Realization of digital filters using input-scaled floating-point arithmetic | IEEE Journals & Magazine | IEEE Xplore