An architecture exploration framework for DSP applications

Advances in chip technology have enabled integrating many functional units on a single chip. This led to the emergence of the concept of system-on-chip (SoC) which is used extensively in the development of advanced embedded systems. Embedded systems are widely used today in different digital signal processing (DSP) applications that usually require high computation power and tight constraints. Using SoC technology increases the challenges facing the designer to choose the optimal design. A tool that helps explore different architectures is required to design an efficient system. In this paper we propose an implementation of an architecture exploration frame-work based on a multi-objective evolutionary algorithm (MOEA) for DSP applications. The design space is based on an experimental core library, and an analytical evaluation approach is used to speedup fitness calculation. Results obtained indicate that the proposed approach is valid and efficient for solving the architecture exploration problem.