CN109947574A - A kind of vehicle big data calculating discharging method based on mist network - Google Patents
A kind of vehicle big data calculating discharging method based on mist network Download PDFInfo
- Publication number
- CN109947574A CN109947574A CN201910246291.5A CN201910246291A CN109947574A CN 109947574 A CN109947574 A CN 109947574A CN 201910246291 A CN201910246291 A CN 201910246291A CN 109947574 A CN109947574 A CN 109947574A
- Authority
- CN
- China
- Prior art keywords
- task
- mist
- layer
- virtual machine
- delay
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- 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
Abstract
The invention discloses a kind of vehicle big datas for calculating network based on mist to calculate discharging method.The present invention can provide more efficient, more reliable calculating environment for analysis vehicle big data.It is proposed that mist calculates network system architecture first, further establish Network Delay Model, then it establishes task and generates model, resettle mist computing resource Optimized model, finally algorithm (Computing Resource-Efficient Task Offloading Algorithm with Load Balancing is unloaded using the effective task of the calculation resources of load balancing proposed by the present invention, CRETOA) Lai Guanli mist calculates the computing resource of Network Load Balance, and road vehicle terminal request processor active task is distributed to optimal mist computing resource.
Description
Technical field
The present invention relates to vehicle networking technical fields more particularly to a kind of vehicle big data based on mist network to calculate unloading side
Method.
Background technique
Cloud computing stores as modern data and one of calculates most important technology, appoints to execute complicated large-scale calculations
Business provide powerful platform, but the big data system architecture based on cloud computing be unable to satisfy it is quick to the delay of ITS application program
Sense requires.And mist calculating provides high flexibility in terms of framework, resource, computing capability, the communication technology and deployment, and has
Support to low latency and high mobility, this becomes the ideal choosing of car networking big data analysis and the unloading of ITS computation
It selects.Mist calculating has high fault tolerance, and mainly there are two reasons: firstly, it is independent of fixed deployment, it can be with the temporarily side of deployment
Formula distributes resource;Secondly, mist, which calculates, can use multi-layer framework, allow to dispose the higher server for calculating specification in higher.
Summary of the invention
It is a primary object of the present invention to solve problems of the prior art, the present invention provides one kind to be based on mist net
The vehicle big data of network calculates discharging method, and specific technical solution is as follows:
A kind of vehicle big data calculating discharging method based on mist network, steps are as follows for specific method:
Step 1: proposing that mist calculates network system architecture;
The mist calculates network system architecture and is divided into three layers: (1) application layer, (2) mist computation layer, including mist calculate node
Equipment, (3) cloud computing layer, including cloud computing equipment are calculated with mist;
Step 2: establishing Network Delay Model;
Specifically, carrying out calculating unloading using cloud and mist cooperation, each task will be off-loaded to cloud or mist;This
Under scene, three data transfer phase are had, including wireless transmission stage and wire transmission stage and calculated result return rank
Section;Postponed the calculating of i.e. response time to the data transfer phase;
Step 3: establishing task and generate model;
Specifically, the length of size of data and task is subjected to distribution appropriate according to the distribution that selected task generates,
Select suitable task length to reduce the quantity of mission failure and the average value of network delay as far as possible;Task arrival time is logical
Cross with exponential distribution and be made of independent same distribution Poisson distribution modeling obtain;
Step 4: establishing mist computing resource Optimized model;
Data are collected, the incoming task set, the virtual machine set in cloud computing layer, mist for obtaining each vehicle termination calculate
The virtual machine set of layer;Then must go out on missions the delay being unloaded on cloud computing layer and the virtual machine of mist computation layer respectively, into one
Step obtains the calculation resources occupancy that vehicle termination executes corresponding task on cloud computing layer and the virtual machine of mist computation layer respectively;It is logical
The formulation of objective function and about beam equation is crossed, balancing delay and calculation resources occupy two elements, obtain optimal calculation resources
Distribution;
Step 5: management mist calculates the computing resource of Network Load Balance, and road vehicle terminal request processor active task is distributed
To optimal mist computing resource;
Specifically, the algorithm that a kind of road vehicle terminal request processor active task distributes to optimal mist computing resource is proposed,
That is the effective task of the calculation resources of load balancing unloads algorithm;It is needed by using expected resource requirement matrix to estimate computing resource
It asks;According to expected resource requirement matrix, virtual machine needs different time and efforts to execute different tasks;Scheduler first
Mist computation layer is offloaded tasks to, if mist computation layer calculation resources occupancy is excessively high, offloads tasks to cloud.
Further, in the step 1, when vehicle is moved to corresponding crossing, when into the overlay area of access point signals,
Vehicle termination, which is added corresponding Wireless LAN and accesses mist, calculates equipment, sends mist calculate node for calculating task;This
Outside, if vehicle termination decision offloads tasks to cloud computing equipment, the WAN connection access provided by Wi-Fi access point is used
Cloud calculates equipment;Mist computation layer is also connect with cloud computing system;In application layer, vehicle termination generates task requests to carry out
It is further processed.
