CN110377352B - Task processing method and device based on mobile device cloud system - Google Patents
Task processing method and device based on mobile device cloud system Download PDFInfo
- Publication number
- CN110377352B CN110377352B CN201910417909.XA CN201910417909A CN110377352B CN 110377352 B CN110377352 B CN 110377352B CN 201910417909 A CN201910417909 A CN 201910417909A CN 110377352 B CN110377352 B CN 110377352B
- Authority
- CN
- China
- Prior art keywords
- task
- node
- path
- probability
- mobile device
- 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
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
- G06F9/44—Arrangements for executing specific programs
- G06F9/445—Program loading or initiating
- G06F9/44594—Unloading
-
- 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
-
- 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
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
- G06F9/5038—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering the execution order of a plurality of tasks, e.g. taking priority or time dependency constraints into consideration
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
A task processing method and device based on a mobile device cloud system are disclosed, and the method comprises the following steps: acquiring a task to be unloaded and an initiating node of the task; performing initial allocation of execution nodes in the mobile device cloud system; reallocating tasks that were not allocated during the initial allocation process; and acquiring a path with the minimum success probability of executing the task, and adjusting the task on the path, so that the task with the maximum probability is successfully executed in the mobile equipment cloud system connected in an opportunistic manner, and the reliability of task execution is ensured.
Description
Technical Field
The application relates to the technical field of task processing, in particular to a task processing method and device based on a mobile device cloud system.
Background
Although performance of mobile devices continues to improve, for computationally intensive tasks, the limited computing power, memory space, and battery life of the tasks still remain difficult to meet the processing requirements of the tasks. Moreover, in practical applications, the computationally complex tasks that a mobile device is required to perform often exceed the processing power of a single device. Therefore, in situations where mobile devices are limited in capabilities and communication status presents an opportunity to be connected, how to handle such tasks becomes a critical issue today.
Disclosure of Invention
The invention aims to provide a task processing method based on a mobile equipment cloud system, so as to realize the successful execution of a task with the maximum probability in an opportunistic connection mobile equipment cloud system and ensure the reliability of task execution.
In order to solve the above problems, a first aspect of the present invention provides a task processing method based on a mobile device cloud system, which obtains a task to be offloaded and an originating node of the task; performing initial allocation of execution nodes in the mobile device cloud system; reallocating tasks that were not allocated during the initial allocation process; and acquiring a path with the minimum success probability for executing the task, and adjusting the task on the path.
Further, the performing initial allocation of execution nodes in the mobile device cloud system includes: obtaining a shortest path for executing the task, wherein the shortest path does not include a direct path; selecting an execution node for executing the task in the shortest path; upon identifying that the energy of the executing node is sufficient to execute the task, assigning the task to the executing node.
Further, the re-allocating the tasks that are not allocated in the initial allocation process includes: sequencing the paths distributed to the tasks in the initial distribution process, wherein the sequencing basis is available probability; and selecting the path with the maximum available probability, and selecting an executing node in the path with the maximum available probability.
Further, the obtaining a path with the minimum probability of success of executing the task and adjusting the task on the path further includes: obtaining the success probability of each path in the mobile device cloud; identifying the path with the maximum success probability and the path with the minimum success probability; identifying a change situation of success probability when any task in the path with the minimum success probability is adjusted to the path with the maximum success probability; and if the success probability is increased, controlling the task to be adjusted to the path with the maximum success probability.
Further, when the task is distributed, the energy and reliability of each node in the system are updated in real time.
Further, after the executing node completes the task, an executing result is fed back to the initiating node.
According to another aspect of the present invention, a task processing apparatus based on a mobile device cloud system includes: the acquisition module is used for acquiring a task to be unloaded and an initiating node of the task; an initial allocation module, configured to perform initial allocation of execution nodes in the mobile device cloud system; the redistribution module is used for redistributing tasks which are not distributed in the initial distribution process; and the adjusting module is used for acquiring the path with the minimum success probability for executing the task and adjusting the task on the path.
According to still another aspect of the present invention, a computer-readable storage medium has stored thereon a program which, when executed, implements the mobile device cloud system-based task processing method.
The technical scheme of the invention has the following beneficial technical effects: data with any size is allowed to be transmitted among nodes for multiple times, and a model for dynamically evaluating task reliability is introduced. Based on the framework provided by the application, the task unloading decision is realized, and the reliability of task execution can be ensured while the probability of successful task execution is maximized.
Drawings
Fig. 1 is a schematic structural diagram of a mobile device cloud system according to an embodiment of the present application;
fig. 2 is a schematic structural diagram of a mobile device cloud system according to another embodiment of the present application;
FIG. 3 is a schematic diagram illustrating a task adjustment process according to an embodiment of the present application;
fig. 4 is a flowchart of a task processing method of a mobile device cloud system according to an embodiment of the present invention;
fig. 5 is a flowchart of a task processing method of a mobile device cloud system according to another embodiment of the present invention;
fig. 6 is a flowchart of a task processing method of a mobile device cloud system according to another embodiment of the present invention;
fig. 7 is a flowchart of a task processing method of a mobile device cloud system according to still another embodiment of the present invention;
FIG. 8(a) is a graph of comparative validation results of successful execution probability as a function of deadline for the first embodiment of the present invention;
FIG. 8(b) is a graph of comparative validation results of task copies as a function of reliability requirements for a second embodiment of the present invention;
FIG. 8(c) is a diagram illustrating the comparison and verification result of the remaining power with the variation of the reliability requirement according to the third embodiment of the present invention;
fig. 9 is a block diagram illustrating a task processing apparatus of a mobile device cloud system according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings in conjunction with the following detailed description. It should be understood that the description is intended to be exemplary only, and is not intended to limit the scope of the present invention. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present invention.
In the drawings a schematic view of a layer structure according to an embodiment of the invention is shown. The figures are not drawn to scale, wherein certain details are exaggerated and possibly omitted for clarity. The shapes of the various regions, layers and their relative sizes, positional relationships are shown in the drawings as examples only, and in practice deviations due to manufacturing tolerances or technical limitations are possible, and a person skilled in the art may additionally design regions/layers with different shapes, sizes, relative positions according to the actual needs.
