This paper presents an analysis of an inexact N -bit ripple carry adder architecture. Results show that a 30 percent power reduction is achieved for several approximate adders while maintaining a root-mean square error of 16 percent.

No abstract

This paper describes a three-year study to introduce nine learner-centered instructional techniques into a two-course electrical engineering graduate course sequence in integrated circuits (ICs) targeted to real-world problems in industry, defense, and security. The study measures the student learning in this two-course sequence with the use of a pre-test/post-test teaching methodology and is carried out through a collaboration of Air Force Research Laboratory Sensors Directorate and Air Force Institute of Technology (AFIT), a government institution. The participants in the study were Air Force officers at AFIT.Results presented in this paper demonstrate the effectiveness of a pre-test/post-test teaching methodology even when data is restricted to samples of small size. Results show that a statistically significant improvement was observed in the first course of the two-course sequence when the Diagnostic and Post-Diagnostic evaluation results were compared. Analysis of the final exam results for one course for Year 2 and Year 3 shows that there is statistically significant improvement in performance. This change is attributed to the improved teaching methodology presented in this paper.

A broadband, electrically controlled polarizationindependent intensity modulator (PIM) based on a Sagnac interferometer is presented. Influences from asymmetric modulation and traveling wave effects are analyzed and methods to eliminate these effects are proposed. Compared to Mach-Zehnder interferometer-based modulators, this modulator is independent of the input polarization state and less sensitive to ambient perturbations due to its reciprocity property. A prototype modulator operating at 1 and 2.488 Gb/s modulation rates is demonstrated.