CN111479242A - Task unloading method for assisting vehicle formation through fog calculation - Google Patents

Task unloading method for assisting vehicle formation through fog calculation Download PDF

Info

Publication number
CN111479242A
CN111479242A CN202010249746.1A CN202010249746A CN111479242A CN 111479242 A CN111479242 A CN 111479242A CN 202010249746 A CN202010249746 A CN 202010249746A CN 111479242 A CN111479242 A CN 111479242A
Authority
CN
China
Prior art keywords
task
time
unloading
energy consumption
calculation
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.)
Pending
Application number
CN202010249746.1A
Other languages
Chinese (zh)
Inventor
崔太平
黄诗宇
杨明绪
陈舟
陈前斌
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.)
Chongqing University of Post and Telecommunications
Original Assignee
Chongqing University of Post and Telecommunications
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 Chongqing University of Post and Telecommunications filed Critical Chongqing University of Post and Telecommunications
Priority to CN202010249746.1A priority Critical patent/CN111479242A/en
Publication of CN111479242A publication Critical patent/CN111479242A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/30—Services specially adapted for particular environments, situations or purposes
    • H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
    • H04W4/44—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P] for communication between vehicles and infrastructures, e.g. vehicle-to-cloud [V2C] or vehicle-to-home [V2H]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/02—Arrangements for optimising operational condition
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/30—Services specially adapted for particular environments, situations or purposes
    • H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
    • H04W4/46—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P] for vehicle-to-vehicle communication [V2V]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Abstract

The invention relates to a task unloading method for assisting vehicle formation through fog calculation, and belongs to the technical field of communication. With the increasing demands of vehicle mobile applications, the independent use of the vehicle itself to complete tasks not only results in too high task processing overhead and consumes a large amount of energy of the vehicle terminal, but also the limited computing resources of the vehicle terminal may not meet the restrictive conditions of the task processing. The invention provides a scheme for dynamically adjusting task unloading between an FC server and vehicle formation members. The method is based on the Lyapunov optimization theory, task unloading judgment is carried out according to the parameters of the arriving task at each moment, the calculation queues of all tasks are guaranteed to be stable, the minimization of the average execution energy consumption of the tasks is realized, and the tasks are guaranteed to be completed within the deadline time.

