CN109067842A - Calculating task discharging method towards car networking - Google Patents

Calculating task discharging method towards car networking Download PDF

Info

Publication number
CN109067842A
CN109067842A CN201810734390.3A CN201810734390A CN109067842A CN 109067842 A CN109067842 A CN 109067842A CN 201810734390 A CN201810734390 A CN 201810734390A CN 109067842 A CN109067842 A CN 109067842A
Authority
CN
China
Prior art keywords
calculating task
task
queue
calculation processing
calculating
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
CN201810734390.3A
Other languages
Chinese (zh)
Other versions
CN109067842B (en
Inventor
张科
杨迪
冷甦鹏
吴凡
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.)
University of Electronic Science and Technology of China
Original Assignee
University of Electronic Science and Technology of China
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 University of Electronic Science and Technology of China filed Critical University of Electronic Science and Technology of China
Priority to CN201810734390.3A priority Critical patent/CN109067842B/en
Publication of CN109067842A publication Critical patent/CN109067842A/en
Application granted granted Critical
Publication of CN109067842B publication Critical patent/CN109067842B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • 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
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/30—Services specially adapted for particular environments, situations or purposes
    • H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
    • H04W4/44—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P] for communication between vehicles and infrastructures, e.g. vehicle-to-cloud [V2C] or vehicle-to-home [V2H]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/30—Services specially adapted for particular environments, situations or purposes
    • H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
    • H04W4/46—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P] for vehicle-to-vehicle communication [V2V]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/50—Queue scheduling
    • H04L47/56—Queue scheduling implementing delay-aware scheduling
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/50—Queue scheduling
    • H04L47/62—Queue scheduling characterised by scheduling criteria
    • H04L47/6245—Modifications to standard FIFO or LIFO

Abstract

The present invention discloses a kind of calculating task discharging method towards car networking, in order to improve the completion time delay for guaranteeing calculating task about the income for calculating resource provider in calculating task uninstall process and to a certain extent in the prior art, the present invention uses the computing resource of base station side MEC serve and the computing resource of road running vehicle to provide unloading scheme for the calculating task of the UE in roadside；The communication delay of task, the processing speed of calculation delay and calculation processing queue and queue length are included in the reference factor of decision, using the variation of the mathematics model analysis state of Markovian decision, an optimisation strategy space is obtained using deep learning algorithm, when i.e. corresponding state uses corresponding behavior, so that the Income Maximum of resource provider, and ensure that the delay requirement of user's calculating task to a certain extent.

