene according to claim 1, special Sign is:
Pheromone concentration described in step 3 includes pheromones of each subtask on migration path when ant group algorithm starts to execute Concentration τc(0)={ τ1c(0)、......、τNcAnd pheromone concentration τ of each subtask on not migration path (0) }1(0)= {τ11(0)、......、τNl(0) }, and forThere is τic(0)=τil(0)=δ, δ ∈ (0,1), forThere is τil(0)=+ ∞.
3. the method for formulating single task migration strategy under a kind of mobile edge calculations scene according to claim 1, special Sign is:
Task computation energy consumption model described in step 2 is as follows: if task is performed locally, mobile device, which is performed locally, appoints The energy consumption of business are as follows: El=PlTl, wherein PlPower when task, task execution time T are executed for local cpul=ωifl -1, ωi The calculation amount of expression task i, flIndicate that the CPU of local device executes rate；
If task is executed in MEC server end, the basic energy consumption E of mobile deviceb=PbTc, wherein PbIndicate mobile device CPU Power when idle, task execution time Tc=ωifc -1, fcIndicate that the CPU of MEC server executes rate；
Multiplexed transport energy consumption model described in step 2 is as follows:
When task j MEC server end execute, and its previous task i mobile device end execute, then the energy consumed are as follows:The wherein transmission time of task
When task j mobile terminal execute, and its previous task MEC server end execute, then the energy consumed are as follows:Wherein, the transmission time of task are as follows:
Wherein, RsAnd RrRespectively indicate the channel speed and data download channels rate of data upload, PsAnd PrRespectively indicate data Send the power with mobile device when data receiver.
Task total energy consumption model described in step 2 is as follows:
Wherein,
Set A={ A1, A2..., ANIndicate the execution position of each task, and
CN201910675567.1A 2019-07-25 2019-07-25 Method for formulating single task migration strategy in mobile edge computing scene Active CN110413392B (en)

Priority Applications (3)

Application Number Priority Date Filing Date Title
CN201910675567.1A CN110413392B (en) 2019-07-25 2019-07-25 Method for formulating single task migration strategy in mobile edge computing scene
PCT/CN2019/124423 WO2021012584A1 (en) 2019-07-25 2019-12-11 Method for formulating single-task migration strategy in mobile edge computing scenario
JP2020533281A JP6942397B2 (en) 2019-07-25 2019-12-11 How to develop a singletasking offload strategy in a mobile edge computing scenario

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910675567.1A CN110413392B (en) 2019-07-25 2019-07-25 Method for formulating single task migration strategy in mobile edge computing scene

Publications (2)

Publication Number Publication Date
CN110413392A true CN110413392A (en) 2019-11-05
CN110413392B CN110413392B (en) 2022-11-29

Family

ID=68363105

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910675567.1A Active CN110413392B (en) 2019-07-25 2019-07-25 Method for formulating single task migration strategy in mobile edge computing scene

Country Status (3)

Country Link
JP (1) JP6942397B2 (en)
CN (1) CN110413392B (en)
WO (1) WO2021012584A1 (en)

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111010295A (en) * 2019-11-28 2020-04-14 国网甘肃省电力公司电力科学研究院 SDN-MEC-based power distribution and utilization communication network task migration method
CN111262906A (en) * 2020-01-08 2020-06-09 中山大学 Method for unloading mobile user terminal task under distributed edge computing service system
CN111274031A (en) * 2020-01-16 2020-06-12 国家电网有限公司信息通信分公司 Method and device for dynamic migration authentication of edge service with cooperation of end and cloud
CN111309460A (en) * 2020-01-15 2020-06-19 南京大学 Task processing method of intelligent mobile equipment in mobile edge computing scene
CN111459662A (en) * 2020-03-18 2020-07-28 湖南大学 Migration management method, device and storage medium in mobile edge calculation
CN111506416A (en) * 2019-12-31 2020-08-07 远景智能国际私人投资有限公司 Computing method, scheduling method, related device and medium of edge gateway
CN111917854A (en) * 2020-07-25 2020-11-10 西安邮电大学 Cooperation type migration decision method and system facing MCC
CN112114960A (en) * 2020-08-06 2020-12-22 河南大学 Scheduling strategy for remote sensing image parallel cluster processing suitable for internet scene
WO2021012584A1 (en) * 2019-07-25 2021-01-28 北京工业大学 Method for formulating single-task migration strategy in mobile edge computing scenario
CN112291335A (en) * 2020-10-27 2021-01-29 上海交通大学 Optimized task scheduling method in mobile edge calculation
CN112988345A (en) * 2021-02-09 2021-06-18 江南大学 Dependency task unloading method and device based on mobile edge calculation
CN113342409A (en) * 2021-04-25 2021-09-03 山东师范大学 Delay sensitive task unloading decision method and system for multi-access edge computing system
CN113687950A (en) * 2021-08-31 2021-11-23 平安医疗健康管理股份有限公司 Priority-based task allocation method, device, equipment and storage medium
CN114448810A (en) * 2021-12-24 2022-05-06 天翼云科技有限公司 Bandwidth traction method and device and electronic equipment
CN114500560A (en) * 2022-01-06 2022-05-13 浙江鼎峰科技股份有限公司 Edge node service deployment and load balancing method for minimizing network delay
CN114979134A (en) * 2022-04-21 2022-08-30 云南大学 Path selection method for service migration in edge computing environment

Families Citing this family (38)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112764936B (en) * 2021-01-29 2022-06-14 北京邮电大学 Edge calculation server information processing method and device based on deep reinforcement learning
CN113452956B (en) * 2021-02-26 2023-03-28 深圳供电局有限公司 Intelligent distribution method and system for power transmission line inspection tasks
CN113128681B (en) * 2021-04-08 2023-05-12 天津大学 Multi-edge equipment-assisted general CNN reasoning acceleration system
CN113076177B (en) * 2021-04-12 2022-03-22 河北大学 Dynamic migration method of virtual machine in edge computing environment
CN113687924B (en) * 2021-05-11 2023-10-20 武汉理工大学 Intelligent dynamic task computing and unloading method based on edge computing system
CN113326112B (en) * 2021-05-11 2023-11-07 武汉理工大学 Multi-unmanned aerial vehicle task unloading and migration method based on block coordinate descent method
CN113342514B (en) * 2021-05-11 2023-11-07 武汉理工大学 Edge calculation model based on near-earth orbit and service placement method thereof
CN113139639B (en) * 2021-05-17 2023-05-30 华侨大学 MOMBI-oriented smart city application multi-target computing migration method and device
CN113422803B (en) * 2021-05-19 2022-11-25 武汉工程大学 Seamless migration method for intelligent unmanned aerial vehicle inspection task based on end edge cloud cooperation
CN113472854A (en) * 2021-06-07 2021-10-01 湖南大学 Method, device and medium for intelligent network collaborative task unloading in complex edge environment
CN113269461A (en) * 2021-06-09 2021-08-17 北京理工大学 Game-based edge computing resource management method
CN113365290B (en) * 2021-06-10 2023-04-14 中山大学 Game theory calculation unloading method based on greedy strategy in world fusion network
CN113543074B (en) * 2021-06-15 2023-04-18 南京航空航天大学 Joint computing migration and resource allocation method based on vehicle-road cloud cooperation
CN113325875B (en) * 2021-06-21 2022-06-07 西安电子科技大学 Unmanned aerial vehicle path planning method for minimizing number of unmanned aerial vehicles
CN113347267B (en) * 2021-06-22 2022-03-18 中南大学 MEC server deployment method in mobile edge cloud computing network
CN113613210B (en) * 2021-07-02 2023-05-26 重庆邮电大学 Internet of vehicles task unloading method based on multi-variable joint optimization
CN113507712B (en) * 2021-07-09 2023-07-28 中国传媒大学 Resource allocation and calculation task unloading method based on alternate direction multiplier
CN113535393B (en) * 2021-07-12 2023-04-04 湖南工学院 Computing resource allocation method for unloading DAG task in heterogeneous edge computing
CN113612843B (en) * 2021-08-02 2022-08-30 吉林大学 MEC task unloading and resource allocation method based on deep reinforcement learning
CN113703391B (en) * 2021-08-11 2023-01-31 广东省科学院智能制造研究所 Machining path planning method based on random pheromone optimization elite ant colony algorithm
CN114090108B (en) * 2021-09-16 2024-02-06 北京邮电大学 Method and device for executing computing task, electronic equipment and storage medium
CN113867348B (en) * 2021-09-26 2023-06-20 河南科技大学 Robot path planning method based on single-population self-adaptive heterogeneous ant colony algorithm
CN113727308B (en) * 2021-10-20 2023-06-30 湖北大学 Edge calculation unloading optimization method based on vehicle position prediction
CN113965961B (en) * 2021-10-27 2024-04-09 中国科学院计算技术研究所 Edge computing task unloading method and system in Internet of vehicles environment
CN114024977B (en) * 2021-10-29 2024-02-06 深圳市高德信通信股份有限公司 Data scheduling method, device and system based on edge calculation
CN114844886B (en) * 2022-03-22 2023-11-24 山东理工大学 Mobile service selection method based on seagull optimization algorithm in cloud edge combined environment
CN115002212B (en) * 2022-04-12 2024-02-27 广州大学 Combined caching and unloading method and system based on cross entropy optimization algorithm
CN114867039B (en) * 2022-04-15 2024-06-21 河海大学 Edge computing unloading method for middle sea area scene
WO2023203731A1 (en) * 2022-04-21 2023-10-26 日本電信電話株式会社 Arithmetic processing offloading system, arithmetic processing offloading method, and program
CN114928609B (en) * 2022-04-27 2023-02-03 南京工业大学 Optimal task unloading method for heterogeneous cloud-edge environment of scene of Internet of things
CN115065727B (en) * 2022-05-19 2023-08-22 南京邮电大学 Task unloading method based on edge computing scene
CN115277692B (en) * 2022-06-21 2023-07-25 华北电力科学研究院有限责任公司 Automatic operation and maintenance method, device and system for edge network computing terminal equipment
CN115361688B (en) * 2022-07-13 2023-11-10 西安电子科技大学 Industrial wireless edge gateway optimization layout scheme based on machine learning
CN116679758B (en) * 2023-08-04 2023-10-17 江西现代职业技术学院 Unmanned aerial vehicle scheduling method, unmanned aerial vehicle scheduling system, computer and readable storage medium
CN117251296B (en) * 2023-11-15 2024-03-12 成都信息工程大学 Mobile edge computing task unloading method with caching mechanism
CN117834643B (en) * 2024-03-05 2024-05-03 南京邮电大学 Deep neural network collaborative reasoning method for industrial Internet of things
CN117873689B (en) * 2024-03-11 2024-05-31 浪潮计算机科技有限公司 Task allocation method, device, equipment and computer readable storage medium
CN117939535B (en) * 2024-03-20 2024-05-31 合肥工业大学智能制造技术研究院 Dependency task unloading method, terminal and storage medium in V2V scene

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2328308A1 (en) * 2009-11-27 2011-06-01 Alcatel Lucent Method for building a path according to adaptation functions using an ant colony
CN109167671A (en) * 2018-07-11 2019-01-08 国网信通亿力科技有限责任公司 A kind of adapted communication system equally loaded dispatching algorithm towards quantum key distribution business
CN109375999A (en) * 2018-10-23 2019-02-22 北京工业大学 A kind of MEC Random Task moving method based on Bayesian network
CN109684075A (en) * 2018-11-28 2019-04-26 深圳供电局有限公司 A method of calculating task unloading is carried out based on edge calculations and cloud computing collaboration

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9536192B2 (en) * 2014-06-23 2017-01-03 International Business Machines Corporation Solving vehicle routing problems using evolutionary computing techniques
US20170026305A1 (en) * 2015-07-23 2017-01-26 Schneider Electric It Corporation System to place virtual machines onto servers based upon backup runtime constraints
CN109062682B (en) * 2018-06-29 2021-07-06 广东工业大学 Resource scheduling method and system for cloud computing platform
CN109101316B (en) * 2018-07-11 2022-10-04 南京邮电大学 Energy efficiency related cloud computing platform resource scheduling method
CN108920280B (en) * 2018-07-13 2021-06-11 哈尔滨工业大学 Mobile edge computing task unloading method under single-user scene
CN110413392B (en) * 2019-07-25 2022-11-29 北京工业大学 Method for formulating single task migration strategy in mobile edge computing scene

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2328308A1 (en) * 2009-11-27 2011-06-01 Alcatel Lucent Method for building a path according to adaptation functions using an ant colony
CN109167671A (en) * 2018-07-11 2019-01-08 国网信通亿力科技有限责任公司 A kind of adapted communication system equally loaded dispatching algorithm towards quantum key distribution business
CN109375999A (en) * 2018-10-23 2019-02-22 北京工业大学 A kind of MEC Random Task moving method based on Bayesian network
CN109684075A (en) * 2018-11-28 2019-04-26 深圳供电局有限公司 A method of calculating task unloading is carried out based on edge calculations and cloud computing collaboration

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
薛宁等: "基于贝叶斯网络的MEC随机任务迁移算法", 《信息通信技术》 *

Cited By (27)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021012584A1 (en) * 2019-07-25 2021-01-28 北京工业大学 Method for formulating single-task migration strategy in mobile edge computing scenario
CN111010295B (en) * 2019-11-28 2022-09-16 国网甘肃省电力公司电力科学研究院 SDN-MEC-based power distribution and utilization communication network task migration method
CN111010295A (en) * 2019-11-28 2020-04-14 国网甘肃省电力公司电力科学研究院 SDN-MEC-based power distribution and utilization communication network task migration method
CN111506416A (en) * 2019-12-31 2020-08-07 远景智能国际私人投资有限公司 Computing method, scheduling method, related device and medium of edge gateway
CN111506416B (en) * 2019-12-31 2023-09-12 远景智能国际私人投资有限公司 Computing method, scheduling method, related device and medium of edge gateway
CN111262906A (en) * 2020-01-08 2020-06-09 中山大学 Method for unloading mobile user terminal task under distributed edge computing service system
CN111262906B (en) * 2020-01-08 2021-05-25 中山大学 Method for unloading mobile user terminal task under distributed edge computing service system
CN111309460A (en) * 2020-01-15 2020-06-19 南京大学 Task processing method of intelligent mobile equipment in mobile edge computing scene
CN111309460B (en) * 2020-01-15 2023-04-07 南京大学 Task processing method of intelligent mobile equipment in mobile edge computing scene
CN111274031B (en) * 2020-01-16 2023-07-25 国家电网有限公司信息通信分公司 Method and device for dynamic migration authentication of end-cloud cooperative edge service
CN111274031A (en) * 2020-01-16 2020-06-12 国家电网有限公司信息通信分公司 Method and device for dynamic migration authentication of edge service with cooperation of end and cloud
CN111459662B (en) * 2020-03-18 2023-07-04 湖南大学 Migration management method, migration management device and storage medium in mobile edge computing
CN111459662A (en) * 2020-03-18 2020-07-28 湖南大学 Migration management method, device and storage medium in mobile edge calculation
CN111917854A (en) * 2020-07-25 2020-11-10 西安邮电大学 Cooperation type migration decision method and system facing MCC
CN111917854B (en) * 2020-07-25 2023-04-07 西安邮电大学 Cooperation type migration decision method and system facing MCC
CN112114960B (en) * 2020-08-06 2022-11-01 河南大学 Scheduling strategy for remote sensing image parallel cluster processing adapting to internet scene
CN112114960A (en) * 2020-08-06 2020-12-22 河南大学 Scheduling strategy for remote sensing image parallel cluster processing suitable for internet scene
CN112291335A (en) * 2020-10-27 2021-01-29 上海交通大学 Optimized task scheduling method in mobile edge calculation
CN112988345A (en) * 2021-02-09 2021-06-18 江南大学 Dependency task unloading method and device based on mobile edge calculation
CN112988345B (en) * 2021-02-09 2024-04-02 江南大学 Dependency task unloading method and device based on mobile edge calculation
CN113342409A (en) * 2021-04-25 2021-09-03 山东师范大学 Delay sensitive task unloading decision method and system for multi-access edge computing system
CN113687950A (en) * 2021-08-31 2021-11-23 平安医疗健康管理股份有限公司 Priority-based task allocation method, device, equipment and storage medium
CN114448810A (en) * 2021-12-24 2022-05-06 天翼云科技有限公司 Bandwidth traction method and device and electronic equipment
CN114500560A (en) * 2022-01-06 2022-05-13 浙江鼎峰科技股份有限公司 Edge node service deployment and load balancing method for minimizing network delay
CN114500560B (en) * 2022-01-06 2024-04-26 浙江鼎峰科技股份有限公司 Edge node service deployment and load balancing method for minimizing network delay
CN114979134A (en) * 2022-04-21 2022-08-30 云南大学 Path selection method for service migration in edge computing environment
CN114979134B (en) * 2022-04-21 2023-01-17 云南大学 Path selection method for service migration in edge computing environment

