Communicating With Farthest Space
courtesy nasa/jpl-caltech2006_4_6-7“It’s like watching grass grow,” says Douglas Mudgway, a longtime engineer and now the historian of NASA’s Deep Space Network. “There’s not very much drama attached to a big antenna tracking a spacecraft millions of miles away. You can’t even see it move. There’s absolutely nothing to see except blinking red lights and guys sitting around drinking coffee and wishing they were somewhere else.”
However, millions of dollars, countless hours of human effort, and irreplaceable scientific knowledge all depend on those people sitting in that control room, monitoring the great antenna dishes outside that loom over the desert like ancient monuments. They’re at one end of an invisible line of communication stretching to the edge of infinity. All the magnificent visions of the planets that our unmanned explorers have brought us from the depths of the solar system, every bit of data collected by the Pioneers and Mariners and Voyagers and Cassinis, all of it first reached Earth as a whisper of photons caressing the face of an antenna dish. For more than 40 years now, maintaining that communication has been the mission of the Deep Space Network, a worldwide system that routinely traffics in the weakest, faintest, and longest-distance calls ever.
Before the space age dawned, in the late 1950s, the idea of sending messages between Earth and deep space was the stuff of science fiction and fringe obsessions. But it became obvious that some means of communicating with spacecraft would be needed. The first steps toward it were made by the military, which was working on intercontinental missiles and needed facilities for tracking and guidance.
In the early 1950s the California Institute of Technology’s Jet Propulsion Laboratory (JPL) was developing such facilities at the White Sands Missile Range for the Army’s Corporal and Sergeant missiles. A team of intensely dedicated engineers, led by Eberhardt Rechtin, created a network called Microlock, a series of ground stations featuring ingenious receivers that could compensate for the shifting frequencies transmitted by moving rockets with a technique known as the phase-locked loop. As perfected by Rechtin and his colleagues Richard Jaffe and Walter Victor, the phase-locked loop system would be the key for communicating to the edge of the solar system.
When America began sending satellites into orbit after the shock of Sputnik , in 1957, Microlock was hastily expanded to include tracking stations in San Diego, Cape Canaveral, Singapore, and Nigeria, from which it followed the Explorer satellites. Some of its equipment was far from sophisticated. “We used little hand-held machines to listen to the satellites,” Rechtin recalls.
courtesy nasa/jpl-caltech2006_4_8But the equipment scaled up rapidly. By mid-1958, as the Pioneer space probes were being readied for America’s first attempts to reach the moon, Rechtin and his team were constructing an 85-foot aluminum parabolic dish on government property in the Mojave Desert in California. Named the Goldstone Pioneer station and built by the Blaw-Knox Company, of Pittsburgh, Pennsylvania, it became the first station of Microlock’s successor, which would be called the Deep Space Instrumentation Facility (DSIF).
The Pioneer station was essentially a radio astronomy antenna based on the Cassegrain system, a configuration already used in many optical astronomy telescopes. Incoming radio waves were reflected by a parabolic main dish to a secondary reflector above the center of the antenna, and the subreflector then bounced the signal down to the focal point of the big dish (reversing the process when the antenna was transmitting). Such a design allowed the antenna’s heavy, bulky electronic transmitting and receiving equipment to be conveniently mounted at the structure’s center of gravity. Modified by two JPL engineers, William Merrick and Robertson Stevens, the antenna used a precision servo-controlled drive mechanism to keep itself pointed at a moving spacecraft, while Rechtin’s supersensitive receivers sifted the spacecraft’s faint voice out of the background radio noise of the cosmos.
On December 6, 1958, only three days after the Mojave antenna, the Goldstone Pioneer tracking station, was declared operational, Pioneer 3 left for the moon—and fell back to Earth about a day later, after getting just over 63,000 miles away. But the Pioneer station and a smaller antenna in Puerto Rico stayed in contact most of the way, bringing back valuable scientific data despite the failure of the main mission. As other missions followed, Rechtin proposed a worldwide network with antennas deployed around the globe at 120-degree intervals of longitude, so that as the Earth rotated, at least one station would always be within sight of a distant spacecraft.
