scispace - formally typeset

Proceedings ArticleDOI

Modified M-timed Petri nets in modelling and performance evaluation of systems

01 Feb 1987-pp 261-268

TL;DR: It is shown that the behavior of modified free-choice M-timed Petri nets can be represented by probabilistic state graphs, stationary probabilities of states can thus be obtained by standard techniques used for analysis of continuous-time homogeneous Markov chains.
Abstract: Modified M-timed Petri nets are Petri nets with exponentially distributed firing times and with generalized inhibitor arcs to interrupt firing transitions. It is shown that the behavior of modified free-choice M-timed Petri nets can be represented by probabilistic state graphs, stationary probabilities of states can thus be obtained by standard techniques used for analysis of continuous-time homogeneous Markov chains. An immediate application of such a model is performance analysis of queueing systems with exponentially distributed service and interarrival times, and with priority and/or preemptive scheduling disciplines. Simple models of computer systems with different scheduling strategies are used as an illustration of modelling and performance analysis.
Topics: Stochastic Petri net (69%), Petri net (67%), Queueing theory (51%), Markov chain (50%)

Content maybe subject to copyright    Report

15-th ACM Annual Computer Science Conference (CSC’87); St. Louis, MO, 1987, February 17–19, 1987, pp.261–268.
Copyright
c
1987 ACM (DOI 10.1145/322917.322959).
MODIFIED M-TIMED PETRI NETS IN MODELLING
AND PERFORMANCE EVALUATION OF SYSTEMS
W.M. Zuberek
Department of Computer Science, Memorial University
St.John’s, NL, Canada A1B 3X5
Abstract
Modified M-timed Petri nets are Petri nets with ex-
ponenti ally distributed firing times and with generalized
inhibitor arcs t o interrupt firing transitions. It is shown
that the behavior of modified free-choice M-timed Petri
nets can be represented by probabilis t ic state graphs,
stationary probabilities of states can thus be obtained
by standard techniques used for analysis of continuous-
time homogeneous Markov chains. An immediate appli-
cation of such a model is performance analysis of queue-
ing systems with exponentially distributed service and
inte r arr i val times, and with priority and/or preemptive
sch ed uli ng disciplines. Simple models of computer sys-
tems with different scheduling strategies are us ed as an
illustration of modelling and performance analysis.
1. INTRODUCTION
Petr i nets [1,7,11] have been successfully used in mod-
elling, validation and analysis of systems of events in
which it is possible for some events to occur concur-
rently, but there are constraints on the occ ur r e nc e,
precedence, or frequency of these occurrences [6,11]. Ba-
sic Petri nets, however, are not complete enough for the
study of systems p e r for mance since no assumption is
made on the duration of systems events. Timed Petri
nets have been introduced by Ramchandani [12] by as-
signing ”firing times” to the transitions of Petri nets.
Sifakis [14] proposed another definition of a timed Petri
net by assigning ”enabling times” to places of a net.
Merlin and Farber [9] discussed timed Petri nets where
a time threshold and maximum delay wer e assigned to
each transition of a net to allow modelling of timeouts
used to recover from failur es in communication sy s tems .
Razouk [13] discussed yet another class of timed Petri
nets with enabling as well as firing times, and demon-
strated the derivation of (symbolic) performance expres-
sions for communication protocols. Another approach
to Petri nets ”with time” assumes that the firing times
are exponentially distributed random variables, and the
corresponding rates are assigned to transitions of a net;
such nets are called stochastic Petri nets [2,10]. In gen-
eralized stochastic Petri nets [2] the set of transitions is
subdivided into two classes of transitions, timed and im-
mediate ones. Timed transitions (as in stochastic nets)
have exponenti ally distributed firing times, while imme-
diate transitions fire in zero time, i.e., they are used
to represent logical conditions which do not contribute
to ”delay” time s. The memoryless property of the ex-
ponenti al distribution simplifies the ”state” description
and analysis of such nets , and, in fact, analysis of (basic
and generalized) stochastic nets is based on the sets of
reach able markings which are generated without timing
constraints. Consequently, the stochastic appr oach can
be used for only such models in which the state space is
isomorphic to the s pace of reachable markings [19], and
this may be a nontrivial task to check.
The formalism des cr i bed in this paper is a continua-
tion of the approach originated by Ramchandani [12,16]
and subsequently extended by multiple and inhibitor
arcs [17,18]. In (basic and extended) M-timed Petri nets,
however, firing transitions cannot be inte r r up te d, and
therefore preemptive disciplines are difficu lt to model
and analyze. In this paper, the basic Petri nets are en-
hanced by ”interrupt” arcs in ord er to suspend the pro-
cess of transition firi ng, as required in strict modelling of
preemptions. Similarly as in [2,10,17,18,19], firing rates
of exponentially distributed firing times are assigned to
transitions of a Petri net, and a ”state” description is
derived which represents the behavior of modifi ed free-
choice M-timed Petr i nets by continuous-time homo-
geneous Markov chains. The stationary probabilities
of the states can thus be obtained by standard tech-
niques, and this pr ovides many pe r for mance measures
such as utilization of systems components, average wait-
ing times and turnaround times or average throughput
rates, which can be derived automaticall y from model
specifications.
This paper is organized in 3 main sections. Section 2
contains definitions of basic concepts for modified free-
choice Petri nets. Modified M-timed Petri nets are in-
troduced in Section 3. Application of modified M-timed
Petr i nets to modelling and performance evaluation is
discussed in Section 4.
2. MODIFIED PETRI NETS
A modified Petr i net N is a quadruple N =
(P, T, A, C) where:
P is a finite, nonempty set of places,

