CN107508698B - Software defined service reorganization method based on content perception and weighted graph in fog calculation - Google Patents
Software defined service reorganization method based on content perception and weighted graph in fog calculation Download PDFInfo
- Publication number
- CN107508698B CN107508698B CN201710597297.8A CN201710597297A CN107508698B CN 107508698 B CN107508698 B CN 107508698B CN 201710597297 A CN201710597297 A CN 201710597297A CN 107508698 B CN107508698 B CN 107508698B
- Authority
- CN
- China
- Prior art keywords
- service
- content
- fog
- strategy
- data
- 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
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
- H04L41/5041—Network service management, e.g. ensuring proper service fulfilment according to agreements characterised by the time relationship between creation and deployment of a service
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0803—Configuration setting
- H04L41/0823—Configuration setting characterised by the purposes of a change of settings, e.g. optimising configuration for enhancing reliability
- H04L41/0833—Configuration setting characterised by the purposes of a change of settings, e.g. optimising configuration for enhancing reliability for reduction of network energy consumption
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0893—Assignment of logical groups to network elements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
- H04L41/5041—Network service management, e.g. ensuring proper service fulfilment according to agreements characterised by the time relationship between creation and deployment of a service
- H04L41/5051—Service on demand, e.g. definition and deployment of services in real time
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention provides a software defined service reorganization method based on content perception and weighted graph in fog calculation, which is operated in a hierarchical fog node architecture, wherein a fog node is composed of three layers, namely a service reconstruction layer, a content-driven control layer and a user-defined reconstruction layer, and three corresponding storage spaces, namely a service resource library, a content-based strategy library and a demand library, are divided in the fog node and are respectively used for storing a service module, a content-operation mapping table and the preprocessing requirements of fog users. Compared with a service providing mechanism in the traditional fog node, the novel software defined service recombination mechanism in the fog node provided by the invention can reduce time cost and calculation cost to a great extent, thereby reducing the overall energy consumption of a fog system.
Description
Technical Field
The invention relates to a software defined service reorganization mechanism based on content perception and a weighted graph in fog calculation.
Background
As an intermediate layer between a remote cloud platform and a large number of bottom-layer devices, the cloud computing concentrates the capabilities of content processing, computing, storage and the like in nodes with certain intelligence at the network edge. In the mode, the original data collected from the bottom layer large-scale sensor network cannot be directly uploaded to the cloud, and the original data are firstly subjected to relevant preprocessing by the local fog node according to the user requirements and then delivered to the cloud and upper layer users for further storage and processing. At present, fog computing has been considered as a trend of next generation Wireless Sensor Networks (WSNs) due to its advantages in terms of low latency and ease of cloud computing burden.
With the gradual application and large-scale deployment of the fog calculation, how to effectively utilize the limited computing capacity and related resources of each fog node is very necessary, so that the energy conservation on the fog calculation level is realized. Under the application and deployment scenes of fog computing, on one hand, different fog computing users have different requirements on the data preprocessing strategies in the fog nodes, and the requirement of each user presents the characteristic of changing along with time; on the other hand, the preprocessing requirements of data of different content types (such as video, audio, etc.) are also greatly different. The dynamics and the differences of the preprocessing requirements on the two levels of the user and the data content enable the user-defined strategy and the preprocessing service based on the content to be recombined into the key for the efficient utilization of the fog node resources. However, the proposed mist node architecture has the characteristics of relative solidification and static property, and cannot meet the requirement.
In previous research aiming at the WSNs energy saving problem, most solutions surround a resource scheduling mechanism in the cloud to realize energy perception in the scene. The research on the optimized fog node architecture only concerns the interaction of each node in the fog node network on the network level, does not relate to the efficient reuse and recombination optimization of resources in a single fog node, and cannot reduce the energy consumption of the whole fog computing system.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a software defined service reorganization method based on content perception and a weighted graph in fog calculation, which realizes the cyclic utilization and reorganization of resources by abstracting, packaging and modularizing fog node calculation resources, applies different data preprocessing strategies according to different perceived content types, and simultaneously provides a programmable interface for upper-layer fog users to allow the users to perform custom control on the preprocessing of bottom-layer data in the fog nodes. On one hand, the mechanism can meet different requirements and simultaneously realize the efficient arrangement and use of the fog node resources; on the other hand, the flexibility and the dynamic property of the processing strategy are also greatly improved.
