Performance Modeling of Embedded Systems for Senior Undergraduates
Virginia Commonwealth University, Richmond VA
Investigators
Abstract
Electrical Engineering (55) The design of embedded systems is inherently complex. A key aspect of embedded systems is that they typically require the operation of numerous individual functions to complete their overall application. Some of these functions may be performed either in software executing on a dedicated microprocessor or on dedicated hardware designed specifically to perform the needed function. During the design process, the designer must decide which functions are to be implemented in software and which are to be implemented in hardware. In addition, embedded systems also must be designed to meet constraints not only in terms of performance, but also in terms of cost, weight, power, area, and many other factors. As might be imagined, designing an embedded system incorporating all of the above aspects is quite challenging. Numerous design alternatives must be considered in the design process. A number of successful tools and techniques for constructing and analyzing performance models of hardware/software components for embedded systems have been developed. Unfortunately, the use of performance modeling to analyze design alternatives for hardware/software systems has not been part of the standard undergraduate curriculum for computer engineering, where embedded systems designers are trained. This project addresses this problem by developing the materials necessary to teach the techniques and benefits of performance modeling for the exploration of design alternatives in a senior undergraduate course in advanced digital design. The result is that undergraduate computer engineering graduates are much better prepared to take on the challenge of designing efficient, effective solutions for the embedded systems applications of the future.
View original record on NSF Award Search →