Universal Access Control Fuzzy Inference System

Ubiquitous computing, proposed by M. Weiser in 1991, is recognized as the computing pattern in future. Ubiquitous computing has become the hotspot in the field of computer science, but most of the researchers of the world focus on its infrastructure. As the research and application of ubiquitous computing develop more deeply, the security in ubiquitous computing is becoming more and more important.Ubiquitous computing is a kind of complicated system, in which lots of devices are connected to each other dynamically, and it should provide anytime and anywhere services, so the access control is thought highly fuzzy. Traditional computing systems mainly enforce centralized access control, and the controlled objects and their relationship could be modeled by precise mathematic model. In ubiquitous computing environment, the information the access control depends, such as . time, location and their relationship, are so complicated. So it is difficult to set up a precise mathematic model for it and if complicated systems were dealt with traditional precise methods, it would be easy to get imprecise result. So traditional access control models are not suitable for ubiquitous computing.Not only according to the requirement of access control in ubiquitous computing but also to the developing trend of access control, fuzziness will be needed. Firstly, as ubiquitous computing is a complicated system, it is fuzzy in its access control which could not be described with precise mathematic model. Secondly, it is well known that all the access control policies are made by system administrators or users and all of them are recorded by a certain language. Because people's thinking and language are fuzzy, so it is clearly that the policies they make are fuzzy.The main research achievements are outlined as follows:(1)Based on the research on the access control and the latest research on the security in ubiquitous computing, the fuzzy access control policies in ubiquitous computing is described and a role-based access control fuzzy model for ubiquitous computing is proposed in the thesis. This model is constructed by making RBAC model support spatio-temporal and fuzzy properties and it could reflect access control policies in ubiquitous computing.(2)A fuzzy inference system is proposed for the access control model based on the interval value fuzzy predicate logic. The fuzzy inference system makes use of reverse inference method and forward inference method. Its inference procedure is as follows. When then access request get into the fuzzy inference system, the reverse inference component will get the inference tree based on the request and the rule base. After that, the forward inference component will get the inference result according to the interval value fuzzy inference method based on all the inference paths. At last, the conflict resolution component will choose a result or calculate a result based on all inferred result.(3)After the theoretical model is constructed, the fuzzy inference system for ubiquitous computing access control is developed. It is designed on the .NET framework platform and B/S pattern is chosen. During the developing procedure, we applied modularization method. The system is divided into several logical part, all of which are coded and compiled separately and integrated together finally. All modules are extensible, which could be changed if needed.The fuzzy model for ubiquitous computing and its inference system could support access control for ubiquitous computing efficiently according our mock tests.