CN113742077A - Calculation migration method based on 5G Internet of vehicles - Google Patents

Calculation migration method based on 5G Internet of vehicles Download PDF

Info

Publication number
CN113742077A
CN113742077A CN202111048558.3A CN202111048558A CN113742077A CN 113742077 A CN113742077 A CN 113742077A CN 202111048558 A CN202111048558 A CN 202111048558A CN 113742077 A CN113742077 A CN 113742077A
Authority
CN
China
Prior art keywords
vehicle
calculation
individuals
population
server
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
CN202111048558.3A
Other languages
Chinese (zh)
Other versions
CN113742077B (en
Inventor
周原
任彩琴
刘明山
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.)
Jilin University
Original Assignee
Jilin University
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 Jilin University filed Critical Jilin University
Priority to CN202111048558.3A priority Critical patent/CN113742077B/en
Publication of CN113742077A publication Critical patent/CN113742077A/en
Application granted granted Critical
Publication of CN113742077B publication Critical patent/CN113742077B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5061—Partitioning or combining of resources
    • G06F9/5072—Grid computing
    • 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
    • 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)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a calculation migration method based on a 5G Internet of vehicles, which comprises the following steps: the first step is to calculate the establishment of a migration scene: selecting a 5G-based vehicle networking scene as a calculation migration scene, wherein the scene mainly comprises an EC server, an RSU, a vehicle and a cloud server, and the EC server, the RSU, the vehicle and the cloud server are communicated by utilizing a 5G technology; secondly, obtaining a computational migration mathematical model according to a computational migration scene; thirdly, obtaining an optimal solution of the mathematical model based on an improved chaos-differential evolution algorithm; and fourthly, taking the calculation migration method corresponding to the optimal solution of the mathematical model as a final calculation migration method. Has the advantages that: the resource limit of the vehicle can be broken through, the calculated amount of the vehicle is reduced, the loss of the battery power of the vehicle is reduced, the vehicle storage resource is saved, and the user experience degree can be improved. By the improved chaos-differential evolution algorithm, the optimal time delay and energy consumption can be obtained, and the requirements of time delay and energy consumption in the environment of the Internet of vehicles can be met.

