Wyniki wyszukiwania - BazTech

Ograniczanie wyników

2 Fundamenta Informaticae

2 2017

Powiadomienia systemowe

Sesja wygasła!

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: sequentialisability

Sortuj według:

Ogranicz wyniki do:

Invariant Structures and Dependence Relations

Janicki R., Kleijn J., Koutny M., Mikulski Ł.

Fundamenta Informaticae

2017

Vol. 155, nr 1/2

1--29

A step trace is an equivalence class of step sequences which can be thought of as different observations of the same underlying concurrent history. Equivalence is determined on basis of a step alphabet that describes the relations between events in terms of potential simultaneity and sequentialisability. Step traces cannot be represented by standard partial orders, but require so-called invariant structures, extended order structures that capture the phenomena of mutual exclusion and weak causality. In this paper, we present an effective way of deciding whether an invariant structure represents a step trace over a given step alphabet. We also describe a method by which one can check whether a given invariant structure can represent a step trace over any step alphabet. Moreover, if the answer is positive, the method provides a suitable step alphabet.

Alphabets of Acyclic Invariant Structures

Janicki R., Kleijn J., Koutny M., Mikulski Ł.

Fundamenta Informaticae

2017

Vol. 154, nr 1/4

207--224

A step trace is an equivalence class of step sequences, where the equivalence is determined by dependencies between pairs of actions expressed as potential simultaneity and sequentialisability. Step traces can be represented by invariant structures with two relations: mutual exclusion and (possibly cyclic) weak causality. An important issue concerning invariant structures is to decide whether an invariant structure represents a step trace over a given step alphabet. For the general case this problem has been solved and an effective decision procedure has been proposed. In this paper, we restrict the class of order structures being considered with the aim of achieving a better characterisation. Requiring that the weak causality relation is acyclic, makes it possible to solve the problem in a purely local way, by considering pairs of events, rather than whole structures.