Description

Task unloading method for assisting vehicle formation through fog calculation
Technical Field
The invention belongs to the technical field of communication, and relates to a task unloading method for assisting vehicle formation through fog calculation.
Background
The limited memory capacity, network bandwidth and processor speed of mobile vehicle terminals make the vehicle unable to meet the performance requirements of partial tasks, such as low latency and low power consumption. These requirements limit the vehicle terminals to running heavy multimedia and digital signal processing related applications. This is not only a temporary hardware limitation of current vehicles, but also an inherent deficiency of moving vehicles.
The pure electric vehicle is a hot topic of the emerging energy vehicle at present, not only the country greatly promotes the purchase of the pure electric vehicle, but also the consumption concept of consumers is more and more inclined to the electric vehicle. The most significant factor affecting the life of an electric vehicle is battery life. In the future, as the processing speed of the Vehicle terminal is faster and faster, the number of the Vehicle-mounted sensors is more and more, and the increasing of the Vehicle-to-Vehicle wireless communication (V2X) service will result in higher energy consumption. Longer battery life is more important than most other functions, including route navigation and information storage. The deployment of autonomous driving, the combination of Augmented Reality (AR) applications with vehicle driving technology, and the advent of other various V2X applications, consume large amounts of power and severely shorten the life of the vehicle battery. Such computationally intensive applications often fail to satisfy the restrictive conditions for their execution when executed in the in-vehicle terminal system.
There are 27 total use cases in the V2X application, including vehicle formation (Platooning). Platoning is the first step of realizing full-automatic driving, and is one of the most representative application cases of the fifth-generation mobile communication technology. In platonic, the distance and speed of travel between vehicles is controlled by a fully automatic control system through real-time updated motion data, using virtual strings to connect two adjacent vehicles, traveling on the road like a train. By adopting the driving mode of vehicle formation, the spacing distance between two vehicles is greatly reduced, the oil consumption is reduced, the number of drivers is reduced, and meanwhile, the utilization rate of vehicle roads can be increased, and the road capacity is increased. Likewise, Platoning also requires support from various V2X applications, consuming a significant amount of battery power to run the relevant applications.
The advent of FC provides a new idea for solving the above-mentioned problems. The FC can provide additional calculation and storage capacity of the roadside BS and V2X service distributed according to the geographical position, so that the vehicle can finish data transmission with less end-to-end delay, a driver can quickly react, safety accidents are reduced, and full-automatic driving is possible.
In summary, the present invention proposes a task offload solution for fleet vehicle assisted by FC to solve the dynamic offload problem between FC servers and PMs. In the scheme, the task is dynamically unloaded by using the Lyapunov optimization algorithm, so that the calculated amount queues of all tasks are stable, the execution energy consumption of the tasks is reduced, and the deadline of task execution is met. The Lyapunov optimization algorithm can dynamically judge the unloading place of the task based on the task parameter initiated at the current moment to obtain the optimal unloading decision of the task, so that all task calculation amount queues are updated under a stable condition, the flexibility is high, and the Lyapunov optimization algorithm has very important significance for minimizing the energy consumption of task execution.
Disclosure of Invention
In view of the above, the present invention provides a task offloading method for assisting vehicle formation by fog calculation.
In order to achieve the purpose, the invention provides the following technical scheme:
a task unloading method for assisting vehicle formation by FC through fog calculation comprises the following steps:
in each time interval, all vehicle formation members PMs in the vehicle formation generate tasks with a certain calculated amount, each task needs to judge the execution place of the task through unloading judgment, can be completed with minimum energy consumption, and simultaneously meets the deadline of the task;
in order to minimize the average task execution energy consumption, a Lyapunov optimization theory is introduced, prediction judgment is carried out according to task parameters requested by PMs at each moment, unloading places of tasks are dynamically adjusted, the stability of all calculation queues is guaranteed, and all tasks are completed within the deadline time.
Optionally, the PMs communicate with each other via vehicle-to-vehicle V2V, and all PMs can communicate directly with the base station BS via vehicle-to-infrastructure V2I;
the FC server and the BS are deployed in a coexistence mode, and a wired transmission connection mode is adopted between the BS and the FC server; when a task needs to be offloaded to the FC server, the task requester needs to send data to the BS first, and then the BS sends the data to the FC server.
Optionally, in the vehicle formation, there are M members in total, and all PMs in each time interval can generate a certain amount of tasks; definition of the invention Am(t) calculating the amount of the task initiated by the mth member of the vehicle formation at the time t, wherein Am(t) obeys a mean value of λmPoisson distribution of (a), i.e. E { A can be obtainedm(t)}＝λmAnd all members satisfy 0. ltoreq.Am(t)；A(t)＝{A1(t)，A2(t)，…，AM(t) }, in which A1(t)，A2(t)，…，AM(t) satisfies independent co-distribution.
