CN108833486B - Hybrid dynamic task scheduling method for complex vehicle-mounted fog computing system environment - Google Patents

Hybrid dynamic task scheduling method for complex vehicle-mounted fog computing system environment Download PDF

Info

Publication number
CN108833486B
CN108833486B CN201810495297.1A CN201810495297A CN108833486B CN 108833486 B CN108833486 B CN 108833486B CN 201810495297 A CN201810495297 A CN 201810495297A CN 108833486 B CN108833486 B CN 108833486B
Authority
CN
China
Prior art keywords
vehicle
fog
task
scheduling
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.)
Active
Application number
CN201810495297.1A
Other languages
Chinese (zh)
Other versions
CN108833486A (en
Inventor
陈潇
林碧玲
王良民
蔡英凤
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.)
Jiangsu University
Original Assignee
Jiangsu 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 Jiangsu University filed Critical Jiangsu University
Priority to CN201810495297.1A priority Critical patent/CN108833486B/en
Publication of CN108833486A publication Critical patent/CN108833486A/en
Application granted granted Critical
Publication of CN108833486B publication Critical patent/CN108833486B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1004—Server selection for load balancing
    • H04L67/101—Server selection for load balancing based on network conditions
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
    • H04L67/62—Establishing a time schedule for servicing the requests
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/30—Services specially adapted for particular environments, situations or purposes
    • H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
    • H04W4/44—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P] for communication between vehicles and infrastructures, e.g. vehicle-to-cloud [V2C] or vehicle-to-home [V2H]

Abstract

The invention discloses a hybrid dynamic task scheduling method for a complex vehicle-mounted fog computing system environment, which comprises the following steps: (1) establishing a vehicle-mounted fog calculation scheduling model, and initializing the vehicleA task set and a fog server set in a fog computing scene; (2) observing the task flow change and the fog system service capability change in the vehicle-mounted fog computing environment through a system environment change measurement function Var (r), and further obtaining a system variation coefficient v1,v2(ii) a (3) According to the discrete difference condition reflected by the variation coefficient, the decision support function DSF (v)1,v2) And generating a decision, and calling a proper scheduling algorithm to distribute the vehicle-mounted task to the fog server. According to the invention, by observing the change condition of the environment of the vehicle-mounted fog computing system and dynamically selecting the scheduling algorithm, the vehicle-mounted tasks can be reasonably and efficiently distributed to each fog server for processing, the processing efficiency of the vehicle-mounted tasks is improved, and the stability of the service capacity of the fog servers is improved.