Description

Calculation migration method based on 5G Internet of vehicles
Technical Field
The invention relates to a calculation and migration method, in particular to a calculation and migration method based on a 5G internet of vehicles.
Background
Currently, as the car networking and automatic driving applications are getting more and more attentions, more Communication Technology (CT) enterprises successively view the car networking as a key application direction for Edge Computing, and furthermore, Information Technology (IT) enterprises are beginning to advance to the field of Edge Computing (EC) based car networking.
With the advent of the age of 5G, enormous amounts of data and connections are being generated. In the scenario of the internet of vehicles, due to the fact that computing power and storage resources of the vehicles are limited, some computing tasks with large computing amount and high safety cannot be independently realized on the vehicle-mounted unit, and therefore the computing tasks which cannot be processed by the vehicles need to be considered to be migrated to other devices with idle resources.
Disclosure of Invention
The invention aims to solve the problem that vehicle-mounted terminal resources are limited and cannot process tasks with large data volume in real time under the edge computing environment, and provides a computing migration method based on a 5G vehicle networking.
The invention provides a calculation migration method based on a 5G Internet of vehicles, which comprises the following steps:
the first step is to calculate the establishment of a migration scene: selecting a 5G-based vehicle networking scene as a calculation migration scene, wherein the scene mainly comprises an EC server, an RSU, a vehicle and a cloud server, and the EC server, the RSU, the vehicle and the cloud server are communicated by utilizing a 5G technology;
secondly, obtaining a computational migration mathematical model according to a computational migration scene;
thirdly, obtaining an optimal solution of the mathematical model based on an improved chaos-differential evolution algorithm;
and fourthly, taking the calculation migration method corresponding to the optimal solution of the mathematical model as a final calculation migration method.
In the second step, the specific steps of obtaining the computational migration mathematical model according to the computational migration scene are as follows:
step one, according to a calculation migration scene, the problem of reasonable distribution of calculation tasks generated by N EC servers and a single vehicle to the vehicles is researched, and it is assumed that the single vehicle has one calculation task K ═ { K ═ K { (K) }1,k2,...ki,...,kMNeeds to be processed, with a subtask of kiThe data request information is ki＝(ai,bi,ci) The vehicle calculation tasks K are all selected to be processed on the vehicle-mounted terminal or the EC server, and P calculation tasks KiProcessed at the vehicle-mounted terminal, M-P calculation tasks kiMigrate toProcessing on EC Server, each subtask kiOne of the EC servers, local or N, must be selected for processing, as follows:
Figure BDA0003251858320000021
wherein, aiIndicating completion of task kiThe number of cycles required, biIndicating the size of the uploaded data, ciIndicating the size of the backtransmission data, M, P represents the number of subtasks, and M > P, xiRepresenting computation migration decisions, determining a computation task kiWhether processing is done locally or on the EC server, yi,nIndicating whether or not to compute task kiAllocating to an EC server n;
xi＝{0,1}，xiwhen 1, the calculation task k is expressediProcessing locally, xiWhen 0, it means that the calculation task is processed on the EC server, yi,n＝{0,1}，yi,nWhen 1, the calculation task k is expressediAssigned to EC servers n, yi,nWhen 0, the calculation task k is representediNot assigned to EC server n;
step two, the vehicle computing task processes the generated time delay and energy consumption in the vehicle-mounted terminal;
if part or all of the subtasks in the vehicle calculation task are processed locally, the calculation delay is expressed as follows:
Figure BDA0003251858320000022
wherein, TVRepresenting the time delay, f, resulting from the processing of the vehicle calculation task at the vehicle-mounted terminalVRepresents the calculation capability of the vehicle V, i.e., the number of CPU cycles per unit time;
if part or all of the subtasks in the vehicle calculation task are processed in the vehicle-mounted terminal, the calculation energy consumption is expressed as follows:
Figure BDA0003251858320000031
wherein E isVProcessing the generated calculation energy consumption, P, for the vehicle calculation task at the vehicle terminalVThe energy consumption power in a unit CPU period of the vehicle is adopted;
step three, the vehicle calculation task processes the generated time delay and energy consumption on the EC server:
if a part or all of the subtasks in the vehicle calculation task need to be migrated to the EC server for processing, the generated time delay includes an uploading time delay, a processing time delay and a returning time delay, and is expressed as follows:
Figure BDA0003251858320000032
wherein, TCThe sum of the time delay generated by uploading the vehicle calculation task to the EC server, the time delay of the EC server for processing the calculation task and the return time delay of the return result to the vehicle is represented as fCDenoted the computing power of the EC server, trVRRepresenting the transmission rate, tr, between the vehicle and the RSURERepresenting the transmission rate, tr, between the RSU and the EC serverEVRepresenting a transmission rate between the EC server and the vehicle;
if part or all of the subtasks in the vehicle calculation task are migrated to the EC server for processing, the wireless transmission energy consumption of the vehicle is expressed as follows:
Figure BDA0003251858320000033
wherein ECRefers to the wireless transmission energy consumption, P, of the vehicleuRefers to the transmission power of the vehicle on the wireless transmission channel, P represents the transmission power between the RSU and the EC, PrThe automobile receives power;
step four, in the calculation and migration scene, the size of two parameters of time delay and energy consumption determines the quality of the selected calculation and migration method, and the vehicle and the EC server are used for parallel processing vehicle calculationTask, so will TVAnd TCThe maximum value of (a) is used as the final time delay, so the final time delay T and the energy consumption E are expressed as follows:
Figure BDA0003251858320000041
E＝EV+EC (7)
step five, the optimal total target of the calculation migration method is expressed as follows:
f＝{minT,minE} (8)。
in the third step, the optimal solution of the mathematical model is obtained based on the improved chaos-differential evolution algorithm, and the steps are as follows:
step one, determining the population size NP to be 50, the scaling factor F to be 0.6 and the maximum iteration number TmaxThe crossover probability factor CR is 0.5;
step two, enabling the iteration number to be 0, and chaotic initialization of population generation individuals
Figure BDA0003251858320000042
Of note are those in which
Figure BDA0003251858320000043
Is a 1 x M-dimensional list of,
Figure BDA0003251858320000044
each element in (1) corresponds to x occurring in a specific step in the second stepiThe value of (a) is selected,
Figure BDA0003251858320000045
is a list of dimensions M x N,
Figure BDA0003251858320000046
each element in (a) corresponds to y appearing in a specific step in the second stepi,nTaking the value of (A);
Figure BDA0003251858320000047
Figure BDA0003251858320000048
Figure BDA0003251858320000049
Figure BDA00032518583200000410
Figure BDA00032518583200000411
