CN113840014B - Distributed task decomposition method adaptive to high-strength weak connection environment - Google Patents
Distributed task decomposition method adaptive to high-strength weak connection environment Download PDFInfo
- Publication number
- CN113840014B CN113840014B CN202111428819.4A CN202111428819A CN113840014B CN 113840014 B CN113840014 B CN 113840014B CN 202111428819 A CN202111428819 A CN 202111428819A CN 113840014 B CN113840014 B CN 113840014B
- Authority
- CN
- China
- Prior art keywords
- information processing
- task
- distributed
- heartbeat detection
- redundancy
- 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
-
- 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
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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
- G06F9/54—Interprogram communication
- G06F9/546—Message passing systems or structures, e.g. queues
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L1/00—Arrangements for detecting or preventing errors in the information received
- H04L1/22—Arrangements for detecting or preventing errors in the information received using redundant apparatus to increase reliability
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
- H04L43/0805—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability
- H04L43/0811—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability by checking connectivity
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/10—Active monitoring, e.g. heartbeat, ping or trace-route
Abstract
The invention discloses a distributed task decomposition method adaptive to a high-strength weak connection environment. By introducing a heartbeat detection mechanism, the link condition is confirmed before the distributed task is decomposed, and information of the information processing node which is interrupted or damaged in communication is fed back in time. After the condition of a communication link is confirmed, a switching area redundancy processing coefficient and a system survivability redundancy increment coefficient are introduced in a task quantity estimation stage to distribute redundant information processing capacity for the distributed tasks, and the information processing tasks can still be effectively carried out under the condition that the information processing nodes are partially damaged through a redundancy strategy. And finally, calling a k-means clustering algorithm to perform regional decomposition on the distributed tasks through the estimated task number to obtain a task decomposition result. The distributed task decomposition requirement under the conditions of unstable communication signals and vulnerable information processing equipment can be met. The problem of the distributed system effectively decompose the task under the unstable or extreme calamity condition of communication is solved.
Description
Technical Field
The invention belongs to the technical field of distributed task decomposition control, and particularly relates to a distributed task decomposition method adaptive to a high-strength weak connection environment.
Background
With the development of the distributed technology, the distributed processing technology can combine a plurality of sets of information processing equipment in different locations and different configurations, coordinate and complete large-scale information processing tasks under the unified management control of the distributed control system, and improve the information processing efficiency. The distributed task decomposition technology aims to decompose a large-scale information processing task into a plurality of executable small-scale tasks and distribute a task range to a specific information processing node, and is a more core component of the distributed processing technology. The technical core is that appropriate tasks are distributed to appropriate information processing nodes in real time, and the purposes of fully utilizing system resources and improving task completion efficiency are achieved.
The current distributed task decomposition technology is mainly divided into a redistribution type and a new task type, wherein the redistribution type mainly solves the problems of redistribution and adjustment of tasks in a closed system, the current mature technologies such as load balancing exist, the new task type mainly solves the problem of decomposition processing of a new input task, and various dynamic task distribution methods such as a multi-Agent system are developed. However, these technologies mostly require the information processing nodes to maintain a high degree of coordination to effectively execute the allocation task, and the information link is complete, reliable and stable. There is a lack of a method for coping with a high-intensity environment and communication interference. In the practical application process, the complex and changeable environmental scene can enable the information processing nodes to keep highly cooperative expectation to be challenged by multiple dimensions, for example, communication under disaster relief or extreme weather generally has interference, and the communication between the information processing nodes is intermittent and cannot establish stable connection; disasters such as earthquake, fire and the like can cause damage to information processing equipment, and result in incapability of information processing nodes, and at the moment, assignment of tasks to the damaged information processing nodes can certainly reduce the information processing efficiency of the whole system.
Disclosure of Invention
The invention provides a distributed task decomposition method adaptive to a high-strength weak connection environment aiming at the problem of distributed task decomposition under the conditions of high-strength environment such as natural disasters and unstable communication signals, and the distributed task decomposition method can meet the distributed task decomposition requirements under the conditions of unstable communication signals and vulnerable information processing equipment.
In order to solve the technical problem, the invention discloses a distributed task decomposition method adaptive to a high-strength weak connection environment, which is operated on a distributed system and comprises the following steps:
step 1: an information processing task is input.
Step 2: and (3) carrying out heartbeat detection on all information processing nodes, if the heartbeat detection finds that all the information processing nodes can not feed back heartbeat signals, notifying that the heartbeat detection of the distributed system is invalid, suspending the information processing task, otherwise, notifying that the heartbeat detection of the distributed system is valid, and executing the step 3.
And step 3: and estimating the number of tasks according to the heartbeat detection result.
And 4, step 4: and (3) calling a k-means clustering algorithm to divide the task processing range, and distributing the number of the tasks estimated in the step (3) to a calculation area.
And 5: and outputting a task decomposition result to finish distributed task decomposition.
In step 1 of the invention, the information processing task queue adopts a priority queue, and the information with high importance degree is endowed with high priority, so that the prior dequeue is ensured.
In step 2 of the invention, by setting a heartbeat detection strategy, link conditions are confirmed before distributed tasks are decomposed, information of information processing nodes with communication interruption or damage is fed back in time, the condition that the tasks are distributed to the information processing nodes with damaged or communication links not communicated for calculation is avoided, and the operation efficiency of a distributed system is improved. The method comprises the following specific steps:
(1) constructing information processing node set under current communication link condition
. Wherein the content of the first and second substances,
it is shown that the information processing node,
indicating the number of current information processing nodes.
(2) For each information processing node
And sending a heartbeat detection signal.
(3) If the information processing node
If the heartbeat detection receipt signal is not received within a preset time (such as 1 ms), the information processing node is separated from the information processing node set
Is deleted.
(4) If the information processing node set is processed after the heartbeat detection is finished
And if the information processing node is null, the information processing node is damaged or cannot communicate, and the heartbeat detection is invalid. Otherwise, outputting the information processing node set.
In step 3 of the invention, redundancy information processing capacity is distributed to the distributed tasks by introducing redundancy processing coefficients and system survivability increment coefficients, and redundancy computing power can be provided to ensure the ordered proceeding of the information processing tasks when the performance of the information processing equipment is unstable or damaged. The invention estimates the task number Sum according to the following formula:
Sum=T total amount of / T Single point ×α×β
Wherein:T total amount of Representing the total processing amount of the whole application task;
T single point Representing a single compute node processing capability.
Alpha represents a redundancy processing coefficient and represents the processing capacity of the distributed redundant information, and the redundancy calculation capacity can ensure the ordered execution of information processing tasks when the performance of the information processing equipment is unstable.
Beta represents a system survivability redundancy increment coefficient, which represents that redundant information processing capacity is provided by setting redundant equipment, and the information processing task can still be effectively carried out under the condition that the information processing node is partially damaged.
In step 4 of the invention, the task processing range is divided by using a k-means clustering algorithm, the clustering number is set as the task decomposition number output in step 3, and the iteration termination condition is set that the central point is not changed any more.
The invention confirms the link condition before the distributed task is decomposed by introducing a heartbeat detection mechanism and feeds back the information of the information processing node which is interrupted or damaged in communication in time. After the condition of a communication link is confirmed, a switching area redundancy processing coefficient and a system survivability redundancy increment coefficient are introduced in a task quantity estimation stage to distribute redundant information processing capacity for the distributed tasks, and the information processing tasks can still be effectively carried out under the condition that the information processing nodes are partially damaged through a redundancy strategy. And finally, calling a k-means clustering algorithm to perform regional decomposition on the distributed tasks through the estimated task number to obtain a task decomposition result.
Compared with the prior art, the invention has the following remarkable advantages: (1) the invention confirms the link condition before the distributed task is decomposed, feeds back information of communication interruption or damaged information processing nodes in time, avoids the condition that tasks are distributed to the information processing nodes with damaged or communication links not communicated for calculation, and improves the operation efficiency of the distributed system. (1) The redundant processing coefficient is introduced to distribute redundant information processing capacity for the distributed tasks, and the redundant computing capacity can ensure the ordered execution of the information processing tasks when the performance of the information processing equipment is unstable. (3) And a system survivability redundancy increment coefficient is introduced, and the information processing task can still be effectively carried out under the condition that the information processing node is partially destroyed through a redundancy strategy. (4) The task area decomposition algorithm adopts a k-means clustering algorithm, and has the characteristics of simple practice, high convergence speed, independence from data input sequence and the like.
Drawings
The foregoing and/or other advantages of the invention will become further apparent from the following detailed description of the invention when taken in conjunction with the accompanying drawings.
FIG. 1 is a flow chart of the implementation steps of the present invention.
Fig. 2 is a flow chart of the heartbeat detection procedure of the present invention.
FIG. 3 is a flowchart of the task processing scope dividing step of the present invention.
Detailed Description
The invention discloses a distributed task decomposition method adaptive to a high-strength weak connection environment. By introducing a heartbeat detection mechanism, the link condition is confirmed before the distributed task is decomposed, and information of the information processing node which is interrupted or damaged in communication is fed back in time. After the condition of a communication link is confirmed, a switching area redundancy processing coefficient and a system survivability redundancy increment coefficient are introduced in a task quantity estimation stage to distribute redundant information processing capacity for the distributed tasks, and the information processing tasks can still be effectively carried out under the condition that the information processing nodes are partially damaged through a redundancy strategy. And finally, calling a k-means clustering algorithm to perform regional decomposition on the distributed tasks through the estimated task number to obtain a task decomposition result. The distributed task decomposition requirement under the conditions of unstable communication signals and vulnerable information processing equipment can be met. The problem of the distributed system effectively decompose the task under the unstable or extreme calamity condition of communication is solved.
Referring to fig. 1, according to an embodiment of the present invention, a distributed task decomposition method for adapting to a high-strength weak connection environment includes the following steps:
step 1: a task is input from an information processing task queue.
Rule 1.1: if the task queue is empty, the process flow will block until a new task arrives.
Step 2: and carrying out heartbeat detection on all information processing nodes.
Rule 2.1: and if the heartbeat detection finds that all information processing nodes can not feed back heartbeat signals, the system is informed that the heartbeat detection is invalid, the information processing task is suspended, and the step 1 is returned.
Rule 2.2: and if the heartbeat detection finds that the information processing node can feed back a heartbeat signal, informing the system that the heartbeat detection is effective, and executing the step 3.
And step 3: and estimating the number of tasks according to the heartbeat detection result. Step 4 is performed.
And 4, step 4: and (3) calling a k-means clustering algorithm to divide the task processing range, and distributing the number of the tasks estimated in the step (3) to a reasonable calculation area. Step 5 is performed.
And 5: and outputting a task decomposition result to finish distributed task decomposition.
In step 1, the information processing task queue can adopt a priority queue, and a high priority is given to the information with high importance degree, so that the prior dequeue is ensured. The priority policy may be combined with the validity period of the task or the time of submission of the task, e.g., giving priority to earlier times of submission in the case of priority, giving priority to closer times of failure in the case of priority and same times of submission, etc.
And 2, the center hop detection aims at confirming the link condition before the distributed tasks are decomposed, feeding back information of the information processing nodes with communication interruption or damage in time, and avoiding the condition that the tasks are distributed to the information processing nodes with damaged or communication links not communicated for calculation. The specific steps are shown in figure 2:
(1) constructing information processing node set under current communication link condition
. Wherein the content of the first and second substances,
it is shown that the information processing node,
indicating the number of current information processing nodes.
(2) For each information processing node
And sending a heartbeat detection signal.
(3) If the information processing node
If the heartbeat detection receipt signal is not received within the preset time (such as 1 ms), the information processing node is connected
From a set of information processing nodes
Is deleted.
(4) If the information processing node set is processed after the heartbeat detection is finished
And if the information processing node is null, the information processing node is damaged or cannot communicate, and the heartbeat detection is invalid. Otherwise, outputting the information processing node set
。
In the step 3, the task quantity estimation aims to reasonably estimate the quantity of the tasks needing to be decomposed according to the total processing quantity of the application tasks and the condition of the information processing nodes, the quantity of the tasks is in direct proportion to the total processing quantity of the whole application tasks and in inverse proportion to the processing capacity of a single computing node, and in addition, a redundancy processing coefficient and a system survivability redundancy increment coefficient after the tasks are divided need to be considered. The number of tasks Sum is estimated according to the following formula:
Sum=T total amount of / T Single point ×α×β
Wherein:
T total amount of Represents the total amount of processing of the entire application task, for example: the processing capacity of a certain task is 20000 batches;
T single point Representing a single compute node processing capability, for example: a single system can process 1000 batches.
Alpha represents a redundancy processing coefficient and represents the processing capacity of the distributed redundant information, and the redundancy calculation capacity can ensure the ordered execution of information processing tasks when the performance of the information processing equipment is unstable. For example: the redundant processing coefficient may be set to 1.2.
Beta represents a system survivability redundancy increment coefficient, which represents that redundant information processing capacity is provided by setting redundant equipment, and the information processing task can still be effectively carried out under the condition that the information processing node is partially damaged. For example: the architectural survivability increment factor may be set to 1.5.
For example, if the processing amount of a certain task is 20000 batches, a single system can process 1000 batches on average, the redundancy processing coefficient is set to 1.2, and the system survivability redundancy increment coefficient is set to 1.5, then the specific calculation process for estimating the number of the certain task is as follows: sum =20000/1000 × 1.2 × 1.5=36, and needs to be divided into 36 processing regions, into 36 processing tasks.
In the step 4, the task processing range is divided by using a k-means clustering algorithm, the k-means algorithm is a typical unsupervised learning method and is commonly used for space clustering analysis, the algorithm needs a criterion function for comparing each object value, the criterion of the wiring harness in the iteration process is the function convergence, and the criterion function is defined as:
wherein
Representing selected points
And a cluster center
J denotes the approximation, i.e. the sum of the squared errors, of the n object points in the case of k cluster centers. According to the criterion function, the steps of the k-means clustering algorithm of the invention are shown in FIG. 3:
(1) constructing information processing node set under current communication link condition
. Wherein the content of the first and second substances,
it is shown that the information processing node,
representing the number of current information processing nodes;
(2) taking the number of the tasks output in the step 3 as a clustering number k;
(3) randomly selecting k points in the information processing node set S as central points of object clustering, and initializing k subsets according to the k points;
(4) assigning all object points in the information processing node set S to k subsets according to the rule of criterion function value minimum principle;
(5) after all the points are assigned to the subsets, recalculating the center point of each subset, and taking the centroid of each subset as a new center point to obtain k new center points;
(6) and (5) judging whether the convergence condition is reached, if the convergence condition is reached, outputting a clustering result, and if not, returning to the step (4).
The convergence condition used by the k-means clustering algorithm in the step 4 is generally to determine whether the central point is changed, and if the central point is not changed, the algorithm is considered to be converged.
The present invention provides a distributed task decomposition method adapted to a high-strength weak connection environment, and a plurality of methods and approaches for implementing the technical solution, the above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, a plurality of improvements and modifications may be made without departing from the principle of the present invention, and these improvements and modifications should also be regarded as the protection scope of the present invention. All the components not specified in the present embodiment can be realized by the prior art.
Claims (2)
1. A distributed task decomposition method adapted to a high-strength weak connection environment, said method operating on a distributed system, comprising the steps of:
step 1: inputting an information processing task;
step 2: performing heartbeat detection on all information processing nodes, if the heartbeat detection finds that all the information processing nodes can not feed back heartbeat signals, notifying that the heartbeat detection of the distributed system is invalid, suspending the information processing task, otherwise, notifying that the heartbeat detection of the distributed system is valid, and executing the step 3;
and step 3: estimating the number of tasks according to the heartbeat detection result;
and 4, step 4: calling a k-means clustering algorithm to divide a task processing range, and distributing the number of the tasks estimated in the step (3) to a calculation area;
and 5: outputting a task decomposition result to complete distributed task decomposition;
in the step 1, a priority queue is adopted by an information processing task queue, a high priority is given to a message with high importance degree, and the prior dequeuing of the message is ensured;
in step 2, by setting a heartbeat detection strategy, link conditions are confirmed before distributed task decomposition, and information of information processing nodes with communication interruption or damage is fed back, which comprises the following steps:
(1) building a currentInformation processing node set under communication link condition
Wherein, in the step (A),Nrepresenting information processing nodes, wherein m represents the number of the current information processing nodes;
(2) for each information processing node
Sending a heartbeat detection signal;
(3) if the information processing node
If the heartbeat detection receipt signal is not received within the preset time and i takes a value of 1-m, the information processing node is connected
Deleted from the information processing node set S;
(4) if the information processing node set is processed after the heartbeat detection is finished
If the information processing node set is empty, the information processing node set is indicated to be damaged or incapable of communicating, the heartbeat detection is invalid, otherwise, the information processing node set is output
；
In step 3, a redundancy processing coefficient and a system survivability redundancy increment coefficient are introduced, and the task number Sum is estimated according to the following formula:
wherein:
representing the total processing amount of the whole application task;
α represents a redundant processing coefficient indicating a redundant information processing capability assigned;
beta represents a system survivability redundancy increment coefficient, and represents the information processing capacity for providing redundancy by arranging redundancy equipment.
2. The distributed task decomposition method adaptive to the high-strength weak connection environment according to claim 1, wherein in step 4, the task processing range is divided by using a k-means clustering algorithm, the number of clusters is set as the number of task decompositions output in step 3, and the iteration termination condition is set such that the central point is not changed any more.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111428819.4A CN113840014B (en) | 2021-11-29 | 2021-11-29 | Distributed task decomposition method adaptive to high-strength weak connection environment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111428819.4A CN113840014B (en) | 2021-11-29 | 2021-11-29 | Distributed task decomposition method adaptive to high-strength weak connection environment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113840014A CN113840014A (en) | 2021-12-24 |
| CN113840014B true CN113840014B (en) | 2022-02-22 |
Family
ID=78971903
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111428819.4A Active CN113840014B (en) | 2021-11-29 | 2021-11-29 | Distributed task decomposition method adaptive to high-strength weak connection environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113840014B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113377535A (en) * | 2021-06-08 | 2021-09-10 | 深圳前海微众银行股份有限公司 | Distributed timing task allocation method, device, equipment and readable storage medium |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9485301B2 (en) * | 2013-03-15 | 2016-11-01 | Live Nation Entertainment, Inc. | Prioritized link establishment for data transfer using task scheduling |
| CN104915250B (en) * | 2015-06-03 | 2018-04-06 | 电子科技大学 | It is a kind of to realize the method for making MapReduce data localization in the industry |
| CN109002966B (en) * | 2018-06-25 | 2019-08-09 | 湖南国科轩宇信息科技有限公司 | A kind of more star mission planning methods based on K mean cluster |
| CN111045808B (en) * | 2019-12-17 | 2023-04-18 | 广东工业大学 | Distributed network task scheduling method and device |
-
2021
- 2021-11-29 CN CN202111428819.4A patent/CN113840014B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113377535A (en) * | 2021-06-08 | 2021-09-10 | 深圳前海微众银行股份有限公司 | Distributed timing task allocation method, device, equipment and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113840014A (en) | 2021-12-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3399426B1 (en) | Method and device for training model in distributed system | |
| Ventura et al. | Grouping parts and tools in flexible manufacturing systems production planning | |
| US20050060608A1 (en) | Maximizing processor utilization and minimizing network bandwidth requirements in throughput compute clusters | |
| CN102271145A (en) | Virtual computer cluster and enforcement method thereof | |
| CN113259469B (en) | Edge server deployment method, system and storage medium in intelligent manufacturing | |
| WO2012000997A1 (en) | An apparatus for processing a batched unit of work | |
| Ozfatura et al. | Gradient coding with clustering and multi-message communication | |
| Ramanathan et al. | Delivery of time-critical messages using a multiple copy approach | |
| US20230244537A1 (en) | Efficient gpu resource allocation optimization method and system | |
| CN110851277A (en) | Task scheduling strategy based on edge cloud cooperation in augmented reality scene | |
| Altiok et al. | Analysis of production lines with general service times and finite buffers: a two-node decomposition approach | |
| CN111443993A (en) | Method for realizing large-scale container cluster | |
| Dai et al. | A multiclass station with Markovian feedback in heavy traffic | |
| CN113840014B (en) | Distributed task decomposition method adaptive to high-strength weak connection environment | |
| US11936758B1 (en) | Efficient parallelization and deployment method of multi-objective service function chain based on CPU + DPU platform | |
| CN102243653A (en) | Method and device for managing database connections | |
| Raz | The dynamic two phase commitment (d2pc) protocol | |
| Zhou et al. | Task offloading strategy of 6G heterogeneous edge-cloud computing model considering mass customization mode collaborative manufacturing environment | |
| Liu et al. | Emergency scheduling optimization simulation of cloud computing platform network public resources | |
| CN114363988B (en) | Clustering method and device and electronic equipment | |
| CN116089079A (en) | Big data-based computer resource allocation management system and method | |
| CN112738225B (en) | Edge calculation method based on artificial intelligence | |
| Miyamoto et al. | FMPA: A Fault-Tolerant Multi-Microprocessor System Based on Autonomous Decentralization Concept | |
| CN111416338A (en) | Economic dispatching method for event-triggered power system under quantitative communication | |
| CN105512087B (en) | Reliability evaluation method of resource-constrained multi-node computing system |
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 | ||
| CB02 | Change of applicant information |
Address after: 210000 No.1, Lingshan South Road, Qixia District, Nanjing City, Jiangsu Province Applicant after: THE 28TH RESEARCH INSTITUTE OF CHINA ELECTRONICS TECHNOLOGY Group Corp. Address before: 210000 1 East Street, alfalfa garden, Qinhuai District, Nanjing, Jiangsu. Applicant before: THE 28TH RESEARCH INSTITUTE OF CHINA ELECTRONICS TECHNOLOGY Group Corp. |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |