A temporal fixpoint calculus

Moshe Y. Vardi

1988 · 205 citations · 21 references

DOIFull text

Open access

Concepts

TL;DR

Temporal logic has recently been extended in two ways—by adding fixpoint operators to reason about arbitrary regular events and by adding past temporal connectives to refer to computation history—both aimed at enabling modular verification. We introduce and study the logic μTL, which extends temporal logic with both fixpoint operators and past temporal connectives. We extend the automata-theoretic paradigm to μTL by showing that any μTL formula can be translated into a finite‑state Büchi automaton whose size is at most exponential in the formula and that accepts exactly the computations satisfying the formula. This translation yields immediate applications, such as an optimal decision procedure and a model‑checking algorithm for μTL.

Abstract

Two distinct extensions of temporal logic has been recently advocated in the literature. The first extension is the addition of fixpoint operators that enable the logic to make assertions about arbitrary regular events. The second extension is the addition of past temporal connectives that enables the logic to refer directly to the history of the computation. Both extensions are motivated by the desire to adapt temporal logic to modular, i.e., compositional, verification (as opposed to global verification). We introduce and study here the logic μTL, which is the extension of temporal logic by fixpoint operators and past temporal connectives. We extend the automata-theoretic paradigm to μTL. That is, we show how, given an μTL formula @@@@, we can produce a finite-state Büchi automaton A@@@@, whose size is at most exponentially bigger than the size of @@@@, such that A@@@@ accepts precisely the computations that satisfy @@@@. The result has immediate applications, e.g., an optimal decision procedure and a model-checking algorithm for μTL.

References

21