wherein the content of the first and second substances,
Figure BDA00032518583200000412
represents the ith chromosome of the 0 th generation in the population,
Figure BDA00032518583200000413
j gene of i chromosome representing 0 generation, M is total number of vehicle calculation subtasks, N is total number of EC servers, rkIs a random number between 0 and 1, mu1Is a control parameter, the value is 3.6;
for the generated population individuals
Figure BDA0003251858320000051
And (3) performing upward rounding to adapt to the solution of the formula (8) in the concrete step of the second step, thus obtaining:
Figure BDA0003251858320000052
Figure BDA0003251858320000053
step three, generating variant individuals by carrying out variant operation
Figure BDA0003251858320000054
Wherein
Figure BDA0003251858320000055
Is a 1 x M-dimensional list of,
Figure BDA0003251858320000056
each element in (1) corresponds to x appearing in the second step of the concrete processiThe value of (a) is selected,
Figure BDA0003251858320000057
is a list of dimensions M x N,
Figure BDA0003251858320000058
each element in the second step corresponds to y appearing in the second detailed stepi,nTaking the value of (A);
realizing individual variation through a differential strategy, namely firstly, taking the variant individual as a differential vector of a parent, wherein each vector is formed by the parent, namely a Tth generation population
Figure BDA0003251858320000059
Two different individuals are subjected to vector synthesis with the individual to be mutated after the vectors of the two different individuals are scaled, namely:
Figure BDA00032518583200000510
Figure BDA00032518583200000511
wherein the content of the first and second substances,
Figure BDA00032518583200000512
is a variant individual, and is a human,
Figure BDA00032518583200000513
are all provided withA parent individual, wherein i is not equal to r1, not equal to r2, not equal to r3, F represents the influence degree of the differential vector on the next generation individual, and the value is 0.6;
for the generated variant individuals
Figure BDA00032518583200000514
Rounding off and rounding up to adapt to the solution of the formula (8) to obtain:
Figure BDA00032518583200000515
Figure BDA00032518583200000516
step four, performing cross operation to generate experimental individuals
Figure BDA00032518583200000517
Population of individuals as defined herein
Figure BDA00032518583200000518
The constraint requirement of the formula (1) in the concrete step of the second step is satisfied;
for the ith individual in the T generation population
Figure BDA0003251858320000061
And variant individuals thereof
Figure BDA0003251858320000062
Performing cross operation among the populations, wherein the cross operation is expressed by the following formula:
Figure BDA0003251858320000063
wherein rand () is a random decimal between 0-1;
step five, carrying out selection operation to generate next generation individual
Figure BDA0003251858320000064
Generating population individuals
Figure BDA0003251858320000065
Between NP and 2 NP;
the selection operation is carried out by making the experimental individuals
Figure BDA0003251858320000066
With parents
Figure BDA0003251858320000067
Competition is carried out, if one party Pareto the other party, the dominant party enters a new population, and if two individuals do not dominate each other, the dominant party will enter a new population
Figure BDA0003251858320000068
And
Figure BDA0003251858320000069
simultaneously adding a new population, and selecting the operation expression as the following formula:
Figure BDA00032518583200000610
Figure BDA00032518583200000611
is a fitness function in the algorithm and is also a formula (8) in the concrete step of the second step;
step six, adopting rapid non-dominant sorting and crowding degree calculation in the intensity pareto algorithm to select the front NP individuals
Figure BDA00032518583200000612
Composing a new population, and recording the number np1 of individuals with the grade of 1 in the new generation population;
step seven, when NP1 ≠ NP, judging the population diversity lambdaT1,λT2Whether or not less than lambdamin，λminIs the threshold value, if so,a chaotic replacement operation is performed and a chaotic standby new population is generated
Figure BDA00032518583200000613
Population diversity lambdaT1，λT2Such as the following equation:
Figure BDA00032518583200000614
Figure BDA0003251858320000071
Figure BDA0003251858320000072
Figure BDA0003251858320000073
the specific chaotic operation process is as follows:
in the evolution process, when the population diversity is low, replacing the individuals in the current population by the individuals in the generated chaotic standby population according to a set probability so as to guide an algorithm to pick out the current local optimum, wherein the replacement probability P of the ith individual is determined by the sequence of the individuals in the population, namely the following formula:
P＝i/NP (26)
wherein i is the ordering of the individuals in the population and NP is the size of the population
Figure BDA0003251858320000074
On the basis of the above-mentioned information, a new chaos-producing stand-by individual is produced
Figure BDA0003251858320000075
The jth component of
Figure BDA0003251858320000076
Wherein
Figure BDA0003251858320000077
Is obtained by the following equations (27), (28), (29), wherein
Figure BDA0003251858320000078
Is a 1 x M-dimensional list of,
Figure BDA0003251858320000079
each element in (1) corresponds to x appearing in the second step of the concrete processiThe value of (a) is selected,
Figure BDA00032518583200000710
is a list of dimensions M x N,
Figure BDA00032518583200000711
each element in the second step corresponds to y appearing in the second detailed stepi,nThe specific formula of the value is as follows:
Figure BDA00032518583200000712
Figure BDA00032518583200000713
Figure BDA00032518583200000714
wherein, mu2Is a control parameter, the value is 3.8;
step eight, if T is less than TmaxIf T is T +1, and return to step three; if T ═ TmaxForming a non-inferior optimal solution of the single-target optimization problem by all individuals with the rank of 1 in the population, and outputting all fitness functions corresponding to all the optimal individuals, namely the value of a formula (8), as a final set G, wherein T refers to the number of iterations;
and step nine, outputting the optimal individual corresponding to the minimum value in the set G obtained in the step eight, and finishing the operation, wherein the calculation migration method corresponding to the selected optimal individual is used as the final calculation migration method.
The invention has the beneficial effects that:
the calculation migration method based on the 5G Internet of vehicles provided by the invention can break through the resource limitation of the vehicles, reduce the calculated amount of the vehicles, reduce the loss of the battery electric quantity of the vehicles, save the storage resources of the vehicles and the like, and can also improve the user experience. The EC server is close to the vehicle side, and compared with the vehicle, the EC server has more sufficient computing and storing resource resources and can assist the vehicle to better process the vehicle computing task. When the vehicle computing task amount is too large, the vehicle-mounted terminal and the EC server can independently process the vehicle computing task in parallel, and the problem that the vehicle computing task amount is large so that the vehicle computing task cannot be processed in time is avoided to a certain extent. According to the invention, through an improved chaos-differential evolution algorithm, optimal time delay and energy consumption can be obtained, and the requirements of time delay and energy consumption in the environment of the Internet of vehicles can be met.
Drawings
