Nima's research aims to assist designers of distributed object and
component-based software architectures in detecting deadlocks, as well
as violations of liveness and safety properties. Nima has defined a
UML profile for distributed objects using UML extension mechanisms,
defined the semantics of stereotypes and tagged values in terms of
a mapping to a process algebra and is then able to use a model checker
for the process algebra to detect design flaws.
Nima's research is partly self-funded and partly supported by the
Software Systems Engineering group.
- N. Kaveh and W. Emmerich (2003).
Validating Distributed Object and Component Designs.
In M. Bernardo and P. Inverardi (ed): Formal Methods for Software Architecture, pp. 63-91, Lecture Notes in Computer Science 2804, Springer Verlag.
- N. Kaveh and W. Emmerich (2001).
Deadlock Detection in Distributed Object Systems.
In V. Gruhn (ed): Joint Proc. of the 8th European Software Engineering Conference and the 9th ACM SIGSOFT Symposium on the Foundations of Software Engineering, Vienna, Austria, pp. 44-51, ACM Press.
Daniel is investigating the problem of engineering XML-based
languages in a systematic way. The problem is motivated by the surge
in number of XML languages, both standardised and languages that have
been defined within an organisation. Daniel is investigating how the
evolution of these languages can be managed in such a way that
compatibility to previous language versions is maintained.
Daniel benefits from a generous studentship from UBS.
- D. Dui and W. Emmerich (2003).
Compatibility of XML Language Versions.
In B. Westfechtel and A. v. d. Hoek (ed): Software Configuration Management. Selected Papers of ICSE Workshops SCM 2001 and SCM 2003, pp. 148-162, Lecture Notes in Computer Science 2649, Springer Verlag.
- D. Dui and W. Emmerich and C. Nentwich and B. Thal (2003).
Consistency Checking of Financial Derivative Transactions.
In M. Aksit and M. Menzini and R. Unland (ed): Objects, Components, Architectures, Services and Applications for a Networked World, pp. 166-183, Lecture Notes in Computer Science 2591, Springer Verlag.
Clovis works on distributed resource allocation and scheduling in
autonomous computational grids. Clovis has established the notion of
overlay grids - multiple virtual organizations sharing the same
computational resources. The research question that he is pursuing is
how resource allocation and scheduling policies can be defined in the
presence of overlay grids and how these policies can be enforced.
Clovis is a research assistant on the e-Minerals project funded by the
National Environment Research
Council and a part-time PhD student.
- C. Chapman, P. Wilson, T. Tannenbaum, M. Farrellee, M. Livny,
J. Brodholt and W. Emmerich (2004). Condor Services for the Global
Grid: Interoperability between Condor and OGSA. In S. Cox (ed): Proc
of the 2004 UK E-Science All Hands Meeting,
Nottingham. pp. 870-877. UK Engineering and Physical Science Research
C. Chapman, M. Musolesi, W. Emmerich and C. Mascolo (2007). Predictive
Resource Scheduling in Computational Grids. In Proc. of the 21st
Int. Parallel and Distributed Processing Symposium, Long Beach,
CA. pp. 1-10. IEEE Computer Society Press.
Andy is working at the intersection between programming languages,
databases and software engineering. The particular problem he is
interested in is the impact that a change of a database schemas has on
application programs that are written against that schema.
Andy's research is funded by a studentship from
Torsten works on devising systematic ways for providing incentives
to sharing resources in peer-to-peer and grid networks. He observed
that current resource sharing is provided just on a voluntary basis in
current peer-to-peer and grid systems. Monetary and virtual currency
mechanisms are deemed too heavyweight and generally require a "central
bank", which is incompatible with the philosophy of peer-to-peer
networks. Torsten's hypothesis is that the overal quality of service
will be improved if systematic bartering is facilitated. To achieve
that peer-to-peer systems could detecting bartering rings and then
provide improved quality of service to members of such bartering
rings. Torsten's aim is to devise efficient and light-weight
distributed algorithms to detect, maintain and exploit these bartering
rings and then to show that they indeed provide an incentive to
increase resource sharing in practice.
Torsten benefits from a PhD studentship from Kodak Research.
The overall aim of Bruno's PhD is to investigate techniques to
increase the visibility of the system components relevant for the
reliable execution of a given service composition and provide
automated support to administrators during failure detection and
diagnosis. The main aspect of the proposed contribution is a failure
diagnosis mechanism, which extends existing approaches to problem
determination and root cause localisation based on machine learning in
several ways. Bruno aims to characterise certain quality attributes of the
root cause localisation mechanism through a series of controlled
experiments and then establish the effect of these attributes on the
availability and time to repair on real systems in practice.
Bruno is funded through a BT/EPSRC Case Studentship.