His plan was quickly approved by the newly formed NASA, and two more 85-foot antennas, identical to the Mojave one, were shipped overseas, one to a missile test range in Woomera, a town in the Australian outback, and the other to a site about 40 miles north of Johannesburg, South Africa. Spaced almost evenly around Earth and connected by trunk lines, undersea cables, micro-wave relays, and radio links, the three stations were providing 24-hour coverage of the entire sky by mid-1961.
courtesy nasa/jpl-caltech2006_4_11NASA and the State Department negotiated the appropriate treaties. “It was very easy sailing at the beginning because we were using the British Commonwealth, in South Africa and Australia,” Rechtin recalls. “That was pretty straightforward.” Political considerations aside, the antennas had to be built in relatively isolated sites free of extraneous radio noise. Remote locales in the deserts of California and Australia, with mountains that blocked stray radio signals, were perfect. South Africa didn’t offer such seclusion, so the Johannesburg station was placed in a shallow valley, reasonably removed from nearby towns. Other antennas were soon added at all three stations to enhance the network’s capacity.
This expansion of the DSIF came just in time to support NASA’s ambitious new unmanned voyages. First came the Ranger series of lunar spacecraft, which were initially plagued by failure but finally succeeded with Ranger 7 , thanks partly to the lessons learned from the engineering data collected by the DSIF throughout the previous six missions. Meanwhile, Mariner 2 completed the first successful mission to another planet when it flew by Venus in December 1962, sending 11 million bits of data back to Earth before falling silent in early January 1963. By the time its sister craft Mariner 4 set out for Mars in November 1964, the DSIF was leaving its adolescence behind, undergoing some profound changes.
The first change was simply in name. In December 1963 the Deep Space Instrumentation Facility became the Deep Space Network (DSN). The second change was an ambitious new antenna project to expand the network’s reach as far as Saturn (and years later, to the entire solar system).
Like all electromagnetic waves, radio signals are bound by the inverse square law, which states that the strength of a signal is inversely proportional to the square of the distance from its source. Double the distance and the signal intensity falls four times, triple the distance and it falls nine times, and so on. When radio waves cross millions of miles of space, the signal becomes extremely feeble by the time it reaches Earth, so the sensitivity of the receiving station is paramount. Another problem is beam divergence, the spreading out of the signal. By the time a signal arrives from as far away as Jupiter, for example, it’s spread so thin that only a minute fraction falls on Earth. Obviously, the bigger your antenna, the more of that signal you can capture—and the more signal, the more data.
Rechtin knew that the 85-foot, or 26-meter (NASA eventually adopted the metric system), dishes just wouldn’t be enough to handle the missions being planned to the outer solar system, so he and his engineers began to consider ways of enhancing the network’s hearing. Inspired by Australia’s magnificent 64-meter (210-foot) radio astronomy dish at Parkes, New South Wales, Rechtin, Stevens, and Victor proposed adding three similarly huge radio ears to the DSN. Rechtin calls the site at Parkes a “brilliant design, very well worked out. I said, ‘We should build one of those.’” It didn’t take him long to find support for the idea at NASA.
At a November 1961 meeting of the American Society of Mechanical Engineers, Rechtin and Stevens spelled out some of what their plan entailed. Each big antenna would require a pointing and tracking capability precise to within a few hundredths of a degree; a parabolic surface with no more than a quarter-inch of deviation; and the ability to withstand extreme wind and weather conditions. Considering the weight of such a huge dish, the requirements were enough to give engineers nightmares. How do you control within a fraction of an inch the movement of something that weighs thousands of tons? How do you maintain its parabolic shape, so crucial to the proper focusing of radio waves, when gravity alone deforms that shape because of the dish’s weight?
More feasibility studies and design proposals followed until January 1963, when NASA awarded a $12 million contract to the Rohr Corporation of California to build the network’s first “big” dish. It would be 210 feet (64 meters) across, just the same as the Parkes dish, and not because of technical considerations. “We built it at exactly 210 feet so we didn’t upstage the Australians,” Rechtin explains. Not only had the Australians provided invaluable assistance and advice, but NASA wanted to preserve goodwill so that each country could make use of the other’s facilities on occasion. It was a courtesy that would pay big dividends to the DSN over the years.
The construction of the big antenna began in December 1963 under the direction of William Merrick, who had supervised the 26-meter projects. Dubbed the Mars Site because its first official task would be handling missions to that planet, it was essentially complete by January 1966. Two months later it received its first messages from space, locking onto the signal of Mariner 4 , then in solar orbit after making the first successful Mars flyby in July 1965. At its official dedication ceremony, in April 1966, the new Goldstone antenna acquired the distant whisper of Pioneer 6 , launched the previous December and by then in solar orbit 28 million miles away. The Deep Space Network had its big dish working at last.
While Rechtin continued to push NASA to follow up with two more big antennas to complete the network, other changes were overtaking the DSN. So far the network had handled only JPL missions because JPL was the only NASA center operating unmanned planetary spacecraft. That arrangement ended with the new Pioneer project, run by NASA’s Ames Research Center in Northern California. Douglas Mudgway says: “That was a real shakeup because up to that point everything had been managed in-house. Almost the same group of people that were building the Deep Space Network were also building the spacecraft.” Accommodating outside users and projects would be a sometimes contentious matter. “Pretty soon there was great competition among the flight projects for priority,” Mudgway says. “We had some terrible yelling matches where we would get the flight project representatives in one room and have them try to agree as to who would get time and how it would be divided.” The problem was made even worse when NASA began to sell or trade time on the network to foreign space projects.
Meanwhile, the remote site at Woomera had outlived its usefulness. “It got to be too expensive to run the station in this really remote location,” Mudgway says. “It cost too much to pay the staff, to ship equipment there. We looked for another site in Australia and found one near the capital of Canberra.” Another change in the geography of the DSN was necessitated by uglier considerations. By 1965 unrest was growing in South Africa because of that government’s apartheid policies. “The State Department told us we were going to have a bloodbath down there so we should find another place at the same longitude,” Rechtin recalls. “So we went to Spain instead.” The new Deep Space Network station was set up at Robledo, near Madrid, a locale where technical help would be readily available and political stability would be assured. NASA quietly continued to run the South African station until 1974, but Madrid took over most major network operations at that longitude.
With its new 64-meter big dish in California and two more under construction at the new Australian and Spanish stations, the Deep Space Network had become an indispensable resource by the end of the 1960s, supporting Mariner flybys of Venus and Mars, Pioneer missions to solar orbit, even the unmanned moon landings of the Surveyor spacecraft. It even did a little moonlighting (as it were) for the Apollo missions. While the Parkes antenna handled the live television from the moon, Mudgway explains, “we put the big DSN stations, which were of utmost reliability, on the downlink telemetry carrying all the medical and physical data on the astronauts.” But however impressive the Deep Space Network’s achievements had been, they were merely prologue for its biggest challenge yet. Flush with confidence and curiosity, NASA was about to push farther into space than humanity had ever before attempted.
courtesy nasa/jpl-caltech2006_4_13“I was told by Nobel Prize winners that it would not be possible really to communicate to the edge of the solar system,” Rechtin recalled in a 1995 oral-history interview with the Institute of Electrical and Electronics Engineers. “If you could, you couldn’t send back enough interesting information. You would have to have bandwidths of your receiving system wide enough to account for the Doppler shifts. If you did that you had to send megawatts of power back from enormous antennas at the edge of the solar system. Nobody knew how to do that… . Well, I sat down and figured out, ‘Wait a minute. We could track the Doppler.’”
He was referring to the Doppler effect, the change in frequency and wavelength of a radio signal or other waves caused by the relative motion between source and receiver. Because a spacecraft and Earth are always in relative motion, the frequency at which the craft transmits its data won’t be the same one detected back on Earth. It’s a useful phenomenon because in principle, you can calculate the spacecraft’s velocity and trajectory by computing the difference between the transmitted and received frequencies. The catch is that you must know the original frequency with extreme precision, and most radio transmitters, particularly the small and relatively weak ones that spacecraft must carry, aren’t completely stable; their signals can vary slightly. You have to listen simultaneously to a narrow range of frequencies to pick out the right one, and receiving several frequencies at once increases random noise from both outer space and the receiver’s own circuits, making extracting the desired voice even harder.
