CN115550357A - Multi-agent multi-task cooperative unloading method - Google Patents
Multi-agent multi-task cooperative unloading method Download PDFInfo
- Publication number
- CN115550357A CN115550357A CN202211016816.4A CN202211016816A CN115550357A CN 115550357 A CN115550357 A CN 115550357A CN 202211016816 A CN202211016816 A CN 202211016816A CN 115550357 A CN115550357 A CN 115550357A
- Authority
- CN
- China
- Prior art keywords
- task
- unloading
- vehicle
- node
- service node
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 66
- 238000004364 calculation method Methods 0.000 claims abstract description 24
- 235000008694 Humulus lupulus Nutrition 0.000 claims abstract description 8
- 230000006854 communication Effects 0.000 claims description 56
- 238000004891 communication Methods 0.000 claims description 44
- 230000005540 biological transmission Effects 0.000 claims description 8
- 238000005562 fading Methods 0.000 claims description 7
- 241000197727 Euscorpius alpha Species 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 3
- 239000003795 chemical substances by application Substances 0.000 description 19
- 239000013598 vector Substances 0.000 description 7
- 108090000623 proteins and genes Proteins 0.000 description 6
- 210000000349 chromosome Anatomy 0.000 description 5
- 239000000543 intermediate Substances 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000035772 mutation Effects 0.000 description 3
- 230000006855 networking Effects 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003786 synthesis reaction Methods 0.000 description 2
- 108700028369 Alleles Proteins 0.000 description 1
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 229910052731 fluorine Inorganic materials 0.000 description 1
- 125000001153 fluoro group Chemical group F* 0.000 description 1
- 238000009396 hybridization Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The application discloses a multi-agent multi-task cooperative unloading method, which comprises the following steps: setting a task and a service node thereof; then determining the node type, and then representing the tasks generated by any task vehicle by using triples; then, local computation and multi-service node edge unloading computation are selected for task unloading; on the basis of selecting edge unloading calculation, judging the idle state of a service node of a task vehicle TaV, wherein the service node meets the requirement of serving as the relay node; then, based on the result judgment of the idle state, executing the unloading task in a single-hop unloading mode; and if the idle state does not meet the single-hop unloading condition, unloading the calculation unloading task into an idle service node of the relay node for execution in two hops. According to the technical scheme, in a dynamic vehicle-mounted environment, the problem of multi-vehicle multi-task cooperative unloading conflict can be solved by using limited computing resources of task vehicles and service resources of surrounding spaces, and minimum system average time delay is realized.
Description
Technical Field
The application relates to the technical field of vehicle networking communication, in particular to a multi-agent multi-task cooperative unloading method.
Background
The vehicle internet is a typical industrial internet of things technology, and in the technology, ubiquitous information can be exchanged and shared among vehicles without manual intervention. In the car networking environment, a vehicle in driving generates massive sensor data every second, and in order to have an intelligent visual field in a complex driving environment, a large amount of data transmission, storage, processing and other operations need to be completed in a short time. The real three-dimensional mountain-turning road multitask scene belongs to a high-speed traffic accident road section, is limited by terrain conditions and a sheltering environment, is slow in vehicle running speed, and has a blind area covered by RSU signals, so that task collaborative unloading is extremely important.
Currently, there are several deficiencies in the current investigated vehicle networking task collaborative offloading scheme. On the unloading cooperative unloading model, most cooperative unloading modes are executed by adopting a single-process unloading model, so that higher time delay is brought by the unloading mode, and the utilization rate of channel resources is lower; then, in a task splitting mode, the types of tasks cooperatively unloaded are mainly inseparable and equal split at present, and the task splitting scheme often cannot fully utilize the computing resources on the service vehicle of the host vehicle; secondly, in the moving state of the vehicle, the vehicle is in a relatively static state in the current unloading research process, but for the real three-dimensional road scene, the vehicle is in a moving state in the cooperative unloading process; and more critical are: in terms of conflict processing, when a multi-task multi-host vehicle collaborative unloading scene occurs, the problem that a plurality of host vehicles compete for one collaborative node together can be encountered, communication conflict occurs, task unloading efficiency is reduced, task unloading delay is increased, and good service quality cannot be provided for users.
Therefore, how to provide a multi-agent multi-task cooperative unloading method, which can solve the multi-vehicle multi-task cooperative unloading conflict problem by using the limited computing resources of the task vehicle and the service resources of the surrounding space in the dynamic vehicle-mounted environment, and realize the minimum system average time delay has become a technical problem to be urgently solved by the technical personnel in the field.
Disclosure of Invention
In order to solve the technical problem, the application provides a multi-agent multi-task cooperative unloading method, which can solve the problem of multi-vehicle multi-task cooperative unloading conflict and realize minimum system average time delay by utilizing limited computing resources of task vehicles and service resources of surrounding spaces in a dynamic vehicle-mounted environment.
The technical scheme provided by the application is as follows:
the application provides a multi-agent multi-task cooperative unloading method, which comprises the following steps: s1, setting a vehicle task unloading scene: the vehicle producing the mission is regarded as a mission vehicle TaV; the task vehicle is always within a communication range within the maximum tolerable time delay, and a service node for providing unloading service for the task vehicle is configured; s2, determining the node type: service nodes belonging to the communication range of the plurality of task vehicles TaV are determined as candidate service nodes; when the candidate service node is determined to be v in task vehicle TaV i After the service node is obtained, the current candidate service node can be used as a candidate relay node of the remaining task vehicles; then, screening out relay nodes capable of executing two-hop unloading and service nodes of the relay nodes from the candidate relay nodes; s3, unloading task representation: will arbitrarily task the vehicle v i Generated task-use triplets:
represents; wherein D i To unload the size of a task, C i Tmax, the computational resources required for a task i Maximum tolerable latency for executing the task; s4, unloading mode selection: after defining the node state, selecting local computation and multi-service node edge unloading computation for task unloading, wherein the edge unloading computation comprises single-hop unloading and two-hop unloading; s5, executing an unloading task: determining tasks based on select edge offload computationV in vehicle TaV i Service node v j And said serving node v, and j satisfying a requirement to act as the relay node; then, based on the result judgment of the idle state, executing the unloading task in the single-hop unloading mode; if the idle state does not meet the single-hop unloading condition, unloading a calculation unloading task to the relay node v through the two hops j Idle service node v k Is executed.