Further, in the step 2, after request task submits to mist computation layer or cloud is handled, service
Delay i.e. response time can be indicated with the sum of the processing of transmission delay and request task delay;dvfAnd dfcIt is from vehicle
Terminal into access point nearest mist computation layer equipment and node and from mist computation layer to the individual data of cloud computing layer be grouped
Transmission delay;
The N run in mist computation layeriThe average transfer delay d of the data packet of a request task application examplefogBy following formula
It provides:
Wherein, PiAnd pi(Pi> pi) it is NiA task is sent to mist computation layer and is sent in cloud data from mist computation layer
The data packet sum of the heart;brThe quantity of the total data packet sent for the response to b request task;
The average transfer delay of mist computation layer processing request task application exampleIt is given by:
In cloud computing layer, it can indicate are as follows:
The delay of request task application example be by the quantity of the request task of server-side processes before treatment come
It calculates, the request task application example sum that can be handled simultaneously are as follows:
Equal part total bandwidth B gives N number of request task application example, and the frequency for occupying each user is not interfere with each other with same
When send their data to mist computation layer and cloud;Therefore:
Respectively indicate vehicle termination Vi∈{V1,V2,...,VwUplink and downlink transmission rate;Here n0It is
Noise power spectral density, hiIt is base station and user NiBetween channel gain, pd,iAnd pu,iIt is vehicle termination V respectivelyiDownlink chain
Road and up-link power;
Set Δ (Vi,Ii) it is to operate in vehicle termination ViIn request task application example IiService delay, You Wuji
It calculates layer and service is provided;In NiIn a request task application example, it is assumed that ni(Ni> ni) a request task application example is redirected
Cloud computing layer is unloaded to carry out operation;The sum for the request task application example that cloud computing layer is handled in time t are as follows:
And it is for each of this n request task application examples, processing latencies beyond the clouds
In ViThe average treatment of the request task application example of interior operation postpones
Then all vehicle termination ViAverage service delayIt can indicate are as follows:
On the contrary, in cloud computing layer, all request task application examples of user side operation are directly and core calculation module
Interaction, the average treatment delay of request task application example hereIt is given by:
Further, in the step 3, the Poisson distribution modeling are as follows:
One essential characteristic of Poisson distribution is that the probability of x is an independent discrete value, its probability and previous institute
There is value unrelated;
The underlying attribute of exponential process is without memory;The waiting time is identified when task reaches;Memoryless characteristic shows
Since task arrival interval is not to the designated time, the distribution of residual waiting time to it is initial similar, be shown below:
Since vehicle termination will not continuously generate service request, mode is generated using free time/active task to simulate
Real scene;According to this mode, user creates task during activity, is waited for during the free time, i.e., each end
There is a state machine at end, and terminal may be at active state or idle state.
Further, in the step 4, specifically, assume in system model there are one group of vehicle termination, quantity m,
With M={ M1,M2,...,MmIndicate;Each vehicle termination has the incoming task of limited quantity, the input of i-th of vehicle termination
Task-set is expressed asWherein, tijIndicate j-th of incoming task of i-th of vehicle termination, vehicle termination Mi
With niA task;SetThe deadline of i-th of vehicle termination is represented, wherein dijIt is cutting for task
The only time;Set VC={ vc1,vc2,...,vcpP virtual machine of the expression in cloud computing layer, and set VFDC=
{vfdc1,vfdc2,...,vfdcqIndicate all mist computation layers q group virtual machine;There are k mist computation layer, each mists
There is q at calculating centeriA virtual machine (VirtualMachine, VM), wherein 1≤i≤k;
X is defined simultaneouslyijk, YijkTwo variables:
Assuming thatWithIt is by by task t respectivelyijIt is unloaded to k-th of the virtual machine and mist computation layer of cloud computing layer
K-th of virtual machine and the delay that generates, task tijExecute delay are as follows:
Then, the total delay for executing all tasks can indicate are as follows:
Vehicle termination MiTask t is executed on k-th of virtual machine of mist computation layerijOccupancy to calculation resources is cijk,
Task t is executed on k-th of virtual machine of cloud computing layerijOccupancy to calculation resources is c'ijk;
Vehicle termination MiCalculation resources occupy CiIncluding two parts: processor active task (1) is unloaded to the void of mist computation layer
The quasi- occupied calculation resources of machine, processor active task is unloaded to the occupied calculation resources of the virtual machine in cloud by (2), we by its
It indicates are as follows:
Therefore, total calculation resources that all vehicle termination M are occupied are as follows:
The calculation resources in mist calculating environment are minimized for realization, have formulated following objective function and constraint equation:
Subject to
Objective function, that is, formula (21) maintains calculation resources to occupy the tradeoff between delay using η;In certain situations
Under, if the calculation resources of mist computation layer are not critical issue compared with the delay of task, the value of η can be set to the value of very little
Or 0, then problem, which becomes, minimizes one of delay;On the contrary, if calculation resources occupancy is main problem, η compared with delay
Value can be set to bigger value;Formula (22) and formula (23) expression can assign the task in mist computation layer or cloud
A virtual machine;Formula (24) indicates task deadline of the overall delay no more than setting of any task;Formula (25)
Indicate that the transmission bandwidth of the needs of any task must not exceed total bandwidth.