FIG. 1 is a schematic diagram of a computation migration scenario according to the present invention.
Fig. 2 is a schematic flow chart of a calculation migration method based on the 5G internet of vehicles according to the present invention.
Fig. 3 is a schematic diagram of the time delay simulation according to the present invention.
Fig. 4 is a schematic diagram of energy consumption simulation according to the present invention.
Detailed Description
Please refer to fig. 1 to 4:
fig. 1 is a schematic diagram of a computation migration scenario, in fig. 1, there are a vehicle, an RSU, and an EC server, which communicate with each other through 5G, and the present invention only considers the communication between a single vehicle and a plurality of EC servers.
The invention provides a calculation migration method based on a 5G Internet of vehicles, which comprises the following steps:
the first step is to calculate the establishment of a migration scene: selecting a 5G-based vehicle networking scene as a calculation migration scene, wherein the scene mainly comprises an EC server, an RSU, a vehicle and a cloud server, and the EC server, the RSU, the vehicle and the cloud server are communicated by utilizing a 5G technology;
secondly, obtaining a computational migration mathematical model according to a computational migration scene;
thirdly, obtaining an optimal solution of the mathematical model based on an improved chaos-differential evolution algorithm;
and fourthly, taking the calculation migration method corresponding to the optimal solution of the mathematical model as a final calculation migration method.
In the second step, the specific steps of obtaining the computational migration mathematical model according to the computational migration scene are as follows:
step one, according to a calculation migration scene, the problem of reasonable distribution of calculation tasks generated by N EC servers and a single vehicle to the vehicles is researched, and it is assumed that the single vehicle has one calculation task K ═ { K ═ K { (K) }1,k2,...ki,...,kMNeeds to be processed, with a subtask of kiThe data request information is ki＝(ai,bi,ci) The vehicle calculation tasks K are all selected to be processed on the vehicle-mounted terminal or the EC server, and P calculation tasks KiProcessed at the vehicle-mounted terminal, M-P calculation tasks kiMigrating to EC server for processing, each subtask kiOne of the EC servers, local or N, must be selected for processing, as follows:
Figure BDA0003251858320000091
wherein, aiIndicating completion of task kiThe number of cycles required, biIndicating the size of the uploaded data, ciIndicating the size of the backtransmission data, M, P represents the number of subtasks, and M > P, xiRepresenting computation migration decisions, determining a computation task kiWhether processing is done locally or on the EC server, yi,nIndicating whether or not to compute task kiAllocating to an EC server n;
xi＝{0,1}，xiwhen 1, the calculation task k is expressediAt the localZ, xiWhen 0, it means that the calculation task is processed on the EC server, yi,n＝{0,1}，yi,nWhen 1, the calculation task k is expressediAssigned to EC servers n, yi,nWhen 0, the calculation task k is representediNot assigned to EC server n;
step two, the vehicle computing task processes the generated time delay and energy consumption in the vehicle-mounted terminal;
if part or all of the subtasks in the vehicle calculation task are processed locally, the calculation delay is expressed as follows:
Figure BDA0003251858320000101
wherein, TVRepresenting the time delay, f, resulting from the processing of the vehicle calculation task at the vehicle-mounted terminalVRepresents the calculation capability of the vehicle V, i.e., the number of CPU cycles per unit time;
if part or all of the subtasks in the vehicle calculation task are processed in the vehicle-mounted terminal, the calculation energy consumption is expressed as follows:
Figure BDA0003251858320000102
wherein E isVProcessing the generated calculation energy consumption, P, for the vehicle calculation task at the vehicle terminalVThe energy consumption power in a unit CPU period of the vehicle is adopted;
step three, the vehicle calculation task processes the generated time delay and energy consumption on the EC server:
if a part or all of the subtasks in the vehicle calculation task need to be migrated to the EC server for processing, the generated time delay includes an uploading time delay, a processing time delay and a returning time delay, and is expressed as follows:
Figure BDA0003251858320000103
wherein, TCThe representation is generated by uploading a vehicle computing task to an EC serverThe sum of the time delay of the EC server processing the calculation task and the return time delay of the return result to the vehicle, fCDenoted the computing power of the EC server, trVRRepresenting the transmission rate, tr, between the vehicle and the RSURERepresenting the transmission rate, tr, between the RSU and the EC serverEVRepresenting a transmission rate between the EC server and the vehicle;
if part or all of the subtasks in the vehicle calculation task are migrated to the EC server for processing, the wireless transmission energy consumption of the vehicle is expressed as follows:
Figure BDA0003251858320000111
wherein ECRefers to the wireless transmission energy consumption, P, of the vehicleuRefers to the transmission power of the vehicle on the wireless transmission channel, P represents the transmission power between the RSU and the EC, PrThe automobile receives power;
step four, in the calculation migration scene, the size of two parameters of time delay and energy consumption determines the quality of the selected calculation migration method, and the vehicle and the EC server process the vehicle calculation task in parallel, so that T is used for calculating the energy consumption of the vehicleVAnd TCThe maximum value of (a) is used as the final time delay, so the final time delay T and the energy consumption E are expressed as follows:
Figure BDA0003251858320000112
E＝EV+EC (7)
step five, the optimal total target of the calculation migration method is expressed as follows:
f＝{minT,minE} (8)。
in the third step, the optimal solution of the mathematical model is obtained based on the improved chaos-differential evolution algorithm, and the steps are as follows:
step one, determining the population size NP to be 50, the scaling factor F to be 0.6 and the maximum iteration number TmaxThe crossover probability factor CR is 0.5;
step two, enabling the iteration number to be 0, and chaotic initialization of population generation individuals
Figure BDA0003251858320000113
Of note are those in which
Figure BDA0003251858320000114
Is a 1 x M-dimensional list of,
Figure BDA0003251858320000115
each element in (1) corresponds to x occurring in a specific step in the second stepiThe value of (a) is selected,
Figure BDA0003251858320000116
is a list of dimensions M x N,
Figure BDA0003251858320000117
each element in (a) corresponds to y appearing in a specific step in the second stepi,nTaking the value of (A);
Figure BDA0003251858320000121
Figure BDA0003251858320000122
Figure BDA0003251858320000123
Figure BDA0003251858320000124
Figure BDA0003251858320000125
wherein the content of the first and second substances,
Figure BDA0003251858320000126
represents the ith chromosome of the 0 th generation in the population,
Figure BDA0003251858320000127
j gene of i chromosome representing 0 generation, M is total number of vehicle calculation subtasks, N is total number of EC servers, rkIs a random number between 0 and 1, mu1Is a control parameter, the value is 3.6;
for the generated population individuals
Figure BDA0003251858320000128
And (3) performing upward rounding to adapt to the solution of the formula (8) in the concrete step of the second step, thus obtaining:
Figure BDA0003251858320000129
Figure BDA00032518583200001210
