CN108415760B - Crowd sourcing calculation online task allocation method based on mobile opportunity network - Google Patents
Crowd sourcing calculation online task allocation method based on mobile opportunity network Download PDFInfo
- Publication number
- CN108415760B CN108415760B CN201810083243.4A CN201810083243A CN108415760B CN 108415760 B CN108415760 B CN 108415760B CN 201810083243 A CN201810083243 A CN 201810083243A CN 108415760 B CN108415760 B CN 108415760B
- Authority
- CN
- China
- Prior art keywords
- task
- time
- tasks
- user
- candidate user
- 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 45
- 238000012358 sourcing Methods 0.000 title claims abstract description 27
- 238000004364 calculation method Methods 0.000 title claims abstract description 20
- 230000008569 process Effects 0.000 claims abstract description 6
- 238000012163 sequencing technique Methods 0.000 claims description 8
- 238000005457 optimization Methods 0.000 claims description 5
- 230000001419 dependent effect Effects 0.000 claims description 2
- 238000004088 simulation Methods 0.000 abstract description 4
- 238000004422 calculation algorithm Methods 0.000 description 6
- 238000004891 communication Methods 0.000 description 3
- 230000008447 perception Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005295 random walk Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
- G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
- G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a crowd-sourcing calculation online task allocation method based on a mobile opportunity network, which aims at the two problems of minimizing the average feedback time of an independent task and minimizing the longest feedback time of a cooperative task respectively, combines the user encounter rule and the calculation capability difference, orders the tasks to be allocated according to the average execution time, executes a virtual offline global allocation method once when a task requester encounters one user, repeatedly calculates and updates the estimated feedback time of each user to each task in real time, and allocates the tasks to the users with the minimum feedback time in sequence, but only the tasks belonging to the current encountered users in the result are actually allocated. When meeting the next user, the above process is repeated until all task assignments are completed. Theoretical analysis and simulation results prove that the method provided by the invention can minimize the task completion time under the same task and user scale, improve the efficiency and have strong practical value in a crowd-sourcing calculation scene based on the mobile opportunity network.
Description
Technical Field
The invention relates to a crowd sourcing calculation online task allocation method based on a mobile opportunity network, and belongs to the technical field of mobile networks and crowd sourcing calculation.
Background
The idea of crowd-sourcing and crowd-sourcing is derived from crowd-sourcing and crowd-sourcing perception, and is a distributed problem solving mode performed by using mobile users and intelligent devices thereof. The popularization and rapid development of mobile devices broadens the application scene and the realization scale of crowd-sourcing computing, users in a network can participate in the perception, calculation and data distribution of tasks at any time and any place through rich sensors (GPS, camera, accelerometer, compass and the like) and strong storage and calculation capabilities which are built in a handheld device, and the large-scale complex problem that machines or individuals are difficult to complete is completed through reasonable cooperation and sharing.
At present, research on crowd-sourcing calculation mostly considers a model based on a dispersed position, namely, a user needs to move to some fixed positions to complete perception tasks, such as environment and traffic condition monitoring, and the like, receiving and transmitting of task information and result feedback are transmitted through a 3G/4G network, and tasks are guaranteed to be performed in order through a reasonable excitation mechanism and a task allocation algorithm. With the development of wireless communication technology, the crowd-sourcing computing system loaded in the mobile social network allows users to realize near-distance communication through WiFi, Bluetooth and D2D, facilitates large-scale data transmission, saves communication cost, has higher safety and reliability based on opportunistic meeting user cooperation, and can be better applied to large-scale crowd-sourcing computing scenes. The mobile users as the nodes in the network for collaborative distributed computing have the advantages of convenient deployment, flexibility, low cost and the like, and the information carried by the users also has rich mining value, so that the mobile users become a new mode worthy of exploration in the future mobile internet. How to reasonably allocate tasks and schedule resources in a complex crowd-sourcing computing system with multi-task and multi-user random walk by combining with user movement rules and computing differences to ensure that the tasks are finished orderly and efficiently is the problem and innovative contribution of the invention.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the task requester finds the task subset which is most suitable for the user according to the current situation and immediately distributes the task subset when meeting one user, thereby ensuring the efficient completion of the task.
