[go: up one dir, main page]

CN109905859B - Efficient edge computing migration method for Internet of vehicles application - Google Patents

Efficient edge computing migration method for Internet of vehicles application Download PDF

Info

Publication number
CN109905859B
CN109905859B CN201910031387.XA CN201910031387A CN109905859B CN 109905859 B CN109905859 B CN 109905859B CN 201910031387 A CN201910031387 A CN 201910031387A CN 109905859 B CN109905859 B CN 109905859B
Authority
CN
China
Prior art keywords
edge
vehicle
calculation
migration
vehicles
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.)
Active
Application number
CN201910031387.XA
Other languages
Chinese (zh)
Other versions
CN109905859A (en
Inventor
许小龙
刘庆祥
李袁成
薛原
陈裕豪
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.)
Shanghai Changxing Information Technology Co ltd
Shanghai Changxing Software Co ltd
Original Assignee
Nanjing University of Information Science and 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 Nanjing University of Information Science and Technology filed Critical Nanjing University of Information Science and Technology
Priority to CN201910031387.XA priority Critical patent/CN109905859B/en
Publication of CN109905859A publication Critical patent/CN109905859A/en
Application granted granted Critical
Publication of CN109905859B publication Critical patent/CN109905859B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Landscapes

  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention provides an efficient edge computing migration method for Internet of vehicles application, which comprises the following steps: s1, acquiring current position information of the vehicle in the Internet of vehicles and application requirements of a migration task generated by the vehicle; s2, acquiring the position information of all edge calculation nodes in the Internet of vehicles and the calculation resource conditions of the edge nodes; s3, filtering out the edge calculation nodes which do not meet the conditions according to the distance between the vehicle and the edge calculation nodes and the application requirements of the migration tasks; s4, calculating the time and energy consumption required by each migration strategy; and S5, obtaining the optimal calculation migration strategy through a simple weighting method and a multi-standard decision algorithm. The method of the invention reduces the energy consumption of the edge computing node to the maximum extent and improves the utilization rate of the edge computing node while meeting the requirements of computing capacity and computing delay of computing tasks.