Further, in the step 5, the input of the effective task unloading algorithm of the calculation resources of the load balancing includes
Vehicle termination set, set of tasks, the deadline of task, available virtual machine set in cloud, mist computation layer available virtual machine collection
It closes;
If the resources occupation rate of mist computation layer is low, task can be distributed directly to the virtual machine of mist computation layer, further
Ground calls mist computation layer to execute task process, is divided according to the return of network intrinsic fog computation layer virtual machine computing resource occupancy
The virtual machine position of mist computation layer with task calculates for executing the resource occupation of the virtual machine of the task, and updates mist
The computing resource of all virtual machines of computation layer occupies;
If the resources occupation rate of mist computation layer is higher, cloud is offloaded tasks to, then cloud computing layer is called to execute task
Process, the virtual machine position of task, the virtual machine will wherein distributed by executing based on the task by returning in cloud computing layer
The computing resource for calculating all virtual machines of resource occupation and cloud occupies;
The service delay of request task is returned after end whole service process.
Further, it is the total of all virtual machine computing resources occupancy of mist computation layer that the mist computation layer computing resource, which occupies,
With;Similarly, the cloud computing resource occupancy is the summation that all virtual machine computing resources in cloud occupy.
Compared with prior art, the beneficial effects of the present invention are: generating the mistake of task caused by congestion because traffic density is big
The relatively existing calculating discharging method of the quantity lost is less;Because of the relatively existing calculating unloading of failed tasks quantity caused by network problem
Method is less;Because the calculation resources problem relatively existing calculating discharging method of task quantity that leads to the failure is less.
Detailed description of the invention
Fig. 1 is the flow diagram of the method for the invention.
Fig. 2 is that mist calculates network system architecture figure.
Specific embodiment
Technical solution of the present invention is described in further detail with reference to the accompanying drawings of the specification.
A kind of vehicle big data calculating discharging method based on mist network, steps are as follows for specific method:
Step 1: proposing that mist calculates network system architecture.
The mist calculates network system architecture and is divided into three layers: (1) application layer, (2) mist computation layer, including mist calculate node
Equipment, (3) cloud computing layer, including cloud computing equipment are calculated with mist.
When vehicle is moved to corresponding crossing, when into the overlay area of access point signals, corresponding nothing is added in vehicle termination
Line local area network simultaneously accesses mist calculating equipment, sends mist calculate node for calculating task;In addition, if vehicle termination determines to appoint
Business is unloaded to cloud computing equipment, then the WAN connection access cloud provided by Wi-Fi access point is used to calculate equipment;Mist computation layer
Also it is connect with cloud computing system;In application layer, vehicle termination generates task requests to be further processed.
Step 2: establishing Network Delay Model.
Specifically, carrying out calculating unloading using cloud and mist cooperation, each task will be off-loaded to cloud or mist;This
Under scene, three data transfer phase are had, including wireless transmission stage and wire transmission stage and calculated result return rank
Section;Postponed the calculating of i.e. response time to the data transfer phase.
After request task submits to mist computation layer or cloud is handled, service delay i.e. response time can be with
It is indicated with the sum of the processing delay of transmission delay and request task;dvfAnd dfcIt is nearest mist meter from vehicle termination to access point
Calculate equipment and node and the transmission delay being grouped from mist computation layer to the individual data of cloud computing layer in layer.
The N run in mist computation layeriThe average transfer delay d of the data packet of a request task application examplefogBy following formula
It provides:
Wherein, PiAnd pi(Pi> pi) it is NiA task is sent to mist computation layer and is sent in cloud data from mist computation layer
The data packet sum of the heart;brThe quantity of the total data packet sent for the response to b request task.
The average transfer delay of mist computation layer processing request task application exampleIt is given by:
In cloud computing layer, it can indicate are as follows:
The delay of request task application example be by the quantity of the request task of server-side processes before treatment come
It calculates, the request task application example sum that can be handled simultaneously are as follows:
Equal part total bandwidth B gives N number of request task application example, and the frequency for occupying each user is not interfere with each other with same
When send their data to mist computation layer and cloud;Therefore:
Respectively indicate vehicle termination Vi∈{V1,V2,...,VwUplink and downlink transmission rate;Here n0It is
Noise power spectral density, hiIt is base station and user NiBetween channel gain, pd,iAnd pu,iIt is vehicle termination V respectivelyiDownlink chain
Road and up-link power.
