We give a semantics for Message Flow Graphs(MFGs), which play the role for interprocess communication that Program Dependence Graphs play for control ow in parallel processes. MFGs have been used to analyse parallel code, and are closely related to Message Sequence Charts and Time Sequence Diagrams in telecommunications systems. Our requirements are rstly, to determine unambiguously exactly what execution traces are speci ed by anMFG, and secondly, to use a nite-state interpretation. Our methods function for both asynchronous and synchronous communications. From a set of MFGs, we de ne a transition system of global states, and from that a Buchi automaton by considering safety and liveness properties of the system. In order easily to describe liveness properties, we interpret the traces of the transition system as a model of Manna-Pnueli temporal logic. Finally,we describe the expressive power of MFGs by mimicking an arbitrary Buchi automaton by means of a set of MFGs. 1.

syntax of HMSC A hierarchical graph is a mathematical structure that represents the information contents of an HMSC. The set HGid represents the set of all HMSC names. Obviously, this includes the names of BMSCs. Since we did not provide a formal graphical syntax for HMSC we cannot provide a formal mapping from HMSC to hierarchical graphs. However, the intuition is clear. A node in an HMSC contains a reference to another HMSC via its name. Definition 3.3.1 (Hierarchical graphs) A hierarchical graph is either a BMSC or a tuple #id, Nodes, Starts, Ends, Edges, l#, where . id # HGid is the name of the hierarchical graph; . Nodes, Starts, and Ends are pairwise disjoint sets of HMSC reference nodes, start nodes and end nodes respectively with Starts #= ?; . Edges # (Nodes # Starts) × (Nodes # Ends) is a set of edges. An edge (n, n # ) is denoted by n # n # ; . l : Nodes # HGid is a labeling function which associates to a node a reference to an HMSC by means of a...

Introduction Recently, the ITU 1 -standardised specification language Message Sequence Chart (MSC) [IT96] has been extended with constructs for more complete and structured specifications. The new version of the language is called MSC'96. Currently, research is performed on the extension of the old formal semantics towards a semantics for MSC'96. Ideally, the development of a language and its semantics should go hand in hand. There is little use in defining a fancy syntactic construction without a precise understanding of its meaning. As was the case for the old version of the MSC language, first the syntax and an incomplete and informal semantics were developed, while the construction of a formal semantics was deferred until after the acceptance of the language by the ITU bodies. It is obvious that the a-posteriori construction of a formal semantics will reveal many places in which the informal language description is ambiguous, under-specified, inconsistent

Abstract. Synthesizing proper implementations for scenario-based specifications is often impossible, due to the distributed nature of implementations. To be able to detect problematic specifications, realizability criteria have been identified, such as non-local choice. In this work we develop a formal framework to study realizability of compositional MSC [GMP03]. We use it to derive a complete classification of criteria that is closely related to the criteria for MSC from [MGR05]. Comparing specifications and implementations is usually complicated, because different formalisms are used. We treat both of them in terms of a single formalism. Thereto we extend the partial order semantics of [Pra86,KL98] with a way to model deadlocks and with a more sophisticated way to address communication. 1

Message Sequence Charts (MSCs) is a visual notation for expressing requirements on communicating systems. Their expressive power has traditionally been somewhat limited, and additional information is usually needed by tools that manipulate them, e.g., to derive test suites. The new standard MSC-2000, developed by ITU-T, extends earlier versions by constructs for data and high-level control, so that it may be possible to derive test sequences directly from MSC requirements, without need of additional information. Motivated by this, we present an execution semantics for a signicant part of the MSC-2000 standard.

Timed algebraic process theories can be developed with quite different purposes in mind. One can aim for theoretical results about the theory itself (completeness, expressiveness, decidability), or one can aim for practical applicability to non-trivial protocols. Unfortunately, these aims do not go well together. In this paper we take two theories, which are probably of the first kind, and try to find out how well suited they are for practical verifications. We verify Fischer's protocol for mutual exclusion in the settings of discrete-time process algebra (ACP dt ) and real-time process algebra (ACP ur ). We do this by transforming the recursive specification into an equivalent linear specification, and then dividing out the maximal bisimulation relation. The required mutual exclusion result can then be found by reasoning about the obtained process graph. Finally, we consider the ease of the verification, and ways to adapt the theory to make it more practical. It will turn out that the...

Realistic models of computer and communication systems result in large, complex performance models. Compositionality, offered by stochastic process algebra constructs a model from submodels which are smaller and more tractable. We present a technique to exploit this structure in order to enhance the solution of the model by decomposition of the underlying Markov process. The decomposition under consideration is time scale decomposition, based on Courtois's near complete decomposability. This work has been influenced by related work on stochastic Petri nets and presents a major advancement related to a preliminary version of this technique already presented at a workshop of this series. In particular, a transformation technique is developed which is based on a delayed choice operator and preserves equivalence. 1 Introduction Stochastic process algebras (SPA) are formal description techniques by which systems are modelled as an interaction of autonomous agents or components who engage i...

The extension of the MSC language with more advanced data concepts is one of the current topics of discussion in the MSC standardization community. We discuss some problems and possibilities. By means of two case studies we study the practical consequences of our proposed approach.

Abstract. MSC is a visual formalism for specifying the behavior of systems. To obtain implementations for individual processes, the MSC choice construction poses fundamental problems. The best-studied cause is non-local choice, which e.g. is unavoidable in systems with autonomous processes. In this paper we characterize two additional problematic classes of choice nodes. Based on these three classes we point out some errors in related work. Extending our work on pragmatic implementations of nonlocal choice, we motivate a different choice semantics which allows a little more behavior. Finally, inspired by practical case studies, we present the first implementation approach for non-local choice nodes that can handle arbitrary numbers of processes. 1

Abstract. Currently, communication protocols for medical devices are being developed for the IEEE 1073.2 standard. The protocol description in its draft remote control package consists of a collection of intended behaviors in terms of MSCs. We have contributed to actually constructing the protocol, ranging from determining an hMSC for these MSCs, via synthesizing process implementations, to integrating it with the basic underlying IEEE 1073.2 protocol. In this paper we report on the non-local choice problems we encountered. We present a practical solution (i.e. an implementation) which on the one hand is close to the behavior specified in the hMSC, and on the other hand meets correctness properties such as deadlock freedom. These properties have been checked using the Spin model checker. We also give some directions for generalizing and extending this work. 1