Research Areas 

Analog Circuit Design

Our current research in this area is on circuit design techniques for short-channel CMOS technologies. Low voltage and low power-dissipation are often key considerations. Area-efficiency is also a very significant issue in these processes, since it relates to cost of implementation. We are exploring the design of front-end circuits such as low-noise amplifiers and oscillators that seek to minimize die-area while meeting the required performance metrics. The design of high-speed and high-frequency circuits, where classical analog feedback cannot be relied upon for minimizing process sensitivity is also being studied.

RFICs and Wireless Systems

This work explores efficient transceivers, frequency synthesis techniques and low-power radio front-ends for ISM-band and other narrow-band wireless applications, and for emerging systems such as Ultra Wideband, Cognitive Radio and wireless sensor networks, where power minimization is critical.  We are also investigating techniques for efficient signaling and interference cancellation in these applications.

Device technology has a significant influence on the definition of wireless systems. This can be clearly observed in the case of emerging broadband wireless systems such as Ultra Wideband. As part of this research, we are studying the interplay between devices, circuits and systems and the problem of optimizing wireless systems for specific device technologies.

Parasitic Noise and Coupling Mechanisms in ICs

Silicon substrates present a finite distributed impedance to the devices that are fabricated on the surface. This allows devices to communicate through a parasitic coupling path. IC package and board impedance can provide other parasitic coupling paths. Noise coupling and substrate losses can be deleterious and in fact can set the bounds on achievable performance in high-dynamic range systems such as transceivers for cellular telephones. We are studying by means of modeling and measurement the degradation in circuit performance due to these parasitic noise mechanisms and coupling paths. This research also includes techniques for mitigation of these effects. As a part of this effort we are investigating functional macromodeling of mixed-signal cores for rapid and accurate estimation of coupled noise.

Ranjit Gharpurey
Department of Electrical and Computer Engineering
University of Texas at Austin