Optionally, when the task is executed locally, the execution time of the task is local computation time, and the execution energy consumption of the task is the computation energy consumption of the task;
when the task is unloaded, the execution time of the task is the transmission time of the task plus the calculation time of the task, and the execution energy consumption of the task is the transmission energy consumption of the task plus the idle energy consumption;
defining the task execution time of the member m at the time t as
Figure BDA0002435053290000021
Wherein
Figure BDA0002435053290000022
Indicating that at the time t, the member m unloads the task to the kth task computation amount queue to consume the task execution time;
defining the task execution energy consumption of the member m at the time t as
Figure BDA0002435053290000023
Wherein
Figure BDA0002435053290000024
And the energy consumed by the member M to unload the task to the k-th task computation load queue at the time t is shown, wherein M is 1, 2, …, and M, k is 0, 1, …, M.
Optionally, the task has M +1 offloading decisions, that is, offloading to all PMs or FC servers;
defining the offloading decision vector of the task at time t as a (t) ═ am(t)|m∈{1，2，…，M}，am(t)∈{0，1，…，M}](ii) a Is used for obtaining an optimal task unloading decision vector a (t) to ensure that the whole user side completes tasks with the minimum average energy consumption to meet the requirement
Figure BDA0002435053290000031
With the proviso that
Figure BDA0002435053290000032
Wherein
Figure BDA0002435053290000033
The average total energy consumption of task unloading is shown, and the limiting condition shows that all task calculation amount queues are in a stable state.
Optionally, after the task is determined by the unloading decision, the task is unloaded to the corresponding task calculation queue;
defining the calculation amount of the task arriving at the kth queue at the time t as bk(t)，bkThe value of (t) is determined according to the unloading decision vector a (t) and the task calculation amount A (t) of member arrival;
defining the length of the k-th task computation amount queue at the time t as Qk(t); in obtaining bkAfter (t), all queues are updated, and the updating process is represented as Qk(t+1)＝max[Qk(t)-fk’0]+bk(t)；
Wherein f iskWhich represents the amount of task computation consumed by the compute node k in each time interval, i.e., the amount of computational resources provided.
Optionally, the optimal offloading strategy of the task is obtained by using a lyapunov optimization theory;
defining the Lyapunov function as
Figure BDA0002435053290000034
Then obtaining a transfer function
Figure BDA0002435053290000035
According to the Lyapunov optimization method, a penalty term delta (t) + VE { E (a (t)) | Q (t)) } needs to be minimized under the condition that the stability of all task calculated quantity queues is ensured, and then the upper bound of a penalty term function value is obtained according to a Lyapunov function and a transfer function
Figure BDA0002435053290000036
And finding the unloading strategy a (t) which minimizes the upper bound; and V represents the attention degree of the algorithm to the energy consumption, and is used for controlling the algorithm to carry out unloading decision of the task at each moment.
Optionally, the energy consumption is minimized to be an optimization target of the system;
obtaining an upper bound of a function value by the constructed Lyapunov transfer penalty term, and then minimizing a decision criterion by finding a proper unloading strategy, wherein an optimization equation is
Figure BDA0002435053290000037
And (3) obtaining a suboptimal solution by adopting a 1-opt local search algorithm, wherein a candidate vector of the suboptimal solution consists of all possible solutions of unary Hamming distances.
The invention has the beneficial effects that: the invention considers the size of the computing resources provided by the vehicle formation and the FC server computing resources at the same time, dynamically adjusts the unloading decision of each task at each moment according to the parameters of the task initiated by the task requester, and minimizes the energy consumption of the whole system through the obtained unloading strategy. Because the PMs and the FC server can provide computing resources for the task requester, in order to ensure that the obtained unloading strategy enables the energy consumption of the system to be minimum, the invention provides a simple and effective unloading decision scheme, the queue states of task computing quantities of all the PMs and the FC server and the deadline time of task execution are comprehensively considered, and the task unloading decision of all members in each time interval is dynamically adjusted through Lyapunov optimization. By the method, the stability of the task calculation amount queue and the energy consumption for executing the task can be simultaneously ensured, and the deadline time of the task is simultaneously met.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention may be realized and attained by the means of the instrumentalities and combinations particularly pointed out hereinafter.
Drawings
For the purposes of promoting a better understanding of the objects, aspects and advantages of the invention, reference will now be made to the following detailed description taken in conjunction with the accompanying drawings in which:
FIG. 1 is a task offload scenario diagram of the present invention;
FIG. 2 is a diagram of a task off-loading system of the present invention;
FIG. 3 is a diagram of a system task computation workload queue model of the present invention.
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention in a schematic way, and the features in the following embodiments and examples may be combined with each other without conflict.
Wherein the showings are for the purpose of illustrating the invention only and not for the purpose of limiting the same, and in which there is shown by way of illustration only and not in the drawings in which there is no intention to limit the invention thereto; to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of an actual product; it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.
The same or similar reference numerals in the drawings of the embodiments of the present invention correspond to the same or similar components; in the description of the present invention, it should be understood that if there is an orientation or positional relationship indicated by terms such as "upper", "lower", "left", "right", "front", "rear", etc., based on the orientation or positional relationship shown in the drawings, it is only for convenience of description and simplification of description, but it is not an indication or suggestion that the referred device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes, and are not to be construed as limiting the present invention, and the specific meaning of the terms may be understood by those skilled in the art according to specific situations.
The invention considers the task unloading scene of FC server and BS coexistence deployment, namely the FC server is deployed at the edge of an access network, the BS and the FC server adopt a wired connection communication mode, the vehicle and the BS adopt a V2I communication mode, and the vehicle adopt a V2V communication mode. In vehicle formation, PMs can directly communicate with each other, tasks are unloaded to an FC server and need to be forwarded through a BS, and then data are sent to the FC server in a wired transmission mode. Here, the present invention ignores the transmission time of the wired transmission.
All PMs in a vehicle fleet will initiate a certain calculated amount of tasks in each time interval, i.e. all PMs are task requesters. Discretizing the time for unloading the task, wherein t is used for representing the time of task execution, and then in each time, the PMs transmit the task to be unloaded to other PMs or FC servers for execution through V2V or V2I. The FC server is numbered 0, the vehicle formation has a vehicle head number of 1 according to the driving direction, and then the numbers are sequentially increased, and a vehicle tail number of M. Definition of the invention Bm，0Is the data transmission rate between member m and BS，Bm，1Is the data transmission rate between member m and member 1 (head), Bm，MFor the data transmission rate between member M and member M, the transmission rate between member M and the other members is defined similarly.
The Lyapunov optimization theory is mainly used for queuing research, so the invention adds the task calculation amount arriving at each moment into the queue. Definition of Q in the inventionk(t) queue length of task computation at time t for member k or FC server, Q (t) { Q { (t) }0(t)，Q1(t)，…，QM(t) } represents the set of task computation volume queue lengths resulting from the FC server and all PMs task offloads, k being 0, 1, …, M. And the calculated amount of the task arriving at each time member is defined as Am(t)，Am(t) obeying a mean value of λmThe Poisson distribution of E { A ] can be obtained by satisfying the independent distributionm(t)}＝λmAnd all members satisfy 0. ltoreq.Am(t)，A(t)＝{A1(t)，A2(t)，…，AM(t)}。
The invention defines the execution time of the task as
Figure BDA0002435053290000051
Wherein
Figure BDA0002435053290000052
At time t, the task execution time it takes for member m to unload the task to the kth queue (in the unload case, task execution time includes task transmission time and task computation time, when the task is computed locally,
Figure BDA0002435053290000053
representing only the time when the task is executing the task locally). According to the formula:
Figure BDA0002435053290000054
the execution time of the task in each queue can be obtained. Wherein Dm(t) denotes time t, at which member m arrives at Am(t) this meterThe calculated amount corresponds to the size of the generated data amount, D (t) { D ═ D1(t)，D2(t)，…，DM(t) }, in which Dm(t) independent and obey uniform distribution. f. ofkRepresenting the amount of task computation consumed by the kth queue at each time instant, i.e. the computational resources provided are fk。
The invention defines the execution energy consumption of the task as
Figure BDA0002435053290000061
Wherein
Figure BDA0002435053290000062
Representing the energy consumed by member m to unload the task to the kth queue at time t. According to a calculation formula
Figure BDA0002435053290000063
Wherein p istrPower p representing vehicle data transmissioncAnd the power of task calculation is represented, and the task execution energy consumption of the task in each queue can be obtained. The computing mode considering the energy consumption is considered from the perspective of the task requester, only the computing energy consumption executed locally is considered, and the computing energy consumption generated by unloading the task to other vehicles and the FC server is not considered.
The invention defines the unloading decision vector of the task as a (t) ═ am(t)|m∈{1，2，…，M}，am(t)∈{0，1，…，M}]Wherein a ism(t) ═ 0 denotes that at time t, member m offloads the task to FC server execution, am(t)＝1，am(t)＝2，…，am(t) ═ M, meaning that the task is offloaded to PMs for execution, and amAnd m represents that the task is executed locally. The final purpose of the invention is to obtain the optimal unloading decision vector a (t), so that the energy consumption of the task during the execution period is minimum, the task calculation amount queue is stable, and the deadline of the task execution is met. Definition of the invention am，k(t) unload decision for member m to unload to queue k at time t, am，k(t) — 1 denotes task-off-load, am，k(t) — 0 means that the task is not unloaded, am，k(t) 1 or 0. From the above, assume am(t) k, can give am，k(t) 1, so the unload decision of vehicle formation member m can be written as