Description

Efficient edge computing migration method for Internet of vehicles application
Technical Field
The invention relates to an efficient edge calculation migration method, and belongs to the technical field of mobile edge calculation.
Background
With the continuous development of the transportation industry and the wide concern of people on traffic safety, the Internet of Vehicles (IoV) is gradually developed. As one of the main applications of the Internet of things, the Internet of vehicles better realizes information interaction among vehicles, between people and vehicles and between vehicles and Road Side Units (RSUs), and greatly improves traffic safety. With the increasing popularity of the internet of vehicles, applications based on the internet of vehicles are increasing, especially many applications with high requirements on computing power and response delay, and the demands of many applications at present far exceed the computing processing power of the vehicles.
When mobile cloud computing is introduced into an environment of the internet of vehicles, a vehicle can transfer a computing task to a remote cloud platform for execution, which is equivalent to expanding computing resources of the vehicle and meeting the computing capability requirement of the computing task, but the time required for the vehicle to transfer the computing task to the remote cloud platform through a Wide Area Network (WAN) is long, and the low-delay requirement of the computing task cannot be met.
The mobile edge computing well solves the problem of long computing time delay of a remote cloud platform. An Edge Computing Device (ECD) composed of the RSU and the server is a small cloud, which is usually deployed near a vehicle to provide enhanced cloud services, so that migration of a Computing task by the vehicle to the ECD can greatly reduce time delay and improve service experience of a user. However, the resources of the ECD are limited, and the energy consumption of the edge computing nodes for executing the computing migration task is high. Therefore, how to effectively utilize the resources of the ECD is the focus of current scientific research. In the specific operation, the calculation task cannot be randomly migrated to the ECD for execution, and a reasonable calculation migration method is required to realize reasonable distribution of ECD resources, meet the requirement of low time delay of the calculation task and reduce energy consumption of the edge calculation nodes to the maximum extent.
At present, many scientific researchers are dedicated to designing an efficient Computing migration mechanism to improve the utilization rate of cloud, for example, m.wang et al consider to introduce Fog Computing (Fog Computing) into IoV environment in "heated Mobility Support for Information-central IoV in Smart City Using Fog Computing", and design a dynamic service Support mechanism according to different service characteristics; cao et al, in the "QoE-based selection strategy for edge computing enabled Internet-of-Vehicles (EC-IoV)", proposed the concept of edge computing based on vehicle networking (EC-IoV), utilized networked Vehicles as edge computing platforms, and designed a node selection strategy based on Quality of Experience (QoE) of users to select optimal edge computing nodes for users. However, the current research on the computing migration technology in the internet of vehicles hardly considers the problem of energy consumption of the ECD, and the energy consumption of the ECD is not negligible when the migration strategy can meet the delay requirement of the computing task.
Disclosure of Invention
Aiming at the condition that only the time delay requirement is considered but the ECD energy consumption problem is not considered in the existing Internet of vehicles calculation and migration technology, the invention provides an efficient edge calculation and migration method for Internet of vehicles application.
In order to solve the technical problems, the invention adopts the following technical means:
an efficient edge computing migration method for car networking application specifically comprises the following steps:
s1, acquiring current position information of the vehicle in the Internet of vehicles and application requirements of a migration task generated by the vehicle;
s2, acquiring the position information of all edge calculation nodes in the Internet of vehicles and the calculation resource condition of each edge node;
s3, filtering out the edge calculation nodes which do not meet the conditions according to the distance between the vehicle and the edge calculation nodes and the application requirements of the migration tasks;
s4, matching the vehicle calculation tasks with the edge nodes meeting the conditions one by one, and calculating the time and energy consumption required by each migration strategy;
and S5, obtaining the optimal calculation migration strategy through a simple weighting method and a multi-standard decision algorithm.
Further, the specific operation of step S1 is as follows:
s11, M vehicles and M-th vehicle c are contained in the Internet of vehicles area AmCoordinate cp at time im,iThe following were used:
cpm,i＝(cpxm,i,cpym,i) (1)
wherein, cpxm,iVehicle c at time imPosition on the abscissa in area A, cpym,iVehicle c at time imAt the ordinate position in the area a, M is 1,2, …, M.
S12, m-th vehicle cmThe resulting computing task is tm＝(twm,trm,tw'm) Wherein, twmRepresenting a computational task tmTask amount of, trmIndicates the execution of tmRequired resource, tw'mIndicating the amount of data returned by the end of execution.
