CN113162968B - Unified characterization framework of network knowledge in intelligent networking and implementation method - Google Patents
Unified characterization framework of network knowledge in intelligent networking and implementation method Download PDFInfo
- Publication number
- CN113162968B CN113162968B CN202110083207.XA CN202110083207A CN113162968B CN 113162968 B CN113162968 B CN 113162968B CN 202110083207 A CN202110083207 A CN 202110083207A CN 113162968 B CN113162968 B CN 113162968B
- Authority
- CN
- China
- Prior art keywords
- network
- knowledge
- network knowledge
- module
- intelligent
- 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
- H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
-
- 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
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a unified characterization framework and an implementation method for network knowledge in an intelligent networking. The architecture comprises modules of network knowledge hierarchical classification, network knowledge space construction, local network knowledge representation, cooperative network knowledge representation, network knowledge storage, network knowledge fusion, network knowledge verification and network knowledge reasoning. The method comprises the following steps: setting four layers of network knowledge construction, network knowledge representation, network knowledge management and network knowledge service; the network knowledge construction layer is responsible for determining the scope of network knowledge in the multi-domain environment; the network knowledge representation layer is responsible for systematic identification of network knowledge structure, mode and content; the network knowledge management layer is responsible for network knowledge storage, fusion and verification under the condition of network communication resource limitation; the network knowledge service layer is responsible for network knowledge learning, reasoning, distribution and acquisition on demand. The invention can support network elements to complete the functions of routing addressing, management control, safety and the like by utilizing network knowledge, thereby reasonably deploying network resources.
Description
Technical Field
The invention relates to the technical field of information communication networks, in particular to a unified characterization framework and an implementation method for network knowledge in an intelligent networking.
Background
The intelligent connection method is oriented to three-element and all-thing intelligent connection of the new human-computer things, the technical characteristics of the information communication network are expanded from human-computer-based connection to intelligent connection based on the human-computer things, the technical center of gravity is changed from information service to intelligent information service, and the technical capability is expanded from connection information to connection intelligence.
The maximum characteristic of the future network is to realize cognitive intelligence, the intelligent agents in the network can realize direct communication in the knowledge level, and according to certain rules and mechanisms, the network organization-intelligent networking with self-optimization, self-learning, self-evolution and self-management is formed like the human society. Although the present society has emerged a great number of intelligent agents at various levels, the information communication network lacks an intelligent connection mechanism, and the great number of intelligent agents are not directly connected with each other on the knowledge level.
In the future intelligent networking design, network knowledge such as wireless channel characteristics, network topology, service types and the like drives a network to complete connection and link establishment of network elements, and autonomous realization of functions such as network management control and network behavior tracing is supported. At present, in the field of information communication networks, the concept of "network knowledge" does not appear at home and abroad, and some related researches are independent developments only, for example, cognitive radio is limited in the field of wireless spectrum, and intelligent network management is mainly used for improving the intelligent management level of the network. The network elements are oriented to the requirements of intelligent networking, optimization, operation and maintenance interaction and the like by using network knowledge, how to design and realize a unified characterization framework of the network knowledge in the intelligent networking is provided, theoretical guidance is provided for the flowing and application of the network knowledge in the intelligent networking, and the method is one of key problems to be solved in the design of the intelligent networking in the future.
Disclosure of Invention
The invention aims to provide a network knowledge uniform representation framework in an intelligent network and an implementation method thereof.
The technical solution for realizing the purpose of the invention is as follows: the utility model provides a unified characterization framework of network knowledge in intelligent networking, includes hierarchical classification module of network knowledge, network knowledge space construction module, local network knowledge characterization module, cooperation network knowledge characterization module, network knowledge storage module, network knowledge fusion module, network knowledge verification module, network knowledge reasoning module, wherein:
the network knowledge classification module provides the category information of the network knowledge for the local network knowledge representation module;
the network knowledge space construction module provides network knowledge such as facts, concepts and rules obtained from network cognition and expert cognition for the local network knowledge characterization module, and specifically comprises signal receiving intensity Doppler expansion, network topology, network scale, node density, user geographic position, user movement rate, an anti-interference strategy and a spectrum management strategy;
the collaborative network knowledge representation module provides unified description information of the collaborative network knowledge for the local network knowledge representation module, wherein the unified description information comprises network knowledge concepts, objects, attributes and the names of models;
the local network knowledge representation module provides the network knowledge after representation for the network knowledge storage module, and the network knowledge storage adopts a Mysql database, a file and a graph database to perform centralized or distributed storage of the network knowledge according to actual conditions;
the network knowledge fusion module provides the aligned and disambiguated network knowledge for the network knowledge storage module;
the network knowledge verification module provides the network knowledge after error correction and updating for the network knowledge storage module;
the network knowledge storage module provides a knowledge source for the network knowledge reasoning module;
and the network knowledge reasoning module provides the implicit network knowledge generated by prediction to the network knowledge storage module.