It is to be understood that the embodiments described are only a few embodiments of the present invention, 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 invention.
In the description of the present invention, it should be noted that the terms "first", "second", and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In addition, the technical features involved in the different embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
The invention will be described in more detail below with reference to the accompanying drawings. Like elements in the various figures are denoted by like reference numerals. For purposes of clarity, the various features in the drawings are not necessarily drawn to scale.
In the following description, numerous specific details of the invention, such as structure, materials, dimensions, processing techniques and techniques of the devices are described in order to provide a more thorough understanding of the invention. However, as will be understood by those skilled in the art, the present invention may be practiced without these specific details. Unless otherwise specifically noted below, various portions of the semiconductor device may be composed of materials well known to those skilled in the art.
It should be noted that the mobile device carries important work in a specific field. For example, on a battlefield, a mobile device equipped with a sound sensor may need to recognize a background gunshot from background to locate a sniper or to classify a gun type to provide information support for subsequent action; mobile devices pre-loaded with target recognition software may need to photograph targets in a battlefield and then recognize the targets so as to eliminate potential dangers as early as possible; in consideration of the special environment requirements of a battlefield (the voice information under the gun fire sound cage can not be acquired, and the voice information can be kept silent when confronted with a local place), the soldier can convert the voice command from a commander into a text through the voice recognition software assembled on the mobile device so as to realize information acquisition.
However, these computationally complex tasks often exceed the processing power of a single mobile device, and therefore, the present application proposes a mobile device cloud-based task processing method.
The definitions designed in this application are presented below by table 1.
First, unlike the related art, which assumes that there is a stable network connection between mobile devices and between the mobile devices and a remote data center, the mobile device cloud considers the network connection between the devices to be opportunistic and intermittent.
Therefore, the system constructed by the application comprises a central node (such as a satellite or a base station) and at least one mobile node (mobile device), wherein the central node is connected with the at least one mobile node and periodically collects the state information of each node, so that a decision basis is provided for task migration. Fig. 1 is a schematic structural diagram of a mobile device cloud system according to an embodiment of the present application. As shown in fig. 1, A, B, C and D are mobile nodes that are opportunistically connected and to which a central node is connected, respectively, that constantly collects contact information from each node with other nodes as well as real-time status information (e.g., power, assigned tasks, etc.). Taking the task allocation manner shown in fig. 1 as an example, the node a initiates a task 1 and a task 2, and sends the task 1 to the node C through the node B, so that the node C feeds back the task result to the node a through the node B after completing the task 1. In this mode, nodes in the mobile device cloud perform task computation in a cooperative manner, which is referred to as mobile device cloud system cooperative computation in the present application.
In order to select the optimal mobile node to perform a task and select the most suitable path to transmit data in the mobile cloud system, the following problems are faced. Firstly, mobile nodes are connected in an opportunistic manner, and the decision at the current moment is based on the current connection state, so that the current optimal decision can become a failure decision in the subsequent process; meanwhile, due to the instantaneity and the dynamic property of the contact time between the nodes, data related to the task cannot be completely transmitted in one contact, and multiple contact transmission needs to be considered. Second, this further increases the complexity of the problem, since each task typically has a time deadline for completion and execution results require return to the task originating node. Furthermore, the condition that the node fails is not avoided in consideration of the mobility of the node and the badness of the environment of a typical scene (such as a disaster and a battlefield). For example, a mobile node in a battlefield may be damaged by an enemy attack. Therefore, fault tolerance mechanisms also need to be considered when designing a mobile device cloud system. Finally, due to the limited computational performance and the limited battery capacity of the mobile device, energy savings issues may also need to be considered when assigning tasks to enable the mobile device to remain operational for long periods of time.
To achieve the object of the present application, the present application first performs the following analysis.
Fig. 2 is a diagram of an application scenario according to an embodiment of the present application. As shown in fig. 2, in a specific environment, some mobile nodes move randomly in respective areas, and can communicate with each other through a configured point-to-point communication interface (WiFi or bluetooth, etc.). Due to the mobility of the mobile node and the limitation of the wireless communication range, the nodes are in an opportunistic communication state. The nodes can share resources with each other, and therefore the system can be regarded as an opportunistic mobile equipment cloud system. When the task computing requirement of a certain node exceeds the capacity of the node, the node can unload the local computing task to other nodes. When two nodes move to the communication range, it can be regarded as one contact, and data transmission can be carried out between the nodes. It should be understood that opportunistic communication is considered to be a touch only if two nodes move into communication range. Therefore, the opportunistic contact between the mobile devices is modeled into the contact network
Wherein, the first and the second end of the pipe are connected with each other,
as a collection of nodesAnd epsilon is the set of contact edges between nodes. Is set to eijE ε node i to node j, which is distributed pcon mainly by inter-contact duration distributionijA decision is made (e.g., node a and node b have a node contact probability of 0.1). Since the contacts between the nodes are mutual, the graph G is an undirected graph. Furthermore, pcon ij0 means that there is no possibility of direct contact between node i and node j, and thus when pconij> 0 is that there is an edge e between node i and node jij. However, these non-directly contacting nodes may be considered as relays for task execution. In addition, the same two nodes are different at different contact durations. Node pair is established in this application<i，j>The minimum amount of data transmitted by one-time contact is deltaij. As shown in fig. 2, the task initiating node a can be in direct contact with nodes b, d, e, while node c is not in direct contact with node a, but can act as a data transmitter and a potential task performer. Since the data must be returned to the originating node after the task is executed, all execution paths are from the originating node to the originating node, e.g., path<a，b，a>. Global information in the mobile device cloud system, such as node states, assigned tasks, contact network maps between nodes, etc., is required to allow for cooperation between nodes. The present application assumes that there is a central node (satellite or base station in a mobile device cloud system, etc.) that periodically handsets these data. Since the bandwidth link between the mobile node and the central node is usually at a low rate, usually a satellite link, the task transmitted to the central node and then to the destination mobile node causes intolerable transmission delay, and therefore point-to-point transmission is necessary. Further, since the present application takes specific application scenarios such as war and disaster as an example, some mobile nodes may move according to specific rules and plans, the contact behavior between the mobile nodes in the present application can be predicted.