Further, in step S2, N edge calculation nodes are set in the car networking area a, and the expressions of the edge calculation nodes are as follows:
en＝(epxn,epyn,eqn,ern) (2)
wherein e isnRepresents the N-th edge calculation node, N is 1,2, …, N, epxnDenotes enPosition on the abscissa in area A, epynDenotes enOrdinate position in area A, eqnDenotes enTotal capacity of (1), ernDenotes enThe idle resources of (1).
Further, trm、eqnAnd ernAll in the form of the number of virtual machines.
Further, the specific operation of step S3 is as follows:
s31, comparing the idle resources of all edge calculation nodes with the size of the resources required by executing vehicle tasks, when tr ism＞ernVehicle cmIs calculated task tmThe edge computation node cannot be migrated to.
S32 calculation vehicle cmCalculation node e to edgenThe distance of (c):
Figure BDA0001944361660000031
s33, current dis (c)m,en) < rho, rho being enOf a vehicle cmIs calculated task tmThe movement to the edge calculation node is impossible, and the movement to the edge node on the side opposite to the vehicle traveling direction is impossible.
Further, the specific operation of step S4 is as follows:
and S41, selecting a qualified target edge calculation node and a target vehicle in the action range of the target node.
S42 calculation vehicle cmComputing task tmTo the target vehicle cnTime Tt ofmTarget vehicle cnHandle tmMigration to target edge computation node enTime To ofm、enTime Te of executing a computing taskmAnd enFeeding back the calculation result to cmTime Tf ofmThe concrete formula is as follows:
Figure BDA0001944361660000032
Figure BDA0001944361660000033
Figure BDA0001944361660000034
Figure BDA0001944361660000035
where v denotes the transmission rate between the vehicles, λm,nRepresenting computational tasksFrom cmIs transmitted to cnThe number of vehicles passing by, v' represents the transfer rate between the vehicle and the edge calculation node, and p represents the calculation capability of each virtual machine.
S43, calculating the total time delay T required by the migration strategy:
Figure BDA0001944361660000041
s44, calculating edge calculating node enBase energy consumption EbnIdle energy consumption EinAnd occupy energy consumption Eun：
Ebn＝Tsn·Pα (9)
Figure BDA0001944361660000042
Figure BDA0001944361660000043
Wherein, TsnDenotes enRun time of the server of (1), PαRepresenting edge calculation node enPower of the server, Bm,nRepresenting a computational task tmWhether or not in enUpper run, PβDenotes enPower of the resource unoccupied, PγDenotes enPower of the occupied resource.
S45, calculating edge calculating node enTotal energy consumption E:
Figure BDA0001944361660000044
further, the specific operation of step S5 is as follows:
s51, respectively normalizing the total time delay and the total energy consumption of each calculation migration strategy into the following values by a simple weighting method and a multi-standard decision algorithm:
Figure BDA0001944361660000045
Figure BDA0001944361660000046
wherein, TmaxAnd TminRespectively representing the maximum and minimum delays resulting from the computation of the migration, EmaxAnd EminRespectively representing the maximum and minimum energy consumption resulting from computing migration.
S52, calculating utility values of all migration strategies, and obtaining a calculated migration strategy with the maximum utility value:
UV＝V(T)·ωT+V(E)·ωE(ωT+ωE＝1) (15)
wherein, ω isTDenotes the weight, ω, of V (T)EThe weights of V (E) and (E) are shown respectively.
The following advantages can be obtained by adopting the technical means:
the invention provides an efficient edge calculation migration method for vehicle networking application, which is used for acquiring information of vehicles and edge calculation nodes in a vehicle networking, filtering out the edge calculation nodes which do not meet conditions according to the application resource requirements of calculation migration tasks, planning migration strategies according to the nodes which meet the requirements, calculating the time required by each migration strategy and the generated energy consumption, and finally selecting the optimal calculation migration strategy according to SAW and MCDM. According to the method, the vehicle driving condition is considered, the transmission between vehicles and the migration technology between the vehicle and the side calculation node are applied in the calculation and migration process, the calculation task can be normally migrated, and the efficiency of the migration process is improved; meanwhile, the calculation migration strategy of the method is dynamically changed along with the specific vehicle information and the information of the side calculation nodes, so that the calculation migration result is more objective and credible. Compared with the traditional calculation migration method, the method comprehensively considers the delay of executing the calculation task and the generated energy consumption, reduces the energy consumption of the calculation nodes to the maximum extent on the premise of meeting the requirements of the calculation capacity and the calculation delay of the calculation task, accords with the theme of green calculation, reasonably plans the migration of the calculation task, and maximizes the utilization efficiency of the calculation nodes.
Drawings
FIG. 1 is a flow chart of the steps of an efficient edge computing migration method for Internet of vehicles applications in accordance with the present invention.
FIG. 2 is a diagram of an example of an efficient edge computing migration method for Internet of vehicles applications in accordance with the present invention.
