Research On Model And Method Of Establishing Network Interaction Patterns Based On Trust

With the rapid development of computer network technology, kinds of new technologies and new applications spring up one after another. For examples, Web2.0 focuses on interaction with users and in Web2.0 interaction patterns have changed from only read to write and establish together. In self-organizing networks, the role of server and the role of client have mixed together and transformed each other dynamically. Network interaction has been a core problem of computer network technology.At present, network interaction patterns (NIPs) can be classified by time sequence of interaction events, flowing method of interaction information, and with or without agents. Among these classification methods, the first and the second classification methods do not consider trust between interaction entities. Although the last classification method avoids this disadvantage, it does not consider interaction patterns from the perspective of the overall.The method of establishing NIPs has a close relationship with NIPs. Now, there are three main methods to establish trustworthy NIPs, including choosing direct interaction entity based on trust, introducing new middle-organizations, and automatically checking and controlling interaction patterns. These methods are all not systemative and have not consider the static and automatical relationship among trust on entities, trust of direct interaction events, trust of basic interaction patterns and trust of NIPs. Therefore, these methods can only solve trustworthiness of local interaction patterns and do not take the relevance between interaction time complexity and interaction trustworthiness into consideration.In the new network interactions, interaction roles are changeable and interaction patterns are readability and writability. Therefore, trustworthiness of interaction and interaction pattern stands out. So to create an efficient trustworthy NIP establishment scheme based on interaction pattern models and their crucial characteristics evaluation models is an urgent problem to be solved for the application of trust management and the design of trustworthy interaction patters.By analyzing the research status of NIP modeling and NIP establishment methods, and considering the characteristics of trust management in interaction network and the factors of modeling NIP establishment, we proposed an interaction network model combined with designability, extendibility and evaluation on NIP trust and time complexity. Then we proposed a NIP establishing scheme based on network interaction models and their crucial characteristics evaluation model by using several methods, such as the fuzzy Petri network, inhibitory arc Petri network and fuzzy time Petri network. We also demonstrated the performance of NIP crucial characteristics evaluation model and NIP establishment scheme through theories reason and simulation.The primary contributions of this paper are shown as follows.(1) To propose seven kinds of trust-based basic interaction patterns (BIPs) and basic indirect interaction patterns (BIIPs) and their fuzzy production rules and Petri net models, which lays a foundation for modeling interaction network.(2) To propose a new interaction network model based on BIPs and BIIPs, which effectively solves the problems of extendibility, designability and formal specification of interaction network.(3) To propose a NIP crucial characteristic evaluation scheme. On the base of reputation-based entity trust evaluation method and BIP (BIIP) trust evaluation method that gives full consideration to interaction events'logical relationship, the scheme is integrated with NIP trust evaluation method properly based on reputation-based trust and certificate-based trust and NIP time complex evaluation method based on fuzzy time Petri network. The proposed NIP crucial characteristic evaluation scheme lays a foundation for choosing NIPs during NIP establishment.(4) To propose a NIP establishment scheme based on BIPs, BIIPs and their crucial characteristic evaluation. The NIP establishment algorithm can find out all defined BIPs and BIIPs and search all integrated interaction patterns (IIP) and evaluate IIPs'crucial characteristics. By considering the active refreshing demand requested by interaction entity and the passive refreshing demand requested by interaction environment, NIP refreshing algorithm can quickly refresh IIP crucial characteristics and search new IIPs. The NIP establishment scheme ensures the robustness and the practicability of NIP establishment scheme.(5) To propose a relevance model for entity trust relationship and interaction time complex. In the relevance model, interaction time complex is decreased with entity trust relationship improved. And the course of trust improvement is derived into four stages, initial stage, crucial stage, sustained stage and maintaining stage. And correspondingly, the decreasing speed of interaction time complex can be derived into four stages. The relevance model of entity trust and interaction time complexity solves the problem of modeling qualitative relationship between entity trust relationship improving speed and interaction time complex decreasing speed.