Issue No. 03 - July-September (1997 vol. 14)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/54.605989
Front-end system engineering (SE) typically involves less than 4% of the total prototyping time and cost compared to electrical engineering (EE) related design aspects, but accounts for over 70% of an electronic product's life cycle cost (including manufacturing). In this article, we propose a novel industry-driven front-end automated design approach for embedded electronics systems. Our approach employs parametric cost models to generate reliable estimates for development and maintenance costs as well as time-to-market costs. More specifically, this method involves the creation of a mathematical programming model which produces minimum life cycle cost product designs that satisfy stringent system-level performance and form factor constraints. This research focuses on the design of embedded microsystems which are composed primarily of commercial-off-the-shelf (COTS) components, thereby allowing for increased performance, lower costs, while maximizing opportunities for reuse. Furthermore, our approach, part of DARPA's RASSP program, is seamlessly integrated within a hardware-less VHDL cosimulation and coverification environment for rapid prototyping. We demonstrate this method by applying it to the front-end design of a Synthetic Aperture Radar (SAR) system. Case study results show how greater than 6X cost savings could be accrued over traditional design processes. The results can also be translated to the commercial marketplace, through the effective interface between system engineering and electrical engineering within an automated environment.
embedded systems, parametric cost modeling, cosimulation, mathematical programming
A. J. Gadient, V. K. Madisetti and J. A. Debardelaben, "Incorporating Cost Modeling in Embedded-System Design," in IEEE Design & Test of Computers, vol. 14, no. , pp. 24-35, 1997.