Research

The major area of focus for Dr. Jeffs' research includes the related disciplines of digital signal processing, sensor array processing, and digital image restoration. The following list describes research topics Dr. Jeffs has been involved with, with the most recent activities listed first.

1. Phased array feeds for radio telescopes dishes: A new trend in developing the next generation of radio telescopes is to use a compact array of antennas at the focal plane of a large dish reflector. Compared to traditional single waveguide feeds, these phased array feeds (PAFs) can increase the instrument field of view
and sky survey speed. Unique challenges associated with PAF observations,
including extremely low signal levels, long-term system gain stability requirements, spatially correlated noise due to mutual coupling, and tight
beamshape tolerances, require the development of new array signal processing
techniques for this application. Dr. Jeffs has studied calibration and beamforming strategies for PAFs including interference mitigation with power spectral density (PSD) estimation bias correction. Key efficiency metrics for single-feed instruments have been extended to the array case and are using these to verify performance of the algorithms. In collaboration with the National Radio Astronomy Observatory (NRAO) and Cornell University, these techniques have been validated in experiments on the Robert C. Byrd Green Bank Telescope, the Green Bank 20-Meter Telescope, and the Arecibo Telescope.

2. Self calibration for low frequency radio astronomical arrays: There is growing
interest in radio astronomical observations in the low frequency range where it may be possible to detect the most distant, and highly red shifted celestial objects yet observed. Instruments must detect signals at unusually low frequencies (10-350 MHz) and over large apertures (100 km). Major projects include the Dutch Low Frequency Array (LOFAR), the Giant Metre-wave Radio Telescope (GMRT), the Long Wavelength Array (LWA), and low-frequency retrofits to the Very Large Array (VLA) in New Mexico, and the Murchison Wide Field Array in Australia. Our research addresses the most significant outstanding calibration challenge for large low-frequency arrays, i.e. correction for ionospheric phase distortion. At these frequencies the Earth's ionosphere acts as a random refractive sheet which over the large aperture induces source-direction-dependent gain and phase errors that must be estimated and calibrated out. Due to direction dependence, existing self calibration techniques cannot be applied. We have studied the direction dependent calibration problem in detail for the LOFAR and VLA arrays from a parameter estimation theoretic perspective. Self calibration algorithms have been proposed, and Cramer-Rao lower bounds (CRB) have been developed to guide further algorithm development and array geometry design.

3. Adaptive interference mitigation for radio astronomy: Radio astronomical
observation is increasingly plagued by man-made interference from ground based broadcast, mobile wireless communications, and satellite downlink sources. Dr. Jeffs has studied several mitigation approaches. Real-time signal processing tools have been developed to cancel space-based interference using adaptive filtering techniques. This has been successfully demonstrated at the National Radio Astronomy Observatory (NRAO) 100m Green Bank Telescope (GBT). New algorithms have been developed and analyzed for algebraically projecting out interference seen in imaging arrays like the Very Large Array in New Mexico. Kalman tracking techniques have been adapted to improve data time blanking for removing aviation radar interference seen at the GBT. Adaptive beamforming cancellation has been employed with a phased array feed on the Green Bank 20 meter Telescope dish to remove mobile interference. Estimation bias caused by array interference cancellation has been studied and correction algorithms proposed.

4. Multiple antenna arrays for wireless communication: Recently developed
algorithms for space-time coded Multiple Input, Multiple Output (MIMO) wireless systems have been shown to be theoretically capable of increasing channel capacity by an order of magnitude or more. This gain is achieved using
multiple antennas at both the transmit and receive ends of a wireless link to exploit the presence of multipath scattering, without increasing radio frequency
bandwidth requirements. We have studied both the indoor and outdoor MIMO
channel environments with an experimental channel sounding platform. Based on these observations, statistical models to enable realistic channel capacity calculations were proposed.