The invention adopts the following technical scheme for solving the technical problems:
a crowd sourcing computation online task allocation method based on a mobile opportunity network comprises the following steps:
step 1, initializing a task J to be distributed, which is issued by a task requester, as J1,j2,...,jm,...,jMWhere M is 1, …, M, jmRepresenting the mth task to be distributed, wherein M is the total number of the tasks to be distributed; according to the type of the tasks to be distributed and the optimization target, sequencing the tasks to be distributed according to average execution time to obtain an ordered task list;
step 2, initializing candidate users U ═ U1,u2,...,un,...,uNWhere N is 1, …, N, unRepresenting the nth candidate user willing to participate in the crowd sourcing calculation task, wherein N is the total number of the candidate users; calculating the meeting parameter [ lambda ] of each candidate user and the task requester according to the historical transaction records1,λ2,...,λn,...,λNAnd historical task execution time ratio of each candidate user { p }1,p2,...,pn,...,pNIn which λ isnAnd pnRespectively representing the encountering parameters and the historical task execution time ratio of the nth candidate user and the task requester;
step 3, when the task requester and a candidate user u in the mobile opportunity networknWhen meeting, the user is subjected to online task allocation, which specifically comprises the following steps:
3-1, executing an offline global virtual allocation method, starting from the first task in the ordered task list obtained in the step 1, calculating the estimated feedback time of all candidate users to the task, sequencing the tasks in a sequence from small to large, allocating the first task to the user with the minimum current estimated feedback time, updating the execution waiting time of the user, recalculating the estimated feedback time of all candidate users to the second task, sequencing the tasks in a sequence from small to large, allocating the second task to the user with the minimum current estimated feedback time, and so on until all tasks are virtually allocated;
3-2, belonging to the candidate user u in the distribution result of the step 3-1nIs actually assigned to the candidate user unExecuting, and not distributing other tasks; marking the tasks and users which are actually distributed, and deleting the tasks and the users from the ordered task list and the candidate users;
and 4, when the task requester meets other candidate users, repeating the process of the step 3 until all the tasks to be distributed are really distributed.
As a preferred scheme of the present invention, step 1, according to the type of the task to be allocated and the optimization goal, the tasks to be allocated are sorted according to the average execution time, specifically:
for tasks to be distributed which are independent from each other, namely the feedback time of each task is not related to each other, the average feedback time is optimized:
where M is the total number of tasks to be assigned, FT (j)m) For the mth task j to be allocatedmThe estimated feedback time, II is the final task allocation formulaUse { tau ] as a measure1,τ2,...,τm,...,τMDenotes the average execution time of the M tasks, τmThe average execution time of the mth task to be distributed is obtained based on the experience of the task requester and the historical performance of the user, and the tasks to be distributed are sequenced from small to large according to the average execution time during initialization, namely:
J＝{j1,j2,...,jM:τ1≤τ2≤...≤τM}
for the tasks to be distributed which are mutually dependent, namely the last task is fed back successfully, the whole task is completed, and the longest feedback time is optimized:
LT(∏)＝Max{FT(j1)|∏,FT(j2)|∏,...,FT(jm),...,FT(jM)|∏},m＝1,…,M
when initializing, the tasks to be distributed are sequenced from large to small according to the average execution time, namely:
J＝{j1,j2,...,jM:τ1≥τ2≥...≥τM}。
as a preferred scheme of the invention, the meeting parameter lambda of the nth candidate user and the task requester in the step 2nAnd historical task execution time ratio pnThe calculation method comprises the following steps:
adopting a random-based mobile model, and considering the meeting time compliance parameter of the nth candidate user and the task requester as lambdanIs integrated over time t to obtain the expected encounter time deltanComprises the following steps:
according to the history encounter data, if the nth candidate user and the task requester encounter for L times within the time T, then:
λn＝LT
wherein, tnmIs a candidate user unPast execution task jmTime of (e), tnmRepresentative candidate user unTotal time consumed by all tasks participating in execution in the past, τ m being task jmIs carried out on the average of the execution time, sigma taumRepresenting the average total time consumed by all persons to perform all tasks that the candidate user has engaged in performing in the past.