Detailed Description
The technical scheme of the invention is further explained by combining the accompanying drawings as follows:
an efficient edge computing migration method for car networking applications is shown in fig. 1, and specifically includes the following steps:
s1, acquiring current position information of the vehicle in the Internet of vehicles and application requirements of a migration task generated by the vehicle; the specific operation is as follows:
s11, assuming that M vehicles are driven on one road, adopting an X-Y coordinate system to represent the road area, defining an internet of vehicles area A { (X, Y) |0 ≦ X ≦ X,0 ≦ Y ≦ Y }, and C ═ { C { (X, Y) |0 ≦ X ≦ Y }, wherein1,c2,…,cMDenotes a set of vehicles, cmRepresenting the m-th vehicle in area A, cmCoordinate cp at time im,iComprises the following steps:
cpm,i＝(cpxm,i,cpym,i) (16)
wherein, cpxm,iVehicle c at time imPosition on the abscissa in area A, cpym,iVehicle c at time imAt the ordinate position in the area a, M is 1,2, …, M.
S12, each vehicle may generate a computing task to be performed, so that in the car networking environment, M computing tasks are generated, and the computing tasks may be represented as a set T ═ { T ═ M1,t2,…,tMWhere t ismRepresenting the computational tasks generated by the m-th vehicle.
M vehicle cmThe resulting computational task may be denoted as tm＝(twm,trm,tw'm) Wherein, twmRepresenting a computational task tmTask amount of, trmIndicates the execution of tmRequired resource, tw'mIndicating the amount of data returned by the end of execution. In the method, the resources required by the computing task are the number of the virtual machine instances occupied by the execution of the computing task.
S2, acquiring the position information of all edge calculation nodes in the Internet of vehicles and the calculation resource condition of each edge node; setting N edge calculation nodes in the Internet of vehicles area A, wherein E is { E ═ E1,e2,…,eNThe expression of the edge calculation node is as follows:
en＝(epxn,epyn,eqn,ern) (17)
wherein e isnRepresents the N-th edge calculation node, N is 1,2, …, N, epxnDenotes enPosition on the abscissa in area A, epynDenotes enOrdinate position in area A, eqnDenotes enTotal capacity of (1), ernDenotes enThe idle resources of (1).
In the method, the idle resources and the total capacity of the edge computing nodes are both given in the form of the number of the virtual machines, the physical resources of the edge computing nodes are averagely divided into a plurality of virtual machines, and when a computing task is migrated to the edge computing nodes for execution, the corresponding virtual machines are instantiated.
S3, filtering out the edge calculation nodes which do not meet the conditions according to the distance between the vehicle and the edge calculation nodes and the application requirements of the migration tasks; the specific operation is as follows:
s31, comparing the free resources of all edge computing nodes with the size of the resources needed by executing vehicle tasks, when the free resources of the edge computing nodes are smaller than the computing resources needed by the computing tasks, namely trm＞ernTime, vehicle cmIs calculated task tmThe edge computation node cannot be migrated to.
S32 calculation vehicle cmCalculation node e to edgenThe distance of (c):
Figure BDA0001944361660000061
s33, the edge calculation node has an action range rho called effective road section, when dis (c)m,en) < rho, then cmAt enValid road section of cmCan calculate it to task tmDirect migration to enIs performed, but in practice, because of cmWhile driving, if the calculation task is migrated to enWhen e isnFinish t executionmReturning the calculation result to cmWhen c is greater thanmMay have been out of enIn order to avoid this, vehicle cmIs calculated task tmCannot migrate to the edge calculation node, cmComputing task t that can only be generatedmTo move in the direction of travel of the vehicle and at enThe subsequent edge calculation nodes are executed.
S4, matching the vehicle calculation tasks with the edge nodes meeting the conditions one by one, and calculating the time and energy consumption required by each migration strategy; the specific operation is as follows:
s41, selecting a qualified target edge calculation node and a target vehicle in the action range of the target node; the migration process of the computing task comprises two parts: one is cmT is transmitted in a vehicle-to-vehicle transmission modemA process of transmitting to a target vehicle; and secondly, the target vehicle migrates the calculation task to the target side calculation node.
S42, calculating task tmIncluding the vehicle cmComputing task tmTo the target vehicle cnTime Tt ofmTarget vehicle cnHandle tmMigration to target edge computation node enTime To ofm、enTime Te of executing a computing taskmAnd enFeeding back the calculation result to cmTime Tf ofm。
Calculating Ttm、Tom、TemAnd TfmThe formula of (1) is as follows:
Figure BDA0001944361660000071
Figure BDA0001944361660000072
Figure BDA0001944361660000073
Figure BDA0001944361660000074
where v denotes the transmission rate between the vehicles, λm,nRepresenting a computing task from cmIs transmitted to cnThe number of vehicles passing by, v' represents the transfer rate between the vehicle and the edge calculation node, and p represents the calculation capability of each virtual machine.
S43, calculating the total time delay T required by the migration strategy:
Figure BDA0001944361660000081
