Direct RF-to-Digital Data Converters Using Time-Varying Circuits

University Of California-Los Angeles, Los Angeles CA

Investigators

Abstract

Applications demand our portable devices to provide an ever-increasing number of diverse wireless communications such as cellular, WiFi, GPS, Bluetooth, and proprietary communications that allow wearable and implantable devices to link efficiently with a smartphone. As such, it is desirable to have a compact single universal radio, similar to a Dick Tracy wristwatch radio, to seamlessly handle the multitude of such communications. Unfortunately, such a universal radio has proven notoriously hard to build, especially for use in a portable communications device. This project aims to develop the technology for the central piece of this universal radio: an analog-to-digital converter that precisely digitizes a large swath of the electromagnetic signal spectrum incident on an antenna while consuming low battery power. The technology will enable radio hardware that can be reconfigured under convenient software control. In addition, it will enable the so-called cognitive radios that increase the utilization of scarce electromagnetic spectrum by dynamically identifying and opportunistically using vacant frequency bands. Both universal and cognitive radios are critically important to the further advancement of wireless connectivity in the country. The developed technology can find use in defense electronics and medical equipment where electromagnetic spectrum sensing and awareness is important for imaging and security applications. In addition to the technical impact, the proposed research activities will provide the diverse undergraduate and graduate students at the University of California, Los Angeles, with invaluable training in the strong interplay between signal processing and electronic circuit design techniques. Such cross-disciplinary training will help engineers better tackle many of the evolving challenges in electrical engineering. The objective of this project is to develop circuit and signal processing techniques to enable wide bandwidth, high resolution, analog-to-digital converters (ADCs) that can be placed at, or very close to, the antenna in a receiver. The proposed approach employs multi-rate filter-bank concepts and time-varying circuits. In a nutshell, a bank of analog filters, made of specially chosen, periodically time-varying circuit components, will be employed to channelize the RF signal into individual, but overlapping, frequency bands that are then digitized using a bank of slower ADCs. A bank of digital filters will intelligently combine the ADC outputs to faithfully reconstruct the original signal. Prototype integrated circuits will be designed, fabricated, and tested to verify the feasibility and utility of the proposed direct RF-to-digital converters and demonstrate record-setting ADC bandwidth and resolution combinations. The intellectual merits of this project are three-fold. Foremost, it will develop a new frequency-channelization ADC approach that is less sensitive to circuit mismatch and has relaxed dynamic range requirements as compared to the conventional time-interleaved ADC approach. Secondly, the time-varying circuit technique, which is a radical and beneficial departure from the conventional principle of making circuits linear and time-invariant, will lead to significant performance and power consumption benefits. The time-varying approach enables very high linearity and precision control of the frequency channelization filter to solve the central problem of all prior ADC approaches that use a bank of analog pre-filters. Finally, the proposed research represents a very tight coupling between integrated circuit design and signal processing theory that is becoming increasingly essential to solve the outstanding problems in integrated circuit design. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.