Antonio Cerone's Website
Welcome >> Home >> Research >> Projects >>
Research Project

Formal Modelling and Analysis of Complex Systems

This project is based on the use of formal methodologies for the modelling and analysis of complex systems, which may involve one or more of the following components: software, hardware, humans, groupware and bioware.

The projects builds up on the two projects run during the biennium 2006-2007, Formal Models for Security and Formal Analysis of Human Behaviour in Interactive Systems., which made use the same modelling methodologies, based on automata and process calculi, and the same analysis tools, based on model-checking.

This project will further extend the applicability of these methodologies and tools, by involving quantitative analysis using probabilistic model-checking and the integration between model-checking and theorem-proving methodologies. Particular emphasis is given at the concept of interaction, which may involve software/hardware and humans/groupware (interactive systems), or individual molecules or cells in biological systems. In fact, interaction causes the emergence of system behaviours that affect the evolution, safety and security of the overall system.

Interactive systems are characterised by complex patterns of behaviours involving highly interleaved, concurrent, ongoing cognitive activities (e.g., air traffic control) or determined by social factors (e.g., collaborative systems, e-voting, group authentication protocols). Patterns of behaviours leading to security or safety failures will be decomposed into more elementary patterns which can be given a psychological interpretation. Model-checking is used to verify the soundness and completenss of the decomposition taxonomy.

Interaction in biological systems usually involves very sophisticated mechanisms (e.g, the movement of molecules across membranes), which cannot be easily modeled using the low-level primitives offered by traditional process calculi. On the other hand, the use of ad hoc calculi for the kind of phenomenum under investigation is not a good practice in a domain such as biology, where new kinds of phenomena are frequently discovered. The project involves the study of flexible formalisms based on basic primitives that can be combined together at a higher abstraction level to define sophisticated operators suitable for describing new kinds of phenomena. Analysis involves both qualitative aspects, such as existence of stable states, and quantitative aspects, such as the probability of reaching a certain state in a given time.

The project includes collaborations with: Bond University, Australia; Queensland University of Technology, Australia; the University of Queensland, Australia; University of Pisa, Italy; CNR-ISTI, Italy; Queen Mary University London, UK; the Interaction Centre of the University College London, UK.

BACK TO TOP
Created: Wed Jan 16 07:55:09 CST 20o8 Feedback Page
Updated: Thu May 27 12:01:03 CST 2010