Set Δ (Vi,Ii) it is to operate in vehicle termination ViIn request task application example IiService delay, You Wuji
It calculates layer and service is provided;In NiIn a request task application example, it is assumed that ni(Ni> ni) a request task application example is redirected
Cloud computing layer is unloaded to carry out operation;The sum for the request task application example that cloud computing layer is handled in time t are as follows:
And it is for each of this n request task application examples, processing latencies beyond the clouds
In ViThe average treatment of the request task application example of interior operation postpones
Then all vehicle termination ViAverage service delayIt can indicate are as follows:
On the contrary, in cloud computing layer, all request task application examples of user side operation are directly and core calculation module
Interaction, the average treatment delay of request task application example hereIt is given by:
Step 3: establishing task and generate model.
Specifically, the length of size of data and task is subjected to distribution appropriate according to the distribution that selected task generates,
Select suitable task length to reduce the quantity of mission failure and the average value of network delay as far as possible;Task arrival time is logical
Cross with exponential distribution and be made of independent same distribution Poisson distribution modeling obtain.
In the step 3, the Poisson distribution modeling are as follows:
One essential characteristic of Poisson distribution is that the probability of x is an independent discrete value, its probability and previous institute
There is value unrelated.
The underlying attribute of exponential process is without memory;The waiting time is identified when task reaches;Memoryless characteristic shows
Since task arrival interval is not to the designated time, the distribution of residual waiting time to it is initial similar, be shown below:
Since vehicle termination will not continuously generate service request, mode is generated using free time/active task to simulate
Real scene;According to this mode, user creates task during activity, is waited for during the free time, i.e., each end
There is a state machine at end, and terminal may be at active state or idle state.
Step 4: establishing mist computing resource Optimized model.
Data are collected, the incoming task set, the virtual machine set in cloud computing layer, mist for obtaining each vehicle termination calculate
The virtual machine set of layer;Then must go out on missions the delay being unloaded on cloud computing layer and the virtual machine of mist computation layer respectively, into one
Step obtains the calculation resources occupancy that vehicle termination executes corresponding task on cloud computing layer and the virtual machine of mist computation layer respectively;It is logical
The formulation of objective function and about beam equation is crossed, balancing delay and calculation resources occupy two elements, obtain optimal calculation resources
Distribution.
In the step 4, specifically, assuming that there are one group of vehicle termination, quantity m, with M={ M in system model1,
M2,...,MmIndicate;Each vehicle termination has the incoming task of limited quantity, the incoming task collection table of i-th of vehicle termination
It is shown asWherein, tijIndicate j-th of incoming task of i-th of vehicle termination, vehicle termination MiWith niIt is a
Task;SetThe deadline of i-th of vehicle termination is represented, wherein dijIt is the deadline of task;
Set VC={ vc1,vc2,...,vcpP virtual machine of the expression in cloud computing layer, and set VFDC={ vfdc1,
vfdc2,...,vfdcqIndicate all mist computation layers q group virtual machine;There are k mist computation layer, each mist calculates center
There is qiA virtual machine (VirtualMachine, VM), wherein 1≤i≤k.
X is defined simultaneouslyijk, YijkTwo variables:
Assuming thatWithIt is by by task t respectivelyijK-th of the virtual machine and mist for being unloaded to cloud computing layer calculate
K-th of virtual machine of layer and the delay that generates, task tijExecute delay are as follows:
Then, the total delay for executing all tasks can indicate are as follows:
Vehicle termination MiTask t is executed on k-th of virtual machine of mist computation layerijOccupancy to calculation resources is cijk,
Task t is executed on k-th of virtual machine of cloud computing layerijOccupancy to calculation resources is c'ijk。
Vehicle termination MiCalculation resources occupy CiIncluding two parts: processor active task (1) is unloaded to the void of mist computation layer
The quasi- occupied calculation resources of machine, processor active task is unloaded to the occupied calculation resources of the virtual machine in cloud by (2), we by its
It indicates are as follows:
Therefore, total calculation resources that all vehicle termination M are occupied are as follows:
The calculation resources in mist calculating environment are minimized for realization, have formulated following objective function and constraint equation:
Subject to
Objective function, that is, formula (21) maintains calculation resources to occupy the tradeoff between delay using η;In certain situations
Under, if the calculation resources of mist computation layer are not critical issue compared with the delay of task, the value of η can be set to the value of very little
Or 0, then problem, which becomes, minimizes one of delay;On the contrary, if calculation resources occupancy is main problem, η compared with delay
Value can be set to bigger value;Formula (22) and formula (23) expression can assign the task in mist computation layer or cloud
A virtual machine;Formula (24) indicates task deadline of the overall delay no more than setting of any task;Formula (25)
Indicate that the transmission bandwidth of the needs of any task must not exceed total bandwidth.
