The Knowledge Representation And Intelligent Fusion On Cognitive Network QoS Situation

With the ever-changing network development, data transmission in the network presents anexplosive growth, the difficulty of obtaining knowledge in the network is further increased, andnetwork transmission quality of service (QoS) can not be guaranteed too. As the research focus ofthe next-generation network, cognitive networks is able to obtain network information throughsituation awareness and make real-time configuration adjustment according to the network situation,then provide a guarantee for network QoS transmission.Currently, majority of the research concentrate their efforts on the area of assessment incognitive network situation, only a few researches for QoS situation fusion. Modeling and storageon domain information are needed before situation fusion, but the current study about domainontology construction of cognitive network QoS situation is rare. Based on the analysis of existingknowledge representation and data fusion mechanism, the article studies the knowledgerepresentation and QoS situation fusion of the cognitive network.Firstly, from the perspective of knowledge representation, the article extracts the specificconcepts about cognitive network QoS, analyzes the relationship between them, and defines therelevant properties, a cognitive network QoS situation ontology model is proposed, then. Based onthis model, protégé Software is used for the ontology model constructing. Then the form of triples isintroduced to record the real-time data during the network running, and the data is stored in theontology storage combined with ontology model. At the same time, the query process aboutontology storage is designed, and the query results on various concepts of time T1is given.Ontology construction of cognitive network QoS situation provides a theoretical basis for thesubsequent fusion.Secondly, from the perspective of local fusion, the article proposes a fusion algorithm aboutthe cognitive network local QoS situation based on fuzzy reasoning. First, the cognitive networkQoS situation level is provided, and then the membership functions of the link average QoSparameters are defined. Secondly, the fuzzy inference rules are given, and a variety of QoSparameters are fused. The simulation results show that the algorithm is able to determine the QoSsituation level of various service transmissions, and could achieve better fusion effect.Finally, from the perspective of global fusion, the article proposes a fusion algorithm about thecognitive network global QoS situation based on D-S evidence theory. On the basis of original D-S evidence theory model, the article derives a fusion formula which can resolve conflict evidenceinstances. The voting mechanism is adopted to solve the initial probability of the D-S evidencetheory formula. Then the improved D-S formula is used to fuse multiple local network QoSsituation, and the global network QoS situation level is received. The simulation shows that thefusion results are more objective and reasonable when conflict between evidences is large, and itcan also provide some theoretical basis for network global QoS situation analysis.