SOFTWARE DESIGN METHODS AND REPRESENTATIONS - Software Design for Six-Sigma: A Roadmap for Excellence - page 93

system. A variety of MOC have been developed to represent heterogeneous systems.
Researchers have classified MOCs according to different criteria.
Gajski et al. (1997) classifies MOCs according to their orientation into five
classes:
State-oriented models use states to describe systems and events trigger transition
between states.
Activity-oriented models do not use states for describing systems, but instead
they use data or control activities.
Structural-oriented models are used to describe the physical aspects of systems.
Examples are as follows: block diagrams and RT netlists.
Data-oriented models describe the relations between data that are used by the
systems. The entity relationship diagram (ERD) is an example of data-oriented
models.
Heterogeneous models merge features of different models into a heterogeneous
model. Examples of heterogeneous models are program state machine (PSM)
and control/data flow graphs (CDFG).
In addition to the classes described above, Bosman et al. (2003) propose a time-
oriented class to capture the timing aspect of MOCs. Jantsch and Sander et al. (2005)
group MOCs based on their timing abstractions. They define the following groups
of MOCs: c ontinuous time models, discrete time models, synchronous models , and
untimed models. Continuous and discrete time models use events with a time stamp. In
the case of continuous time models, time stamps correspond to a set of real numbers,
whereas the time stamps correspond to a set of integer numbers in the case of discrete
time models. Synchronous models are based on the synchrony hypothesis . 27
Cortes et al. (2002) group MOCs based on common characteristics and the original
model they are based on. The following is an overview of common MOCs based on
the work by Cortes et al. (2002), and Bosman et al. (2003).
4.5.6.1 Finite State Machines (FSM). The FSM model consists of a set of
states, a set of inputs, a set of outputs, an output function, and a next-state function
(Gajski et al., 2000). A system is described as a set of states, and input values can
trigger a transition from one state to another. FSMs commonly are used for modeling
control-flow dominated systems. The main disadvantage of FSMs is the exponential
growth of the number of the states as the system complexity rises because of the lack of
hierarchy and concurrency. To address the limitations of the classic FSM, researcher's
have proposed several derivates of the classic FSM. Some of these extensions are
described as follows:
SOLAR (Jerraya & O'Brien, 1995) is based on the Extended FSM Model
(EFSM), which can support hierarchy and concurrency. In addtion, SOLAR
supports high-level communication concepts, including channels and global
27 Outputs are produced instantly in reaction to inputs, and no observable delay occurs in the outputs.