UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

Model checking multi-agent systems

Raimondi, F.; (2006) Model checking multi-agent systems. Doctoral thesis , University of London. Green open access

[thumbnail of 5627.pdf]

Download (1MB)


A multi-agent system (MAS) is usually understood as a system composed of interacting autonomous agents. In this sense, MAS have been employed successfully as a modelling paradigm in a number of scenarios, especially in Computer Science. However, the process of modelling complex and heterogeneous systems is intrinsically prone to errors: for this reason, computer scientists are typically concerned with the issue of verifying that a system actually behaves as it is supposed to, especially when a system is complex. Techniques have been developed to perform this task: testing is the most common technique, but in many circumstances a formal proof of correctness is needed. Techniques for formal verification include theorem proving and model checking. Model checking techniques, in particular, have been successfully employed in the formal verification of distributed systems, including hardware components, communication protocols, security protocols. In contrast to traditional distributed systems, formal verification techniques for MAS are still in their infancy, due to the more complex nature of agents, their autonomy, and the richer language used in the specification of properties. This thesis aims at making a contribution in the formal verification of properties of MAS via model checking. In particular, the following points are addressed: • Theoretical results about model checking methodologies for MAS, obtained by extending traditional methodologies based on Ordered Binary Decision Diagrams (OBDDS) for temporal logics to multi-modal logics for time, knowledge, correct behaviour, and strategies of agents. Complexity results for model checking these logics (and their symbolic representations). • Development of a software tool (MCMAS) that permits the specification and verification of MAS described in the formalism of interpreted systems. • Examples of application of MCMAS to various MAS scenarios (communication, anonymity, games, hardware diagnosability), including experimental results, and comparison with other tools available.

Type: Thesis (Doctoral)
Title: Model checking multi-agent systems
Open access status: An open access version is available from UCL Discovery
Language: English
URI: https://discovery.ucl.ac.uk/id/eprint/5627
Downloads since deposit
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item