As can be seen from the foregoing analysis, the task of the task initiating node may not be executed by one mobile node due to large content or due to deadline limitation, and therefore, the task initiated by the task initiating node needs to be decomposed into a plurality of childrenFor example, in a speech-to-text application, a large speech file may be divided into a plurality of small files, and then the plurality of small files are converted by a plurality of mobile nodes simultaneously, so as to shorten the speech conversion time. This application sets subtask set
Each subtask
Is described as containing information such as the execution load Wn(number of CPU processes), deadline TnInput data quantity DinnAnd amount of output data DoutnAnd the like. Given that the amount of data before and after processing is usually different, for a particular application, Din is assumednAnd DoutnWith a parameter factor delta, i.e. Dinn＝δDoutn. Such detailed information on how to obtain tasks is prior art and is not described in detail in this application. The method and the device mainly solve the problem of how to distribute the plurality of subtasks initiated by the task initiating node to each node in the mobile equipment cloud so as to accelerate the task completion speed.
Generally speaking, only the result of task execution needs to be returned to the task initiating node, and the intermediate result of task execution does not need to be returned. However, due to the opportunistic connections between mobile nodes and the consideration of minimizing task completion time, fault tolerance, and energy efficiency, it is very challenging to optimally assign these tasks to each dynamically changing mobile node.
Further, in the present application, each subtask is set to be irrelevant, so that a star-shaped communication network structure is presented in the execution process of the subtask. For subtasks
We schedule according to the priority order of the tasks. For a single subtask, the offload network model of subtasks between nodes is shown in FIG. 2, and when encountering other nodes, the task initiatesThe node may offload its load to the encountered node and return the result to the task initiating node after task execution is complete.
Due to the mobility of the nodes and the badness of the environment, especially in a disaster or a field environment, the nodes inevitably fail, which leads to the effective task execution. Therefore, considering fault tolerance mechanisms in the sub-task migration process is particularly important to guarantee successful execution of the tasks. In this regard, the present application analyzes both the failure and the task reliability of the mobile node.
And (5) analyzing task failure. The failure of the node usually comes from hardware errors or external environment attacks. The incidence of node errors is known to obey poisson distribution and has been widely used in system reliability studies. Thus, for a single node i, the node error probability distribution over time T can be formalized as:
wherein u isiIs a parameter of the poisson function, and s represents the number of errors occurring simultaneously within time T.
Therefore, when s is equal to 0, the probability that the node i does not go wrong within the time T is expressed as:
for the task reliability model. In the present application, the reliability of the input task is R, N subtasks are included in the input task, and the importance of each subtask is equal, so that the reliability of each subtask can be considered to be the same, and can be calculated as
In order to improve the reliability of task execution, each task needs to be distributed to a plurality of nodes for execution (the tasks are decomposed into a plurality of subtasks and then distributed to a plurality of nodes for execution). Also, consider the finite nature of the node resourcesA node may also assume multiple subtasks to guarantee the backup of a task. Let on (i) be the set of subtasks already allocated in node i, and at the current time, if the subtask n is allocated to node i, the total execution time (TT) of the subtask n can be represented as:
wherein ET (n, i) is the execution time of the task n on the node i, nkIs the kth element in the on (i) set. Load W of known task nnAnd the processing speed c of the node iiIf ET (n, i) is Wn/ci. Referring to equation (2), it can be deduced that the current reliability of the node is:
where R (n, i) is the probability that task n will not be in error during the execution of node i. Thus, when R (n, i) is greater than the required reliability R, the assignment of task n is stopped.
Further, after the task initiating node initiates the task, the task initiating node is divided into a plurality of subtasks by a certain task description tool, and the subtasks are recorded as
The problem to be solved is how to assemble the subtasks
Each subtask in (a) is offloaded to other nodes in the mobile device cloud system to speed up the execution of the task. It should be noted that, during the sub-task unloading process, the following requirements also need to be satisfied: 1) the deadline requirements of each subtask are met; 2) ensuring the reliability requirement of task execution; 3) the energy consumption for executing the task is minimized as much as possible. Finally, a task offloading problem in a collaborative mobile device cloud system may be defined to maximize the probability of successful execution of all subtasks and satisfy all constraints. IntoRather, the problem can be numerically formalized as:
wherein the content of the first and second substances,
is a set of paths from the task initiating node to the result receiving node. PS (n, p, i) is the probability that task n will be transmitted over path p and executed successfully by node i within the deadline. x is a radical of a fluorine atomn,p,iIs a variable that indicates whether a subtask is assigned to node i in path p. R (n) represents the actual reliability of the execution of the subtask n, and equation (6) ensures that the actual execution reliability of the subtask n is greater than the required reliability. Equation (7) ensures that the power level of node i is not lower than the set threshold. Formula (8) represents xn,p,iIs a binary variable. Equation (9) indicates that any task is allocated to the same node at most once and each task is allocated at least once. Therefore, the problem is a joint optimization problem by optimizing the distribution path of the tasks, the execution nodes and the scores of the tasks under the conditions of reliability and energy constraint. Since the execution path set grows exponentially with the number of nodes, the time complexity of solving the optimal solution to the problem is very highHigh, the optimization algorithms involved are also not suitable to run in mobile device cloud systems.
Further, since the nodes are usually carried or controlled by people, the moving pattern of the mobile node may exhibit certain regularity. For example, drones typically cruise on a fixed route set by a commander when acquiring battlefield situations. Thus, given that the behavior of contacts between nodes is predictable, the present application analyzes the successful execution of tasks on a particular path in a probabilistic manner, since different paths have different data transfer and task execution capabilities and are largely determined by the probability of their opportunistic connections. Most of the prior related technologies mainly consider the single-contact condition of two nodes when performing path probability analysis, and assume that data can be successfully transmitted regardless of the contact time. However, in the present application, it is assumed that the transmission capability of a single contact between nodes is limited, and for larger data, multiple contacts may be required to complete the transmission.
The probability of availability of the opportunistic path is a measure of the connection state of the path in a specific time, and is a key factor for the successful execution of a task on the path. The method measures the availability of the path by using the contact interval duration of each node on the path. The contact interval duration refers to the time between the end of the previous contact and the beginning of the next contact. Intuitively, for a node, the smaller the contact interval, the longer the contact time, and thus the greater the probability of availability of a path. Based on this, the probability of availability of the opportunity path may be determined by the contact interval duration factor. To analyze the availability of the paths, the contact process between the nodes needs to be studied in depth.
