A Context-Aware Based Authorization System For Pervasive Grid Computing

Loading...
Thumbnail Image
Date
2015-02
Authors
CHIEN HUI, MARILYN LIM
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The purpose of this study was to address the limitation of pervasive grid particularly on the area of authorization. The advance in wireless technologies had accelerated the evolution from grid technologies to pervasive grid. This brings challenges to the authorization mechanism, as the conventional authorization does not support the environment context during the authorizing process. The aim of this research is to enhance the authorization framework for better trustworthiness in order to work in pervasive grid. From the review, the existing authorization mechanisms have limitations of only supporting static attributes (which value is unchanged throughout the authorization session), and also lack of mechanism to re-verify and confirming on user’s environment information. This resulted in decrease of grid security level, such as the degrees of the grid system resistance to unauthorized user whom purposely misuse the grid infrastructure. One of the solutions to address this limitation is introducing context-awareness into the authorization mechanism. In this thesis, a context-awareness authorization framework was proposed to improve the security level of pervasive grid infrastructure and services by tightening the access control level with context information as additional authorization criterion. However additional checking will prolong the authorization time, which will also prolong the toleration of unauthorized user during change in environment. Thus the introduction of session access token was also proposed to simplify the process and improve the efficiency of authorization mechanism. Simulation design and implementation platform were discussed to understand the development of simulation model for xiv proposed authorization framework. Various scenarios and experiments were designed and tested with the simulation model to evaluate the performance of proposed framework. The experimental results show that, with additional contextual data, the simulation model was able to produce desired authorization decision result effectively, with the context acquisition time of 0.0014s for small size environment context. The experiment also demonstrated that, implementation of session access token has improved the overall efficiency of authorization process by 90% for test case with re-authorization request. The significance of these results shows that the authorization mechanism was able to disconnect the unauthorized user from the service in timely manner to protect the security of grid. The scalability analysis shows that the framework was able to support maximum size of 423 active users to operate without overhead occurrence. The research findings in this thesis have proved that the proposed solution had overcome the limitation of existing authorization system with improvement in efficiency and scalability. We can conclude that the proposed framework was able to improve the trustworthiness level of authorization in pervasive grid.
Description
Keywords
Context-Aware Based Authorization System , For Pervasive Grid Computing
Citation