A realization method of a network knowledge unified representation architecture in an intelligent networking is characterized in that a network knowledge construction layer, a network knowledge representation layer, a network knowledge management layer and a network knowledge service layer are arranged, wherein:
the network knowledge construction layer is used for determining the scope of network knowledge in multi-domain environments such as external environment, network state and user behavior, creating a network knowledge space and supporting network knowledge to drive network elements to complete network autonomous opening, routing addressing, management control and link connection;
the network knowledge representation layer is used for systematically identifying the structure, mode and content of the network knowledge, realizing formal description of the network knowledge and supporting a network intelligent agent to realize interaction of semantic levels;
the network knowledge management layer is used for finishing network knowledge storage, alignment, disambiguation, error correction and updating under the condition of network communication resource limitation, and ensuring the consistency, accuracy and integrity of network knowledge;
and the network knowledge service layer provides the capabilities of knowledge calculation and semantic search, realizes network knowledge reasoning, network knowledge sharing distribution and on-demand acquisition, and drives the network to finish the behaviors of connection and routing addressing of massive intelligent agents, operation and maintenance management and control and endogenous safety.
Compared with the prior art, the invention has the following remarkable advantages:
(1) the abstract model for constructing the network knowledge system is described by the thought summary of the hierarchical structure, so that the structure, the function and the principle of the network knowledge can be known from a deeper level; moreover, the framework has good and clearly defined hierarchical division, can be repeatedly used, supports isomerism and has expandability;
(2) the communication network is facilitated to form a standard formalized language so as to realize the flow and application of network knowledge in the communication network;
(3) the network knowledge system covers three domains of external environment, network state and user behavior, and is beneficial to enhancing the intelligent connection efficiency of the network.
Drawings
FIG. 1 is a schematic diagram of a network knowledge unified characterization architecture in an intelligent networking.
Fig. 2 is a schematic diagram of interface definition of the main functional modules.
Detailed Description
The invention relates to a network knowledge unified representation architecture in an intelligent networking, which comprises a network knowledge hierarchical classification module, a network knowledge space construction module, a local network knowledge representation module, a cooperative network knowledge representation module, a network knowledge storage module, a network knowledge fusion module, a network knowledge verification module and a network knowledge reasoning module, wherein:
the network knowledge classification module provides the category information of the network knowledge for the local network knowledge representation module;
the network knowledge space construction module provides network knowledge such as facts, concepts and rules obtained from network cognition and expert cognition for the local network knowledge characterization module, and specifically comprises signal receiving intensity Doppler expansion, network topology, network scale, node density, user geographic position, user movement rate, an anti-interference strategy and a spectrum management strategy;
the collaborative network knowledge representation module provides unified description information of the collaborative network knowledge for the local network knowledge representation module, wherein the unified description information comprises network knowledge concepts, objects, attributes and the names of models;
the local network knowledge representation module provides the network knowledge after representation for the network knowledge storage module, and the network knowledge storage adopts a Mysql database, a file and a graph database to perform centralized or distributed storage of the network knowledge according to actual conditions;
the network knowledge fusion module provides the aligned and disambiguated network knowledge for the network knowledge storage module;
the network knowledge verification module provides the network knowledge after error correction and updating for the network knowledge storage module;
the network knowledge storage module provides a knowledge source for the network knowledge inference module;
and the network knowledge reasoning module provides the implicit network knowledge generated by prediction to the network knowledge storage module.
Further, the network knowledge hierarchical classification module provides the local network knowledge characterization module with category information of network knowledge, including:
external environment domain, network state domain or user behavior domain network knowledge; dynamic network knowledge or static network knowledge; declarative, procedural, or decision-type network knowledge; public and private network knowledge.