Figure BDA0002435053290000064
Wherein
Figure BDA0002435053290000065
To ensure that a task is processed at only one compute node.
Then, according to the calculated task execution time and energy consumption and the system optimization target, the following optimization equation is obtained:
Figure BDA0002435053290000066
Figure BDA0002435053290000067
Figure BDA0002435053290000068
(c3)f0＞fm＞0，Bm，k(t)＞0
wherein, V is a weight control parameter, which represents the attention degree for minimizing energy consumption. The constraint (c1) is to ensure that the task completes within the deadline, including the task transmission time, the task computation time, and the waiting time of the task in the queue. The constraint (c2) indicates the unload decision point for a task, and a task can only be unloaded to one place. The constraints (c3) are computational resources and network constraints, and the computational resources provided by the FC servers are always larger than the computational resources provided by the PMs. In order to minimize the value of the optimization equation and obtain the unloading strategy a (t), the invention adopts a 1-opt local search algorithm to obtain the optimal solution.
After the optimal unloading strategy is obtained, all queues need to be updated. The invention defines a dynamic unloading system as Qk(t+1)＝max[Qk(t)-fk，0]+bk(t) wherein Qk(t) represents the system state of the kth task computation volume queue at time t, i.e. the queue length. bk(t) represents the amount of task computation for the k-th queue reached at time t. The calculated quantity A generated at time t from PMsm(t) and the offloading strategy yields bk(t)，bkThe expression of (t) is:
Figure BDA0002435053290000071
examples
Consider a total of M PMs in a formation of vehicles, with number 1 representing the number of the first vehicle, i.e., the head, of the formation; then the number of the serial numbers is increased in sequence, and M represents the serial number of the vehicles at the tail of the formation; the FC server and the BS are deployed in a coexistence mode, a wired connection communication mode is adopted between the FC server and the BS, and the vehicle formation is always within the coverage range of the base station. The vehicle-to-vehicle communication employs V2V, and the vehicle-to-BS communication employs V2I, as shown in fig. 1. If the task is offloaded to PMs, the data can be sent directly to the destination. If the task needs to be unloaded to the FC server, the information needs to be sent to the BS, and then the BS forwards the information to the FC server.
And a specific unloading system, as shown in figure 2. The base station is numbered 0, and assuming that the member numbered M is the requester of the task, M is 1, 2, …, M, Bm，0Is the data transmission rate between member m and the base station, Bm，1Is the data transmission rate between member m and member 1, Bm，MBeing the data transfer rate between member M and member M, the data transfer rate between member M and the other PMs is defined in the same manner as described above, since not all PMs are drawn.
The system has M +1 task calculation amount queues, namely M vehicle formation members and an FC server queue, and the state of the queue is updated in real time along with the change of the time t. Each member in the vehicle formation is a requester of a task, and a certain calculated amount of tasks are generated at each moment, and then are unloaded to a corresponding task calculated amount queue through judgment of an unloading decision, as shown in fig. 3. bk(t) watchIndicating the amount of task computation that reaches the kth queue after the task offload decision is made at time t, e.g., if members 1 and 2 decide to offload tasks to the FC server at time t, then b0(t)＝A1(t)+A2(t)。f0Representing computing resources that the FC server can provide, f1，…，fMIndicating the computational resources that PMs can provide and also indicating the task computation that the corresponding task computation queue can consume at each time.
Each task is unloaded to different places, and the execution time and the energy consumption of the task are different. At time t, the task execution time of member m is
Figure BDA0002435053290000072
Wherein
Figure BDA0002435053290000073
At the time t, the task execution time consumed by the member m task in the kth queue for processing (when the task is unloaded to other computing nodes for processing, the execution time comprises the task transmission time and the task computing time; when the task is processed locally, the execution time is only the local computing time),
Figure BDA0002435053290000074
the task execution energy consumption of the member m is
Figure BDA0002435053290000075
Wherein
Figure BDA0002435053290000076
Indicating the energy consumption consumed by the task of the member m to process in the kth queue at the moment t,
Figure BDA0002435053290000081
the ultimate goal of the invention is to minimize the average execution energy consumption of the tasks
Figure BDA0002435053290000082
With the proviso that
Figure BDA0002435053290000083
And an unloading strategy is obtained in each time interval, so that the energy consumption for executing the task is reduced to the greatest extent possible. Wherein
Figure BDA0002435053290000084
Represents the sign of a mathematical operation, i.e. the averaging,
Figure BDA0002435053290000085
the average energy consumption of task unloading is shown, and the limiting condition shows that all queues are in a stable condition.