step three, generating variant individuals by carrying out variant operation
Figure BDA00032518583200001211
Wherein
Figure BDA00032518583200001212
Is a 1 x M-dimensional list of,
Figure BDA00032518583200001213
each element in (1) corresponds to x appearing in the second step of the concrete processiThe value of (a) is selected,
Figure BDA00032518583200001214
is a list of dimensions M x N,
Figure BDA00032518583200001215
each element in the second step corresponds toOccurrence of yi,nTaking the value of (A);
realizing individual variation through a differential strategy, namely firstly, taking the variant individual as a differential vector of a parent, wherein each vector is formed by the parent, namely a Tth generation population
Figure BDA00032518583200001216
Two different individuals are subjected to vector synthesis with the individual to be mutated after the vectors of the two different individuals are scaled, namely:
Figure BDA00032518583200001217
Figure BDA00032518583200001218
wherein the content of the first and second substances,
Figure BDA00032518583200001219
is a variant individual, and is a human,
Figure BDA00032518583200001220
all are parent individuals, wherein i is not equal to r1, not equal to r2, not equal to r3, F represents the influence degree of the differential vector on the next-generation individuals, and the value is 0.6;
for the generated variant individuals
Figure BDA0003251858320000131
Rounding off and rounding up to adapt to the solution of the formula (8) to obtain:
Figure BDA0003251858320000132
Figure BDA0003251858320000133
step four, performing cross operation to generate experimental individuals
Figure BDA0003251858320000134
Population of individuals as defined herein
Figure BDA0003251858320000135
The constraint requirement of the formula (1) in the concrete step of the second step is satisfied;
for the ith individual in the T generation population
Figure BDA0003251858320000136
And variant individuals thereof
Figure BDA0003251858320000137
Performing cross operation among the populations, wherein the cross operation is expressed by the following formula:
Figure BDA0003251858320000138
wherein rand () is a random decimal between 0-1;
step five, carrying out selection operation to generate next generation individual
Figure BDA0003251858320000139
Generating population individuals
Figure BDA00032518583200001310
Between NP and 2 NP;
the selection operation is carried out by making the experimental individuals
Figure BDA00032518583200001311
With parents
Figure BDA00032518583200001312
Competition is carried out, if one party Pareto the other party, the dominant party enters a new population, and if two individuals do not dominate each other, the dominant party will enter a new population
Figure BDA00032518583200001313
And
Figure BDA00032518583200001314
simultaneously adding a new population, and selecting the operation expression as the following formula:
Figure BDA00032518583200001315
Figure BDA00032518583200001316
is a fitness function in the algorithm and is also a formula (8) in the concrete step of the second step;
step six, adopting rapid non-dominant sorting and crowding degree calculation in the intensity pareto algorithm to select the front NP individuals
Figure BDA0003251858320000141
Composing a new population, and recording the number np1 of individuals with the grade of 1 in the new generation population;
step seven, when NP1 ≠ NP, judging the population diversity lambdaT1,λT2Whether or not less than lambdamin，λminIf the value is the threshold value, performing chaotic replacement operation and generating a chaotic standby new population
Figure BDA0003251858320000142
Population diversity lambdaT1，λT2Such as the following equation:
Figure BDA0003251858320000143
Figure BDA0003251858320000144
Figure BDA0003251858320000145
Figure BDA0003251858320000146
the specific chaotic operation process is as follows:
in the evolution process, when the population diversity is low, replacing the individuals in the current population by the individuals in the generated chaotic standby population according to a set probability so as to guide an algorithm to pick out the current local optimum, wherein the replacement probability P of the ith individual is determined by the sequence of the individuals in the population, namely the following formula:
P＝i/NP (26)
wherein i is the ordering of the individuals in the population and NP is the size of the population
Figure BDA0003251858320000147
On the basis of the above-mentioned information, a new chaos-producing stand-by individual is produced
Figure BDA0003251858320000148
The jth component of
Figure BDA0003251858320000149
Wherein
Figure BDA0003251858320000151
Is obtained by the following equations (27), (28), (29), wherein
Figure BDA0003251858320000152
Is a 1 x M-dimensional list of,
Figure BDA0003251858320000153
each element in (1) corresponds to x appearing in the second step of the concrete processiThe value of (a) is selected,
Figure BDA0003251858320000154
is a list of dimensions M x N,
Figure BDA0003251858320000155
each element in the second step corresponds to a second stepY occurring in step (ii)i,nThe specific formula of the value is as follows:
Figure BDA0003251858320000156
Figure BDA0003251858320000157
Figure BDA0003251858320000158
wherein, mu2Is a control parameter, the value is 3.8;
step eight, if T is less than TmaxIf T is T +1, and return to step three; if T ═ TmaxForming a non-inferior optimal solution of the single-target optimization problem by all individuals with the rank of 1 in the population, and outputting all fitness functions corresponding to all the optimal individuals, namely the value of a formula (8), as a final set G, wherein T refers to the number of iterations;
and step nine, outputting the optimal individual corresponding to the minimum value in the set G obtained in the step eight, and finishing the operation, wherein the calculation migration method corresponding to the selected optimal individual is used as the final calculation migration method.
In the above embodiment, assuming that the vehicles are present randomly and independently and the speed of the vehicles is fixed, the vehicles on the straight road are subject to the poisson distribution. The vehicle and the EC server, and the EC server communicate with each other by using 5G. The vehicle selects one EC server among N EC servers to process per computation task i, each EC server being abundant in computation storage resources.
The simulation diagram is obtained by carrying out Python language programming on the window7 system by using a spyder platform, and is shown in figures 3 and 4. Fig. 3 is a simulation diagram of the obtained time delay, and fig. 4 is a simulation diagram of the energy consumption.
In FIG. 3, the abscissa is the amount of calculation required for the task, i.e., the second step of the detailed stepsNow aiThe ordinate represents the value of the time delay T appearing in the formula (8). From the figure, a can be seeniWhen the value is between 400Megacycles and 1800Megacycles, the change of the time delay T is between 0 and 0.24 s. It can be seen that the time delay of the computational migration method obtained by using the improved chaos-differential evolution method is controlled within 0.24s, which meets the requirement on time delay in the internet of vehicles.
In FIG. 4, the abscissa is the amount of calculation required for the task, i.e., a appearing in the second detailed stepiThe ordinate represents the value of the energy consumption E occurring in the formula (8). From the figure, a can be seeniWhen the value is between 400Megacycles and 1800Megacycles, the change of the energy consumption E is between 0 and 0.37J. It can be seen that under the condition of meeting the requirement of the internet of vehicles on time delay, the energy consumption is not large, and the calculation migration method obtained by using the improved chaos-differential evolution method is excellent.

