The Study Of Dynamic Awareness Of Component Behavior And Component Clustering Mechanism In Future Network

With the rapid development of computer and communication technology, many disadvantagesof the existing network are gradually revealed, which make it difficult to meet communicationneeds of high speed, high efficiency, high capacity and ubiquity for future network, and alsodifficult to achieve efficient use of network resources. There is an urgent need to explore newnetwork architecture to meet the increasingly diverse needs of network services, and therefore theconcept of future network comes into being.Intelligent network is the research direction of future network. In this thesis, we suppose thatthe future network has the ability to self-awareness, and can adjust the internal configuration ofnetwork to adapt to changes of the environment intelligently, therefore a network model based onthe MDE(Monitor, Decision and Execution) ring has been introduced. Meanwhile, an architecturemodel of collaborative component has also been designed, so that the awareness of componentbehavior and component clustering can be achieved on the network component layer.As for the study of dynamic awareness of component behavior, a model of componentbehavior awareness has been designed. In this model, component behavior can be awared byinformation access, information processing and information prediction. In addition, a fuzzycomponent clustering mechanism has also been designed for component clustering. And in order toverify the validity of this algorithm, matlab has been used to simulate it, resulting in a range ofclustering results which have different accuracy and a dynamic clustering chart.