Summary and Future Work. This paper describes a way of providing a sustainable and secure economy in Pleiades. The solution utilizes trust management, an established secure system, to create contracts for consumers and providers. Using trust management, users are allowed to write, share, and use trust management assertions to issue, buy and consume resources. However, the way the trust management has been combined to create contracts, and the way it is used for trust management checks, reduces the cryptographic overhead associated with the general trust management process. An additional benefit for the consumer, is increased level of quality of service. The system scales well for future development, and also is suitable for the distributed nature of the MA. Details of how the software provided by the MA is created, how contract freshness is maintained, and a set of pre-defined rules on how to correctly write the credentials, i.e., in terms of variable names, is needed. Simulations of the system in order to show, amongst others, tradeoffs between in- stantiating short lived VMs and its coupling with the creation and usage of KeyNote credentials, how well the system scales when being exposed to evolving organisations, and more importantly the effectiveness between different organisations in relation to communication overheads, requires further investigation. Crossing different organisa- tions poses other problems too, especially in terms of user’s data security. Since this could be an inhibitor when sharing resources beetwen different organisations, though potentially not an issue when contained within the organization itself, it warrants the investigation into a separate component, that would address this issue, for the Pleiades system. Revocation has been addressed by the creators of KeyNote in [19]. Revocation may be used, when addressing the subject of freshness of contracts, it relies on the synchro- nization of clocks between principals, which may be difficult to achieve. The work done in [10] show what would appear to be, an inevitable increase in compute time when making trust management checks on fine grained jobs, due to the number of TM checks needed. A major contribution of the system described here is that this performance degradation is not inevitable. In fact, since the solution described here is general, it could be applied to systems other than Pleiades.
Summary and Future Work. This deliverable represents the conceptual foundation of the OPTIC project. In particular, it has sought to provide the theoretical resources necessary to develop our understanding of the complex causal processes involved in contemporary policy-making at the European level. Essentially, this concluding chapter has two complementary purposes; it is intended to provide a concise summary of the preceding chapters and, importantly, it is intended to situate the contributions that these chapters have made in the wider context of the OPTIC project and forthcoming work packages. In Chapter 2, we firstly examined the nature of transport policies and processes of policy analysis in the European Union at both the supra-national and national scales. Attention was directed at a number of areas, including the principal institutional actors involved in agenda- setting and decision-making, together with an overview of some of the key instrumental mechanisms upon which political interventions in the transport system typically depend. Of particular significance was the range of overarching EU policy documents, collectively stating the main problems related to transport in the Union, the goals for the EU transport policy and favoured approaches to achieving these goals. We secondly complemented this with a broad overview of policy measures that have been implemented across a range of EU member states in recent years. This exercise demonstrated that non-intentional effects and barriers to successful implementation are frequent occurrences in such interventions, regardless of measures‘ specific objectives, types and geographical scales. Recognising the complex and heterogeneous nature of the non-intentional effects highlighted in Chapter 2, and those documented elsewhere in the published literature, Chapter 3 sought to develop a systematic typology of such effects. As noted, the purpose of this exercise was not to establish a universally applicable taxonomy for all aspects of intentions and effects of transport policies, but rather to create a schema tailored to the specific needs of policy-makers and associated actors. As a result, the typology was concerned with establishing conceptual clarity, providing a useful categorisation of types of non-intentional policy effects and supporting the design of policy packages later discussed in Chapter 4. The final typology is extremely detailed and undertakes to analyse the nature of policy effects with respect to numerous variables. ...
Summary and Future Work. We have described an efficient iterative approach for solving separable non- linear inverse problems. Many researchers have studied the separable non- linear least squares problem, but few have specifically applied it to large- scale ill-posed inverse problems. We have addressed this problem and shown that by combining a Xxxxx-Xxxxxx approach for minimizing a reduced cost functional with a sophisticated iterative solver for computing Tikhonov reg- ularized solutions for linear ill-posed inverse problems, one can efficiently solve large-scale nonlinear inverse problems, with relatively little user input required. Nonlinear inverse problems of this form arise in many applications, and we have provided two examples from image deblurring in which the proposed algorithm can successfully update both the image and the imaging parame- ters simultaneously. Future work includes understanding how the additional regularization term affects the theoretical convergence properties of this al- gorithm when applied to ill-posed problems and developing an automated way to select the stopping iteration.
Summary and Future Work. In this paper we have elaborated a concept that combines WCF with code contracts. As a consequence, WCF appli- cation developers – both on server and client side – can now profit from the additional expressive power of code contracts including runtime and tool support. It has been stressed elsewhere that there does not exist a generic solution yet. Our novel approach exploits well-known standards such as WSDL and WS-Policy. We have described how to transform code contracts expressions contained in the WCF service into a programming language independent representation. This representation will be used to generate an enhanced client proxy infrastructure, thus allowing to evaluate the WCF service’s code contracts already on client side. We see several areas for future work. One direction is concerned with a precise definition of “WCF code contracts expressions.” When defining code contracts for WCF ser- vices, only those variables should be referred that are visible to the service consumer. While service parameters are public and hence meaningful for a service consumer, it is not useful for the client when members of the service implementation class are included into the created code contracts assertions. Therefore, rules should be defined that i) characterize valid expressions (similar to the ones presented in Section 5 on contract extraction in [10]) and ii) translate the code con- tracts statements into corresponding WS-Policy assertions embedded into the service’s WSDL description. Additional tool support for WCF code contracts is an- other topic. We have shown how a custom binding can be defined such that code contracts expressions are exported to (resp. imported from) the WSDL. For a WCF developer, it would be helpful to have a specific “WCF code contracts” project type for VisualStudio that automatically introduces the required elements in the WCF configuration files. This work is concerned with making code contracts available for a WCF client environment. Another interesting question is how a WCF service consumer developed with an alternative technology such as Java (see e.g., [11]) can process the code contracts expressions.
