Reisig, W.: Petri net models for distributed algorithms. In Computer Science Today | Recent Trends and Developments, Lecture Notes in  Home/Search   Document Not in Database   Summary   Related Articles   Check

....logics seems to be a non trivial task. Some partial results may be found in [39] In [34] the authors present proof rules for the logic ISTL with a trace semantics together with a relative expressive completeness result. Reisig has also developed a kit of proof rules for a version of UNITY logic [40, 41]. The models of this logic are the non sequential processes of a net system and the proof rules are mainly designed to help reason about distributed algorithms modelled using net systems. At present not much is known about corresponding logics in a branching time setting. Most of the attempts in ....

Reisig, W.: Petri net models for distributed algorithms. In Computer Science Today | Recent Trends and Developments, Lecture Notes in

....In general the concept of progress is associated with actions (which do or do not take place) In Petri net modelling, actions are represented by transitions, hence the concept of progress is connected with transitions. The concept of progress has been investigated and discussed by Reisig [95], where the set of transitions was divided into two disjoint sets: the set of progressing transitions and the set of quiescent transitions. The work in [95] illustrated how the progress property formed an important part of an approach to verifying distributed algorithms using Petri nets and ....

....hence the concept of progress is connected with transitions. The concept of progress has been investigated and discussed by Reisig [95] where the set of transitions was divided into two disjoint sets: the set of progressing transitions and the set of quiescent transitions. The work in [95] illustrated how the progress property formed an important part of an approach to verifying distributed algorithms using Petri nets and Temporal Logic. In our work we assume that all transitions satisfy the progress propertywhich states that a transition can not remain enabled forever without ....

W. Reisig. Petri net models of distributed algorithms. Lecture Notes in Computer Science, vol. 1000, Springer Verlag, 1995.

....view. Most temporal logics introduced for Petri nets do not support both views these logics are either state based or event based. In this paper we introduce a temporal logic which supports both views. This logic is an extension of a set of related state based temporal logics for Petri nets [Rei95,WVV 96,Rei97b] which have been developed for the verification of distributed algorithms and are called Distributed Algorithms Working Notation (DAWN) We call the extension Event and State based Temporal Logic (ESTL) ESTL can be used in the early design phases for formalizing informal ....

....they will be formally introduced in Sect. 3. 3 In contrast to the state based logic, we also allow transitions in the expressions; moreover, we use different but related temporal operators. In this paper we formally define the meaning of these formulas and show how the state based logics of [Rei95,WVV 96] smoothly extend to an event based logic. Note that we use different symbols for the same temporal operators in the event based version and the state based version because we want to integrate both views in a single temporal logic. The use of different operators allows for a nesting ....

[Article contains additional citation context not shown here]

Wolfgang Reisig. Petri net models of distributed algorithms. In J. van Leeuven, editor, Computer Science Today: Recent Trends and Developments, LNCS 1000, pages 441--454. Springer-Verlag, Berlin, 1995.

....[8] or a vector of terms [15] Rather, we represent a place invariant as a multiset valued linear expression in which place names may occur as bag valued variables. Though this difference is only syntactical, it allows a smoother transition between Petri net properties and temporal logic (cf. [16, 11, 21, 10]) Moreover, it gives rise to a generalization: We can use expressions which evaluate to an arbitrary commutative monoid equipped with some affine preorder. We call this generalization simulation algebraically, a simulation is a homomorphism from the occurrence graph of the net to the ....

....a place invariant as a linear expression in which places may occur as variables. This representation was inspired by verification techniques for algebraic system nets, since linear expressions allow a smooth transition form Petri net concepts such as place invariants to temporal properties (cf. [16, 11, 21, 10]) 4. Reisig [15] introduces a firing rule as semantics for algebraic nets, only. In this paper we also introduce the non sequential behaviour for algebraic system nets, which we call processes of the algebraic system net. This is justified, since we have shown that the set of processes of an ....

W. Reisig. Petri net models of distributed algorithms. In J. van Leeuwen, editor, Computer Science Today: Recent Trends and Developments, volume 1000 of LNCS, pages 441--454. Springer-Verlag, 1995.

....or event based. In this paper we introduce a temporal logic which supports both views. This logic is an extension of a set of related state based temporal logics for Petri nets supported by the DFG Forschergruppe Petri Net Technology y e mail: fkindler,vesperg informatik.hu berlin.de [Rei95, KW95, WVV 96, Rei97b] which have been developed for the verification of distributed algorithms and are called DAWN 1 for this reason. The proposed extension of the logic can be used in the early design phases for formalizing informal requirements as well as in the later phases to verify ....

....logic: Sigma 1 j= check get visa ) 2) In contrast to the state based logic, we also allow transitions in the expressions; moreover, we use different but related temporal operators. In this paper we formally define the meaning of these formulas and show how the state based logics of [Rei95, KW95] smoothly extend to an event based logic. Note that we use different symbols for the same temporal operators in the event based version and the state based version because we want to integrate both views in a single temporal logic. The use of different operators allows for a nesting of ....

[Article contains additional citation context not shown here]

Wolfgang Reisig. Petri net models of distributed algorithms. In J. van Leeuven, editor, Computer Science Today: Recent Trends and Developments, LNCS 1000, pages 441--454. Springer-Verlag, Berlin, 1995.

....of PTL studied in [T95] For coherent formulas, weak model checking is seen to be decidable as well. Apart from formal interest, is weak model checking relevant in the context of specifying and reasoning about distributed systems Rather interestingly, the p causes q modality discussed by [R95] has the same flavour as asserting reachability in some interleaving. PP94] studies a logic which is more expressive than PTL where these situations are easily described. However, we have no detailed examples of system verification necessitating weak model checking. A word about the syntax of ....

Reisig, W., "Petri net models of distributed algorithms", Tech. Rep. 58, Humboldt Univ., Berlin, 1995 (to appear in LNCS 1000).

.... a labelled occurrence net, obeying some additional requirements. In our definition we weaken these requirements as much as possible. Thus we do not require, that elements in a concurrent run can t occur concurrently to themselves, as it is ordinary done in the case of contact free nets (cf. [6]) Moreover, we do not require, that every transition in a concurrent run must be reachable via some occurrence sequence, as in [8] Definition 3.3 Let Sigma = P; T; F ; S Sigma ) be an EN system. A concurrent run (co run for short) of Sigma is a pair (K; l) where K is an occurrence net ....

Reisig, W.: Petri Net Models of Distributed Algorithms. Lecture Notes in Computer Science, 1000 (1995).

....known message passing algorithm for mutual exclusion in an arbitrary connected graph, formulated in terms of a high level Petri net. We give a rigorous correctness proof for this algorithm, exploiting its representation as a Petri net. Petri nets are well suited for modeling distributed algorithms [10, 2, 15, 21, 14, 22]: a place can be interpreted as a channel from its input transitions to its output transitions in a message passing architecture or it can be interpreted as a semaphore in a shared memoryarchitecture. Moreover, places can represent local states of sequential components. Figure 1 shows a simplified ....

....we make intensive use of unless properties that are immediate consequences of local dependencies explicitly represented in the net. This work is part of a project that develops a general framework for the specification and verification of distributed algorithms with Petri nets (see e.g. [21, 14, 22, 13]) In the present paper, we only extract the part of the formalism which is necessary for understanding and proving our example (and similar ones) 2 Petri nets We roughly follow the notions for high level Petri nets given in [9] called Colored Petri nets there. However, since our class of ....

[Article contains additional citation context not shown here]

Wolfgang Reisig. Petri net models of distributed algorithms. In Jan van Leeuwen, editor, Computer Science Today: Recent Trends and Developments, LNCS 1000, pages 441--454. Springer-Verlag, 1995.

....semantics. Semantical: Our semantics is fully abstract with respect to the composition operation and a reference semantics for closed components to be defined. Essentially, the reference semantics is the set of processes of a place transition system [GR83] equipped with the progress assumption [Rei87, KW95, Rei95]. Practical: The semantics can be used to design systems in a modular way in combination with a temporal logic interpreted on processes of place transition systems. In particular, we can specify components in a rely guarantee style. 2 Basic definitions Now, we present the basic notions known ....

....the previous section. 3.1 Components A component is a net with some addition: As shown in the example, we distinguish some places as input and output places of the component. Moreover, we distinguish two kinds of transitions: progress transitions and quiescent transitions. A progress transition [Rei95, KW95] of a component which is enabled forever must eventually occur. Most transitions of systems and components are assumed to be progress transitions. Therefore, we set off the quiescent transitions T q in the formal definition. Definition 6 (Component) Let Sigma = N; M) be a system net with N = ....

Wolfgang Reisig. Petri net models of distributed algorithms. In Jan van Leeuwen, editor, Computer Science Today: Recent Trends and Developments, LNCS 1000, pages 441--454. Springer-Verlag, 1995.

....fact, the fairness solutions we considered were far away from reality since they would have involved global constraints, such as queues on every place. The work by Reisig et al. contributes a notion of fair transitions. Transitions cannot be enabled infinitely often without occurring (see, e.g. [93]) Clarke et al. 22] takes a different approach where fairness is enforced on the (execution) paths considered in the state space in the context of model checking temporal properties (expressed in CTL) Jensen [55] uses the strongly connected component graph (SCC graph) in order to determine ....

W. Reisig. Petri net models of distributed algorithms. In J. van Leeuven, editor, Computer Science Today: Recent Trends and Developments, volume 1000 of Lecture Notes in Computer Science, pages 441--454. SpringerVerlag, 1995.

....algebraic system nets. The P invariant calculus of Reisig [23] can be adapted to algebraic system nets, which will be demonstrated in a forthcoming paper [16] We add two more features to algebraic system nets, which have been shown to be necessary for adequately modelling distributed systems [24]: the distinction of progress and external transitions and fair arcs [18] When talking about a distributed algorithm we actually do not talk about a single algebraic system net, but a class of algebraic system nets. The reason is that distributed algorithms usually work for different ....

Wolfgang Reisig. Petri net models of distributed algorithms. In Jan van Leeuwen, editor, Computer Science Today: Recent Trends and Developments, LNCS 1000, pages 441--454. Springer-Verlag, 1995.

....The use of Petri nets supports this aim, abstracting from a concrete programming notation together with the use of techniques based on multisets and state expressions over the places of the Petri net. Studies have shown that a lot of distributed algorithms fit to the proposed proof style, e.g. [7, 10]. ....

W. Reisig: "`Petri Net Models of Distributed Algorithms"'. In Leeuven, Jan van (Ed.): Computer Science Today. Recent Trends and Developments. Lecture Notes in Computer Science 1000. Berlin: Springer-Verlag: 1995: 441--454.

No context found.

Reisig W.: Petri Net Models of Distributed Algorithms. In Computer Science Today: Recent Trends and Developments, Jan van Leeuven (ed.), Springer Verlag, LNCS vol 1000, 1995.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC