Skip to main content
QUICK REVIEW

[Paper Review] Expressive Completeness of Metric Temporal Logic

Paul Hunter, Joël Ouaknine|arXiv (Cornell University)|Aug 24, 2012
Formal Methods in Verification12 references3 citations
TL;DR

This paper establishes that Metric Temporal Logic (MTL) with rational constants is expressively complete for first-order logic with order and rational successor functions ($\mathit{FO}(<,+\mathbb{Q})$). By generalizing Gabbay's separation principle to the metric setting and leveraging bounded quantifier relativization, the authors construct a translation from $\mathit{FO}(<,+\mathbb{Q})$ to MTL, resolving a long-standing open question about the expressive power of quantitative temporal logics.

ABSTRACT

Metric Temporal Logic (MTL) is a generalisation of Linear Temporal Logic in which the Until and Since modalities are annotated with intervals that express metric constraints. A seminal result of Hirshfeld and Rabinovich shows that over the reals, first-order logic with binary order relation &lt; and unary function +1 is strictly more expressive than MTL with integer constants. Indeed they prove that no temporal logic whose modalities are definable by formulas of bounded quantifier depth can be expressively complete for FO(

Motivation & Objective

  • To resolve the expressive completeness of MTL for $\mathit{FO}(<,+\mathbb{Q})$ in the context of real-time verification.
  • To extend Gabbay's separation principle to metric temporal logics, enabling inductive translation from first-order logic to MTL.
  • To overcome the limitations of MTL with integer constants, which are known to be strictly less expressive than $\mathit{FO}(<,+1)$.
  • To provide a canonical, variable-free temporal logic counterpart to $\mathit{FO}(<,+\mathbb{Q})$, mirroring Kamp’s theorem for LTL.

Proposed method

  • Generalize Gabbay’s notion of separation to MTL, showing every MTL formula is equivalent to a Boolean combination of formulas restricted to near present, distant future, or distant past.
  • Define $N$-bounded $\mathit{FO}(<,+\mathbb{Q})$ formulas, where all quantifiers are relativized to $(x-N,x+N)$, and construct a translation from such formulas to MTL.
  • Introduce auxiliary monadic predicates ($P_{=}, P_{<}, P_{>}, P_{+}, P_{-}$) to encode relative position of a variable $x$ with respect to other variables in $\mathit{FO}(<,+\mathbb{Q})$ formulas.
  • Use a scaling argument to lift a translation from $\mathit{FO}(<,+1)$ to MTL to a full translation from $\mathit{FO}(<,+\mathbb{Q})$ to MTL.
  • Apply the bounded formula translation to each component of the separated formula, then eliminate auxiliary predicates by substituting atomic conditions.
  • Prove that the resulting MTL formula is equivalent to the original $\mathit{FO}(<,+\mathbb{Q})$ formula by combining the separation, boundedness, and predicate substitution techniques.

Experimental results

Research questions

  • RQ1Is there a temporal logic with rational constants that is expressively complete for $\mathit{FO}(<,+\mathbb{Q})$?
  • RQ2Can Gabbay’s separation principle be generalized to the metric setting to support inductive translation from first-order logic to MTL?
  • RQ3Why does MTL with integer constants fail to be expressively complete for $\mathit{FO}(<,+1)$, and how does allowing rational constants resolve this?
  • RQ4Can the expressive power of MTL be extended to match the full expressive power of $\mathit{FO}(<,+\mathbb{Q})$ via a systematic translation procedure?

Key findings

  • MTL with rational constants is expressively complete for $\mathit{FO}(<,+\mathbb{Q})$, establishing the first full analog of Kamp’s theorem in the quantitative setting.
  • The proof relies on a generalized separation principle for MTL, where formulas are rewritten as Boolean combinations of near-present, distant-future, and distant-past components.
  • Bounded $\mathit{FO}(<,+\mathbb{Q})$ formulas—those with quantifiers restricted to a finite interval around the current point—can be translated into equivalent MTL formulas.
  • The use of auxiliary monadic predicates allows the translation of $\mathit{FO}(<,+\mathbb{Q})$ formulas to MTL by encoding relative position and arithmetic constraints.
  • The translation process preserves equivalence and eliminates auxiliary variables through substitution, resulting in a pure MTL formula.
  • The result resolves an open question about the expressive relationship between MTL and TPTL in the presence of both past and future modalities, showing that MTL is as expressive as $\mathit{FO}(<,+\mathbb{Q})$.

Better researchstarts right now

From reading papers to final review, dramatically reduce your research time.

No credit card · Free plan available

This review was created by AI and reviewed by human editors.