A method for realizing a network knowledge unified representation architecture in an intelligent network is provided with a network knowledge construction layer, a network knowledge representation layer, a network knowledge management layer and a network knowledge service layer, wherein:
the network knowledge construction layer is used for determining the scope of network knowledge in multi-domain environments such as external environment, network state and user behavior, creating a network knowledge space and supporting network knowledge to drive network elements to complete network autonomous opening, routing addressing, management control and link connection;
the network knowledge representation layer is used for systematically identifying the structure, the mode and the content of network knowledge, realizing formal description of the network knowledge and supporting a network intelligent agent to realize interaction of semantic levels;
the network knowledge management layer is used for finishing network knowledge storage, alignment, disambiguation, error correction and updating under the condition of network communication resource limitation, and ensuring the consistency, accuracy and integrity of network knowledge;
and the network knowledge service layer provides the capabilities of knowledge calculation and semantic search, realizes network knowledge reasoning, network knowledge sharing distribution and on-demand acquisition, and drives the network to finish the behaviors of connection and routing addressing of massive intelligent agents, operation and maintenance management and control and endogenous safety.
Further, there are three sources of the network knowledge: the method is derived from a network cognition process, community interaction of an intelligent agent and traditional expert knowledge.
Further, the network knowledge construction layer is provided with a network knowledge hierarchical classification module and a network knowledge space construction module, wherein:
the network knowledge classification module is used for completing the division of the network knowledge from different angles of multi-domain environment, the state of the network knowledge, the function of a network intelligent agent, cognitive cycle and the public and private properties of the knowledge so as to support the completion of network tasks;
and the network knowledge space construction module is used for completing construction of the network knowledge space based on the network tasks and the knowledge granularity.
Further, the network knowledge classification module completes the division of the network knowledge from different angles to support the completion of the network task, and specifically comprises the following steps:
from the perspective of multi-domain environment, the network knowledge is divided into external environment domain knowledge, network state domain knowledge and user behavior domain knowledge; the signal receiving strength, the time delay expansion and the Doppler expansion of the electromagnetic environment signal characteristics belong to the knowledge of an external environment domain; knowledge that network topology, network scale, node density and network traffic belong to a network state domain; the user geographical position, the user moving rate, the user service type and the user service quality requirement belong to the knowledge of the user behavior domain;
from the perspective of the state of the network knowledge, the network knowledge is divided into dynamic network knowledge and static network knowledge;
dividing the network knowledge of the user terminal, the network knowledge of the traditional network element, the network knowledge of the management control agent and the network knowledge of the intelligent composite equipment according to the function angle of the network agent;
from the perspective of cognitive cycle, the method comprises the following steps of dividing the cognitive cycle into statement type knowledge, process type knowledge and decision type knowledge;
from the perspective of public and private attributes of knowledge, the network knowledge is divided into public network knowledge and private network knowledge.
Further, the network knowledge space construction module completes construction of a network knowledge space based on network tasks and knowledge granularity, wherein:
the network knowledge construction based on the network tasks is divided into network knowledge construction oriented to network autonomous opening and network knowledge construction oriented to tasks of anti-interference and topology reconstruction in the network operation process;
the network knowledge construction based on the knowledge granularity is to construct a network knowledge space by utilizing a set theory and a granularity calculation theory according to a network knowledge object and the attribute of the network knowledge object.
Further, the network knowledge representation layer comprises a local network knowledge representation module, a collaborative network knowledge representation module and a representation form conversion module capable of relating network knowledge, wherein:
the local network knowledge characterization module is used for characterizing the network knowledge in various intelligent agents by using corresponding characterization methods aiming at the network knowledge classified in different grades; the method comprises the following steps: describing the concepts, attributes and basic network knowledge elements of the network knowledge by using a representation method of the biased ontologies such as frame representation, connection meaning representation and object-oriented representation; describing high-level network knowledge processes such as tasks, strategies and behaviors by using a representation method of bias relations such as a semantic network and a conceptual diagram;
the collaborative network knowledge representation module is used for realizing the uniform description of the network knowledge objects and concepts of all the intelligent agents through a set mechanism and a set method aiming at the network knowledge of the interaction of the intelligent agents; the adopted mechanism is that one or a group of intelligent agents in the network are utilized, and the nominal descriptions of network knowledge objects and concepts needing interaction are unified in a broadcasting and education mode;
and the characterization form conversion module capable of associating the network knowledge completes the unification of the characterization forms of the network knowledge which are associated with each other, and realizes the joint network knowledge calculation.
Further, the network knowledge management layer comprises a network knowledge storage module, a network knowledge fusion module and a network knowledge verification module, wherein:
the network knowledge storage module is used for storing various network knowledge in different network environments;
the network knowledge fusion module is used for aligning and disambiguating network knowledge and fusing scattered, redundant and heterogeneous network knowledge;
the network knowledge verification module is used for completing the error correction and updating of the dynamic network knowledge, the network knowledge of different sources and different forms; network knowledge error correction includes error detection of network knowledge; the network knowledge updating comprises periodic updating and active updating of the network knowledge.