Step 5: management mist calculates the computing resource of Network Load Balance, and road vehicle terminal request processor active task is distributed
To optimal mist computing resource.
Specifically, the algorithm that a kind of road vehicle terminal request processor active task distributes to optimal mist computing resource is proposed,
That is the effective task of the calculation resources of load balancing unloads algorithm;It is needed by using expected resource requirement matrix to estimate computing resource
It asks;According to expected resource requirement matrix, virtual machine needs different time and efforts to execute different tasks;Scheduler first
Mist computation layer is offloaded tasks to, if mist computation layer calculation resources occupancy is excessively high, offloads tasks to cloud.
In the step 5, the input of the effective task unloading algorithm of the calculation resources of the load balancing includes vehicle termination
Gather, set of tasks, the deadline of task, available virtual machine set in cloud, mist computation layer available virtual machine set.
If the resources occupation rate of mist computation layer is low, task can be distributed directly to the virtual machine of mist computation layer, further
Ground calls mist computation layer to execute task process, is divided according to the return of network intrinsic fog computation layer virtual machine computing resource occupancy
The virtual machine position of mist computation layer with task calculates for executing the resource occupation of the virtual machine of the task, and updates mist
The computing resource of all virtual machines of computation layer occupies.
If the resources occupation rate of mist computation layer is higher, cloud is offloaded tasks to, then cloud computing layer is called to execute task
Process, the virtual machine position of task, the virtual machine will wherein distributed by executing based on the task by returning in cloud computing layer
The computing resource for calculating all virtual machines of resource occupation and cloud occupies.
The mist computation layer computing resource occupancy is the summation that all virtual machine computing resources of mist computation layer occupy;It is similar
Ground, the cloud computing resource occupancy are the summations that all virtual machine computing resources in cloud occupy.
The service delay of request task is returned after end whole service process.
The foregoing is merely better embodiment of the invention, protection scope of the present invention is not with above embodiment
Limit, as long as those of ordinary skill in the art's equivalent modification or variation made by disclosure according to the present invention, should all be included in power
In the protection scope recorded in sharp claim.
Claims (7)
1. a kind of vehicle big data based on mist network calculates discharging method, it is characterised in that: steps are as follows for specific method:
Step 1: proposing that mist calculates network system architecture;
The mist calculates network system architecture and is divided into three layers: (1) application layer, (2) mist computation layer, including mist calculate node and mist
Calculate equipment, (3) cloud computing layer, including cloud computing equipment;
Step 2: establishing Network Delay Model;
Specifically, carrying out calculating unloading using cloud and mist cooperation, each task will be off-loaded to cloud or mist;In this scene
Under, three data transfer phase are had, including wireless transmission stage and wire transmission stage and calculated result return the stage;
Postponed the calculating of i.e. response time to the data transfer phase;
Step 3: establishing task and generate model;
Specifically, the length of size of data and task is carried out distribution appropriate according to the distribution that selected task generates, to the greatest extent may be used
Suitable task length can be selected to reduce the quantity of mission failure and the average value of network delay;Task arrival time passes through tool
By exponential distribution and be made of independent same distribution Poisson distribution modeling obtain;
Step 4: establishing mist computing resource Optimized model;
It collects data, obtains the incoming task set of each vehicle termination, the virtual machine set in cloud computing layer, mist computation layer
Virtual machine set;Then must go out on missions the delay being unloaded on cloud computing layer and the virtual machine of mist computation layer respectively, further
The calculation resources for executing corresponding task on cloud computing layer and the virtual machine of mist computation layer respectively to vehicle termination occupy;Pass through mesh
The formulation of scalar functions and constraint equation, balancing delay and calculation resources occupy two elements, obtain optimal calculation resources distribution;
Step 5: management mist calculates the computing resource of Network Load Balance, and road vehicle terminal request processor active task is distributed to most
Excellent mist computing resource;
Specifically, propose the algorithm that a kind of road vehicle terminal request processor active task distributes to optimal mist computing resource, i.e., it is negative
It carries the effective task of balanced calculation resources and unloads algorithm;Computational resource requirements are estimated by using expected resource requirement matrix;
According to expected resource requirement matrix, virtual machine needs different time and efforts to execute different tasks;Scheduler will first
Task is unloaded to mist computation layer, if mist computation layer calculation resources occupancy is excessively high, offloads tasks to cloud.