As a preferred scheme of the present invention, the estimating feedback time in step 3 specifically includes:
(1) task requester waits for a candidate user unTime delta of first meeting and transmitting task informationnWhen the two have met, δ n0; when the two have not met, δn＝1/λn，λnIs a candidate user unAn encounter parameter with a task requester;
(2) candidate user unFor task jmIs TnmAverage execution time tau from historical tasksmAnd the user task execution time ratio pnCandidate user unFor task jmThe estimated execution time is:
if the candidate user unIn executing task jmIf there is a preceding unfinished task in the previous hand, the part of time also includes the time for executing the preceding task, i.e. task jmThe execution waiting time of the user u is obtained by summationnFor task jmThe estimated completion time is:
wherein, Σ tnmRepresentative candidate user unThe total time consumed by all tasks participating in execution in the past, ΣτmRepresents the average total time consumed by all persons performing all tasks that the candidate user has engaged in performing in the past,
as task jmExecution latency of (1);
(3) time delta for task requester to wait for candidate user to return resultn' this action needs to be done when the two meet again after the task is completed, this time is the expected time waiting for the last meeting minus the time taken to execute the task in this period, namely:
δn'＝1/λn-Tnm％(1/λn)
therefore, candidate user treats task j to be distributedmThe estimated feedback time is as follows: FTn(m)＝δn+Tnm+δn'。
As a preferred scheme of the present invention, the specific process of step 4 is as follows:
and executing an offline global virtual allocation method once when the task requester meets a candidate user, immediately allocating the most appropriate task for the meeting candidate user according to the current situation, and gradually approaching to the optimal state until the real allocation of all the tasks to be allocated is completed.
Compared with the prior art, the invention adopting the technical scheme has the following technical effects:
1. the invention provides a reasonable online task allocation method aiming at the problems of minimizing the average feedback time of an independent task and minimizing the longest feedback time of a cooperative task in crowd-sourcing calculation and combining the user meeting probability and the difference of the calculation capacity of the user to the task. And executing an off-line virtual global distribution method once when the task requester meets one user, traversing all tasks to be distributed, searching the task which is most suitable for the user according to the current condition, and immediately distributing.
2. The method can ensure that the task is finished and returned at the fastest speed under the same problem scale, improves the overall execution efficiency, and has high practical value in a crowd-sourcing calculation scene based on the mobile opportunity network.
Drawings
FIG. 1 is a flow chart of the crowd sourcing computation online task allocation method based on the mobile opportunity network of the present invention.
FIG. 2 is a graph comparing the performance of an embodiment of the method of the present invention with other methods at different numbers of tasks.
Fig. 3 is a graph comparing the performance of the method embodiment of the present invention with other methods for different numbers of users.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.
The invention provides a crowd-sourcing calculation online task allocation method based on a mobile opportunity network, which aims at the problems of minimizing the average feedback time of an independent task and minimizing the longest feedback time of a cooperative task respectively and provides a reasonable online task allocation method by combining the user encounter probability and the calculation capability difference of the user to the task. Under the method, each time a task requester meets one user, the task subset which is most suitable for the user is searched according to the current situation and immediately distributed, and the task is ensured to be efficiently completed. The whole frame is shown in figure 1, and comprises the following steps:
step 1: initialization task information J ═ J1,j2,...,jm,...,jMWhere j ismRepresenting the mth task, and M is the total number of tasks to be distributed. According to the task type and the optimization target, sorting the tasks to be distributed according to the average execution time;
in the embodiment, the crowd sourcing computing task is assumed to be downloading and processing of multiple videos, and a user with better network conditions assists a task requester to download multiple video segments of different types and perform annotation processing. The feedback time of each task is not related to each other, and the average feedback time is optimized:
where M is the total number of tasks, FT (j)m) For the estimated feedback time of the mth task, (J) for the final task assignment scheme1,J2,...,Jn,...,JN}，JnRepresenting the sequential subset of tasks assigned to the nth user. By { τ1,τ2,...,τm,...,τMDenotes the average execution time of the M tasks, τmThe average execution time of the mth task is obtained based on the self experience of the task requester and the historical performance of the user, and the tasks to be distributed are sequenced from small to large according to the average execution time during initialization, namely:
J＝{j1,j2,...,jM:τ1≤τ2≤...≤τM}
step 2: initializing user information U ═ U0,u1,...,un,...,uNIn which u0Being a task requester, u1...uNFor candidate users willing to participate in a crowd-sourcing computing task, unRepresenting the nth candidate user, wherein N is the total number of the candidate users. According to the stored historical transaction records, the meeting parameters of each user and the task requester in the computer opportunity network are { lambda1,λ2,...,λn,...,λNThe historical task execution time ratio of the users p1,p2,...,pn,...,pN}. Corresponding, λnAnd pnRespectively representing the encounter parameter and the historical task execution time ratio of the nth user and the task requester. If two user nodes meet L times within time T, then:
λn＝L/T
the expected meeting time of the two is:
historical task completion time ratio represents user unThe actual time consumed during the execution of the historical taskThe average ratio is calculated by the following steps of (1) the ratio of the average levels, due to the difference among the network speed owned by the user, the hardware condition of the mobile equipment, the participation enthusiasm and the reliability of the user, and the like:
wherein, tnmFor user unPast execution task jmTime of (e), tnmRepresentative candidate user unThe total time consumed by all tasks participating in execution in the past, ∑ τmRepresenting the average total time consumed by all persons to perform all tasks that the candidate user has engaged in performing in the past.
And step 3: when task requester and a user unWhen meeting in the opportunity network, performing online task allocation on the user;
step 3-1: and executing an offline global virtual allocation method, starting from the first task in the ordered task list, calculating the estimated feedback time of all users in the current opportunity network to the task, sequencing the tasks according to the sequence from small to large, allocating the first task to the user with the minimum estimated feedback time at present, and updating the execution waiting time of the user. And calculating and sequencing the estimated feedback time of all the users to the second task, distributing the second task to the user with the minimum estimated feedback time, and so on until all the tasks are virtually distributed.
The estimated feedback time of each task comprises three parts:
(1) task requester waits for a user unTime delta of first meeting and transmitting task informationnAccording to the exponential model mentioned above, δ when the two have metn0; when the two have not met, δn＝1/λn；
(2) User unFor task jmIs estimated to be the completion time Tnm. Average execution time tau according to historical tasksmAnd the user task execution time ratio pnUser unFor task jmIs predicted to executeThe method comprises the following steps:
if the user is performing task jmIf there is a preceding unfinished task in the previous hand, the part of time also includes the time for executing the preceding task, i.e. task jmExecution latency of (1). Summing to obtain user unFor task jmThe estimated completion time is:
(3) time delta for task requester to wait for result returned by usern'The action needs to be performed when the two meet again after the task is completed, and the time is the expected time for waiting for the last meeting minus the time occupied by executing the task in the period of time, namely:
δn'＝1/λn-Tnm％(1/λn)
wherein,
estimate completion time, T, for the last part of the calculated tasknm％(1/λn) Is TnmFor 1/lambdanA remainder is taken to mean the time taken to execute the task during the last encountered wait time.
So from the user's perspective, for a subset of tasks that have been assigned a value of jn1,jn2,...,jnkU users ofnThe estimated feedback time of the next task x to be distributed is as follows: FTn(x)＝δn+Tnx+δn'。
More specifically, assume that the task requester first interacts with user u1And (4) meeting, traversing each task to perform virtual global distribution. For task j1Calculating the current FT of all usersn(1) And the sequence is as follows:
FTn(1)＝{FT1(1),FT2(1),...,FTN(1)；FT1(1)≤FT2(1)≤...≤FTN(1)}
wherein FTn(1) The calculation formula of (2) is as follows:
will task j1Allocating to user u with minimum estimated feedback time1Update u1Execution latency for subsequent tasks:
T1m＝t11+t1m
for task j2Recalculating FT for all usersn(2) And ordering to continue with task j2Is assigned to FTn(2) The smallest users until all tasks are virtually allocated.
Step 3-2: belonging to the user u in the distribution result of the last stepnIs actually assigned to user unAnd executing, and not distributing other tasks. And marking the tasks and users which are actually distributed, and deleting the tasks and users from the task and user list.
