Professor's antenna arrays steer clear of interference

For a long time, Jennifer Bernhard has had an idea for a new kind of antenna array that would enhance communication among devices in a wireless office network. Now the ECE assistant professor has funding to begin working in earnest on the idea, or a version of it, but her sponsor plans to deploy the arrays in outer space rather than an office space.

And that is fine with Bernhard. She'll use the three-year, $312,000 NASA contract not only to build a prototype for the agency, but also to develop what she calls a "coherent design methodology" for the arrays, which will make them easy to modify for different applications. Meanwhile, she'll keep talking to companies about commercial uses.

Bernhard's plan is to design and build arrays whose elements (individual antennas) are reconfigurable and thus can "steer" their transmitted beams with unprecedented agility. "Usually a beam is steered by phasing the elements in an array," said Bernhard, referring to a technique in which identical elements are excited at different times. "But this kind of steering is limited by the fact that all the elements are the same. We will give each individual element the ability to change its radiation pattern, so you have another degree of freedom in steering the beam."

That new degree of freedom will help a satellite communicate more effectively with other satellites, with spacecraft, and with ground stations, scoring a victory for perspicuity amid the jabber of cosmic noise and flux that too often shout down the cooler heads of space.

Of course, interference and environmental unpredictability can gum up just about any wireless system, but only those networks whose units are big enough to accommodate an array are candidates for Bernhard's solution. The compact design of a mobile phone, for example, usually allows for just one element. Bernhard's prototypes for NASA will consist of 4 X 4 arrays (16 elements), and an actual working array on, say, a future Mars mission might consist of 200 X 300 elements and occupy several square feet.

Bernhard's co-investigator on the project is ECE Associate Professor Eric Michielssen, an expert in numerical analysis of antennas. Michielssen will develop an array simulator that can account for the individual radiation patterns of all the elements-no mean computational feat.

For her part, Bernhard will oversee development and fabrication of the arrays, using Michielssen's simulations as well as measurements of actual designs. She already has some promising schemes employing off-the-shelf pin diodes and varactors as "switches" to reconfigure the planar, spiral elements best suited for high-data-rate wireless transmissions. However, the devices she expects to deliver to NASA in about two years will use microelectromechanical systems (MEMS) to carry out the switching function.

"Right now, there are no production quality, radio-frequency MEMS switches available, so we're using something that's available and that we know works," said Bernhard. "But by the third year of the grant, we should be able to get good MEMS samples." One possible source of those samples: ECE Professor Milton Feng.

Applied in an office environment, such arrays could enable wireless parallel computing and data processing, according to Bernhard. "You could have wireless, ad hoc networks that use just the units you need for a given purpose. The radiation patterns of various machines can be constrained so that they don't interfere with each other."