Sophisticated signal processing techniques and simple proof-of-principle antenna arrays built from PVC pipe, aluminium foil and copper wire could revolutionise the way NASA obtains data from its Earth observing satellites.
If the adaptive array system being studied by NASA and Georgia Institute of Technology researchers ultimately proves feasible, it could dramatically decrease the cost of building and maintaining ground stations, enabling the cost-effective construction of many more ground stations. Ultimately, that could make information from the space agency’s Earth-observing satellites more widely and rapidly available.
The “off-the-shelf” technology has already demonstrated that it can successfully receive one satellite telemetry frequency.
“The dream would be to provide total global coverage with these antenna systems and to network the systems together to make these NASA information services available to anybody sitting at a computer, almost like video-on-demand,” explained Mary Ann Ingram, a professor in Georgia Tech’s School of Electrical and Computer Engineering. “Timely information from Earth-observing satellites could be useful in many ways, such as directing operations to fight a forest fire, for instance.”
Graduate Research Assistant Billy Barott adjusts an antenna array being tested on the roof of the
Information from satellites such as Earth Observing-1 (EO-1) is now downlinked to various 11-metre dishes, primarily in the Arctic Circle, where sub-zero temperatures create maintenance and reliability issues for their complex aiming mechanisms. Typically, satellites such as EO-1 are in contact with these antenna systems between five and eight times per day for ten minutes at a time. The present antenna systems cost about $4 million each, and require resident crews to operate and maintain them.
The NASA/Georgia Tech project envisions replacing these antennas with a network of inexpensive antenna arrays that would have no moving parts and use sophisticated software, instead of careful aiming, to gather data from the satellites. The network could lower operational costs while improving access to the information.
“When people use cell phones to make calls, there are no moving parts on the antennas,” noted Dan Mandl, mission director for NASA’s EO-1 program at Goddard Space Flight Center near Washington. “What I would like to do is build a continuous cell-like network around the world that would provide almost unlimited opportunities to downlink data.”
Mandl compared NASA’s existing downlink system to old-fashioned pay phones located off expressway exits. “If you witness an accident, you can open your cell phone and call for assistance,” he said. “But if you don’t have a cell phone, you have to get off the highway at the next exit and hunt for a pay phone. What we would like to do is give these satellites the equivalent of cell phones to allow anytime, anywhere contact.”
The proof-of-principle adaptive arrays being tested by Ingram and her research team are built from inexpensive components, including common PVC piping and aluminium foil. Signals from the four antennas are analysed using a processing technique that learns to improve its performance by constructively combining scattered and reflected versions of the signal, and by suppressing noise and interference. That eliminates the need for costly front-end hardware and precise aiming of the antenna arrays, and allows flexibility in the location of the ground station.
“Instead of one big aperture from an 11-metre dish, we’re going to use several smaller apertures being developed by NASA Glenn Research Center in Cleveland, in collaboration with industry and academia. We will connect them with digital signal processing,” Ingram explained. “A smaller aperture has a wider beam, so the tracking requirement won’t be as great. They may pick up interference, especially in tracking a satellite at a low elevation angle, but because we combine multiple apertures, we can null out the interference.”
The arrays individually won’t provide the same data rate as NASA’s large structures, but having more of them spread out around the world will compensate for that. Network capacity studies show that two ground stations, each with seven 0.75-metre dishes or eight electronically steered antennas, could equal the data capacity of NASA’s existing 11-metre dish at Poker Flats, Alaska – at a significantly lower cost.
Because they don’t depend on precisely aiming a dish, each array could potentially communicate with more than one satellite at a time. “What we’d really like to have is a shared antenna resource in which software is used to separate out the signals,” Mandl explained. “As we get more satellites up in space, this will become more important.”
In testing done on the Georgia Tech campus in Atlanta, the researchers were able to downlink EO-1 information in the S-band, a frequency used for transmissions at low data rates. They had to develop a special filter to eliminate interference from terrestrial repeater stations of popular satellite radio services.
“We have demonstrated the lower rates in S-band, and during the upcoming year we will work on X-band for higher rates,” Mandl said. “Ultimately, we would like to target data rates in the range of 300 megabits per second. The technology could also be used in future research to implement Ka, thus enabling even higher data rates.”
To extend satellite reception time, the researchers are also examining several technical issues such as array-based synchronisation and optimisation of the tilt angles of the planar apertures of the electronically steered antennas. For example, this optimisation could quadruple the download capacity for a ground station with eight co-located electronically steered antennas.
If successful, the adaptive array project would give NASA more flexibility in design of future high data rate satellites that may generate terabits of data on each orbit of the earth. Reliably downlinking that amount of information will require a new approach, Mandl noted.
As the project enters its third year, the following activities are planned: Testing of inflatable dish antennas; demonstration of a six megabit-per-second downlink from NASA’s SAC-C satellite, and testing of electronically steered antennas being developed at the University of Colorado under a grant from NASA Glenn Research Center.
“If you are in the Arctic and the motor moving your dish breaks down, it may take a few weeks to fix it,” Mandl said. “If this could be done with no moving parts using techniques of digital signal processing and software radio, one of the most desirable features will be a high level of reliability. That’s important for space applications and locations where you can just put equipment out there and not require an operator or maintenance crew.”
Georgia Tech researchers analyse data from an antenna array being testing on the roof of the