Sabbatical Research Plan
I am jointly appointed in the Departments of Electrical Engineering and Earth and Ocean Sciences. As can be seen from the
publication list,
my research interests span both wave propagation and digital signal processing. My current reserach focus has evolved from the combination of techniques in Geophysics and Electrical Engineering. My work in phased arrays has been supported by Amoco Canda and my research in geophysical tomography has been supported for the last four years by the Lawrence Berkeley National Laboratory. While on sabbatical I plan to focus on the following two research projects:
 Phased Array Applications to Narrowband Cellular Telephony
In conventional cellular telephony, a major problem is allocation of sufficient bandwidth to serve as many clients as possible from the transmitting base station. One way to overcome the bandwidth limitation is to use multiple transmitting antennae in an optimal fashion. The means of achieving this goal is to create a phased array of antennae to focus multiple beams from the base station to the multiplicity of receivers. The summation of multiple noninterfering beams provides a means of overcoming the bandwidth limitations.
I have been involved with Amoco Canada in the creation of a seismic antenna to focus seismic energy in preferential directions. The principles learned in this work can be modifed to obviate the problem of adequate bandwidth in cellular telephony. In that case, it is desired to define a desired radiation pattern for each receiver allotted to a basestation. The novelty in the research is the application of techniques from Geophysical Inverse Theory in solving the least squares optimization problem by minimizing a model norm subject to the desired data as constraints. The resulting algorithm will be optimized, with the view of
making it adaptive in realtime. Using a Rayleigh or Rician fading channel, which are models of multiple interference, the effect of imperfect beam focussing on the bit error transmission rate will also be investigated.
 Application of Statistical Scattering Models to Geophysical Tomography
Seismic crosswell tomography is a method for imaging the subsurface between two boreholes. The images are obtained by exciting a piezoelectric transmitter in one borehole and receiving the signal in the receiver borehole. For each transmission, the traveltime of the earliest arriving seismic energy, determined by the subsurface seismic velociy, is picked. Redundancy is provided by repeating the experiment at multiple sourcereceiver offsets. The complete set of traveltime picks is used to obtain a map or image of the seismic velocity , which is diagnostic of the structure of the subsurface. The quality of the image (tomogram) is critically dependent on the proper traveltime picks.
For the last four years, in collaboration with the Earth Sciences Division at the Lawrence Berkeley National Laboratory, I have been involved in the application of Golaycode phase modulation as a source sequence for seismic crosswell tomography. The identification of the first energy arrivals, obtained by picking the first correlation peak, is difficult due to interfering waves (diffracted and reflected waves). The application of both the Rayleigh and Rician fading models, used in telecommunications, to this interference problem is novel enough to merit serious investigation, with the intention of improving the quality of the correlation picks and hence the resultant tomograms. Comparison between a statistical representation of the multiple scattered waves and a deterministic representation, via finite difference simulation where appropriate, will also be undertaken.