The technical scheme adopted by the invention is as follows:
a software defined service recombination method based on content perception and weighted graph in fog calculation is operated in a hierarchical fog node architecture, the fog node is composed of three layers which are a service reconstruction layer, a content-driven control layer and a user-defined reconfiguration layer respectively, and at the same time, three corresponding storage spaces, namely a service resource library, a content-based strategy library and a demand library, are divided and are used for storing a service module, a content-operation mapping table and preprocessing requirements of fog users respectively.
The fog node collects and extracts important data attributes from the data packet header and the load so as to complete the classification of the content, and then utilizes the content tag technology to identify a large amount of original data generated by the wireless sensor network based on the difference of the content types, so as to execute different preprocessing strategies on the content data of different tags in the following process.
The fog user controls a specific bottom layer preprocessing strategy by utilizing an upward interface provided by a user-defined reconfiguration layer, so that the calling of a service module in the service arrangement process conforms to the actual requirement of the user, and meanwhile, the fog user completes flexible management of the service module through a programming interface, including rewriting, changing, deleting and increasing of the service module, so as to improve the flexibility of dynamic service reconfiguration, ensure that various preprocessing requirements can be met through arrangement of modularized resources, and further reduce the energy consumption generated by redeployment of fog nodes.
When the selection module is instantiated, one preprocessing strategy is composed of one or more operations according to a certain logic sequence, the same operation corresponds to one or more service modules which realize the operation with different calculation cost and algorithm, data of each content type corresponds to different preprocessing strategies, namely corresponds to the combination of different service modules, the original data labeled by the CB L technology searches the processing strategies corresponding to the data in a strategy library stored in a content driving control layer and completes the service modules required to be called of the strategies, then the required service modules are arranged according to the pre-defined preprocessing flow in the strategy library in a certain logic sequence, and the data are instantiated by a service reconstruction layer execution module of a fog node, parameters are transmitted, and finally output is obtained, and the processed data or valuable results are delivered to upper-layer fog users.
For data of different content types, different upper-layer fog users and time-varying preprocessing requirements of the users, one set of independent preprocessing program does not need to be deployed and configured for each strategy, one or more service modules are called only on the basis of output of content perception and actual service requirements of the users, and when the requirements change, the service modules are recycled and dynamically recombined, so that the utilization efficiency of service resources can be improved and the energy consumption of fog nodes can be reduced while different requirements are flexibly met.
The service reconstruction layer comprises a service recombination unit for calling, instantiating and controlling the required service modules and executing the service recombination unit, a content sensing unit for extracting and classifying the characteristics of the raw data collected from the wireless sensor network, a module management unit for registering, adding, deleting and updating each service module, a service resource library arranged in the service reconstruction layer, a data cache region for temporarily storing the raw data and a service module pool for storing abstract packaging resources, a large amount of raw data collected from the bottom wireless sensor network are temporarily placed in the data cache region for waiting processing after being uploaded to a fog node, a data packet entering a preprocessing flow is firstly transmitted to the content sensing unit, the unit extracts and analyzes key attributes from a data packet header and a load, and a classification result of the content is output, and then based on different content types, the original data generated by the wireless sensor network is identified by using a content tag technology, so that reasonable service reconstruction can be completed according to the content.
When a specific preprocessing strategy and a required service module issued by a content-driven control layer are received, a service reconfiguration layer selects a corresponding module from a service module pool, instantiates the corresponding module by using preset or parameters specified in the strategy, arranges and reconfigures the modules according to a given logic sequence so as to complete a specific data processing requirement, and can replace or reconfigure the modules on the basis of the existing service modules and arrangement to meet the new preprocessing requirement when data processing strategies are updated by different types of data packets collected from a bottom layer network or upper layer fog users.
The content-driven control layer comprises a requirement analysis unit for analyzing user requirements and a strategy formulation unit for generating and optimizing a data preprocessing strategy, and meanwhile, a content-based strategy library arranged in the content-driven control layer comprises a service logic sequence table for storing preset data processing procedures and a content-operation mapping table for defining different content type data and corresponding operations required by the content type data.