s44, edge calculation node enThe energy consumption mainly comprises basic energy consumption EbnIdle energy consumption EinAnd occupy energy consumption Eun. Energy consumption of edge computing node and operation time Ts of server thereofnRelated, TsnThe calculation expression is as follows:
Figure BDA0001944361660000082
wherein B ism,nRepresenting a computational task tmWhether or not in enUpper execution, Bm,nThe expression of (a) is:
Figure BDA0001944361660000083
calculating Ebn、EinAnd EunThe formula of (1) is as follows:
Ebn＝Tsn·Pα (26)
Figure BDA0001944361660000084
Figure BDA0001944361660000085
wherein, PαRepresenting edge calculation node enPower of the server, PβDenotes enPower of the resource unoccupied, PγDenotes enPower of the occupied resource.
S45, calculating edge calculating node enTotal energy consumption E:
Figure BDA0001944361660000086
s5, obtaining an optimal calculation migration strategy through a simple weighting method and a multi-standard decision algorithm; the specific operation is as follows:
s51, for the computation migration technology, the lower the time delay and the better the energy consumption. According to the simple weighting method and the multi-standard decision algorithm, the total time delay and the total energy consumption of each computational migration strategy are negative standards, which can be respectively normalized as follows:
Figure BDA0001944361660000087
Figure BDA0001944361660000091
wherein, TmaxAnd TminRespectively representing the maximum and minimum delays resulting from the computation of the migration, EmaxAnd EminRespectively representing the maximum and minimum energy consumption resulting from computing migration.
S52, calculating utility values of all migration strategies, and obtaining a calculated migration strategy with the maximum utility value:
UV＝V(T)·ωT+V(E)·ωE(ωT+ωE＝1) (32)
wherein, ω isTDenotes the weight, ω, of V (T)EThe weights of V (E) and (E) are shown respectively.
The method of the present invention is further explained by referring to a specific embodiment, as shown in fig. 2, a section of a road in the internet of vehicles is used as a research area, 4 edge calculation nodes are set in the area, and E ═ E { (E) } is set1,e2,e3,e4Effective areas of the calculation nodes of the respective sides are shown by dotted lines in fig. 1, a total of 11 vehicles travel in the areas, and C ═ C1,c2,c3,...,c11}. The relevant information of the edge calculation node is shown in table 1:
TABLE 1
Edge calculation node e1/e2/e3/e4
Number of virtual machines (total capacity eq)n) 10
Computing power per virtual machine (p) 2000MHz
In the present embodiment, the vehicle c2，c5And c7Respectively generating computing tasks t2，t5And t7The information of each calculation task is shown in table 2:
TABLE 2
Computing tasks t2 t5 t7
Number of required virtual machines 3 4 3
Task volume (Kb) 300 500 400
Feedback task volume (Kb) 400 300 500
According to the criterion of step S3 of the method according to the invention, task t2Can migrate to e2、e3And e4Executing; t is t5Can migrate to e3And e4Executing; t is t7Can only migrate to e4Executing; in addition, as can be seen from tables 1 and 2, the edge calculation node e2、e3And e4All have 10 idle virtual machine resources, task t2、t5And t7The required virtual machine resources are 3, 4 and 3 respectively, so the migration path in 6 above is feasible.
When the calculation task is transmitted between vehicles, in order to ensure the transmission efficiency, there are strict requirements on the relative positions of the vehicles, and in this embodiment, only the transmission of the calculation task between adjacent and closer vehicles is considered, for example, the task t2Migration to e3Execution, c2Can not directly handle t2Is transmitted to c7Or c8But must pass through path c2→c4→c6→c8Realization of t2By c through8Handle t2Migrate to e3And (6) executing.
In step S4, the target vehicle shifts the calculation task To the time To of the target-side calculation nodemThe time Te of the edge node executing the calculation taskmAnd the time Tf when the edge node feeds back the calculation resultmIs certain and does not change with the difference of the edge calculation nodes for executing the task, but the transmission time of the calculation task between the vehicles is related to the number of vehicles on the transmission path. The values of the parameters involved in step S4 are shown in table 3:
TABLE 3
Parameter(s) Value of
Transmission rate v between vehicles 1Gbps
Vehicle and edge calculation knotTransmission rate v 'between points' 600Mbps
Computing node server power Pα 300W
Power P of unoccupied virtual machinesβ 30W
Power P of occupied virtual machineγ 50W
By a migration path c2→c4→c6→c8→e3For example, the transfer time Tt of a task between vehicles is calculated2：
Figure BDA0001944361660000101
Time delays and energy consumptions corresponding to the 6 migration paths are sequentially calculated, and the calculation results are shown in table 4:
TABLE 4
Figure BDA0001944361660000102
Figure BDA0001944361660000111
6 calculated migration policies can be obtained according to the migration paths in 6 in table 4, and the utility value of each migration policy is calculated according to the formula (32), as shown in table 5:
TABLE 5
Figure BDA0001944361660000112
As can be seen from the data in table 5, the utility value of the calculated migration policy 2 is the highest, and the optimal calculated migration policy in this embodiment is: will calculate task t2Migration to e2Executing, computing task t5And t7Migration to e4And (6) executing.
The embodiments of the present invention have been described in detail with reference to the drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the gist of the present invention.