Each member generates a certain amount of computation task at each time interval, and the place where the task is computed needs to be determined by the unloading decision, so the computation amount of the task arriving at each queue at each time is determined by the unloading decision. bk(t) represents the amount of computation of the task arriving at the kth queue at time t, and then the amount of computation A arriving at time t based on PMsm(t) can obtain:
Figure BDA0002435053290000086
the present invention defines T (a) (T) ═ T1(t)，T2(t)，…，TM(t)]，Tm(t) represents the execution time of the task after the task of the member m carries out the unloading decision judgment at the time t, and the specific execution time is
Figure BDA0002435053290000087
The energy consumption generated by the local execution of the task and the transmission energy consumption and the idle energy consumption generated when the task is unloaded are the total energy consumption of the task. Definition of the invention
Figure BDA0002435053290000088
Represents the sum of the energy consumption of the whole system for executing tasks with a (t) unloading decisions at the time t.
The invention defines a dynamic unloading system: qk(t+1)＝max[Qk(t)-fk，0]+bk(t) of (d). Wherein Qk(t) represents the system state of the kth task computation volume queue at time t, i.e., the queue length. Therefore, Q can be knownkThe larger (t) the longer the queue latency. Based on the Lyapunov optimization theory, the invention defines the Lyapunov function as
Figure BDA0002435053290000089
Then Q is addedkSubstitution of (t +1) into L (t +1) -L (t) yields:
Figure BDA00024350532900000810
wherein the derivation of the function refers to a theorem: let W, U, μ and A be non-negative positive real numbers, and W ═ max [ U- μ, 0]+ A, W can be obtained2≤U2+A2+μ2-2U(μ-A)。
The invention defines the Lyapunov transfer function as:
Figure BDA00024350532900000811
according to the lyapunov optimization theory, it is required to minimize penalty items, i.e., minimize energy consumption, while ensuring queue stability. Under the system model of the invention, the Lyapunov penalty item is the energy consumption of the user to execute the task, namely E { E (a (t)) | Q (t)) }, and the Lyapunov transfer penalty item can be obtained as follows:
Δ(t)+VE{E(a(t))|Q(t)}
wherein, V is a weight control parameter, which represents the attention degree for minimizing energy consumption. In other words, V can be considered a threshold for the system state in which the control algorithm makes the unloading decision.
In order to minimize the overall system energy consumption, an upper bound of the lyapunov transfer penalty term needs to be found, and the upper bound is minimized by finding a suitable unloading decision a (t), wherein the upper bound of the lyapunov transfer penalty term is:
Figure BDA0002435053290000091
the right side of the equation is an upper bound function, and then the fixed value items are ignored, so that the following optimization equation can be obtained according to the optimization target of the system:
Figure BDA0002435053290000092
Figure BDA0002435053290000093
Figure BDA0002435053290000094
(c3)f0＞fm＞0，Bm，k(t)＞0
in order to obtain the minimum value of the optimization equation, the invention adopts a 1-opt local search algorithm to obtain the suboptimal solution, the candidate vector of the suboptimal solution consists of all possible solutions of unary Hamming distance, the algorithm has lower computational complexity, and the running time of the algorithm is O (| d | D | Y3). The obtained unloading strategy needs to judge whether the execution time of the task meets the deadline time of the task, otherwise, the unloading decision is abandoned, and another suboptimal solution is searched. The method comprises the following specific steps:
1) initializing task calculation quantity queue state and time;
2) establishing a task unloading request of PMs, and acquiring the calculated amount of the task arriving at the current time, wherein A (t) { A }1(t)，A2(t)，…，AM(t), data size D (t) ═ D1(t)，D2(t)，…，DM(t) }, and the network transmission rate;
3) within each time interval t, calculating the execution time of unloading the task to each place or executing locally according to the acquired data
Figure BDA0002435053290000095
Energy consumption associated with task execution
Figure BDA0002435053290000096
4) Substituting the obtained calculation result into the optimization equation of the system in each time interval t
Figure BDA0002435053290000097
Setting the value of a parameter V, and obtaining an optimal task unloading strategy a (t) by using a 1-opt local search algorithm;
5) calculating task calculation amount b of arriving queue by the obtained task unloading strategy a (t)k(t) use the formula Qk(t+1)＝max[Qk(t)-fk，0]+bk(t) updating the calculated amount queues of all tasks;
in summary, the invention mainly aims to provide a task unloading method based on lyapunov optimization under the condition that an FC server assists in a cellular internet of vehicles, so that the energy consumption for task execution is reduced, and the task deadline is met. According to the invention, reasonable control parameters V are set for the constructed Lyapunov optimization function, so that the energy consumption is well controlled, the stability of a task calculation amount queue can be ensured by the algorithm, and a task unloading strategy can be dynamically adjusted according to the change of the task data parameters at the current moment. In cellular internet of vehicles queuing systems, the algorithm used herein is one of the effective and feasible ways to achieve dynamic offloading of tasks.
Finally, the above embodiments are only intended to illustrate the technical solutions of the present invention and not to limit the present invention, and although the present invention has been described in detail with reference to the preferred embodiments, it will be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions, and all of them should be covered by the claims of the present invention.