And 4, step 4: when the task requester meets other users, the process of step 3 is repeated until all tasks are actually distributed.
In step 4, the task requester executes an offline virtual allocation method every time when meeting with a user, immediately allocates the most appropriate task to the meeting user according to the current situation, and gradually approaches to the optimal state until all tasks are really allocated.
In this embodiment, the comparison algorithms implemented by simulation are respectively: the method is based on a simple time series-user homogeneity task allocation algorithm (SFSU), a classic Waterfilling task allocation algorithm (namely a random allocation algorithm WFDU) in the mobile social network, and a reverse-ordering large task priority allocation algorithm (LFDU). The meeting probability parameter lambda of each user obeys the random distribution in [0,0.1], the user task execution time ratio p obeys the random distribution in [0.5,1.5], and the task average execution time tau obeys the random distribution in [0,50 ]. Fig. 2 is a diagram showing the performance of the embodiment of the method (SFDU) of the present invention compared with other methods when the number of users N is 100 and the number of tasks M is {100,200, …,1000}, and also showing the influence trend of the number of tasks on the simulation result; fig. 3 is a graph showing the performance of an embodiment of the method of the present invention compared with other methods when the number of tasks M is 1000 and the number of users N is {50,100, …,500}, and also shows the influence trend of the number of users on the simulation result. It can be seen that under the same conditions, the time gain of the Online (Online) method is generally superior to that of the corresponding Offline (Offline) method, and the method provided by the invention is also superior to other existing distribution methods.
The above embodiments are only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited thereby, and any modifications made on the basis of the technical scheme according to the technical idea of the present invention fall within the protection scope of the present invention.
Claims (4)
1. A crowd sourcing computation online task allocation method based on a mobile opportunity network is characterized by comprising the following steps:
step 1, initializing a task J to be distributed, which is issued by a task requester, as J1,j2,...,jm,...,jMWhere M is 1, …, M, jmRepresenting the mth task to be distributed, wherein M is the total number of the tasks to be distributed; according to the type of the tasks to be distributed and the optimization target, sequencing the tasks to be distributed according to average execution time to obtain an ordered task list;
step 2, initializing candidate users U ═ U1,u2,...,un,...,uNWhere N is 1, …, N, unRepresenting the nth candidate user willing to participate in the crowd sourcing calculation task, wherein N is the total number of the candidate users; calculating the meeting parameter [ lambda ] of each candidate user and the task requester according to the historical transaction records1,λ2,...,λn,...,λNAnd historical task execution time ratio of each candidate user { p }1,p2,...,pn,...,pNIn which λ isnAnd pnRespectively representMeeting parameters of the n candidate users and the task requester and historical task execution time ratio;
step 3, when the task requester and a candidate user u in the mobile opportunity networknWhen meeting, the user is subjected to online task allocation, which specifically comprises the following steps:
3-1, executing an offline global virtual allocation method, starting from the first task in the ordered task list obtained in the step 1, calculating the estimated feedback time of all candidate users to the task, sequencing the tasks in a sequence from small to large, allocating the first task to the user with the minimum current estimated feedback time, updating the execution waiting time of the user, recalculating the estimated feedback time of all candidate users to the second task, sequencing the tasks in a sequence from small to large, allocating the second task to the user with the minimum current estimated feedback time, and so on until all tasks are virtually allocated;
3-2, belonging to the candidate user u in the distribution result of the step 3-1nIs actually assigned to the candidate user unExecuting, and not distributing other tasks; marking the tasks and users which are actually distributed, and deleting the tasks and the users from the ordered task list and the candidate users;
the estimating the feedback time specifically comprises:
(1) task requester waits for a candidate user unInitial encounter, expected encounter time deltanWhen the two have met, δn0; when the two have not met, δn＝1/λn，λnIs a candidate user unAn encounter parameter with a task requester;
(2) candidate user unFor task jmIs TnmAverage execution time tau from historical tasksmAnd historical task execution time ratio pnCandidate user unFor task jmThe estimated execution time is:
if the candidate user unIn executing task jmIf there is a task before the completion of the previous task, the completion time T is estimatednmAlso including the time to perform the preceding task, i.e. task jmThe execution waiting time of the user u is obtained by summationnFor task jmThe estimated completion time is:
wherein, Σ tnmRepresentative candidate user unThe total time consumed by all tasks participating in execution in the past, ∑ τmRepresents the average total time consumed by all persons performing all tasks that the candidate user has engaged in performing in the past,
as task jmExecution latency of (1);
(3) time delta for task requester to wait for candidate user to return resultn'Waiting for the candidate user to return the result needs to be performed when the two meet again after the task is completed, where the time is the time occupied by executing the task within the expected time of waiting for the last meeting minus the expected time of waiting for the last meeting, that is:
δn'＝1/λn-Tnm％(1/λn)
therefore, candidate user treats task j to be distributedmThe estimated feedback time is as follows: FTn(m)＝δn+Tnm+δn'；
And 4, when the task requester meets other candidate users, repeating the process of the step 3 until all the tasks to be distributed are really distributed.
2. The mobile opportunistic network-based crowd computing online task allocation method according to claim 1, wherein the step 1 of sorting the tasks to be allocated according to the types of the tasks to be allocated and the optimization objective by average execution time specifically comprises:
for tasks to be distributed which are independent from each other, namely the feedback time of each task is not related to each other, the average feedback time is optimized:
where M is the total number of tasks to be assigned, FT (j)m) For the mth task j to be allocatedmPredicted feedback time of (n) is the final task allocation scheme, using { τ }1,τ2,...,τm,...,τMDenotes the average execution time of the M tasks, τmThe average execution time of the mth task to be distributed is obtained based on the experience of the task requester and the historical performance of the user, and the tasks to be distributed are sequenced from small to large according to the average execution time during initialization, namely:
J＝{j1,j2,...,jM:τ1≤τ2≤...≤τM}
for the tasks to be distributed which are mutually dependent, namely the last task is fed back successfully, the whole task is completed, and the longest feedback time is optimized:
LT(Π)＝Max{FT(j1)|Π,FT(j2)|Π,...,FT(jm),...,FT(jM)|Π},m＝1,…,M
when initializing, the tasks to be distributed are sequenced from large to small according to the average execution time, namely:
J＝{j1,j2,...,jM:τ1≥τ2≥...≥τM}。
3. the mobile opportunistic network based crowd-sourcing computing online task distribution method according to claim 1, wherein step 2 is performed based on an encounter parameter λ between the nth candidate user and the task requesternAnd historical task execution time ratio pnThe calculation method comprises the following steps:
considering the nth candidate by using a random-based motion modelThe meeting time obeying parameter of the user and the task requester is lambdanIs integrated over time t to obtain the expected encounter time deltanComprises the following steps:
according to the history encounter data, if the nth candidate user and the task requester encounter for L times within the time T, then:
λn＝L/T
wherein, tnmIs a candidate user unPast execution task jmTime of (e), tnmRepresentative candidate user unTotal time consumed by all tasks participating in execution in the past, τmAs task jmIs carried out on the average of the execution time, sigma taumRepresenting the average total time consumed by all persons to perform all tasks that the candidate user has engaged in performing in the past.
