Designing for low power in complex embedded DSP systems

Abstract

This paper presents an empirical methodology for low power driven complex DSP embedded systems design. Unlike DSP design for high performance, research of low power DSP design has received little attention, yet power dissipation is an increasingly important and growing problem. Highly accurate power prediction models for DSP software are derived. Unlike previous techniques, the methodology derives software power prediction models using statistical optimization and it is verified with real power measurements. The approach is general enough to be applied to any embedded DSP processor. Results from two different DSP processors and over 180 power measurements of DSP code show that power can be predicted far embedded systems design with less than 4% error. This result is important for developing a general methodology for power characterization of embedded DSP software since low power is critical to complex DSP applications in many cost sensitive markets.