Claims (8)

1. A task unloading method for assisting vehicle formation by FC through fog calculation is characterized by comprising the following steps: the method comprises the following steps:
in each time interval, all vehicle formation members PMs in the vehicle formation generate tasks with a certain calculated amount, each task needs to judge the execution place of the task through unloading judgment, can be completed with minimum energy consumption, and simultaneously meets the deadline of the task;
in order to minimize the average task execution energy consumption, a Lyapunov optimization theory is introduced, prediction judgment is carried out according to task parameters requested by PMs at each moment, unloading places of tasks are dynamically adjusted, the stability of all calculation queues is guaranteed, and all tasks are completed within the deadline time.
2. The method of claim 1, wherein the task offloading method is implemented by a fog computer to assist vehicle formation: the PMs communicate with each other via vehicle-to-vehicle V2V, all PMs can communicate directly with the base station BS via vehicle-to-infrastructure V2I;
the FC server and the BS are deployed in a coexistence mode, and a wired transmission connection mode is adopted between the BS and the FC server; when a task needs to be offloaded to the FC server, the task requester needs to send data to the BS first, and then the BS sends the data to the FC server.
3. The method of claim 2, wherein the task offloading method is implemented by a fog computer to assist vehicle formation: the vehicle formation comprises M members in total, and all PMs in each time interval can generate tasks with a certain calculated amount; definition of the invention Am(t) calculating the amount of the task initiated by the mth member of the vehicle formation at the time t, wherein Am(t) obeys a mean value of λmPoisson distribution of (i) can be obtained
Figure FDA0002435053280000015
And all the members satisfy 0. ltoreq. Am(t)；A(t)＝{A1(t)，A2(t)，…，AM(t) }, in which A1(t)，A2(t)，…，AM(t) satisfies independent co-distribution.
4. A method of task offloading by fog calculation assisted vehicle formation according to claim 2 or 3, characterized by:
when the task is executed locally, the execution time of the task is local calculation time, and the execution energy consumption of the task is the calculation energy consumption of the task;
when the task is unloaded, the execution time of the task is the transmission time of the task plus the calculation time of the task, and the execution energy consumption of the task is the transmission energy consumption of the task plus the idle energy consumption;
defining the task execution time of the member m at the time t as
Figure FDA0002435053280000011
Wherein
Figure FDA0002435053280000012
Indicating that at the time t, the member m unloads the task to the kth task computation amount queue to consume the task execution time;
defining the task execution energy consumption of the member m at the time t as
Figure FDA0002435053280000013
Wherein
Figure FDA0002435053280000014
And the energy consumed by the member M to unload the task to the k-th task computation load queue at the time t is shown, wherein M is 1, 2, …, and M, k is 0, 1, …, M.
5. The method of claim 4, wherein the task offloading method is implemented by a fog computer to assist vehicle formation: the task has M +1 unloading decisions, namely unloading to all PMs or FC servers;
defining the offloading decision vector of the task at time t as a (t) ═ am(t)|m∈{1，2，…，M}，am(t)∈{0，1，…，M}](ii) a Is used for obtaining an optimal task unloading decision vector a (t) to ensure that the whole user side completes tasks with the minimum average energy consumption to meet the requirement
Figure FDA0002435053280000021
With the proviso that
Figure FDA0002435053280000022
Wherein
Figure FDA0002435053280000023
The average total energy consumption of task unloading is shown, and the limiting condition shows that all task calculation amount queues are in a stable state.
6. The method of claim 4, wherein the task offloading method is implemented by a fog computer to assist vehicle formation: after the task is judged by the unloading decision, unloading the task to a corresponding task calculation queue;
defining the calculation amount of the task arriving at the kth queue at the time t as bk(t)，bkThe value of (t) is determined according to the unloading decision vector a (t) and the task calculation amount A (t) of member arrival;
defining the length of the k-th task computation amount queue at the time t as Qk(t); in obtaining bkAfter (t), all queues are updated, and the updating process is represented as Qk(t+1)＝max[Qk(t)-fk’0]+bk(t)；
Wherein f iskWhich represents the amount of task computation consumed by the compute node k in each time interval, i.e., the amount of computational resources provided.
7. The method of claim 6, wherein the task offloading method is implemented by a fog computer to assist vehicle formation: the optimal unloading strategy of the task is obtained by using the Lyapunov optimization theory;
defining the Lyapunov function as
Figure FDA0002435053280000024
Then obtaining a transfer function
Figure FDA0002435053280000025
According to Li yaThe Ponuov optimization method needs to minimize a penalty term delta (t) + VE { E (a (t)) | Q (t)) } under the condition of ensuring the stability of all task calculated quantity queues, and then obtains the upper bound of a penalty term function value according to a Lyapunov function and a transfer function
Figure FDA0002435053280000026
Figure FDA0002435053280000027
And finding the unloading strategy a (t) which minimizes the upper bound; and V represents the attention degree of the algorithm to the energy consumption, and is used for controlling the algorithm to carry out unloading decision of the task at each moment.
8. The method of claim 7, wherein the task offloading method is implemented by a fog computer to assist vehicle formation: the minimization of the energy consumption is an optimization target of the system;
obtaining an upper bound of a function value by the constructed Lyapunov transfer penalty term, and then minimizing a decision criterion by finding a proper unloading strategy, wherein an optimization equation is
Figure FDA0002435053280000028
And (3) obtaining a suboptimal solution by adopting a 1-opt local search algorithm, wherein a candidate vector of the suboptimal solution consists of all possible solutions of unary Hamming distances.
CN202010249746.1A 2020-04-01 2020-04-01 Task unloading method for assisting vehicle formation through fog calculation Pending CN111479242A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010249746.1A CN111479242A (en) 2020-04-01 2020-04-01 Task unloading method for assisting vehicle formation through fog calculation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010249746.1A CN111479242A (en) 2020-04-01 2020-04-01 Task unloading method for assisting vehicle formation through fog calculation

Publications (1)

Publication Number Publication Date
CN111479242A true CN111479242A (en) 2020-07-31

Family

ID=71749393

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010249746.1A Pending CN111479242A (en) 2020-04-01 2020-04-01 Task unloading method for assisting vehicle formation through fog calculation

Country Status (1)

Country Link
CN (1) CN111479242A (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109656703A (en) * 2018-12-19 2019-04-19 重庆邮电大学 A kind of mobile edge calculations auxiliary vehicle task discharging method
CN109829332A (en) * 2019-01-03 2019-05-31 武汉理工大学 A kind of combined calculation discharging method and device based on energy collection technology
CN110234127A (en) * 2019-06-11 2019-09-13 重庆邮电大学 A kind of mist network task discharging method based on SDN
CN110377353A (en) * 2019-05-21 2019-10-25 湖南大学 Calculating task uninstalling system and method

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109656703A (en) * 2018-12-19 2019-04-19 重庆邮电大学 A kind of mobile edge calculations auxiliary vehicle task discharging method
CN109829332A (en) * 2019-01-03 2019-05-31 武汉理工大学 A kind of combined calculation discharging method and device based on energy collection technology
CN110377353A (en) * 2019-05-21 2019-10-25 湖南大学 Calculating task uninstalling system and method
CN110234127A (en) * 2019-06-11 2019-09-13 重庆邮电大学 A kind of mist network task discharging method based on SDN

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
YUYU HU: "Task Offloading Based on Lyapunov Optimization for MEC-assisted Platooning", 《2019 11TH INTERNATIONAL CONFERENCE ON WIRELESS COMMUNICATIONS AND SIGNAL PROCESSING》 *

Similar Documents

Publication Publication Date Title
CN109391681B (en) MEC-based V2X mobility prediction and content caching offloading scheme
CN111918311B (en) Vehicle networking task unloading and resource allocation method based on 5G mobile edge computing
CN109656703B (en) Method for assisting vehicle task unloading through mobile edge calculation
CN109067842A (en) Calculating task discharging method towards car networking
Chen et al. Efficiency and fairness oriented dynamic task offloading in internet of vehicles
CN112367640B (en) V2V mode multi-task unloading method and system based on mobile edge calculation
Kazmi et al. A novel deep reinforcement learning-based approach for task-offloading in vehicular networks
CN112929849A (en) Reliable vehicle-mounted edge calculation unloading method based on reinforcement learning
Liu et al. Energy-efficient joint computation offloading and resource allocation strategy for isac-aided 6g v2x networks
CN113344255A (en) Vehicle-mounted network application data transmission and charging optimization method based on mobile edge calculation and block chain
CN114980029A (en) Unloading method based on task relevance in Internet of vehicles
Wu et al. Delay constrained hybrid task offloading of internet of vehicle: a deep reinforcement learning method
CN111479242A (en) Task unloading method for assisting vehicle formation through fog calculation
CN116204315A (en) Track-based dynamic task unloading method for vehicle resource pool in Internet of vehicles
CN113709249B (en) Safe balanced unloading method and system for driving assisting service
CN115002123A (en) Fast adaptive task unloading system and method based on mobile edge calculation
CN105933855B (en) Car networking data download method based on user satisfaction
CN115103313A (en) Intelligent road task cooperative processing method and system based on position prediction
CN113448707A (en) Online batch parallel task scheduling method in edge computing
CN115065683B (en) Vehicle edge network task allocation and unloading method based on vehicle clustering
Li et al. Mobility and dependency-aware task offloading for intelligent assisted driving in vehicular edge computing networks
Toyoda et al. A dynamic task offloading method with centralized controller to improve task success rate
CN111625360B (en) Vehicle big data edge calculation unloading method based on plane separation
Pan et al. A Route Planning for Autonomous Vehicle in 5G and Edge Computing Environment
Shi et al. RSU-assisted Proactive Perception and Edge Computing for Autonomous Driving

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20200731

RJ01 Rejection of invention patent application after publication