Synchronous and Asynchronous Interactions in Concurrent Distributed Systems
ICALP 2008 affiliated workshop
6th of July 2008 - Reykjavik, Iceland
The University of Pisa.
The programme for the workshop and the papers from the pre-proceedings can be found here.
The scope of this first experience is to include theoretical and applied aspects of interactions and the synchronization mechanisms used among actors of concurrent or distributed systems. The workshop intends to attract researchers interested in models, verification, tools, and programming primitives concerning such complex interactions.
Synchronisation mechanisms are one of the key aspects in concurrency and they are becoming enormously relevant in modern distributed systems. Theoretical models, design and verification of interaction protocols and programming practice must take synchronisations into account for specifying, implementing and reasoning on systems where computations are spread across possibly many actors that interact within a precise interaction framework.
At a low level of abstraction, systems can be classified according to a wide spectrum, ranging between the two extremes of (completely) synchronous or asynchronous interactions. In fact, such a classification can be given according to the assumptions made on, e.g., the number of participants or the time interactions need to be effected. Significantly, the behaviour of such systems can be investigated using different assumptions that yield different expressiveness or complexity results.
Several recent theoretical results shed light on the interrelations between synchronous and asynchronous interaction mechanisms (e.g., expressiveness results for distributed algorithms, relations among observational semantics of (a)synchronous models). Interaction mechanisms have also been studied in relation to other features of systems such as mobility (e.g., name passing process calculi, graph-based models).
Recent models of distributed systems (e.g., Service Oriented or Overlay Computing, multi-agent systems) pose further challenging problems. For instance, overlay computing envisages systems as abstract computations among loosely coupled partners relying on heterogeneous communication infrastructures which typically exploit different interaction mechanisms. In this context, several interaction mechanisms based on different underlying concepts have been proposed (e.g., sessions, data driven coordination, probabilistic/timed models). Interestingly, recent lines of research have (re)considered synchronisation mechanisms proposed in the past (for instance, the CCS- or CSP-like synchronisation and weaker variants of them) as coordination mechanisms that can suitably represent interactions of distributed systems at different levels of abstraction. For instance, abstract interactions can be suitably modelled using CSP-like synchronisation and then mapped, at a lower level, to CCS-like ones that correspond more closely to e.g., the synchronisation in TCP/IP protocols.
Additionally, recent models of computations (e.g., those inspired by other disciplines like biology or chemistry) utilise novel kinds of interactions that typically interlace with other important characteristics of systems (e.g., spatial or environmental conditions).
Finally, synchronisation aspects of interactions also affect the implementation and verification of systems. Theoretical results can indeed suggest new techniques and methods for implementing systems and the efficient verification of their properties. Examples of this include the recent works proposing type systems or algorithms based on checking behavioural equivalences of systems whose specification exploits a different interaction mechanism with respect to the implementation.