Description

Calculating task discharging method towards car networking
Technical field
The invention belongs to field of wireless communication, in particular to a kind of deep learning algorithm unloads edge calculating task Technology.
Background technique
With the quickening of intelligent movable equipment growth rate and continuing to bring out for mobile application, traditional centralized network frame Structure can not adapt to low time delay, and high-performance etc. calculates demand.These ever-increasing application demands excite mobile edge calculations (MEC) development.Mobile edge calculations are as a kind of promising technology, since computing resource is closer to mobile device end, because This can provide edge calculations service at cellular network edge, by reducing network operation and service offering time delay, promote user Quality of Service Experience, therefore have received widespread attention, and MEC has been used as key technology to be included into 5G standard.
MEC is intended to eliminate delay by providing the computing capability closer to equipment.It enhances virtual/augmented reality etc. Various applications, for needing the technologies such as unmanned of response time of Millisecond, this point is most important.Edge Calculating will play key effect in 5G, and 5G is to be expected to welcome robot, pilotless automobile, immersion media on the spot in person Experience, the next-generation mobile connection technology of AI and the internet of things era.
Carrier of the vehicle as computing resource can provide connecing for computing resource in edge side for the mobile subscriber in roadside Enter, data do not have to pass to remote cloud again, just can solve in edge side, are more suitable at real-time data analysis and intelligence Reason, it is also more efficiently and safe.In the communication process of car networking, MEC server is widely deployed network edge base station side To improve the connectivity of network, base station itself also can be used as the supplier of computing resource, and possess higher downlink communication speed Rate.Consider that the mobile subscriber of roadside walking is the demander of resource, the move vehicle on edge MEC server and road is meter The supplier of resource is calculated, vehicle provides the access processing of calculating task as the mobile vehicle of computing resource for user.Increasingly More applications needs sufficient and timely calculation processing, and vehicle plays a significant role in reducing calculation delay.It is deployed in base station On mobile Edge Server may be that the computation requests of UE provide calculation processing service, and downlink with higher Traffic rate.Calculating task does not need to be transferred to distal end Cloud Server, it can be unloaded to local mobile Edge Server (such as Base station) or vehicle, edge calculations are suitable for short distance real-time data analysis and Intelligent treatment, therefore calculation processing more safety is high Effect.
Many previous research work are dedicated to research calculating task and are unloaded to MEC server by wireless cellular network (such as base station).However, less research work, which is dedicated to research calculating task, to be unloaded in move vehicle.Vehicle and base The computing resource supplier important as two that stand is usually independent analysis in existing research, and individually vehicle unloading Or the computing resource processing capacity of base station unloading is usually limited.Recognize the limitation of existing research, in the present invention, jointly Consider that move vehicle and base station work together to unload calculating task, by dispatching algorithm to improve computing resource to the maximum extent The income of provider, and ensure that the delay requirement of user's calculating task to a certain extent.
Summary of the invention
In order to improve in the prior art about the income of calculating resource provider in calculating task uninstall process and one Determine the completion time delay for guaranteeing calculating task in degree, the present invention proposes a kind of calculating task discharging method towards car networking.
The technical solution of the present invention is as follows: the calculating task discharging method towards car networking, comprising:
S1, when UE generates calculating task, initiate to calculate unloading solicited message to base station, include pair inside the solicited message The calculating task illustrates information；
After S2, base station receive the calculating unloading request that UE sends, according to the state of multiple calculation processing queues of base station, The state and UE of the available calculation processing queue of road vehicle calculate the size combination Markovian decision model of unloading task Determine unloading queue of some calculation processing queue as UE calculating task；
S3, UE establish the communication linkage with calculate node according to the scheduling result of feedback, and unload calculating task to accordingly In the calculation processing queue of calculate node；
S4, calculation processing queue and feed back to calculation processing result according to the calculating task of the sequential processes UE of FIFO UE。
Further, illustrate that information includes: communication task amount size, the calculating task amount of calculating task described in step S1 Size and delay requirement.
Further, the unloading queue determination process of UE calculating task described in step S2 specifically include it is following step by step:
A1, moment t system mode beScheduling decision is
Wherein,Expression is put into corresponding calculation processing queue i in the calculating task that time t is reached；For base The m calculation processing queue stood t moment calculating task amount size,For the calculating task amount of vehicle calculation processing queue Size,It is vehicle calculation processing queue in the available amount of computational resources of t moment, dtFor the t moment UE calculating task generated Calculating task amount size, ftFor the communication task amount size of the t moment UE calculating task generated；
The treating capacity of A2, each calculation processing queue actual calculating task in time interval τ are as follows:
Wherein, ViFor the calculation processing rate of calculation processing queue i,Indicate scheduling probability；
The system shape of A3, moment t+1 are as follows:
A4, by scheduling behaviorThe completion time delay of caused task j are as follows:
Wherein, CbIndicate the uplink communication rate of UE and base station, CvIndicate the traffic rate of UE and vehicle；fjIt indicates to calculate The communication work amount of the fixation of task j；djIndicate the amount of calculation of the fixation of calculating task j, TjIndicate consolidating for calculating task j Fixed time-constrain；
