Abstract
Whereas most of the work that analyses Synchronous Dataflow (SDF) stays in the dataflow framework, this work pushes its analysis into another framework level, thereby addressing issues that are not well addressed or are even unexplored in SDF. In this manner, the paper proposes a model-driven engineering (MDE) method, combining Synchronous Dataflow (SDF) and Petri nets, to highlight and reinforce their interoperability in digital signal processing applications, cyber-physical systems, or industrial applications. Improvements regarding the settlement and exploitation of the initial conditions associated with SDF are demonstrated; this issue is crucial for every cyber-physical system, since a system’s initial conditions are crucial to ensuring the system’s liveness. The improvements outlined in this work exploit an innovating mapping in the Place/Transition (P/T) Petri net domain that is intended to reduce and predict the total amount of initial data in SDF channels. The relevance of the firing semantics engaged with the equivalent Petri net model is discussed. This paper proposes a new approach to estimate whether an SDF has a static schedule by performing simulation and property verification of the equivalent-based P/T Petri net system achieved, framed by a Petri net invariant analysis and based on the stubborn set method of Petri nets. In this way, this new approach will allow mitigating the state explosion problem. Finally, a strategy is applied to two case studies to discover all the elementary circuits (static schedules) associated with the generated model’s state-space.
Highlights
Signal processing-based applications typically demand initial conditions to start a model’s execution, which can be associated with the system’s initialization and configuration
If a dataflow graph includes delays, an equivalent Petri net delay structure must be inserted in the final Petri net model
It should be clarified that Petri nets do not need to be cyclic, but in this case, what is required is to find static schedules; they should be cyclic
Summary
Signal processing-based applications typically demand initial conditions to start a model’s execution, which can be associated with the system’s initialization and configuration. In this paper, this period is referred to as the start-up phase. The major goal of the proposed work is to find initial markings such that system’s model enjoys liveness and consistency properties, in addition to minimizing the runtime and reducing the buffer requirement in the system start-up phase. This strategy potentially will have a strong impact on reducing a system’s energy consumption
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