The data packet identified by the content label technology is uploaded to the strategy making unit, the unit searches the processing operation needed by the unit in the content-operation mapping table, and meanwhile, if the upper-layer fog user has special requirements, the requirements are also issued to the requirement analyzing unit and are used for controlling the execution of specific operation or the selection of modules.
In the process of generating the strategy, the strategy generation process comprises two substeps, namely strategy formulation and strategy optimization based on the weighted graph, wherein the substeps comprise the following steps: the strategy making refers to that a service reconstruction layer selects a needed module for initialization and recombination according to a service strategy issued by a content-driven control layer so as to meet the requirements of different fog users and content processing and obtain all data preprocessing strategies which can be executed on certain types of content; the strategy optimization refers to optimizing a data preprocessing strategy according to energy consumption and specific requirements issued by a user, and selecting a relatively reasonable strategy to guide module recombination of a service reconstruction layer.
Compared with the prior art, the invention has the following beneficial effects:
the invention meets the requirements of fog users and different data contents on differentiated processing in a fog calculation scene, and realizes efficient utilization of resources in a single fog node. The invention abstracts and encapsulates the fog node computing resources to allow service recombination, namely when the processing requirement changes, the invention can realize resource reuse and strategy updating only by rearranging the modularized processing operation without reconfiguring and deploying all data processing flows. The service recombination mechanism defined by the software promotes the energy-saving optimization of a single fog node, and further reduces the energy consumption of the whole fog system.
Drawings
Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:
FIG. 1 is an infrastructure of the present invention;
FIG. 2 is a schematic diagram of service reconfiguration according to the present invention;
FIG. 3 is a service policy making and demand resolution interaction model of the present invention;
FIG. 4 is a diagram illustrating delay comparison between different service providing mechanisms;
fig. 5 is a comparison of complexity of service provisioning for different devices under the mechanism of the present invention.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that it would be obvious to those skilled in the art that various changes and modifications can be made without departing from the spirit of the invention. All falling within the scope of the present invention.
The software defined service reorganization mechanism based on content awareness and weighted graph in fog computing provided by the invention has the basic architecture as shown in fig. 1.
The software-defined service reorganization mechanism under the fog computing is operated in a hierarchical fog node architecture. The cloud nodes are marginalized locally, and service functions of computing, storing, preprocessing and the like of data of different content types are provided for upper-layer cloud users. The service arrangement mechanism of the fog node is different from the traditional fog node pretreatment service in the greatest way, namely, the computing service resources of the fog node are packaged in a modularization mode, so that the service recombination and resource reuse of software definition are realized, the resource utilization rate is further improved, and the purpose of reducing the overall energy consumption of a fog system is achieved.
As shown in fig. 1, the system-haze node corresponding to this service is composed of three layers, namely a service reconfiguration layer, a content-driven control layer, and a user-defined reconfiguration layer. Meanwhile, the fog node needs to be divided into three corresponding storage spaces, namely a service resource library, a content-based policy library and a requirement library, which are respectively used for storing the service module, the content-operation mapping table and the preprocessing requirement of the fog user.
The fog user utilizes an interface provided upwards by a reconfiguration layer defined by a user to control a specific bottom layer preprocessing strategy, so that the module calling in the service arranging process conforms to the actual requirement of the fog user, and simultaneously, the fog user finishes flexible management on a service module through a programming interface, including rewriting, changing, deleting and increasing the module, so as to improve the flexibility of dynamic service recombination, ensure that various preprocessing requirements can be met through arranging modular resources, and further reduce the energy consumption generated by re-arranging the fog node.
The original data after CB L technical labeling searches the corresponding processing strategy in a strategy library stored in a control layer of a content drive and completes the module required to be called of the strategy, then arranges the required service modules in a certain logic sequence according to the pre-defined preprocessing flow in the strategy library, instantiates the service reconstruction layer execution module of a fog node, transmits parameters, finally obtains output, and delivers the processed data or the valuable result to an upper-layer fog user.
In the system, for data of different content types, different upper-layer fog users and time-varying preprocessing requirements thereof, a set of independent preprocessing program does not need to be deployed and configured for each strategy, one or more service modules are called only based on output of content perception and actual service requirements of the users, and when the requirements change, the service modules are recycled and dynamically recombined, so that the utilization efficiency of service resources can be improved and the energy consumption of fog nodes can be reduced while different requirements are flexibly met.
