23.08.2013 Views

Specification of Reactive Hardware/Software Systems - Electronic ...

Specification of Reactive Hardware/Software Systems - Electronic ...

Specification of Reactive Hardware/Software Systems - Electronic ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Pro<strong>of</strong>s <strong>of</strong> Propositions and Transformations 429<br />

¢<br />

BSpec e<br />

¢<br />

¡<br />

¡<br />

£<br />

£<br />

application <strong>of</strong> 7<br />

because <strong>of</strong> (1) ChSort(BSpec e ¥ Sys p ) §<br />

application <strong>of</strong> 4<br />

because <strong>of</strong> (2) f ¤ ChSort(BSpec e<br />

L§ f ¥ envs¥ Sys p ¥ Sys<br />

<br />

e<br />

£ BSpec<br />

application <strong>of</strong> 1 ¦<br />

¢<br />

¡<br />

e BSpec<br />

L§ Id ¥ envs¥ Sys p ¥ Sys<br />

£<br />

¡ ¡ ¡ ¡ ¤ (L)¦ ¦<br />

ch L f (ch) f<br />

L¥ Sys p ) = Id ¤ ChSort(BSpec e<br />

L¥ Sys p ) ¦<br />

L¥ envs¥ Sys p ¥ Sys<br />

This ends the pro<strong>of</strong> <strong>of</strong> Transformation 10. Note that the pro<strong>of</strong> implicitly uses the<br />

congruence property <strong>of</strong> ¢<br />

¡ as stated in Proposition 4. ¨<br />

Pro<strong>of</strong> <strong>of</strong> Transformation 11<br />

The pro<strong>of</strong> <strong>of</strong> the observation equivalence part consists <strong>of</strong> showing that = £<br />

( (BSpece 1 ¢<br />

BSpece 2) envs1 ¥ envs2 Sys L¥ p <br />

¥<br />

e (BSpec1 L) ¢ (BSpec ¥ Sys e 2 envs1 ¥ envs2 Sys L)¥ p ) ¡<br />

£<br />

¥ Sys<br />

Chan(la ) l ¡ a ¡ AASort(BSpece 1 ¥ Sysp ) la ¡ AASort(BSpece 2 ¥ Sysp L ¤ ¦ is a weak<br />

)¦ §<br />

bisimulation. The pro<strong>of</strong> requires (iii) and (v) <strong>of</strong> Proposition 2. The transformation<br />

equivalence part follows from the definition <strong>of</strong> Reset and from ¨<br />

Lemma 2.<br />

Pro<strong>of</strong> <strong>of</strong> Transformation 12<br />

The result for transformation equivalence directly follows from the definition <strong>of</strong> Reset<br />

and from the result for observation equivalence. For observation equivalence we will<br />

show that = £<br />

( BSpece ¥ envs¥ Sys p<br />

1 ¥ Sys , BSpece ¥ envs¥ Sys p<br />

2 ¡ ) Sysp1<br />

and Sysp2<br />

are non-<br />

¥ Sys<br />

conflicting is a weak bisimulation. Let ¦ (conf p p<br />

p a<br />

1 conf ¡ 2 ) and suppose £ conf1 ¥ conf p<br />

1 ¡ .<br />

p a<br />

The pro<strong>of</strong> is by rule induction on the shape <strong>of</strong> the derivation £ tree <strong>of</strong> conf conf p<br />

1 ¡ . We<br />

argue by cases on the applied axioms and rules.<br />

Case axiom (9’)<br />

Then conf p<br />

1 = Cc <br />

¡ ¥ Er)¥ ¥ (E1 Sys ¥¡ ¡ p<br />

1 ¥ Sys , a = ¡ , conf p<br />

1 ¡ = <br />

BSpec § p P1 ¥¡ ¡ ¡ Er© Pr <br />

§ § E1©<br />

Cc <br />

¡£ Er¢ ¥ ¥ Sys E1£ ¡ ¡ p<br />

p <br />

1 , Sysp1<br />

CD1 ¥ Sys ¡ ¡ CDp ¡ ¡ CD p r , CDp cluster class Cc ¡ ¥ Pr P1 ¥¡ ¡<br />

behaviour specification ¡ ¡ BSpecp , BSpecp P1 ¥¡ ¡ ¡ ¥ Er© Pr ¡ B<strong>Specification</strong>s, and conf § E1© § p<br />

= Cc <br />

¥ Er)¥ ¥ ¥¡ ¡ ¡ (E1 Sys p<br />

Sys ¥ 2 . Since Sysp1<br />

and Sys p<br />

2 are non-conflicting, Sys p<br />

tains CDp and thus by axiom (9’) conf p<br />

£ 2 conf p<br />

2 ¡ = <br />

<br />

¥ Sys p<br />

¡<br />

¤ Sys ¥<br />

p<br />

2 . But then also conf2 conf p<br />

2 ¡ and clearly (conf p<br />

1 ¡ ¥ conf p<br />

.<br />

2 ¡ ) ¡<br />

2<br />

2 also con-<br />

§ BSpec p § § E1© P1 ¥¡ ¡ ¡ Er© Pr C c E1£ ¡ ¡ ¡ £ Er¢ ,<br />

Case rule (s’)<br />

Then conf p<br />

1 = BSpece 1 ¢ BSpece 2 ¥ envs1 envs2 ¥ Sys p<br />

p<br />

1 , conf1 ¥ Sys ¡ = BSpece 1 ¡ ¢ BSpece2 ,<br />

envs2 envs1¡ Sys ¥ p<br />

1 ¥ Sys , and BSpece 1 ¥ envs1 Sys ¥ p a<br />

£<br />

e<br />

1 BSpec ¡ ¥ envs1¡ ¥ 1 Sys ¥ Sys<br />

p<br />

1 . Since<br />

¥ Sys<br />

( BSpece 1 ¥ envs1 Sys ¥ p <br />

e<br />

1 , BSpec ¥ 1 ¥ envs1 Sys ¥ Sys p<br />

2 ¡ ) , we have by induction that<br />

¥ Sys<br />

<br />

e BSpec1 , envs1 , Sys p<br />

¡<br />

a<br />

¤ 2 , conf Sys<br />

p , for some conf p with ( BSpece 1 ¡ , envs1¡ , Sys p<br />

1 ,<br />

Sys ¥ conf p ) . But then necessarily ¡<br />

conf p = BSpece 1 ¡ ¥ envs1¡ Sys ¥ p<br />

2 . By applying rule<br />

¥ Sys<br />

(s’) we then have BSpece 1 ¢ BSpece2 ¥ envs1 envs2 ¥ Sys p<br />

¡<br />

a<br />

¤ 2 conf ¥ Sys<br />

p<br />

2 ¡ = BSpece 1 ¡ ¢ BSpece2 ,<br />

envs2 envs1¡ Sys ¥ p<br />

p<br />

2 , and evidently (conf1 ¥ Sys ¡ conf ¥ p<br />

2 ¡ ) ¡ .<br />

The pro<strong>of</strong>s <strong>of</strong> the remaining (and symmetric) cases are <strong>of</strong> a similar complexity. This ends

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!