Claims (3)

1. A calculation migration method based on a 5G Internet of vehicles is characterized in that: the method comprises the following steps:
the first step is to calculate the establishment of a migration scene: selecting a 5G-based vehicle networking scene as a calculation migration scene, wherein the scene mainly comprises an EC server, an RSU, a vehicle and a cloud server, and the EC server, the RSU, the vehicle and the cloud server are communicated by utilizing a 5G technology;
secondly, obtaining a computational migration mathematical model according to a computational migration scene;
thirdly, obtaining an optimal solution of the mathematical model based on an improved chaos-differential evolution algorithm;
and fourthly, taking the calculation migration method corresponding to the optimal solution of the mathematical model as a final calculation migration method.
2. The calculation migration method based on the 5G Internet of vehicles according to claim 1, wherein: the second step of obtaining a computational migration mathematical model according to the computational migration scenario comprises the following specific steps:
step one, according to a calculation migration scene, the vehicle generation of N EC servers and a single vehicle is researchedThe problem of reasonable distribution of the calculation tasks of (1) is to assume that a single vehicle has one calculation task K ═ K1,k2,...ki,...,kMNeeds to be processed, with a subtask of kiThe data request information is ki＝(ai,bi,ci) The vehicle calculation tasks K are all selected to be processed on the vehicle-mounted terminal or the EC server, and P calculation tasks KiProcessed at the vehicle-mounted terminal, M-P calculation tasks kiMigrating to EC server for processing, each subtask kiOne of the EC servers, local or N, must be selected for processing, as follows:
Figure FDA0003251858310000011
wherein, aiIndicating completion of task kiThe number of cycles required, biIndicating the size of the uploaded data, ciIndicating the size of the backtransmission data, M, P represents the number of subtasks, and M > P, xiRepresenting computation migration decisions, determining a computation task kiWhether processing is done locally or on the EC server, yi,nIndicating whether or not to compute task kiAllocating to an EC server n;
xi＝{0,1}，xiwhen 1, the calculation task k is expressediProcessing locally, xiWhen 0, it means that the calculation task is processed on the EC server, yi,n＝{0,1}，yi,nWhen 1, the calculation task k is expressediAssigned to EC servers n, yi,nWhen 0, the calculation task k is representediNot assigned to EC server n;
step two, the vehicle computing task processes the generated time delay and energy consumption in the vehicle-mounted terminal;
if part or all of the subtasks in the vehicle calculation task are processed locally, the calculation delay is expressed as follows:
Figure FDA0003251858310000021
wherein, TVRepresenting the time delay, f, resulting from the processing of the vehicle calculation task at the vehicle-mounted terminalVRepresents the calculation capability of the vehicle V, i.e., the number of CPU cycles per unit time;
if part or all of the subtasks in the vehicle calculation task are processed in the vehicle-mounted terminal, the calculation energy consumption is expressed as follows:
Figure FDA0003251858310000022
wherein E isVProcessing the generated calculation energy consumption, P, for the vehicle calculation task at the vehicle terminalVThe energy consumption power in a unit CPU period of the vehicle is adopted;
step three, the vehicle calculation task processes the generated time delay and energy consumption on the EC server:
if a part or all of the subtasks in the vehicle calculation task need to be migrated to the EC server for processing, the generated time delay includes an uploading time delay, a processing time delay and a returning time delay, and is expressed as follows:
Figure FDA0003251858310000023
wherein, TCThe sum of the time delay generated by uploading the vehicle calculation task to the EC server, the time delay of the EC server for processing the calculation task and the return time delay of the return result to the vehicle is represented as fCDenoted the computing power of the EC server, trVRRepresenting the transmission rate, tr, between the vehicle and the RSURERepresenting the transmission rate, tr, between the RSU and the EC serverEVRepresenting a transmission rate between the EC server and the vehicle;
if part or all of the subtasks in the vehicle calculation task are migrated to the EC server for processing, the wireless transmission energy consumption of the vehicle is expressed as follows:
Figure FDA0003251858310000031
wherein ECRefers to the wireless transmission energy consumption, P, of the vehicleuRefers to the transmission power of the vehicle on the wireless transmission channel, P represents the transmission power between the RSU and the EC, PrThe automobile receives power;
step four, in the calculation migration scene, the size of two parameters of time delay and energy consumption determines the quality of the selected calculation migration method, and the vehicle and the EC server process the vehicle calculation task in parallel, so that T is used for calculating the energy consumption of the vehicleVAnd TCThe maximum value of (a) is used as the final time delay, so the final time delay T and the energy consumption E are expressed as follows:
Figure FDA0003251858310000032
E＝EV+EC (7)
step five, the optimal total target of the calculation migration method is expressed as follows:
f＝{minT,minE} (8)。
3. the calculation migration method based on the 5G Internet of vehicles according to claim 1, wherein: in the third step, the step of obtaining the optimal solution of the mathematical model based on the improved chaos-differential evolution algorithm is as follows:
step one, determining the population size NP to be 50, the scaling factor F to be 0.6 and the maximum iteration number TmaxThe crossover probability factor CR is 0.5;
step two, enabling the iteration number to be 0, and chaotic initialization of population generation individuals
Figure FDA0003251858310000033
Of note are those in which
Figure FDA0003251858310000034
Is a 1 x M-dimensional list of,
Figure FDA0003251858310000035
each element in (1) corresponds to x occurring in a specific step in the second stepiThe value of (a) is selected,
Figure FDA0003251858310000036
is a list of dimensions M x N,
Figure FDA0003251858310000037
each element in (a) corresponds to y appearing in a specific step in the second stepi,nTaking the value of (A);
Figure FDA0003251858310000041
Figure FDA0003251858310000042
Figure FDA0003251858310000043
Figure FDA0003251858310000044
Figure FDA0003251858310000045
wherein the content of the first and second substances,
Figure FDA0003251858310000046
represents the ith chromosome of the 0 th generation in the population,
Figure FDA0003251858310000047
the j-th gene of the i-th chromosome of the 0 th generation,m is the total number of vehicle calculation subtasks, N is the total number of EC servers, rkIs a random number between 0 and 1, mu1Is a control parameter, the value is 3.6;
for the generated population individuals
Figure FDA0003251858310000048
And (3) performing upward rounding to adapt to the solution of the formula (8) in the concrete step of the second step, thus obtaining:
Figure FDA0003251858310000049
Figure FDA00032518583100000410
step three, generating variant individuals by carrying out variant operation
Figure FDA00032518583100000411
Wherein
Figure FDA00032518583100000412
Is a 1 x M-dimensional list of,
Figure FDA00032518583100000413
each element in (1) corresponds to x appearing in the second step of the concrete processiThe value of (a) is selected,
Figure FDA00032518583100000414
is a list of dimensions M x N,
Figure FDA00032518583100000415
each element in the second step corresponds to y appearing in the second detailed stepi,nTaking the value of (A);
the individual mutation is realized by a differential strategy, firstly, the differential vector of which the mutated individual is a parent is adopted, each vector is composed of the parent,i.e. in the T generation population
Figure FDA00032518583100000416
Two different individuals are subjected to vector synthesis with the individual to be mutated after the vectors of the two different individuals are scaled, namely:
Figure FDA00032518583100000417
Figure FDA00032518583100000418
wherein the content of the first and second substances,
Figure FDA00032518583100000419
is a variant individual, and is a human,
Figure FDA00032518583100000420
all are parent individuals, wherein i is not equal to r1, not equal to r2, not equal to r3, F represents the influence degree of the differential vector on the next-generation individuals, and the value is 0.6;
for the generated variant individuals
Figure FDA0003251858310000051
Rounding off and rounding up to adapt to the solution of the formula (8) to obtain:
Figure FDA0003251858310000052
Figure FDA0003251858310000053
step four, performing cross operation to generate experimental individuals
Figure FDA0003251858310000054
Population of individuals as defined herein
Figure FDA0003251858310000055
The constraint requirement of the formula (1) in the concrete step of the second step is satisfied;
for the ith individual in the T generation population
Figure FDA0003251858310000056
And variant individuals thereof
Figure FDA0003251858310000057
Performing cross operation among the populations, wherein the cross operation is expressed by the following formula:
Figure FDA0003251858310000058
wherein rand () is a random decimal between 0-1;
step five, carrying out selection operation to generate next generation individual
Figure FDA0003251858310000059
Generating population individuals
Figure FDA00032518583100000510
Between NP and 2 NP;
the selection operation is carried out by making the experimental individuals
Figure FDA00032518583100000511
With parents
Figure FDA00032518583100000512
Competition is carried out, if one party Pareto the other party, the dominant party enters a new population, and if two individuals do not dominate each other, the dominant party will enter a new population
Figure FDA00032518583100000513
And
Figure FDA00032518583100000514
simultaneously adding a new population, and selecting the operation expression as the following formula:
Figure FDA00032518583100000515
Figure FDA00032518583100000516
is a fitness function in the algorithm and is also a formula (8) in the concrete step of the second step;
step six, adopting rapid non-dominant sorting and crowding degree calculation in the intensity pareto algorithm to select the front NP individuals
Figure FDA0003251858310000061
Composing a new population, and recording the number np1 of individuals with the grade of 1 in the new generation population;
step seven, when NP1 ≠ NP, judging the population diversity lambdaT1,λT2Whether or not less than lambdamin，λminIf the value is the threshold value, performing chaotic replacement operation and generating a chaotic standby new population
Figure FDA0003251858310000062
Population diversity lambdaT1，λT2Such as the following equation:
Figure FDA0003251858310000063
Figure FDA0003251858310000064
Figure FDA0003251858310000065
Figure FDA0003251858310000066
the specific chaotic operation process is as follows:
in the evolution process, when the population diversity is low, replacing the individuals in the current population by the individuals in the generated chaotic standby population according to a set probability so as to guide an algorithm to pick out the current local optimum, wherein the replacement probability P of the ith individual is determined by the sequence of the individuals in the population, namely the following formula:
P＝i/NP (26)
wherein i is the ordering of the individuals in the population and NP is the size of the population
Figure FDA0003251858310000067
On the basis of the above-mentioned information, a new chaos-producing stand-by individual is produced
Figure FDA0003251858310000068
The jth component of
Figure FDA0003251858310000069
Wherein
Figure FDA0003251858310000071
Figure FDA0003251858310000072
Is obtained by the following equations (27), (28), (29), wherein
Figure FDA0003251858310000073
Is a 1 x M-dimensional list of,
Figure FDA0003251858310000074
each element in (1)Corresponding to x occurring in the second concrete stepiThe value of (a) is selected,
Figure FDA0003251858310000075
is a list of dimensions M x N,
Figure FDA0003251858310000076
each element in the second step corresponds to y appearing in the second detailed stepi,nThe specific formula of the value is as follows:
Figure FDA0003251858310000077
Figure FDA0003251858310000078
Figure FDA0003251858310000079
wherein, mu2Is a control parameter, the value is 3.8;
step eight, if T is less than TmaxIf T is T +1, and return to step three; if T ═ TmaxForming a non-inferior optimal solution of the single-target optimization problem by all individuals with the rank of 1 in the population, and outputting all fitness functions corresponding to all the optimal individuals, namely the value of a formula (8), as a final set G, wherein T refers to the number of iterations;
and step nine, outputting the optimal individual corresponding to the minimum value in the set G obtained in the step eight, and finishing the operation, wherein the calculation migration method corresponding to the selected optimal individual is used as the final calculation migration method.
CN202111048558.3A 2021-09-08 2021-09-08 Computing migration method based on 5G Internet of vehicles Active CN113742077B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111048558.3A CN113742077B (en) 2021-09-08 2021-09-08 Computing migration method based on 5G Internet of vehicles