Fig. 2 shows a specific design of a service reconfiguration mechanism in the proposed cloud node architecture. The key components in the mechanism comprise a service reorganization unit used for calling, instantiating and controlling the required service modules to execute, a content perception unit used for carrying out feature extraction and classification on the raw data collected from the WSNs network, and a module management unit used for registering, adding, deleting and updating each service module. Meanwhile, a service resource library is set, and the resource library comprises a data cache region for temporarily storing original data and a service module pool for storing abstracted encapsulated resources.
The data packet entering the preprocessing flow is firstly transmitted to a content sensing unit, the unit extracts and analyzes key attributes from a data packet header and a load, the content classification result is output, and then the original data generated by the WSNs network is identified by utilizing a CB L technology based on different content types so as to complete reasonable service reconstruction according to the content in the following.
On the other hand, after receiving the specific preprocessing strategy and the required service module issued by the content-driven control layer, the service reconfiguration unit selects the corresponding module from the service module pool, instantiates the module by using the preset or the parameters specified in the strategy, and arranges and reconfigures the module according to the given logic sequence so as to complete the specific data processing requirement.
Here, the operations that make up a certain processing policy correspond to one or more service modules implemented by different algorithms, so that selecting different modules for data preprocessing has different computational costs and energy consumption. The relationship between a service module and its corresponding operations is defined as:
wherein, DPOi represents the collection of in different service modules contained in the ith data processing operation.
Each service module and the interaction relationship between them can be expressed as:
wherein the content of the first and second substances,
o represents the specific operation performed on the data packet by the algorithm encapsulated in the module;
in represents a series of inputs to the service module;
out represents a series of outputs of the service module; the parameters are specified by the service reconstruction unit in a numeric or character string format, the parameters have meanings including but not limited to data compression rate, filtering conditions, packet size threshold and the like, and the parameters are dynamically adjusted according to the current state of the network and the requirements of the fog users.
When different types of data packets are collected from the underlying network or the upper layer mist user updates the data processing strategy, the service reconfiguration mechanism can replace or reconfigure the existing service modules on the basis of arrangement to meet the new preprocessing requirement.
The content-driven control mechanism is the core for realizing software defined service recombination in the fog node, and comprises a requirement analysis unit for analyzing the requirement of a user and a strategy formulation unit for generating and optimizing a data preprocessing strategy. Meanwhile, a content-based policy library is set, wherein the policy library comprises a service logic sequence table for storing preset data processing procedures and a content-operation mapping table for defining different content type data and corresponding operations required by the different content type data. The interaction model of the mechanism with other units is shown in fig. 3.
The data packet identified by the CB L technology is uploaded to a policy making unit, and the unit searches the processing operation needed by the data packet in a content-operation mapping table, meanwhile, if the upper-layer fog user has special requirements, the requirements are also issued to a requirement analyzing unit and used for controlling the execution of specific operation or the selection of modules.
1) Policy making
The service reconfiguration layer selects the required modules to initialize and recombine according to the service strategy issued by the content-driven control layer, so that the requirements of different fog users and content processing are met. The service policy is defined as:
DP＝{mod,para,ord}, (3)
wherein mod represents the required service modules; para represents the initial parameters used during the pretreatment; ord denotes the execution order of the respective operations.
In the process of policy making, the service modules involved in the policy and the conversion relation between the service modules are abstracted into a global weighted graph in the policy making unit
Wherein
The method comprises the following steps that (1) the data are a limited non-empty set consisting of S points, and each node represents one possible intermediate state of the data in the preprocessing process;
is composed of
A binary subset of (a) represents a particular service module, corresponding to an edge in the weighted graph. An edge eij connecting ni and nj indicates that the operation executed by the module corresponding to the edge eij is required to be called to make the data to be processed transfer from the state i to the critical state j. Typically, there are multiple edges (service modules) between two nodes (states) that have a tandem relationship, because the same operation can be performed by modules that are implemented by multiple algorithms.
After receiving the original data Cg subjected to analysis and feature extraction, the strategy making unit searches all operations required by the data according to the label to content-operation mapping table, and determines a starting point nl and an end point nk on the way of the strategy by combining the current state of the data packet; then, construct the corresponding pre-processing strategy subgraph for the data corresponding to the content type, i.e. follow the service logic sequence table and the processing operation to be included to generate subgraph E ═ Vs,Es) The sub-graph contains all the strategies that can satisfy the preprocessing requirement of the content Cg processing.
