CN117042051B - Task unloading strategy generation method, system, equipment and medium in Internet of vehicles - Google Patents
Task unloading strategy generation method, system, equipment and medium in Internet of vehicles Download PDFInfo
- Publication number
- CN117042051B CN117042051B CN202311096130.5A CN202311096130A CN117042051B CN 117042051 B CN117042051 B CN 117042051B CN 202311096130 A CN202311096130 A CN 202311096130A CN 117042051 B CN117042051 B CN 117042051B
- Authority
- CN
- China
- Prior art keywords
- task
- processing
- road side
- representing
- side unit
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 64
- 238000012545 processing Methods 0.000 claims abstract description 109
- 238000005457 optimization Methods 0.000 claims abstract description 57
- 238000005265 energy consumption Methods 0.000 claims abstract description 44
- 238000004364 calculation method Methods 0.000 claims abstract description 17
- 230000007774 longterm Effects 0.000 claims abstract description 7
- 230000008569 process Effects 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 9
- 230000005540 biological transmission Effects 0.000 claims description 8
- 239000013598 vector Substances 0.000 claims description 7
- 238000000342 Monte Carlo simulation Methods 0.000 claims description 5
- 230000006870 function Effects 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 claims description 3
- 230000006855 networking Effects 0.000 claims description 3
- 241000764238 Isis Species 0.000 claims description 2
- 238000001914 filtration Methods 0.000 claims description 2
- 230000003139 buffering effect Effects 0.000 claims 1
- 238000004891 communication Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000003190 augmentative effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W28/00—Network traffic management; Network resource management
- H04W28/02—Traffic management, e.g. flow control or congestion control
- H04W28/08—Load balancing or load distribution
- H04W28/09—Management thereof
- H04W28/0925—Management thereof using policies
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W28/00—Network traffic management; Network resource management
- H04W28/02—Traffic management, e.g. flow control or congestion control
- H04W28/10—Flow control between communication endpoints
- H04W28/14—Flow control between communication endpoints using intermediate storage
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W28/00—Network traffic management; Network resource management
- H04W28/16—Central resource management; Negotiation of resources or communication parameters, e.g. negotiating bandwidth or QoS [Quality of Service]
- H04W28/18—Negotiating wireless communication parameters
- H04W28/20—Negotiating bandwidth
-
- 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]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Quality & Reliability (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a method, a system, equipment and a medium for generating a task unloading strategy in the Internet of vehicles, and relates to the technical field of Internet of vehicles resource management. The method comprises the following steps: acquiring a vehicle equipment task; according to the task processing network architecture, system performance calculation of different processing network layers is carried out on the vehicle equipment tasks to obtain multi-layer system performance parameters; the task processing network architecture is constructed according to vehicle equipment, road side units and remote clouds; the system performance parameters include: time delay, energy consumption and cost; constructing an optimization problem by taking minimum time delay as a target and combining energy consumption and cost, and solving the optimization problem to obtain a task unloading strategy in the Internet of vehicles; the optimization problem is an optimization problem based on bandwidth allocation and cache allocation. The invention can solve the problems of bandwidth allocation and buffer allocation in task unloading through a long-term online allocation strategy, and can realize the guarantee of the service quality of the system.
Description
Technical Field
The invention relates to the technical field of internet of vehicles resource management, in particular to a method, a system, equipment and a medium for generating a task unloading strategy in the internet of vehicles.
Background
Under the technologies of big data, artificial intelligence, etc., the internet of vehicles (Internet of Vehicles, ioV) has become a promising paradigm for supporting high capacity and large scale connections. In recent years, the maturation of 5g technology and large-scale commercial use have led to a number of intelligent vehicle applications in internet of vehicles, such as autopilot, augmented reality and collision detection. However, these high-resource and low-latency demand computing applications provide people with living convenience, while presenting a significant challenge to vehicle terminals with inadequate computing resources.
Cloud computing can be used as a key technology to solve the problem of insufficient terminal resources. Cloud computing allows a terminal device to offload tasks to the cloud to use rich computing resources of the cloud, but tasks in a car networking scene often have the characteristic of high delay sensitivity, and uploading to the cloud can bring unavoidable high transmission delay, and meanwhile the terminal device also needs to pay high computing resource use cost to the cloud, so that a conventional cloud computing model cannot be applicable in the car networking scene.
To address the latency and cost issues of cloud computing, mobile edge computing (Mobile Edge Computing, MEC) was introduced into the internet of vehicles, known as vehicle edge computing (Vehicular Edge Computing, VEC), which techniques sink computing resources onto edge nodes close to end users. The edge processor is placed in a Road Side Unit (RSU) on the Road Side, and tasks generated by the terminal device can run on the edge processor, which not only can relieve the resource deficiency of the vehicle terminal, but also can reduce the task transmission delay and ensure the reliability of the service quality (Quality of Service, qoS). Since the capability of the edge server to process tasks is smaller than that of the cloud server, if too many tasks are offloaded to the edge server, a large computing pressure is brought to the edge server, and then the processing time of the tasks on the edge server may be longer than that of the tasks on the cloud server. To further improve the quality of service, the edge server may offload tasks to the cloud server, thereby alleviating the computing pressure of the edge server, which is referred to as a cloud-edge collaborative policy. The strategy fully utilizes the advantages of cloud computing and edge computing, and many researches under the strategy aim to solve the problem of task offloading in the internet of vehicles. However, only one or two types of tasks, such as latency-intensive tasks and computation-intensive tasks, are considered in the prior art, but in a real scenario, a vehicle may generate multiple types of tasks, and it is not accurate to simply classify the vehicle-generated tasks into one or two types. In view of the vehicle communicating wirelessly with the drive test unit, a reasonable bandwidth allocation is necessary in order to avoid competing communication resources from task to task at the time of the multitasking upload. The service cache serves as a database necessary for task processing and if a task is to be processed on the RSU, the RSU must cache the database necessary for the task processing. Unlike the huge storage capacity of the cloud, the limited cache capacity of the RSU only allows a small amount of data sets to be cached, and as the cache can significantly affect the performance of edge calculation, how to design a proper cache allocation strategy and ensure the service quality of the vehicle-mounted user is important.
