11institutetext: Department of Mathematics and Statistics, University of Helsinki, Helsinki, Finland 11email: {juha.kontinen, max.sandstrom}@helsinki.fi 22institutetext: Department of Computer Science, University of Sheffield, Sheffield, UK 22email: [email protected]

A Remark on the Expressivity of Asynchronous TeamLTL and HyperLTL

Juha Kontinen 11 0000-0003-0115-5154    Max Sandström 1122 0000-0002-6365-2562    Jonni Virtema 1122 0000-0002-1582-3718
(June 7, 2024)
Abstract

Linear temporal logic (LTL) is used in system verification to write formal specifications for reactive systems. However, some relevant properties, e.g. non-inference in information flow security, cannot be expressed in LTL. A class of such properties that has recently received ample attention is known as hyperproperties. There are two major streams in the research regarding capturing hyperproperties, namely hyperlogics, which extend LTL with trace quantifiers (HyperLTL), and logics that employ team semantics, extending truth to sets of traces. In this article we explore the relation between asynchronous LTL under set-based team semantics (TeamLTL) and HyperLTL. In particular we consider the extensions of TeamLTL with the Boolean disjunction and a fragment of the extension of TeamLTL with the Boolean negation, where the negation cannot occur in the left-hand side of the Until-operator or within the Global-operator. We show that TeamLTL extended with the Boolean disjunction is equi-expressive with the positive Boolean closure of HyperLTL restricted to one universal quantifier, while the left-downward closed fragment of TeamLTL extended with the Boolean negation is expressively equivalent with the Boolean closure of HyperLTL restricted to one universal quantifier.

Keywords:
Hyperproperties Temporal Logic Team Semantics HyperLTL Verification

1 Introduction

In 1977 Amir Pnueli [15] introduced a core concept in verification of reactive and concurrent systems: model checking of formulae of linear temporal logic (LTLLTL\mathrm{LTL}roman_LTL). The idea is to view the accepting executions of the system as a set of infinite sequences, called traces, and check whether this set satisfies specifications expressed in LTLLTL\mathrm{LTL}roman_LTL. The properties that can be checked by observing every execution of the system in isolation are called trace properties. An oft-cited example of a trace property is termination, which states that a system terminates if each of its computations terminates. Classical LTLLTL\mathrm{LTL}roman_LTL is fit for the verification of such propositional trace properties, however some properties relevant in, for instance, information flow security are not trace properties. These properties profoundly speak of relations between traces. Clarkson and Schneider [3] coined the term hyperproperties to refer to such properties that lie beyond what LTLLTL\mathrm{LTL}roman_LTL can express. Bounded termination is an easy to grasp example of a hyperproperty: whether every computation of a system terminates within some bound common for all traces, cannot be determined by looking at traces in isolation. In information flow security, dependencies between public observable outputs and secret inputs constitute possible security breaches; checking for hyperproperties becomes invaluable. Two well-known examples of hyperproperties from this field are noninterference  [17, 14], where a high-level user cannot affect what low-level users see, and observational determinism [20], meaning that if two computations are in the same state according to a low-level observer, then the executions will be indistinguishable. However, hyperproperties are not limited to information flow security; examples from different fields include distributivity and other system properties such as fault tolerance [6].

Given this background, several approaches to formally specifying hyperproperties have been proposed since 2010, with families of logics emerging from these approaches. The two major streams in the research regarding capturing hyperproperties are hyperlogics and logics that employ team semantics. In the hyperlogics approach, logics that capture trace properties are extended with trace quantification, extending logics such as LTLLTL\mathrm{LTL}roman_LTL, computation tree logic (CTLCTL\mathrm{CTL}roman_CTL) or quantified propositional temporal logic (QPTLQPTL\mathrm{QPTL}roman_QPTL), into HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL [2], HyperCTLsuperscriptHyperCTL\mathrm{HyperCTL^{*}}roman_HyperCTL start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT [2], and HyperQPTLHyperQPTL\mathrm{HyperQPTL}roman_HyperQPTL [16, 4], respectively. An alternative approach is to lift the semantics of the temporal logics from being defined on traces to sets of traces, by using what is known as team semantics. This approach yields logics such as TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL [13, 8] and TeamCTLTeamCTL\mathrm{Team}\mathrm{CTL}roman_TeamCTL[12, 8]. Since its conception, TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL has been considered in two distinct variants: a synchronous semantics, where the team of traces agrees on the time step of occurrence when evaluating temporal operators; and an asynchronous semantics, where the temporal operators are evaluated independently on each trace. An example that illustrates the difference between these two semantics is the aforementioned termination and bound termination pair of properties. If we write 𝖥𝖥\operatorname{\mathsf{F}}sansserif_F for the future-operator and terminateterminate\mathrm{terminate}roman_terminate for a proposition symbol representing the trace terminating, we can write the formula 𝖥terminate𝖥terminate\operatorname{\mathsf{F}}\mathrm{terminate}sansserif_F roman_terminate, which under the synchronous semantics expresses the hyperproperty “bounded termination”, while under the asynchronous semantics the same formula defines the trace property “termination”. Not only is the above formulation of bounded termination clear and concise, it also illuminates a key difference between hyperlogics and team logics: while each formula of hyperlogic has a fixed number of quantifiers, which restricts the number of traces that can be referred to in a formula, which restricts the number of traces between which dependencies can be characterised by formulae, team logics have the ability to refer to an unbounded number of traces, even an infinite collection.

One of the original motivations behind team semantics [18] was to enable the definition of novel atomic formulae, and this is another important defining feature of team temporal logics as well. Among these atoms the dependence atom dep(x¯,y¯)dep¯𝑥¯𝑦\operatorname{\mathrm{dep}}(\bar{x},\bar{y})roman_dep ( over¯ start_ARG italic_x end_ARG , over¯ start_ARG italic_y end_ARG ) and inclusion atom x¯y¯¯𝑥¯𝑦\bar{x}\subseteq\bar{y}over¯ start_ARG italic_x end_ARG ⊆ over¯ start_ARG italic_y end_ARG stand out as the most influential. They respectively state that the variables y¯¯𝑦\bar{y}over¯ start_ARG italic_y end_ARG are functionally dependent on the variables x¯¯𝑥\bar{x}over¯ start_ARG italic_x end_ARG, and that the values of the variables x¯¯𝑥\bar{x}over¯ start_ARG italic_x end_ARG also occur among the values of variables y¯¯𝑦\bar{y}over¯ start_ARG italic_y end_ARG. As an example of the use of the inclusion atom, let the proposition symbols o1,,onsubscript𝑜1subscript𝑜𝑛o_{1},\dots,o_{n}italic_o start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , … , italic_o start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT denote public observable bits and assume that the proposition symbol s𝑠sitalic_s is a secret bit. The atomic formula (o1,on,s)(o1,on,¬s)subscript𝑜1subscript𝑜𝑛𝑠subscript𝑜1subscript𝑜𝑛𝑠(o_{1},\dots o_{n},s)\subseteq(o_{1},\dots o_{n},\neg s)( italic_o start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , … italic_o start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT , italic_s ) ⊆ ( italic_o start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , … italic_o start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT , ¬ italic_s ) expresses a form of non-inference by stating that an observer cannot infer the value of the confidential bit from the outputs.

While the expressivity of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL and other hyperlogics has been studied extensively, where the many extensions of TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL lie in relation the hyperlogics is still not completely understood. The connections for the logics without extensions were already established in Krebs et al. [13], where they showed that synchronous TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL and HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL are expressively incomparable and that the asynchronous variant collapses to LTLLTL\mathrm{LTL}roman_LTL. With regards to the expressivity of synchronous semantics, Virtema et al. [19] showed that the extensions of TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL can be translated to HyperQPTL+HyperQPTL+\mathrm{HyperQPTL\textsuperscript{\hskip-2.0pt\small+}}roman_HyperQPTL +, which in turn extends HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL with (non-uniform) quantification of propositions. Relating the logics to the first-order context, Kontinen and Sandström [10] defined Kamp-style translations from extensions of both semantics of TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL to the three-variable fragment of first-order team logic. It is worth noting that recently asynchronous hyperlogics have been considered also in several other articles (see, e.g., [9, 1]). An example of the significant rift between asynchronous and synchronous TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL is that the asynchronous semantics is essentially a first-order logic, while the synchronous semantics has second-order aspects. Especially the set-based variant of asynchronous TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL can be translated, using techniques in [10], into first-order logic under team semantics, which is known to be first-order logic [18]. Similarly, HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL is equally expressive as the guarded fragment of first-order logic with the equal level predicate, as was shown by Finkbeiner and Zimmermann [7].

In this article we focus on exploring the connections between fragments of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL and extensions of TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL. The set-based asynchronous semantics that we consider here was defined in Kontinen et al. [11] in order to further study the complexity of the model checking problem for these logics. Prior to that, the literature on temporal team semantics employed a semantics based on multisets of traces. In the wider team semantics literature, this often carries the name strict semantics, in contrast to lax semantics which is de facto a set-based semantics. This relaxation of the semantics enabled the definition of normal forms for the logics, which we use in this article to explore the connection with HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL.

Our contribution. We show correspondences in expressivity between the set-based variant of linear temporal logic under asynchronous team semantics and fragments of the Boolean closure of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL. In particular we show that LTLLTL\mathrm{LTL}roman_LTL under team semantics with the Boolean disjunction, TeamLTL(∨⃝)TeamLTL∨⃝\mathrm{TeamLTL}(\varovee)roman_TeamLTL ( ∨⃝ ), is equi-expressive with the positive Boolean closure of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL restricted to only one universal quantifier, while the left downward closed fragment of TeamLTL()TeamLTLsimilar-to\mathrm{TeamLTL}(\sim)roman_TeamLTL ( ∼ ) is equi-expressive with the Boolean closure of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL restricted to one universal quantifier.

2 Preliminaries

We begin by defining the variant of TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL and its extensions, as in [11].

Let APAP\mathrm{AP}roman_AP be a set of atomic propositions. The formulae of LTLLTL\mathrm{LTL}roman_LTL (over APAP\mathrm{AP}roman_AP) is attained by the grammar:

φ::=p¬pφφφφφ𝖦φφ𝖴φ,\varphi\mathrel{\mathop{{\mathop{:}}{\mathop{:}}}}=p\mid\neg p\mid\varphi\lor% \varphi\mid\varphi\land\varphi\mid\operatorname{\leavevmode\hbox to7.41pt{% \vbox to7.41pt{\pgfpicture\makeatletter\hbox{\hskip 3.70276pt\lower-1.11943pt% \hbox to0.0pt{\pgfsys@beginscope\pgfsys@invoke{ }\definecolor{pgfstrokecolor}{% rgb}{0,0,0}\pgfsys@color@rgb@stroke{0}{0}{0}\pgfsys@invoke{ }% \pgfsys@color@rgb@fill{0}{0}{0}\pgfsys@invoke{ }\pgfsys@setlinewidth{0.4pt}% \pgfsys@invoke{ }\nullfont\hbox to0.0pt{\pgfsys@beginscope\pgfsys@invoke{ }{ {{}{}}{{}}{}{{{}}{}{}{}{}{}{}{}{}}\pgfsys@beginscope\pgfsys@invoke{ }% \pgfsys@setlinewidth{0.51666pt}\pgfsys@invoke{ }{}\pgfsys@moveto{0.0pt}{2.5833% 3pt}\pgfsys@moveto{3.44443pt}{2.58333pt}\pgfsys@curveto{3.44443pt}{4.48566pt}{% 1.90233pt}{6.02776pt}{0.0pt}{6.02776pt}\pgfsys@curveto{-1.90233pt}{6.02776pt}{% -3.44443pt}{4.48566pt}{-3.44443pt}{2.58333pt}\pgfsys@curveto{-3.44443pt}{0.681% pt}{-1.90233pt}{-0.8611pt}{0.0pt}{-0.8611pt}\pgfsys@curveto{1.90233pt}{-0.8611% pt}{3.44443pt}{0.681pt}{3.44443pt}{2.58333pt}\pgfsys@closepath\pgfsys@moveto{0% .0pt}{2.58333pt}\pgfsys@stroke\pgfsys@invoke{ } \pgfsys@invoke{\lxSVG@closescope }\pgfsys@endscope} \pgfsys@invoke{\lxSVG@closescope }\pgfsys@endscope{}{}{}\hss}% \pgfsys@discardpath\pgfsys@invoke{\lxSVG@closescope }\pgfsys@endscope\hss}}% \lxSVG@closescope\endpgfpicture}}}\varphi\mid\operatorname{\mathsf{G}}\varphi% \mid\varphi\operatorname{\mathsf{U}}\varphi,italic_φ start_RELOP : : end_RELOP = italic_p ∣ ¬ italic_p ∣ italic_φ ∨ italic_φ ∣ italic_φ ∧ italic_φ ∣ OPFUNCTION italic_φ ∣ sansserif_G italic_φ ∣ italic_φ sansserif_U italic_φ ,

where pAP𝑝APp\in\mathrm{AP}italic_p ∈ roman_AP. We follow the convention that all formulae of TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL are given in negation normal form, where ¬\neg¬ is only allowed before atomic propositions, as is customary when dealing with team semantics.

We will consider the extensions of TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL with the Boolean disjunction ∨⃝∨⃝\varovee∨⃝, denoted TeamLTL(∨⃝)TeamLTL∨⃝\mathrm{TeamLTL}(\varovee)roman_TeamLTL ( ∨⃝ ), and Boolean negation similar-to\sim, denoted TeamLTL()TeamLTLsimilar-to\mathrm{TeamLTL}(\sim)roman_TeamLTL ( ∼ ).

A trace t𝑡titalic_t over APAP\mathrm{AP}roman_AP is an infinite sequence of sets of proposition symbols from (2AP)ωsuperscriptsuperscript2AP𝜔(2^{\mathrm{AP}})^{\omega}( 2 start_POSTSUPERSCRIPT roman_AP end_POSTSUPERSCRIPT ) start_POSTSUPERSCRIPT italic_ω end_POSTSUPERSCRIPT. Given a natural number i𝑖i\in\mathbb{N}italic_i ∈ blackboard_N, we denote by t[i]𝑡delimited-[]𝑖t[i]italic_t [ italic_i ] the (i+1)𝑖1(i+1)( italic_i + 1 )th element of t𝑡titalic_t and by t[i,]𝑡𝑖t[i,\infty]italic_t [ italic_i , ∞ ] the suffix (t[j])jisubscript𝑡delimited-[]𝑗𝑗𝑖(t[j])_{j\geq i}( italic_t [ italic_j ] ) start_POSTSUBSCRIPT italic_j ≥ italic_i end_POSTSUBSCRIPT of t𝑡titalic_t. We call a set of traces a team.

We write 𝒫()+𝒫superscript\mathcal{P}(\mathbb{N})^{+}caligraphic_P ( blackboard_N ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT to denote 𝒫(){}𝒫\mathcal{P}(\mathbb{N})\setminus\{\emptyset\}caligraphic_P ( blackboard_N ) ∖ { ∅ }. For a team T(2AP)ω𝑇superscriptsuperscript2AP𝜔T\subseteq(2^{\mathrm{AP}})^{\omega}italic_T ⊆ ( 2 start_POSTSUPERSCRIPT roman_AP end_POSTSUPERSCRIPT ) start_POSTSUPERSCRIPT italic_ω end_POSTSUPERSCRIPT a function f:T𝒫()+:𝑓𝑇𝒫superscriptf\colon T\rightarrow\mathcal{P}(\mathbb{N})^{+}italic_f : italic_T → caligraphic_P ( blackboard_N ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT, we set T[f,]:={t[s,]tT,sf(t)}T[f,\infty]\mathrel{\mathop{:}}=\{t[s,\infty]\mid t\in T,s\in f(t)\}italic_T [ italic_f , ∞ ] : = { italic_t [ italic_s , ∞ ] ∣ italic_t ∈ italic_T , italic_s ∈ italic_f ( italic_t ) }. For TTsuperscript𝑇𝑇T^{\prime}\subseteq Titalic_T start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT ⊆ italic_T, f:T𝒫()+:𝑓𝑇𝒫superscriptf\colon T\to\mathcal{P}(\mathbb{N})^{+}italic_f : italic_T → caligraphic_P ( blackboard_N ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT, and f:T𝒫()+:superscript𝑓superscript𝑇𝒫superscriptf^{\prime}\colon T^{\prime}\to\mathcal{P}(\mathbb{N})^{+}italic_f start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT : italic_T start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT → caligraphic_P ( blackboard_N ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT, we define that f<fsuperscript𝑓𝑓f^{\prime}<fitalic_f start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT < italic_f if and only if

tT::for-all𝑡superscript𝑇absent\displaystyle\forall t\in T^{\prime}:∀ italic_t ∈ italic_T start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT : min(f(t))min(f(t)) and,superscript𝑓𝑡𝑓𝑡 and,\displaystyle\min(f^{\prime}(t))\leq\min(f(t))\text{ and, }roman_min ( italic_f start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT ( italic_t ) ) ≤ roman_min ( italic_f ( italic_t ) ) and,
if max(f(t)) exists, max(f(t))<max(f(t)).if 𝑓𝑡 exists, superscript𝑓𝑡𝑓𝑡\displaystyle\text{ if }\max(f(t))\text{ exists, }\max(f^{\prime}(t))<\max(f(t% )).if roman_max ( italic_f ( italic_t ) ) exists, roman_max ( italic_f start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT ( italic_t ) ) < roman_max ( italic_f ( italic_t ) ) .
Definition 1 (TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL)

Let T𝑇Titalic_T be a team, and φ𝜑\varphiitalic_φ and ψ𝜓\psiitalic_ψ TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL-formulae. The lax semantics is defined as follows.

Tlmodels𝑇𝑙\displaystyle T\models litalic_T ⊧ italic_l \displaystyle\Leftrightarrow tl for all tT, where l{p,¬ppAP}formulae-sequencemodels𝑡𝑙 for all 𝑡𝑇 where l{p,¬ppAP}\displaystyle t\models l\text{ for all }t\in T,\text{ where $l\in\{p,\neg p% \mid p\in\mathrm{AP}\}$}italic_t ⊧ italic_l for all italic_t ∈ italic_T , where italic_l ∈ { italic_p , ¬ italic_p ∣ italic_p ∈ roman_AP }
is a literal and “t” refers to LTL-satisfactionmodelsis a literal and “𝑡” refers to LTL-satisfaction\displaystyle\text{ is a literal and ``}t\models\text{'' refers to $\mathrm{% LTL}$-satisfaction}is a literal and “ italic_t ⊧ ” refers to roman_LTL -satisfaction
Tφψmodels𝑇𝜑𝜓\displaystyle T\models\varphi\wedge\psiitalic_T ⊧ italic_φ ∧ italic_ψ \displaystyle\Leftrightarrow Tφ and Tψmodels𝑇𝜑 and 𝑇models𝜓\displaystyle T\models\varphi\text{ and }T\models\psiitalic_T ⊧ italic_φ and italic_T ⊧ italic_ψ
Tφψmodels𝑇𝜑𝜓\displaystyle T\models\varphi\vee\psiitalic_T ⊧ italic_φ ∨ italic_ψ \displaystyle\Leftrightarrow T1,T2 s.t. T1T2=T and T1φ and T2ψsubscript𝑇1subscript𝑇2 s.t. subscript𝑇1subscript𝑇2𝑇 and subscript𝑇1models𝜑 and subscript𝑇2models𝜓\displaystyle\exists T_{1},T_{2}\text{ s.t. }T_{1}\cup T_{2}=T\text{ and }T_{1% }\models\varphi\text{ and }T_{2}\models\psi∃ italic_T start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , italic_T start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT s.t. italic_T start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ∪ italic_T start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT = italic_T and italic_T start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ⊧ italic_φ and italic_T start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ⊧ italic_ψ
Tφmodels𝑇𝜑\displaystyle T\models\operatorname{\leavevmode\hbox to7.41pt{\vbox to7.41pt{% \pgfpicture\makeatletter\hbox{\hskip 3.70276pt\lower-1.11943pt\hbox to0.0pt{% \pgfsys@beginscope\pgfsys@invoke{ }\definecolor{pgfstrokecolor}{rgb}{0,0,0}% \pgfsys@color@rgb@stroke{0}{0}{0}\pgfsys@invoke{ }\pgfsys@color@rgb@fill{0}{0}% {0}\pgfsys@invoke{ }\pgfsys@setlinewidth{0.4pt}\pgfsys@invoke{ }\nullfont\hbox to% 0.0pt{\pgfsys@beginscope\pgfsys@invoke{ }{ {{}{}}{{}}{}{{{}}{}{}{}{}{}{}{}{}}\pgfsys@beginscope\pgfsys@invoke{ }% \pgfsys@setlinewidth{0.51666pt}\pgfsys@invoke{ }{}\pgfsys@moveto{0.0pt}{2.5833% 3pt}\pgfsys@moveto{3.44443pt}{2.58333pt}\pgfsys@curveto{3.44443pt}{4.48566pt}{% 1.90233pt}{6.02776pt}{0.0pt}{6.02776pt}\pgfsys@curveto{-1.90233pt}{6.02776pt}{% -3.44443pt}{4.48566pt}{-3.44443pt}{2.58333pt}\pgfsys@curveto{-3.44443pt}{0.681% pt}{-1.90233pt}{-0.8611pt}{0.0pt}{-0.8611pt}\pgfsys@curveto{1.90233pt}{-0.8611% pt}{3.44443pt}{0.681pt}{3.44443pt}{2.58333pt}\pgfsys@closepath\pgfsys@moveto{0% .0pt}{2.58333pt}\pgfsys@stroke\pgfsys@invoke{ } \pgfsys@invoke{\lxSVG@closescope }\pgfsys@endscope} \pgfsys@invoke{\lxSVG@closescope }\pgfsys@endscope{}{}{}\hss}% \pgfsys@discardpath\pgfsys@invoke{\lxSVG@closescope }\pgfsys@endscope\hss}}% \lxSVG@closescope\endpgfpicture}}}\varphiitalic_T ⊧ OPFUNCTION italic_φ \displaystyle\Leftrightarrow T[1,]φmodels𝑇1𝜑\displaystyle T[1,\infty]\models\varphiitalic_T [ 1 , ∞ ] ⊧ italic_φ
T𝖦φmodels𝑇𝖦𝜑\displaystyle T\models\operatorname{\mathsf{G}}\varphiitalic_T ⊧ sansserif_G italic_φ \displaystyle\Leftrightarrow f:T𝒫()+ it holds that T[f,]φ:for-all𝑓𝑇𝒫superscript it holds that 𝑇𝑓models𝜑\displaystyle\forall f\colon T\rightarrow\mathcal{P}(\mathbb{N})^{+}\text{ it % holds that }T[f,\infty]\models\varphi∀ italic_f : italic_T → caligraphic_P ( blackboard_N ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT it holds that italic_T [ italic_f , ∞ ] ⊧ italic_φ
Tφ𝖴ψmodels𝑇𝜑𝖴𝜓\displaystyle T\models\varphi\operatorname{\mathsf{U}}\psiitalic_T ⊧ italic_φ sansserif_U italic_ψ \displaystyle\Leftrightarrow f:T𝒫()+ such that T[f,]ψ and:𝑓𝑇𝒫superscript such that 𝑇𝑓models𝜓 and\displaystyle\exists f\colon T\rightarrow\mathcal{P}(\mathbb{N})^{+}\text{ % such that }T[f,\infty]\models\psi\mbox{ and }∃ italic_f : italic_T → caligraphic_P ( blackboard_N ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT such that italic_T [ italic_f , ∞ ] ⊧ italic_ψ and
f:T𝒫()+s.t. f<f, it holds that T[f,]φ:for-allsuperscript𝑓superscript𝑇𝒫superscripts.t. f<f, it holds that superscript𝑇superscript𝑓models𝜑\displaystyle\forall f^{\prime}\colon T^{\prime}\rightarrow\mathcal{P}(\mathbb% {N})^{+}\text{s.t. $f^{\prime}<f$, it holds that }T^{\prime}[f^{\prime},\infty% ]\models\varphi∀ italic_f start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT : italic_T start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT → caligraphic_P ( blackboard_N ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT s.t. italic_f start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT < italic_f , it holds that italic_T start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT [ italic_f start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT , ∞ ] ⊧ italic_φ
or T=, where T:={tTmax(f(t))0}\displaystyle\text{ or }T^{\prime}=\emptyset,\text{ where }T^{\prime}\mathrel{% \mathop{:}}=\{t\in T\mid\max(f(t))\neq 0\}or italic_T start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT = ∅ , where italic_T start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT : = { italic_t ∈ italic_T ∣ roman_max ( italic_f ( italic_t ) ) ≠ 0 }

The semantics for the Boolean disjunction and Boolean negation, used in the extensions TeamLTL(∨⃝)TeamLTL∨⃝\mathrm{TeamLTL}(\varovee)roman_TeamLTL ( ∨⃝ ) and TeamLTL()TeamLTLsimilar-to\mathrm{TeamLTL}(\sim)roman_TeamLTL ( ∼ ), are given by:

Tφ∨⃝ψmodels𝑇𝜑∨⃝𝜓\displaystyle T\models\varphi\varovee\psiitalic_T ⊧ italic_φ ∨⃝ italic_ψ \displaystyle\Leftrightarrow Tφ or Tψmodels𝑇𝜑 or 𝑇models𝜓\displaystyle T\models\varphi\text{ or }T\models\psiitalic_T ⊧ italic_φ or italic_T ⊧ italic_ψ
Tφ\displaystyle T\models\sim\varphiitalic_T ⊧ ∼ italic_φ \displaystyle\Leftrightarrow T⊧̸φnot-models𝑇𝜑\displaystyle T\not\models\varphiitalic_T ⊧̸ italic_φ

Note that the Boolean disjunction is definable in TeamLTL()TeamLTLsimilar-to\mathrm{TeamLTL}(\sim)roman_TeamLTL ( ∼ ), as the dual of conjunction, i.e. Tlφ∨⃝ψsuperscriptmodels𝑙𝑇𝜑∨⃝𝜓T\mathrel{\models^{l}}\varphi\varovee\psiitalic_T start_RELOP ⊧ start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT end_RELOP italic_φ ∨⃝ italic_ψ if and only if Tl(φψ)T\mathrel{\models^{l}}\sim(\sim\varphi\wedge\sim\psi)italic_T start_RELOP ⊧ start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT end_RELOP ∼ ( ∼ italic_φ ∧ ∼ italic_ψ ).

Two important properties of team logics are flatness and downward closure. A logic has the flatness property if Tlφsuperscriptmodels𝑙𝑇𝜑T\mathrel{\models^{l}}\varphiitalic_T start_RELOP ⊧ start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT end_RELOP italic_φ if and only if {t}lφsuperscriptmodels𝑙𝑡𝜑\{t\}\mathrel{\models^{l}}\varphi{ italic_t } start_RELOP ⊧ start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT end_RELOP italic_φ for all tT𝑡𝑇t\in Titalic_t ∈ italic_T, holds for all formulae φ𝜑\varphiitalic_φ of the logic. A logic is downward closed if for all formulae φ𝜑\varphiitalic_φ of the logic if Tlφsuperscriptmodels𝑙𝑇𝜑T\mathrel{\models^{l}}\varphiitalic_T start_RELOP ⊧ start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT end_RELOP italic_φ and ST𝑆𝑇S\subseteq Titalic_S ⊆ italic_T then Slφsuperscriptmodels𝑙𝑆𝜑S\mathrel{\models^{l}}\varphiitalic_S start_RELOP ⊧ start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT end_RELOP italic_φ. The following Proposition was proven in [11].

Proposition 1

TeamLTLlsuperscriptTeamLTL𝑙\mathrm{TeamLTL}^{l}roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT has both the flatness and the downward closure properties, while TeamLTLl(∨⃝)superscriptTeamLTL𝑙∨⃝\mathrm{TeamLTL}^{l}(\varovee)roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∨⃝ ) only has the downward closure property.

We consider the left-downward closed fragment of TeamLTLl()superscriptTeamLTL𝑙similar-to\mathrm{TeamLTL}^{l}(\sim)roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∼ ), denoted left-dc–TeamLTLl()left-dc–superscriptTeamLTL𝑙similar-to\text{left-dc--}\mathrm{TeamLTL}^{l}(\sim)left-dc– roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∼ ), where every subformula of the form 𝖦ψ𝖦𝜓\operatorname{\mathsf{G}}\psisansserif_G italic_ψ or ψ𝖴θ𝜓𝖴𝜃\psi\operatorname{\mathsf{U}}\thetaitalic_ψ sansserif_U italic_θ, the subformula ψ𝜓\psiitalic_ψ is a TeamLTL(∨⃝)TeamLTL∨⃝\mathrm{TeamLTL}(\varovee)roman_TeamLTL ( ∨⃝ )-formula

It was established in [11] that any formula of TeamLTLl(∨⃝)superscriptTeamLTL𝑙∨⃝\mathrm{TeamLTL}^{l}(\varovee)roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∨⃝ ) can be equivalently expressed in ∨⃝∨⃝\varovee∨⃝-disjunctive normal form, i.e. in the form

\scalerel∨⃝iIαi,\operatorname*{\scalerel*{\ovee}{\sum}}_{i\in I}\alpha_{i},start_OPERATOR ∗ ∨⃝ ∑ end_OPERATOR start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ,

where αisubscript𝛼𝑖\alpha_{i}italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT are LTLLTL\mathrm{LTL}roman_LTL-formulae.

Similarly by [11], every formula of left-dc–TeamLTLl()left-dc–superscriptTeamLTL𝑙similar-to\text{left-dc--}\mathrm{TeamLTL}^{l}(\sim)left-dc– roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∼ ) can be equivalently stated in quasi-flat normal form, which means in the form

\scalerel∨⃝iI(αijJiβi,j),\operatorname*{\scalerel*{\ovee}{\sum}}_{i\in I}(\alpha_{i}\wedge\bigwedge_{j% \in J_{i}}\exists\beta_{i,j}),start_OPERATOR ∗ ∨⃝ ∑ end_OPERATOR start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ( italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ∧ ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_POSTSUBSCRIPT ∃ italic_β start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ) ,

where αisubscript𝛼𝑖\alpha_{i}italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT and βi,jsubscript𝛽𝑖𝑗\beta_{i,j}italic_β start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT are LTLLTL\mathrm{LTL}roman_LTL-formulae, and βi,jsubscript𝛽𝑖𝑗\exists\beta_{i,j}∃ italic_β start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT is an abbreviation for the formula βi,jdsimilar-toabsentsubscriptsuperscript𝛽𝑑𝑖𝑗\sim\beta^{d}_{i,j}∼ italic_β start_POSTSUPERSCRIPT italic_d end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT, where βi,jdsubscriptsuperscript𝛽𝑑𝑖𝑗\beta^{d}_{i,j}italic_β start_POSTSUPERSCRIPT italic_d end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT is the formula obtained from ¬β𝛽\neg\beta¬ italic_β, after ¬\neg¬ has been pushed down to the atomic level.

Next we state the syntax and semantics of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL, as defined in [2], as well as the Boolean closure concepts we are concerned with.

Definition 2 (Syntax of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL)

Let APAP\mathrm{AP}roman_AP be a set of propositional variables and 𝒱𝒱\mathcal{V}caligraphic_V the set of all trace variables. Formulas of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL are generated by the following grammar:

ψ𝜓\displaystyle\psiitalic_ψ ::=::absent\displaystyle\mathrel{\mathop{{\mathop{:}}{\mathop{:}}}}=start_RELOP : : end_RELOP = π.ψformulae-sequence𝜋𝜓\displaystyle\exists\pi.\psi∃ italic_π . italic_ψ \displaystyle\mid π.ψformulae-sequencefor-all𝜋𝜓\displaystyle\forall\pi.\psi∀ italic_π . italic_ψ \displaystyle\mid φ𝜑\displaystyle\varphiitalic_φ
φ𝜑\displaystyle\varphiitalic_φ ::=::absent\displaystyle\mathrel{\mathop{{\mathop{:}}{\mathop{:}}}}=start_RELOP : : end_RELOP = aπsubscript𝑎𝜋\displaystyle a_{\pi}italic_a start_POSTSUBSCRIPT italic_π end_POSTSUBSCRIPT \displaystyle\mid ¬φ𝜑\displaystyle\neg\varphi¬ italic_φ \displaystyle\mid φφ𝜑𝜑\displaystyle\varphi\vee\varphiitalic_φ ∨ italic_φ \displaystyle\mid φ𝜑\displaystyle\operatorname{\leavevmode\hbox to7.41pt{\vbox to7.41pt{% \pgfpicture\makeatletter\hbox{\hskip 3.70276pt\lower-1.11943pt\hbox to0.0pt{% \pgfsys@beginscope\pgfsys@invoke{ }\definecolor{pgfstrokecolor}{rgb}{0,0,0}% \pgfsys@color@rgb@stroke{0}{0}{0}\pgfsys@invoke{ }\pgfsys@color@rgb@fill{0}{0}% {0}\pgfsys@invoke{ }\pgfsys@setlinewidth{0.4pt}\pgfsys@invoke{ }\nullfont\hbox to% 0.0pt{\pgfsys@beginscope\pgfsys@invoke{ }{ {{}{}}{{}}{}{{{}}{}{}{}{}{}{}{}{}}\pgfsys@beginscope\pgfsys@invoke{ }% \pgfsys@setlinewidth{0.51666pt}\pgfsys@invoke{ }{}\pgfsys@moveto{0.0pt}{2.5833% 3pt}\pgfsys@moveto{3.44443pt}{2.58333pt}\pgfsys@curveto{3.44443pt}{4.48566pt}{% 1.90233pt}{6.02776pt}{0.0pt}{6.02776pt}\pgfsys@curveto{-1.90233pt}{6.02776pt}{% -3.44443pt}{4.48566pt}{-3.44443pt}{2.58333pt}\pgfsys@curveto{-3.44443pt}{0.681% pt}{-1.90233pt}{-0.8611pt}{0.0pt}{-0.8611pt}\pgfsys@curveto{1.90233pt}{-0.8611% pt}{3.44443pt}{0.681pt}{3.44443pt}{2.58333pt}\pgfsys@closepath\pgfsys@moveto{0% .0pt}{2.58333pt}\pgfsys@stroke\pgfsys@invoke{ } \pgfsys@invoke{\lxSVG@closescope }\pgfsys@endscope} \pgfsys@invoke{\lxSVG@closescope }\pgfsys@endscope{}{}{}\hss}% \pgfsys@discardpath\pgfsys@invoke{\lxSVG@closescope }\pgfsys@endscope\hss}}% \lxSVG@closescope\endpgfpicture}}}\varphiOPFUNCTION italic_φ \displaystyle\mid φ𝖴φ,𝜑𝖴𝜑\displaystyle\varphi\operatorname{\mathsf{U}}\varphi,italic_φ sansserif_U italic_φ ,

where aAP𝑎APa\in\mathrm{AP}italic_a ∈ roman_AP and π𝒱𝜋𝒱\pi\in\mathcal{V}italic_π ∈ caligraphic_V.

We denote the set of all traces by TRTR\mathrm{TR}roman_TR and the set of all trace variables by 𝒱𝒱\mathcal{V}caligraphic_V. For a trace assignment function Π:𝒱TR:Π𝒱TR\Pi\colon\mathcal{V}\to\mathrm{TR}roman_Π : caligraphic_V → roman_TR, we write Π[i,]Π𝑖\Pi[i,\infty]roman_Π [ italic_i , ∞ ] for the trace assignment defined through Π[i,]=Π(π)[i,]Π𝑖Π𝜋𝑖\Pi[i,\infty]=\Pi(\pi)[i,\infty]roman_Π [ italic_i , ∞ ] = roman_Π ( italic_π ) [ italic_i , ∞ ], and Π[πt]Πdelimited-[]maps-to𝜋𝑡\Pi[\pi\mapsto t]roman_Π [ italic_π ↦ italic_t ] for the assignment that assigns t𝑡titalic_t to π𝜋\piitalic_π, but otherwise is identical to ΠΠ\Piroman_Π.

Definition 3 (Semantics of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL)

Let aAP𝑎APa\in\mathrm{AP}italic_a ∈ roman_AP be a proposition symbol, π𝒱𝜋𝒱\pi\in\mathcal{V}italic_π ∈ caligraphic_V be a trace variable, T𝑇Titalic_T be a set of traces, and let Π:𝒱TR:Π𝒱TR\Pi\colon\mathcal{V}\to\mathrm{TR}roman_Π : caligraphic_V → roman_TR be a trace assignment.

ΠTπ.ψformulae-sequencesubscriptmodels𝑇Π𝜋𝜓\displaystyle\Pi\models_{T}\exists\pi.\psiroman_Π ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ∃ italic_π . italic_ψ \displaystyle\Leftrightarrow there exists tT:Π[πt]Tψ:there exists 𝑡𝑇subscriptmodels𝑇Πdelimited-[]maps-to𝜋𝑡𝜓\displaystyle\text{there exists }t\in T\colon\Pi[\pi\mapsto t]\models_{T}\psithere exists italic_t ∈ italic_T : roman_Π [ italic_π ↦ italic_t ] ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_ψ
ΠTπ.ψformulae-sequencesubscriptmodels𝑇Πfor-all𝜋𝜓\displaystyle\Pi\models_{T}\forall\pi.\psiroman_Π ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ∀ italic_π . italic_ψ \displaystyle\Leftrightarrow for all tT:Π[πt]Tψ:for all 𝑡𝑇subscriptmodels𝑇Πdelimited-[]maps-to𝜋𝑡𝜓\displaystyle\text{for all }t\in T\colon\Pi[\pi\mapsto t]\models_{T}\psifor all italic_t ∈ italic_T : roman_Π [ italic_π ↦ italic_t ] ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_ψ
ΠTaπsubscriptmodels𝑇Πsubscript𝑎𝜋\displaystyle\Pi\models_{T}a_{\pi}roman_Π ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_a start_POSTSUBSCRIPT italic_π end_POSTSUBSCRIPT \displaystyle\Leftrightarrow aΠ(π)[0]𝑎Π𝜋delimited-[]0\displaystyle a\in\Pi(\pi)[0]italic_a ∈ roman_Π ( italic_π ) [ 0 ]
ΠT¬φsubscriptmodels𝑇Π𝜑\displaystyle\Pi\models_{T}\neg\varphiroman_Π ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ¬ italic_φ \displaystyle\Leftrightarrow Π⊧̸Tφsubscriptnot-models𝑇Π𝜑\displaystyle\Pi\not\models_{T}\varphiroman_Π ⊧̸ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_φ
ΠTφ1φ2subscriptmodels𝑇Πsubscript𝜑1subscript𝜑2\displaystyle\Pi\models_{T}\varphi_{1}\vee\varphi_{2}roman_Π ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ∨ italic_φ start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT \displaystyle\Leftrightarrow ΠTφ1 or ΠTφ2subscriptmodels𝑇Πsubscript𝜑1 or Πsubscriptmodels𝑇subscript𝜑2\displaystyle\Pi\models_{T}\varphi_{1}\text{ or }\Pi\models_{T}\varphi_{2}roman_Π ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT or roman_Π ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT
ΠTφsubscriptmodels𝑇Π𝜑\displaystyle\Pi\models_{T}\operatorname{\leavevmode\hbox to7.41pt{\vbox to% 7.41pt{\pgfpicture\makeatletter\hbox{\hskip 3.70276pt\lower-1.11943pt\hbox to0% .0pt{\pgfsys@beginscope\pgfsys@invoke{ }\definecolor{pgfstrokecolor}{rgb}{% 0,0,0}\pgfsys@color@rgb@stroke{0}{0}{0}\pgfsys@invoke{ }\pgfsys@color@rgb@fill% {0}{0}{0}\pgfsys@invoke{ }\pgfsys@setlinewidth{0.4pt}\pgfsys@invoke{ }% \nullfont\hbox to0.0pt{\pgfsys@beginscope\pgfsys@invoke{ }{ {{}{}}{{}}{}{{{}}{}{}{}{}{}{}{}{}}\pgfsys@beginscope\pgfsys@invoke{ }% \pgfsys@setlinewidth{0.51666pt}\pgfsys@invoke{ }{}\pgfsys@moveto{0.0pt}{2.5833% 3pt}\pgfsys@moveto{3.44443pt}{2.58333pt}\pgfsys@curveto{3.44443pt}{4.48566pt}{% 1.90233pt}{6.02776pt}{0.0pt}{6.02776pt}\pgfsys@curveto{-1.90233pt}{6.02776pt}{% -3.44443pt}{4.48566pt}{-3.44443pt}{2.58333pt}\pgfsys@curveto{-3.44443pt}{0.681% pt}{-1.90233pt}{-0.8611pt}{0.0pt}{-0.8611pt}\pgfsys@curveto{1.90233pt}{-0.8611% pt}{3.44443pt}{0.681pt}{3.44443pt}{2.58333pt}\pgfsys@closepath\pgfsys@moveto{0% .0pt}{2.58333pt}\pgfsys@stroke\pgfsys@invoke{ } \pgfsys@invoke{\lxSVG@closescope }\pgfsys@endscope} \pgfsys@invoke{\lxSVG@closescope }\pgfsys@endscope{}{}{}\hss}% \pgfsys@discardpath\pgfsys@invoke{\lxSVG@closescope }\pgfsys@endscope\hss}}% \lxSVG@closescope\endpgfpicture}}}\varphiroman_Π ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT OPFUNCTION italic_φ \displaystyle\Leftrightarrow Π[1,]Tφsubscriptmodels𝑇Π1𝜑\displaystyle\Pi[1,\infty]\models_{T}\varphiroman_Π [ 1 , ∞ ] ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_φ
ΠTφ1𝖴φ2subscriptmodels𝑇Πsubscript𝜑1𝖴subscript𝜑2\displaystyle\Pi\models_{T}\varphi_{1}\operatorname{\mathsf{U}}\varphi_{2}roman_Π ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT sansserif_U italic_φ start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT \displaystyle\Leftrightarrow there exists i0:Π[i,]Tφ2:there exists 𝑖0subscriptmodels𝑇Π𝑖subscript𝜑2\displaystyle\text{there exists }i\geq 0\colon\Pi[i,\infty]\models_{T}\varphi_% {2}there exists italic_i ≥ 0 : roman_Π [ italic_i , ∞ ] ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT
 and for all 0j<i we have Π[j,]Tφ1 and for all 0𝑗𝑖 we have Π𝑗subscriptmodels𝑇subscript𝜑1\displaystyle\quad\text{ and for all }0\leq j<i\text{ we have }\Pi[j,\infty]% \models_{T}\varphi_{1}and for all 0 ≤ italic_j < italic_i we have roman_Π [ italic_j , ∞ ] ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT
Definition 4 (Universal Fragments)

The universal fragment of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL, denoted by HyperLTLsuperscriptfor-allHyperLTL\forall^{*}\mathrm{HyperLTL}∀ start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT roman_HyperLTL, is the fragment of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL with no existential quantification. We write HyperLTLfor-allHyperLTL\forall\mathrm{HyperLTL}∀ roman_HyperLTL for the one variable universal fragment of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL, and QHyperLTL𝑄HyperLTLQ\mathrm{HyperLTL}italic_Q roman_HyperLTL for the one variable fragment of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL.

Definition 5 ((Positive) Boolean Closure)

The Boolean closure of a logic \mathcal{L}caligraphic_L, denoted by BC()BC\mathrm{BC}(\mathcal{L})roman_BC ( caligraphic_L ), is the extension of \mathcal{L}caligraphic_L that is closed under \wedge, \vee and ¬\neg¬. The positive Boolean closure of a logic \mathcal{L}caligraphic_L, denoted by PBC()PBC\mathrm{PBC}(\mathcal{L})roman_PBC ( caligraphic_L ), is the extension of \mathcal{L}caligraphic_L that is closed under \wedge and \vee.

The semantics for the Boolean closures are attained by relaxing the definition of conjunction \wedge, disjunction \vee, and ¬\neg¬ to apply to any formula of the Boolean closure.

Using a suitable algorithm, all BC()BC\mathrm{BC}(\mathcal{L})roman_BC ( caligraphic_L )-formulae can be equivalently expressed in disjunctive normal form, i.e. as a disjunction of conjunctions with possibly a negation in front of each formula of \mathcal{L}caligraphic_L. Similarly, all PBC()PBC\mathrm{PBC}(\mathcal{L})roman_PBC ( caligraphic_L )-formulae can be equivalently expressed as

iIjJφi,jsubscript𝑖𝐼subscript𝑗𝐽subscript𝜑𝑖𝑗\bigvee_{i\in I}\bigwedge_{j\in J}\varphi_{i,j}⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT

for some formulae φi,jsubscript𝜑𝑖𝑗\varphi_{i,j}\in\mathcal{L}italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ∈ caligraphic_L and index sets I𝐼Iitalic_I and J𝐽Jitalic_J. From here on we use I𝐼Iitalic_I and J𝐽Jitalic_J to denote arbitrary index sets.

3 Correspondence between TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL and HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL

In this section we will explore the relationship between the logics by proving some correspondence theorems. First, however, we prove some pertinent propositions regarding the Boolean closure of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL, showing that conjunction, disjunction and negation distribute over the quantifiers in a manner analogous to first-order logic. We go through these propositions in some detail as, although they appear familiar from the first-order setting, HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL is usually considered only in the prenex normal form and thus these basic results are not explicitly addressed in the literature. Moreover, the proofs feature arguments that will be useful in subsequent proofs.

As usual, for logics \mathcal{L}caligraphic_L and superscript\mathcal{L}^{\prime}caligraphic_L start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT, we write superscript\mathcal{L}\leq\mathcal{L}^{\prime}caligraphic_L ≤ caligraphic_L start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT, if for every \mathcal{L}caligraphic_L-formula there exists an equivalent superscript\mathcal{L}^{\prime}caligraphic_L start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT-formula. We write superscript\mathcal{L}\equiv\mathcal{L}^{\prime}caligraphic_L ≡ caligraphic_L start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT, if both superscript\mathcal{L}\leq\mathcal{L}^{\prime}caligraphic_L ≤ caligraphic_L start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT and superscript\mathcal{L}^{\prime}\leq\mathcal{L}caligraphic_L start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT ≤ caligraphic_L.

Proposition 2

PBC(HyperLTL)HyperLTLPBCsuperscriptfor-allHyperLTLsuperscriptfor-allHyperLTL\mathrm{PBC}(\forall^{*}\mathrm{HyperLTL})\equiv\forall^{*}\mathrm{HyperLTL}roman_PBC ( ∀ start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT roman_HyperLTL ) ≡ ∀ start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT roman_HyperLTL

Proof

Let iIjJψi,jsubscript𝑖𝐼subscript𝑗𝐽subscript𝜓𝑖𝑗\bigvee_{i\in I}\bigwedge_{j\in J}\psi_{i,j}⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_ψ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT be an arbitrary formula of PBC(HyperLTL)PBCsuperscriptfor-allHyperLTL\mathrm{PBC}(\forall^{*}\mathrm{HyperLTL})roman_PBC ( ∀ start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT roman_HyperLTL ). If all ψi,jsubscript𝜓𝑖𝑗\psi_{i,j}italic_ψ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT are quantifier free, we are done, as then iIjJψi,jsubscript𝑖𝐼subscript𝑗𝐽subscript𝜓𝑖𝑗\bigvee_{i\in I}\bigwedge_{j\in J}\psi_{i,j}⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_ψ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT is a HyperLTLsuperscriptfor-allHyperLTL\forall^{*}\mathrm{HyperLTL}∀ start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT roman_HyperLTL-formula. Thus, we may assume that ψi,j=π1πnφi,jsubscript𝜓𝑖𝑗for-allsubscript𝜋1for-allsubscript𝜋𝑛subscript𝜑𝑖𝑗\psi_{i,j}=\forall\pi_{1}\cdots\forall\pi_{n}\varphi_{i,j}italic_ψ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT = ∀ italic_π start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ⋯ ∀ italic_π start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT for some LTLLTL\mathrm{LTL}roman_LTL-formula φi,jsubscript𝜑𝑖𝑗\varphi_{i,j}italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT. Suppose

ΠTiIjJπ1πnφi,j.subscriptmodels𝑇Πsubscript𝑖𝐼subscript𝑗𝐽for-allsubscript𝜋1for-allsubscript𝜋𝑛subscript𝜑𝑖𝑗\Pi\models_{T}\bigvee_{i\in I}\bigwedge_{j\in J}\forall\pi_{1}\cdots\forall\pi% _{n}\varphi_{i,j}.roman_Π ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT ∀ italic_π start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ⋯ ∀ italic_π start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT .

Without loss of generality, we may assume a uniform quantifier block in each conjunct, as one can rename variables and take the largest quantifier block as the common one, since redundant quantifiers do not effect evaluation. The previous is therefore equivalent with

ΠTiIπ1πnjJφi,j.subscriptmodels𝑇Πsubscript𝑖𝐼for-allsubscript𝜋1for-allsubscript𝜋𝑛subscript𝑗𝐽subscript𝜑𝑖𝑗\Pi\models_{T}\bigvee_{i\in I}\forall\pi_{1}\cdots\forall\pi_{n}\bigwedge_{j% \in J}\varphi_{i,j}.roman_Π ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ∀ italic_π start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ⋯ ∀ italic_π start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT .

At this point, we wish to push the disjunction past the quantifier block, but the variables would become entangled and different traces could satisfy different disjuncts. We need to distinguish the variables of the disjuncts from each other, so we rename the trace quantifiers. The previous evaluation is therefore equivalent with

ΠTπ11π1iπn1πniiIjJφi,j(π11,,πni).subscriptmodels𝑇Πfor-allsuperscriptsubscript𝜋11for-allsuperscriptsubscript𝜋1𝑖for-allsuperscriptsubscript𝜋𝑛1for-allsuperscriptsubscript𝜋𝑛𝑖subscript𝑖𝐼subscript𝑗𝐽subscript𝜑𝑖𝑗superscriptsubscript𝜋11superscriptsubscript𝜋𝑛𝑖\Pi\models_{T}\forall\pi_{1}^{1}\cdots\forall\pi_{1}^{i}\cdots\forall\pi_{n}^{% 1}\cdots\forall\pi_{n}^{i}\bigvee_{i\in I}\bigwedge_{j\in J}\varphi_{i,j}(\pi_% {1}^{1},\cdots,\pi_{n}^{i}).roman_Π ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ∀ italic_π start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 1 end_POSTSUPERSCRIPT ⋯ ∀ italic_π start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_i end_POSTSUPERSCRIPT ⋯ ∀ italic_π start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 1 end_POSTSUPERSCRIPT ⋯ ∀ italic_π start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_i end_POSTSUPERSCRIPT ⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ( italic_π start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 1 end_POSTSUPERSCRIPT , ⋯ , italic_π start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_i end_POSTSUPERSCRIPT ) .

This is a formula of HyperLTLsuperscriptfor-allHyperLTL\forall^{*}\mathrm{HyperLTL}∀ start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT roman_HyperLTL. ∎

The following remark, familiar from first-order logics, can be proven with a straight-forward induction over the length of the quantifier block.

Remark 1

For HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL-formula Q1π1Qnπnψsubscript𝑄1subscript𝜋1subscript𝑄𝑛subscript𝜋𝑛𝜓Q_{1}\pi_{1}\cdots Q_{n}\pi_{n}\psiitalic_Q start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT italic_π start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ⋯ italic_Q start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT italic_π start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT italic_ψ it holds that

¬Q1π1QnπnψQ1π1Qnπn¬ψ,subscript𝑄1subscript𝜋1subscript𝑄𝑛subscript𝜋𝑛𝜓superscriptsubscript𝑄1subscript𝜋1superscriptsubscript𝑄𝑛subscript𝜋𝑛𝜓\neg Q_{1}\pi_{1}\cdots Q_{n}\pi_{n}\psi\equiv Q_{1}^{-}\pi_{1}\cdots Q_{n}^{-% }\pi_{n}\neg\psi,¬ italic_Q start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT italic_π start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ⋯ italic_Q start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT italic_π start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT italic_ψ ≡ italic_Q start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT italic_π start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ⋯ italic_Q start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT italic_π start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT ¬ italic_ψ ,

where for every index i𝑖iitalic_i, Qisubscript𝑄𝑖Q_{i}italic_Q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT are quantifiers for-all\forall or \exists, and Qisuperscriptsubscript𝑄𝑖Q_{i}^{-}italic_Q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT is \exists if Qisubscript𝑄𝑖Q_{i}italic_Q start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT is for-all\forall and vice versa.

Proposition 3

BC(HyperLTL)HyperLTLBCHyperLTLHyperLTL\mathrm{BC}(\mathrm{HyperLTL})\equiv\mathrm{HyperLTL}roman_BC ( roman_HyperLTL ) ≡ roman_HyperLTL

Proof

Consider a BC(HyperLTL)BCHyperLTL\mathrm{BC}(\mathrm{HyperLTL})roman_BC ( roman_HyperLTL )-formula iIjJφi,jsubscript𝑖𝐼subscript𝑗𝐽subscript𝜑𝑖𝑗\bigvee_{i\in I}\bigwedge_{j\in J}\varphi_{i,j}⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT in disjunctive normal form, with either φi,jHyperLTLsubscript𝜑𝑖𝑗HyperLTL\varphi_{i,j}\in\mathrm{HyperLTL}italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ∈ roman_HyperLTL or φi,j=¬ψi,jsubscript𝜑𝑖𝑗subscript𝜓𝑖𝑗\varphi_{i,j}=\neg\psi_{i,j}italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT = ¬ italic_ψ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT for some formula ψi,jHyperLTLsubscript𝜓𝑖𝑗HyperLTL\psi_{i,j}\in\mathrm{HyperLTL}italic_ψ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ∈ roman_HyperLTL. By Remark 1 ¬ψi,jQ1i,jπ1i,jQni,jπni,jθi,jsubscript𝜓𝑖𝑗superscriptsubscript𝑄1𝑖𝑗superscriptsubscript𝜋1𝑖𝑗superscriptsubscript𝑄𝑛𝑖𝑗superscriptsubscript𝜋𝑛𝑖𝑗subscript𝜃𝑖𝑗\neg\psi_{i,j}\equiv Q_{1}^{i,j}\pi_{1}^{i,j}\cdots Q_{n}^{i,j}\pi_{n}^{i,j}% \theta_{i,j}¬ italic_ψ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ≡ italic_Q start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_i , italic_j end_POSTSUPERSCRIPT italic_π start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_i , italic_j end_POSTSUPERSCRIPT ⋯ italic_Q start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_i , italic_j end_POSTSUPERSCRIPT italic_π start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_i , italic_j end_POSTSUPERSCRIPT italic_θ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT, where θi,jLTLsubscript𝜃𝑖𝑗LTL\theta_{i,j}\in\mathrm{LTL}italic_θ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ∈ roman_LTL. Thus we may assume that φi,jsubscript𝜑𝑖𝑗\varphi_{i,j}italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT only appears positively. By a similar argument to that of the proof of Proposition 2 we get the following:

iIjJQ1i,jπ1i,jQni,jπni,jψi,jQ11,1π11,1Q11,jπ11,jQni,jπni,jiIjJψi,j.subscript𝑖𝐼subscript𝑗𝐽superscriptsubscript𝑄1𝑖𝑗superscriptsubscript𝜋1𝑖𝑗superscriptsubscript𝑄𝑛𝑖𝑗superscriptsubscript𝜋𝑛𝑖𝑗subscript𝜓𝑖𝑗superscriptsubscript𝑄111superscriptsubscript𝜋111superscriptsubscript𝑄11𝑗superscriptsubscript𝜋11𝑗superscriptsubscript𝑄𝑛𝑖𝑗superscriptsubscript𝜋𝑛𝑖𝑗subscript𝑖𝐼subscript𝑗𝐽subscript𝜓𝑖𝑗\bigvee_{i\in I}\bigwedge_{j\in J}Q_{1}^{i,j}\pi_{1}^{i,j}\cdots Q_{n}^{i,j}% \pi_{n}^{i,j}\psi_{i,j}\equiv Q_{1}^{1,1}\pi_{1}^{1,1}\cdots Q_{1}^{1,j}\pi_{1% }^{1,j}\cdots Q_{n}^{i,j}\pi_{n}^{i,j}\bigvee_{i\in I}\bigwedge_{j\in J}\psi_{% i,j}.⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_Q start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_i , italic_j end_POSTSUPERSCRIPT italic_π start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_i , italic_j end_POSTSUPERSCRIPT ⋯ italic_Q start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_i , italic_j end_POSTSUPERSCRIPT italic_π start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_i , italic_j end_POSTSUPERSCRIPT italic_ψ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ≡ italic_Q start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 1 , 1 end_POSTSUPERSCRIPT italic_π start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 1 , 1 end_POSTSUPERSCRIPT ⋯ italic_Q start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 1 , italic_j end_POSTSUPERSCRIPT italic_π start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 1 , italic_j end_POSTSUPERSCRIPT ⋯ italic_Q start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_i , italic_j end_POSTSUPERSCRIPT italic_π start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_i , italic_j end_POSTSUPERSCRIPT ⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_ψ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT .

One last remark before we get to the core results of this article, this time relating quantifier-free HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL-formulae with LTLLTL\mathrm{LTL}roman_LTL-formulae. The remark can again be proven by induction on the structure of the formula.

Remark 2

Let T𝑇Titalic_T be a team, ΠΠ\Piroman_Π be a trace assignment, π𝜋\piitalic_π be a trace variable, φ𝜑\varphiitalic_φ be a LTLLTL\mathrm{LTL}roman_LTL-formula, and let φ(π)𝜑𝜋\varphi(\pi)italic_φ ( italic_π ) be the HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL formula identical to φ𝜑\varphiitalic_φ, except every proposition symbol p𝑝pitalic_p is replaced by pπsubscript𝑝𝜋p_{\pi}italic_p start_POSTSUBSCRIPT italic_π end_POSTSUBSCRIPT. Suppose Π(π)=tΠ𝜋𝑡\Pi(\pi)=troman_Π ( italic_π ) = italic_t for some tT𝑡𝑇t\in Titalic_t ∈ italic_T. Now the following equivalence holds

ΠTφ(π)tφ.iffsubscriptmodels𝑇Π𝜑𝜋models𝑡𝜑\Pi\models_{T}\varphi(\pi)\iff t\models\varphi.roman_Π ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_φ ( italic_π ) ⇔ italic_t ⊧ italic_φ .

Using the above propositions we may now proceed with proving our main results: correspondence theorems between team logics and the Boolean closures of hyperlogics.

Note that TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL has no separation between closed and open formulae, and has no features to encode trace assignments. Thus, when φ𝜑\varphiitalic_φ is a formula of some team based logic \mathcal{L}caligraphic_L and ψ𝜓\psiitalic_ψ is a formula of a hyper logic superscript\mathcal{L}^{\prime}caligraphic_L start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT without free variables, we say that φ𝜑\varphiitalic_φ and ψ𝜓\psiitalic_ψ are equivalent, if the equivalence TφTψmodels𝑇𝜑subscriptmodels𝑇𝜓T\models\varphi\Leftrightarrow\emptyset\models_{T}\psiitalic_T ⊧ italic_φ ⇔ ∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_ψ, holds for all sets of traces T𝑇Titalic_T. The notations superscript\mathcal{L}\leq\mathcal{L}^{\prime}caligraphic_L ≤ caligraphic_L start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT and superscript\mathcal{L}\equiv\mathcal{L}^{\prime}caligraphic_L ≡ caligraphic_L start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT are then defined in the obvious way, by restricting superscript\mathcal{L}^{\prime}caligraphic_L start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT to formulae without free variables.

Theorem 3.1

TeamLTLl(∨⃝)PBC(HyperLTL)superscriptTeamLTL𝑙∨⃝PBCfor-allHyperLTL\mathrm{TeamLTL}^{l}(\varovee)\equiv\mathrm{PBC}(\forall\mathrm{HyperLTL})roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∨⃝ ) ≡ roman_PBC ( ∀ roman_HyperLTL )

Proof

Let T𝑇Titalic_T be an arbitrary team and φ𝜑\varphiitalic_φ an arbitrary TeamLTLl(∨⃝)superscriptTeamLTL𝑙∨⃝\mathrm{TeamLTL}^{l}(\varovee)roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∨⃝ )-formula. By [11, Theorem 10], we may assume that φ𝜑\varphiitalic_φ is in the form \scalerel∨⃝iIαi\operatorname*{\scalerel*{\ovee}{\sum}}_{i\in I}\alpha_{i}start_OPERATOR ∗ ∨⃝ ∑ end_OPERATOR start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT, where I𝐼Iitalic_I in an index set and αisubscript𝛼𝑖\alpha_{i}italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT are LTLLTL\mathrm{LTL}roman_LTL-formulae. We let αi(π)subscript𝛼𝑖𝜋\alpha_{i}(\pi)italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_π ) denote the HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL-formulae obtained from αisubscript𝛼𝑖\alpha_{i}italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT, by replacing every proposition symbol p𝑝pitalic_p by pπsubscript𝑝𝜋p_{\pi}italic_p start_POSTSUBSCRIPT italic_π end_POSTSUBSCRIPT. We obtain the following chain of equivalences:

T\scalerel∨⃝iIαi\displaystyle T\models\operatorname*{\scalerel*{\ovee}{\sum}}_{i\in I}\alpha_{i}italic_T ⊧ start_OPERATOR ∗ ∨⃝ ∑ end_OPERATOR start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT there is iI such that Tαiiffabsentthere is 𝑖𝐼 such that 𝑇modelssubscript𝛼𝑖\displaystyle\iff\text{there is }i\in I\text{ such that }T\models\alpha_{i}⇔ there is italic_i ∈ italic_I such that italic_T ⊧ italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT
there is iI such that tαi for all tTiffabsentthere is 𝑖𝐼 such that 𝑡modelssubscript𝛼𝑖 for all 𝑡𝑇\displaystyle\iff\text{there is }i\in I\text{ such that }t\models\alpha_{i}% \text{ for all }t\in T⇔ there is italic_i ∈ italic_I such that italic_t ⊧ italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT for all italic_t ∈ italic_T
there is iI such that Tπαi(π)iffabsentthere is 𝑖𝐼 such that subscriptmodels𝑇for-all𝜋subscript𝛼𝑖𝜋\displaystyle\iff\text{there is }i\in I\text{ such that }\emptyset\models_{T}% \forall\pi\alpha_{i}(\pi)⇔ there is italic_i ∈ italic_I such that ∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ∀ italic_π italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_π )
TiIπαi(π),iffabsentsubscriptmodels𝑇subscript𝑖𝐼for-all𝜋subscript𝛼𝑖𝜋\displaystyle\iff\emptyset\models_{T}\bigvee_{i\in I}\forall\pi\alpha_{i}(\pi),⇔ ∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ∀ italic_π italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_π ) ,

where the first equivalence follows from the semantics of ∨⃝∨⃝\varovee∨⃝, the second equivalence holds by the flatness of αisubscript𝛼𝑖\alpha_{i}italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT, the third equivalence is due to the semantics of for-all\forall and Remark 2, and the final equivalence follows from the semantics of \vee.

For the converse direction, consider an arbitrary PBC(HyperLTL)PBCfor-allHyperLTL\mathrm{PBC}(\forall\mathrm{HyperLTL})roman_PBC ( ∀ roman_HyperLTL )-sentence ψ𝜓\psiitalic_ψ. As noted above, ψ𝜓\psiitalic_ψ is equivalent to a sentence iIjJπφi,j(π)subscript𝑖𝐼subscript𝑗𝐽for-all𝜋subscript𝜑𝑖𝑗𝜋\bigvee_{i\in I}\bigwedge_{j\in J}\forall\pi\varphi_{i,j}(\pi)⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT ∀ italic_π italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ( italic_π ), where φi,j(π)subscript𝜑𝑖𝑗𝜋\varphi_{i,j}(\pi)italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ( italic_π ), for every pair i𝑖iitalic_i and j𝑗jitalic_j, is a HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL-formula with π𝜋\piitalic_π as the only free variable. Now by an argument similar to the proof of Proposition 2, TiIjJπφi,j(π)subscriptmodels𝑇subscript𝑖𝐼subscript𝑗𝐽for-all𝜋subscript𝜑𝑖𝑗𝜋\emptyset\models_{T}\bigvee_{i\in I}\bigwedge_{j\in J}\forall\pi\varphi_{i,j}(\pi)∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT ∀ italic_π italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ( italic_π ) if and only if TiIπjJφi,j(π)subscriptmodels𝑇subscript𝑖𝐼for-all𝜋subscript𝑗𝐽subscript𝜑𝑖𝑗𝜋\emptyset\models_{T}\bigvee_{i\in I}\forall\pi\bigwedge_{j\in J}\varphi_{i,j}(\pi)∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ∀ italic_π ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ( italic_π ). Equivalently then by the definition of the semantics of the disjunction, we may fix iIsuperscript𝑖𝐼i^{\prime}\in Iitalic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT ∈ italic_I such that TπjJφi,j(π)subscriptmodels𝑇for-all𝜋subscript𝑗𝐽subscript𝜑superscript𝑖𝑗𝜋\emptyset\models_{T}\forall\pi\bigwedge_{j\in J}\varphi_{i^{\prime},j}(\pi)∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ∀ italic_π ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT italic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT , italic_j end_POSTSUBSCRIPT ( italic_π ). By the definition of the universal quantifier then we get that the previous is equivalent with [πt]TjJφi,j(π)subscriptmodels𝑇delimited-[]maps-to𝜋𝑡subscript𝑗𝐽subscript𝜑superscript𝑖𝑗𝜋\emptyset[\pi\mapsto t]\models_{T}\bigwedge_{j\in J}\varphi_{i^{\prime},j}(\pi)∅ [ italic_π ↦ italic_t ] ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT italic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT , italic_j end_POSTSUBSCRIPT ( italic_π ) for all tT𝑡𝑇t\in Titalic_t ∈ italic_T. Now by Remark 2, the previous holds if and only if tjJφi,j for all tTmodels𝑡subscript𝑗𝐽subscript𝜑superscript𝑖𝑗 for all 𝑡𝑇t\models\bigwedge_{j\in J}\varphi_{i^{\prime},j}\text{ for all }t\in Titalic_t ⊧ ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT italic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT , italic_j end_POSTSUBSCRIPT for all italic_t ∈ italic_T, which is equivalent to TljJφi,jsuperscriptmodels𝑙𝑇subscript𝑗𝐽subscript𝜑superscript𝑖𝑗T\mathrel{\models^{l}}\bigwedge_{j\in J}\varphi_{i^{\prime},j}italic_T start_RELOP ⊧ start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT end_RELOP ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT italic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT , italic_j end_POSTSUBSCRIPT, due to the flatness property of TeamLTLTeamLTL\mathrm{TeamLTL}roman_TeamLTL. Finally, by the semantics of the Boolean disjunction, the previous is equivalent with T\scalerel∨⃝iIjJφi,jT\models\operatorname*{\scalerel*{\ovee}{\sum}}_{i\in I}\bigwedge_{j\in J}% \varphi_{i,j}italic_T ⊧ start_OPERATOR ∗ ∨⃝ ∑ end_OPERATOR start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT. ∎

As a corollary we get that TeamLTLl(∨⃝)superscriptTeamLTL𝑙∨⃝\mathrm{TeamLTL}^{l}(\varovee)roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∨⃝ ) is subsumed by the universal fragment of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL, which follows from Theorem 3.1 and the observations made in the proof of Proposition 2.

Corollary 1

TeamLTLl(∨⃝)HyperLTLsuperscriptTeamLTL𝑙∨⃝superscriptfor-allHyperLTL\mathrm{TeamLTL}^{l}(\varovee)\leq\forall^{*}\mathrm{HyperLTL}roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∨⃝ ) ≤ ∀ start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT roman_HyperLTL

Note that another consequence of Theorem 3.1 is that HyperLTLfor-allHyperLTL\forall\mathrm{HyperLTL}∀ roman_HyperLTL is strictly less expressive than PBC(HyperLTL)PBCfor-allHyperLTL\mathrm{PBC}(\forall\mathrm{HyperLTL})roman_PBC ( ∀ roman_HyperLTL ), as the former is equivalent with LTLLTL\mathrm{LTL}roman_LTL [5] and thus has the flatness property, while the latter is equivalent with TeamLTL(∨⃝)TeamLTL∨⃝\mathrm{TeamLTL}(\varovee)roman_TeamLTL ( ∨⃝ ), which does not satisfy flatness. This stands in contrast to the unrestricted universal fragment HyperLTLsuperscriptfor-allHyperLTL\forall^{*}\mathrm{HyperLTL}∀ start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT roman_HyperLTL, which by Proposition 2 is equivalent to its positive Boolean closure.

Theorem 3.2

left-dc–TeamLTLl()BC(QHyperLTL)BC(HyperLTL)left-dc–superscriptTeamLTL𝑙similar-toBC𝑄HyperLTLBCfor-allHyperLTL\text{left-dc--}\mathrm{TeamLTL}^{l}(\sim)\equiv\mathrm{BC}(Q\mathrm{HyperLTL}% )\equiv\mathrm{BC}(\forall\mathrm{HyperLTL})left-dc– roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∼ ) ≡ roman_BC ( italic_Q roman_HyperLTL ) ≡ roman_BC ( ∀ roman_HyperLTL )

Proof

Let φ𝜑\varphiitalic_φ be an arbitrary left-dc–TeamLTLl()left-dc–superscriptTeamLTL𝑙similar-to\text{left-dc--}\mathrm{TeamLTL}^{l}(\sim)left-dc– roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∼ )-formula. Now by the quasi-flat normal form Tlφsuperscriptmodels𝑙𝑇𝜑T\mathrel{\models^{l}}\varphiitalic_T start_RELOP ⊧ start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT end_RELOP italic_φ if and only if Tl\scalerel∨⃝iI(αijJβi,j)T\mathrel{\models^{l}}\operatorname*{\scalerel*{\ovee}{\sum}}_{i\in I}\big{(}% \alpha_{i}\wedge\bigwedge_{j\in J}\exists\beta_{i,j}\big{)}italic_T start_RELOP ⊧ start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT end_RELOP start_OPERATOR ∗ ∨⃝ ∑ end_OPERATOR start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ( italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ∧ ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT ∃ italic_β start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ). Equivalently, by the semantics of ∨⃝∨⃝\varovee∨⃝, we may fix an index iIsuperscript𝑖𝐼i^{\prime}\in Iitalic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT ∈ italic_I such that TlαijJβi,jsuperscriptmodels𝑙𝑇subscript𝛼superscript𝑖subscript𝑗𝐽subscript𝛽superscript𝑖𝑗T\mathrel{\models^{l}}\alpha_{i^{\prime}}\wedge\bigwedge_{j\in J}\exists\beta_% {i^{\prime},j}italic_T start_RELOP ⊧ start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT end_RELOP italic_α start_POSTSUBSCRIPT italic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT end_POSTSUBSCRIPT ∧ ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT ∃ italic_β start_POSTSUBSCRIPT italic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT , italic_j end_POSTSUBSCRIPT. By the semantics of the logic, flatness, and the interpretation of the shorthand \exists, the previous evaluation is equivalent with that tαimodels𝑡subscript𝛼superscript𝑖t\models\alpha_{i^{\prime}}italic_t ⊧ italic_α start_POSTSUBSCRIPT italic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT end_POSTSUBSCRIPT for all tT𝑡𝑇t\in Titalic_t ∈ italic_T and for all jJ𝑗𝐽j\in Jitalic_j ∈ italic_J there is a tjTsubscript𝑡𝑗𝑇t_{j}\in Titalic_t start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ∈ italic_T such that tjβi,jmodelssubscript𝑡𝑗subscript𝛽superscript𝑖𝑗t_{j}\models\beta_{i^{\prime},j}italic_t start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ⊧ italic_β start_POSTSUBSCRIPT italic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT , italic_j end_POSTSUBSCRIPT. By Remark 2 the previous holds if and only if [πt]Tαi(π)subscriptmodels𝑇delimited-[]maps-to𝜋𝑡subscript𝛼superscript𝑖𝜋\emptyset[\pi\mapsto t]\models_{T}\alpha_{i^{\prime}}(\pi)∅ [ italic_π ↦ italic_t ] ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_α start_POSTSUBSCRIPT italic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT end_POSTSUBSCRIPT ( italic_π ) for all tT𝑡𝑇t\in Titalic_t ∈ italic_T and for all jJ𝑗𝐽j\in Jitalic_j ∈ italic_J there is a tjTsubscript𝑡𝑗𝑇t_{j}\in Titalic_t start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ∈ italic_T such that [πtj]Tβi,j(π)subscriptmodels𝑇delimited-[]maps-to𝜋subscript𝑡𝑗subscript𝛽superscript𝑖𝑗𝜋\emptyset[\pi\mapsto t_{j}]\models_{T}\beta_{i^{\prime},j}(\pi)∅ [ italic_π ↦ italic_t start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ] ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_β start_POSTSUBSCRIPT italic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT , italic_j end_POSTSUBSCRIPT ( italic_π ). Equivalently, by the semantics of for-all\forall and \exists, we have that Tπαi(π)subscriptmodels𝑇for-all𝜋subscript𝛼superscript𝑖𝜋\emptyset\models_{T}\forall\pi\alpha_{i^{\prime}}(\pi)∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ∀ italic_π italic_α start_POSTSUBSCRIPT italic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT end_POSTSUBSCRIPT ( italic_π ) and TjJπβi,j(π)subscriptmodels𝑇subscript𝑗𝐽𝜋subscript𝛽superscript𝑖𝑗𝜋\emptyset\models_{T}\bigwedge_{j\in J}\exists\pi\beta_{i^{\prime},j}(\pi)∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT ∃ italic_π italic_β start_POSTSUBSCRIPT italic_i start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT , italic_j end_POSTSUBSCRIPT ( italic_π ), which, finally by the semantics of \vee and \wedge, holds if and only if TiI(παi(π)jJπβi,j(π))subscriptmodels𝑇subscript𝑖𝐼for-all𝜋subscript𝛼𝑖𝜋subscript𝑗𝐽𝜋subscript𝛽𝑖𝑗𝜋\emptyset\models_{T}\bigvee_{i\in I}\big{(}\forall\pi\alpha_{i}(\pi)\wedge% \bigwedge_{j\in J}\exists\pi\beta_{i,j}(\pi)\big{)}∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ( ∀ italic_π italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_π ) ∧ ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT ∃ italic_π italic_β start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ( italic_π ) ).

On the other hand, let ψ𝜓\psiitalic_ψ be an arbitrary sentence of BC(HyperLTL)BCHyperLTL\mathrm{BC}(\mathrm{HyperLTL})roman_BC ( roman_HyperLTL ). Now we get the following chain of equivalences, where Qi,j{,}subscript𝑄𝑖𝑗for-allQ_{i,j}\in\{\exists,\forall\}italic_Q start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ∈ { ∃ , ∀ }:

Tψsubscriptmodels𝑇𝜓\displaystyle\emptyset\models_{T}\psi∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_ψ iff\displaystyle\iff TiI(jJQi,jπφi,j)subscriptmodels𝑇subscript𝑖𝐼subscript𝑗𝐽subscript𝑄𝑖𝑗𝜋subscript𝜑𝑖𝑗\displaystyle\emptyset\models_{T}\bigvee_{i\in I}\big{(}\bigwedge_{j\in J}Q_{i% ,j}\pi\varphi_{i,j}\big{)}∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ( ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT italic_Q start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT italic_π italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT )
iff\displaystyle\iff TiI(παijJπφi,j)subscriptmodels𝑇subscript𝑖𝐼for-all𝜋subscript𝛼𝑖subscript𝑗𝐽𝜋subscript𝜑𝑖𝑗\displaystyle\emptyset\models_{T}\bigvee_{i\in I}\big{(}\forall\pi\alpha_{i}% \wedge\bigwedge_{j\in J}\exists\pi\varphi_{i,j}\big{)}∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ⋁ start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ( ∀ italic_π italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ∧ ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT ∃ italic_π italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT )
iff\displaystyle\iff there is iI such that Tπαi, and for all jJformulae-sequencethere is 𝑖𝐼 such that subscriptmodels𝑇for-all𝜋subscript𝛼𝑖 and for all 𝑗𝐽\displaystyle\text{there is }i\in I\text{ such that }\emptyset\models_{T}% \forall\pi\alpha_{i},\text{ and for all }j\in Jthere is italic_i ∈ italic_I such that ∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ∀ italic_π italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , and for all italic_j ∈ italic_J
it holds that Tπφi,jsubscriptmodels𝑇it holds that 𝜋subscript𝜑𝑖𝑗\displaystyle\text{it holds that }\emptyset\models_{T}\exists\pi\varphi_{i,j}it holds that ∅ ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT ∃ italic_π italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT
iff\displaystyle\iff there is iI such that [πt]Tαi for all tT, and for allformulae-sequencethere is 𝑖𝐼 such that delimited-[]maps-to𝜋𝑡subscriptmodels𝑇subscript𝛼𝑖 for all 𝑡𝑇 and for all\displaystyle\text{there is }i\in I\text{ such that }\emptyset[\pi\mapsto t]% \models_{T}\alpha_{i}\text{ for all }t\in T,\text{ and for all }there is italic_i ∈ italic_I such that ∅ [ italic_π ↦ italic_t ] ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT for all italic_t ∈ italic_T , and for all
jJ there exists tjT such that [πtj]Tφi,j𝑗𝐽 there exists subscript𝑡𝑗𝑇 such that delimited-[]maps-to𝜋subscript𝑡𝑗subscriptmodels𝑇subscript𝜑𝑖𝑗\displaystyle j\in J\text{ there exists }t_{j}\in T\text{ such that }\emptyset% [\pi\mapsto t_{j}]\models_{T}\varphi_{i,j}italic_j ∈ italic_J there exists italic_t start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ∈ italic_T such that ∅ [ italic_π ↦ italic_t start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ] ⊧ start_POSTSUBSCRIPT italic_T end_POSTSUBSCRIPT italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT
iff\displaystyle\iff there is iI such that tαi for all tT, and for all jJformulae-sequencethere is 𝑖𝐼 such that 𝑡modelssubscript𝛼𝑖 for all 𝑡𝑇 and for all 𝑗𝐽\displaystyle\text{there is }i\in I\text{ such that }t\models\alpha_{i}\text{ % for all }t\in T,\text{ and for all }j\in Jthere is italic_i ∈ italic_I such that italic_t ⊧ italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT for all italic_t ∈ italic_T , and for all italic_j ∈ italic_J
there exists tjT such that tjφi,jthere exists subscript𝑡𝑗𝑇 such that subscript𝑡𝑗modelssubscript𝜑𝑖𝑗\displaystyle\text{there exists }t_{j}\in T\text{ such that }t_{j}\models% \varphi_{i,j}there exists italic_t start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ∈ italic_T such that italic_t start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT ⊧ italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT
iff\displaystyle\iff there is iI such that Tαi, and for all jJ it holds thatformulae-sequencethere is 𝑖𝐼 such that 𝑇modelssubscript𝛼𝑖 and for all 𝑗𝐽 it holds that\displaystyle\text{there is }i\in I\text{ such that }T\models\alpha_{i},\text{% and for all }j\in J\text{ it holds that }there is italic_i ∈ italic_I such that italic_T ⊧ italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , and for all italic_j ∈ italic_J it holds that
Tφi,jmodels𝑇subscript𝜑𝑖𝑗\displaystyle T\models\exists\varphi_{i,j}italic_T ⊧ ∃ italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT
iff\displaystyle\iff there is iI such that TαijJφi,jthere is 𝑖𝐼 such that 𝑇modelssubscript𝛼𝑖subscript𝑗𝐽subscript𝜑𝑖𝑗\displaystyle\text{there is }i\in I\text{ such that }T\models\alpha_{i}\wedge% \bigwedge_{j\in J}\exists\varphi_{i,j}there is italic_i ∈ italic_I such that italic_T ⊧ italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ∧ ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT ∃ italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT
iff\displaystyle\iff T\scalerel∨⃝iI(αijJφi,j),\displaystyle T\models\operatorname*{\scalerel*{\ovee}{\sum}}_{i\in I}\big{(}% \alpha_{i}\wedge\bigwedge_{j\in J}\exists\varphi_{i,j}\big{)},italic_T ⊧ start_OPERATOR ∗ ∨⃝ ∑ end_OPERATOR start_POSTSUBSCRIPT italic_i ∈ italic_I end_POSTSUBSCRIPT ( italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ∧ ⋀ start_POSTSUBSCRIPT italic_j ∈ italic_J end_POSTSUBSCRIPT ∃ italic_φ start_POSTSUBSCRIPT italic_i , italic_j end_POSTSUBSCRIPT ) ,

where the first equivalence is due to the normal form for a Boolean closure, the second equivalence is holds because the universally quantified conjuncts can equivalently be evaluated simultaneously, the third equivalence follows from the semantics of \wedge and \vee, the fourth equivalence holds by the semantics of for-all\forall and \exists, the fifth equivalence holds by Remark 2, the sixth equivalences is due to flatness and the definition of the shorthand \exists, the seventh equivalence holds by the semantics of \wedge, and the final equivalence follows from the semantics of ∨⃝∨⃝\varovee∨⃝.

The other equivalence in the theorem is a direct consequence of Remark 1.∎

4 Conclusion

In this article we explored the connections in expressivity between extensions of linear temporal logic under set-based team semantics (TeamLTLlsuperscriptTeamLTL𝑙\mathrm{TeamLTL}^{l}roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT) and fragments of linear temporal logic extended with trace quantifiers (HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL). We showed that TeamLTLlsuperscriptTeamLTL𝑙\mathrm{TeamLTL}^{l}roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT, when extended with the Boolean disjunction, corresponds to the positive Boolean closure of the one variable universal fragment of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL. Furthermore we considered a fragment of TeamLTLlsuperscriptTeamLTL𝑙\mathrm{TeamLTL}^{l}roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT extended with the Boolean negation, where the formulae are restricted to not to contain the Boolean negation on the left-hand side of the ‘until’ operator (𝖴𝖴\operatorname{\mathsf{U}}sansserif_U) or under the ‘always going to’ (𝖦𝖦\operatorname{\mathsf{G}}sansserif_G) operator. We showed a correspondence between that fragment and the Boolean closure of the one variable universal fragment of HyperLTLHyperLTL\mathrm{HyperLTL}roman_HyperLTL. From our results it follows that the logics considered are all true extensions of LTLLTL\mathrm{LTL}roman_LTL. Decidability of the model checking and satisfaction problems for the team based logics was shown in [11], and by our correspondence results (and the translation implied by the proofs of the theorems), decidability of the problems extends to the hyperlogics in question as well.

It is fascinating to see that the restriction to left downward closed formulae in the latter correspondence on the team logic side disappears on the hyperlogic side. This hints at that the fragment considered is intuitive. It is an open question whether the downward closed fragment of TeamLTLl()superscriptTeamLTL𝑙similar-to\mathrm{TeamLTL}^{l}(\sim)roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∼ ) is TeamLTLl(∨⃝)superscriptTeamLTL𝑙∨⃝\mathrm{TeamLTL}^{l}(\varovee)roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∨⃝ ), or if some restricted use of the Boolean negation could be allowed and still maintain downward closure. Another open question is whether an analogous correspondence exists for the full logic TeamLTLl()superscriptTeamLTL𝑙similar-to\mathrm{TeamLTL}^{l}(\sim)roman_TeamLTL start_POSTSUPERSCRIPT italic_l end_POSTSUPERSCRIPT ( ∼ ), or even for some lesser restriction of the logic than the left-downward closed fragment.

{credits}

Acknowledgements. This work was supported by the Academy of Finland grant 345634 and the DFG grant VI 1045/1-1.

References

  • [1] Jan Baumeister, Norine Coenen, Borzoo Bonakdarpour, Bernd Finkbeiner, and César Sánchez. A temporal logic for asynchronous hyperproperties. In CAV (1), volume 12759 of Lecture Notes in Computer Science, pages 694–717. Springer, 2021.
  • [2] Michael R. Clarkson, Bernd Finkbeiner, Masoud Koleini, Kristopher K. Micinski, Markus N. Rabe, and César Sánchez. Temporal logics for hyperproperties. In POST 2014, pages 265–284, 2014.
  • [3] Michael R. Clarkson and Fred B. Schneider. Hyperproperties. Journal of Computer Security, 18(6):1157–1210, 2010.
  • [4] Norine Coenen, Bernd Finkbeiner, Christopher Hahn, and Jana Hofmann. The hierarchy of hyperlogics. In LICS 2019, pages 1–13. IEEE, 2019.
  • [5] Bernd Finkbeiner. Temporal hyperproperties. Bull. EATCS, 123, 2017. URL: http://eatcs.org/beatcs/index.php/beatcs/article/view/514.
  • [6] Bernd Finkbeiner, Christopher Hahn, Philip Lukert, Marvin Stenger, and Leander Tentrup. Synthesis from hyperproperties. Acta Informatica, 57(1-2):137–163, 2020. doi:10.1007/s00236-019-00358-2.
  • [7] Bernd Finkbeiner and Martin Zimmermann. The first-order logic of hyperproperties. In Heribert Vollmer and Brigitte Vallée, editors, STACS 2017, volume 66 of LIPIcs, pages 30:1–30:14. Schloss Dagstuhl - Leibniz-Zentrum für Informatik, 2017.
  • [8] Jens Oliver Gutsfeld, Arne Meier, Christoph Ohrem, and Jonni Virtema. Temporal team semantics revisited. In Christel Baier and Dana Fisman, editors, LICS ’22: 37th Annual ACM/IEEE Symposium on Logic in Computer Science, Haifa, Israel, August 2 - 5, 2022, pages 44:1–44:13. ACM, 2022. doi:10.1145/3531130.3533360.
  • [9] Jens Oliver Gutsfeld, Markus Müller-Olm, and Christoph Ohrem. Automata and fixpoints for asynchronous hyperproperties. Proc. ACM Program. Lang., 5(POPL):1–29, 2021. doi:10.1145/3434319.
  • [10] Juha Kontinen and Max Sandström. On the expressive power of teamltl and first-order team logic over hyperproperties. In WoLLIC, volume 13038 of Lecture Notes in Computer Science, pages 302–318. Springer, 2021.
  • [11] Juha Kontinen, Max Sandström, and Jonni Virtema. Set semantics for asynchronous teamltl: Expressivity and complexity. In Jérôme Leroux, Sylvain Lombardy, and David Peleg, editors, 48th International Symposium on Mathematical Foundations of Computer Science, MFCS 2023, August 28 to September 1, 2023, Bordeaux, France, volume 272 of LIPIcs, pages 60:1–60:14. Schloss Dagstuhl - Leibniz-Zentrum für Informatik, 2023. URL: https://doi.org/10.4230/LIPIcs.MFCS.2023.60, doi:10.4230/LIPICS.MFCS.2023.60.
  • [12] Andreas Krebs, Arne Meier, and Jonni Virtema. A team based variant of CTL. In Fabio Grandi, Martin Lange, and Alessio Lomuscio, editors, 22nd International Symposium on Temporal Representation and Reasoning, TIME 2015, Kassel, Germany, September 23-25, 2015, pages 140–149. IEEE Computer Society, 2015. doi:10.1109/TIME.2015.11.
  • [13] Andreas Krebs, Arne Meier, Jonni Virtema, and Martin Zimmermann. Team Semantics for the Specification and Verification of Hyperproperties. In Igor Potapov, Paul Spirakis, and James Worrell, editors, MFCS 2018, volume 117, pages 10:1–10:16, Dagstuhl, Germany, 2018. Schloss Dagstuhl–Leibniz-Zentrum fuer Informatik.
  • [14] John McLean. Proving noninterference and functional correctness using traces. Journal of Computer Security, 1(1):37–58, 1992. doi:10.3233/JCS-1992-1103.
  • [15] Amir Pnueli. The temporal logic of programs. In 18th Annual Symposium on Foundations of Computer Science, pages 46–57. IEEE Computer Society, 1977.
  • [16] Markus N. Rabe. A Temporal Logic Approach to Information-Flow Control. PhD thesis, Saarland University, 2016.
  • [17] A. W. Roscoe. CSP and determinism in security modelling. In Proceedings of the 1995 IEEE Symposium on Security and Privacy, Oakland, California, USA, May 8-10, 1995, pages 114–127. IEEE Computer Society, 1995. doi:10.1109/SECPRI.1995.398927.
  • [18] Jouko Väänänen. Dependence Logic. Cambridge University Press, 2007.
  • [19] Jonni Virtema, Jana Hofmann, Bernd Finkbeiner, Juha Kontinen, and Fan Yang. Linear-time temporal logic with team semantics: Expressivity and complexity. In Mikolaj Bojanczyk and Chandra Chekuri, editors, 41st IARCS Annual Conference on Foundations of Software Technology and Theoretical Computer Science, FSTTCS 2021, December 15-17, 2021, Virtual Conference, volume 213 of LIPIcs, pages 52:1–52:17. Schloss Dagstuhl - Leibniz-Zentrum für Informatik, 2021. doi:10.4230/LIPIcs.FSTTCS.2021.52.
  • [20] Steve Zdancewic and Andrew C. Myers. Observational determinism for concurrent program security. In 16th IEEE Computer Security Foundations Workshop (CSFW-16 2003), 30 June - 2 July 2003, Pacific Grove, CA, USA, page 29. IEEE Computer Society, 2003. doi:10.1109/CSFW.2003.1212703.