## A Local Algorithm for Checking Probabilistic Bisimilarity

### Cached

### Download Links

Citations: | 1 - 1 self |

### BibTeX

@MISC{Deng_alocal,

author = {Yuxin Deng and Wenjie Du},

title = {A Local Algorithm for Checking Probabilistic Bisimilarity},

year = {}

}

### OpenURL

### Abstract

Abstract—Bisimilarity is one of the most important relations for comparing the behaviour of formal systems in concurrency theory. Decision algorithms for bisimilarity in finite state systems are usually classified into two kinds: global algorithms are generally efficient but require to generate the whole state spaces in advance, and local algorithms combine the verification of a system’s behaviour with the generation of the system’s state space, which is often more effective to determine that one system fails to be related to another. Although local algorithms are well established in the classical concurrency theory, the study of local algorithms in probabilistic concurrency theory is not mature. In this paper we propose a polynomial time local algorithm for checking probabilistic bisimilarity. With mild modification, the algorithm can be easily adapted to decide probabilistic similarity with the same time complexity. Keywords-concurrency; probabilistic bisimilarity; local algorithm; probabilistic labelled transition systems; I.

### Citations

3204 |
Communication and Concurrency
- Milner
- 1989
(Show Context)
Citation Context ...algorithm; probabilistic labelled transition systems; I. INTRODUCTION In the last three decades a wealth of behavioural equivalences have been proposed in concurrency theory. Among them, bisimilarity =-=[38]-=-, [41] is probably the most studied one as it admits a suitable semantics and an elegant co-inductive proof technique. It can also be given an efficient decision procedure [40]. Given a labelled trans... |

655 |
Concurrency and automata on infinite sequences
- Park
- 1981
(Show Context)
Citation Context ...thm; probabilistic labelled transition systems; I. INTRODUCTION In the last three decades a wealth of behavioural equivalences have been proposed in concurrency theory. Among them, bisimilarity [38], =-=[41]-=- is probably the most studied one as it admits a suitable semantics and an elegant co-inductive proof technique. It can also be given an efficient decision procedure [40]. Given a labelled transition ... |

581 |
Markov Decision Processes
- Puterman
- 1994
(Show Context)
Citation Context ...[27], probabilistic processes [28], simple probabilistic automata [43], probabilistic transition systems [29] etc. Furthermore, there are strong structural similarities with Markov Decision Processes =-=[42]-=-, [17]. i∈I 1) ∆ = ∑ i∈I pi · si, where I is a countable index set and ∑ i∈I pi = 1 2) For each i ∈ I there is a state ti such that si R ti 3) Θ = ∑ i∈I pi · ti. Note that in the decomposition of ∆, t... |

403 | Bisimulation through probabilistic testing
- Larsen, Skou
- 1991
(Show Context)
Citation Context ...en the pLTS is finitary. In the probabilistic setting, the definitions of bisimulationlike equivalences are somewhat complicated by the fact that transitions go from states to distributions (see e.g. =-=[31]-=-). So we need to lift relations between states to relations between distributions (see e.g. [12]). Definition 2.2: Given two sets S and T and a relation R ⊆ S ×T. We lift R to a relation R † ⊆ D(S) × ... |

353 |
Three partition refinement algorithms
- Paige, Tarjan
- 1987
(Show Context)
Citation Context ...mong them, bisimilarity [38], [41] is probably the most studied one as it admits a suitable semantics and an elegant co-inductive proof technique. It can also be given an efficient decision procedure =-=[40]-=-. Given a labelled transition system (LTS) with n states and m transitions, the partition refinement algorithm of Paige and Tarjan takes time O(m log n) to generate all bisimulation equivalence classe... |

337 |
Graph algorithms
- Even
- 1979
(Show Context)
Citation Context ...relation R † . The latter can be solved by using network-based techniques, as already observed in [1]. Networks: We briefly recall the basic definitions of networks. More details can be found in e.g. =-=[20]-=-. A network is a tuple N = (N, E, ⊥, ⊤, c) where (N, E) is a finite directed graph (i.e. N is a set of nodes and E ⊆ N × N is a set of edges) with two special nodes ⊥ (the source) and ⊤ (the sink) and... |

265 | Probabilistic simulations for probabilistic processes
- Segala, Lynch
- 1995
(Show Context)
Citation Context ...lacing nondeterministic with probabilistic constructs. The reconciliation of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the literature [48], =-=[44]-=-, [34], [43], [25], [36], [4], [29], [39], [9], [46], [37], [16], [17], [15]. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms fo... |

256 |
Results on the propositional mu–calculus
- Kozen
- 1983
(Show Context)
Citation Context ...ency theory. In the nonprobabilistic setting there is an approach to checking bisimilarity between two states by first constructing a characteristic formula [45] for one state in the modal µ-calculus =-=[30]-=- and check if the other state satisfies the formula. This approach yields efficient algorithms for checking behavioural relations [11]. In the probabilistic setting, characteristic formulae also exist... |

230 |
Modeling and verification of randomized distributed real-time systems
- Segala
- 1995
(Show Context)
Citation Context ...terministic with probabilistic constructs. The reconciliation of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the literature [48], [44], [34], =-=[43]-=-, [25], [36], [4], [29], [39], [9], [46], [37], [16], [17], [15]. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilis... |

95 |
A calculus for communicating systems with time and probabilities
- Hansson, Jonsson
- 1985
(Show Context)
Citation Context ...n probabilistic concurrency theory [23], [8], [32] proceed by replacing nondeterministic with probabilistic constructs. The reconciliation of nondeterministic and probabilistic constructs starts with =-=[24]-=- and has received a lot of attention in the literature [48], [44], [34], [43], [25], [36], [4], [29], [39], [9], [46], [37], [16], [17], [15]. We shall also work in a framework that features the co-ex... |

61 | Quantitative analysis and model checking
- Huth, Kwiatkowska
- 1997
(Show Context)
Citation Context ...istic with probabilistic constructs. The reconciliation of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the literature [48], [44], [34], [43], =-=[25]-=-, [36], [4], [29], [39], [9], [46], [37], [16], [17], [15]. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bi... |

60 |
Testing Probabilistic and Nondeterministic Processes
- Wang, Larsen
- 1992
(Show Context)
Citation Context ...by replacing nondeterministic with probabilistic constructs. The reconciliation of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the literature =-=[48]-=-, [44], [34], [43], [25], [36], [4], [29], [39], [9], [46], [37], [16], [17], [15]. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorit... |

56 | Weak bisimulation for fully probabilistic processes - Baier, Hermanns - 1997 |

47 | 2002): Testing preorders for probabilistic processes can be characterized by simulations. Theoret
- Jonsson, Wang
(Show Context)
Citation Context ...e lift R to a relation R † ⊆ D(S) × D(T) by letting ∆ R † Θ whenever 1 Essentially the same model has appeared in the literature under different names such as NP-systems [27], probabilistic processes =-=[28]-=-, simple probabilistic automata [43], probabilistic transition systems [29] etc. Furthermore, there are strong structural similarities with Markov Decision Processes [42], [17]. i∈I 1) ∆ = ∑ i∈I pi · ... |

41 |
Testing Equivalences and fully abstract models for Probabilistic Processes
- Christoff
- 1990
(Show Context)
Citation Context ... bisimulation. In recent years, probabilistic constructs have been proven useful for giving quantitative specifications of system behaviour. The first papers on probabilistic concurrency theory [23], =-=[8]-=-, [32] proceed by replacing nondeterministic with probabilistic constructs. The reconciliation of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in ... |

40 |
Axiomatizations for probabilistic bisimulation
- Bandini, Segala
- 2001
(Show Context)
Citation Context ...robabilistic constructs. The reconciliation of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the literature [48], [44], [34], [43], [25], [36], =-=[4]-=-, [29], [39], [9], [46], [37], [16], [17], [15]. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity ... |

39 |
Compositional verification of probabilistic processes
- Larsen, Skou
- 1992
(Show Context)
Citation Context ...mulation. In recent years, probabilistic constructs have been proven useful for giving quantitative specifications of system behaviour. The first papers on probabilistic concurrency theory [23], [8], =-=[32]-=- proceed by replacing nondeterministic with probabilistic constructs. The reconciliation of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the li... |

32 | An efficient algorithm for computing bisimulation equivalence
- Dovier, Piazza, et al.
- 2004
(Show Context)
Citation Context ...ransitions, the partition refinement algorithm of Paige and Tarjan takes time O(m log n) to generate all bisimulation equivalence classes. This algorithm belongs to a class of global algorithms (e.g. =-=[19]-=-, [22]), as classified in [10], which require an LTS to be fully generated a priori. However, in many cases, one may be able to determine that one process fails to be related to another by examining o... |

28 | B.: Computing behavioural relations, logically
- Cleaveland, Steffen
- 1991
(Show Context)
Citation Context ...characteristic formula [45] for one state in the modal µ-calculus [30] and check if the other state satisfies the formula. This approach yields efficient algorithms for checking behavioural relations =-=[11]-=-. In the probabilistic setting, characteristic formulae also exist in the probabilistic modal µ-calculus [14]. We believe that it is promising to check probabilistic bisimilarity in a logical way, alo... |

28 |
Characteristic formulae for processes with divergence. Information and Computation 110:149–163
- Steffen, Ingólfsdóttir
- 1994
(Show Context)
Citation Context ...igated in the area of probabilistic concurrency theory. In the nonprobabilistic setting there is an approach to checking bisimilarity between two states by first constructing a characteristic formula =-=[45]-=- for one state in the modal µ-calculus [30] and check if the other state satisfies the formula. This approach yields efficient algorithms for checking behavioural relations [11]. In the probabilistic ... |

25 | Deciding bisimilarity and similarity for probabilistic processes
- Baier, Engelen, et al.
(Show Context)
Citation Context .... We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity and similarity have been studied in [18], [2], =-=[1]-=-, [3], [49] for fully probabilsitic processes and in [26], [5], [1], [49] for nondeterministic probabilistic processes. However, all these algorithms are global because they require the whole state sp... |

25 | Probabilistic and prioritized models of Timed CSP - Lowe - 1995 |

24 | Verifying Bisimulations “On the Fly
- Fernandez, Mounier
- 1990
(Show Context)
Citation Context ...er by examining only a fraction of the state space. One would like to have a verification algorithm that exploits this fact. Another class of algorithms, called local or “on the fly” algorithms (e.g. =-=[21]-=-, [6], [33], [35]), combine the verification of a system’s behaviour with the generation of the system’s state space. Fernandez and Mouier [21] first proposed an “on the fly” algorithm for checking be... |

23 | Axioms for probability and nondeterminism
- Mislove, Ouaknine, et al.
(Show Context)
Citation Context ...c constructs. The reconciliation of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the literature [48], [44], [34], [43], [25], [36], [4], [29], =-=[39]-=-, [9], [46], [37], [16], [17], [15]. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity and similari... |

21 | Decision algorithms for probabilistic bisimulation
- Cattani, Segala
- 2002
(Show Context)
Citation Context ...ence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity and similarity have been studied in [18], [2], [1], [3], [49] for fully probabilsitic processes and in [26], =-=[5]-=-, [1], [49] for nondeterministic probabilistic processes. However, all these algorithms are global because they require the whole state space of a system to be fully generated in advance. In [47] a lo... |

18 | Semantic domains for combining probability and nondeterminism
- Tix, Keimel, et al.
- 2005
(Show Context)
Citation Context ...s. The reconciliation of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the literature [48], [44], [34], [43], [25], [36], [4], [29], [39], [9], =-=[46]-=-, [37], [16], [17], [15]. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity and similarity have bee... |

17 |
On some equivalence relations for probabilistic processes
- Huynh, Tian
- 1992
(Show Context)
Citation Context ...-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity and similarity have been studied in [18], [2], [1], [3], [49] for fully probabilsitic processes and in =-=[26]-=-, [5], [1], [49] for nondeterministic probabilistic processes. However, all these algorithms are global because they require the whole state space of a system to be fully generated in advance. In [47]... |

16 | Can a maximum flow be computed in o(nm) time
- Cheriyan, Hagerup, et al.
- 1990
(Show Context)
Citation Context ...ns ∆, Θ ∈ D(S) and R⊆ S × S Output: If ∆ R † Θ then “yes” else “no” Method: Construct the network N(∆, Θ, R) Compute the maximum flow F in N(∆, Θ, R) If F < 1 then return “no” else “yes”. As shown in =-=[7]-=-, computing the maximum flow in a network can be done in time O(n 3 / log n) and space O(n 2 ), where n is the number of nodes in the network. So we immediately have the following result. Lemma 3.3: T... |

15 | Optimal state-space lumping in Markov chains
- DERISAVI, HERMANNS, et al.
- 2003
(Show Context)
Citation Context ... [17], [15]. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity and similarity have been studied in =-=[18]-=-, [2], [1], [3], [49] for fully probabilsitic processes and in [26], [5], [1], [49] for nondeterministic probabilistic processes. However, all these algorithms are global because they require the whol... |

14 |
Equivalence and preorder checking for finitestate system
- Cleaveland, Sokolsky
- 2001
(Show Context)
Citation Context ...nement algorithm of Paige and Tarjan takes time O(m log n) to generate all bisimulation equivalence classes. This algorithm belongs to a class of global algorithms (e.g. [19], [22]), as classified in =-=[10]-=-, which require an LTS to be fully generated a priori. However, in many cases, one may be able to determine that one process fails to be related to another by examining only a fraction of the state sp... |

14 | Characterising testing preorders for finite probabilistic processes
- Deng, Glabbeek, et al.
- 2007
(Show Context)
Citation Context ... nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the literature [48], [44], [34], [43], [25], [36], [4], [29], [39], [9], [46], [37], [16], [17], =-=[15]-=-. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity and similarity have been studied in [18], [2], ... |

14 | An expectation-based model for probabilistic temporal logic
- Morgan, McIver
- 1999
(Show Context)
Citation Context ...with probabilistic constructs. The reconciliation of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the literature [48], [44], [34], [43], [25], =-=[36]-=-, [4], [29], [39], [9], [46], [37], [16], [17], [15]. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimila... |

13 | W.: Testing and refinement for nondeterministic and probabilistic processes
- Jonsson, Ho-Stuart, et al.
- 1994
(Show Context)
Citation Context ...d T and a relation R ⊆ S ×T. We lift R to a relation R † ⊆ D(S) × D(T) by letting ∆ R † Θ whenever 1 Essentially the same model has appeared in the literature under different names such as NP-systems =-=[27]-=-, probabilistic processes [28], simple probabilistic automata [43], probabilistic transition systems [29] etc. Furthermore, there are strong structural similarities with Markov Decision Processes [42]... |

12 | Probabilistic temporal logics via the modal mu-calculus
- Narasimha, Cleaveland, et al.
- 1999
(Show Context)
Citation Context ...tructs. The reconciliation of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the literature [48], [44], [34], [43], [25], [36], [4], [29], [39], =-=[9]-=-, [46], [37], [16], [17], [15]. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity and similarity ha... |

12 | D.: Flow Faster: Efficient Decision Algorithms For Probabilistic Simulations
- Zhang, Hermanns, et al.
- 2008
(Show Context)
Citation Context ... also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity and similarity have been studied in [18], [2], [1], [3], =-=[49]-=- for fully probabilsitic processes and in [26], [5], [1], [49] for nondeterministic probabilistic processes. However, all these algorithms are global because they require the whole state space of a sy... |

8 |
Remarks on testing probabilistic processes
- Deng, Glabbeek, et al.
(Show Context)
Citation Context ...ciliation of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the literature [48], [44], [34], [43], [25], [36], [4], [29], [39], [9], [46], [37], =-=[16]-=-, [17], [15]. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity and similarity have been studied in... |

7 |
Probabilistic weak simulation is decidable in polynomial time
- Baier, Hermanns, et al.
(Show Context)
Citation Context ...shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity and similarity have been studied in [18], [2], [1], =-=[3]-=-, [49] for fully probabilsitic processes and in [26], [5], [1], [49] for nondeterministic probabilistic processes. However, all these algorithms are global because they require the whole state space o... |

7 | Scalar outcomes suffice for finitary probabilistic testing
- Deng, Glabbeek, et al.
- 2007
(Show Context)
Citation Context ...ion of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the literature [48], [44], [34], [43], [25], [36], [4], [29], [39], [9], [46], [37], [16], =-=[17]-=-, [15]. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity and similarity have been studied in [18],... |

4 |
Probabilistic barbed congruence
- Deng, Du
- 2007
(Show Context)
Citation Context ...alences are somewhat complicated by the fact that transitions go from states to distributions (see e.g. [31]). So we need to lift relations between states to relations between distributions (see e.g. =-=[12]-=-). Definition 2.2: Given two sets S and T and a relation R ⊆ S ×T. We lift R to a relation R † ⊆ D(S) × D(T) by letting ∆ R † Θ whenever 1 Essentially the same model has appeared in the literature und... |

3 | Symbolic bisimulations for probabilistic systems
- Wu, Palamidessi, et al.
- 2007
(Show Context)
Citation Context ...[26], [5], [1], [49] for nondeterministic probabilistic processes. However, all these algorithms are global because they require the whole state space of a system to be fully generated in advance. In =-=[47]-=- a local algorithm in the style of [33] is proposed to decide probabilistic bisimilarity, but it is tailored to check equivalence relations and thus cannot handle probabilistic similairty which is onl... |

2 |
Semantic Preorders in the Automated Verification of Concurrent Systems
- Celikkan
- 1995
(Show Context)
Citation Context ...examining only a fraction of the state space. One would like to have a verification algorithm that exploits this fact. Another class of algorithms, called local or “on the fly” algorithms (e.g. [21], =-=[6]-=-, [33], [35]), combine the verification of a system’s behaviour with the generation of the system’s state space. Fernandez and Mouier [21] first proposed an “on the fly” algorithm for checking behavio... |

2 | Characterising probabilistic processes logically
- Deng, Glabbeek
- 2010
(Show Context)
Citation Context ... the formula. This approach yields efficient algorithms for checking behavioural relations [11]. In the probabilistic setting, characteristic formulae also exist in the probabilistic modal µ-calculus =-=[14]-=-. We believe that it is promising to check probabilistic bisimilarity in a logical way, along the line of consideration proposed in [11]. There are other local algorithms for checking nonprobabilistic... |

2 |
2007): Results on the Quantitative Mu-Calculus
- McIver, Morgan
(Show Context)
Citation Context ... reconciliation of nondeterministic and probabilistic constructs starts with [24] and has received a lot of attention in the literature [48], [44], [34], [43], [25], [36], [4], [29], [39], [9], [46], =-=[37]-=-, [16], [17], [15]. We shall also work in a framework that features the co-existence of probability and nondeterminism. Decision algorithms for probabilistic bisimilarity and similarity have been stud... |

1 |
Algorithm for verifying strong open bisimulation in full pi-calculus
- Deng, Fu
(Show Context)
Citation Context ...eover, the basic idea illustrated above is also applicable to checking similarity. In [33] Lin lifted Fernandez and Mouier’s algorithm to handle value-passing processes. The idea was later on used in =-=[13]-=- to verify open bisimulation. In recent years, probabilistic constructs have been proven useful for giving quantitative specifications of system behaviour. The first papers on probabilistic concurrenc... |

1 |
Bisimulation minimization and symbolic model checking. Formal Methods in System Design
- Fisler, Vardi
- 1990
(Show Context)
Citation Context ...ions, the partition refinement algorithm of Paige and Tarjan takes time O(m log n) to generate all bisimulation equivalence classes. This algorithm belongs to a class of global algorithms (e.g. [19], =-=[22]-=-), as classified in [10], which require an LTS to be fully generated a priori. However, in many cases, one may be able to determine that one process fails to be related to another by examining only a ... |

1 | On-the-fly instantiation” of value-passing processes
- Lin
- 1998
(Show Context)
Citation Context ...ning only a fraction of the state space. One would like to have a verification algorithm that exploits this fact. Another class of algorithms, called local or “on the fly” algorithms (e.g. [21], [6], =-=[33]-=-, [35]), combine the verification of a system’s behaviour with the generation of the system’s state space. Fernandez and Mouier [21] first proposed an “on the fly” algorithm for checking behavioural e... |

1 | Improved on-the-fly equivalence checking using boolean equation systems
- Mateescu, Oudot
- 2008
(Show Context)
Citation Context ...nly a fraction of the state space. One would like to have a verification algorithm that exploits this fact. Another class of algorithms, called local or “on the fly” algorithms (e.g. [21], [6], [33], =-=[35]-=-), combine the verification of a system’s behaviour with the generation of the system’s state space. Fernandez and Mouier [21] first proposed an “on the fly” algorithm for checking behavioural equival... |