Further, in a preferred embodiment of the present invention, the step S4 further includes:
if the local computation is selected, namely TaV selects to use local computing resources to process tasks, the local computation time delay is directly obtained
Wherein,
representing the local computing power of TaV,
representing the computational load of the local task.
Further, in a preferred mode of the present invention, in the step S4, the defining node status specifically includes: using a variable Δ of 0-1 i，j Representing said mission vehicle v in single hop unloading i Of a serving node v within communication range j The state of (2):
using a variable beta of 0-1 j，k Indicating a relay node v when performing two-hop offloading j Service node v k The state of (2):
further, in a preferred embodiment of the present invention, in the step S5, the step of executing the unloading task specifically includes:
judging v in task vehicle TaV i Service node v j Idle state of (2):
if the task vehicle v i Service node v j In an idle state, i.e. alpha i，j =0; offloading tasks directly to service node v through single hop offloading j Calculating to obtain single-hop calculation time delay;
wherein,
for task vehicles v i To the service node v j A distributed sub-computation task; r is i，j Is v i Transmitting a computing task to v j The upload rate of (d); mu.s 1 The overlapping factor in the communication process represents the proportion of data which can be unloaded to the edge node;
representing a service node v j For task vehicle v i The calculated force of the distribution;
if the task vehicle v i Service node v j In an occupied state, i.e. alpha i，j =1, description serving node v j Has been selected as a serving node by the remaining TaV; at a service node v j Under the condition of meeting the requirement of serving as a relay node, the node v j Serving node v offloaded to its idle as a relay node over two hops k Performing off-load computing tasks, i.e. beta j，k =0; then two-hop computation delay is obtained:
wherein,
denotes v i Will calculate the subtask
Upload to relay node v j ；
Indicating a relay node v j Will calculate the task
Two-hop offload to relay node v j Service node v k (ii) a Vt is represented as a set of mission vehicles TaV; vs is represented as a collection of service nodes.
Further, in a preferred mode of the present invention, in step S2, after the determining the node type is completed, the method further includes: constructing a vehicle motion model, passing a mission vehicle v α Position information at time t
Predicting position information at t + delta t moment
Further, in a preferred embodiment of the present invention, a vehicle motion model is constructed and position information is estimated
The method specifically comprises the following steps:
s1, obtaining task vehicle v α Velocity at time t
The included angles with the x-axis, the y-axis and the z-axis are theta x ，θ y ，θ z ∈[0，π]；
S2, according to the speed
And included angle acquisition speed
Velocity components in the x, y and z axes;
s3, obtaining a vehicle running delta t through the direction cosine law, and then obtaining a task vehicle v α Distance components traveled in the x, y and z directions;
cos is obtained by the direction cosine theorem 2 θ x +cos 2 θ y +cos 2 θ z =1, i.e.:
s4, subsequent application of L 2 Calculating norm to obtain task vehicle v α To v β Euclidean distance at any instant:
namely, it is
Further, in a preferred mode of the present invention, in the step S2 of determining the node type, the method further includes: clustering the running roads of the vehicles based on the task vehicles, wherein N +1 edge servers are provided, and M task vehicles Vt are provided, so that service node clustering of the multi-task vehicles is performed;
the service node clustering step specifically includes:
s1, from a vehicle set V = { V = 0 ，v 1 ，v 2 ，...，v N Randomly selecting M task vehicles as centroids;
s2, sampling in the service node, and calculating Euclidean distances from each sample to M centroids;
s3, if the sample is within the range of delta t epsilon from 0 to Tmax i ]Euclidean distance from any time to centroid
Entering the subsequent step; wherein R is the coverage radius of a roadside unit RSU which is arranged on a road and provides service for all equipment;
s4, if the sample is in the communication range of the centroids, alpha i，j =1 using the sample as the selected serving node; otherwise alpha i，j =0, the sample is divided into clusters corresponding to the centroid;
and S5, finally, outputting a clustering result to obtain a service node set Vs of the task vehicle.
Further, in a preferred embodiment of the present invention, before the unloading manner is selected in step S4, the method further includes: constructing a communication model, and acquiring the rate of a communication uploading link; the method comprises the following steps:
s1, setting a communication mode between a task vehicle and a service node: a short-range wireless communication mode; orthogonal frequencies are adopted to reduce the mutual influence in the communication process;
s2, setting an uplink in the communication process as a flat Rayleigh fading channel without considering channel interference;
s3, calculating the task vehicle v according to the Shannon formula i To the service node v j Average transmission rate of the uplink:
wherein, may be v2v or v2r, W v2v And W v2r Respectively representing V2V and V2R channel bandwidths; pi is the task vehicle v i Of the transmitted power, p 0 Representing the white Gaussian noise power, d i，j Is v i To v j The transmission distance of (a);
represents a path loss exponent; the channel fading coefficient is denoted by h.
Further, in a preferred embodiment of the present invention, before unloading the task representation in step S3, the method further includes the following steps: and determining an unequal task set C of the task vehicle distributed to the service nodes.
Further, in a preferred mode of the present invention, the specific steps of task allocation include:
s1, setting an objective function, wherein the objective function aims to solve a calculation time delay problem of multi-vehicle multi-task unloading;
s2, converting the target function into an absolute value of a parallel difference between unloading of the edge node and local computation
Minimization problems, i.e.
