440 likes | 585 Views
Mobile Calculi. Prof. Diletta Romana Cacciagrano. Operational Semantics based on reduction. Reduction semantics. Reduction semantics. :. ( red-cong ). Alpha-conversion. Structural congruence. Structural congruence. Operational Semantics based on labels.
E N D
Mobile Calculi Prof. Diletta Romana Cacciagrano
Reductionsemantics : (red-cong)
Labeledsemantics(Lateoperationalsemantics+alphaconv +structcong) α CONG α
Labeledsemantics(Earlyoperationalsemantics+alphaconv +structcong) α CONG α
Labeledsemantics Early and late LTSs
Input L
Congruencew.r.t.parallel(proofforearly. Similarlyfor late) Theorem:
Strong bisimilarityisnot a congruence(proofforearly. Similarlyfor late))
Open bisimulation(Bisimulationfor Pi) Nameinstantationismoved inside the definitionofbisimulation. The open bisimilarity, written , is the largest open bisimulation.
Testingmachinery A set ofprocessestobe test. A set oftests or observers. These are obtainedbyextending the syntaxofprocessesto generate processeswhich can perform a particularaction (omega) reporting success. A way toexercise a process on a given test: itisdonebyletting the process and the test torun in parallel and bylooking at the computationswhich the embeddedprocess can perform. Thesecomputations can besuccessful or failing, depending on whether or nottheyallow the executionof omega. A generalcriterion (semantics) forinterpreting the resultsoftheseexercises.
Testingmachinery Observer (Tests) Experiments
Testing and Bisimulationequivalences Bisimulationequivalences are usuallyratherstrict: theydepend on the whole branching structureofprocesseswhich, in some cases, are notrelevant. Weak bisimulation incorporates a particular notion of fairness: it abstracts from the tau-loops (i.e infinite sequences of tau-moves): the “normal” behavior can be resumed each time after a finite sequence of tau-moves. Must testing semantics is based on the interpretation of tau-loops as divergences, making them quasi-observable as a chaotic or under-specified behavior. Forthis, ithasbeendefinedfair-testingsemantics. The standard testing equivalences are coarser than weak bisimulation in the case of divergence-free processes, and they are incomparable in general.