Error analysis of digital filters realized with floating-point arithmetic | IEEE Journals & Magazine | IEEE Xplore