Confining Data and Processes in Global Computing Applications
Abstract
A programming notation is introduced that can be used for protecting secrecy and integrity of data in global computing applications. The approach is based on the explicit annotations of data and network nodes. Data are tagged with information about the allowed movements, network nodes are tagged with information about the nodes that can send data and spawn processes to them. The annotations are used to confine movements of data and processes. The approach is illustrated by applying it to three paradigmatic calculi for global computing, namely cKLAIM (a calculus at the basis of KLAIM), Dpi (a distributed version of the pi-calculus) and Mobile Ambients Calculus. For all of these formalisms, it is shown that their semantics guarantees that computations proceed only while respecting confinement constraints. Namely, it is proven that, after successful static type checking, data can reside at and cross only authorised nodes. `Local' formulations of this property where only relevant sub-nets type check are also presented. Finally, the theory is tested by using it to model secure behaviours of a UNIX-like multiuser system.
@Article\{nicola.gorla.pugliese:confining-data-processes-gc-applications,
author = \{R. {De Nicola} and D. Gorla and R. Pugliese},
title = \{Confining Data and Processes in Global Computing Applications},
journal = \{Science of Computer Programming},
year = \{2004},
pages = \{Elsevier},
url = \{http://mikado.di.fc.ul.pt/repository/nicola.gorla.pugliese_confining-data-processes-gc-applications.pdf}
}
About this site. Last modified: Sat Apr 20 16:59:49 CEST 2024