Claims (4)

1. An efficient edge computing migration method for car networking applications, comprising the steps of:
s1, acquiring current position information of the vehicle in the Internet of vehicles and application requirements of a migration task generated by the vehicle;
s2, acquiring the position information of all edge calculation nodes in the Internet of vehicles and the calculation resource condition of each edge node;
s3, filtering out the edge calculation nodes which do not meet the conditions according to the distance between the vehicle and the edge calculation nodes and the application requirements of the migration tasks; the specific operation is as follows:
s31, comparing the idle resources er of all edge calculation nodesnAnd resources tr needed for executing the vehicle taskmWhen tr ism＞ernVehicle cmIs calculated task tmCannot migrate to the edge compute node;
s32 calculation vehicle cmCalculation node e to edgenThe distance of (c):
Figure FDA0003206140590000011
wherein, the vehicle cmThe coordinate at the present time is cpm,i＝(cpxm,i,cpym,i) Edge ofCalculation node enHas the coordinate of en＝(epxn,epyn)；
S33, current dis (c)m,en) < rho, rho being enOf a vehicle cmIs calculated task tmThe side calculation node cannot be migrated to, and the side calculation node cannot be migrated to an edge node on the side opposite to the vehicle traveling direction;
s4, matching the vehicle calculation tasks with the edge nodes meeting the conditions one by one, and calculating the time and energy consumption required by each migration strategy; the specific operation is as follows:
s41, selecting a qualified target edge calculation node and a target vehicle in the action range of the target node;
s42 calculation vehicle cmComputing task tmTo the target vehicle cnTime Tt ofmTarget vehicle cnHandle tmMigration to target edge computation node enTime To ofm、enTime Te of executing a computing taskmAnd enFeeding back the calculation result to cmTime Tf ofmThe concrete formula is as follows:
Figure FDA0003206140590000012
Figure FDA0003206140590000013
Figure FDA0003206140590000014
Figure FDA0003206140590000015
wherein, twmRepresents tmV represents the vehicle's mission quantityInter transmission rate, λm,nRepresenting a computing task from cmIs transmitted to cnThe number of passing vehicles, v' represents the transmission rate between the vehicle and the edge calculation node, trmIndicates the execution of tmRequired resources, p represents the computing power of each virtual machine, tw'mIndicating the amount of data returned by the end of execution;
s43, calculating the total time delay T required by the migration strategy:
Figure FDA0003206140590000021
the Internet of vehicles contains M vehicles, wherein M is 1,2, … and M;
s44, calculating edge calculating node enBase energy consumption EbnIdle energy consumption EinAnd occupy energy consumption Eun：
Ebn＝Tsn·Pα
Figure FDA0003206140590000022
Figure FDA0003206140590000023
Wherein, TsnDenotes enRun time of the server of (1), PαRepresenting edge calculation node enPower of the server eqnDenotes enTotal capacity of (B)m,nRepresenting a computational task tmWhether or not in enUpper run, PβDenotes enPower of the resource unoccupied, PγDenotes enPower of the occupied resource;
s45, calculating edge calculating node enTotal energy consumption E:
Figure FDA0003206140590000024
the internet of vehicles has N edge calculation nodes, where N is 1,2, …, and N is
S5, obtaining an optimal calculation migration strategy through a simple weighting method and a multi-standard decision algorithm; the specific operation is as follows:
s51, respectively normalizing the total time delay and the total energy consumption of each calculation migration strategy into the following values by a simple weighting method and a multi-standard decision algorithm:
Figure FDA0003206140590000031
Figure FDA0003206140590000032
wherein, TmaxAnd TminRespectively representing the maximum and minimum delays resulting from the computation of the migration, EmaxAnd EminRespectively representing the maximum energy consumption and the minimum energy consumption generated by the calculation migration;
s52, calculating utility values of all migration strategies, and obtaining a calculated migration strategy with the maximum utility value:
UV＝V(T)·ωT+V(E)·ωE(ωT+ωE＝1)
wherein, ω isTDenotes the weight, ω, of V (T)EThe weights of V (E) and (E) are shown respectively.
2. The method for efficient edge computing migration in car networking applications according to claim 1, wherein the specific operations of step S1 are as follows:
s11, M vehicles and M-th vehicle c are contained in the Internet of vehicles area AmCoordinate cp at time im,iThe following were used:
cpm,i＝(cpxm,i,cpym,i)
wherein, cpxm,iVehicle c at time imTransverse in the area ACoordinate position, cpym,iVehicle c at time imAt the ordinate position in region a, M is 1,2, …, M;
s12, m-th vehicle cmThe resulting computing task is tm＝(twm,trm,tw'm) Wherein, twmRepresenting a computational task tmTask amount of, trmIndicates the execution of tmRequired resource, tw'mIndicating the amount of data returned by the end of execution.
3. The method for efficient edge computing migration in internet of vehicles applications according to claim 1, wherein in step S2, N edge computing nodes are set in the area a of the internet of vehicles, and the expressions of the edge computing nodes are as follows:
en＝(epxn,epyn,eqn,ern)
wherein e isnRepresents the N-th edge calculation node, N is 1,2, …, N, epxnDenotes enPosition on the abscissa in area A, epynDenotes enOrdinate position in area A, eqnDenotes enTotal capacity of (1), ernDenotes enThe idle resources of (1).
4. The method of any of claims 2 or 3, wherein tr is a measure of the migration of the edge calculations in the Internet of vehiclesm、eqnAnd ernAll in the form of the number of virtual machines.
CN201910031387.XA 2019-01-14 2019-01-14 Efficient edge computing migration method for Internet of vehicles application Active CN109905859B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910031387.XA CN109905859B (en) 2019-01-14 2019-01-14 Efficient edge computing migration method for Internet of vehicles application

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910031387.XA CN109905859B (en) 2019-01-14 2019-01-14 Efficient edge computing migration method for Internet of vehicles application

Publications (2)

Publication Number Publication Date
CN109905859A CN109905859A (en) 2019-06-18
CN109905859B true CN109905859B (en) 2021-09-28

Family

ID=66943781

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910031387.XA Active CN109905859B (en) 2019-01-14 2019-01-14 Efficient edge computing migration method for Internet of vehicles application

Country Status (1)

Country Link
CN (1) CN109905859B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111338760B (en) * 2020-02-27 2023-04-25 长沙市源本信息科技有限公司 Service instance cross-node telescoping method and device for edge computing
CN111614745B (en) * 2020-05-18 2022-04-26 安徽中科美络信息技术有限公司 Vehicle task data scheduling method based on edge calculation
US11803413B2 (en) 2020-12-03 2023-10-31 International Business Machines Corporation Migrating complex legacy applications
CN115002155B (en) * 2022-05-25 2023-02-21 安徽江淮汽车集团股份有限公司 Time delay optimization-based road side unit service migration method
CN115100898B (en) * 2022-05-31 2023-09-12 东南大学 A collaborative computing task offloading method for urban intelligent parking management systems

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015171800A1 (en) * 2014-05-09 2015-11-12 Amazon Technologies, Inc. Migration of applications between an enterprise-based network and a multi-tenant network
EP2976860A1 (en) * 2013-03-20 2016-01-27 Wolting Holding B.V. Compiler for and method for software defined networks
WO2016188438A1 (en) * 2015-05-27 2016-12-01 International Business Machines Corporation Updating networks with migration information for a virtual machine
CN106844051A (en) * 2017-01-19 2017-06-13 河海大学 The loading commissions migration algorithm of optimised power consumption in a kind of edge calculations environment
CN107317700A (en) * 2017-06-09 2017-11-03 湖北理工学院 Vehicle-mounted edge calculations node selects system and method
CN108632813A (en) * 2018-05-21 2018-10-09 北京邮电大学 The motion management method and system of mobile edge calculations
CN108810849A (en) * 2018-06-19 2018-11-13 广东工业大学 A kind of moving method, device and the unmanned plane of unmanned plane calculating task
CN108845885A (en) * 2018-07-04 2018-11-20 济南浪潮高新科技投资发展有限公司 A kind of edge calculations method for managing resource towards automatic Pilot
CN109005538A (en) * 2018-07-27 2018-12-14 安徽大学 A message authentication method between unmanned vehicles and multi-mobile edge computing servers

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050273668A1 (en) * 2004-05-20 2005-12-08 Richard Manning Dynamic and distributed managed edge computing (MEC) framework

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2976860A1 (en) * 2013-03-20 2016-01-27 Wolting Holding B.V. Compiler for and method for software defined networks
WO2015171800A1 (en) * 2014-05-09 2015-11-12 Amazon Technologies, Inc. Migration of applications between an enterprise-based network and a multi-tenant network
WO2016188438A1 (en) * 2015-05-27 2016-12-01 International Business Machines Corporation Updating networks with migration information for a virtual machine
CN106844051A (en) * 2017-01-19 2017-06-13 河海大学 The loading commissions migration algorithm of optimised power consumption in a kind of edge calculations environment
CN107317700A (en) * 2017-06-09 2017-11-03 湖北理工学院 Vehicle-mounted edge calculations node selects system and method
CN108632813A (en) * 2018-05-21 2018-10-09 北京邮电大学 The motion management method and system of mobile edge calculations
CN108810849A (en) * 2018-06-19 2018-11-13 广东工业大学 A kind of moving method, device and the unmanned plane of unmanned plane calculating task
CN108845885A (en) * 2018-07-04 2018-11-20 济南浪潮高新科技投资发展有限公司 A kind of edge calculations method for managing resource towards automatic Pilot
CN109005538A (en) * 2018-07-27 2018-12-14 安徽大学 A message authentication method between unmanned vehicles and multi-mobile edge computing servers

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
A service migration strategy based on multiple attribute decision in mobile edge computing;Dan Zhao 等;《 2017 IEEE 17th International Conference on Communication Technology (ICCT)》;20180517;全文 *
移动边缘环境下面向工作流管理的计算迁移方法;伏舒存 等;《计算机应用》;20181127;全文 *
移动边缘环境下面向工作流管理的计算迁移方法;郭俊;《中国优秀硕士学位论文全文数据库 信息科技辑》;20181115;全文 *

Also Published As

Publication number Publication date
CN109905859A (en) 2019-06-18

Similar Documents

Publication Publication Date Title
CN109905859B (en) Efficient edge computing migration method for Internet of vehicles application
CN111711666B (en) Internet of vehicles cloud computing resource optimization method based on reinforcement learning
CN110035410A (en) Federated resource distribution and the method and system of unloading are calculated in a kind of vehicle-mounted edge network of software definition
Zhang et al. An approach of multi-objective computing task offloading scheduling based NSGS for IOV in 5G
CN113344255B (en) Vehicle-mounted network application data transmission and charging optimization method based on mobile edge calculation and block chain
CN111831427B (en) A distributed inter-vehicle task offloading method based on mobile edge computing
CN107948083B (en) SDN data center congestion control method based on reinforcement learning
CN111352731A (en) Method, system, apparatus and medium for distributing tasks in edge computing network
CN109803292B (en) Multi-level user moving edge calculation method based on reinforcement learning
CN111240821B (en) A collaborative cloud computing migration method based on security classification of Internet of Vehicles applications
CN110502321A (en) A kind of resource regulating method and system
Liu et al. QoS-aware task offloading and resource allocation optimization in vehicular edge computing networks via MADDPG
Wu et al. Delay-aware edge-terminal collaboration in green internet of vehicles: A multiagent soft actor-critic approach
CN112084035A (en) A task scheduling method and system based on ant colony algorithm
CN104811466B (en) The method and device of cloud media resource allocation
Sutagundar et al. Resource allocation for fog enhanced vehicular services
Gong et al. Slicing-based resource optimization in multi-access edge network using ensemble learning aided DDPG algorithm
Ma et al. Advance-FL: A3C-based adaptive asynchronous online federated learning for vehicular edge cloud computing networks
Cao et al. A reservation-based multi-source distributed offloading strategy in dynamic environments
He et al. A DDPG hybrid of graph attention network and action branching for multi-scale end-edge-cloud vehicular orchestrated task offloading
Sun et al. Energy Optimization of Multi-Task DNN Inference in MEC-Assisted XR Devices: A Lyapunov-Guided Reinforcement Learning Approach
Zhang et al. Latency minimization for MEC-V2X assisted autonomous vehicles task offloading
Gowri et al. Adaptive probabilistic neural network based edge data center authentication for secure load balancing in fog computing
Chen et al. A hybrid deep reinforcement learning approach for jointly optimizing offloading and resource management in vehicular networks
Huo et al. Collaborative offloading strategy for dependent tasks in mobile edge computing

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
TR01 Transfer of patent right
TR01 Transfer of patent right

Effective date of registration: 20221221

Address after: 201203 zone a, floor 3, building 5, No. 888, Bibo Road, pilot Free Trade Zone, Pudong New Area, Shanghai

Patentee after: SHANGHAI CHANGXING SOFTWARE Co.,Ltd.

Patentee after: Shanghai Changxing Information Technology Co.,Ltd.

Address before: No.219, ningliu Road, Jiangbei new district, Nanjing, Jiangsu Province, 210032

Patentee before: Nanjing University of Information Science and Technology