Further, the network knowledge service layer comprises a network knowledge reasoning module, a network knowledge distribution module, a network knowledge learning module and a network knowledge on-demand acquisition module, wherein:
the network knowledge reasoning module is used for predicting the network knowledge type and relationship and generating more implicit network knowledge;
the network knowledge on-demand acquisition module monitors knowledge requests of other network agents, sends required knowledge to the monitored network agents and establishes a standardized and uniform on-demand service model;
the network knowledge distribution module is responsible for sharing and distributing service of network knowledge, and comprises a strategy knowledge sending module which sends strategy knowledge to a designated intelligent agent and is used for guiding the cognitive decision of the intelligent agent;
and the network knowledge learning module is used for providing network knowledge learning service, learning by using current and historical knowledge in a knowledge base according to the requirements of users or network management, wherein the learning comprises network flow learning, network jitter learning, network delay learning and network packet loss rate learning, and establishing a prediction and optimization mechanism according to the network performance and the change trend of a multi-domain environment.
Examples
The technical solution of the present invention is further described in detail with reference to the accompanying drawings and specific embodiments, and before the detailed description, the concept of network knowledge is described:
the network knowledge is a behavior main body, and a plurality of facts, concepts, rules and the like obtained by sensing, learning and analyzing multi-domain characteristic information such as external environment, user behavior, network state and the like are uniformly characterized as formatted 'appointments' which can be mutually understood by network intelligent bodies, and the transmission of the formatted 'appointments' between the network intelligent bodies represents the interaction of the network knowledge. The form of the network knowledge is a mixture of flowing and dynamic states, and is continuously changed and updated along with stimulation and learning; the network intelligent agent obtains the network knowledge through observation, judgment and iterative learning, also can obtain the network knowledge of other intelligent agents through interaction, and meanwhile, can obtain the network knowledge from experts.
Firstly, fig. 1 shows a system architecture implementation process of network knowledge unified representation in the intelligent networking, a network knowledge construction layer, a network knowledge representation layer, a network knowledge management layer and a network knowledge service layer are arranged, and all the layers are expanded.
(1) Network knowledge construction
The network knowledge construction mainly faces to multi-domain environments such as external environments, network states, user behaviors and the like, determines the scope of network knowledge, creates a network knowledge space, and supports network knowledge to drive network elements to complete network autonomous opening, routing addressing, management control, link connection and the like. There are three main sources of network knowledge: the method is derived from a network cognition process, community interaction of an intelligent agent and traditional expert knowledge. The layer mainly comprises functional modules of network knowledge hierarchical classification, network knowledge space construction and the like.
The network knowledge classification module is mainly responsible for completing the division of network knowledge from different angles so as to better support the completion of network tasks. From the perspective of multi-domain environment, the network knowledge can be divided into external environment domain knowledge, network state domain knowledge and user behavior domain knowledge; from the state perspective of network knowledge, the method can be divided into: dynamic network knowledge and static network knowledge; according to the angle of the function of the network agent, the method can be divided into the following steps: network knowledge of a user terminal, network knowledge of a traditional network element, network knowledge of a management control agent, and network knowledge of an agent with two or more functions; from the perspective of cognitive cycle, the method can be divided into statement type knowledge, process type knowledge and decision type knowledge; from the perspective of public and private attributes of knowledge, the knowledge can be divided into public network knowledge and private network knowledge. For example, the signal reception intensity network knowledge space construction module mainly completes construction of a network knowledge space based on network tasks, knowledge granularity and the like. The network knowledge construction based on the network task can be mainly divided into network knowledge construction oriented to network autonomous opening and network knowledge construction oriented to tasks such as anti-interference and topology reconstruction in the network operation process; the network knowledge construction based on the knowledge granularity is mainly to construct a network knowledge space by utilizing a set theory and a granularity calculation theory according to a network knowledge object and various attributes of the network knowledge object.
(2) Network knowledge characterization
The network knowledge representation is mainly characterized in that network knowledge such as facts, concepts and rules obtained by network cognition is effectively simplified under the constraint of network communication resources, so that systematic identification of the structure, the mode and the content of the network knowledge is realized, and interaction of network intelligent agent semantic levels is supported. The layer mainly comprises: the system comprises a local network knowledge characterization module, a collaborative network knowledge characterization module, a characterization form conversion module capable of associating network knowledge and the like.