(1,2 ... m), if the completion time delay of the calculating task j reached in queue i meets by A5, i ∈Then recognize It can be completed for the calculating task of arrival；On the contrary, if the completion time delay of the calculating task reached meetsThen think The calculating task of arrival cannot be completed in time；
I=m+1, if the calculating task j reached is discharged into queue m+1, whenThen think The calculating task of arrival can be completed；
A6, behaviour decision making atCaused StTo St+1State transfer bring return rtAre as follows:
Wherein, rtFirst itemIt indicates to be provided by each service queue in a time interval and calculates money Source bring whole income, rtSection 2It represents and avoids the serious uneven and right of service queue length The punishment of queue length square, rtSection 3Punishment of expression task whether timely complete is to optimize UE User experience,α is the penalty factor of queue length, and β is that task is completed The penalty factor of time delay；
A7, it is calculated according to the following formula when accumulation of discount rewards π*When maximum, corresponding decision at；
Wherein, η is discount factor, o≤η < 1.
Further, the step A7 solves following formula using double DQN deep learning algorithm；Obtain accumulation of discount Reward π*When maximum, corresponding decision at；
Beneficial effects of the present invention: method of the invention uses the computing resource and road driving of base station side MEC serve The computing resource of vehicle provides unloading scheme for the calculating task of the UE (mobile devices such as mobile phone) in roadside；When by the communication of task Prolong, the reference factor of decision is included in the processing speed of calculation delay and calculation processing queue and queue length, using Markov The variation of the mathematics model analysis state of decision obtains an optimisation strategy space, i.e., corresponding shape using deep learning algorithm When state uses corresponding behavior, so that the Income Maximum of resource provider, and ensure that user's calculating task to a certain extent Delay requirement.
Detailed description of the invention
Fig. 1 is car networking schematic diagram of a scenario provided in an embodiment of the present invention；
Fig. 2 is the solution of the present invention flow chart；
Fig. 3 is the Unloading Model that vehicle provided in an embodiment of the present invention calculates queue.
Specific embodiment
For convenient for those skilled in the art understand that technology contents of the invention, with reference to the accompanying drawing to the content of present invention into one Step is illustrated.
The computing resource of the MEC serve of base station side is seen as the n calculation processing teams run parallel by method of the invention Column, while the driving vehicle in roadside can be seen as a dynamic calculation processing queue, the distribution meter when the computation requests of UE arrive In calculation task to corresponding calculation processing queue, so that the benefit of resource provider is maximum, and guarantee user to a certain extent The delay requirement of calculating task.
As shown in Figure 1: in systems, base station can serve as control centre to the car networking scene that the present invention applies, because it Calculating task can be arranged into corresponding service queue.Vehicle and base station are two different computing resource providers, base It stands and static computing resource is provided, vehicle provides dynamic computing resource.Due to base station be each service queue distribute it is different Fixed computing resource, therefore base station can provide the calculating service queue of m kind different rates.Equally, vehicle can provide one A service queue, because of the mobility of vehicle, service speed has changeability, but can be predicted, each calculating service The income of queue unit time when handling calculating task is ρi(1≤i≤m+1)。
As long as the calculating task of mobile terminal reaches, task will be scheduled to base station a service queue or vehicle one A service queue.Once calculating task enters a queue, it will wait in line at the end of the queue.Simultaneously, it is assumed that user A total of n different calculating tasks, for each calculating task j, there are a fixed communication work amount fj, a fixation Amount of calculation djT is constrained with a regular timej, as shown in table 1, wherein calculating task amount can by the revolution of CPU Lai It indicates.In addition, dividing time into minimum interval τ, and at most there is a calculating task to arrive in each time interval It reaches.So the problem of present invention research, how is calculated distribution of computation tasks to corresponding when the computation requests of user reach In service queue, so that resource provider has maximum value and guarantees that the time delay of UE calculating task is wanted to a certain extent It asks.
1 calculating task of table illustrates information
Based on scene as shown in Figure 1, the present invention provides a kind of edge calculations task discharging method towards car networking, such as Shown in Fig. 2 the following steps are included:
S1, when UE generates calculating task, initiate to calculate unloading solicited message to base station, include pair inside the solicited message The communication task amount size for illustrating the i.e. calculating task of information, the calculating task amount size, delay requirement of the calculating task；
After S2, base station receive the unloading request that UE is sent, the state analysis of system is carried out, according to multiple calculating of base station It handles the state of queue, the state of the available calculation processing queue of road vehicle and UE and calculates the size selection of unloading task properly Unloading queue of the calculation processing queue as UE calculating task, and scheduling result is fed back into UE；
S3, UE establish the communication linkage with calculate node, and unload calculating task and arrive according to the scheduling result of computation requests In the calculation processing queue of corresponding calculate node；
S4, calculation processing queue and feed back to calculation processing result according to the calculating task of the sequential processes UE of FIFO UE。
Step S2 specifically include it is following step by step:
S21, base station obtain the state that current time base station calculates queue according to the occupancy situation of base station calculation processing queue Information.Specifically:
The state of base station calculating queue are as follows:WhereinFor the m of base station A calculation processing queue t moment queue length (calculating task amount size),For the length of vehicle calculation processing queue,It is vehicle calculation processing queue in the available amount of computational resources of t moment, dtCalculating for the t moment UE calculating task generated is appointed Business amount size, ftFor the communication task amount size of the t moment UE calculating task generated.And due to each calculation processing of base station The calculation processing rate of queue is inconsistent, and in each time interval τ, the processing income of calculation processing queue i is ρi(1≤i≤ m+1)。
S22, base station obtain the available calculating of current time vehicle according to the position of UE user and the beacon information of vehicle Handle the status information of queue.Specifically:
Base station obtains the available calculation processing of current time vehicle according to the position of UE user and the beacon information of vehicle The status information of queueSince base station is a control centre, it can be communicated with the vehicle of traveling.Such as vehicle Computing capability, position, driving direction, the vehicle beacon message cycle of speed are interacted with base station.UE has and vehicle Establish the fixed communication radius R of D2D transmission connection.
Since base station has global information of vehicles, it can dispatch applicable vehicle to unload the calculating task of UE.Due to Speed of the vehicle with constant speed V, UE is compared with the vehicle of mobile traveling it is considered that static.Then available vehicle Communication linkage settling time (CCET) between UE, the CCET of UE and the vehicle in transmission range can be assessed as:
Wherein, P is position vector of the vehicle with respect to UE, and V is the velocity vector of vehicle driving.
If the calculating task of UE is scheduled to vehicle, selection is had maximum t by base stationccetVehicle unloaded, this Ensure the communication connection time long enough between vehicle and UE to keep the reliability communicated.At the same time, the vehicle of selection will The auxiliary vehicle of surrounding is assigned the task within the scope of single-hop communication.As shown in figure 3, VsIt is that there is maximum tccetVehicle, Va, Vb, VcIt is in VsTransmission range in auxiliary vehicle, and available computational resources Rv=(Ra+Rb+Rc).Base station will be periodical Status information of the update about vehicle calculation processing queue
S23, base station are according to the state of the calculation processing queue of base station, the transition of vehicle calculation processing queue, current needs The calculating task of scheduling illustrates information, selects suitable calculation processing queue unloading as the calculating task according to dispatching algorithm Carry queue.
The scheduling problem of calculating task can be considered as Markovian decision model in step S23.I.e. base station is according to base station The state of calculation processing queue, vehicle calculation processing queue transition, currently need the calculating task dispatched illustrate information, root Select suitable calculation processing queue as the unloading queue of the calculating task according to dispatching algorithm.The following steps are included:
The system mode S of A1, moment ttForScheduling decision is
Wherein,Mean that the calculating task reached in time t is put into corresponding calculation processing queue i.
A2, each calculation processing queue treating capacity of actual calculating task in time interval τ areI ∈ (1,2 ... m+1), Vi are the calculation processing rate of calculation processing queue i, dispatch probabilityWhen being 1, indicate the moment t calculating task reached to be scheduled to calculating task processing queue i,It indicates moment t to arrive when being 0 The calculating task reached is not scheduled to calculating task processing queue i.
A3, so the system mode of moment t+1 can be derived as following formula:
A4, analyze the state transfer bring return before, need the completion time delay of analysis task, the uplink of UE and base station Traffic rate is believed that as Cb, the traffic rate of UE and vehicle is believed that as Cv.Scheduling behaviorMean in time t arrival Calculating task is put into corresponding calculation processing queue i, then by scheduling behaviorThe completion time delay of caused task j are as follows:
Each calculating task j has fixed communication work amount fj, amount of calculation djWith time-constrain Tj。
If A5, at queue i (i ∈ (1,2 ... m))) in reach calculating task j completion time delay meetThen Think that the calculating task reached can be completed.On the contrary, if the completion time delay of the calculating task reached meetsThen think The calculating task of arrival cannot be completed in time.If the calculating task j reached is discharged into queue m+1, task can be complete At and if only if
A6, due to behaviour decision making atCaused StTo St+1State transfer bring return rtAre as follows:
Wherein,
Wherein, α is the penalty factor of queue length, and β is the penalty factor that task completes time delay.About rtFirst item be Computing resource bring whole income is provided by each service queue in a time interval, Section 2 is in order to avoid service Queue length serious uneven and the punishment to queue length square, last punishment for task whether timely complete To optimize the user experience of UE.
A7, computing resource provider will not only consider the return at current time, also want to obtain better achievement for a long time Consider the future returns that will be obtained.Final target is one optimal scheduling strategy of study to maximize accumulation of discount prize It encourages, i.e.,Wherein η (o≤η < 1) is discount factor.When t is sufficiently large, ηtTend to 0, represents rtThere is minor impact to Total Return.
It is solved in step A7Double DQN deep learning algorithm can be used, specifically Algorithm steps it is as follows:
S3, UE according to base station about the scheduled feedback of computation requests as a result, establish and the communication linkage of calculate node, and unload Calculating task is carried into the calculation processing queue of corresponding calculate node.
S4, calculation processing queue according to the sequential processes UE of first in, first out calculating task, and it is calculation processing result is anti- Feed UE.
Those of ordinary skill in the art will understand that the embodiments described herein, which is to help reader, understands this hair Bright principle, it should be understood that protection scope of the present invention is not limited to such specific embodiments and embodiments.For ability For the technical staff in domain, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made Any modification, equivalent substitution, improvement and etc. should be included within scope of the presently claimed invention.

