Research On Access Control Strategy Of Smart Device Network

In recent years,with the development of Internet of Things,more and more smart devices have come into people's lives,and the network of smart devices becomes huge.The development of smart devices not only brings a lot of convenience to people's life,but also gives rise to some new security risks.Access control is an important part of the network security system,it is also an important means to protect the security of the network and assure the device resources,users' privacy and network data to be accessed legally.At present,the research on access control strategy mainly focuses on traditional network or sensor network,which is not suitable for smart device network.Smart devices are different from sensor nodes,which have powerful computing capability.Moreover,smart device network has the characteristics of high heterogeneity,large number of terminals,and complex scenarios,which put forward some new demands for access control strategy.This paper analyzes these requirements and takes into account the shortage of traditional access control strategy,then adopts trust-based access control strategy and makes the following works:(1)After analyzing the demand of trust-based access control strategy in smart device network,this paper improves the classical trust model(EigenTrust model)in P2P network.In the improved model,we take into the factors such as time decay,service reward assessment,and device activity into account,then propose the scheme of direct trust calculation which is based on favorability evaluation and forgetting curve.In addition,in the improved model,we define the social relationships among the devices,use the grey system theory and put forward the recommended trust calculation method based on social relationships and grey relational analysis.Compared with the traditional calculation method,the simulation experiment results show that the two methods have higher sensitivity to the malicious behavior of the device and have lower recommended trust error in the case of different recommended set.(2)The improved EigenTrust model and the trust model in the traditional access control strategy have the following deficiencies:less dimensions of the influence factors,insufficient scalability and inadequate utilization of all kinds of influence factors.Based on four dimensions including of direct trust,recommended trust,security attribute and social relationship,this paper utilizes fuzzy inference system and presents an access control strategy based on fuzzy aggregation and multi-dimension trust model in smart device network.The simulation results show that the new trust model has higher accuracy,better benefits and well adaptability,and it also improves the scalability and influence dimensions.