Specifically, since the contact process between nodes follows poisson distribution, the contact interval duration between nodes follows exponential distribution. Consider a two-hop path (two data migrations occur) < a, v, a >, which means that node a offloads the task to node v and then returns the result to a. Therefore, the available probability corresponding to the path can be calculated as follows:
setting a random variable T1And T2A time length distribution of contact for the first hop and the second hop nodes, respectively, whichThe probability density function corresponds to f (t, λ) respectively1)＝Exp(t,λ1) And f (t, λ)2)＝Exp(t,λ2). Then, the total path available duration may be represented as T ═ T1+T2The corresponding probability density function can also pass through f (t, λ)1) And f (t, λ)2) Is obtained, i.e. f (t, λ)1)·f(t,λ2). The availability probability corresponding to the path can be expressed as:
in fact, the above formula calculates the path availability probability within one touch. However, in practice, transmitting a particular datum may require multiple transmissions, and thus equation (10) does not fully mine the path availability. Suppose a data transfer takes m contacts in the first hop to complete the data transfer, and T1q～Exp(λq) Let the q ═ 1 in the first hop, the contact duration of the a transmissions, then the total contact duration of the first hop is then
Theoretical proof of T1Obeying a gamma distribution, i.e. T1～Gamma(a,λ1). In the same way, T2～Gamma(a,λ2). Based on this, the availability probability of a path < a, v, a > can be expressed as:
more generally, for a path p with K hops, the corresponding total available duration is then
The corresponding probability density function is:
wherein λ isk(K ═ 1,. K), and mkAnd (K-1., K) represents the probability function parameter factor of the K-th hop and the number of contacts required to transmit data, respectively. Wherein, when k is very large, f (t) is a high latitude measure, the calculation complexity is very high, and the approximation can be simplified into a single gamma random variable by using Satterthwaite (Sautersvaite) approximation theory, namely:
wherein the content of the first and second substances,
to this end, the available probability of our path may be represented by the formula P (T ≦ T) ═ ^ P0 Tf (t) dt calculation.
Because, within a certain time period, even if a path is available, it is not necessarily guaranteed that the achievement transmits the relevant data. The following application presents a probability analysis to measure the success of data transmission. Noting that the contact time duration within a single contact and being assumed to follow the pareto distribution, the amount of data transmitted within a single contact is also readily attainable to approximately follow the pareto distribution, since the transmission rate between nodes is relatively stable. For node pair < v, w >, set DiPareto (α, β) is a random variable representing the amount of data transmitted at the ith transmission, where α is the shape factor and β is the scale factor (which can be considered the minimum amount of data for a single transmission). Assuming that the node pair has c contacts within time T, the total amount of data transmitted by the node pair during time T may be expressed as
The probability that data with the size of D is successfully transmitted in the time T is P (D is larger than or equal to D). Taking into account the variables D and DiA maximum value M of (i ═ 1., c) (i.e., M ═ max { D })iI 1.., c), the variable D may be approximated by M. Thus, P (D.gtoreq.D) can be further simplifiedComprises the following steps:
the following conditions are required due to the unloading of a task successfully achieved within a given time T: 1) the path must be available within time T; 2) all relevant data, including the result data, must be transmitted along the target path within time T. Thus, given a task n, the corresponding input and output data sizes are respectively
And
at its path < a, v1,...,vK-1The probability of successful offload, a > can be calculated in the following manner.
Since any node in the path may be responsible for the task execution, there is a choice for any task n to execute in k on the path p. Also, for the k-th hop, the minimum transmission amount per transmission is considered to be βkAt least is provided with
To transmit the relevant data. Dk(n, i) represents the amount of data that task n needs to transmit at the ith node when it is executed at the kth node. Thus, the probability of successfully executing task n on path p is the sum of the probabilities of all possible cases, namely:
wherein f isi(m1,..,mK) When the representative task n is executed at the ith node, each hop respectively needs m1,..,mKProbability of secondary contact to fully transmit data. Note that if case m1,..,mKTransmission loss only when one of the hops was in previous contactFailure to beat, i.e. m1-1,..,mKA certain hop of data transmission failure in-1. Is provided with
For data, the number of transmission times is m1,..mk-1,mKWhen transmission fails and m is the case1,..mk,mKThe probability of success, i.e.:
wherein the content of the first and second substances,
represents the sum of the time of the task n executed at the ith node and the time required for the k-th hop to transmit the related data, 1i＝kDenotes that when i ═ k, it has a value of 1, otherwise it is 0;
k contact m when the execution node of the task n is ikThe number of next possible transmissions. Obviously, in a path, the data that the node before the executing node needs to transmit is the related data carried by the task, i.e. the data input
The data to be transmitted by the nodes behind the executing node is the result of the task execution, namely the data transmission quantity is
Formally can be expressed as:
wherein the content of the first and second substances,
is m before the k-th jumpkProbability of secondary transmission failure. It can be analytically concluded that there are two main aspects of transmission failure, one is that the connection of the path is not available, and the other is that the data is not transmitted over the deadline. If m iskThe sub-transmission is successful, meaning that m isk1 transmission path must be available. Thus, it results in the m-thkThe reason for-1 transmission failure is mainly that the data is not completely transmitted. Thus, the number of the first and second electrodes,
can be calculated by the following formula:
then, assume that h tasks are allocated to path p and set the task allocation result of path p as
Wherein the content of the first and second substances,
for the task combination assigned to the route,
and executing nodes corresponding to the paths p for all the tasks. In order to ensure that the tasks are executed according to the priority sequence, the tasks distributed to the path p are arranged in a stack structure, and the deletion and the addition of the tasks are in a last-in first-out sequence so as to facilitate the adjustment of the subsequent tasks. Fig. 3 gives an example of the adjustment of task 3 from node d in path u to node c in path v. Thus for path p, to calculate its probability of successful execution
The data transmission amount per hop and the task execution time need to be calculated. Is given by1,d2,...,dK> and < t1,t2,...,tKThe transmission data amount and the task execution time of each hop are respectively. Wherein the content of the first and second substances,
represents the total data transmission amount of the k-th hop,
is in the path
Each node executes tasks
The amount of data that needs to be transmitted for the k-th hop. DkThe definition of (n, i) can refer to formula (19). Similarly, the total execution time of the k-th hop may be calculated
Wherein, Tk(n, i) the time taken by the k-th hop for task n to execute on node i, which includes the execution time of the task and the transmission time of the data, can be computed as Tk(n,i)＝1i＝k·wn/ck+dk/rk. Thus, D in equation (18) can be expressedk(n, i) and T'n,i,lBy substitution of dkAnd tkThat is, the probability of successful execution of path p under a multitask assignment may be calculated
Therefore, in consideration of reliability of task execution of the nodes, energy consumption of the nodes and other constraints, the following rules need to be followed during specific design: 1) if the power of the node i is less than the predetermined threshold (e.g., the power of the node i is less than the predetermined threshold)
) If so, no task is distributed to the node i; 2) if it is
Task n should try to be allocated to the node closest to the initiating node to reduce the network communication pressure and vice versa; 3) each task cannot be assigned to the same node more than once. It should be appreciated that if a node fails, the tasks previously assigned to that node will fail, and thus multiple backups at the same node will not achieve the goal of improved reliability.