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111048558.3A CN113742077B (en) 2021-09-08 2021-09-08 Computing migration method based on 5G Internet of vehicles

Publications (2)

Publication Number Publication Date
CN113742077A true CN113742077A (en) 2021-12-03
CN113742077B CN113742077B (en) 2024-03-01

Family

ID=78736925

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111048558.3A Active CN113742077B (en) 2021-09-08 2021-09-08 Computing migration method based on 5G Internet of vehicles

Country Status (1)

Country Link
CN (1) CN113742077B (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109710374A (en) * 2018-12-05 2019-05-03 重庆邮电大学 The VM migration strategy of task unloading expense is minimized under mobile edge calculations environment
CN109840154A (en) * 2019-01-08 2019-06-04 南京邮电大学 A kind of computation migration method that task based access control relies under mobile cloud environment
CN111984419A (en) * 2020-08-28 2020-11-24 华侨大学 Complex task computing and transferring method for marginal environment reliability constraint
CN112995289A (en) * 2021-02-04 2021-06-18 天津理工大学 Internet of vehicles multi-target computing task unloading scheduling method based on non-dominated sorting genetic strategy
CN113139639A (en) * 2021-05-17 2021-07-20 华侨大学 MOMBI-based smart city application-oriented multi-target calculation migration method and device

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109710374A (en) * 2018-12-05 2019-05-03 重庆邮电大学 The VM migration strategy of task unloading expense is minimized under mobile edge calculations environment
CN109840154A (en) * 2019-01-08 2019-06-04 南京邮电大学 A kind of computation migration method that task based access control relies under mobile cloud environment
CN111984419A (en) * 2020-08-28 2020-11-24 华侨大学 Complex task computing and transferring method for marginal environment reliability constraint
CN112995289A (en) * 2021-02-04 2021-06-18 天津理工大学 Internet of vehicles multi-target computing task unloading scheduling method based on non-dominated sorting genetic strategy
CN113139639A (en) * 2021-05-17 2021-07-20 华侨大学 MOMBI-based smart city application-oriented multi-target calculation migration method and device

Also Published As

Publication number Publication date
CN113742077B (en) 2024-03-01

Similar Documents

Publication Publication Date Title
CN111507601B (en) Resource optimization allocation decision method based on deep reinforcement learning and block chain consensus
Chen et al. Efficiency and fairness oriented dynamic task offloading in internet of vehicles
CN112822234B (en) Task unloading method based on deep reinforcement learning in Internet of vehicles
CN111585816A (en) Task unloading decision method based on adaptive genetic algorithm
CN109191052B (en) Multi-vehicle type vehicle path optimization method, server and system
CN112788605B (en) Edge computing resource scheduling method and system based on double-delay depth certainty strategy
CN113781002B (en) Low-cost workflow application migration method based on agent model and multiple group optimization in cloud edge cooperative network
CN107679750A (en) A kind of cloud manufacturing service reso urce matching method based on adaptation coefficient genetic algorithm
CN114585006B (en) Edge computing task unloading and resource allocation method based on deep learning
CN109803292B (en) Multi-level user moving edge calculation method based on reinforcement learning
CN115374853A (en) Asynchronous federal learning method and system based on T-Step polymerization algorithm
CN112083967B (en) Cloud edge computing task unloading method, computer equipment and storage medium
CN116541106B (en) Computing task unloading method, computing device and storage medium
CN112307667A (en) Method and device for estimating state of charge of storage battery, electronic equipment and storage medium
CN114567933A (en) Resource allocation method in heterogeneous cloud and mist cooperative network based on improved genetic algorithm
CN111429014A (en) Electric vehicle charging station selection method based on double-queue dynamic pricing model
CN113139639B (en) MOMBI-oriented smart city application multi-target computing migration method and device
He Optimization of edge delay sensitive task scheduling based on genetic algorithm
CN112231117B (en) Cloud robot service selection method and system based on dynamic vector hybrid genetic algorithm
CN113742077A (en) Calculation migration method based on 5G Internet of vehicles
CN112380006A (en) Data center resource allocation method and device
CN116843500A (en) Charging station planning method, neural network model training method, device and equipment
CN116645130A (en) Automobile order demand prediction method based on combination of federal learning and GRU
CN114444240B (en) Delay and service life optimization method for cyber-physical system
CN113452625B (en) Deep reinforcement learning-based unloading scheduling and resource allocation method

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