Also Published As

Publication number Publication date
CN110413392B (en) 2022-11-29
JP6942397B2 (en) 2021-09-29
WO2021012584A1 (en) 2021-01-28
JP2021525402A (en) 2021-09-24

Similar Documents

Publication Publication Date Title
CN110413392A (en) The method of single task migration strategy is formulated under a kind of mobile edge calculations scene
CN109684075B (en) Method for unloading computing tasks based on edge computing and cloud computing cooperation
Li et al. Deep reinforcement learning based computation offloading and resource allocation for MEC
Fadlullah et al. HCP: Heterogeneous computing platform for federated learning based collaborative content caching towards 6G networks
Shu et al. Multi-user offloading for edge computing networks: A dependency-aware and latency-optimal approach
Chen et al. ThriftyEdge: Resource-efficient edge computing for intelligent IoT applications
Cheng et al. Space/aerial-assisted computing offloading for IoT applications: A learning-based approach
Sun et al. Learning-based task offloading for vehicular cloud computing systems
CN111405568B (en) Computing unloading and resource allocation method and device based on Q learning
Dab et al. Q-learning algorithm for joint computation offloading and resource allocation in edge cloud
Lee et al. An online secretary framework for fog network formation with minimal latency
CN111475274B (en) Cloud collaborative multi-task scheduling method and device
CN109951873B (en) Task unloading mechanism under asymmetric and uncertain information in fog computing of Internet of things
WO2023040022A1 (en) Computing and network collaboration-based distributed computation offloading method in random network
CN109286664A (en) A kind of computation migration terminal energy consumption optimization method based on Lagrange
Shu et al. Dependency-aware and latency-optimal computation offloading for multi-user edge computing networks
Nath et al. Multi-user multi-channel computation offloading and resource allocation for mobile edge computing
CN110069341B (en) Method for scheduling tasks with dependency relationship configured according to needs by combining functions in edge computing
Misra et al. Multiarmed-bandit-based decentralized computation offloading in fog-enabled IoT
CN111552564A (en) Task unloading and resource optimization method based on edge cache
CN108600299B (en) Distributed multi-user computing task unloading method and system
CN114205353B (en) Calculation unloading method based on hybrid action space reinforcement learning algorithm
CN111988787B (en) Task network access and service placement position selection method and system
CN110287034A (en) The dynamic task allocation method of energy-delay balance in a kind of chargeable mobile edge calculations
CN113992678A (en) Calculation migration method for offshore MEC load balancing and resource allocation joint optimization

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