Equivalence for Stochastic Systems


Nathanaël Fijalkow


Formerly: Research Fellow @ Simons, Berkeley
Now: Alan Turing Institute, affiliated to Warwick

Computer Science Department, Warwick, Thursday 9th, 2017

1 0 1 1/3 1/2 5/6 4/9 7/24 0.736111
A strategy induces a Markov Chain

Reasoning about Stochastic Models

Probabilistic Bisimulation

(due to Larsen, Skou 91)

Bisimulation for Non-Deterministic Systems


Labelled Transition Systems: $\Delta : Q \times A \to \mathcal{P}(Q)$

Bisimulation game: Duplicator claims that $p \equiv q$ We say that $p$ and $q$ are bisimilar if Duplicator can play forever


Coinductive definition: If $p \equiv q$, then for all $a$ in $A$, for all $p' \in \Delta(p,a)$, there exists $q' \in \Delta(q,a)$ such that $p' \equiv q'$

Bisimulation for Labelled Markov Chains


Labelled Markov Chains: $\Delta : Q \times A \to \mathcal{D}(Q)$

Coinductive definition: If $p \equiv q$, then for all $a$ in $A$, for all $C$ equivalence class of $\equiv$, $$\Delta(p,a,C) = \Delta(q,a,C)$$


Equivalent game: Duplicator claims that $p \equiv q$
  • Spoiler: pick $a$ in $A$ and $C \subseteq Q$ such that $\Delta(p,a,C) \neq \Delta(q,a,C)$
  • Duplicator: $p'$ in $C$, $q'$ not in $C$ and claim that $p' \equiv q'$
Duplicator claims they are bisimilar Duplicator claims they are bisimilar Spoiler picks a set of green states Duplicator claims they are bisimilar Spoiler picks a set of green states Duplicator picks a green state and a non-green state Duplicator claims they are bisimilar Spoiler picks a set of green states Duplicator picks a green state and a non-green state Spoiler's move was stupid! Duplicator claims they are bisimilar Spoiler picks a set of green states Duplicator claims they are bisimilar Spoiler picks a set of green states Duplicator picks a green state and a non-green state

25 years of research on Probabilistic Bisimulation


How to prove that two systems are not bisimilar?


Find a formula satisfying one but not the other!

Logical Characterisation

\[ \phi\ =\ \top\ \mid\ \phi \wedge \phi\ \mid\ \langle a \rangle_{> p} \phi\ \] Theorem: (Desharnais, Edalat, Panangaden 98)
Assumptions: the state space is a Polish space and the set of labels is countable

Two states are bisimiliar if, and only if, they satisfy the same formulas of this logic

Why countability of the set of labels?



Logical Characterisation, take 2

\[ \phi\ =\ \top\ \mid\ \phi \wedge \phi\ \mid\ \langle a \rangle_{> p} \phi\ \] Theorem: (F., Klin, Panangaden)
Assumptions: the state space is a Polish space and the transition function is continuous

Two states are bisimiliar if, and only if, they satisfy the same formulas of this logic