z-logo
Premium
Performance prediction of networked information systems via Petri nets and queuing nets
Author(s) -
Shin Insub,
Levis Alexander H.
Publication year - 2002
Publication title -
systems engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.474
H-Index - 50
eISSN - 1520-6858
pISSN - 1098-1241
DOI - 10.1002/sys.10031
Subject(s) - executable , correctness , petri net , computer science , queueing theory , distributed computing , programming language , computer network
Abstract An approach is presented for generating a performance prediction model so that both qualitative (logical correctness) and quantitative (timeliness) properties of a real‐time system can be evaluated. The architecture of a system is layered into a functional layer and a physical one. Both architectural layers are developed as executable models: the executable functional model is a Petri net and the executable physical model is a queuing net. The two‐layered executable models are then connected to develop a performance prediction model. A message‐passing pattern is generated from the Petri net using a state space analysis technique. Then, the queuing net model processes these messages preserving the pattern. Once the network delays are obtained from the queuing model, their values are inserted back into the Petri net model. Since the communication service demands are isolated from the executable functional model, the communications network can be specified independently at any preferred level of detail. This enables the executable functional model to be invariant with respect to the executable physical model resulting in additional flexibility in designing a large‐scale information system. This property, together with the synthesis technique, enables both formal and simulation methods to be used together, when each one utilizes a different self‐contained commercial‐off‐the‐shelf software application. © 2002 Wiley Periodicals, Inc. Syst Eng 6: 1–18, 2003

This content is not available in your region!

Continue researching here.

Having issues? You can contact us here