Temporal Logic of Stochastic Actions for Verification of
Probabilistic Systems
LI Jun-tao
Information College
Guizhou University of Finance and Economic
Guiyang, china
E-mail: jtxq@qq.com
LONG Shi-gong,
Computer Science and Technology College
Guizhou University
Guiyang, china
E-mail: 526796467@qq.com
Abstract-The specification and verification
of probabilistic systems were usually based
on Computational Tree Logic, and systems
and properties were specified by different
language respectively. This paper extends
and reforms Temporal Logic of Actions,
puts foreword Temporal Logic of Stochastic
Actions (TLSA), which can use additional
state-action probabilistic distribution and
probabilistic operator to specify
probabilistic systems and their properties in
the same logic.
Keywords-Specifying systems; System
verification; TLA; Probabilistic systems
I. I
NTRODUCTION
As soon as the Model Checking was
invented in 1980’s, researchers had started
applying it to the study of Probabilistic
Systems’ verification. In the beginning, people
focus on the qualitative properties of system,
for example, the program is whether or not
terminating with probability 1. Afterwards, the
algorithms for verifying quantitative properties
have been progressed also. At present, the
verification technology of Probabilistic
Systems is mainly used for the field of security,
distributed algorithms, systems biology, and
system performance analysis, and so on.
In past thirty years, the implementation
model of Probabilistic Systems is mainly based
on Markov decision processes [1][2](MDPs),
there are also some others import timed
automata[6], putdown automata[7], or
two-player game. They specify the property of
system with linear temporal logic (LTL),
-regular properties or probabilistic
computation tree logic (PCTL), the latter
imports probability distribution to computation
tree logic (CTL); it is most used description
language of Probabilistic Systems. Obviously,
the traditional research method used different
description language to specify the models or
properties of Probabilistic Systems; this is not
well for the property verification and design
implementation of system.
This paper puts forward Temporal Logic of
Stochastic Actions (TLSA), it is extension of
Temporal Logic of Actions [4] (TLA) with
probability, the latter is based on linear
temporal logic, it defined the actions and
operators, and can achieve specifying and
verification of concurrent systems and their
properties; the former inherit the most
prominent feature of TLA: system and its
properties can be described using TLSA at the
same time.
II.
PROBABILISTIC TRANSITION SYSTEM
Probabilistic Transition Systems (PTSs) is
abstract model of Probabilistic Systems, its
definition is followed:
Defination1.1 Probabilistic Transition
systems is a 5-tuple:(= {+,!,, ,
},among them:
+˖States set of system;
!˖Initial states set of system;
˖Actions set of system;
˖ +×+relationship of states
trasation;
˖˄+×+ ˅ [0,1] is probability
distribution of system actions, and meet the
condition:
s∀∈+ , (, , ') 1
A
psAs
∈
=
¦
.
We can see that probabilistic transition
systems are Label Transition Systems (LTS)
adding a probability distribution of actions.
III. S
YNTAX AND SEMANTICS OFTLSA
3.1 Syntax
TLSA’s symbol includes:
˄1˅Probabilistic values: pr(ę[0,1]);
˄2˅Constant symbol: c, c
1
, c
2
…
˄3˅Rigid variables: u, u
1
, u
2
…
˄4˅Flexible variable: x, x
1
, x
2
…
˄5˅Atomic proposition: p, p
1
, p
2
…
˄6˅Constant element symbol: m, m
1
,
m
2
…
˄7˅Arithmetic operator: +,-, *;
˄8˅Relational operators: <, =;
˄9˅logical operator:
∧ , ¬ ,
><pr;
˄10˅Quantifier:
∃ ˗
˄11˅Temporal operator˖, [].
The other conjunctions, for example Ģ,
2015 14th International Symposium on Distributed Computing and Applications for Business Engineering and Science
978-1-4673-6593-2/15 $31.00 © 2015 IEEE
DOI 10.1109/DCABES.2015.23
62