2) Policy optimization
Strategy sub-graph E ═ V (V) corresponding to original data Cg of certain type of content generated in strategy making processs,Es) Then, all data preprocessing strategies that can be executed on Cg are obtained, and each strategy contains different service modules, namely, corresponding to different calculation cost and module scheduling requirements specific to fog users. Then root should beOptimizing the initial strategy set according to energy consumption and specific requirements issued by users, and selecting a relatively reasonable strategy to guide module recombination of a service reconstruction layer.
In order to reduce the overall energy consumption, an adjacency matrix is introduced in the strategy optimization process to measure and compare different strategies in the strategy set. And G is the total amount of data packets input to the fog node for preprocessing in a certain time window, wherein Cg is recorded as G, and G is 1,2, … and G. H is the number of the needed preprocessing operations, K is the number of modules which can realize a specific operation by using various algorithms, and hk represents the kth service module corresponding to the H operation. It is apparent that there is a relationship H-S-1 where S represents the number of all states it may be in during data processing. Marking global processing strategy map with E ═ eij
A corresponding adjacency matrix, wherein:
similarly, a submatrix of a global policy is defined
The matrix consists of G × K Boolean values
The composition is as follows:
matrix array
To identify global policy maps
Edge { e ] of (1)ijWhether it is chosen into the subgraph, i.e. whether the h-th operation isData C for a particular content typegAs needed.
In order to evaluate the total resource consumption of each strategy in the strategy set generated in the last step, a weight matrix is defined
Each element in the matrix
The calculation cost of the k module corresponding to the h operation. The purpose of policy optimization is to select a module sequence with relatively minimum resource consumption on the premise of meeting the requirements of users. Here, a parameter Res is introduced to represent the total resource consumption for all original packets within a certain time window, where:
in principle, the policy making unit always selects the preprocessing policy with the least total resource consumption to reorganize the modules, but in some cases, the fog user may specify some special requirements according to the actual business scenario through the external interface provided by the user-defined reconfiguration layer, for example, the preprocessing of some kind of content must call some or some specific service modules, etc. The strategy making unit sets the user-defined events as the highest priority, and optimizes the resource use as much as possible on the premise of ensuring the user requirements.
As shown in fig. 4, the figure illustrates the delay penalty of the service mechanism of the fog node according to the present invention and other conventional service providing mechanisms when different numbers of service modules are recombined.
As shown in fig. 5, the figure describes the calculation cost of different devices under the fog node service mechanism according to the present invention when different numbers of service modules are recombined.
Experiments show that compared with a service providing mechanism in the traditional fog node, the novel software defined service recombination mechanism in the fog node provided by the invention can reduce the time cost and the calculation cost to a great extent, thereby reducing the overall energy consumption of a fog system.
The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes or modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention. The embodiments and features of the embodiments of the present application may be combined with each other arbitrarily without conflict.
