Distributed Resource Discovery in UniFrame
                 

Resource Discovery is defined as the process of locating, accessing,
retrieving and managing pertinent resources from distributed, possibly
heterogeneous repositories. The resource discovery is an integral part of the
UniFrame -- a framework that aims at allowing an interoperation of
heterogeneous and geographically scattered software components. The
development of URDS (UniFrame Resource Discovery System) involves many
challenges. These challenges can be grouped into three categories:
a) Publishing and Promotion of components, b) access and discovery of
components and c) infrastructure for hosting components and mapping their
characteristics with the resource queries.

Specifically, the objectives of the URDS research are:

a) To explore different approaches for describing the necessary features of
distributed software components,

b) To create the necessary infrastructure for hosting and locating
heterogeneous software components,

and

c) To provide appropriate formalism for querying, mapping and matching of
software components stored in distributed repositories.

A successful creation of URDS will provide the appropriate software components,
along with the necessary formalism, as inputs to the UniFrame system generator,
which will compose these components into an integrated distributed system.

                Composition of Security-aware Components In UniFrame
                    

UniFrame aims at providing a framework that enables a seamless
interoperation of heterogeneous distributed software components. One possible
source of heterogeneity is based on the security characteristics of individual
components. Thus, while composing a system dynamically from distributed
components, each having its specific security features, predicting the
security features of the final system is crucial. The predication of the
security properties of the integrated system requires a reasoning based on
the specification of individual security features and their possible
interactions. To achieve this goal, it is necessary to examine how components
are characterized by various security properties such as access control or
encryption and how such properties are advertised by components in order to
search for components meeting certain requirements. The specific objectives of
this research are:

a) To characterize different security properties as applicable to the
development of software components,

b) To identify and classify security features provided by different component
models,

c) To investigate alternatives for specifying the security properties of
software components,

and

d) To propose a compositional model that will enable the prediction of
security features of the integrated system and facilitate an interoperation of
heterogeneous components.

The end goal of this research is to provide the system integrator with some
level of assurance that the system being generated has known and predictable
security characteristics.