Description

Hybrid dynamic task scheduling method for complex vehicle-mounted fog computing system environment
Technical Field
The invention belongs to a vehicle-mounted communication technology, and particularly relates to a hybrid dynamic task scheduling method for a complex vehicle-mounted fog computing system environment.
Background
The Internet of vehicles is a new field taking a vehicle as a main body to participate in Internet connection, and is a typical application of the Internet of things in the traffic field. By means of the interconnection of the intelligence of the single vehicles, road traffic information is shared, wireless communication and information interaction between vehicles and X (X: vehicles, roads, pedestrians, the Internet and the like) are improved, and development of an intelligent traffic system is promoted. With the development of the car networking technology and industry, the number of intelligent cars is increased, the car-mounted applications are diversified, the quantity of car-mounted information and the information transmission speed are increased in a geometric progression, and how to process the huge data and the processing efficiency of the data becomes a problem which needs to be solved in the car networking.
The traditional Internet of vehicles is a 'end management cloud' three-layer system from the network perspective, wherein an end refers to an intelligent sensor which is responsible for collecting vehicle-mounted information and sensing a driving state and an environment; the management refers to solving the interconnection and intercommunication of the vehicle-X and realizing the communication and roaming among networks; the cloud refers to an information platform of a cloud architecture and provides cloud computing functions such as virtualization, security authentication, real-time interaction and mass storage. Although the cloud technology solves the problem of data storage and processing efficiency in the internet of vehicles to a certain extent, when the cloud system in the internet of vehicles faces a large number of widely distributed vehicle-mounted terminal devices and a large amount of collected vehicle-mounted data, the cloud system in the internet of vehicles inevitably encounters some problems.
Firstly, mass vehicle-mounted data are gushed into a core network cloud, which causes serious network congestion; secondly, the vehicle-mounted equipment is far away from the cloud data center, so that data transmission is delayed; finally, there is a risk of long distance communication from the on-board device to the cloud, and the cost of backup in the cloud is also high.
On the other hand, with the development of the vehicle industry technology, a vehicle has certain computing capacity and storage capacity, but the capacity of a single vehicle cannot well process all vehicle-mounted data and tasks, and transmitting all vehicle-mounted data and tasks to a cloud end for processing also causes the idle of vehicle-mounted physical resources, so that the load of a cloud service system is overlarge.
Therefore, in the internet of vehicles, it is not necessarily an effective method to transmit a large amount of data processing processes to the cloud center, and a scheduling scheme suitable for the environment of the internet of vehicles needs to be applied, so that the utilization rate of vehicle-mounted resources can be improved while vehicle-mounted data processing can be completed timely and efficiently.
In the existing scheme, there are many scheduling algorithms based on security, task scheduling focusing on failure recovery, scheduling considering energy and efficiency, and scheduling architecture of data parallel processing. However, these solutions consider the scenario that an onboard task queue and server follow a random distribution with a stable average; in a real vehicle-mounted environment, the vehicle-mounted task queue and the service capability are variable, and generally represent a random process with different values. Therefore, the existing scheduling scheme does not consider the diversification of the vehicle-mounted environment, and only assumes the vehicle-mounted environment to be in a fixed state. Therefore, in consideration of diversification of vehicle-mounted environments, the invention provides a hybrid dynamic task scheduling method for a complex vehicle-mounted fog computing system environment.
Disclosure of Invention
The purpose of the invention is as follows: the invention aims to solve the defects in the prior art and provides a hybrid dynamic task scheduling method for a complex vehicle-mounted fog computing system environment.
The technical scheme is as follows: the invention relates to a hybrid dynamic task scheduling method facing to a complex vehicle-mounted fog computing system environment, which comprises two dynamic scheduling algorithms, is suitable for vehicle-mounted task scheduling in different vehicle-mounted fog computing systems, and comprises the following steps:
(1) establishing a vehicle-mounted fog computing scheduling model (namely, firstly, utilizing idle physical resources in a vehicle to construct a fog computing environment, then applying a scheduling algorithm in the environment so as to improve the processing efficiency of data and tasks in the vehicle), and initializing a task set and a fog server set in a vehicle-mounted fog computing scene;
(2) observing the task flow change and the fog system service capability change in the vehicle-mounted fog computing environment through a system environment change measurement function Var (r), and further obtaining a system variation coefficient v1And v2(ii) a The coefficient of variation is the ratio of the standard deviation of the original data to the average of the original data, and is used for comparing the dispersion degrees of the two groups of data, and the coefficient of variation reflects the absolute value of the dispersion degree of the data; the data size is not only influenced by the discrete degree of the variable values, but also influenced by the average level of the variable values; in addition, the original data refers to corresponding task flow data generated by communication among vehicles, roads and networks in the vehicle-mounted environment and related data of the capacity of each server in the vehicle-mounted fog computing environment; in the vehicle-mounted fog calculation scheduling model, the variation coefficient comprises the arrival variation v of the vehicle-mounted task queue workflow1With fog server capability change v2；
(3) According to the discrete condition of the variation coefficient, the decision support function DSF (v)1,v2) Generating a decision, and calling a corresponding scheduling algorithm to distribute the vehicle-mounted task to the fog server; and the dynamic scheduling algorithm selects a corresponding one of the following two methods according to actual conditions:
a) the queue-based dynamic scheduling algorithm QDS has a scheduling process related to the length of a waiting queue in a server, and a vehicle-mounted task is allocated to the server with the shortest waiting queue length to be processed and completed within a certain time range;
b) the time-based dynamic scheduling algorithm TDS has a scheduling process related to task completion time in a queue, and at a specified time point, a vehicle-mounted task is processed and completed by a server which is less in time for completing the task.
Further, the queue-based dynamic scheduling algorithm QDS is represented by p (Q), and p (Q) is represented by a decision function DFQDS(t) construction, DF, according to the principles of the QDS algorithmQDSThe (t) function is expressed as:
Figure GDA0002483618290000031
where C is a constant coefficient of the system environment, t is the transmission time point, and q (t) is the waiting queue length generated by the system at time point t, since the value of q (t) changes with the change of time point t, the queue length is the instantaneous queue length. For clarity, the decision function is based on the QDS algorithm, and the instantaneous queue length based on the QDS algorithm is denoted by T [ q (T) ].
Further, said time-based dynamic scheduling algorithm TDS is denoted by p (T '), and p (T') is represented by a decision function DFTDS(t) composition, DF according to the principle of the TDS algorithmTDSThe (t) function is expressed as:
Figure GDA0002483618290000032
wherein K is the number of system users; r (T) is the response time of the system at the time T, and T [ R (T) ] is used as the response time is the instantaneous value of the response time at the time T, and the decision function is based on the TDS algorithm principle for the clear expression]Representing the transient response time based on the TDS algorithm; rS(K) Is an average response time and is a fixed value due to RS(K) Is obtained by approximate calculation of M/M/1// K queue model, so it is approximate average response time, denoted as E [ R ]S(K)]I.e. E [ R ]S(K)]Is the average response time and is a fixed value, T [ R (T)]The transient response time at time t (based on the TDS algorithm).
Further, the different in-vehicle fog computing system includes: the vehicle-mounted fog computing system with various task requirements and the condition that the service capability of the system is unstable exist.
Further, the specific method for initializing in step (1) is as follows:
(1.1) the elements in the task set J are tasks J in the vehicle-mounted fog computing systemkIn the constructed vehicle-mounted fog calculation scheduling model, each vehicle-mounted task is a quintuple, namely
jk＝＜k,v,λ,l,p＞,jk∈J(k＝1,2,...,n)
Wherein k is the ID of the vehicle-mounted task, v is the index of a source vehicle, lambda is the generation rate of the vehicle-mounted task, l is the length of the vehicle-mounted task, and p is the scheduling algorithm selected by the vehicle-mounted task;
(1.2) fog Server set S, wherein the elements in S are fog servers S in vehicle-mounted fog computing systemiIn the constructed vehicle-mounted fog calculation scheduling model, each fog server is a quadruple, namely
si＝＜i,v,μ,t＞,si∈S(i＝1,2,...,m)
Wherein i is the ID of the fog server, v is the vehicle index providing the physical resource, μ is the service capability of the given fog server, and t is the remaining service time of the fog server during the movement of the vehicle.
Further, the process of the step (2) is specifically as follows:
(2.1) calculating a standard deviation σ based on the on-vehicle task generation rate λ and the fog server service capacity μ:
Figure GDA0002483618290000041
(2.2) calculating coefficient of variation CV:
Var(λ)＝CVλ；Var(μ)＝CVμ；
wherein, CV isλFor task flow to reach changes, CVμServing the fog server for a change in capacity;
(2.3) obtaining the coefficient of variation v of the system1,v2：
v1＝CVλ；v2＝CVμ；
Figure GDA0002483618290000042
Further, the decision support function DSF (v) in the step (3)1,v2) The process is as follows:
Figure GDA0002483618290000043
wherein v isoutFor the selection of the scheduling algorithm, if the value is 1, the discrete degree of the arrival of the work task flow is large, the dynamic scheduling algorithm QDS based on the queue is selected, and if the value is 0, the discrete degree of the service capability of the fog server is large, the dynamic scheduling algorithm TDS based on the time is selected.
Has the advantages that: according to the invention, a vehicle-mounted fog computing environment is constructed by utilizing vehicle-mounted physical resources, so that the utilization of the vehicle-mounted resources is improved, and the problem of load balance in the Internet of vehicles is improved; based on the constructed vehicle-mounted fog computing and scheduling model, the invention provides a dynamic hybrid scheduling method, aiming at the changing situation of a vehicle-mounted task queue in a vehicle-mounted fog computing environment and the stable situation of the service capacity of a fog server, different algorithms are called to carry out vehicle-mounted task allocation, and therefore the processing efficiency of vehicle-mounted tasks is improved.
Drawings
FIG. 1 is a model diagram of a vehicle-mounted fog calculation scheduling of the present invention;
FIG. 2 is a diagram illustrating a specific process of a queue-based QDS in the present invention;
FIG. 3 is a diagram illustrating a specific process of a queue-based QDS of the present invention;
fig. 4 is an experimental result diagram of the hybrid scheduling algorithm and the comparison method of the present invention.
Detailed Description
The technical solution of the present invention is described in detail below, but the scope of the present invention is not limited to the embodiments.
As shown in fig. 1, the hybrid dynamic task scheduling method for a complex vehicle-mounted fog computing system environment of the present invention first establishes a vehicle-mounted fog computing scheduling model, and then provides a hybrid dynamic scheduling method suitable for a diversified vehicle-mounted fog computing environment on the basis of the established model. The method specifically comprises the following steps: (1) establishing a vehicle-mounted fog computing scheduling model, and initializing a task set and a fog server set in a vehicle-mounted fog computing scene; (2) observing the task flow change and the fog system service capability change in the vehicle-mounted fog computing environment through a system environment change measurement function Var (r), and further obtaining a system variation coefficient v1,v2(ii) a (3) According to the discrete condition of the variation coefficient, the decision support function DSF (v)1,v2) And generating a decision, and calling a proper scheduling algorithm to distribute the vehicle-mounted task to the fog server.
As shown in fig. 2 and fig. 3, two specific processes of the dynamic scheduling algorithm in the method are described as follows:
a) a queue-based dynamic scheduling algorithm QDS, as shown in fig. 2, the scheduling process of which is related to the length of the waiting queue in the server, and within a certain time range, the vehicle-mounted task will be allocated to the server with the shortest waiting queue length for processing and completing;
b) as shown in fig. 3, the time-based dynamic scheduling algorithm TDS is related to the task completion time in the queue in the scheduling process, and at a specific time point, even if the number of tasks in the processing queue is large, the in-vehicle task allocates a new task to a server with a small task completion time for processing and completion.
In the vehicle-mounted fog computing environment, a queue-based dynamic scheduling algorithm QDS in the hybrid scheduling method is represented by p (Q), and p (Q) is represented by a decision function DFQDS(t) construction, DF, according to the principles of the QDS algorithmQDSThe (t) function is described in detail as follows:
Figure GDA0002483618290000051
where C is a constant coefficient of the system environment, t is the transmission time point, and q (t) is the waiting queue length generated by the system at time point t, since the value of q (t) changes with the change of time point t, the queue length is the instantaneous queue length. For clarity, the decision function is based on the QDS algorithm, and the instantaneous queue length based on the QDS algorithm is denoted by T [ q (T) ].
In the vehicle fog computing environment, the time-based dynamic scheduling algorithm TDS in the hybrid scheduling method is represented by p (T '), and p (T') is represented by a decision function DFTDS(t) composition, DF according to the principle of the TDS algorithmTDSThe (t) function is described in detail as follows:
Figure GDA0002483618290000061
wherein K is the number of system users; r (T) is the response time of the system at the time T, and T [ R (T) ] is used as the response time is the instantaneous value of the response time at the time T, and the decision function is based on the TDS algorithm principle for the clear expression]Representing the transient response time based on the TDS algorithm; rS(K) Is an average response time and is a fixed value due to RS(K) Is obtained by approximate calculation of M/M/1// K queue model, so it is approximate average response time, denoted as E [ R ]S(K)]I.e. E [ R ]S(K)]Is the average response time and is a fixed value, T [ R (T)]The transient response time at time t (based on the TDS algorithm).
In an in-vehicle fog computing environment, said different in-vehicle fog computing system comprising: the vehicle-mounted fog computing system with various task requirements and the condition that the service capability of the system is unstable exist.