Based on the analysis, the application provides a task processing method and device based on a mobile device cloud system.
It should be noted that the task mentioned in the present application may be understood as the aforementioned subtask n, that is, the task in the initiating task node has been divided into a plurality of subtasks, and the assignment problem of the subtasks is studied in the present application.
It should be noted that the larger amount of data carried by the task means that the longer time is required for unloading, and thus a higher error rate is caused during the task execution process. It can be seen that it is necessary to give different priorities to different tasks in order to increase the overall success rate of task execution. Based on the idea, the priority of the task is set according to the size of the task data, and the priority is higher when the data volume is larger, and vice versa. Given a subtask, it is necessary to decide on which path it should transmit data and which node on that path to select for task processing. As discussed above, without loss of generality, let the result of task assignment be a triplet
Representing a set of tasks
Is distributed to a path p for data transmission, and the corresponding task execution node is
The ideal solution is to find one path among all possible paths to maximize the probability of task execution, including the path availability probability and the dataThe probability of success of the input. However, finding the optimal method in so many feasible sets is necessarily inefficient, as the feasible sets grow exponentially with increasing number of nodes. Note that in the foregoing analysis of the present application, once a path is determined, a task execution node may be determined according to the above-described rule.
Fig. 4 is a flowchart of a task processing method based on a mobile device cloud system according to an embodiment of the present invention. As shown in fig. 4, the task processing method based on the mobile device cloud system according to the embodiment of the present invention includes the following steps:
s101: and acquiring at least one task to be unloaded and an initiating node of the task to be unloaded.
S102: for each task to be offloaded, an initial allocation of execution nodes is made in the mobile device cloud system.
According to an embodiment of the present invention, step S102 is shown in fig. 5, and further includes:
s201: and acquiring the shortest path with the maximum probability of executing the task.
Wherein the shortest path does not include a direct path, i.e. a path directed to the originating node itself.
S202: and selecting an execution node for executing the task according to the data volume of the task in the shortest path.
Further, judging whether the input data volume of the task is larger than the output data volume of the task; if so, selecting a node close to the initiating node as an executing node.
S203: upon identifying that the energy of the executing node is sufficient to execute the task, the task is assigned to the executing node.
That is to say, the method and the device find the path with the maximum available probability, then select the optimal task execution node in the path, and consider the energy consumption factor. And after each path is distributed, repeating the steps according to the task priority in a descending order to perform the process processing on all the tasks until all the tasks are distributed for the first time or no available node exists, and stopping at the stage.
For example, in the cloud system of terminal devices shown in fig. 2, a node a has a subtask n to be unloaded to other nodesAnd the deadline of the subtask is T. In the initial distribution stage, all nodes are all available execution nodes, and the available node set is recorded
Let all paths from available nodes a to a be aggregated as
The path of the assigned task is
The available probability of path k within time T is
The available probability of a direct path from a task trigger node to a task result receiving node is Pr'av1. Then, Dijsktra algorithm is used to find out the out-of-direct-path messenger
Shortest path k, subtask n is assigned to path k and added to the assigned set of paths
In (1). Once the path k is assigned, the corresponding edge along the path is deleted from the original probability map G. If it is
The subtask n should be distributed to the node closest to the initiating node as much as possible to reduce the network communication pressure; otherwise, the subtask n should be allocated as far as possible to the node farthest from the originating node. The executing node of the task is selected from the available nodes of the path k according to the above rule, and the selected node i needs to satisfy
Wherein e isiRepresenting the energy required by node i to perform the subtask n. This means that only nodes whose energy is sufficient for task execution can do soIs selected. The state of each node, such as energy and the current reliability of each task, is then updated. If the energy is less than a certain threshold, the node is selected from the available set
And (5) deleting. Repeating the above steps until all paths have been assigned tasks, i.e.
Or all tasks are assigned and the reliability condition is met.
S103: tasks that were not allocated during the initial allocation process are reallocated.
According to an embodiment of the present invention, step S103 is shown in fig. 6, and further includes:
s301: and sequencing the paths allocated to the tasks in the initial allocation process.
Wherein the ranking is based on the current availability probability, i.e., the availability probability that the path can still reach after the initial allocation.
S302: and selecting the path with the highest available probability, and selecting the executive node in the path with the highest available probability.
That is, after the initial allocation, there may be a case where the task is not fully allocated or the number of copies of the task has not yet reached the reliability requirement, and the purpose of this stage is to further ensure the full allocation of the task and guarantee the reliability requirement. The most basic idea is to continually assign incomplete tasks to the path with the highest probability of successful execution through iteration.
In particular, for unallocated task sets
The task in (1), first, the allocated paths are collected
The paths in (1) are sorted according to the current available probability, and the path with the maximum available probability is recorded as pm. Then, after that,assigning subtask n to path pmAnd according to the foregoing
And
and a node for executing the node energy determination task.
S104: and acquiring the success probability of each path in the cloud system of the mobile equipment, and adjusting the task according to the success probability.
