| |
| Investigators: | Prof P Henderson, Southampton University |
| Collaborators: | ICL DERA Above Water Systems, Portsmouth |
EPSRC Funding: £424,841
Abstract:
Wiith everything connected to everything else in the coming information age, and with errors and temporal delays, the totality of information, even within subsystems, will almost inevitably contain inconsistencies. The project aims to establish a combination of an architecture for ‘Enterprise Application Integration’, ‘inconsistency-tolerant components’, and an approach to reasoning about consistency across a large-scale system. This is a three-year project, following an initial one-year feasibility study.
Our belief is that it is possible to develop system architectures which are sufficiently powerful for realistic Enterprise Application Integration but about which it is possible to demonstrate their a priori validity. An example of the type of information consistency problems which concern us are those that arise when a payment from one enterprise to another gives a valid reference number but an invalid account number. This could be a logical error (a simple mistake) or a temporal error (one of the numbers has changed recently) or an attempted fraud, or one of many possibilities. Any action the system takes unilaterally in these circumstances could be wrong and may need to be reversed. Interesting effects arise when many such inconsistent transactions interact. Reasoning that the systems behaviour will be consistent in some local sense may need to accept that the system will never be consistent in a global sense.
| Further information: | http://www.ecs.soton.ac.uk/~ph/Rices.htm |
GO TO .... | ideo home | SII home | EPSRC SI programme home | SII call for proposals | SII proposers' guidance | SII contacts