2. a kind of vehicle big data based on mist network according to claim 1 calculates discharging method, it is characterised in that: institute
It states in step 1, when vehicle is moved to corresponding crossing, when into the overlay area of access point signals, vehicle termination is added corresponding
Wireless LAN simultaneously accesses mist calculating equipment, sends mist calculate node for calculating task;In addition, if vehicle termination decision will
Task is unloaded to cloud computing equipment, then the WAN connection access cloud provided by Wi-Fi access point is used to calculate equipment;Mist calculates
Layer is also connect with cloud computing system;In application layer, vehicle termination generates task requests to be further processed.
3. a kind of vehicle big data based on mist network according to claim 1 calculates discharging method, it is characterised in that: institute
It states in step 2, after request task submits to mist computation layer or cloud is handled, service delay i.e. response time can
To be indicated with the sum of the processing of transmission delay and request task delay;dvfAnd dfcIt is nearest mist from vehicle termination to access point
Equipment and node and the transmission delay being grouped from mist computation layer to the individual data of cloud computing layer in computation layer;
The N run in mist computation layeriThe average transfer delay d of the data packet of a request task application examplefogIt is given by:
Wherein, PiAnd pi(Pi> pi) it is NiA task is sent to mist computation layer and is sent to cloud data center from mist computation layer
Data packet sum;brThe quantity of the total data packet sent for the response to b request task;
The average transfer delay of mist computation layer processing request task application exampleIt is given by:
In cloud computing layer, it can indicate are as follows:
The delay of request task application example is calculated by the quantity of the request task of server-side processes before treatment
, the request task application example sum that can be handled simultaneously are as follows:
Equal part total bandwidth B gives N number of request task application example, and the frequency for occupying each user does not interfere with each other with while inciting somebody to action
Its data is sent to mist computation layer and cloud;Therefore:
Respectively indicate vehicle termination Vi∈{V1,V2,...,VwUplink and downlink transmission rate;Here n0It is noise
Power spectral density, hiIt is base station and user NiBetween channel gain, pd,iAnd pu,iIt is vehicle termination V respectivelyiDownlink and
Up-link power;
Set Δ (Vi,Ii) it is to operate in vehicle termination ViIn request task application example IiService delay, by mist computation layer
Service is provided;In NiIn a request task application example, it is assumed that ni(Ni> ni) a request task application example is redirected unloading
To cloud computing layer to carry out operation;The sum for the request task application example that cloud computing layer is handled in time t are as follows:
And it is for each of this n request task application examples, processing latencies beyond the cloudsIn Vi
The average treatment of the request task application example of interior operation postpones
Then all vehicle termination ViAverage service delayIt can indicate are as follows:
On the contrary, all request task application examples of user side operation are directly interacted with core calculation module in cloud computing layer,
Here the average treatment of request task application example postponesIt is given by:
4. a kind of vehicle big data based on mist network according to claim 1 calculates discharging method, it is characterised in that: institute
It states in step 3, the Poisson distribution modeling are as follows:
One essential characteristic of Poisson distribution is that the probability of x is an independent discrete value, its probability and previous all values
It is unrelated;
The underlying attribute of exponential process is without memory;The waiting time is identified when task reaches;Memoryless characteristic show due to
Task arrival interval not to the designated time, the distribution of residual waiting time to it is initial similar, be shown below:
Since vehicle termination will not continuously generate service request, mode is generated using free time/active task to simulate reality
Scene;According to this mode, user creates task during activity, is waited for during the free time, i.e., each terminal
There is a state machine, terminal may be at active state or idle state.
5. a kind of vehicle big data based on mist network according to claim 1 calculates discharging method, it is characterised in that: institute
It states in step 4, specifically, assuming that there are one group of vehicle termination, quantity m, with M={ M in system model1,M2,...,Mm}
It indicates;Each vehicle termination has the incoming task of limited quantity, and the incoming task set representations of i-th of vehicle termination areWherein, tijIndicate j-th of incoming task of i-th of vehicle termination, vehicle termination MiWith niA
Business;SetThe deadline of i-th of vehicle termination is represented, wherein dijIt is the deadline of task;Collection
Close VC={ vc1,vc2,...,vcpP virtual machine of the expression in cloud computing layer, and set VFDC={ vfdc1,
vfdc2,...,vfdcqIndicate all mist computation layers q group virtual machine;There are k mist computation layer, each mist calculates center
There is qiA virtual machine (VirtualMachine, VM), wherein 1≤i≤k;
X is defined simultaneouslyijk, YijkTwo variables:
Assuming thatWithIt is by by task t respectivelyijIt is unloaded to k-th of the virtual machine and mist computation layer of cloud computing layer
K-th of virtual machine and the delay generated, task tijExecute delay are as follows:
Then, the total delay for executing all tasks can indicate are as follows:
Vehicle termination MiTask t is executed on k-th of virtual machine of mist computation layerijOccupancy to calculation resources is cijk, in cloud meter
It calculates and executes task t on k-th of virtual machine of layerijOccupancy to calculation resources is c'ijk;
Vehicle termination MiCalculation resources occupy CiIncluding two parts: processor active task (1) is unloaded to the virtual machine of mist computation layer
Processor active task is unloaded to the occupied calculation resources of the virtual machine in cloud by occupied calculation resources, (2), we are indicated
Are as follows:
Therefore, total calculation resources that all vehicle termination M are occupied are as follows:
The calculation resources in mist calculating environment are minimized for realization, have formulated following objective function and constraint equation:
Subject to
Objective function, that is, formula (21) maintains calculation resources to occupy the tradeoff between delay using η;In some cases, such as
The calculation resources of fruit mist computation layer are not critical issue compared with the delay of task, and the value of η can be set to the value or 0 of very little,
Then problem, which becomes, minimizes one of delay;On the contrary, if calculation resources occupancy is main problem, the value of η compared with delay
It can be set to bigger value;Formula (22) and formula (23) indicate can to assign the task to one in mist computation layer or cloud
A virtual machine;Formula (24) indicates task deadline of the overall delay no more than setting of any task;Formula (25) indicates
The transmission bandwidth of the needs of any task must not exceed total bandwidth.
6. a kind of vehicle big data based on mist network according to claim 1 calculates discharging method, it is characterised in that: institute
It states in step 5, the input of the effective task unloading algorithm of the calculation resources of the load balancing includes vehicle termination set, task-set
It closes, the deadline of task, available virtual machine set in cloud, mist computation layer available virtual machine set;
If the resources occupation rate of mist computation layer is low, task can be distributed directly to the virtual machine of mist computation layer, further, adjusted
Task process is executed with mist computation layer, distributed task is returned to according to network intrinsic fog computation layer virtual machine computing resource occupancy
Mist computation layer virtual machine position, calculate for executing the resource occupation of the virtual machine of the task, and update mist computation layer
All virtual machines computing resource occupy;
If the resources occupation rate of mist computation layer is higher, cloud is offloaded tasks to, then cloud computing layer is called to execute task process,
Returning in cloud computing layer will be used to execute the computing resource of the task in the wherein virtual machine position of distribution task, the virtual machine
It occupies and the computing resource of all virtual machines of cloud occupies;
The service delay of request task is returned after end whole service process.
7. a kind of vehicle big data based on mist network according to claim 6 calculates discharging method, it is characterised in that: institute
Stating mist computation layer computing resource occupancy is the summation that all virtual machine computing resources of mist computation layer occupy;Similarly, the cloud
Computing resource occupancy is the summation that all virtual machine computing resources in cloud occupy.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910246291.5A CN109947574B (en) | 2019-03-29 | 2019-03-29 | Fog network-based vehicle big data calculation unloading method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910246291.5A CN109947574B (en) | 2019-03-29 | 2019-03-29 | Fog network-based vehicle big data calculation unloading method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109947574A true CN109947574A (en) | 2019-06-28 |
CN109947574B CN109947574B (en) | 2023-05-30 |
Family
ID=67012724
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910246291.5A Active CN109947574B (en) | 2019-03-29 | 2019-03-29 | Fog network-based vehicle big data calculation unloading method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109947574B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110519776A (en) * | 2019-08-07 | 2019-11-29 | 东南大学 | Balanced cluster and federated resource distribution method in a kind of mist computing system |
CN111010434A (en) * | 2019-12-11 | 2020-04-14 | 重庆工程职业技术学院 | Optimized task unloading method based on network delay and resource management |
CN111124531A (en) * | 2019-11-25 | 2020-05-08 | 哈尔滨工业大学 | Dynamic unloading method for calculation tasks based on energy consumption and delay balance in vehicle fog calculation |
CN112416603A (en) * | 2020-12-09 | 2021-02-26 | 北方工业大学 | Combined optimization system and method based on fog calculation |
CN112685186A (en) * | 2021-01-08 | 2021-04-20 | 北京信息科技大学 | Method and device for unloading computing tasks, electronic equipment and storage medium |
CN113015109A (en) * | 2021-02-23 | 2021-06-22 | 重庆邮电大学 | Wireless virtual network access control method in vehicle fog calculation |
CN114363215A (en) * | 2021-12-27 | 2022-04-15 | 北京特种机械研究所 | Train communication network time delay analysis method based on supply and demand balance |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160182639A1 (en) * | 2014-12-17 | 2016-06-23 | University-Industry Cooperation Group Of Kyung-Hee University | Internet of things network system using fog computing network |
CN109343904A (en) * | 2018-09-28 | 2019-02-15 | 燕山大学 | A kind of mist calculating dynamic offloading method based on Lyapunov optimization |
-
2019
- 2019-03-29 CN CN201910246291.5A patent/CN109947574B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160182639A1 (en) * | 2014-12-17 | 2016-06-23 | University-Industry Cooperation Group Of Kyung-Hee University | Internet of things network system using fog computing network |
CN109343904A (en) * | 2018-09-28 | 2019-02-15 | 燕山大学 | A kind of mist calculating dynamic offloading method based on Lyapunov optimization |
Non-Patent Citations (1)
Title |
---|
李波: "云机器人系统研究综述", 《计算机工程与应用》 * |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110519776A (en) * | 2019-08-07 | 2019-11-29 | 东南大学 | Balanced cluster and federated resource distribution method in a kind of mist computing system |
CN110519776B (en) * | 2019-08-07 | 2021-09-17 | 东南大学 | Balanced clustering and joint resource allocation method in fog computing system |
CN111124531A (en) * | 2019-11-25 | 2020-05-08 | 哈尔滨工业大学 | Dynamic unloading method for calculation tasks based on energy consumption and delay balance in vehicle fog calculation |
CN111124531B (en) * | 2019-11-25 | 2023-07-28 | 哈尔滨工业大学 | Method for dynamically unloading calculation tasks based on energy consumption and delay balance in vehicle fog calculation |
CN111010434A (en) * | 2019-12-11 | 2020-04-14 | 重庆工程职业技术学院 | Optimized task unloading method based on network delay and resource management |
CN111010434B (en) * | 2019-12-11 | 2022-05-27 | 重庆工程职业技术学院 | Optimized task unloading method based on network delay and resource management |
CN112416603A (en) * | 2020-12-09 | 2021-02-26 | 北方工业大学 | Combined optimization system and method based on fog calculation |
CN112416603B (en) * | 2020-12-09 | 2023-04-07 | 北方工业大学 | Combined optimization system and method based on fog calculation |
CN112685186A (en) * | 2021-01-08 | 2021-04-20 | 北京信息科技大学 | Method and device for unloading computing tasks, electronic equipment and storage medium |
CN112685186B (en) * | 2021-01-08 | 2023-04-28 | 北京信息科技大学 | Method and device for unloading computing task, electronic equipment and storage medium |
CN113015109A (en) * | 2021-02-23 | 2021-06-22 | 重庆邮电大学 | Wireless virtual network access control method in vehicle fog calculation |
CN114363215A (en) * | 2021-12-27 | 2022-04-15 | 北京特种机械研究所 | Train communication network time delay analysis method based on supply and demand balance |
Also Published As
Publication number | Publication date |
---|---|
CN109947574B (en) | 2023-05-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109947574A (en) | A kind of vehicle big data calculating discharging method based on mist network | |
CN113950066B (en) | Single server part calculation unloading method, system and equipment under mobile edge environment | |
CN109714219B (en) | Virtual network function rapid mapping method based on satellite network | |
CN109684075B (en) | Method for unloading computing tasks based on edge computing and cloud computing cooperation | |
CN108566659B (en) | 5G network slice online mapping method based on reliability | |
CN109788046B (en) | Multi-strategy edge computing resource scheduling method based on improved bee colony algorithm | |
CN111818168A (en) | Self-adaptive joint calculation unloading and resource allocation method in Internet of vehicles | |
CN111930436A (en) | Random task queuing and unloading optimization method based on edge calculation | |
CN113364850B (en) | Software-defined cloud-edge collaborative network energy consumption optimization method and system | |
CN109669768A (en) | A kind of resource allocation and method for scheduling task towards side cloud combination framework | |
CN110489176B (en) | Multi-access edge computing task unloading method based on boxing problem | |
CN114363984B (en) | Cloud edge collaborative optical carrier network spectrum resource allocation method and system | |
CN113156992B (en) | Three-layer architecture collaborative optimization method for unmanned aerial vehicle in edge environment | |
US20210006459A1 (en) | Network and Method for Servicing a Computation Request | |
CN111124662A (en) | Fog calculation load balancing method and system | |
CN108111335A (en) | A kind of method and system dispatched and link virtual network function | |
CN109005211B (en) | Micro-cloud deployment and user task scheduling method in wireless metropolitan area network environment | |
CN110519776A (en) | Balanced cluster and federated resource distribution method in a kind of mist computing system | |
CN111711962A (en) | Cooperative scheduling method for subtasks of mobile edge computing system | |
CN112162789A (en) | Edge calculation random unloading decision method and system based on software definition | |
CN116893861A (en) | Multi-agent cooperative dependency task unloading method based on space-ground cooperative edge calculation | |
CN113114738A (en) | SDN-based optimization method for internet of vehicles task unloading | |
CN114564312A (en) | Cloud edge-side cooperative computing method based on adaptive deep neural network | |
Tseng et al. | An mec-based vnf placement and scheduling scheme for ar application topology | |
CN115134371A (en) | Scheduling method, system, equipment and medium containing edge network computing resources |
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 |