Claims (4)

1. the calculating task discharging method towards car networking characterized by comprising
S1, when UE generates calculating task, initiate to calculate unloading solicited message to base station, include to the meter inside the solicited message Calculation task illustrates information；
After S2, base station receive the calculating unloading request that UE is sent, according to the state of multiple calculation processing queues of base station, road The size combination Markovian decision model that the state and UE of the available calculation processing queue of vehicle calculate unloading task determines Unloading queue of some calculation processing queue as UE calculating task；
S3, UE establish the communication linkage with calculate node according to the scheduling result of feedback, and unload calculating task and calculate to corresponding In the calculation processing queue of node；
Calculation processing result and is fed back to UE according to the calculating task of the sequential processes UE of FIFO by S4, calculation processing queue.
2. the calculating task discharging method according to claim 1 towards car networking, which is characterized in that described in step S1 Illustrate that information includes: communication task amount size, calculating task amount size and the delay requirement of calculating task.
3. the calculating task discharging method according to claim 2 towards car networking, which is characterized in that UE described in step S2 The unloading queue determination process of calculating task specifically include it is following step by step:
A1, moment t system mode beScheduling decision is
Wherein,Expression is put into corresponding calculation processing queue i in the calculating task that time t is reached；For the m of base station A calculation processing queue t moment calculating task amount size,For the calculating task amount size of vehicle calculation processing queue,It is vehicle calculation processing queue in the available amount of computational resources of t moment, dtCalculating for the t moment UE calculating task generated is appointed Business amount size, ftFor the communication task amount size of the t moment UE calculating task generated；
The treating capacity of A2, each calculation processing queue actual calculating task in time interval τ are as follows:
Wherein, ViFor the calculation processing rate of calculation processing queue i,Indicate scheduling probability；
The system shape of A3, moment t+1 are as follows:
A4, by scheduling behaviorThe completion time delay of caused task j are as follows:
Wherein, CbIndicate the uplink communication rate of UE and base station, CvIndicate the traffic rate of UE and vehicle；fjIndicate calculating task j Fixation communication work amount；djIndicate the amount of calculation of the fixation of calculating task j, TjIndicate calculating task j fixation when Between constrain；
(1,2 ... m), if the completion time delay of the calculating task j reached in queue i meets T by A5, i ∈a t≤Tj, then it is assumed that it arrives The calculating task reached can be completed；On the contrary, if the completion time delay of the calculating task reached meets Ta t> Tj, then it is assumed that arrival Calculating task cannot be completed in time；
I=m+1 works as T if the calculating task j reached is discharged into queue m+1a t≤min(Tj,tccet), then it is assumed that it arrives The calculating task reached can be completed；
A6, behaviour decision making atCaused StTo St+1State transfer bring return rtAre as follows:
Wherein, rtFirst itemIndicate that providing computing resource by each service queue in a time interval brings Whole incomes, rtSection 2It represents and avoids the serious uneven and long to queue of service queue length The punishment of degree square, rtSection 3Punishment of expression task whether timely complete is to optimize the user of UE Experience,α is the penalty factor of queue length, and β is that task completes time delay Penalty factor；
A7, it is calculated according to the following formula when accumulation of discount rewards π*When maximum, corresponding decision at；
Wherein, η is discount factor, o≤η < 1.
4. the calculating task discharging method according to claim 3 towards car networking, which is characterized in that the step A7 is adopted Following formula is solved with double DQN deep learning algorithm；Obtain accumulation of discount reward π*When maximum, corresponding decision at；
CN201810734390.3A 2018-07-06 2018-07-06 Calculation task unloading method facing Internet of vehicles Active CN109067842B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810734390.3A CN109067842B (en) 2018-07-06 2018-07-06 Calculation task unloading method facing Internet of vehicles

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810734390.3A CN109067842B (en) 2018-07-06 2018-07-06 Calculation task unloading method facing Internet of vehicles

Publications (2)

Publication Number Publication Date
CN109067842A true CN109067842A (en) 2018-12-21
CN109067842B CN109067842B (en) 2020-06-26

Family

ID=64818820

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810734390.3A Active CN109067842B (en) 2018-07-06 2018-07-06 Calculation task unloading method facing Internet of vehicles

Country Status (1)

Country Link
CN (1) CN109067842B (en)

Cited By (30)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109756378A (en) * 2019-01-12 2019-05-14 大连理工大学 A kind of intelligence computation discharging method under In-vehicle networking
CN109862086A (en) * 2019-01-21 2019-06-07 华北电力大学 Task Assigned Policy based on matching algorithm in a kind of vehicle-mounted edge calculations
CN109933427A (en) * 2019-02-19 2019-06-25 广东工业大学 A kind of vehicle mist calculate in the car networking task immigration method based on direction
CN109992387A (en) * 2019-04-01 2019-07-09 北京邮电大学 A kind of task processing method, device and the electronic equipment of terminal collaboration
CN110058928A (en) * 2019-03-25 2019-07-26 南京工程学院 Based on the calculation task allocating method of martingale theory in isomery car networking
CN110347495A (en) * 2019-07-24 2019-10-18 张�成 A kind of task immigration method carrying out mobile edge calculations using deeply study
CN110505165A (en) * 2019-08-26 2019-11-26 北京邮电大学 A kind of edge calculations task discharging method and device based on two way auction mechanism
CN110647403A (en) * 2019-10-31 2020-01-03 桂林电子科技大学 Cloud computing resource allocation method in multi-user MEC system
CN110928599A (en) * 2019-11-06 2020-03-27 浙江理工大学 Task unloading method and system based on data flow in edge computing network
CN110928691A (en) * 2019-12-26 2020-03-27 广东工业大学 Traffic data-oriented edge collaborative computing unloading method
CN111160755A (en) * 2019-12-26 2020-05-15 西北工业大学 DQN-based real-time scheduling method for aircraft overhaul workshop
CN111212128A (en) * 2019-12-30 2020-05-29 北京邮电大学 Edge-assisted-based collaborative Web AR multi-user service system and method
CN111355779A (en) * 2020-02-18 2020-06-30 湖南大学 Service-based Internet of vehicles task unloading method and unloading device thereof
CN111479238A (en) * 2020-04-14 2020-07-31 东华大学 Internet of vehicles information processing time delay optimization method
CN111538583A (en) * 2020-06-23 2020-08-14 中国人民解放军国防科技大学 Low-delay collaborative task processing method and device for Internet of vehicles in mobile environment
CN111611063A (en) * 2020-05-27 2020-09-01 江南大学 Cloud-aware mobile fog computing system task unloading method based on 802.11p
CN111786839A (en) * 2020-07-15 2020-10-16 南通大学 Calculation unloading method and system for energy efficiency optimization in vehicle-mounted edge calculation network
WO2020216135A1 (en) * 2019-04-25 2020-10-29 南京邮电大学 Multi-user multi-mec task unloading resource scheduling method based on edge-end collaboration
CN111866807A (en) * 2020-06-22 2020-10-30 江苏大学 Software definition vehicle-mounted task fine-grained unloading method based on deep reinforcement learning
CN111984383A (en) * 2020-08-18 2020-11-24 薛东 Service data processing method and cloud platform based on cloud network fusion and artificial intelligence
CN112422346A (en) * 2020-11-19 2021-02-26 北京航空航天大学 Variable-period mobile edge computing unloading decision method considering multi-resource limitation
CN112445617A (en) * 2020-11-25 2021-03-05 中国石油大学(华东) Load strategy selection method and system based on mobile edge calculation
CN112512018A (en) * 2020-07-24 2021-03-16 北京航空航天大学 Method for dynamically unloading tasks among cooperative vehicles based on mobile edge calculation
CN112584351A (en) * 2020-12-08 2021-03-30 重庆邮电大学 Communication-calculation integrated resource allocation method for vehicle-to-vehicle fog calculation
CN112822234A (en) * 2020-12-29 2021-05-18 华北电力大学 Task unloading method based on deep reinforcement learning in Internet of vehicles
CN112888021A (en) * 2021-01-29 2021-06-01 重庆邮电大学 Task unloading method for avoiding interruption in Internet of vehicles
CN113364859A (en) * 2021-06-03 2021-09-07 吉林大学 MEC-oriented joint computing resource allocation and unloading decision optimization scheme in Internet of vehicles
CN114363857A (en) * 2022-03-21 2022-04-15 山东科技大学 Method for unloading edge calculation tasks in Internet of vehicles
CN115002123A (en) * 2022-05-25 2022-09-02 西南交通大学 Fast adaptive task unloading system and method based on mobile edge calculation
CN117221950A (en) * 2023-11-08 2023-12-12 南京邮电大学 Vehicle task unloading method and system based on deep reinforcement learning