The local network knowledge representation mainly aims at the network knowledge of different classification, and the effective representation method is utilized to represent the network knowledge in various intelligent agents. For example, basic network knowledge elements such as concepts, attributes and objects of network knowledge are described by using a representation method of a biased ontology such as a framework representation, a connection meaning representation and an object-oriented representation; and describing a high-level network knowledge process such as tasks, strategies and behaviors by using a representation method of a bias relationship such as a semantic network and a conceptual diagram.
The cooperative network knowledge representation is mainly used for realizing the uniform description of network knowledge objects, concepts and other designations of all intelligent agents through a certain mechanism and method aiming at the network knowledge of intelligent agent interaction. The adopted mechanism is to unify the named descriptions of the network knowledge objects, concepts and the like needing interaction by using one or a group of intelligent agents in the network in a broadcasting, education and other modes.
The characteristic form conversion of the associable network knowledge is mainly responsible for finishing the unification of the mutually associated network knowledge characteristic forms so as to facilitate the joint network knowledge calculation.
(3) Network knowledge management
The network knowledge management is mainly responsible for completing network knowledge storage, fusion, verification and the like facing to complex and changeable network environments and network resource constraints, and the consistency, accuracy and integrity of the network knowledge are ensured. The layer mainly comprises: the system comprises a network knowledge storage module, a network knowledge fusion module, a network knowledge verification module and the like.
The network knowledge storage mainly aims at survivability, safety and the like, and various network knowledge can be stored in a complex and changeable network environment.
The network knowledge fusion is mainly responsible for alignment, disambiguation and the like of network knowledge, and the network knowledge with dispersion, redundancy and isomerism is fused.
The network knowledge verification is mainly responsible for completing the error correction, updating and the like of dynamic network knowledge, network knowledge of different sources and different forms; the network knowledge error correction mainly completes the error detection of the network knowledge and the like; the network knowledge updating mainly completes the periodic updating, the active updating and the like of the network knowledge.
(4) Network knowledge service
The network knowledge service is responsible for providing capabilities of knowledge calculation, semantic search and the like, realizing network knowledge reasoning, efficient network knowledge distribution, on-demand acquisition and the like, and driving a network to complete the connection and routing addressing of massive intelligent agents, management control, safety and other behaviors. The method mainly comprises network knowledge reasoning, network knowledge distribution, network knowledge learning, network knowledge acquisition according to requirements and the like.
The network knowledge reasoning mainly completes the prediction of the category, the relationship and the like of the network knowledge, and generates more implicit network knowledge.
Network knowledge acquires knowledge requests mainly responsible for monitoring other network agents as required, sends required knowledge to the network agents, and establishes a standardized and unified on-demand service model;
the network knowledge distribution is mainly responsible for the efficient sharing and distribution service of the network knowledge, for example, the strategy knowledge can be sent to a designated intelligent agent for guiding the cognitive decision of the intelligent agent;
the network knowledge learning mainly provides network knowledge learning service, and can utilize current knowledge and historical knowledge in a knowledge base to learn according to the needs of users or network management, such as network traffic learning, network jitter learning, network delay learning, network packet loss rate learning and the like, and a prediction and optimization mechanism is established according to the network performance and the change trend of a multi-domain environment.
Secondly, fig. 2 shows the interface definition of the main functional module in the framework aiming at the network knowledge unified representation system framework in the intelligent networking, and the functional modules are expanded below.
The main functional modules for realizing the network knowledge system architecture comprise: the system comprises a network knowledge classification module, a network knowledge space construction module, a local network knowledge representation module, a cooperative network knowledge representation module, a network knowledge storage module, a network knowledge fusion module, a network knowledge verification module, a network knowledge reasoning module and the like.
(1) Network knowledge hierarchical classification-local network knowledge characterization interface:
the hierarchical classification of network knowledge provides the local network knowledge representation with category information of network knowledge, such as: external environment domain, network state domain or user behavior domain network knowledge; dynamic network knowledge or static network knowledge; declarative, procedural, or decision-making network knowledge; public, private network knowledge, etc.
(2) Network knowledge space construction-local network knowledge characterization interface:
the network knowledge space is constructed to provide network knowledge such as signal receiving strength Doppler expansion, network topology, network scale, node density, user geographic position, user movement rate, anti-interference strategy, spectrum management strategy and the like for local network knowledge representation, wherein the network knowledge represents facts, concepts, rules and the like obtained from network cognition and expert cognition.