Closer to 0 represents higher parallelism; the transformed objective function is:
wherein, C represents a computation subtask with unequal splitting; constraint C 1 Constraint boundaries representing local and offload computation subtasks; c 2 Indicating the maximum tolerance time limit Tmax of the task i Inner mission vehicle v i To the service node v j Relative Euclidean distance therebetween is less than or equal to v i The communication range R of (2); c 3 Indicating that the sum of local and multi-service node offload computation subtasks equals task vehicle v i Total calculation task of
And S3, distributing the unloading task to the service node by utilizing a differential evolution algorithm.
Compared with the prior art, the multi-agent multi-task cooperative unloading method provided by the invention comprises the following steps: s1, setting a vehicle task unloading scene: identifying the vehicle producing the mission as mission vehicle TaV; the task vehicle is always within a communication range within the maximum tolerable time delay, and a service node for providing unloading service for the task vehicle is configured; s2, determining the node type: service nodes belonging to the communication range of the plurality of task vehicles TaV are determined as candidate service nodes; when the candidate service node is determined to be v in task vehicle TaV i After the service node is obtained, the current candidate service node can be used as a candidate relay node of the remaining task vehicles; then, screening out relay nodes capable of executing two-hop unloading and service nodes of the relay nodes from the candidate relay nodes; s3, unloading task representation: will arbitrarily task the vehicle v i Generated task-use triplets:
represents; wherein D i To unload the size of a task, C i Tmax, a computing resource required for a task i Maximum tolerable latency for executing the task; s4, unloading mode selection: after defining the node state, selecting local computation and multi-service node edge unloading computation for task unloading, wherein the edge unloading computation comprises single-hop unloading and two-hop unloading; s5, executing an unloading task: on the basis of selecting edge unloading calculation, judging v in task vehicle TaV i Service node v j And said serving node v, and j satisfying a requirement to act as the relay node; then, based on the result judgment of the idle state, executing the unloading task in the single-hop unloading mode; if the idle state does not meet the single-hop unloading condition, unloading a calculation unloading task to the relay node v through the two hops j Idle service node v k Is executed. By utilizing the multi-agent multi-task cooperative unloading method provided by the application, when contradiction between limited computing resources and user required experience occurs, local execution is carried outThe parallel computing task is executed by the row and multi-service node multi-hop distributed unloading, the computing resource of the task vehicle and the service resource of the surrounding space are fully utilized on the premise of guaranteeing the service requirement of the user, and the overall computing time delay of the minimum system is realized. When the number of task vehicles is large, the overall time delay of the system is better by adopting multi-hop distributed unloading compared with the traditional single-hop single-node unloading system. According to the technical scheme, the problem of multi-vehicle multi-task cooperative unloading conflict can be solved by using limited computing resources of the task vehicle and service resources of the surrounding space, and the minimum system average time delay is realized.
Has the beneficial effects that:
1. the multi-agent multi-task cooperative unloading method provided by the application can solve the problem of multi-vehicle multi-task cooperative unloading conflict by adopting a multi-hop serial unloading distributed execution unloading strategy at multiple tasks and multiple nodes;
2. the multi-agent multi-task collaborative unloading method has the obvious low-delay characteristic in solving the problem of multi-task optimized unloading, and the prediction of the vehicle moving track under the three-dimensional road scene is realized;
3. the multi-agent multi-task collaborative unloading method provided by the application allocates the unloading tasks of the task vehicles to the service nodes based on the differential evolution algorithm, and has good convergence when solving the high-dimensional nonlinear problem of unequal task splitting.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flowchart illustrating steps of a multi-agent multi-task cooperative offloading method according to an embodiment of the present invention;
FIG. 2 is a flowchart of a framework for solving a service node by a clustering algorithm according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a first clustering structure of a service node set according to an embodiment of the present invention;
fig. 4 is a schematic diagram of a second cluster structure of a service node set according to an embodiment of the present invention;
fig. 5 is a schematic diagram of a third cluster structure of a service node set according to an embodiment of the present invention;
fig. 6 is a comparison diagram of task offloading computation latencies provided by the embodiment of the present invention.
Detailed Description
In order to make those skilled in the art better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It will be understood that when an element is referred to as being "fixed" or "disposed" on another element, it can be directly on the other element or be indirectly disposed on the other element; when an element is referred to as being "connected to" another element, it can be directly connected to the other element or be indirectly connected to the other element.
It will be understood that the terms "length," "width," "upper," "lower," "front," "rear," "first," "second," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like, as used herein, refer to an orientation or positional relationship indicated in the drawings that is solely for the purpose of facilitating the description and simplifying the description, and do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus should not be construed as limiting the application.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or to implicitly indicate the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the present application, "plurality" or "a plurality" means two or more unless specifically limited otherwise.
It should be understood that the structures, ratios, sizes, and the like shown in the drawings are only used for matching the disclosure of the present disclosure to be understood and read by those skilled in the art, and are not used for limiting the practical limitations of the present disclosure, so they do not have the essential technical meaning, and any modifications of the structures, changes of the ratio relationships, or adjustments of the sizes, should still fall within the scope of the technical disclosure of the present disclosure without affecting the function and the achievable purpose of the present disclosure.
The invention provides a multi-agent multi-task cooperative unloading method, which comprises the following steps: s1, setting a vehicle task unloading scene: the vehicle producing the mission is regarded as a mission vehicle TaV; the service node is configured to provide unloading service for the task vehicle when the task vehicle is always within the communication range within the maximum tolerable time delay; s2, determining the node type: service nodes belonging to the communication range of the plurality of task vehicles TaV are determined as candidate service nodes; when the candidate service node is determined to be v in task vehicle TaV i After the service node is obtained, the current candidate service node can be used as a candidate relay node of the remaining task vehicles; then, screening out relay nodes capable of executing two-hop unloading and service nodes of the relay nodes from the candidate relay nodes; s3, unloading task representation: will arbitrarily task the vehicle v i Generated task-use triplets:
representing; wherein D i To unload the size of a task, C i Tmax, a computing resource required for a task i Maximum tolerable latency for executing the task; s4, unloading mode selection: after defining node state, selecting local computation and multi-service node edge unloading computationLine task offloading, wherein the edge offloading computation comprises single hop offloading and two hop offloading; s5, executing an unloading task: based on the calculation of the selective edge unloading, v in the task vehicle TaV is judged i Service node v j And said serving node v, and j satisfying a requirement to act as the relay node; then, based on the result judgment of the idle state, executing the unloading task in the single-hop unloading mode; if the idle state does not meet the single-hop unloading condition, unloading a calculation unloading task to the relay node v through the two hops j Idle service node v k Is executed. The invention provides a multi-agent multi-task cooperative unloading method which can solve the problem of multi-vehicle multi-task cooperative unloading conflict and realize minimum system average time delay by utilizing limited computing resources of task vehicles and service resources of surrounding spaces.
Referring to fig. 1 to fig. 6, the following describes steps of a multi-agent multitask cooperative offloading method disclosed in the present application with reference to specific embodiments.
The application takes the ten turns in Enshi in Hubei as a specific embodiment, and provides a multi-agent multi-task cooperative unloading method which specifically comprises the following steps:
s1, setting a vehicle task unloading scene: the vehicle producing the mission is regarded as a mission vehicle TaV; and configuring a service node for providing unloading service for the task vehicle when the task vehicle is always in a communication range within the maximum tolerable time delay.
In the embodiment of the application, a ten-turn road is taken as an unloading scene; firstly, a Road Side Unit (RSU) with a coverage radius R and capable of providing service for all devices is deployed on a Road, and an MEC server is deployed on the RSU, denoted by v 0 . With set V = { V = } 1 ，v 2 ，...，v N The method comprises the steps that N vehicle-mounted terminals in Poisson distribution in the RSU coverage range are represented, positioning equipment such as a Beidou satellite navigation system (BDS) and the like are arranged in each vehicle-mounted terminal, and track information of vehicles can be obtained in real time through the equipment; vehicles producing tasks are called Task vehicles (Task Veh)icle, taV), expressed as a collection
5363 vehicles and RSUs within communication range of TaV, collectively referred to as Service Nodes (SNs), are used in sets
And (4) showing. Collection of
Marking real-time track information of all vehicles on the road, wherein the track information of the alpha vehicle at the t moment is
Wherein
In order to be the speed of the vehicle,
is the position coordinates of the alpha-th vehicle.
S2, determining the node type: service nodes belonging to the communication range of the plurality of task vehicles TaV are determined as candidate service nodes; when the candidate service node is determined to be v in task vehicle TaV i After the service node is obtained, the current candidate service node can be used as a candidate relay node of the rest task vehicles; and then screening out the relay nodes capable of executing two-hop unloading and the service nodes of the relay nodes from the candidate relay nodes.
Wherein, intersection vehicles when SN belongs to the communication range of a plurality of task vehicles TaV are collectively called Candidate service Nodes (CsN) and are used as a set
When the candidate service node has been determined to be TaV i After the service node(s), the current candidate service node may serve as a remaining mission vehicle (Vs- { v) i }) Candidate relay Nodes (CrN), using sets
Subsequently, by further screening out Relay Nodes (RN) capable of executing two-hop unloading from the CrN, the set is used
Serving Nodes (RsN) of Relay Nodes, using sets
Specifically, in an embodiment of the present invention, the multi-agent multitask collaborative offloading method further includes: after the determining the node type is completed, the method further comprises the following steps: constructing a vehicle motion model, passing through a task vehicle v α Position information at time t
Predicting position information at t + delta t moment
Specifically, in the embodiment of the invention, a vehicle motion model is constructed, and position information is estimated
The method specifically comprises the following steps: s1, obtaining task vehicle v α Velocity at time t
The included angles with the x-axis, the y-axis and the z-axis are theta x ，θ y ，θ z ∈[0，π]；
S2, according to the speed
And included angle acquisition speed
On the x-axis,Velocity components of the y-axis and z-axis;
s3, obtaining a vehicle running delta t through the direction cosine law, and then obtaining a task vehicle v α Distance components traveled in the x-axis, y-axis, and z-axis directions;
cos is obtained by the direction cosine theorem 2 θ x +cos 2 θ y +cos 2 θ z ＝1，
Namely:
s4, subsequent application of L 2 Calculating norm to obtain task vehicle v α To v β Euclidean distance at any instant:
s3, unloading task representation: will arbitrarily task the vehicle v i The generated task-use triplet:
represents; wherein D i To unload the size of a task, C i Tmax, a computing resource required for a task i To the maximum tolerable delay for executing the task.
Specifically, in an embodiment of the present invention, the multi-agent multitask collaborative offloading method further includes: constructing a communication model, and acquiring the rate of a communication uploading link; the method comprises the following steps:
s1, setting a communication mode between a task vehicle and a service node: a short-range wireless communication mode; the orthogonal frequency is adopted to reduce the mutual influence in the communication process;
s2, setting an uplink in the communication process as a flat Rayleigh fading channel without considering channel interference;
s3, calculating the task vehicle v according to the Shannon formula i To the service node v j Average transmission rate of the uplink:
wherein, may be v2v or v2r, W v2v And W v2r Respectively representing V2V and V2R channel bandwidths; p i Is a task vehicle v i Of the transmitted power, p 0 Representing the white Gaussian noise power, d i，j Is v i To v j The transmission distance of (a);
represents a path loss exponent; the channel fading coefficient is denoted by h.
In an embodiment, the communication between the vehicles and the RSU adopts IEEE 802.11 protocol in short-range wireless communication mode. And orthogonal frequencies are adopted in the communication process, so that the mutual influence in the communication process can be reduced. And setting an uplink in the communication process as a flat Rayleigh fading channel without considering channel interference.
S4, selecting an unloading mode: after the node state is defined, local computation and multi-service node edge unloading computation are selected for task unloading, wherein the edge unloading computation comprises single-hop unloading and two-hop unloading.
In the embodiment of the present application, there are 4 ways to process tasks: local processing, wherein the method comprises the steps of unloading to an MEC server of an RUS in a single-hop to Infrastructure (V2I) mode, unloading to an adjacent idle service Vehicle in a single-hop to V2V (V2V) mode and unloading to an idle service Vehicle of a relay node in a two-hop mode; the following three offloading methods are edge offload computation.
Specifically, in the embodiment of the present invention, the defining node state specifically includes: use of0-1 variable α i ，j Representing said single-hop off-load mission vehicle v i Of a serving node v within communication range j The state of (2):
using the variable beta from 0 to 1 j，k Indicating a relay node v performing two-hop offloading j Service node v k The state of (2):
s5, executing an unloading task: on the basis of selecting edge unloading calculation, judging v in task vehicle TaV i Service node v j And said serving node v, and j satisfying a requirement to act as the relay node; then, based on the result judgment of the idle state, executing the unloading task in the single-hop unloading mode; if the idle state does not meet the single-hop unloading condition, unloading a calculation unloading task to the relay node v through the two hops j Idle service node v k Is executed.
In this embodiment, task vehicle TaV may not only process tasks locally, but also may utilize the same to take full advantage of the vehicle's computing resources
The SN in TaV communication range is always in time. Tasks can be handled in 2 ways, local computation and edge offload computation.
1) Computation offload with local computation:
when TaV chooses to process the task using its own local computing resources, the local computing latency is:
wherein TaV has a local computing power of
Representing the computational load of the local task.
2) Edge offload computation
When TaV is performing the service node offload computation, the following stages are involved: and uploading the task, calculating the task and feeding back the result. In the result feedback stage, the downlink transmission speed of the task is much higher than the uplink speed, so the delay of result feedback is ignored in this embodiment. Suppose u of TaV i Selecting an offloaded set of service nodes VS i The time delays of different stages are analyzed respectively:
case 1: task vehicle v i Service node v j In idle state, i.e. alpha i，j And =0. In this case, the offloading of tasks to the service node may compute the task directly:
wherein,
for task vehicles v i To the service node v j A distributed sub-computation task; r is i，j Is v is i Transmitting a computing task to v j The upload rate of (c); mu.s 1 The overlapping factor in the communication process represents the proportion of data which can be unloaded to the edge node;
representing a service node v j For task vehicle v i The calculated force of the distribution.
Case 2: task vehicle v i Service node v j In the occupied state, i.e. alpha i，j And =1. This is due to the fact that the service node v j Has been selected as a serving node by the remaining TaV, in this case v j Vehicle v for mission only i Candidate relay node CrN of. Suppose v k Is always at v j Within a communication range of
V is then j The vehicle can be used as a relay node to execute unloading calculation tasks in a two-hop unloading mode, namely beta j，k ＝0。
Wherein,
denotes v i Will calculate the subtask
Upload to relay node v j ；
Indicating a relay node v j Will calculate the task
Two-hop offload to relay node v j Service node v k 。
Specifically, in the embodiment of the present invention, the general expression of the multitask unloading is:
based on the method, when contradiction between limited computing resources and user experience occurs, the multi-hop distributed unloading is adopted to execute the parallel computing task through local execution and multi-service node multi-hop distributed unloading, and the overall time delay of the multi-hop distributed unloading system is better than that of a traditional single-hop single-node unloading system. The objective function is designed as follows:
wherein, C represents a calculation sub-task with unequal splitting of different Vs, vrn and Vrs.
Constraint C 1 Constraint boundaries representing local and offload computation subtasks; c 2 Indicating the maximum tolerance time limit Tmax of the task i Inner mission vehicle v i To the service node v j Relative Euclidean distance therebetween is less than or equal to v i The communication range R of (2); c 3 Indicating that the sum of local and multi-service node offload computation subtasks equals task vehicle v i Total calculation task of
In step S2, not all vehicles on the three-dimensional road can be used as service nodes of the task vehicle, considering that the vehicle position is dynamically changed, so that the network topology is also changed, and the communication range of the task vehicle is limited. Therefore, the clustering algorithm is used for determining the service node set and the relay nodes of vehicles with different tasks, and the two-hop service node.
Specifically, in the real-time method of the present invention, in the step S2 of determining the node type, the method further includes: clustering the running roads of the vehicles based on the task vehicles, wherein N +1 edge servers are provided, and M task vehicles Vt are provided, so that service node clustering of the multi-task vehicles is performed;
the service node clustering step specifically includes:
s1, from a vehicle set V = { V = 0 ，v 1 ，v 2 ，...，v N Randomly selecting M task vehicles as centroids;
s2, sampling in the service node, and calculating Euclidean distances from each sample to M centroids;
s3, if the sample is within the range of delta t epsilon from 0 to Tmax i ]Euclidean distance from any time to centroid
Entering the subsequent step; wherein R is the coverage radius of a roadside unit RSU which is arranged on a road and provides service for all equipment;
s4, if the sample is in the communication range of the centroids, alpha i，j =1 using the sample as the selected serving node; otherwise alpha i，j =0, the sample is divided into clusters corresponding to the centroid;
and S5, finally, outputting a clustering result to obtain a service node set Vs of the task vehicle.
Wherein, in order to verify the quality of the clustering degree, the similarity metric is selected to be the reciprocal of the Euclidean distance
That is, the smaller the distance between the task vehicle and the service node is, the greater the similarity between the task vehicle and the service node is, and the smaller the similarity is otherwise.
The optimized objective function is the problem of minimizing the average time delay of the system, and if the candidate service node is used as the service node of a certain task vehicle, the candidate service node can serve as a candidate relay node of other Tavs in a communication range. In order to ensure that the system delay is minimized, the number of the clustered clusters is determined by comparing TaV, namely, the selected service node is used as a TaV service node with less clusters.
For clearer expression, we assume V 0 And V 4 For the task vehicle, a and B are their respective service node sets, C = a &' B is a candidate service node, and C should be a TaV service node or an optional relay node in different cases by way of example. The specific case analysis is as follows:
case 1: as shown in fig. 3, the service node set first cluster structure: if | A | is > | B |, then V 3 Is a V 4 Service node of V 0 The selected relay node of (1).
Case 2: as shown in fig. 4, the service node set has a second cluster structure: if | A | is less than | B |, then V 3 Is a V 0 Service node of V 4 Selected relay node。
Case 3: as shown in fig. 5, the service node set has a third cluster structure: if | A | = | B |, if
Then V 3 Is a V 0 Service node of V 4 And vice versa.
Next, RN and RsN are determined from the candidate relay node CrN, and the adopted strategy is to determine whether the communication range of the candidate relay node has idle vehicles, namely beta j，k And =0. If beta is present i，k And if not, the node can not be used as a relay node to carry out two-hop unloading. Sets Vrn and Vrs can be calculated by the above method.
Specifically, in the embodiment of the present invention, before the step S3 of unloading the task representation, the method further includes the following steps: and determining an unequal task set C of the task vehicles to be distributed to the service nodes.
In order to minimize the average delay of the system, in this embodiment, the optimization objective function is converted into the absolute value of the parallel difference between the unloading of the edge node and the local computation
Minimization problem, i.e.
Closer to 0 represents higher parallelism; the transformed objective function is as follows:
then, distributing the unloading task to the service node by using a differential evolution algorithm (DE algorithm):
the DE algorithm generates population individuals by encoding with floating point vectors. In the optimization process of the DE algorithm, firstly, two individuals are selected from parent individuals to carry out vector differencing to generate a differential vector; secondly, another individual is selected to be summed with the difference vector to generate an experimental individual; then, carrying out cross operation on the parent individuals and the corresponding experimental individuals to generate new filial individuals; and finally, selecting between the parent individuals and the child individuals, and storing the individuals meeting the requirements into a next generation group.
The unloading of the multi-task vehicle only has influence on the unloading strategy and does not have influence on the task distribution, and only the task vehicle v is used i For example, a task unequal split process is illustrated. C 2 Only the offload policy is affected, and the influence is not considered in the unequal split set C of this section of the computing task. The mathematical model of the optimization problem is simplified to:
where D is the dimension of the solution space, x 1 ，x 2 ，...，x D Respectively, the allocation calculation sub-task quantities,
respectively representing the upper and lower bounds of the value range of the computation subtask. The DE algorithm flow is as follows:
1) Initializing a population: initial population
Randomly generating:
wherein x is i (0) The ith "chromosome" (or individual) representing the 0 th generation in the population; x is a radical of a fluorine atom j，i (0) The j-th "gene" representing the i-th "chromosome" of the 0 th generation; NP means population sizeSmall, rand (0,1) represents random numbers uniformly distributed in a section (0,1).
2) Mutation operation: DE realizes individual variation through a differential strategy, and the differential strategy is to randomly select two different individuals in a population, and vector synthesis is carried out on the two different individuals after vector difference of the two different individuals is scaled and then the two different individuals are subjected to vector synthesis with an individual to be varied, namely
Wherein, F is a scaling factor,
represents the ith individual in the population of the g generation.
In the evolution process, in order to ensure the validity of the solution, it is necessary to judge that each "gene" in the "chromosome" is regenerated by a random method (the same method as the generation method of the initial population).
Population of the g generation
By mutation, an intermediate is produced
3) And (3) cross operation: for the g generation population { x i (g) And variant intermediates { h } i (g + 1) } Cross-manipulations between individuals:
wherein CR is the crossover probability, j rand Is [1,2,.., D.)]Is a random integer of (a).
To ensure variant intermediates { h } i (g + 1) } at least one "gene" per "chromosome" is inherited by the next generation. The first gene to cross-operate was randomly taken h i J in (g + 1) rand The locus "gene" as the crossed "chromosome" u i (g+1) J th rand Allelic "genes". The subsequent cross operation process selects x by the cross probability CR i (g) H is to be i (g + 1) as u i (g + 1) allele.
3) Selecting operation: DE employs a greedy algorithm to select individuals for entry into the next generation population:
the specific evolution process is as follows:
(1) Determining a control parameter of a differential evolution algorithm and determining a fitness function. The control parameters of the differential evolution algorithm comprise: population size NP, scaling factor F and hybridization probability CR.
(2) An initial population is randomly generated.
(3) And evaluating the initial population, namely calculating the fitness value of each individual in the initial population.
(4) And judging whether a termination condition is reached or the evolution algebra reaches the maximum. If so, terminating the evolution, and outputting the obtained optimal individual as an optimal solution; if not, continuing.
(5) And carrying out mutation and cross operation to obtain an intermediate population.
(6) And selecting individuals from the original population and the intermediate population to obtain a new generation population.
(7) Evolution algebra g = g +1, go to step (4).
From the above, the multi-agent multi-task cooperative offloading method according to the embodiment of the present invention includes: s1, setting a vehicle task unloading scene: identifying the vehicle producing the mission as mission vehicle TaV; the task vehicle is always within a communication range within the maximum tolerable time delay, and a service node for providing unloading service for the task vehicle is configured; s2, determining the node type: service nodes belonging to the communication range of the plurality of task vehicles TaV are determined as candidate service nodes; when the candidate service node is determined to be v in task vehicle TaV i After the service node is obtained, the current candidate service node can be used as a candidate relay node of the remaining task vehicles; then fromScreening out relay nodes capable of executing two-hop unloading and service nodes of the relay nodes from the candidate relay nodes; s3, unloading task representation: will arbitrarily task the vehicle v i Generated task-use triplets:
representing; wherein D i To unload the size of a task, C i Tmaxi is the maximum tolerable delay for executing the task; s4, unloading mode selection: after defining the node state, selecting local computation and multi-service node edge unloading computation for task unloading, wherein the edge unloading computation comprises single-hop unloading and two-hop unloading; s5, executing an unloading task: on the basis of selecting edge unloading calculation, judging v in task vehicle TaV i Service node v j And said serving node v, and j satisfying a requirement to act as the relay node; then, based on the result judgment of the idle state, executing the unloading task in the single-hop unloading mode; if the idle state does not meet the single-hop unloading condition, unloading a calculation unloading task to the relay node v through the two hops j Idle service node v k Is executed. The invention provides a multi-agent multi-task cooperative unloading method which can solve the problem of multi-vehicle multi-task cooperative unloading conflict and realize minimum system average time delay by utilizing limited computing resources of task vehicles and service resources of surrounding spaces.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A multi-agent multi-task cooperative unloading method is characterized by comprising the following steps:
s1, setting a vehicle task unloading scene: identifying the vehicle producing the mission as mission vehicle TaV; the task vehicle is always within a communication range within the maximum tolerable time delay, and a service node for providing unloading service for the task vehicle is configured;
s2, determining the node type: service nodes belonging to the communication range of the plurality of task vehicles TaV are determined as candidate service nodes; when the candidate service node is determined to be v in task vehicle TaV i After the service node is obtained, the current candidate service node can be used as a candidate relay node of the rest task vehicles; then, screening out relay nodes capable of executing two-hop unloading and service nodes of the relay nodes from the candidate relay nodes;
s3, unloading task representation: will arbitrarily task the vehicle v i Generated task-use triplets:
represents; wherein D i To unload the size of a task, C i Tmax, the computational resources required for a task i Maximum tolerable latency for executing the task;
s4, unloading mode selection: after defining the node state, selecting local computation and multi-service node edge unloading computation for task unloading, wherein the edge unloading computation comprises single-hop unloading and two-hop unloading;
s5, executing an unloading task: on the basis of selecting edge unloading calculation, judging v in task vehicle TaV i Service node v j And said serving node v, and j satisfying a requirement to act as the relay node; then, based on the result judgment of the idle state, executing the unloading task in the single-hop unloading mode;
if the idle state does not meet the single-hop unloading condition, the calculation unloading task is unloaded to the relay through the two hopsNode v j Idle service node v k Is executed.
2. The multi-agent multi-tasking cooperative offloading method of claim 1, wherein in step S4, further comprising:
if the local computation is selected, namely TaV selects to use local computing resources to process tasks, the local computation time delay is directly obtained
Wherein,
representing the local computing power of TaV,
representing the computational load of the local task.
3. The multi-agent multi-task cooperative offloading method of claim 2, wherein in the step S4, the defining node state is specifically: using a variable alpha of 0-1 i,j Representing said mission vehicle v in single hop unloading i Of a serving node v within communication range j The state of (2):
using a variable beta of 0-1 j,k Indicating a relay node v when performing two-hop offloading j Service node v k The state of (1):
4. the multi-agent multi-task cooperative offloading method of claim 3, wherein in the step S5, the step of performing an offloading task is specifically:
judging v in task vehicle TaV i Service node v j Idle state of (2):
if the task vehicle v i Service node v of j In an idle state, i.e. alpha i,j =0; offloading tasks directly to service node v through single hop offloading j Calculating to obtain single-hop calculation time delay;
wherein,
for task vehicles v i To the service node v j A distributed sub-computation task; r is i,j Is v is i Transmitting a computing task to v j The upload rate of (d); mu.s 1 The overlapping factor in the communication process represents the proportion of data which can be unloaded to the edge node;
representing a service node v j For task vehicle v i The calculated force of the distribution;
if the task vehicle v i Service node v j In an occupied state, i.e. alpha i,j =1, description serving node v j Has been selected as a serving node by the remaining TaV; at a service node v j Under the condition of meeting the requirement of serving as a relay node, the node v j Serving node v offloaded to its idle as a relay node over two hops k Performing off-load computing tasks, i.e. beta j,k =0; then obtain twoJump calculation delay:
wherein,
denotes v i Will calculate the subtask
Upload to relay node v j ；
Indicating a relay node v j Will calculate the task
Two-hop offload to relay node v j Service node v k (ii) a Vt is represented as a set of mission vehicles TaV; vs is represented as a collection of service nodes.
5. The multi-agent multi-task cooperative offloading method of claim 3, wherein in step S2, after completing the determining the node type, further comprising: constructing a vehicle motion model, passing a mission vehicle v α Position information at time t
Predicting position information at t + delta t moment
6. The multi-agent multi-task collaborative offloading method of claim 5, wherein a vehicle motion model is constructed and location information is predicted
The method specifically comprises the following steps:
s1, obtaining task vehicle v α Velocity at time t
The included angles with the x-axis, the y-axis and the z-axis are theta x ,θ y ,θ z ∈[0,π]；
S2, according to the speed
And included angle acquisition speed
Velocity components in the x, y and z axes;
s3, obtaining a vehicle running delta t through the direction cosine law, and then obtaining a task vehicle v α Distance components traveled in the x, y and z directions;
cos is obtained by the direction cosine theorem 2 θ x +cos 2 θ y +cos 2 θ z =1, i.e.:
s4, subsequent application of L 2 Calculating norm to obtain task vehicle v α To v β Euclidean distance at any instant:
namely that
7. The multi-agent multitask coordinated offload method according to claim 6, wherein in said step S2 determining node type, further comprising: clustering the running roads of the vehicles based on the task vehicles, wherein N +1 edge servers are provided, and M task vehicles Vt are provided, so that service node clustering of the multi-task vehicles is performed;
wherein the step of clustering the service nodes specifically comprises:
s1, from a vehicle set V = { V = 0 ,v 1 ,v 2 ,…,v N Randomly selecting M task vehicles as mass centers;
s2, sampling in the service node, and calculating Euclidean distances from each sample to M centroids;
s3, if the sample is within the range of delta t epsilon from 0 to Tmax i ]Euclidean distance from any time to centroid
Entering the subsequent step; wherein R is the coverage radius of a roadside unit RSU which is arranged on a road and provides service for all equipment;
s4, if the sample is in the communication range of the centroids, alpha i,j =1 using the sample as the selected serving node; otherwise alpha i,j =0, the sample is divided into clusters corresponding to the centroid;
and S5, finally, outputting a clustering result to obtain a service node set Vs of the task vehicle.
8. The multi-agent multi-task cooperative offloading method of claim 1, further comprising, before the offloading mode selection in step S4: constructing a communication model, and acquiring the rate of a communication uploading link; the method comprises the following steps:
s1, setting a communication mode between a task vehicle and a service node: a short-range wireless communication mode; orthogonal frequencies are adopted to reduce the mutual influence in the communication process;
s2, setting an uplink in the communication process as a flat Rayleigh fading channel without considering channel interference;
s3, calculating the task vehicle v according to the Shannon formula i To the service node v j Average transmission rate of the uplink:
wherein, v2v or v2r, W v2v And W v2r Respectively representing V2V and V2R channel bandwidths; p is i Is a task vehicle v i Of the transmitted power, p 0 Representing the white Gaussian noise power, d i,j Is v i To v j The transmission distance of (a); θ represents a path loss exponent; the channel fading coefficient is denoted by h.
9. The multi-agent multi-task collaborative offload method according to claim 6, further comprising, before offloading the task representation in the step S3: and determining an unequal task set C of the task vehicle distributed to the service nodes.
10. The multi-agent multi-task cooperative work method according to claim 9, wherein the specific step of task assignment comprises:
s1, setting an objective function, wherein the objective function aims to solve a calculation time delay problem of multi-vehicle multi-task unloading;
s2, converting the target function into an absolute value of a parallel difference between unloading of the edge node and local computation
Minimization problem, i.e.
Closer to 0 represents higher parallelism; the transformed objective function is:
st.
C 1 :
C 2 :
C 3 :
wherein, C represents a computation subtask with unequal splitting; constraint C 1 Representing constraint boundaries for local and offload computation subtasks; c 2 Indicating the maximum tolerance time limit Tmax of the task i Inner mission vehicle v i To the service node v j Relative Euclidean distance therebetween is less than or equal to v i The communication range R of (2); c 3 Indicating that the sum of local and multi-service node offload computation subtasks equals task vehicle v i Total calculation task of
And S3, distributing the unloading task to the service node by utilizing a differential evolution algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211016816.4A CN115550357A (en) | 2022-08-24 | 2022-08-24 | Multi-agent multi-task cooperative unloading method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211016816.4A CN115550357A (en) | 2022-08-24 | 2022-08-24 | Multi-agent multi-task cooperative unloading method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115550357A true CN115550357A (en) | 2022-12-30 |
Family
ID=84725292
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211016816.4A Pending CN115550357A (en) | 2022-08-24 | 2022-08-24 | Multi-agent multi-task cooperative unloading method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115550357A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115988462A (en) * | 2023-03-17 | 2023-04-18 | 中电建市政建设集团山东工程有限公司 | Debugging method of edge computing module based on vehicle-road cooperation |
-
2022
- 2022-08-24 CN CN202211016816.4A patent/CN115550357A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115988462A (en) * | 2023-03-17 | 2023-04-18 | 中电建市政建设集团山东工程有限公司 | Debugging method of edge computing module based on vehicle-road cooperation |
| CN115988462B (en) * | 2023-03-17 | 2023-06-30 | 中电建市政建设集团山东工程有限公司 | Debugging method of edge computing module based on vehicle-road cooperation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109756378B (en) | Intelligent computing unloading method under vehicle-mounted network | |
| Yang et al. | A parallel intelligence-driven resource scheduling scheme for digital twins-based intelligent vehicular systems | |
| Chen et al. | Efficiency and fairness oriented dynamic task offloading in internet of vehicles | |
| CN111445111B (en) | Electric power Internet of things task allocation method based on edge cooperation | |
| Ahmed et al. | A survey on vehicular task offloading: Classification, issues, and challenges | |
| CN108777852A (en) | A kind of car networking content edge discharging method, mobile resources distribution system | |
| CN112188442A (en) | Vehicle networking data-driven task unloading system and method based on mobile edge calculation | |
| Zhao et al. | A digital twin-assisted intelligent partial offloading approach for vehicular edge computing | |
| CN115209426B (en) | Dynamic deployment method for digital twin servers in edge car networking | |
| CN113891477A (en) | Resource allocation method based on MEC calculation task unloading in Internet of vehicles | |
| CN112511614B (en) | Intelligent networking vehicle task unloading method under vehicle-mounted edge environment based on reinforcement learning | |
| Lan et al. | Deep reinforcement learning for computation offloading and caching in fog-based vehicular networks | |
| Li et al. | DNN partition and offloading strategy with improved particle swarm genetic algorithm in VEC | |
| CN113163365A (en) | Unmanned aerial vehicle support networking resource optimization method based on alternating direction multiplier algorithm | |
| CN112115505A (en) | New energy automobile charging station charging data transmission method based on mobile edge calculation and block chain technology | |
| Hou et al. | Hierarchical task offloading for vehicular fog computing based on multi-agent deep reinforcement learning | |
| Wei et al. | OCVC: An overlapping-enabled cooperative vehicular fog computing protocol | |
| CN115550357A (en) | Multi-agent multi-task cooperative unloading method | |
| CN115629873A (en) | System and method for controlling unloading of vehicle-road cloud cooperative tasks and stability of task queue | |
| CN117221951A (en) | Task unloading method based on deep reinforcement learning in vehicle-mounted edge environment | |
| CN114980029A (en) | Unloading method based on task relevance in Internet of vehicles | |
| Wang et al. | Joint offloading decision and resource allocation in vehicular edge computing networks | |
| Gao et al. | Fast Adaptive Task Offloading and Resource Allocation in Large-Scale MEC Systems via Multiagent Graph Reinforcement Learning | |
| CN113821346A (en) | Computation uninstalling and resource management method in edge computation based on deep reinforcement learning | |
| CN116916272A (en) | Resource allocation and task unloading method and system based on automatic driving automobile network |
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 |