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090124241A1 (en) * 2007-11-14 2009-05-14 Qualcomm Incorporated Method and system for user profile match indication in a mobile environment
CN104869151A (en) * 2015-04-07 2015-08-26 北京邮电大学 Business unloading method and system
CN105391782A (en) * 2015-11-16 2016-03-09 上海交通大学 V2I/V2V Internet-of-things system based on cognitive OFDM and method thereof
CN106210152A (en) * 2016-09-27 2016-12-07 桂林电子科技大学 A kind of vehicle-mounted cloud system based on Internet of Things and resource acquiring method
CN106534333A (en) * 2016-11-30 2017-03-22 北京邮电大学 Bidirectional selection computing unloading method based on MEC and MCC
CN206060828U (en) * 2016-09-27 2017-03-29 桂林电子科技大学 A kind of vehicle-mounted cloud system based on Internet of Things
CN107145387A (en) * 2017-05-23 2017-09-08 南京大学 A kind of method for scheduling task learnt under vehicle-mounted net environment based on deeply
CN107295109A (en) * 2017-08-16 2017-10-24 重庆邮电大学 Task unloading and power distribution joint decision method in self-organizing network cloud computing
CN107333267A (en) * 2017-06-23 2017-11-07 电子科技大学 A kind of edge calculations method for 5G super-intensive networking scenes
CN107995660A (en) * 2017-12-18 2018-05-04 重庆邮电大学 Support Joint Task scheduling and the resource allocation methods of D2D- Edge Servers unloading

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090124241A1 (en) * 2007-11-14 2009-05-14 Qualcomm Incorporated Method and system for user profile match indication in a mobile environment
CN104869151A (en) * 2015-04-07 2015-08-26 北京邮电大学 Business unloading method and system
CN105391782A (en) * 2015-11-16 2016-03-09 上海交通大学 V2I/V2V Internet-of-things system based on cognitive OFDM and method thereof
CN106210152A (en) * 2016-09-27 2016-12-07 桂林电子科技大学 A kind of vehicle-mounted cloud system based on Internet of Things and resource acquiring method
CN206060828U (en) * 2016-09-27 2017-03-29 桂林电子科技大学 A kind of vehicle-mounted cloud system based on Internet of Things
CN106534333A (en) * 2016-11-30 2017-03-22 北京邮电大学 Bidirectional selection computing unloading method based on MEC and MCC
CN107145387A (en) * 2017-05-23 2017-09-08 南京大学 A kind of method for scheduling task learnt under vehicle-mounted net environment based on deeply
CN107333267A (en) * 2017-06-23 2017-11-07 电子科技大学 A kind of edge calculations method for 5G super-intensive networking scenes
CN107295109A (en) * 2017-08-16 2017-10-24 重庆邮电大学 Task unloading and power distribution joint decision method in self-organizing network cloud computing
CN107995660A (en) * 2017-12-18 2018-05-04 重庆邮电大学 Support Joint Task scheduling and the resource allocation methods of D2D- Edge Servers unloading

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
于博文等: "移动边缘计算任务卸载和基站关联协同决策问题研究", 《计算机研究与发展》 *
周芳芳: "基于移动P2P的车联网下载方法的研究与实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *
张强: "移动云计算中的任务卸载技术研究", 《中国博士学位论文全文数据库(电子期刊)》 *
张科等: "车云网络中的移动边缘计算的延迟卸载", 《 2016年第八届弹性网络设计与建模国际研讨会（RNDM）》 *
王秋宁等: "移动边缘计算的移动性管理研究", 《中兴通讯技术》 *

Cited By (44)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109756378B (en) * 2019-01-12 2021-07-16 大连理工大学 Intelligent computing unloading method under vehicle-mounted network
CN109756378A (en) * 2019-01-12 2019-05-14 大连理工大学 A kind of intelligence computation discharging method under In-vehicle networking
CN109862086A (en) * 2019-01-21 2019-06-07 华北电力大学 Task Assigned Policy based on matching algorithm in a kind of vehicle-mounted edge calculations
CN109862086B (en) * 2019-01-21 2020-10-09 华北电力大学 Task allocation strategy based on matching algorithm in vehicle-mounted edge calculation
CN109933427A (en) * 2019-02-19 2019-06-25 广东工业大学 A kind of vehicle mist calculate in the car networking task immigration method based on direction
CN110058928A (en) * 2019-03-25 2019-07-26 南京工程学院 Based on the calculation task allocating method of martingale theory in isomery car networking
CN109992387A (en) * 2019-04-01 2019-07-09 北京邮电大学 A kind of task processing method, device and the electronic equipment of terminal collaboration
CN109992387B (en) * 2019-04-01 2021-09-24 北京邮电大学 Terminal collaborative task processing method and device and electronic equipment
WO2020216135A1 (en) * 2019-04-25 2020-10-29 南京邮电大学 Multi-user multi-mec task unloading resource scheduling method based on edge-end collaboration
CN110347495A (en) * 2019-07-24 2019-10-18 张�成 A kind of task immigration method carrying out mobile edge calculations using deeply study
CN110347495B (en) * 2019-07-24 2023-04-28 张�成 Task migration method for performing mobile edge calculation by using deep reinforcement learning
CN110505165A (en) * 2019-08-26 2019-11-26 北京邮电大学 A kind of edge calculations task discharging method and device based on two way auction mechanism
CN110647403A (en) * 2019-10-31 2020-01-03 桂林电子科技大学 Cloud computing resource allocation method in multi-user MEC system
CN110928599A (en) * 2019-11-06 2020-03-27 浙江理工大学 Task unloading method and system based on data flow in edge computing network
CN110928599B (en) * 2019-11-06 2023-04-18 浙江理工大学 Task unloading method and system based on data flow in edge computing network
CN111160755B (en) * 2019-12-26 2023-08-18 西北工业大学 Real-time scheduling method for aircraft overhaul workshop based on DQN
CN111160755A (en) * 2019-12-26 2020-05-15 西北工业大学 DQN-based real-time scheduling method for aircraft overhaul workshop
CN110928691A (en) * 2019-12-26 2020-03-27 广东工业大学 Traffic data-oriented edge collaborative computing unloading method
CN111212128A (en) * 2019-12-30 2020-05-29 北京邮电大学 Edge-assisted-based collaborative Web AR multi-user service system and method
CN111212128B (en) * 2019-12-30 2021-05-04 北京邮电大学 Edge-assisted-based collaborative Web AR multi-user service system and method
CN111355779A (en) * 2020-02-18 2020-06-30 湖南大学 Service-based Internet of vehicles task unloading method and unloading device thereof
CN111479238A (en) * 2020-04-14 2020-07-31 东华大学 Internet of vehicles information processing time delay optimization method
CN111611063B (en) * 2020-05-27 2023-04-11 江南大学 Cloud-aware mobile fog computing system task unloading method based on 802.11p
CN111611063A (en) * 2020-05-27 2020-09-01 江南大学 Cloud-aware mobile fog computing system task unloading method based on 802.11p
CN111866807A (en) * 2020-06-22 2020-10-30 江苏大学 Software definition vehicle-mounted task fine-grained unloading method based on deep reinforcement learning
CN111538583B (en) * 2020-06-23 2020-10-20 中国人民解放军国防科技大学 Low-delay collaborative task processing method and device for Internet of vehicles in mobile environment
CN111538583A (en) * 2020-06-23 2020-08-14 中国人民解放军国防科技大学 Low-delay collaborative task processing method and device for Internet of vehicles in mobile environment
US11445400B2 (en) 2020-07-15 2022-09-13 Nantong University Energy-efficient optimized computing offloading method for vehicular edge computing network and system thereof
CN111786839A (en) * 2020-07-15 2020-10-16 南通大学 Calculation unloading method and system for energy efficiency optimization in vehicle-mounted edge calculation network
CN111786839B (en) * 2020-07-15 2021-09-07 南通大学 Calculation unloading method and system for energy efficiency optimization in vehicle-mounted edge calculation network
CN112512018A (en) * 2020-07-24 2021-03-16 北京航空航天大学 Method for dynamically unloading tasks among cooperative vehicles based on mobile edge calculation
CN112512018B (en) * 2020-07-24 2022-03-04 北京航空航天大学 Method for dynamically unloading tasks among cooperative vehicles based on mobile edge calculation
CN111984383B (en) * 2020-08-18 2021-05-14 上海悟景信息科技有限公司 Service data processing method and cloud platform based on cloud network fusion and artificial intelligence
CN111984383A (en) * 2020-08-18 2020-11-24 薛东 Service data processing method and cloud platform based on cloud network fusion and artificial intelligence
CN112422346A (en) * 2020-11-19 2021-02-26 北京航空航天大学 Variable-period mobile edge computing unloading decision method considering multi-resource limitation
CN112445617A (en) * 2020-11-25 2021-03-05 中国石油大学(华东) Load strategy selection method and system based on mobile edge calculation
CN112584351A (en) * 2020-12-08 2021-03-30 重庆邮电大学 Communication-calculation integrated resource allocation method for vehicle-to-vehicle fog calculation
CN112822234A (en) * 2020-12-29 2021-05-18 华北电力大学 Task unloading method based on deep reinforcement learning in Internet of vehicles
CN112888021A (en) * 2021-01-29 2021-06-01 重庆邮电大学 Task unloading method for avoiding interruption in Internet of vehicles
CN113364859A (en) * 2021-06-03 2021-09-07 吉林大学 MEC-oriented joint computing resource allocation and unloading decision optimization scheme in Internet of vehicles
CN114363857B (en) * 2022-03-21 2022-06-24 山东科技大学 Method for unloading edge calculation tasks in Internet of vehicles
CN114363857A (en) * 2022-03-21 2022-04-15 山东科技大学 Method for unloading edge calculation tasks in Internet of vehicles
CN115002123A (en) * 2022-05-25 2022-09-02 西南交通大学 Fast adaptive task unloading system and method based on mobile edge calculation
CN117221950A (en) * 2023-11-08 2023-12-12 南京邮电大学 Vehicle task unloading method and system based on deep reinforcement learning

