Figure 1. TRW's passive millimeter wave demonstration camera includes real-time video, filed-of-view 15 X 10 deg., 18-in. aperture, and 1040-w-band direct detection MMIC receivers.
No one likes to look up at the departures board and see that their flight has been canceled. Fog, mist, and rain cost airlines around the world billions of dollars each year in delays and rerouting -- not to mention the intangible cost of customer dissatisfaction. Instrument landing systems (ILS) help reduce weather-related flight cancellations at major airports, where direction, glide path, and other information can be transmitted from ground stations to incoming aircraft. Still, thousands of flights each year are canceled or delayed at smaller airports that do not have the benefit of extensive radar and communication systems. Even when ILS is present, landing in inclement weather is no easy task; the job would be vastly safer and easier if pilots had a system that could see through the rain to the ground below.
Several sensor technologies have been considered as potential solutions for the problems associated with landing aircraft in bad weather situations. Infrared (IR) imagers are one technique, but these systems suffer from shadows and absorption in wet environments. Active techniques that emit a signal and then capture the reflected signal also offer possibilities, but airports are already inundated with stray signals, radiation, and transmissions. But passive millimeter wave (MMW) imagers offer a potential solution that would beat most weather conditions, not add to an already crowded electromagnetic spectrum, and would display land and object features in a way that is intuitively understood by laymen as well as professionals. Furthermore, the technology is particularly effective at revealing metal objects without the use of active radiation, which means it would deliver the same benefits to security personnel at airports searching for concealed weapons as it would to soldiers searching for camouflaged tanks and other war machines. "The reason we're looking at passive MMW is it's a great metal detector," said Roger Smith of the Army Research Laboratory. "The metal pops right out at you from different backgrounds. It's really unlike any other system." Beating the expense
Unlike IR imagers, however, passive MMW imagers are extremely few in number and, therefore, expensive. Roger Appleby of the UK's Defense Evaluation and Research Agency (DERA) believes the potential benefits of passive MMW certainly justify the cost of development. "One major difference between millimeter wave imaging and infrared imaging is that the contrast in a millimeter wave scene can be as high as 200 K, as the sky is very cold," he said. "(In addition to this greater dynamic range), the sensitivities of passive millimeter wave imagers can be similar to that of thermal imagers."
Figure 2. PMMW image of industrial area.
Figure 3. Aircraft could be identified under fog, smoke, clouds, sandstorms, or snow.
Although sensors that detect radiation in the 30- to 300-GHz range have been around for some time, it's only in recent years that companies have developed imagers that use this part of the spectrum. "The two major challenges in this field are getting imagery in real time and with sufficient spatial resolution," Appleby said. "I remember one military office was quoted as saying, 'When the performance is good the equipment is too big, and when the equipment is the right size the performance is poor!'" Researchers around the world have recently developed concepts that satisfy both the space envelope and the required performance.
Two U.S. companies, ThermoTrex Corp. and TRW, are pursuing very different paths to find a balance between resolution, speed, size, and cost. TRW's work depends on a special focal plane array and generally operates like a camera model. ThermoTrex, however, has developed a pupil plane system that uses a flat panel phased array sensor with fewer individual detectors.
According to John Lovberg, manager of the Passive Remote Sensing Group at ThermoTrex, the original version of the ThermoTrex Passive Millimeter Camera captured the incoming millimeter radiation through a 32-channel, 3-ft.2 frequency-scanned sparse phased array antenna. The antenna signal was downconverted from the input RF signal (between 91 and 97 GHz) to an IF range between 6 and 12 GHz, in 1-GHz bands. This signal was used to modulate a patented acousto-optic (Bragg crystal) image converter. As the RF signal fluctuated it changed the diffractive index of the crystal, which in turn deflected an incident laser beam to a particular segment of a CCD camera, producing an optical image. ThermoTrex has upgraded the original PMC by replacing the Bragg cell with an MMW beamforming lens that converts wave phase into azimuthal position. Discrete lens outputs are then directed to other MMW lenses that separate the signal into 100-MHz-wide frequency bins along the elevation axis, resulting in a two-dimensional image.
In this method, wave frequency becomes the y axis and the phase of the incoming wave represents the x axis. "It's a very efficient way to use electronics, but a somewhat less efficient way to use the thermal (input) signal bandwidth because you're splitting one signal into a whole bunch of pixels. This allows you to generate imagery through interference of signals from a smaller number of receivers relative to a focal plane array. So you can make large pixel arrays from a small amount of electronics, which is a definite cost advantage," Lovberg said.
Initially, the ThermoTrex PMC had a sensitivity of approximately 6 K and collected one image frame every 10 seconds. Signal processing improvements have greatly improved the performance; the PMC now delivers 60- X 75-pixel images at 30-Hz update rates. The improved system has recently completed two months of flight testing from a helicopter platform.
A planned second-generation PMC will use more than 200 parallel signal channels in a linear array with ultra-low-noise indium phosphide monolithic millimeter-wave integrated circuit (MMIC) amplifiers and a new planar antenna design. The result, Lovberg said, should be an imager with better than 2 K sensitivity, capable of generating 192- X 128-pixel images in real time, scalable to very large aperture sizes. Recently, ThermoTrex completed a concept study for the Defense Advanced Research Projects Agency (DARPA) that showed how a 25-ft. antenna array could be mounted under the wings of an unmanned Predator surveillance vehicle. "With a pupil plane array you can build large antenna panels," Lovberg said. "That's something we're uniquely able to do because of the pupil plane versus the focal plane architecture."
DERA Malvern has pioneered a mechanically scanned real-time imager that has a 20- X 40-deg. field of view and uses only 32 receivers to generate a 125- X 68-pixel image. By using such a small number of receivers the design offers a large reduction in the total cost of the system. DERA achieved this low number of receivers by using a mechanical scanned concept based on folded geometry utilizing polarization techniques. This design greatly reduces system volume and is based on low-cost materials, such as polystyrene, ferrite (the material refrigerator magnets are made from), and printed grids similar to printed circuit boards. The current prototype operates at 35 GHz; however, a 94-GHz imager based on a refined concept is well advanced, Appleby said. Like the ThermoTrex PMC, DERA's imager will use indium phosphide devices.
"I think these imagers will enter the market in two areas: security scanning and poor-weather flying aids," Appleby said. "MMIC technology is vital to these applications. In the longer term, research will lead to electronic scanning." MMIC (pronounced 'mimic') is the product of 17 years of development at TRW. It combines several elements, including a high-gain/low-noise amplifier chain, integrated diode detector, and a switch for Dicke mode operation into a single 2- X 7-mm2 MMIC chip.
TRW not only beat the material and technological hurdles of creating the MMIC chip, the company also fully realized a demonstration camera model and an automated assembly line for manufacturing the MMIC receiver modules within the $5 million budget set by DARPA.
To create the array, groups of four chips are integrated into a single module and 10 modules are then strung together and attached to a single water-cooled card. These cards -- each analogous to one line of a CCD camera-are then stacked into a group of 26, creating a streaming video with 1-K sensitivity at 17 fps. And, said TRW project leader Merit Shoucri, the estimated production cost per detector is in the tens of dollars.
"Everyone is working on two elements: aperture size and real-time imaging. TRW is working on both," Shoucri said.
The system's centerline is at 89 GHz, with a 10-GHz bandwidth. At these frequencies, packaging and production systems become crucial. As a result, TRW had to develop pick-and-place capabilities that met positioning specifications to within 0.001 in.
TRW is currently working on a production prototype model. According to Shoucri, the PMMW camera system has achieved the holy grail of adverse weather imaging systems: airplane landing, guidance, and navigation. "Basically, technology is introduced into the cockpit of an airplane very slowly," Shoucri said. "HUDs (head- up displays) have been around for 30 years and it's just now that most of the major carriers are starting to install them in new aircraft. So the next step will probably be something like the millimeter camera sometime in the next five or 10 years."
TRW Space & Electronics Group
Redondo Beach CA 90278
Phone: (1) 310/812-5161
Fax: (1) 310/813-3213
Air Force Research Lab.
Phone: (1) 850/882-4631
Fax: (1) 850/882-4034
Malvern Worcsershire UK
Phone: 44 1684 89 4222
Fax: 44 1684 89 4498
San Diego, CA
Phone: (1) 858/646-5300
Fax: (1) 858/646-5301
R. Winn Hardin
R. Winn Hardin is a science and technology based in Jacksonville, FL.