Transforming Models of Computation to Automotive Component Models

Abstract: Due to the performance and capability enhancements of computer-based ser-vices in the last couple of decades, software has started to control and operate most of the functionalities in a vehicular system and offered a faster devel-opment process with the flexibility of managing more advanced computer- controlled functionality. Nevertheless, the size of the software, complexity, and safety verification issues have dramatically increased with this enormous workload as well. To overcome the excessive real-time requirements and soft- ware complexity, notions of Component-Based Software Engineering (CBSE) and Model-Driven Engineering (MDE) have been conceived and tools have been developed based on these approaches.The Rubus approach, its component model Rubus Component Model (RCM), and its toolchain Rubus Integrated Component Model Development Environ- ment (Rubus ICE) have been presented to support the development of embed- ded control systems for vehicles by aiming to cover most crucial processes in real-time development that are design, analysis, and synthesis. The tool is based on the notions of CBSE, MDE, and real-time scheduling theory. On the other hand, the Synchronous Data Flow (SDF) Model of Computation (MoC)is another component model that is widely used in the design and modeling processes of applications in telecommunications, avionics, and automotive in- dustries.Even though these two approaches have many common features and they both refer to similar industry issues, there is no transition defined between these domains so far. By constituting a relation, compatibility between two widely used domains can be procured and versatile analysis tools of Rubus ICE can also be used for Synchronous Data Flow Graph (SDFG) applications. This work proposes a transformation methodology along with its steps and rules to transform an application modeled as SDFG to its RCM equivalent with the purpose of enabling the advanced analysis and synthesis Rubus ICE implements for the SDFG designs. Additionally, two benchmark examples are tested and analyzed to support the applicability and feasibility of the proposed transformation approach.