According to an embodiment of the present invention, step S104 is shown in fig. 7, and further includes:
s401: and acquiring the success probability of each path in the mobile device cloud.
It should be understood that, at this time, the mobile device cloud system has undergone two task allocations, and all the sub-tasks to be offloaded in the current state have been already allocated to each node, but each path has a different success probability according to the performance parameters (task data amount, energy, etc.) of each node.
S402: the path with the highest probability of success is identified with the path with the lowest probability of success.
S403: and identifying the change condition of the success probability when any task in the path with the minimum success probability is adjusted to the path with the maximum success probability.
S404: and if the success probability is increased, controlling the task to adjust to the path with the maximum success probability.
That is, after reallocation, since the assignment of the new task is added to each path, the probability of successful execution of the original task is affected. This means that the re-allocated solution is not optimal. This stage increases the probability of successful execution by adjusting the assignment of tasks.
It should be understood that, in the embodiment of the present application, the available probability is a probability value of whether the paths are connected, and the success probability is a probability value of successfully executing the task on the paths, where successful execution of the task requires that not only the paths are connected, but also the functions of calculation, transmission, and the like of the task are completed. In other words, the available probability is a precondition for the probability of success.
Specifically, paths u and v with the maximum and minimum success probabilities in the mobile device cloud system are found, and the task of allocating the path with the minimum probability is adjusted to the path with the maximum probability until the success execution probability is not increased.
For example, the path assignment set to the maximum and minimum probability of success is respectively
And
extracting the task h from the task queue of the path v if
The task h is added to the path u with the highest probability of being currently executed. If there is no task executed after the task h in the path v is adjusted to the path u, that is,
the path v is deleted from the path set P'. And repeating the steps until the adjustment of the task distribution does not improve the overall execution success probability any more.
In order to prove the effect of the task processing method based on the mobile device cloud system, the method of the present application is compared with a Random strategy, a MaxPro (maximum contact probability) strategy and a MaxRate (maximum transmission rate) strategy. The Random strategy randomly selects the node contacted by the task initiating node as a target node for task unloading. And the MaxPrO strategy selects the node with the maximum contact probability with the task initiating node as a task unloading object. And the MaxRate strategy selects the node with the maximum transmission rate with the task initiating node as the unloading purpose of the task. Obviously, this strategy results in minimal transmission delay.
As shown in fig. 8(a), PSE indicators versus task deadline. With the increase of the deadline, besides the Random policy, the probability of successful execution of the task PSE increases accordingly. This is because the longer the deadline is, the more likely the node is to come into contact with the task initiating node and thus it is easier to return the execution result within a specified time. Because the Random strategy randomly selects the encountered nodes as task execution nodes, the corresponding curves also present Random variation conditions. Note that when Deadline is greater than 1500, the probability of successful execution of the task no longer increases because it is possible that all tasks can be completed within 1500 time slots, and the increase in Deadline does not affect the successful execution of the task. From this picture, it is also found that the PSE index always maintains the highest level in the change of the algorithm UNION compared with other algorithms with the deadline, thus indicating the superiority of the algorithm compared with other algorithms.
As shown in fig. 8(b), the average number of task copies varies with the reliability requirement. As can be seen, as reliability requirements increase, the number of copies of a task continues to increase. This is because the improvement in reliability requires more copies to ensure successful execution of the task. From this result, it is also observed that the number of copies produced by UNION, an algorithm herein, is more stable than other algorithms, and is kept approximately stable at the number of 2 copies, which means UNION has the advantage of more resource saving without reducing reliability than other algorithms.
Comparing the remaining energy of each strategy with the reliability requirement, as shown in fig. 8(c), it can be seen that UNION maintains the highest and most stable remaining energy. This illustrates that the algorithm presented herein can also optimize the allocation of tasks to balance the energy consumption among the nodes. In a real system, this is essential to delay the operation of the mobile system for as long a time as possible. As other policies have less energy remaining because under these policies it is likely that a node will be assigned multiple tasks.
In summary, the task processing method of the cloud system of the mobile device in the embodiment of the application allows data of any size to be transmitted among nodes for many times, and introduces a model for dynamically evaluating task reliability. Based on the framework provided by the application, the task unloading decision is realized, and the reliability of task execution can be ensured while the probability of task successful execution is maximized.
Fig. 9 is a block diagram illustrating a task processing apparatus based on a mobile device cloud system according to an embodiment of the present invention. As shown in fig. 9, a task processing apparatus 100 based on a mobile device cloud system according to an embodiment of the present invention includes: an acquisition module 10, an initial allocation module 20, a re-allocation module 30 and an adjustment module 40.
The acquiring module 10 is configured to acquire a task to be offloaded and an initiating node of the task; an initial allocation module 20 is configured to perform initial allocation of execution nodes in the mobile device cloud system; the reassignment module 30 is used to reassign tasks that were not assigned during the initial assignment process; the adjusting module 40 is configured to obtain a path with the minimum success probability for executing the task, and adjust the task on the path.
In order to achieve the above object, the present application further proposes a computer-readable storage medium, on which a program is stored, and when the program is executed, the program implements the foregoing task processing method based on a mobile device cloud system.
It is to be understood that the above-described embodiments of the present invention are merely illustrative of or explaining the principles of the invention and are not to be construed as limiting the invention. Therefore, any modification, equivalent replacement, improvement and the like made without departing from the spirit and scope of the present invention should be included in the protection scope of the present invention. Further, it is intended that the appended claims cover all such variations and modifications as fall within the scope and boundaries of the appended claims or the equivalents of such scope and boundaries.
In the above description, the technical details of patterning, etching, and the like of each layer are not described in detail. It will be understood by those skilled in the art that layers, regions, etc. of the desired shape may be formed by various means known in the art. In addition, in order to form the same structure, those skilled in the art can also design a method which is not exactly the same as the method described above.
The invention has been described above with reference to embodiments thereof. However, these examples are for illustrative purposes only and are not intended to limit the scope of the present invention. The scope of the invention is defined by the appended claims and equivalents thereof. Various alternatives and modifications can be devised by those skilled in the art without departing from the scope of the invention, and these alternatives and modifications are intended to be within the scope of the invention.