(3) Collaborative network knowledge characterization-local network knowledge characterization interface:
the collaborative network knowledge representation provides the local network knowledge representation with descriptive information of the object, concept, etc. that need interactive network knowledge.
(4) Local network knowledge characterization-network knowledge storage interface:
the local network knowledge representation provides network knowledge after representation for network knowledge storage, the network knowledge storage adopts a Mysql database, a file, a graph database and the like, and centralized or distributed storage of the network knowledge is carried out according to actual conditions.
(5) Network knowledge fusion-network knowledge storage interface:
network knowledge fusion provides aligned and disambiguated network knowledge for network knowledge storage.
(6) Network knowledge verification-network knowledge storage interface:
the network knowledge verification provides the network knowledge after error correction and updating for the network knowledge storage.
(7) Network knowledge storage-network knowledge inference interface:
the network knowledge storage provides a knowledge source for network knowledge reasoning; the network knowledge inference provides the implicit network knowledge generated by prediction to a network knowledge storage;
the invention can support network elements to complete functions of routing addressing, management control, safety and the like by using network knowledge, realize reasonable deployment of network resources, provide theoretical guidance for a communication network to form standard formal languages and realize flow of the network knowledge in the network.
The above description is only for the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention.
Claims (10)
1. The utility model provides a unified characterization system of network knowledge in intelligent networking which characterized in that, includes hierarchical classification module of network knowledge, network knowledge space construction module, local network knowledge characterization module, collaborative network knowledge characterization module, network knowledge storage module, network knowledge fusion module, network knowledge verification module, network knowledge reasoning module, wherein:
the network knowledge classification module provides the category information of the network knowledge for the local network knowledge representation module;
the network knowledge space construction module provides network knowledge such as facts, concepts and rules obtained from network cognition and expert cognition for the local network knowledge characterization module, and specifically comprises signal receiving intensity Doppler expansion, network topology, network scale, node density, user geographic position, user movement rate, an anti-interference strategy and a spectrum management strategy;
the collaborative network knowledge representation module provides unified description information of the collaborative network knowledge for the local network knowledge representation module, wherein the unified description information comprises network knowledge concepts, objects, attributes and the names of models;
the local network knowledge representation module provides the network knowledge after representation for the network knowledge storage module, and the network knowledge storage adopts a Mysql database, a file and a graph database to perform centralized or distributed storage of the network knowledge according to actual conditions;
the network knowledge fusion module provides the aligned and disambiguated network knowledge for the network knowledge storage module;
the network knowledge verification module provides the network knowledge after error correction and updating for the network knowledge storage module;
the network knowledge storage module provides a knowledge source for the network knowledge reasoning module;
and the network knowledge reasoning module provides the implicit network knowledge generated by prediction to the network knowledge storage module.
2. The system for uniformly characterizing network knowledge in an intelligent network according to claim 1, wherein the network knowledge hierarchical classification module provides the local network knowledge characterization module with category information of network knowledge, comprising:
external environment domain, network state domain or user behavior domain network knowledge; dynamic network knowledge or static network knowledge; declarative, procedural, or decision-type network knowledge; public, private network knowledge.
3. A realization method of a network knowledge uniform representation system in an intelligent networking is characterized in that a network knowledge construction layer, a network knowledge representation layer, a network knowledge management layer and a network knowledge service layer are arranged, wherein:
the network knowledge construction layer is used for determining the scope of network knowledge in multi-domain environments such as external environment, network state and user behavior, creating a network knowledge space and supporting network knowledge to drive network elements to complete network autonomous opening, routing addressing, management control and link connection;
the network knowledge representation layer is used for systematically identifying the structure, mode and content of the network knowledge, realizing formal description of the network knowledge and supporting a network intelligent agent to realize interaction of semantic levels;
the network knowledge management layer is used for finishing network knowledge storage, alignment, disambiguation, error correction and updating under the condition of network communication resource limitation, and ensuring the consistency, accuracy and integrity of network knowledge;
and the network knowledge service layer provides the capabilities of knowledge calculation and semantic search, realizes network knowledge reasoning, network knowledge sharing distribution and acquisition as required, and drives the network to complete the behaviors of connection and routing addressing of massive intelligent agents, operation and maintenance management and control and endogenous safety.
4. The method for implementing the system for uniformly characterizing network knowledge in the intelligent network according to claim 3, wherein the network knowledge has three sources: the method is derived from a network cognition process, community interaction of an intelligent agent and traditional expert knowledge.