4. The mobile opportunistic network-based crowd sourcing computing online task allocation method according to claim 1, wherein the specific process of step 4 is as follows:
and executing an offline global virtual allocation method once when the task requester meets a candidate user, immediately allocating the tasks allocated according to the offline global virtual allocation method for the meeting candidate user according to the current situation, and gradually approaching to the optimal state until all the tasks to be allocated are really allocated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810083243.4A CN108415760B (en) | 2018-01-29 | 2018-01-29 | Crowd sourcing calculation online task allocation method based on mobile opportunity network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810083243.4A CN108415760B (en) | 2018-01-29 | 2018-01-29 | Crowd sourcing calculation online task allocation method based on mobile opportunity network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108415760A CN108415760A (en) | 2018-08-17 |
| CN108415760B true CN108415760B (en) | 2021-11-30 |
Family
ID=63126554
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810083243.4A Active CN108415760B (en) | 2018-01-29 | 2018-01-29 | Crowd sourcing calculation online task allocation method based on mobile opportunity network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108415760B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111353574B (en) * | 2019-11-04 | 2021-04-30 | 清华大学 | Group intelligent level calculation method and system based on action relation network |
| CN112000485B (en) * | 2020-09-01 | 2024-01-12 | 北京元心科技有限公司 | Task allocation method, device, electronic equipment and computer readable storage medium |
| CN112363822B (en) * | 2020-09-22 | 2024-02-13 | 长沙理工大学 | Collaborative task allocation method for crowdsourcing platform |
| CN112732433A (en) * | 2021-03-30 | 2021-04-30 | 骊阳(广东)节能科技股份有限公司 | Data processing system capable of carrying out priority allocation |
| CN118396358A (en) * | 2024-07-01 | 2024-07-26 | 国科恒泰（北京）医疗科技股份有限公司 | Method, device, equipment and medium for processing tasks to be handled in medical instrument system |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103731844B (en) * | 2014-01-03 | 2015-05-13 | 苏州大学 | Method for distributing tasks of mobile perception system relevant to geographic positions |
| IN2014MU00042A (en) * | 2014-02-06 | 2015-09-25 | Tata Consultancy Services Ltd | |
| CN105183543B (en) * | 2015-08-28 | 2018-07-10 | 中国科学技术大学苏州研究院 | A kind of gunz based on mobile social networking calculates online method for allocating tasks |
| CN107066322B (en) * | 2017-02-28 | 2018-02-27 | 吉林大学 | A kind of online task allocating method towards self-organizing intelligent perception system |
-
2018
- 2018-01-29 CN CN201810083243.4A patent/CN108415760B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN108415760A (en) | 2018-08-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108415760B (en) | Crowd sourcing calculation online task allocation method based on mobile opportunity network | |
| Wang et al. | Computation offloading in multi-access edge computing using a deep sequential model based on reinforcement learning | |
| Goudarzi et al. | A distributed deep reinforcement learning technique for application placement in edge and fog computing environments | |
| CN107911478B (en) | Multi-user calculation unloading method and device based on chemical reaction optimization algorithm | |
| CN111932027B (en) | Cloud service comprehensive scheduling optimization system and method fusing edge facilities | |
| CN109561148B (en) | Distributed task scheduling method based on directed acyclic graph in edge computing network | |
| CN105729491B (en) | The execution method, apparatus and system of robot task | |
| WO2021147353A1 (en) | Order dispatch | |
| CN110069341B (en) | Method for scheduling tasks with dependency relationship configured according to needs by combining functions in edge computing | |
| CN110134146A (en) | A kind of distributed multiple no-manned plane method for allocating tasks under uncertain environment | |
| CN113867843B (en) | Mobile edge computing task unloading method based on deep reinforcement learning | |
| Gao et al. | Fast adaptive task offloading and resource allocation via multiagent reinforcement learning in heterogeneous vehicular fog computing | |
| CN116339932A (en) | Resource scheduling method, device and server | |
| CN116893861A (en) | Multi-agent cooperative dependency task unloading method based on space-ground cooperative edge calculation | |
| Semwal et al. | On ordering multi-robot task executions within a cyber physical system | |
| CN115392776A (en) | Spatial crowdsourcing task allocation method based on multi-skill cooperation | |
| CN113190342A (en) | Method and system architecture for multi-application fine-grained unloading of cloud-edge cooperative network | |
| CN112445617A (en) | Load strategy selection method and system based on mobile edge calculation | |
| CN117436627A (en) | Task allocation method, device, terminal equipment and medium | |
| CN116795553A (en) | Method and device for scheduling computing power resources, storage medium and electronic device | |
| CN109101329B (en) | Fine-grained task allocation method and system for acquiring data through multiple mobile terminals | |
| Han et al. | Opportunistic coded distributed computing: An evolutionary game approach | |
| CN103164747B (en) | Battlefield first-aid repair resource recombination optimization decision method | |
| CN112416579B (en) | Time-sensitive multiparty data fusion excitation method | |
| CN115208892A (en) | Vehicle-road cooperative online task scheduling method and system based on dynamic resource demand |
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 |