Establishing a vehicle-mounted fog computing scheduling model, and initializing specific implementation steps of a task set and a fog server set in a system, wherein in a vehicle-mounted network environment, due to rich application and networking cooperation of vehicles, a large number of data tasks of a source vehicle need to be processed, so that the tasks in the source vehicle are defined as a task set J; in addition, the free physical resources in the vehicle are utilized to construct a cluster of vehicles adjacent to the road into a fog server, and different fog servers form a fog server set S, such asFog server S in FIG. 11With fog server S2A fog server set S is formed, and the specific implementation steps are as follows:
a) the elements in the task set J and J are tasks J in the vehicle-mounted fog computing systemkIn the constructed vehicle-mounted fog calculation scheduling model, each vehicle-mounted task is a quintuple and is specifically represented as follows:
jk＝＜k,v,λ,l,p＞,jk∈J(k＝1,2,...,n)
wherein k is the ID of the vehicle-mounted task, v is the index of a source vehicle, lambda is the generation rate of the vehicle-mounted task, l is the length of the vehicle-mounted task, and p is the scheduling algorithm selected by the vehicle-mounted task;
b) the fog server set S is the fog server S in the vehicle-mounted fog computing systemiIn the constructed vehicle-mounted fog computing and scheduling model, each fog server is a quadruple, and the specific expression is as follows:
si＝＜i,v,μ,t＞,si∈S(i＝1,2,...,m)
wherein i is the ID of the fog server, i.e., S is represented when i is 11Fog Server, i is 2 and represents S2A fog server; v is the vehicle index providing the physical resources, μ is the service capacity of a given fog server, and t is the remaining service time of the fog server during the movement of the vehicle.
The coefficient of variation in the step (2) is a ratio of a standard deviation of the original data to an average of the original data, and is used for comparing the dispersion degrees of the two groups of data, and the coefficient of variation reflects an absolute value of the dispersion degree of the data; the data size is not only affected by the degree of dispersion of the variable values, but also by the average level of the variable values.
The method comprises the following specific steps of:
a) calculating a standard deviation sigma based on the vehicle-mounted task generation rate lambda and the service capacity mu of the fog server:
Figure GDA0002483618290000071
b) calculating the coefficient of variation CV:
Var(λ)＝CVλ；Var(μ)＝CVμ；
wherein, CV isλFor task flow to reach changes, CVμServing the fog server for a change in capacity;
c) obtaining the coefficient of variation v of the system1,v2：
v1＝CVλ；v2＝CVμ；
Figure GDA0002483618290000072
Decision support function DSF (v)1,v2) The method comprises the following specific steps:
Figure GDA0002483618290000073
wherein v isoutFor the selection of the scheduling algorithm, if the value is 1, the discrete degree of the arrival of the work task flow is large, the dynamic scheduling algorithm QDS based on the queue is selected, and if the value is 0, the discrete degree of the service capability of the fog server is large, the dynamic scheduling algorithm TDS based on the time is selected.
Implementation example:
in the hybrid dynamic task scheduling method for the complex vehicle-mounted fog computing system environment, fig. 1 shows a scene after a vehicle-mounted fog computing scheduling model is established and a task set and a fog server set in a vehicle-mounted fog computing scene are initialized; according to the scene constructed by the method shown in the figure 1, the defined queue-based dynamic scheduling algorithm QDS and the defined time-based dynamic scheduling algorithm TDS are deployed in a scheduling model, so that different scheduling algorithms can be dynamically applied to the vehicle-mounted task queues in the diversified vehicle-mounted fog computing environment, and the vehicle-mounted task queues can be effectively scheduled to different fog servers for processing and completing.
Furthermore, in specific experiments, the efficiency of different task scheduling algorithms in a diversified on-board fog computing environment was considered. In the experiment, the invention utilizes PEPA (Performance Evaluation Process Algebra) performance Evaluation Process algebra to carry out modeling. PEPA is a high-level model specification language that differs from classical process algebra in that each activity follows an exponential distribution of a random time. In addition, two comparison schemes are selected in the experiment to be compared with the scheme provided by the invention, so that the high efficiency of the hybrid scheduling method is proved. The specific comparison scheme is as follows:
comparison 1: using only a queue-based dynamic scheduling algorithm QDS in a diversified vehicle-borne fog computing environment;
comparison 2: using only a time-based dynamic scheduling algorithm TDS in a diverse on-board fog computing environment;
as shown in fig. 4, the experimental results show that the hybrid scheduling method of the present invention is more efficient and stable than the other two methods, and the hybrid scheduling method can generate better system performance.
According to the embodiment, the change condition of the vehicle-mounted fog computing system environment is observed, and the scheduling algorithm is dynamically selected, so that the vehicle-mounted tasks can be reasonably and efficiently distributed to each fog server for processing, the processing efficiency of the vehicle-mounted tasks is improved, and the stability of the service capacity of the fog servers is improved.