Disclosure of Invention
The invention aims to provide a method, a system, equipment and a medium for generating a task unloading strategy in the Internet of vehicles, which can solve the problems of bandwidth allocation and buffer allocation in task unloading through a long-term online allocation strategy and realize the guarantee of the service quality of the system.
In order to achieve the above object, the present invention provides the following solutions:
a task unloading strategy generation method in the Internet of vehicles comprises the following steps:
acquiring a vehicle equipment task;
according to the task processing network architecture, system performance calculation of different processing network layers is carried out on the vehicle equipment tasks to obtain multi-layer system performance parameters; the task processing network architecture is constructed according to vehicle equipment, road side units and remote clouds; the system performance parameters include: time delay, energy consumption and cost;
constructing an optimization problem by taking minimum time delay as a target and combining energy consumption and cost, and solving the optimization problem to obtain a task unloading strategy in the Internet of vehicles; the optimization problem is an optimization problem based on bandwidth allocation and cache allocation.
Optionally, according to the task processing network architecture, performing system performance calculation of different processing network layers on the vehicle equipment task to obtain a multi-layer system performance parameter, which specifically includes:
in the t time slot, calculating the vehicle equipment tasks by utilizing a system performance model, wherein the time delay, the energy consumption and the cost are respectively generated during the processing of a road side unit and a remote cloud;
and determining the performance parameters of the multi-layer system according to the calculation results of the road side unit and the remote cloud.
Optionally, the system performance model includes:
wherein,representing the average processing time delay of a task; />An arrival rate indicating a task to be processed at the roadside unit; />Representing the total task arrival rate; />Representing the average processing time delay of the task road side unit n; />The arrival rate of the task uploaded by the road side unit n is represented; />Representing the average time delay of a task uploaded by a road side unit; />Representing average processing delay of a task in a remote cloud; t (T) r Representing propagation delay between the roadside unit and the remote cloud; />The task switching delay of the road side unit is represented;
wherein,representing the average processing energy consumption of a task; />Representing the average processing energy consumption of the task road side unit n; />Representing average processing energy consumption of a task in a remote cloud;
wherein,representing the average processing cost of a task; />Representing the cost of a roadside unit to process a task;representing average cache update costs for the roadside units; />Representing the cost of remote cloud processing of a task.
Optionally, the minimum time delay is taken as a target, and an optimization problem is constructed by combining energy consumption and cost, and the optimization problem is solved to obtain a task unloading strategy in the internet of vehicles, which specifically comprises:
converting the optimization problem into a bandwidth allocation sub-problem and a buffer allocation sub-problem by using a Lyapunov optimization framework;
solving a task unloading strategy of the bandwidth allocation sub-problem by using an alternate direction multiplication method;
solving a task unloading strategy of the buffer allocation sub-problem by using a Monte Carlo simulation method;
and determining a task unloading strategy in the Internet of vehicles according to the task unloading strategy of the bandwidth allocation sub-problem and the task unloading strategy of the buffer allocation sub-problem.
Optionally, the optimization problem is used for filtering the cache task according to time delay, energy consumption and cost, when the road side unit caches one task, the task is processed in the road side unit, and when the road side unit selects the task which is not cached, the task is uploaded to the remote cloud for processing through the wired network.
The invention also provides a task unloading strategy generation system in the Internet of vehicles, which comprises the following steps:
the task acquisition module is used for acquiring a task of the vehicle equipment;
the performance calculation module is used for calculating the system performance of different processing network layers for the vehicle equipment tasks according to the task processing network architecture to obtain multi-layer system performance parameters; the task processing network architecture is constructed according to vehicle equipment, road side units and remote clouds; the system performance parameters include: time delay, energy consumption and cost;
the optimization solving module is used for constructing an optimization problem by taking the minimum time delay as a target and combining energy consumption and cost, and solving the optimization problem to obtain a task unloading strategy in the Internet of vehicles; the optimization problem is an optimization problem based on bandwidth allocation and cache allocation.
The invention also provides electronic equipment, which comprises a memory and a processor, wherein the memory is used for storing a computer program, and the processor runs the computer program to enable the electronic equipment to execute the task unloading strategy generation method in the Internet of vehicles.
The invention also provides a computer readable storage medium storing a computer program which when executed by a processor implements the task offloading policy generation method in the internet of vehicles as described above.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention discloses a method, a system, equipment and a medium for generating a task unloading strategy in the Internet of vehicles, wherein the method comprises the steps of acquiring a task of vehicle equipment; according to the task processing network architecture, system performance calculation of different processing network layers is carried out on the vehicle equipment tasks to obtain multi-layer system performance parameters; the task processing network architecture is constructed according to vehicle equipment, road side units and remote clouds; the system performance parameters include: time delay, energy consumption and cost; constructing an optimization problem by taking minimum time delay as a target and combining energy consumption and cost, and solving the optimization problem to obtain a task unloading strategy in the Internet of vehicles; the optimization problem is an optimization problem based on bandwidth allocation and cache allocation. The invention can solve the problems of bandwidth allocation and buffer allocation in task unloading through a long-term online allocation strategy, and can realize the guarantee of the service quality of the system.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that are needed in the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a task offloading policy generation method in the Internet of vehicles of the present invention;
fig. 2 is a schematic diagram of a task processing network architecture in the present embodiment.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention aims to provide a method, a system, equipment and a medium for generating a task unloading strategy in the Internet of vehicles, which can solve the problems of bandwidth allocation and buffer allocation in task unloading through a long-term online allocation strategy and realize the guarantee of the service quality of the system.
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.
As shown in fig. 1, the present invention provides a method for generating a task offloading policy in the internet of vehicles, including:
step 100: acquiring a vehicle equipment task;
step 200: according to the task processing network architecture, system performance calculation of different processing network layers is carried out on the vehicle equipment tasks to obtain multi-layer system performance parameters; the task processing network architecture is constructed according to vehicle equipment, road side units and remote clouds; the system performance parameters include: time delay, energy consumption and cost; the method specifically comprises the following steps:
in the t time slot, calculating the vehicle equipment tasks by utilizing a system performance model, wherein the time delay, the energy consumption and the cost are respectively generated during the processing of a road side unit and a remote cloud; and determining the performance parameters of the multi-layer system according to the calculation results of the road side unit and the remote cloud.
Wherein the system performance model comprises:
in the method, in the process of the invention,representing the average processing time delay of a task; />An arrival rate indicating a task to be processed at the roadside unit; />Representing the total task arrival rate; />Representing the average processing time delay of the task road side unit n; />The arrival rate of the task uploaded by the road side unit n is represented; />Representing the average time delay of a task uploaded by a road side unit; />Representing average processing delay of a task in a remote cloud; t (T) r Representing propagation delay between the roadside unit and the remote cloud; />The task switching delay of the road side unit is represented;
in the method, in the process of the invention,representing the average processing energy consumption of a task; />Representing the average processing energy consumption of the task road side unit n; />Representing average processing energy consumption of a task in a remote cloud;
in the method, in the process of the invention,representing the average processing cost of a task; />Representing the cost of a roadside unit to process a task;representing average cache update costs for the roadside units; />Representing the cost of remote cloud processing of a task.
Step 300: constructing an optimization problem by taking minimum time delay as a target and combining energy consumption and cost, and solving the optimization problem to obtain a task unloading strategy in the Internet of vehicles; the optimization problem is an optimization problem based on bandwidth allocation and cache allocation. The method specifically comprises the following steps:
converting the optimization problem into a bandwidth allocation sub-problem and a buffer allocation sub-problem by using a Lyapunov optimization framework; solving a task unloading strategy of the bandwidth allocation sub-problem by using an alternate direction multiplication method; solving a task unloading strategy of the buffer allocation sub-problem by using a Monte Carlo simulation method; and determining a task unloading strategy in the Internet of vehicles according to the task unloading strategy of the bandwidth allocation sub-problem and the task unloading strategy of the buffer allocation sub-problem.
Based on the above technical solution, the following embodiments are provided, and the policy mainly includes the following contents:
the internet of vehicles architecture is comprised of vehicle equipment, road Side Units (RSUs) and remote clouds. Both the roadside units and the remote cloud are equipped with computing servers. The vehicle equipment may generate various types of tasks and transmit to the roadside units via the wireless network and receive processing.
In the MEC system, the task offloading strategy is as follows:
(1) Considering vehicle devices based on actual scenarios may generate multiple types of tasks, and considering the unavoidable competing bandwidth resources between different types of tasks, reasonable bandwidth allocation is required for the vehicle devices during task uploading.
(2) Because the storage capacity of the road side unit is limited, the road side unit cannot cache the database necessary for running all types of tasks, so that the task of which type is cached needs to be comprehensively considered according to time delay, energy consumption and cost. If the road side unit caches the task, the task is processed by the road side unit; otherwise, uploading the cloud processing result to a remote cloud through a wired network.
And step 1, solving time delay, energy consumption and cost generated by processing a task at a Road Side Unit (RSU) in a t-th time slot.
Consider that vehicle-to-Road Side Unit (RSU) communications employ orthogonal frequency division multiplexing techniques and that the communication bandwidths of each road side unit do not overlap. The number of vehicles covered by the roadside unit n at the time of the time slot t can be calculated asThe road side unit will divide the bandwidth evenly to each vehicle, so the bandwidth divided by each vehicle can be calculated as: />Where B is the total bandwidth of the roadside unit. Defining bandwidth allocation vectorsThe upload rate of the vehicle k upload task S (S e {1,2,., S }) at time slot t is:
wherein the method comprises the steps ofIs the bandwidth of the vehicle k sending task s, +.>Is the transmission power of the transmission task s of vehicle k, < >>Is the channel gain, sigma, of vehicle k and road side unit 2 Is gaussian white noise. Considering that the bandwidth allocated to the vehicle must be used entirely for the uploading task, the month constraint: />The uploading delay of the task s is assumed to follow an exponential distribution, so that various types of task offloading processes can be regarded as a plurality of M/1 queues. The time for the vehicle k to upload the task s on average at time slot t can be calculated as:
wherein the method comprises the steps ofIs the upload rate of task s, d s Is the data size of task s. To ensure stability of the task upload queue, there are:
the average latency of the vehicle k upload task can be calculated as:
another constraint in view of the mobility of the vehicle is as follows:
wherein the method comprises the steps ofIs the residence time of vehicle k in its road side unit coverage.
The time to upload a task to vehicle k at time slot t may be calculated as:
the task arrival rate of the task s at the road side unit n is:
the total task arrival rate may be determined byGiven.
Cache decision for roadside unit nConstrained by storage capacity, can be expressed as:
wherein the method comprises the steps ofIndicating whether task s is cached, a s Is the buffer size of task s, H n Indicating the storage capacity size of the roadside unit s. Assume that the processing time compliance parameter of task s at roadside unit n is μ n,s =d s α s /f n An exponential distribution of d s Is the data size of task s, α s Is the processing density of task s, f n Is the processing frequency of the roadside unit processor.
The arrival rate of the task processed by the road side unit can be calculatedThe probability that a newly arrived task belongs to task s is +.>The processing time of the task at the road side unit follows an over-exponential distribution. Let random variable X n Representing the processing time of the task at the roadside unit n. X is X n Is a probability function f (x n ) Can be calculated as:
X n the expectations and variances of (1) may be calculated as:
to ensure that the task processing queues remain stable at the roadside unit n, there are the following constraints:
according to the Pollaczek-Khinachine formula, the average processing delay of the task roadside unit n may be calculated as:
the average consumed energy can be expressed as:
wherein P is n,c Is the power at which the processor of the roadside unit n processes the task.
The cost of processing a task at the time slot t-roadside unit n can be calculated as:
wherein eta n,s Is the cost of processing tasks s by the roadside unit n.
If two adjacent time slots have different buffer decisions, the road side unit needs to pay buffer update cost. The average cache update cost is as follows:
wherein eta n,s′ Is the cost, L, of the database of the roadside unit n cache tasks s t Is the length of one slot.
And 2, solving time delay, energy consumption and cost generated by the task in the remote cloud processing in the t-th time slot.
Assume that the uploading delay obeying parameter of the task s uploaded from the road side unit n to the remote cloud center isIs an exponential distribution of>r n Is the transmission rate. Similar to the analysis of the processing delay of a task at the roadside unit n, the uploading delay of a task may also be considered as it obeys the hyperexponential distribution. Let->The arrival rate of the task uploaded for the roadside unit n.
Let random variable Y n Representing the time a task is uploaded from a roadside unit to a remote cloud, the probability density function may be expressed as:
the expectations and variances are respectively:
considering the stability of the task upload queue, there are the following constraints:
the average delay of a road side unit uploading a task can be calculated as:
the average energy consumption of uploading a task is:
wherein P is n,r Is the power consumption of the road side unit upload task.
The number of servers is generally considered to be infinite in the remote cloud, and a G/M/≡model is established for simulating the processing process of tasks in the remote cloud, and the processing rate of the tasks s can be expressed asWherein f r Is the processing frequency of the remote cloud processor. The average processing delay of a task at the remote cloud can be expressed as:
the energy consumption to process a task can be expressed as:
wherein P is r Is the power consumption of the processor to process tasks.
The cost of remote cloud processing a task can be expressed as:
wherein eta s Is the average cost of remote cloud processing of a task.
Furthermore, if the vehicle device exits the coverage of the original road side unit before the calculation result is obtained, a handover delay will occur. The switching delay will be calculated by:
wherein D is h Is a switching time delay L r Is the road length.
Based on the analysis of the roadside unit processing model and the remote cloud processing model, the average processing delay of a task, the energy consumption and the system cost can be calculated as:
wherein T is r Is the propagation delay between the roadside unit and the remote cloud.
Step 3, constructing an optimization problem by taking the long-term average time delay of the minimized task as a target
Definition of the definitionAnd->The allocation decision vector is cached for the bandwidth allocation decision vector. The minimization is as follows:
step 4, the original minimization problem of the Lyapunov optimization framework can be decomposed into two sub-problems, one is a bandwidth allocation sub-problem and the other is a cache allocation sub-problem. The bandwidth allocation sub-problem is a convex optimization problem, and the problem is large in scale and is solved by adopting an alternate direction multiplication method, the method can update the solution vector in a distributed mode to obtain an optimal solution, wherein each update is equivalent to solving a quadratic programming problem, and the problem can be solved by a convex optimization tool box; the buffer allocation sub-problem is a nonlinear integer programming problem, which is generally regarded as an NP-hard problem, and can be solved by a monte carlo simulation method, and solution vectors are obtained by generating a large number of random numbers.
Compared with the prior art, the technical scheme of the embodiment has obvious beneficial effects.
The embodiment considers the unloading scene with various task types under the three-layer network architecture, and proposes an unloading strategy to perform bandwidth allocation and buffer allocation in the Internet of vehicles network. Different queue models are established for different network layers, and further average processing time delay, energy consumption and system cost of tasks are deduced. The optimization problem is constructed by taking the long-term average processing time delay of the minimized task as a target, and the original optimization problem is decomposed into two optimization problems by means of a Lyapunov optimization framework, and the two optimization problems are respectively solved by an alternate direction multiplier method and a Monte Carlo simulation method.
In addition, the invention also provides a task unloading strategy generation system in the Internet of vehicles, which comprises the following steps:
the task acquisition module is used for acquiring a task of the vehicle equipment;
the performance calculation module is used for calculating the system performance of different processing network layers for the vehicle equipment tasks according to the task processing network architecture to obtain multi-layer system performance parameters; the task processing network architecture is constructed according to vehicle equipment, road side units and remote clouds; the system performance parameters include: time delay, energy consumption and cost;
the optimization solving module is used for constructing an optimization problem by taking the minimum time delay as a target and combining energy consumption and cost, and solving the optimization problem to obtain a task unloading strategy in the Internet of vehicles; the optimization problem is an optimization problem based on bandwidth allocation and cache allocation.
The invention also provides electronic equipment, which comprises a memory and a processor, wherein the memory is used for storing a computer program, and the processor runs the computer program to enable the electronic equipment to execute the task unloading strategy generation method in the Internet of vehicles.
The invention also provides a computer readable storage medium storing a computer program which when executed by a processor implements the task offloading policy generation method in the internet of vehicles as described above.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other.
The principles and embodiments of the present invention have been described herein with reference to specific examples, the description of which is intended only to assist in understanding the core concept of the invention; also, it is within the scope of the present invention to be modified by those of ordinary skill in the art in light of the present teachings. In view of the foregoing, this description should not be construed as limiting the invention.
Claims (6)
1. The task unloading strategy generation method in the Internet of vehicles is characterized by comprising the following steps of:
acquiring a vehicle equipment task;
according to the task processing network architecture, system performance calculation of different processing network layers is carried out on the vehicle equipment tasks to obtain multi-layer system performance parameters; the task processing network architecture is constructed according to vehicle equipment, road side units and remote clouds; the system performance parameters include: time delay, energy consumption and cost;
constructing an optimization problem by taking minimum time delay as a target and combining energy consumption and cost, and solving the optimization problem to obtain a task unloading strategy in the Internet of vehicles; the optimization problem is an optimization problem based on bandwidth allocation and buffer allocation;
the system performance calculation of different processing network layers is carried out on the vehicle equipment tasks according to the task processing network architecture to obtain multi-layer system performance parameters, and the method specifically comprises the following steps:
in the step 1, in the t time slot, calculating the vehicle equipment task by using a system performance model, wherein the time delay, the energy consumption and the cost are respectively generated during road side unit and remote cloud processing, and the method specifically comprises the following steps:
the road side is provided with N road side units, and the number of vehicles covered by the road side units N at the time of a time slot t is equal to the number of vehicles covered by the road side units N; the road side unit distributes the bandwidth evenly to each vehicle, and the bandwidth obtained by each vehicle is as follows:wherein (1)>Indicating that vehicle k can offload tasks to RSU a 1 And RSU A 2 ,/>Representing RSU A 1 Number of vehicles in coverage area>Representing RSU A 2 The number of vehicles in the coverage area, B, is the total bandwidth of the road side unit; defining a bandwidth allocation vector->S represents the number of task categories, S e {1, 2..s }; the uploading rate of the uploading task s of the vehicle k at the time slot t is as follows:
wherein,is the bandwidth of the vehicle k sending task s, +.>Is the transmission power of the transmission task s of vehicle k, < >>Is the channel gain, sigma, of vehicle k and road side unit 2 Is white gaussian noise; considering that the bandwidth allocated to the vehicle must be used entirely for the uploading task, there is therefore a constraint: />Assuming that the uploading time delay of the task s obeys an exponential distribution, a plurality of types of task unloading processes can be regarded as a plurality of M/M/1 queues; the time for the vehicle k to upload the task s on average at the time of the time slot t is calculated as:
wherein,is the upload rate of task s, d s Is the data size of task s; to ensure stability of the task upload queue, there are:
wherein,representing the arrival rate of the task s of the vehicle k;
the average delay of the vehicle k upload task is calculated as:
wherein,the task arrival rate vector of the vehicle k at the time of the time slot t;
another constraint on the mobility of the vehicle is:
wherein the method comprises the steps ofIs the residence time of vehicle k in its road side unit coverage;
the time to upload a task at time slot t vehicle k is calculated as:
the task arrival rate of the task s at the road side unit n is:
the total task arrival rate may be determined byObtaining;
cache decision for roadside unit nConstrained by storage capacity, expressed as:
wherein,indicating whether task s is cached, a s Is the buffer size of task s, H n Representing the storage capacity size of the roadside unit s; assume that the processing time compliance parameter of task s at roadside unit n is μ n,s =d s α s /f n An exponential distribution of d s Is the data size of task s, α s Is the processing density of task s, f n Is the processing frequency of the roadside unit processor;
calculating the arrival rate of a task processed by a road side unitThe probability that a newly arrived task belongs to task s is +.>Therefore, the processing time of the task at the road side unit obeys the hyper-exponential distribution; let random variable X n Representing the processing time of the task at the roadside unit n; x is X n Is a probability function f (x n ) Is calculated as:
wherein mu n,s Representing the processing rate of task s at roadside unit n;
X n the expected and variance of (1) is calculated as:
to ensure that the task processing queues remain stable at the roadside unit n, there are the following constraints:
according to the Pollaczek-Khinachine formula, the average processing delay of the task roadside unit n is calculated as:
the average consumed energy is expressed as:
wherein P is n,c Is the power of the processing task of the processor of the road side unit n;
the cost of processing a task at the time slot t-way side unit n is calculated as:
wherein eta n,s Is the cost of processing task s by roadside unit n;
if two adjacent time slots have different buffer decisions, the road side unit needs to pay buffer updating cost; the average cache update cost is as follows:
wherein eta n,s′ Is the cost, L, of the database of the roadside unit n cache tasks s t Is the length of one slot;
the system performance model includes:
wherein,representing the average processing time delay of a task; />An arrival rate indicating a task to be processed at the roadside unit;representing the total task arrival rate; />Representing the average processing time delay of the task road side unit n; />The arrival rate of the task uploaded by the road side unit n is represented; />Representing the average time delay of a task uploaded by a road side unit; />Representing average processing delay of a task in a remote cloud; t (T) r Representing propagation delay between the roadside unit and the remote cloud; />The task switching delay of the road side unit is represented;
wherein,representing the average processing energy consumption of a task; />Representing the average processing energy consumption of the task road side unit n; />Representing average processing energy consumption of a task in a remote cloud; />Representing the average processing energy consumption of a task uploaded by the road side unit n;
wherein,representing the average processing cost of a task; />Representing the cost of a roadside unit to process a task; />Representing average cache update costs for the roadside units; />Representing the cost of remote cloud processing of a task;
step 2, determining a multi-layer system performance parameter according to the calculation results of the road side unit and the remote cloud, wherein the method specifically comprises the following steps:
solving time delay generated by the task in the remote cloud processing in the t-th time slot, energy consumption and cost:
assume that the uploading delay obeying parameter of the task s uploaded from the road side unit n to the remote cloud center isIs an exponential distribution of>r n Is the transmission rate; let->The arrival rate of the task uploaded by the road side unit n;
let random variable Y n Representing the time a task was uploaded from a roadside unit to a remote cloud, the probability density function is expressed as:
the expectations and variances are respectively:
considering the stability of the task upload queue, there are constraints:
the average time delay of uploading a task by the road side unit is calculated as follows:
the average energy consumption of uploading a task is:
wherein P is n,r Is the power consumption of the road side unit uploading task;
the number of servers is considered to be infinite in the remote cloud, and a G/M/≡model is established for simulating the processing process of tasks in the remote cloud, and the processing rate of tasks s is expressed asWherein f r Is the processing frequency of the remote cloud processor; the average processing delay of a task at the remote cloud is expressed as:
the energy consumption of processing a task is expressed as:
wherein P is r Is the power consumption of the processor to process tasks;
the cost of remote cloud processing of a task is expressed as:
wherein eta s Is the average cost of remote cloud processing of a task;
in addition, if the vehicle device exits the coverage area of the original road side unit before obtaining the calculation result, a switching delay is generated, and the calculation formula of the switching delay is as follows:
wherein D is h Is a switching time delay L r Is the road length;representing the average running speed of the vehicle k;
based on the analysis of the roadside unit processing model and the remote cloud processing model, the average processing time delay of one task, the energy consumption and the system cost are calculated as follows:
wherein T is r Is the propagation delay between the roadside unit and the remote cloud;
the optimization problem is constructed with the goal of minimizing the long-term average latency of the task:
definition of the definitionAnd->Allocating decision vectors for bandwidth and buffering allocation blocksVector policy, minimize:
wherein,indicating whether the database associated with task s is cached, when +.>Indicating that it is not cached when +.>When the representation is cached.
2. The method for generating the task offloading policy in the internet of vehicles according to claim 1, wherein the minimum time delay is used as a target, and an optimization problem is constructed by combining energy consumption and cost, and the optimization problem is solved to obtain the task offloading policy in the internet of vehicles, and the method specifically comprises:
converting the optimization problem into a bandwidth allocation sub-problem and a buffer allocation sub-problem by using a Lyapunov optimization framework;
solving a task unloading strategy of the bandwidth allocation sub-problem by using an alternate direction multiplication method;
solving a task unloading strategy of the buffer allocation sub-problem by using a Monte Carlo simulation method;
and determining a task unloading strategy in the Internet of vehicles according to the task unloading strategy of the bandwidth allocation sub-problem and the task unloading strategy of the buffer allocation sub-problem.
3. The method for generating task offloading policies in the internet of vehicles according to claim 1, wherein the optimization problem is used for filtering the buffered tasks according to time delay, energy consumption and cost, and when a task is buffered by a road side unit, the task is processed by the road side unit, and when a task not buffered by the road side unit is selected, the task is uploaded to a remote cloud for processing through a wired network.
4. A task offloading policy generation system in the internet of vehicles, applied to the method of any one of claims 1 to 3, comprising:
the task acquisition module is used for acquiring a task of the vehicle equipment;
the performance calculation module is used for calculating the system performance of different processing network layers for the vehicle equipment tasks according to the task processing network architecture to obtain multi-layer system performance parameters; the task processing network architecture is constructed according to vehicle equipment, road side units and remote clouds; the system performance parameters include: time delay, energy consumption and cost;
the optimization solving module is used for constructing an optimization problem by taking the minimum time delay as a target and combining energy consumption and cost, and solving the optimization problem to obtain a task unloading strategy in the Internet of vehicles; the optimization problem is an optimization problem based on bandwidth allocation and cache allocation.
5. An electronic device comprising a memory for storing a computer program and a processor that runs the computer program to cause the electronic device to perform the in-car networking task offloading policy generation method of any one of claims 1-3.
6. A computer readable storage medium, characterized in that it stores a computer program which, when executed by a processor, implements the task offloading policy generation method in the internet of vehicles as claimed in any one of claims 1 to 3.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311096130.5A CN117042051B (en) | 2023-08-29 | 2023-08-29 | Task unloading strategy generation method, system, equipment and medium in Internet of vehicles |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311096130.5A CN117042051B (en) | 2023-08-29 | 2023-08-29 | Task unloading strategy generation method, system, equipment and medium in Internet of vehicles |
Publications (2)
Publication Number | Publication Date |
---|---|
CN117042051A CN117042051A (en) | 2023-11-10 |
CN117042051B true CN117042051B (en) | 2024-03-08 |
Family
ID=88628158
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311096130.5A Active CN117042051B (en) | 2023-08-29 | 2023-08-29 | Task unloading strategy generation method, system, equipment and medium in Internet of vehicles |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117042051B (en) |
Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109862086A (en) * | 2019-01-21 | 2019-06-07 | 华北电力大学 | Task Assigned Policy based on matching algorithm in a kind of vehicle-mounted edge calculations |
CN112601197A (en) * | 2020-12-18 | 2021-04-02 | 重庆邮电大学 | Resource optimization method in train-connected network based on non-orthogonal multiple access |
CN112702414A (en) * | 2020-12-21 | 2021-04-23 | 中国电子科技集团公司第五十四研究所 | Edge fleet cloud computing unloading method, system, medium equipment and application |
CN112860350A (en) * | 2021-03-15 | 2021-05-28 | 广西师范大学 | Task cache-based computation unloading method in edge computation |
CN114143891A (en) * | 2021-11-30 | 2022-03-04 | 南京工业大学 | FDQL-based multi-dimensional resource collaborative optimization method in mobile edge network |
CN114153572A (en) * | 2021-10-27 | 2022-03-08 | 中国电子科技集团公司第五十四研究所 | Calculation unloading method for distributed deep learning in satellite-ground cooperative network |
CN114528042A (en) * | 2022-01-30 | 2022-05-24 | 南京信息工程大学 | Energy-saving automatic interconnected vehicle service unloading method based on deep reinforcement learning |
CN114863683A (en) * | 2022-05-11 | 2022-08-05 | 湖南大学 | Heterogeneous Internet of vehicles edge calculation unloading scheduling method based on multi-objective optimization |
CN114973673A (en) * | 2022-05-24 | 2022-08-30 | 华南理工大学 | Task unloading method combining NOMA and content cache in vehicle-road cooperative system |
CN115150893A (en) * | 2022-06-14 | 2022-10-04 | 燕山大学 | MEC task unloading strategy method based on task division and D2D |
CN115604765A (en) * | 2022-08-25 | 2023-01-13 | 东莞理工学院(Cn) | Computing offload optimization method and device, electronic equipment and storage medium |
CN115665859A (en) * | 2022-10-24 | 2023-01-31 | 南京邮电大学 | Multi-agent task unloading and resource scheduling combined optimization method |
CN116260871A (en) * | 2023-03-03 | 2023-06-13 | 南京理工大学 | Independent task unloading method based on local and edge collaborative caching |
CN116321298A (en) * | 2023-03-29 | 2023-06-23 | 西南科技大学 | Multi-objective joint optimization task unloading strategy based on deep reinforcement learning in Internet of vehicles |
CN116360883A (en) * | 2023-03-10 | 2023-06-30 | 沈阳理工大学 | Combined optimization method for unloading of Internet of vehicles computing tasks |
CN116405493A (en) * | 2023-03-01 | 2023-07-07 | 天津理工大学 | Edge cloud collaborative task unloading method based on MOGWO strategy |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CA3177392A1 (en) * | 2020-10-04 | 2022-04-07 | Strong Force Iot Portfolio 2016, Llc | Industrial digital twin systems and methods with echelons of executive, advisory and operations messaging and visualization |
-
2023
- 2023-08-29 CN CN202311096130.5A patent/CN117042051B/en active Active
Patent Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109862086A (en) * | 2019-01-21 | 2019-06-07 | 华北电力大学 | Task Assigned Policy based on matching algorithm in a kind of vehicle-mounted edge calculations |
CN112601197A (en) * | 2020-12-18 | 2021-04-02 | 重庆邮电大学 | Resource optimization method in train-connected network based on non-orthogonal multiple access |
CN112702414A (en) * | 2020-12-21 | 2021-04-23 | 中国电子科技集团公司第五十四研究所 | Edge fleet cloud computing unloading method, system, medium equipment and application |
CN112860350A (en) * | 2021-03-15 | 2021-05-28 | 广西师范大学 | Task cache-based computation unloading method in edge computation |
CN114153572A (en) * | 2021-10-27 | 2022-03-08 | 中国电子科技集团公司第五十四研究所 | Calculation unloading method for distributed deep learning in satellite-ground cooperative network |
CN114143891A (en) * | 2021-11-30 | 2022-03-04 | 南京工业大学 | FDQL-based multi-dimensional resource collaborative optimization method in mobile edge network |
CN114528042A (en) * | 2022-01-30 | 2022-05-24 | 南京信息工程大学 | Energy-saving automatic interconnected vehicle service unloading method based on deep reinforcement learning |
CN114863683A (en) * | 2022-05-11 | 2022-08-05 | 湖南大学 | Heterogeneous Internet of vehicles edge calculation unloading scheduling method based on multi-objective optimization |
CN114973673A (en) * | 2022-05-24 | 2022-08-30 | 华南理工大学 | Task unloading method combining NOMA and content cache in vehicle-road cooperative system |
CN115150893A (en) * | 2022-06-14 | 2022-10-04 | 燕山大学 | MEC task unloading strategy method based on task division and D2D |
CN115604765A (en) * | 2022-08-25 | 2023-01-13 | 东莞理工学院(Cn) | Computing offload optimization method and device, electronic equipment and storage medium |
CN115665859A (en) * | 2022-10-24 | 2023-01-31 | 南京邮电大学 | Multi-agent task unloading and resource scheduling combined optimization method |
CN116405493A (en) * | 2023-03-01 | 2023-07-07 | 天津理工大学 | Edge cloud collaborative task unloading method based on MOGWO strategy |
CN116260871A (en) * | 2023-03-03 | 2023-06-13 | 南京理工大学 | Independent task unloading method based on local and edge collaborative caching |
CN116360883A (en) * | 2023-03-10 | 2023-06-30 | 沈阳理工大学 | Combined optimization method for unloading of Internet of vehicles computing tasks |
CN116321298A (en) * | 2023-03-29 | 2023-06-23 | 西南科技大学 | Multi-objective joint optimization task unloading strategy based on deep reinforcement learning in Internet of vehicles |
Non-Patent Citations (6)
Title |
---|
"JOTE: Joint Offloading of Tasks and Energy in Fog-Enabled IoT Networks";Penghao Cai等;《IEEE Internet of Things Journal》;全文 * |
"Performance Analysis of a VM-PM Repair Strategy in MEC-enabled Wireless Systems with Bursty Traffic";Yuting Wang等;《IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY》;全文 * |
"Performance evaluation and optimization of a task offloading strategy on the mobile edge computing with edge heterogeneity";W Li等;《https://link.springer.com/article/10.1007/s11227-021-03781-w》;全文 * |
"基于单用户和分类任务的MEC任务卸载策略及性能优化";金顺福等;《燕山大学学报》;全文 * |
"面向车辆边缘网络中的任务动态卸载算法";曾耀平等;《计算机工程与应用》;全文 * |
Hao Zhao等. Hao Zhao等,"Performance research on a task ofoading strategy in a two tier edge structure based MEC system".《https://link.springer.com/article/10.1007/s11227-023-05059-9》.2023,Hao Zhao等. * |
Also Published As
Publication number | Publication date |
---|---|
CN117042051A (en) | 2023-11-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112601197B (en) | Resource optimization method in train-connected network based on non-orthogonal multiple access | |
CN111414252B (en) | Task unloading method based on deep reinforcement learning | |
CN112839382B (en) | Video semantic driven communication and computing resource joint allocation method in Internet of vehicles | |
CN111010684B (en) | Internet of vehicles resource allocation method based on MEC cache service | |
CN110312231A (en) | Content caching decision and resource allocation joint optimization method based on mobile edge calculations in a kind of car networking | |
CN111629380B (en) | Dynamic resource allocation method for high concurrency multi-service industrial 5G network | |
CN113268341B (en) | Distribution method, device, equipment and storage medium of power grid edge calculation task | |
CN111711666B (en) | Internet of vehicles cloud computing resource optimization method based on reinforcement learning | |
CN113645273B (en) | Internet of vehicles task unloading method based on service priority | |
CN115297171B (en) | Edge computing and unloading method and system for hierarchical decision of cellular Internet of vehicles | |
CN111132074A (en) | Multi-access edge computing unloading and frame time slot resource allocation method in Internet of vehicles environment | |
CN113382383A (en) | Method for unloading calculation tasks of public transport vehicle based on strategy gradient | |
CN111511028A (en) | Multi-user resource allocation method, device, system and storage medium | |
CN114641041A (en) | Edge-intelligent-oriented Internet of vehicles slicing method and device | |
CN116367231A (en) | Edge computing Internet of vehicles resource management joint optimization method based on DDPG algorithm | |
CN116541106A (en) | Computing task unloading method, computing device and storage medium | |
CN115835294A (en) | RAN slice and task unloading joint optimization method assisted by deep reinforcement learning in Internet of vehicles | |
Ma et al. | Deep reinforcement learning for pre-caching and task allocation in internet of vehicles | |
Qu et al. | Stochastic cumulative DNN inference with RL-aided adaptive IoT device-edge collaboration | |
CN117042051B (en) | Task unloading strategy generation method, system, equipment and medium in Internet of vehicles | |
CN113452625B (en) | Deep reinforcement learning-based unloading scheduling and resource allocation method | |
CN115134829A (en) | Mobile service edge cooperation unloading method, device, equipment and medium | |
Ke et al. | Medley deep reinforcement learning-based workload offloading and cache placement decision in UAV-enabled MEC networks | |
Lieira et al. | Meta-heuristic mechanism based on whale optimization algorithm for tasks allocation in edge computing | |
CN117857559B (en) | Metropolitan area optical network task unloading method based on average field game and edge server |
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 |