Modified M-t imed Petri nets in modelling and performance evaluation of systems 262
T is a finite, nonempty set of transitions,
A is a nonempty set of directed arcs which connect
places with transitions and transitions with places
such that for e ach transition there is at least one
place connec te d with it
t T p P : (p, t) A,
C is a (possi bl y empty) set of interrupt arcs which con-
nect places with transitions, C P × T , and A and
C are disjoint sets.
A place p is an input (or an output) place of a transi-
tion t iff there exists an arc (p, t) (or ( t, p), respectively)
in the set A. The sets of all input and output places of
a transition t are denoted by Inp(t) and Out(t) , respec -
tive ly. S imi larl y, the sets of input and output transitions
of a place p are denoted by Inp(p) and Out(p). Also,
a place p is an interrupting place of a transition t iff
(p, t) C. The set of all interrupting places of t is de-
noted by Int(t), and the set of transitions connected
by interrupt arcs with a place p is denoted by Int(p),
Int(p) = {t|p Int(t)}. The notation is extended on
sets of places and transitions in the usual way.
A modified net N = (P, T, A, C) is singular iff the
input sets of transi ti ons with nonempty interrupting sets
are disjoint with interr up ti ng sets of other transitions
t T : Int(t) = Int(Inp(t)) =
where denotes the empty set. The description of
non-singular nets must take into account ”sequences”
of consecutive interrupts in a net when one interrupted
transition, through its input places, interrupts another
transition(s). In singular nets t he r e is no such ”propa-
gation” of inte r r u pt s.
A marked Petri net M is a pair M = (N, m
0
) where:
N is a modified Petri net, N = (P, T, A, C),
m
0
is an initial marking func tion which assigns a non-
negative integer number of so called tokens t o each
place of the net, m
0
: P {0, 1, ...}.
Let any function m : P {0, 1, ...} be called a mark-
ing in a net N = ( P, T, A, C).
A transition t is enabled by a marking m iff every
input place of this transition contains at least one token
and every interrupting place of t contains zero tokens.
For ordinary nets (i.e., nets without time) the interrupt
arcs are thus equivalent to inhibitor arcs [1,11,17].
The set of all transitions enabled by a marking m is
denoted by En(m).
A place p is shared iff it is an inpu t place for more
than one transition. A shared place p is guarded iff for
each two different transitions t
i
and t
j
sharing p there
is another place p
k
such that p
k
is in the input set of
one of these t r ans iti ons and in the interrupt in g set of
the other one
t
i
Out(p) t
j
Out(p) {t
i
} p
k
P {p} :
((p
k
, t
i
) A (p
k
, t
j
) C) ((p
k
, t
j
)
A (p
k
, t
i
) C),
i.e., no two transitions from the set Out(p) can be en-
abled by the same marking m. A net is conflict-free iff
all its shared p lace s are guarde d.
A shared place p is free-choice (or extended free-choice
[7]) iff the input sets and i nterrupting sets of all transi-
tions shar in g p are ident ic al, i.e., iff:
t
i
, t
j
Out(p)) : Inp(ti) = Inp(t
j
)Int(t
i
) = Int(t
j
).
A net is free-choice iff all i ts s h are d place s are either
guarded or free-choice. Only free-choice Petri nets are
considered in this paper.
Ever y transition enabled by a marking m can fire.
When a transition fires, a token is r emoved from each
of its input places (but not from interr u pt in g places),
and a token is added to each of its output places. This
determines a new marking in a net, a new set of en able d
transitions, and so on.
A marking m
j
is directly reachable (or t
k
-reachable)
from a marking m
i
in a net N iff there exists a transition
t
k
enabled by the marking m
i
, t En(m
i
), such that
p P : m
j
(p) =
m
i
(p) 1, if p Inp(t) Out(t),
m
i
(p) + 1, if p Out(t) Inp(t),
m
i
(p), otherwise.
A marking m
j
is (generally) reachable from a marking
m
i
in a net N if ther e exists a sequence of markings
(m
i
0
m
i
1
m
i
2
...m
i
k
) such that m
i
0
= m
i
, m
i
k
= m
j
, and
each marking m
i
is directly reachable from the marking
m
i
1
for = 1, ..., k.
A set M(M) of reachable markings of a marked Petri
net M = (N, m
0
) is the set of all markings whi ch are
reach able from the initial marking m
0
(including m
0
).
3. MODIFIED M-TIMED PETRI NETS
In timed Petri nets , each transition t takes a ”real”
time to fir e , and therefore it is convenient to distinguish
three phases of transition firings, its initiation, delay,
and termination. When a transition t is enabled, a fir-
ing can be initiated by removing a token from each of
t’s input places. Th is token remains in the transition t
for the ”firing time”, and then the firing ter min ates by
adding a token to each of t’s outpu t places. Each of the
firings is initiated in the same instant of time in which it
becomes enabled. If a transit ion be comes e nab led wh ile
it fires, a new, indepe nd ent firing can be initiated. If
a net contains conflic ts , and there are several different
possibilities of firing transitions for the same marking,

Modified M-t imed Petri nets in modelling and performance evaluation of systems 263
the choice of actual transitions is assumed to be a ran-
dom pr oces s which can be described by corresponding
probabilities or probability distribution functions. In
modified timed Petri nets, a firing of a transition may
be interrupted i f the set of transition interr u pt in g places
becomes nonempty. If, d uring a firing period of such
transition t, all t’s interruptin g places become marked,
the firing of t ceases and the tokens removed from t’s
input places at the beginning of firing, are ”returned”
to their original places. It should be noticed that the
inte r r up ti ng places of t must be empty to initiate a fir-
ing of t; an interrupt can thus occur only as a result of
termination of another firing (or firings).
Since in timed nets all transition firings are initiated
in the same instants of ti me in which the transitions
become enabled, it is convenient to associate with each
marking m a set of all possibilities of new firings (in
nets with conflicts there are usually several such possi-
bilities). The set of selection functions describes all such
possibilities.
A selection function e of a marking m in a net N is
any function e : T {0, 1, ...} such that
(1) there exists a sequence of transitions u =
(t
i
1
, t
i
2
, ..., t
k
) in which t
i
j
En(m
i
j1
) for j =
1, ..., k, and for m
i
0
= m where
p P : m
i
j
(p) = m
i
j1
(p)
1, if p Inp(t
i
j
),
0, otherwise,
(2) the set of transitions enabled by m
i
k
, En(m
i
k
), is
empty,
(3) for each tinT , e(t) is equal to the number of occur-
rences of t in the sequen ce u;
i.e., a selection function e indicates (by nonzero values)
all those transitions which can initiate their firings si-
multane ous ly (and some transitions may initiate several
firings). The set of all selection functi ons of a marking
m is denoted by Sel(m).
A marked net M is simple if all selection f un ct ions of
all reachable markings assign at most one firing to each
transition of M, i.e.
m M(M) e Sel(m)) t T : e(t) 1.
Only simple nets are discussed in this paper. Non-
simple nets can be described by a straightforward ex-
tension of simple nets [ 18,19] that takes into account
combination s of multiple firings and their corresponding
probabilities. It should be noticed that in M-timed nets
without multiple arcs, non-simple behaviour can occur
only in the initial states, when the initial marking func-
tion introduce s ”multiple enablings” of some transitions.
Moreover, such initial markings can usually be converted
to equivalent ones, which preserve the behaviour of a
net, and which provide simple selection fu nc tion s.
An M-timed modified free-choice Petri net T is a
triple T = (M, c, r) where:
M is a modified free-choice marked Pe tr i net, M =
(N, m
0
), N = (P, T, A, C),
c is a choice function which assigns a free -choice prob-
ability to e ach transition t of the net in such a way
that for each free-choice place p:
X
tOut(p)
c(t) = 1,
and for all remaining transitions c(t) = 1,
r is a firing rate function which assigns a positive
real number r(t) to each transition t of the net,
r : T > tarrowR
+
, and R
+
denotes the set of
positive real numbers; the firing time of a transi-
tion t is a random variable v(t) with the distribution
function
Prob(v(t) > x) = e
xr(t)
, x > 0.
The memoryless (or Markov) prop er ty of the (nega-
tive ) exponential distribution is t he basic factor in de-
scriptions of M-timed Petri nets (the exponential dis-
tribution is the only continuous distr i bu tion with the
memoryless prope r ty). It means that if the duration v
of a certain activity (e.g., the firing time) is distributed
exponentially with parameter r, and if that activity is
observed at time y after its beginning, then the remain-
ing duration of the activity is independe nt of y and is
also distr ib ut ed exponentially with parameter r:
Prob(v > y + x|v > y) = Prob(v > x) = e
rx
.
Consequently, the state descriptions of M-timed nets
represent only the actual configuration of a net, and
completely igmores t he history, or the sequence of pre-
vious state s .
The behavior of an M-timed Petri net can be de-
scribed by a set of states (or configurations of a net),
with state transitions corr es pond in g to termin ations of
firings (and initiations of new firings).
A state ss@ of an M-timed Petri net T is a pair of
functions s = (m, f) where:
m is a marking function, m : P {0, 1, ...},
f is a firing function which indicates (for each transi-
tion of the net) the number of active firings, i.e.,
the number of firings which have been initiated but
are not yet terminated, f : T rightarrow{0, 1, ...}.

Modified M-t imed Petri nets in modelling and performance evaluation of systems 264
An initial state s of a net T is a pair s = (m) i, f
i
)
where f
i
is a s ele ct ion function from the set Sel(m
0
),
and the marking m
i
is defined by
p P : m
i
(p) = m
0
(p)
X
tOut(p)
f
i
(t).
A free-choice M-timed net T may have several differ-
ent initial states.
A state s
j
= (m
j
, f
j
) is directly reachable (or (t
k
, e
)-
reach able ) from t he state s
i
= (m
i
, f
i
) iff:
1. f
i
(t
k
) > 0,
2. e
Sel(m
ikj
),
3. p P ) : m
j
(p) = m
ikj
(p)
P
tOut(p)
e
(t),
4. t T : f
j
(t) = f
ik
(t) d
ik
(t) + e
(t),
where
5. p P : m
ikj
(p) = m
ik
(p) +
P
tOut(p)
d
ik
(t),
6. t T : d
ik
(t) = min(f
ik
(t), min
pInt(t)
(m
ik
(p)).
7. p P : m
ik
(p) = m
i
(p) +
1, if p Out(t
k
),
0, otherwise,
8. t T : f
ik
(t) = f
i
(t)
1, if t = t
k
,
0, otherwise.
The state s
j
which is (t
k
, e
)-reachable from the state
s
i
is thus obt aine d by the termination of a t
k
firing
(1), updating the marking and firing functions (7,8),
performing all the interrupts (if any) of firing transi-
tions which correspond to the updated marking, mik (5
and 6), and then initiating new firings (if any) which
are determined by a selection function e
from the set
Sel(m
ikj
) (2,3 and 4). It should be noticed that the for-
mula (6) is a direct consequence of singularity and does
not take ”propagation” of interrupts into account. For
nonsingular nets, the description of state reachability
becomes a little mor e convoluted .
Similarly as for marked nets, a state s
j
is (generally)
reach able from a state s
i
if there is a sequence of directly
reach able states from the state s
i
to the state s
j
. Also,
the set S(T) of reachable states is defined as the set of
all states of a net T which are reach able from the initial
states of the net T (incl ud in g the initial state(s)).
A state graph G of an M-timed Petri net T is a labeled
directed graph G(T) = (V, D, u) where:
V is a set of vertices which is equal to the set of reach-
able states of the net T, V = S(T),
D is a set of directed arcs, D V ×V , such that (s
i
, s
j
)
is in D iff s
j
is direct ly reachable from s
i
,
u is a transition-rate function which assigns the rate of
transition from s
i
to s
j
to each arc (s
i
, s
j
) in the
set D, u : D R
+
, in such a way that if s
j
is
(t
k
, e
)-reachable fr om s
i
and s
i
= (m
i
, f
i
), then
u(s
i
, s
j
) = r(t
k
) f
i
(t
k
)
Y
tT
c(t)
e
(t)
.
It should be observed that the state graph of a mod-
ified free-choice M-timed Petri net is a continuous-time
homogeneous Markov chain. The stationary probabil-
ities x(s) of the states s S(T) can be obtained by
solving a system of simultaneous linear equations [5,8]
X
(s
j
,s
i
)D
u(s
j
, s
i
) x(s
j
) = x(s
i
)
X
(s
i
,s
j
)D
u(s
i
, s
j
);
for i = 1, ..., K 1
X
1iK
x(s
i
) = 1
where K is the number of states in the set S(T).
Example. Th e M-timed Petri net shown in Fig.1a (as
usual, places are represented by circles, transitions by
bars, interru pt arcs by small dots instead of arrowheads,
the initial marking by dots inside places, and the firing
rate function and the choice function are given as an
additional descriptions of transiti ons ) net contains one
guarded place, p
1
. The net is singular since it contains
only one transition (t
3
) with nonempty interrupting set.
The choice function, c, assigns the probability 1 to all
transitions since this net does not c ontain free-choice
places.
The state graph G(T
1
) is shown in Fig.1b, and the
derivation of the set of reachable states S(T
1
) is given
in Tab.1 which also contains the stationary probabilities
x(s
i
) of the states s
i
S(T
1
). .
Fig.1. M-timed net T
1
(a) and its state graph ( b) .

Modified M-t imed Petri nets in modelling and performance evaluation of systems 265
m
i
f
i
d
ik
m
ikj
e
s
i
x(s
i
) 1 2 3 4 5 1 2 3 4 t
k
1 2 3 4 1 2 3 4 5 1 2 3 4 s
j
u(s
i
, s
j
)
1 0.184 0 0 0 2 0 1 1 0 0 1 0 0 0 0 1 0 1 2 0 0 1 1 0 0 2 5.0
2 0 0 0 0 0 1 0 2 0 0 0 0 0 3 2.0
2 0.184 0 0 0 1 0 0 2 1 0 2 0 0 1 0 1 1 0 2 0 1 0 0 0 1 4.0
3 0 0 0 0 1 0 0 1 1 0 0 1 1 4 2.0
3 0.084 0 1 0 2 0 1 0 0 0 1 0 0 0 0 1 1 1 2 0 1 1 0 0 1 5.0
4 0.180 0 0 0 0 0 0 2 1 1 2 0 0 1 0 1 1 0 1 0 1 0 0 0 5 4.0
3 0 0 0 0 1 0 0 0 1 0 0 0 1 6 2.0
4 0 0 0 0 0 0 0 1 0 0 0 0 0 2 1.0
5 0.136 0 0 0 1 0 1 1 0 1 1 0 0 0 0 1 0 1 1 0 0 1 1 0 4 5.0
2 0 0 0 0 0 1 0 1 0 0 0 0 0 7 2.0
4 0 0 0 0 0 0 0 2 0 0 0 0 0 1 1.0
6 0.107 1 0 0 0 0 0 2 0 2 2 0 0 0 0 1 1 0 0 0 1 0 0 0 8 4.0
4 0 0 0 0 1 0 0 1 0 0 0 1 0 4 2.0
7 0.051 0 1 0 1 0 1 0 0 1 1 0 0 0 0 1 1 1 1 0 1 1 0 0 5 5.0
4 0 0 0 0 0 1 0 2 0 0 0 0 0 3 1.0
8 0.057 0 0 0 0 0 1 1 0 2 1 0 0 0 0 1 0 1 0 0 0 1 0 0 6 5.0
2 0 0 0 0 0 1 0 0 0 0 0 0 0 9 2.0
4 0 0 0 0 0 0 0 1 0 0 0 0 0 5 2.0
9 0.016 0 1 0 0 0 1 0 0 2 1 0 0 0 0 1 1 1 0 0 1 1 0 0 8 5.0
4 0 0 0 0 0 1 0 1 0 0 0 0 0 7 2.0
Tab.1. The set of reachable states for T
1
.
4. PERFORMANCE EVALUATION
A very simple closed-network model of an interactive
system with 2 class e s of users (and j obs ) and with a p r i-
ority preemptive s cheduling is shown in Fig.2a. It con-
tains one ce ntral server P
c
with two queues of waiting
jobs, Q
1
and Q
2
(for class-1 and class-2 jobs, respec-
tive ly) , and n
1
(active) users in class-1 and n
2
(active)
users in class-2. All class-1 jobs have higher priority
than the class-2 ones, i.e., the class-1 jobs rece ive the
service before the jobs from class-2.
Fig.2. Closed network model os an interactive systems (a)
and its transition-rate diagram (b).
Suppose that the jobs of the same class are s tati st i-
cally i de ntical, that they are served by the First-Come-
First-Served discipline, and that the service times as well
as the user terminal (or ”thinkin g”) times are exponen-
tially distributed. Under these assumptions the number
of jobs in th e system (i.e., in the server and its waiting
queues) is a finite continuous-time homogeneous Markov
chain [5,8]. For n
1
= n
2
= 2, there are 9 states of this
Markov chain:
0: no jobs in the system;
1: a class-1 job in P
c
, empty queues;
2: a class-2 job in P
c
, empty queues;
3: a class-1 job in P
c
, 1 job in the queue Q
1
;
4: a class-1 job in P
c
, 1 job in the queue Q
2
;
5: a class-2 job in P
c
, 1 job in the queue Q
2
;
6: a class-1 job in P
c
, 1 job in Q
1
, 1 job in Q
2
;
7: a class-1 job in P
c
, 2 jobs in the queue Q
2
;
8: a class-1 job in P
c
, 1 job in Q
1
, 2 jobs in Q
2
.
The corresponding transition-rate diagram [6] is
shown in Fig.2b where a
1
, a
2
, d
1
and d
2
denote the
terminal rates for class-1 and class-2, and the service
rates for class-1 and class-2 j obs , respectively.
The same system can be modeled by the M-timed
Petr i net sh own in Fig.1a. The transitions t
1
and t
3
correspond to the central server processing class-1 jobs
(t
1
) and class-2 jobs (t
3
) with the service rates (or the
firing rates) equal to 5 and 2, respectively. The places
p
2
and p
4
model the waiting queues (Q
1
and Q
2
, re-
spectively). The transitions t
2
and t
4
correspond to the

Citations
More filters

Journal ArticleDOI
TL;DR: A new net which is called Extended High Level Timed Petri Nets (EHLTPN) is introduced and it is shown how to extend High Level Petri nets (HLPN) with time, actions and interrupt mechanism to create the compact and comprehensive models for distributed embedded real-time systems.
Abstract: Many extensions of Petri nets have been proposed to model the behaviors and time relations of embedded system, yet these models are all based on some assumptions about the behaviors of embedded systems. Especially they all do not have the actual ability to model interrupt mechanism of embedded system. A new net which is called Extended High Level Timed Petri Nets (EHLTPN) is introduced in this study. It shows how to extend High Level Petri Nets (HLPN) with time, actions and interrupt mechanism. Interruptible subnets corresponding to different interruptible resources are introduced to model the behaviors and time relations of distributed embedded real-time systems. Each interruptible subnet realizes interrupt mechanism by an Interrupt Switch Transition and a set of Resuming Transitions. We give an informal description of this new model and show how this model be formally defined. A transform rule presented shows that each subnet corresponding to an interruptible resource in EHLTPN can be transformed into a behaviorally equivalent subnet of HLTPNAT with priority. This model makes it possible to create the compact and comprehensive models for distributed embedded real-time systems.

4 citations


Cites background from "Modified M-timed Petri nets in mode..."

  • ...Zuberek([15]) in Modified M-Timed Petri nets firstly introduces interrupt arcs which can easily be represented with inhibitor arcs([16])....

    [...]


Journal ArticleDOI
George Hassapis1, Dimitra Ananidou1Institutions (1)
TL;DR: In this study the template model was used to create a specific for this application HLPN model, and the response times of two automation functions were predicted and compared with those derived from the operating characteristics of the reactor.
Abstract: Homogeneous, shared memory multiprocessors that incorporate real-time operating systems constitute in many corporations the basic platforms for developing applications of plant monitoring and automation. In this work, a template model based on the High Level Petri Net (HLPN) formalism is proposed for this class of computers. Mapping functional and timing requirements of the application software to states of this model and searching for their existence in the reachability tree of the net can verify the satisfaction of these requirements. A state searching algorithm has been developed for the case of a shared memory multiprocessor in which there is a bus-based interconnection network supporting a single communication channel for the exchange of data among the CPUs, the common memory and the computer interfaces. This algorithm groups the infinite number of states of the HLPN to a number of finite regions, identifies the region, which the desired state belongs to, and checks for the existence of a path from the initial state that leads to this region. In order to demonstrate the use of the template in the modeling and verification of timing and functional specifications of a system of the considered class, the implementation of the automation functions of a chemical reactor by a VME-bus based multiprocessor with two CPUs and running under the control of the OS-9 operating system was studied. In this study the template model was used to create a specific for this application HLPN model. The response times of two automation functions were predicted by the use of this model and compared with those derived from the operating characteristics of the reactor.

2 citations


Cites background from "Modified M-timed Petri nets in mode..."

  • ...Other models (Zuberek, 1985; Zuberek, 1987; Zuberek et al., 1998; Lindemann, 1998; Koriem, 1999) have been proposed for assessing the probability of meeting task deadlines in distributed and bus-based multiprocessor architectures when various task allocation schemes to different processing elements…...

    [...]

  • ...Other models (Zuberek, 1985; Zuberek, 1987; Zuberek et al., 1998; Lindemann, 1998; Koriem, 1999) have been proposed for assessing the probability of meeting task deadlines in distributed and bus-based multiprocessor...

    [...]


Proceedings ArticleDOI
Cheng Guangming1, Liao Ming-hong1Institutions (1)
16 Dec 2005-
TL;DR: By introducing two special transition types, interrupt switch transition and resuming transition to an interruptible subnet, Petri nets has the ability to model the interrupt mechanism of embedded systems.
Abstract: Petri nets as an important tool are more and more used in the specification, formal analysis, validation, simulation and software test of embedded system. While as an important component of embedded system, Petri nets can't model the behaviors of interrupt system directly. This paper firstly extends high level Petri nets with time and actions, and then gives the definition of interruptible subnets. By introducing two special transition types, interrupt switch transition and resuming transition to an interruptible subnet, Petri nets has the ability to model the interrupt mechanism of embedded systems. This model greatly decreases the complexity of interrupt system model based on Petri nets and has highly descriptive ability.

1 citations


Cites methods from "Modified M-timed Petri nets in mode..."

  • ...HLTPN has the similar representing mechanism as time Petri net, but in HLTPN, each place can have more than one data types, and transitions are annotated with an action inscription to describe the actions of corresponding transition....

    [...]


References
More filters

Book
01 Jan 1981-

3,504 citations


Journal ArticleDOI
TL;DR: It is shown that GSPN are equivalent to continuous-time stochastic processes, and solution methods for the derivation of the steady state probability distribution are presented.
Abstract: Generalized stochastic Petri nets (GSPNs) are presented and are applied to the performance evaluation of multiprocessor systems. GSPNs are derived from standard Petri nets by partitioning the set of transitions into two subsets comprising timed and immediate transitions. An exponentially distributed random firing time is associated with each timed transition, whereas immediate transitions fire in zero time. It is shown that GSPN are equivalent to continuous-time stochastic processes, and solution methods for the derivation of the steady state probability distribution are presented. Examples of application of gspn models to the performance evaluation of multiprocessor systems show the usefulness and the effectiveness of this modeling tool. 15 references.

1,369 citations


Journal ArticleDOI

1,209 citations


"Modified M-timed Petri nets in mode..." refers background or methods in this paper

  • ...Petri nets [1,7, 11 ] have been successfully used in modelling, validation and analysis of systems of events in which it is possible for some events to occur concurrently, but there are constraints on the occurrence, precedence, or frequency of these occurrences [6,11]....

    [...]

  • ...Petri nets [1,7,11] have been successfully used in modelling, validation and analysis of systems of events in which it is possible for some events to occur concurrently, but there are constraints on the occurrence, precedence, or frequency of these occurrences [6, 11 ]....

    [...]

  • ...It should be noticed that the pair of interrupt arcs in Fig.3 is equivalent to an inhibitor arc [1, 11 ,17] from pz to t,. In fact, modified nets easily represent inhibitor arcs since interrupts occur only when all interrupting places of a transition are nonempty; an additional “empty” place (like ps in Fig.3) is sufficient to “block” the interrupts....

    [...]


Journal ArticleDOI
Molloy1Institutions (1)
TL;DR: An isomorphism between the behavior of Petri nets with exponentially distributed transition rates and Markov processes is presented and this work solves for the steady state average message delay and throughput on a communication link when the alternating bit protocol is used for error recovery.
Abstract: An isomorphism between the behavior of Petri nets with exponentially distributed transition rates and Markov processes is presented. In particular, k-bounded Petri nets are isomorphic to finite Markov processes and can be solved by standard techniques if k is not too large. As a practical example, we solve for the steady state average message delay and throughput on a communication link when the alternating bit protocol is used for error recovery.

1,059 citations


"Modified M-timed Petri nets in mode..." refers background in this paper

  • ...Another approach to Petri nets “with time” assumes that the firing times are exponentially distributed random variables, and the corresponding rates are assigned to transitions of a net; such nets are called stochastic Petri nets [2, 10 ]....

    [...]


Journal ArticleDOI
P. Merlin1, D. Farber2Institutions (2)
TL;DR: The time-Petri net (TPN) appears to be a suitable model for the study of practical recoverable processes and several practical communication protocols are formally designed and analyzed using this new model.
Abstract: A study is presented which permits the formal analysis and synthesis of recoverable computer communication protocols. This study is based on a formal representation of processes by a model of computation, the Petri nets (PN's). The PN model is generalized to include a representation of the possible failures, and then, the concept of "recoverability" is formally defined. A set of necessary and sufficient conditions which a process must satisfy in order to be recoverable is derived. In the PN model, the processes that satisfy these conditions are shown to have some practical limitations. A new model, the time-Petri net (TPN), is introduced to remove these limitations. This new model allows the introduction of constraints in the execution times of its part. As shown in this paper, the TPN appears to be a suitable model for the study of practical recoverable processes. Several practical communication protocols are formally designed and analyzed using this new model, and some interesting properties of these protocols are formally derived.

900 citations


Performance
Metrics
No. of citations received by the Paper in previous years
YearCitations
20061
20051
20031