Claims (6)

1. A hybrid dynamic task scheduling method for a complex vehicle-mounted fog computing system environment is characterized by comprising the following steps: the method comprises two dynamic scheduling algorithms, is suitable for vehicle-mounted task scheduling of different system environments in a vehicle-mounted fog computing system, and comprises the following steps:
(1) establishing a vehicle-mounted fog computing scheduling model, and initializing a task set and a fog server set in a vehicle-mounted fog computing scene;
(2) by defining an environment change measurement function Var (r) by the system, wherein r is an independent variable of the Var () function, the task flow change and the fog system service capability change in the vehicle-mounted fog computing environment are observed, and then the variation coefficients of the task flow and the service capability in the system are respectively obtained, namely the variation coefficient of the vehicle-mounted task queue workflow reaching change v1With fog server capability change v2(ii) a The coefficient of variation is the standard deviation of the original data and the originalThe ratio of the mean of the starting data; the original data refers to corresponding task flow data generated by communication among vehicles, roads and networks in the vehicle-mounted environment and related data of the capacity of each server in the vehicle-mounted fog computing environment;
(3) decision support function DSF (v) according to discrete conditions reflected by coefficient of variation1,v2) Generating a decision, and calling a corresponding scheduling algorithm to distribute the vehicle-mounted task to the fog server; and the dynamic scheduling algorithm selects a corresponding one of the following two methods according to actual conditions:
a) the queue-based dynamic scheduling algorithm QDS has a scheduling process related to the length of a waiting queue in a server, and a vehicle-mounted task is allocated to the server with the shortest waiting queue length to be processed and completed within a certain time range;
b) the time-based dynamic scheduling algorithm TDS is characterized in that the scheduling process of the TDS is related to the task completion time in a queue, and at a specified time point, the vehicle-mounted task processes and completes the servers which are less in time for completing the task;
decision support function DSF (v)1,v2) The process is as follows:
Figure FDA0002543679070000011
CVλfor the case of change of task flow, CVμServing the fog server for a change in capacity;
wherein v isoutFor the selection of the scheduling algorithm, if the value is 1, the discrete degree of the arrival of the work task flow is large, the dynamic scheduling algorithm QDS based on the queue is selected, and if the value is 0, the discrete degree of the service capability of the fog server is large, the dynamic scheduling algorithm TDS based on the time is selected.
2. The complex vehicle fog computing system environment-oriented hybrid dynamic task scheduling method of claim 1, wherein: the queue-based dynamic scheduling algorithm QDS is represented by p (Q), Q represents the queue length, and p (Q) is represented by a decision functionDFQDS(t) construction, DF, according to the principles of the QDS algorithmQDSThe (t) function is expressed as:
Figure FDA0002543679070000021
wherein, C is a constant coefficient of the system environment, t is the transmission time point, q (t) is the waiting queue length generated by the system at the time point t, since the value of q (t) changes with the change of the time point t, the queue length is the instantaneous queue length; for clarity, the decision function is based on the QDS algorithm, and the instantaneous queue length based on the QDS algorithm is denoted by T [ q (T) ].
3. The complex vehicle fog computing system environment-oriented hybrid dynamic task scheduling method of claim 1, wherein: the time-based dynamic scheduling algorithm TDS is composed of a decision function DFTDS(t) composition, DF according to the principle of the TDS algorithmTDSThe (t) function is expressed as:
Figure FDA0002543679070000022
wherein K is the number of system users; r (T) is the response time of the system at the time T, and T [ R (T) ] is used because the response time is the instantaneous value of the response time at the time T and the decision function is based on the TDS algorithm principle for clear expression]Representing the transient response time based on the TDS algorithm; rS(K) Is an average response time and is a fixed value due to RS(K) Is obtained by approximate calculation of M/M/1// K queue model, so it is approximate average response time, denoted as E [ R ]S(K)]I.e. E [ R ]S(K)]Is the average response time and is a fixed value, T [ R (T)]Is the instantaneous response time at time t.
4. The complex vehicle fog computing system environment-oriented hybrid dynamic task scheduling method of claim 1, wherein: the on-board fog computing system includes: the vehicle-mounted fog computing system with various task requirements and the condition that the service capability of the system is unstable exist.
5. The complex vehicle fog computing system environment-oriented hybrid dynamic task scheduling method of claim 1, wherein: the specific method for initializing in the step (1) comprises the following steps:
(1.1) the elements in the task set J are tasks J in the vehicle-mounted fog computing systemkIn the constructed vehicle-mounted fog calculation scheduling model, each vehicle-mounted task is a quintuple, namely
jk＝＜k,v,λ,l,p＞,jk∈J(k＝1,2,...,n)
Wherein k is the ID of the vehicle-mounted task, v is the index of a source vehicle, lambda is the generation rate of the vehicle-mounted task, l is the length of the vehicle-mounted task, and p is the scheduling algorithm selected by the vehicle-mounted task;
(1.2) fog Server set S, wherein the elements in S are fog servers S in vehicle-mounted fog computing systemiIn the constructed vehicle-mounted fog calculation scheduling model, each fog server is a quadruple, namely
si＝＜i,v',μ,t＞,si∈S(i＝1,2,...,m)
Where i is the ID of the fog server, v' is the vehicle index providing the physical resources, μ is the service capacity of a given fog server, and t is the remaining service time of the fog server during movement of the vehicle.
6. The complex vehicle fog computing system environment-oriented hybrid dynamic task scheduling method of claim 1, wherein: the process of the step (2) is specifically as follows:
(2.1) calculating a standard deviation sigma including a standard deviation sigma of a task queue generation rate based on the on-board task generation rate lambda and the fog server service capability muλStandard deviation sigma from fog server service capabilityμThe specific calculation is as follows:
Figure FDA0002543679070000031
(2.2) calculating the coefficient of variation CV, coefficient of variation CV including task flowλCoefficient of variation CV with fog server capabilityμThe specific calculation formula is as follows:
Var(λ)＝CVλ；Var(μ)＝CVμ；
wherein, CV isλFor the case of change of task flow, CVμServing the fog server for a change in capacity;
(2.3) obtaining a system variation coefficient which is expressed as the arrival variation v of the workflow of the vehicle-mounted task queue1With fog server capability change v2：
v1＝CVλ；v2＝CVμ；
Figure FDA0002543679070000032
CN201810495297.1A 2018-05-22 2018-05-22 Hybrid dynamic task scheduling method for complex vehicle-mounted fog computing system environment Active CN108833486B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810495297.1A CN108833486B (en) 2018-05-22 2018-05-22 Hybrid dynamic task scheduling method for complex vehicle-mounted fog computing system environment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810495297.1A CN108833486B (en) 2018-05-22 2018-05-22 Hybrid dynamic task scheduling method for complex vehicle-mounted fog computing system environment

Publications (2)

Publication Number Publication Date
CN108833486A CN108833486A (en) 2018-11-16
CN108833486B true CN108833486B (en) 2020-08-25

Family

ID=64148955

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810495297.1A Active CN108833486B (en) 2018-05-22 2018-05-22 Hybrid dynamic task scheduling method for complex vehicle-mounted fog computing system environment

Country Status (1)

Country Link
CN (1) CN108833486B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110162390B (en) * 2019-05-15 2021-03-02 清华大学 Task allocation method and system for fog computing system
CN111614754B (en) * 2020-05-20 2022-04-05 重庆邮电大学 Fog-calculation-oriented cost-efficiency optimized dynamic self-adaptive task scheduling method

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102495758A (en) * 2011-12-05 2012-06-13 中南大学 Scheduling method of real-time tasks in distributing type high performance calculation environment
CN104035819B (en) * 2014-06-27 2017-02-15 清华大学深圳研究生院 Scientific workflow scheduling method and device
KR101574026B1 (en) * 2014-12-17 2015-12-03 경희대학교 산학협력단 Internet of things using fog computing network
CN105791371B (en) * 2014-12-22 2019-01-04 中国科学院上海高等研究院 A kind of cloud storage service system and method

Also Published As

Publication number Publication date
CN108833486A (en) 2018-11-16

Similar Documents

Publication Publication Date Title
Zhou et al. BEGIN: Big data enabled energy-efficient vehicular edge computing
Feng et al. AVE: Autonomous vehicular edge computing framework with ACO-based scheduling
CN111711666B (en) Internet of vehicles cloud computing resource optimization method based on reinforcement learning
CN111142883B (en) SDN architecture-based vehicle computing task unloading method
CN111831427B (en) Distributed inter-vehicle task unloading method based on mobile edge calculation
Pereira et al. A novel fog-based resource allocation policy for vehicular clouds in the highway environment
CN108833486B (en) Hybrid dynamic task scheduling method for complex vehicle-mounted fog computing system environment
CN110489218A (en) Vehicle-mounted mist computing system task discharging method based on semi-Markovian decision process
Fizza et al. PASHE: Privacy aware scheduling in a heterogeneous fog environment
de Souza et al. Exploring computation offloading in vehicular clouds
Siddiqi et al. Dynamic priority-based efficient resource allocation and computing framework for vehicular multimedia cloud computing
CN111611063A (en) Cloud-aware mobile fog computing system task unloading method based on 802.11p
Midya et al. Pso based optimized resource allocation in three tier cloud architecture for vanet
Choo et al. Reliable vehicle selection algorithm with dynamic mobility of vehicle in vehicular cloud system
CN111480349B (en) Control device and method for determining data format
CN115964178A (en) Internet of vehicles user computing task scheduling method and device and edge service network
Lee et al. Deep reinforcement learning-based resource allocation in vehicular fog computing
CN115208892A (en) Vehicle-road cooperative online task scheduling method and system based on dynamic resource demand
CN114138466A (en) Task cooperative processing method and device for intelligent highway and storage medium
Lee et al. A study of mobile edge computing system architecture for connected car media services on highway
CN114827946A (en) Task similarity-based edge calculation method and system in Internet of vehicles scene
CN114900870A (en) Cache content updating algorithm based on vehicle clustering
CN113448707A (en) Online batch parallel task scheduling method in edge computing
Pasha Opportunistic task offloading in vehicular networks
CN111615082A (en) Heterogeneous automatic driving vehicle formation calculation task collaborative execution 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