Claims (6)
1. A software defined service reorganization method based on content perception and weighted graph in fog computing is characterized in that the service reorganization method is operated in a hierarchical fog node architecture, a fog node is composed of three layers which are a service reconfiguration layer, a content-driven control layer and a user-defined reconfiguration layer respectively, and meanwhile, three corresponding storage spaces, namely a service resource library, a content-based strategy library and a demand library, are divided from the fog node and are used for storing a service module, a content-operation mapping table and a preprocessing requirement of a fog user respectively;
the service reconstruction layer comprises a service recombination unit for calling, instantiating and controlling the required service modules and executing the service recombination unit, a content sensing unit for extracting and classifying the characteristics of the raw data collected from the wireless sensor network, a module management unit for registering, adding, deleting and updating each service module, a service resource library arranged in the service reconstruction layer, a data cache region for temporarily storing the raw data and a service module pool for storing abstract packaging resources, a large amount of raw data collected from the bottom wireless sensor network are temporarily placed in the data cache region for waiting processing after being uploaded to a fog node, a data packet entering a preprocessing flow is firstly transmitted to the content sensing unit, the unit extracts and analyzes key attributes from a data packet header and a load, and a classification result of the content is output, then based on different content types, the original data generated by the wireless sensor network is identified by using a content tag technology, so that reasonable service reconstruction can be completed according to the content;
when a specific preprocessing strategy and a required service module issued by a content-driven control layer are received, a service reconfiguration layer selects a corresponding module from a service module pool, instantiates the corresponding module by using preset or parameters specified in the strategy, and arranges and reconfigures the modules according to a given logic sequence so as to complete a specific data processing requirement, and when different types of data packets are collected from a bottom layer network or upper layer mist users update the data processing strategy, the service reconfiguration layer can replace or reconfigure the data processing strategy on the basis of the existing service modules and arrangement so as to meet the new preprocessing requirement;
the content-driven control layer comprises a requirement analysis unit for analyzing user requirements and a strategy formulation unit for generating and optimizing a data preprocessing strategy, and meanwhile, a content-based strategy library arranged in the content-driven control layer comprises a service logic sequence table for storing a preset data processing flow and a content-operation mapping table for defining different content type data and corresponding operations required by the different content type data;
the data packet identified by the content label technology is uploaded to the strategy making unit, the unit searches the processing operation needed by the unit in the content-operation mapping table, and meanwhile, if the upper-layer fog user has special requirements, the requirements are also issued to the requirement analyzing unit and are used for controlling the execution of specific operation or the selection of modules.
2. The method as claimed in claim 1, wherein the fog node collects and extracts important data attributes from the data packet header and the payload, so as to complete the classification of the content, and then identifies a large amount of original data generated by the wireless sensor network by using a content tagging technique based on the difference of the content types, so as to subsequently execute different preprocessing strategies on the content data of different tags.
3. The method for reorganizing software defined services based on content awareness and weighted graph in fog computing according to claim 1, wherein a fog user controls a specific underlying preprocessing policy by using an interface provided upwards by a user-defined reconfiguration layer, so that a service module call in a service arrangement process conforms to the actual requirement of the user, and the fog user completes flexible management of the service module through a programming interface, including rewriting, changing, deleting and adding of the service module.
4. The method for software-defined service reorganization based on content awareness and weighted graph in fog computing according to claim 1, wherein when the selection module is instantiated, a preprocessing strategy is composed of one or more operations according to a certain logic sequence, the same operation corresponds to one or more service modules which realize the operation with different computation costs and algorithms, data of each content type corresponds to different preprocessing strategies, namely, combinations of different service modules, the original data labeled by the CB L technology searches a corresponding processing strategy in a strategy library stored in a content-driven control layer and completes a service module which needs to be called of the strategy, then the needed service module is instantiated in a certain logic sequence according to a pre-defined preprocessing flow in the strategy library, and the instantiated and parameter-transferred service reorganization layer execution module of the fog node is delivered to the upper-layer fog user, and finally the processed data or valuable results are delivered to the upper-layer fog user.
5. The method for recombining software-defined services based on content awareness and weighted graph in fog computing according to claim 1, wherein for data of different content types, different upper-layer fog users and their time-varying preprocessing requirements, only one or more service modules need to be invoked based on the output of content awareness and the actual business needs of the users, and when the requirements change, the service modules are recycled and dynamically recombined.
