Algorithms Go from Floating- to Fixed-Point

by Marc Barberis EE Times, 4 Apr 2005 Migrating signal-processing algorithms from floating- to fixed-point is often necessary to meet various design constraints, including real-time performance, cost and power dissipation. The migration process can be challenging for developers creating fixed-point implementations based on a floating-point reference model. Read the article