Although the embodiments of the present invention have been described in detail, it should be understood that various changes, substitutions, and alterations can be made hereto without departing from the spirit and scope of the invention.
It should be understood that the above examples are only for clarity of illustration and are not intended to limit the embodiments. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. And obvious variations or modifications derived therefrom are intended to be within the scope of the invention.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Claims (7)
1. A task processing method based on a mobile device cloud system is characterized by comprising the following steps:
acquiring at least one task to be unloaded and an initiating node of the task to be unloaded;
performing initial allocation of execution nodes in the mobile device cloud system for each task to be offloaded;
reallocating tasks that were not allocated during the initial allocation process;
acquiring the success probability of each path in the mobile equipment cloud system, and adjusting the task according to the success probability;
the obtaining of the success probability of each path in the cloud system of the mobile device and the adjusting of the task according to the success probability further include:
obtaining the success probability of each path in the mobile device cloud;
the success probability is calculated as：
Wherein the content of the first and second substances,
representing tasks
In the first place
When each node executes, each hop is respectively required
Probability of secondary contact to fully transmit data;
at the same time arrange
For data at the second
The number of times of transmission is respectively
When the transmission is successful, only when in the second place
Number of hops transmission is
Probability of transmission failure;
the following can be obtained:
and is
Wherein the content of the first and second substances,
is a task
In the first place
A node is
The process of jumping is carried out,
is a task
In the first place
A node is
Is jumped and
，
is the first
Jump to the first
The length of the contact time of the secondary transmission,
is the first
Jump to the first
The length of the contact time of the secondary transmission,
is the first
Before jumping
The probability of a failure of a secondary transmission,
for the set of available paths to be used,
for a task
(ii) a deadline;
and the number of the first and second electrodes,
and the number of the first and second electrodes,
wherein the content of the first and second substances,
in when
If so, the value is 1, otherwise, the value is 0;
at task
Is an execution node of
Is first of
Jump contact
After the next timeThe number of the data that can be transmitted,
to be a task
Is detected by the load of (a) a load,
is a node
The computing power of (a) is calculated,
to be a task
In the first place
When a node executes
The rate at which data is being hopped for transmission;
while
Wherein the content of the first and second substances,
in (1)
Is a task
In the first place
A node is
The process of jumping is carried out,
to be a task
The amount of data input of (a) is,
to be a task
The amount of data output, and, correspondingly,
in (1)
Is a task
In the first place
A node is
Jump and are
；
Identifying the path with the maximum success probability and the path with the minimum success probability;
identifying a change situation of success probability when any task in the path with the minimum success probability is adjusted to the path with the maximum success probability;
and if the success probability is increased, controlling the task to adjust to the path with the maximum success probability.
2. The mobile device cloud system-based task processing method of claim 1, wherein the performing of the initial allocation of the execution nodes in the mobile device cloud system comprises:
obtaining a shortest path with the highest probability of executing the task, wherein the shortest path does not include a direct path;
selecting an execution node for executing the task according to the data volume of the task in the shortest path;
upon identifying that the energy of the executing node is sufficient to execute the task, assigning the task to the executing node.
3. The task processing method based on the mobile device cloud system according to claim 2, wherein the selecting, in the shortest path, an execution node for executing the task according to a data amount of the task further comprises:
judging whether the input data volume of the task is larger than the output data volume of the task;
and if so, selecting a node close to the initiating node as the executing node.
4. The mobile device cloud system based task processing method of claim 1, wherein said re-allocating tasks that were not allocated during said initial allocation process comprises:
sequencing the paths distributed to the tasks in the initial distribution process, wherein the sequencing basis is the current available probability;
and selecting the path with the maximum available probability, and selecting an executing node in the path with the maximum available probability.
5. The task processing method based on the mobile device cloud system according to any one of claims 1 to 4, further comprising:
and updating the energy and reliability of each node in the system in real time when the task is distributed.
6. The task processing method based on the mobile device cloud system according to any one of claims 1 to 4, further comprising:
and after the executing node finishes the task, feeding back an executing result to the initiating node.
7. A computer-readable storage medium having a program stored thereon, wherein the program, when executed, implements the mobile device cloud system-based task processing method of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910417909.XA CN110377352B (en) | 2019-05-20 | 2019-05-20 | Task processing method and device based on mobile device cloud system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910417909.XA CN110377352B (en) | 2019-05-20 | 2019-05-20 | Task processing method and device based on mobile device cloud system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110377352A CN110377352A (en) | 2019-10-25 |
| CN110377352B true CN110377352B (en) | 2022-06-14 |
Family
ID=68248567
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910417909.XA Active CN110377352B (en) | 2019-05-20 | 2019-05-20 | Task processing method and device based on mobile device cloud system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110377352B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112181665A (en) * | 2020-10-21 | 2021-01-05 | 中国联合网络通信集团有限公司 | Task optimization method and device for low-earth-orbit satellite |
| CN112114648B (en) * | 2020-11-23 | 2021-02-09 | 中国人民解放军国防科技大学 | Wearable device power management method and system and computer device |
| CN112866381B (en) * | 2021-01-17 | 2022-07-01 | 湘潭大学 | Task redundancy allocation method based on facility site selection problem in edge calculation |
| CN113872850A (en) * | 2021-09-27 | 2021-12-31 | 东莞市亚太未来软件有限公司 | Real-time communication method and system |
| CN114697209B (en) * | 2022-03-30 | 2023-12-22 | 广州穗华能源科技有限公司 | Cloud edge collaborative computing resource configuration method and configuration system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106462542A (en) * | 2014-03-10 | 2017-02-22 | 英特尔公司 | Mobile application acceleration via fine-grain offloading to cloud computing infrastructures |
| CN106936892A (en) * | 2017-01-09 | 2017-07-07 | 北京邮电大学 | A kind of self-organizing cloud multi-to-multi computation migration method and system |
| CN108958916A (en) * | 2018-06-29 | 2018-12-07 | 杭州电子科技大学 | Workflow unloads optimization algorithm under a kind of mobile peripheral surroundings |
| CN109005211A (en) * | 2018-06-29 | 2018-12-14 | 福建师范大学 | Thin cloud deployment and scheduling user task method under a kind of wireless MAN environment |
| CN109074265A (en) * | 2016-03-28 | 2018-12-21 | 甲骨文国际公司 | The preformed instruction of mobile cloud service |
| CN109756578A (en) * | 2019-02-26 | 2019-05-14 | 上海科技大学 | A kind of low time delay method for scheduling task calculating network towards dynamic mist |
Family Cites Families (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101978352B (en) * | 2007-12-13 | 2017-11-03 | 先进微装置公司 | For with multiple graphics subsystem, the driver framework of the computing device of the power dissipation modes of reduction, software and method |
| US8271974B2 (en) * | 2008-10-08 | 2012-09-18 | Kaavo Inc. | Cloud computing lifecycle management for N-tier applications |
| US9268545B2 (en) * | 2011-03-31 | 2016-02-23 | Intel Corporation | Connecting mobile devices, internet-connected hosts, and cloud services |
| KR20140023674A (en) * | 2012-08-17 | 2014-02-27 | 삼성전자주식회사 | Method for using and creating an shortcut object of contents based on a cloud service, and device supporting the same |
| US10154095B2 (en) * | 2014-05-21 | 2018-12-11 | N.Io Innovation, Llc | System and method for aggregating and acting on signals from one or more remote sources in real time using a configurable platform instance |
| US11546428B2 (en) * | 2014-08-19 | 2023-01-03 | Hand Held Products, Inc. | Mobile computing device with data cognition software |
| US9832081B2 (en) * | 2014-09-29 | 2017-11-28 | International Business Machines Corporation | Allocating physical nodes for processes in an execution plan |
| US20160164986A1 (en) * | 2014-12-08 | 2016-06-09 | Google Inc. | Multi-purpose application launching interface |
| CN106597881A (en) * | 2016-11-03 | 2017-04-26 | 深圳量旌科技有限公司 | Cloud service robot based on distributed decision-making algorithm |
| EP3355188B1 (en) * | 2017-01-31 | 2021-08-25 | OpenSynergy GmbH | Instrument display on a car dashboard by checking frames of a gui by a realtime os |
| CN107087019B (en) * | 2017-03-14 | 2020-07-07 | 西安电子科技大学 | Task scheduling method and device based on end cloud cooperative computing architecture |
| CN107766135B (en) * | 2017-09-29 | 2021-04-27 | 东南大学 | Task allocation method based on particle swarm optimization and simulated annealing optimization in moving cloud |
-
2019
- 2019-05-20 CN CN201910417909.XA patent/CN110377352B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106462542A (en) * | 2014-03-10 | 2017-02-22 | 英特尔公司 | Mobile application acceleration via fine-grain offloading to cloud computing infrastructures |
| CN109074265A (en) * | 2016-03-28 | 2018-12-21 | 甲骨文国际公司 | The preformed instruction of mobile cloud service |
| CN106936892A (en) * | 2017-01-09 | 2017-07-07 | 北京邮电大学 | A kind of self-organizing cloud multi-to-multi computation migration method and system |
| CN108958916A (en) * | 2018-06-29 | 2018-12-07 | 杭州电子科技大学 | Workflow unloads optimization algorithm under a kind of mobile peripheral surroundings |
| CN109005211A (en) * | 2018-06-29 | 2018-12-14 | 福建师范大学 | Thin cloud deployment and scheduling user task method under a kind of wireless MAN environment |
| CN109756578A (en) * | 2019-02-26 | 2019-05-14 | 上海科技大学 | A kind of low time delay method for scheduling task calculating network towards dynamic mist |
Non-Patent Citations (1)
| Title |
|---|
| ACO-BASED SOLUTION FOR COMPUTATION OFFLOADING IN MOBILE CLOUD COMPUTING;Weidong Bao 等;《Big Data and Information Analytics (BDIA)》;20160101;第1-13页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110377352A (en) | 2019-10-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110377352B (en) | Task processing method and device based on mobile device cloud system | |
| Liu et al. | Online computation offloading and traffic routing for UAV swarms in edge-cloud computing | |
| US8918792B2 (en) | Workflow monitoring and control system, monitoring and control method, and monitoring and control program | |
| CN110493360B (en) | Mobile edge computing unloading method for reducing system energy consumption under multiple servers | |
| EP2493118B1 (en) | Information processing system | |
| WO2015001850A1 (en) | Task allocation determination device, control method, and program | |
| CN113784373B (en) | Combined optimization method and system for time delay and frequency spectrum occupation in cloud edge cooperative network | |
| JP2015070614A (en) | Relay apparatus and method of selecting relay node based on learning in wireless network | |
| CN115454589A (en) | Task scheduling method and device and Kubernetes scheduler | |
| CN115629865A (en) | Deep learning inference task scheduling method based on edge calculation | |
| Fan et al. | Joint DNN partition and resource allocation for task offloading in edge-cloud-assisted IoT environments | |
| Sun et al. | Task allocation for multi-APs with mobile edge computing | |
| CN116954905A (en) | Task scheduling and migration method for large Flink data | |
| CN115955479A (en) | Task rapid scheduling and resource management method in cloud edge cooperation system | |
| CN115809148A (en) | Load balancing task scheduling method and device facing edge computing | |
| KR102279184B1 (en) | Task offloading method in fog networks and design method for fog networks architecture | |
| Kinoshita et al. | Joint bandwidth scheduling and routing method for large file transfer with time constraint and its implementation | |
| CN114610503A (en) | Multi-application online computing unloading resource scheduling method and system based on overall resource requirements of task graph in edge computing | |
| CN113918318A (en) | Joint optimization method and system for mobile edge calculation | |
| US20230008176A1 (en) | A Multi-Tenant Real-Time Process Controller for Edge Cloud Environments | |
| US20220383116A1 (en) | METHOD AND SYSTEM FOR TASK MANAGEMENT IN AN INTERNET OF THINGS (IoT) EDGE NETWORK | |
| Wei et al. | A novel scheduling mechanism for hybrid cloud systems | |
| Sookoor et al. | Offload destination selection to enable distributed computation on battlefields | |
| Shah et al. | Resource allocation scheme to reduce communication cost in mobile ad hoc computational grids | |
| CN116760454B (en) | Relay satellite resource allocation method and device |
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 |