6. The method for restructuring software defined services based on content awareness and weighted graph in fog calculation according to claim 1, wherein the policy generation process comprises two sub-steps, namely policy making and policy optimization based on weighted graph, wherein: the strategy making refers to that a service reconstruction layer selects a needed module for initialization and recombination according to a service strategy issued by a content-driven control layer so as to meet the requirements of different fog users and content processing and obtain all data preprocessing strategies which can be executed on certain types of content; the strategy optimization refers to optimizing a data preprocessing strategy according to energy consumption and specific requirements issued by a user, and selecting a relatively reasonable strategy to guide module recombination of a service reconstruction layer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710597297.8A CN107508698B (en) | 2017-07-20 | 2017-07-20 | Software defined service reorganization method based on content perception and weighted graph in fog calculation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710597297.8A CN107508698B (en) | 2017-07-20 | 2017-07-20 | Software defined service reorganization method based on content perception and weighted graph in fog calculation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107508698A CN107508698A (en) | 2017-12-22 |
| CN107508698B true CN107508698B (en) | 2020-07-24 |
Family
ID=60678914
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710597297.8A Active CN107508698B (en) | 2017-07-20 | 2017-07-20 | Software defined service reorganization method based on content perception and weighted graph in fog calculation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107508698B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110473120A (en) * | 2018-05-10 | 2019-11-19 | 深圳富桂精密工业有限公司 | Micro services isomery regenerative system, method and storage medium based on industry internet |
| EP3407194A3 (en) | 2018-07-19 | 2019-03-20 | Erle Robotics, S.L. | Method for the deployment of distributed fog computing and storage architectures in robotic modular components |
| CN111857886A (en) * | 2019-04-26 | 2020-10-30 | 张明明 | Software running method, system, computing equipment and storage medium |
| CN110555138B (en) * | 2019-08-05 | 2022-09-13 | 慧镕电子系统工程股份有限公司 | Hybrid cloud storage method under cloud computing architecture |
| CN114296947B (en) * | 2022-03-09 | 2022-07-08 | 四川大学 | Multi-calculation model management method for complex scene |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103688489A (en) * | 2012-12-03 | 2014-03-26 | 华为技术有限公司 | Method for strategy processing and network equipment |
| CN106533806A (en) * | 2016-12-26 | 2017-03-22 | 上海交通大学 | Method for providing cross-layer quality of service (QoS) based on application awareness in multi-tenant software defined network (SDN) |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10139279B2 (en) * | 2015-05-12 | 2018-11-27 | BioSensing Systems, LLC | Apparatuses and methods for bio-sensing using unmanned aerial vehicles |
-
2017
- 2017-07-20 CN CN201710597297.8A patent/CN107508698B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103688489A (en) * | 2012-12-03 | 2014-03-26 | 华为技术有限公司 | Method for strategy processing and network equipment |
| CN106533806A (en) * | 2016-12-26 | 2017-03-22 | 上海交通大学 | Method for providing cross-layer quality of service (QoS) based on application awareness in multi-tenant software defined network (SDN) |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107508698A (en) | 2017-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107508698B (en) | Software defined service reorganization method based on content perception and weighted graph in fog calculation | |
| CN110163016B (en) | Hybrid computing system and hybrid computing method | |
| TW202026858A (en) | Exploiting activation sparsity in deep neural networks | |
| CN109997154A (en) | Information processing method and terminal device | |
| CN111339027B (en) | Automatic design method of reconfigurable artificial intelligent core and heterogeneous multi-core chip | |
| CN102541640A (en) | Cluster GPU (graphic processing unit) resource scheduling system and method | |
| KR102259927B1 (en) | Workflow engine framework | |
| CN114915629B (en) | Information processing method, device, system, electronic equipment and storage medium | |
| CN108364061A (en) | Arithmetic unit, operation execute equipment and operation executes method | |
| CN110650347B (en) | Multimedia data processing method and device | |
| US20200409744A1 (en) | Workflow engine framework | |
| CN104657149A (en) | Software framework implementation method of management module of storage system | |
| CN107633298A (en) | A kind of hardware structure of the recurrent neural network accelerator based on model compression | |
| Mora et al. | Distributed architectures for intensive urban computing: a case study on smart lighting for sustainable cities | |
| CN114915630A (en) | Task allocation method based on Internet of things equipment, network training method and device | |
| CN102054108B (en) | Lower hardware mapping method of integrated circuit, and time-space diagram compression method and device | |
| CN108345934A (en) | A kind of activation device and method for neural network processor | |
| CN111611062A (en) | Cloud-edge collaborative hierarchical computing method and cloud-edge collaborative hierarchical computing system | |
| CN103164476A (en) | Execution method and execution device of applying metadata to describe files in business intelligence (BI) | |
| CN114841345A (en) | Distributed computing platform based on deep learning algorithm and application thereof | |
| CN107360031B (en) | Virtual network mapping method based on optimized overhead-to-revenue ratio | |
| CN110324204A (en) | A kind of high speed regular expression matching engine realized in FPGA and method | |
| WO2020062277A1 (en) | Management method and apparatus for computing resources in data pre-processing phase of neural network | |
| CN109343947A (en) | A kind of resource regulating method and device | |
| CN114912618A (en) | Quantum computing task scheduling method and device and quantum computer operating 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |