Johan Lewi

A Linear Local Model Checking Algorithm for CTL

Transition systems play a central role as formal models for reactive and concurrent systems. A promising technique for automatically verifying transition systems is model checking. In the model-theoretic approach one mechanically determines whether a given transition system S0 meets a specification f by checking whether S0 is a model for f. In this paper Computation Tree Logic, a form of branching time temporal logic, is used as a specification tool. The paper presents a local model checking algorithm ALMC for CTL. Given a system state S0 (typically the initial system state) and a CTL formula S0, algorithm ALMC mechanically checks whether S0 satisfies formula S0. ALMC has time complexity linear in both the size of the system and the size of the formula to be checked. Furthermore ALMC is goal-oriented, i.e., in order to decide whether a formula S0 is true in a state S0, only a necessary part of the system-formula space is investigated. This is to be contrasted with the well-known global labeling algorithm of [4,5], where a priori all subformulae of S0 are checked for all states.

A Linear Algorithm for Solving Fixed-Point Equations on Transition Systems

In this paper we present an algorithm for effectively computing extremal fixed-points of a system of mutually recursive equations over a finite transition system. The proposed algorithm runs in time linear in the size of the transition system and linear in the size of the system of equations, thereby improving on [AC].

Efficient FixPoint computation

Most of the algorithms presented for computing fixpoints have been developed for very specific application areas, e.g. dataflow analysis, abstract interpretation, denotational semantics, system verification, to mention but a few. Surprisingly little attention has been devoted to the construction of general purpose, application independent fixpoint algorithms (one notable exception being [17]). The goal of this paper is to put known ideas and algorithms into a more general and abstract setting. More precisely we discuss a variety of efficient general purpose algorithms for computing (part of) the least solution of a monotone system of equations over a complete partial order structure. The advantage of having general purpose fixpoint algorithms is twofold. Firstly, once proven correct, they can be instantiated to a wide variety of application domains. Secondly, they separate the essentials of fixpoint computation from concrete application domain aspects. We consider algorithms based on (chaotic) fixpoint approximation, as well as algorithms based on fixpoint induction techniques. The algorithms are constructed in a stepwise fashion: First a basic schema, capturing the essence of the algorithm, is discussed, which is then subsequently refined using a number of optimisation steps. As a sample application, we sketch how an algorithm for computing the prebisimulation preorder is obtained, matching the complexity of the so far best known ’ad hoc’ constructed algorithm for this preorder.

Object-oriented programming: concepts

The authors present the result of a comparative analysis of different programming languages and systems that claim the label object-oriented. The concepts of persistence, concurrency, composition, perspective, distribution, location and mobility, communication, encapsulation, autonomy, and mutation are presented as characteristics of physical objects. Classifications are used by people to organize their knowledge of physical objects. The authors discuss the appearance of the concepts in programming languages and illustrate them through the different investigated languages. They survey a conceptual framework that gives the physical-world metaphor its due and that will be supported in the TIE-system.<<ETX>>

Constraints in Object-Oriented Analysis

Object-oriented analysis methods can incorporate the concept of constraints to express rules of the problem domain in the specification model, restricting the possible instances of the model. As such, constraints describe properties that must be true at each moment in time for the entire system, without determining how they are to be preserved. The ways in which these constraints are introduced in the model differ from method to method, and even between distinct constraint types in a single method. Different ways in which constraints can be described, are illustrated and compared.

Generation of interactive parsers with error handling

The generation scheme discussed, produces interactive transducers in the form of Ada programs, with an underlying parser that is of type ELL(1). The emphasis is on error recovery in interactive parsers. A generation scheme is proposed containing powerful error-recovery generation capabilities. The interaction between syntactic and semantic error recovery is also discussed. The generation scheme has been implemented as part of the MIRA transducer writing system. With MIRA, a number of industrial case studies have been worked out, from which considerable feedback has been obtained to test and improve the adopted error-recovery strategy. One of the case studies worked out with MIRA consists of the design and implementation of an interactive software package called ABACUS. A subset called MINI-ABACUS is used as an illustration of the error-recovery principles discussed throughout this work. >

Efficient Computation of Nested Fix-Points, with Applications to Model Checking

The paper presents a general algorithm for computing nested fix-points over complete lattices of finite height. The method presented relies on techniques familiar from the realm of functional programming languages, such as e.g. lazy evaluation. The algorithm is constructed in a stepwise fashion: We start with a schema based on some simple facts of fix-point theory. As such this schema is easily seen to be correct. It is, however, rather inefficient. We then trace the sources of inefficiency and refine the basic schema resulting in a correct and more efficient algorithm. After presenting the general algorithm, we apply it, by means of illustration, to the field of model checking.

Formal specifications: an industrial case study

The authors present an algebraic specification of two-party voice calls. Abstraction is made from technical details. Only relevant characteristics are specified. Emphasis is put on modularity and on a so-called object-oriented design methodology. As a vehicle for this exposition, the authors use a constructive algebraic specification language that allows rapid prototyping.<<ETX>>

Linear local and global model checking algorithms for a kernal temporal logic language

Finite transition systems play a central role as formal models for reactive and concurrent systems. A promising technique for automatically verifying transition systems is model checking. In the model-theoretic approach one mechanically determines whether a system ts meets a specification f (expressed as a formula of some temporal logic) by checking whether ts is a model for f. The authors present a kernel specification formalism BTL. They discuss efficient (linear-time) global and local model checking algorithms for BTL. The global algorithm will compute all states that satisfy a given formula whereas the local algorithm will only check a formula for one state (typically the initial system state). Both algorithms run in time linear in the size of the transition system and the length of the formula.<<ETX>>

A Case Study: the Mini-PABX

One of the most interesting case studies we made is the formal specification of a substantial part of a call handling system, the ITT 5400 BCS (Business Communication System) [Bell85b]. The ITT 5400 BCS is a modern Private Automatic Branch Exchange (PABX for short), which has been developed and produced by Bell Telephone Mfg. Co. Geel (Belgium) in the context of ITT’s Office 2000 concept. Because voice communication accounts for some 80% of all office communications [Bell85a], the ITT 5400 BCS is supplied with a wide range of features. These features considerably improve the flow of information, provide more ease of operation and save time and costs. The range of features for voice communications includes extension features (i.e. features for the ordinary users), operator features and system features.