5. The method for implementing the network knowledge unified characterization system in the intelligent network according to claim 3 or 4, wherein the network knowledge building layer is provided with a network knowledge hierarchical classification module and a network knowledge space building module, wherein:
the network knowledge classification module is used for completing the division of the network knowledge from different angles of multi-domain environment, the state of the network knowledge, the function of a network intelligent agent, cognitive cycle and the public and private properties of the knowledge so as to support the completion of network tasks;
and the network knowledge space construction module is used for completing construction of the network knowledge space based on the network tasks and the knowledge granularity.
6. The method for implementing a unified characterization system for network knowledge in an intelligent network according to claim 5, wherein the hierarchical classification module for network knowledge is used for partitioning network knowledge from different angles to support the completion of network tasks, and specifically comprises the following steps:
from the perspective of multi-domain environment, the network knowledge is divided into external environment domain knowledge, network state domain knowledge and user behavior domain knowledge; the signal receiving strength, the time delay expansion and the Doppler expansion of the electromagnetic environment belong to the knowledge of an external environment domain; knowledge that network topology, network scale, node density and network traffic belong to a network state domain; the user geographical position, the user moving rate, the user service type and the user service quality requirement belong to the knowledge of the user behavior domain;
from the perspective of the state of the network knowledge, the network knowledge is divided into dynamic network knowledge and static network knowledge;
dividing the network knowledge of the user terminal, the network knowledge of the traditional network element, the network knowledge of the management control agent and the network knowledge of the intelligent composite equipment according to the function angle of the network agent;
from the perspective of cognitive cycle, the method comprises the following steps of dividing the cognitive cycle into statement type knowledge, process type knowledge and decision type knowledge;
from the perspective of public and private attributes of knowledge, the network knowledge is divided into public network knowledge and private network knowledge.
7. The method for implementing the unified characterization system for network knowledge in the intelligent network according to claim 5, wherein the network knowledge space construction module completes construction of the network knowledge space based on network tasks and knowledge granularity, wherein:
the network knowledge construction based on the network tasks is divided into network knowledge construction oriented to network autonomous opening and network knowledge construction oriented to tasks of anti-interference and topology reconstruction in the network operation process;
the network knowledge construction based on the knowledge granularity is to construct a network knowledge space by utilizing a set theory and a granularity calculation theory according to a network knowledge object and the attribute of the network knowledge object.
8. The method for implementing the system for uniformly characterizing network knowledge in the intelligent network according to claim 6 or 7, wherein the network knowledge characterization layer comprises a local network knowledge characterization module, a collaborative network knowledge characterization module, and a characterization form conversion module capable of associating network knowledge, wherein:
the local network knowledge characterization module is used for characterizing the network knowledge in various intelligent agents by using corresponding characterization methods aiming at the network knowledge classified in different grades; the method comprises the following steps: describing the concepts, attributes and basic network knowledge elements of the network knowledge by using a representation method of the biased ontologies such as frame representation, connection meaning representation and object-oriented representation; describing high-level network knowledge processes such as tasks, strategies and behaviors by using a representation method of bias relations such as a semantic network and a concept graph;
the collaborative network knowledge representation module is used for realizing the uniform description of network knowledge objects and concepts of all the intelligent agents by a set mechanism and a set method aiming at the network knowledge interacted by the intelligent agents; the adopted mechanism is that one or a group of intelligent agents in the network are utilized, and the finger descriptions of network knowledge objects and concepts needing interaction are unified in a broadcasting and education mode;
and the characterization form conversion module capable of associating the network knowledge completes the unification of the characterization forms of the network knowledge which are associated with each other, and realizes the joint network knowledge calculation.
9. The method for implementing the system for uniformly characterizing network knowledge in the intelligent network according to claim 6 or 7, wherein the network knowledge management layer comprises a network knowledge storage module, a network knowledge fusion module and a network knowledge verification module, wherein:
the network knowledge storage module is used for storing various network knowledge in different network environments;
the network knowledge fusion module is used for aligning and disambiguating network knowledge and fusing scattered, redundant and heterogeneous network knowledge;
the network knowledge verification module is used for completing the error correction and updating of the dynamic network knowledge, the network knowledge of different sources and different forms; network knowledge error correction includes error detection of network knowledge; the network knowledge updating comprises periodic updating and active updating of the network knowledge.
10. The method for implementing the unified characterization system of network knowledge in the intelligent network according to claim 6 or 7, wherein the network knowledge service layer comprises a network knowledge inference module, a network knowledge distribution module, a network knowledge learning module, and a network knowledge on-demand acquisition module, wherein:
the network knowledge reasoning module is used for predicting the network knowledge type and relationship and generating more implicit network knowledge;
the network knowledge on-demand acquisition module monitors knowledge requests of other network agents, sends required knowledge to the monitored network agents, and establishes a standardized and uniform on-demand service model;
the network knowledge distribution module is responsible for sharing and distributing service of network knowledge, and comprises a strategy knowledge sending module which sends strategy knowledge to a designated intelligent agent and is used for guiding the cognitive decision of the intelligent agent;
and the network knowledge learning module is used for providing network knowledge learning service, learning by using current and historical knowledge in a knowledge base according to the requirements of users or network management, wherein the learning comprises network flow learning, network jitter learning, network delay learning and network packet loss rate learning, and establishing a prediction and optimization mechanism according to the network performance and the change trend of a multi-domain environment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110083207.XA CN113162968B (en) | 2021-01-21 | 2021-01-21 | Unified characterization framework of network knowledge in intelligent networking and implementation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110083207.XA CN113162968B (en) | 2021-01-21 | 2021-01-21 | Unified characterization framework of network knowledge in intelligent networking and implementation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113162968A CN113162968A (en) | 2021-07-23 |
| CN113162968B true CN113162968B (en) | 2022-09-13 |
Family
ID=76878702
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110083207.XA Active CN113162968B (en) | 2021-01-21 | 2021-01-21 | Unified characterization framework of network knowledge in intelligent networking and implementation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113162968B (en) |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100505970B1 (en) * | 2003-11-28 | 2005-08-30 | 한국전자통신연구원 | System and method for knowledge modeling using ontology |
| CN110795806B (en) * | 2019-10-16 | 2022-08-30 | 南京邮电大学 | Intelligent networking system architecture |
-
2021
- 2021-01-21 CN CN202110083207.XA patent/CN113162968B/en active Active
Non-Patent Citations (1)
| Title |
|---|
| 基于知识融合过程的大数据知识服务框架研究;周利琴等;《图书馆学研究》;20171115(第21期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113162968A (en) | 2021-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | Wireless network design for emerging IIoT applications: Reference framework and use cases | |
| CN113316118B (en) | Unmanned aerial vehicle cluster network self-organizing system and method based on task cognition | |
| CN106502266B (en) | Multi-machine collaborative operation command control system | |
| Li et al. | Zoning for hierarchical network optimization in software defined networks | |
| Zhang et al. | AI-enabled space-air-ground integrated networks: Management and optimization | |
| Han et al. | Time-varying topology model for dynamic routing in LEO satellite constellation networks | |
| CN114745286B (en) | Intelligent network situation awareness system oriented to dynamic network based on knowledge graph technology | |
| Zhou et al. | Intelligence-endogenous networks: Innovative network paradigm for 6G | |
| Hu et al. | Intelligent decision making framework for 6G network | |
| CN105354634A (en) | Active power distribution network antenna integration apparatus | |
| Zhao et al. | Socialized learning for smart cities: Cognitive paradigm, methodology, and solution | |
| CN113162968B (en) | Unified characterization framework of network knowledge in intelligent networking and implementation method | |
| Duan et al. | Cognitive intelligence based 6G distributed network architecture | |
| Xia et al. | Digital Twin and Artificial Intelligence for Intelligent Planning and Energy-Efficient Deployment of 6G Networks in Smart Factories | |
| Bariah et al. | Digital twin-empowered smart cities: A new frontier of wireless networks | |
| Bariah et al. | Digital twin-empowered communications: A new frontier of wireless networks | |
| Coelho et al. | Semantic behavior modeling and event-driven reasoning for urban system of systems | |
| Tan et al. | Uav control in smart city based on space-air-ground integrated network | |
| Harvey et al. | Evolutionary autonomous networks | |
| Huang et al. | A brief survey and implementation on ai for intent-driven network | |
| Dridi et al. | Reinforcement Learning Vs ILP Optimization in IoT support of Drone assisted Cellular Networks | |
| Nagasundaram et al. | Analysis of the requirement and artificial intelligence-based resource management system in cloud | |
| Lin et al. | Hypergraph-Based Autonomous Networks: Adaptive Resource Management and Dynamic Resource Scheduling | |
| Srinivas et al. | Functional overview of integration of AIML with 5G and beyond the network | |
| CN102075555A (en) | Spatial information processing service automatic composition method with dynamics |
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 |