Publications

• An Adaptive Nulling Antenna for Military Satellite Communications
• Real-Time Radar Image Understanding: A Machine-Intelligence Approach
• Adaptive Nulling in the Hyperthermia Treatment of Cancer
• A 35-GHz Beam Waveguide System for the Millimeter-Wave Radar
• Scaled Atmospheric Blooming Experiments (SABLE)

An Adaptive Nulling Antenna for Military Satellite Communications
William C. Cummings

Using adaptive nulling, an antenna system can reconfigure the receive antenna to produce nulls at the locations of any interference sources. Typically, two generic types of adaptive nulling antennas are considered: the thinned phased array and the multiple-beam antenna. The thinned phased array has excellent nulling resolution but is not a good candidate for area coverage. The multiple/beam antenna, on the other hand, can provide good area coverage but has poorer nulling resolution.

Lincoln Laboratory has constructed and demonstrated an antenna system that combines the desirable qualities of both types of antennas with the added advantages of multiple simultaneous beam service and extremely lightweight. Designed to withstand the rigors of both a launch and a space environment, the antenna produces three simultaneous electronically agile beams, any one of which can be used for area coverage. The antenna operates in the 43.5-to-45.5-GHz band and is capable of producing broadband nulls at least 30 dB deep with a resolution of about 0.1 degrees. Weighing slightly less than 56 lb, the antenna is a viable candidate for military satellite communications.

Real-Time Radar Image Understanding: A Machine-Intelligence Approach
Ann Marie Aull, Robert A. Gabel, and Thomas J. Goblick

Machine intelligence (MI) techniques have been combined with conventional signal processing and image processing techniques to build a software package that automatically recognizes reentry vehicles from a sequence of radar images. This software package, called ROME (Radar Object Modeling Environment), takes as its input a time sequence of range-Doppler images of a target produced by the Lincoln Laboratory Discrimination System from raw radar-data samples. ROME then processes this image sequence to extract radar features and track them from image to image. From these feature tracks, ROME then constructs a three-dimensional model of the target. The object model derived from the image sequence is then compared to a catalog of models of known objects to find the best match. If no sufficiently close match is found, the observed object is declared unrecognized. The object-model catalog is constructed by adding new models derived from radar data that do not match any model already in the catalog. Thus ROME is "trained" to recognize objects by using real radar data from known objects.

Because of the strong interest in real-time recognition of reentry vehicles, the entire ROME system was recoded for parallel execution on the MX-l multiprocessor, which was developed in the Machine Intelligence group at Lincoln Laboratory. This machine was designed for MI applications that involve intensive numeric as well as symbolic computation, and that have real-time processing requirements. The final version of the ROME object-recognition system, coded in a combination of parallel Common LISP and C, runs in real time on the 16-node MX-1 multiprocessor.

Adaptive Nulling in the Hyperthermia Treatment of Cancer
Alan J. Fenn and Gerald A. King

We have investigated the use of adaptive array antenna techniques to maximize the applied electric field at a tumor site in a target body while simultaneously minimizing or reducing the field at the locations of undesired high-temperature regions, or hot spots. Computer simulations have shown that adaptive nulling can prevent undesired hot spots from occurring during the heating of a deep-seated tumor. The simulation results have been supported by experimental measurements with a commercial hyperthermia phased-array antenna system that was modified to implement adaptive-nulling and adaptive-focusing algorithms. The experiments were conducted at the State University of New York (SUNY) Health Science Center in Syracuse, N.Y. Two types of phantom targets—a saline-filled polyethylene bottle and a beef sample—were used in the experiments. Results indicate that adaptive nulling can be used to reduce the electric field at one or more target positions while simultaneously maintaining a focus at a deep-seated location within the target.

A 35-GHz Beam Waveguide System for the Millimeter-Wave Radar
Willliam D. Fitzgerald

The millimeter-wave radar is a broadband, dual-polarized Cassegrain system operating in the Ka band (35 GHz) and W band (95 GHz). To upgrade system sensitivity and bandwidth, we replaced the 35-GHz microwave system with a reflecting beam waveguide (BWG), which is a quasi-optical system. This article describes the design and performance of the BWG retrofit. The goal is to increase sensitivity of the millimeter-wave radar by 10 dB. Two new high-power amplifiers, each of which produces double the power of the existing source, are paralleled with a quasi-optical combiner for a 6-dB increase in total power (to 100 kW). In addition, the BWG system reduces the microwave transmit-and-receive line losses by 4 dB.

The new BWG system is the first quasi-optical, high-power, dual-polarized, angle-tracking radar; its advantages include lower losses, broader bandwidths, and power levels well beyond the capabilities of conventional waveguide systems. In particular, the quasi-optical components are superior to their waveguide counterparts. As a rule of thumb, if the main aperture in a Cassegrain system is greater than 400 wavelengths, then a BWG system becomes a viable option.

Scaled Atmospheric Blooming Experiments (SABLE)
Daniel G. Fouche, Charles Higgs, and C. Frederick Pearson

The SABLE field experiments investigated the use of phase-conjugate adaptive optics to compensate for strong thermal blooming and turbulence. The experiments, conducted at the TRW Capistrano Test Site in southern California over horizontal paths of 100 and 400 m, were designed to explore issues related to the propagation of high-power lasers from the ground to space. The laser used in the experiments was the 10-kW hydrogen-fluoride (HF) Alpha Verification Module (AVM).

Considerable diagnostic instrumentation was incorporated into SABLE to enable detailed measurements of the characteristics of the beam at both ends of the propagation path, the atmosphere along the path, and the adaptive optics system. The measurements were used both to interpret the experimental results and to test the accuracy of MOLLY, a four-dimensional computer propagation code developed at Lincoln Laboratory.

SABLE demonstrated that (1) phase-conjugate adaptive optics can successfully compensate for severe blooming in the real atmosphere over a wide range of atmospheric conditions, (2) wind variations suppress the phase-compensation instability (PCI) and improve system performance significantly, and (3) the computer propagation code can accurately predict system performance for adaptive optics compensation of turbulence and severe thermal blooming in the real atmosphere.