Also Published As

Publication number Publication date
CN109067842B (en) 2020-06-26

Similar Documents

Publication Publication Date Title
CN109067842A (en) Calculating task discharging method towards car networking
Xiong et al. Deep reinforcement learning for mobile 5G and beyond: Fundamentals, applications, and challenges
Ning et al. Intelligent edge computing in internet of vehicles: A joint computation offloading and caching solution
Liu et al. Path planning for UAV-mounted mobile edge computing with deep reinforcement learning
Liang et al. A novel adaptive resource allocation model based on SMDP and reinforcement learning algorithm in vehicular cloud system
CN111464976A (en) Vehicle task unloading decision and overall resource allocation method based on fleet
Kazmi et al. A novel deep reinforcement learning-based approach for task-offloading in vehicular networks
Yuan et al. A survey on computation offloading for vehicular edge computing
Ju et al. Computation offloading and tasks scheduling for the internet of vehicles in edge computing: A deep reinforcement learning-based pointer network approach
Nguyen et al. Flexible computation offloading in a fuzzy-based mobile edge orchestrator for IoT applications
Puligheddu et al. SEM-O-RAN: Semantic and flexible O-RAN slicing for NextG edge-assisted mobile systems
Mu et al. Deep reinforcement learning based adaptive threshold multi-tasks offloading approach in mec
Nguyen et al. EdgePV: collaborative edge computing framework for task offloading
Zhang et al. Vehicular multi-slice optimization in 5G: Dynamic preference policy using reinforcement learning
Cui et al. Cooperative vehicles‐assisted task offloading in vehicular networks
Zhang et al. A UCB-based dynamic CoAP mode selection algorithm in distribution IoT
Mirza et al. MCLA task offloading framework for 5G-NR-V2X-based heterogeneous VECNs
Ma et al. Edge computing and UAV swarm cooperative task offloading in vehicular networks
Zhao et al. A research of task-offloading algorithm for distributed vehicles
Khani et al. ARL-RA: efficient resource allocation in 5G edge networks: a novel intelligent solution using approximate reinforcement learning algorithm
CN113783959A (en) Internet of vehicles joint optimization calculation task unloading rate and resource allocation method
Gupta et al. Global edge bandwidth cost gradient-based heuristic for fast data delivery to connected vehicles under vehicle overlaps
CN113448707A (en) Online batch parallel task scheduling method in edge computing
Wang et al. Research on task offloading optimization strategies for vehicular networks based on game theory and deep reinforcement learning
Chai et al. Task scheduling of computation-intensive graph jobs in UAV-assisted hybrid vehicular networks

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