Rechtin and Victor solved this problem by developing the phase-locked loop receiver for the Deep Space Network. The basic idea had been around for a while, but they perfected it, first during their Army missile-tracking days and later on for the network. To compensate for the frequency changes in the spacecraft signal, the phase-locked loop system sends an extremely stable carrier signal from Earth to the spacecraft; it is then modulated with data and retransmitted back at a predetermined multiple of the original frequency. Whatever Doppler shifts the downlink from the spacecraft may undergo, the receiver remains locked onto it. This basic principle enables the network’s huge radio dishes to maintain contact with and extract data from a signal that may be no stronger than several billionths of a watt. The phase-locked loop concept is so useful that it has become part of many commercial high-frequency radio receivers.
After the incredibly weak signal is detected, it must be amplified, and the Deep Space Network employs the most sensitive radio receivers ever designed. Conventional radio amplifiers would lose the meager signal in the circuit noise generated within the amplifier itself. When something else was required, Rechtin and his deputy Nicholas Renzetti found it in the maser.
In a maser (which is an acronym for microwave amplification by stimulated emission of radiation ), electromagnetic energy (such as a faint radio signal) is applied to atoms in an excited energy state, causing them to fall to a lower energy level while in the process emitting more waves at precisely the same frequency as the applied signal, thus amplifying it. (Lasers operate exactly the same way, except at light rather than microwave frequencies.) This makes the maser an almost zero-noise, extremely sensitive amplifier. Solid-state ruby-crystal masers were adopted by the Deep Space Network in 1962. Because they used large permanent magnets and needed to be cryogenically cooled by liquid helium at a temperature close to absolute zero, they were expensive, difficult to maintain, and temperamental to operate, but they attained a sensitivity and noiseless amplifying power that no alternative could match. “The maser turned out to be the most sensitive receiving system, even today,” says Charles Stelzried, a former network engineer and 52-year veteran of the space program. “But they’re working to get away from that because of the special cooling and other requirements.”
If the great antennas can’t be pointed in the proper direction to begin with and remain locked onto the constantly moving spacecraft, nothing else matters. The way that’s done is simple in concept, if complicated in execution. In much the way an optical telescope focuses on a single star by using a low-power “spotter” telescope, a smaller radio dish with a wider angle of view finds the general area of sky from which the spacecraft signal is emanating. With this approximation, the big dish is then directed by the small dish to home in on the signal precisely, through a complex system of servomechanisms and optical equipment, and kept locked on as long as the spacecraft is in the antenna’s sky.
Before Pioneer 10 was launched, in March 1972, no spacecraft had attempted to venture beyond Mars. Some doubted that, even if contact could be maintained, a craft could survive passage through the asteroid belt between Mars and Jupiter. Pioneer 10 proved the naysayers wrong, flying by Jupiter in December 1973 and sending back the first close-up pictures and data from that giant world. Pioneer 11 followed a year later and then made the first visit to Saturn in 1979.
The Pioneer missions showed how, despite the efficiency and sophistication of the Deep Space Network, the most mundane events on Earth could affect its operations. At one point while Pioneer 10 was cruising toward Jupiter, a fishing trawler snagged and broke a transatlantic cable connecting the Pioneer control center with the South African station, disrupting communication with the craft until rerouting could be accomplished. At another point workers at Canberra threatened to strike, closing down the station there just when Pioneer 11 was due to fly by Jupiter. “It was a running joke that [at] every single launch or encounter, the Australian tracking station would go on strike or threaten to strike because they knew they had us by the balls,” recalls the Pioneer operations manager Fred Wirth. Fortunately the Canberra labor problems were resolved in time, and Pioneer 11 fulfilled its Jupiter mission.
Pioneer 10 and 11 proved to be warm-ups for the crown jewel of the unmanned exploration of the solar system, the Voyager project. Following the trails blazed by the Pioneers to Jupiter and Saturn, Voyager 1 and 2 not only gave us encyclopedias of new scientific data on those worlds but brought us sights of unimagined, breathtaking beauty: the rings of Saturn, the intricate impressionism of Jupiter’s cloud layers, the mind-boggling variety of the outer moons. Closer to Earth, Mariner 10 explored Venus and Mercury, and the Viking spacecraft landed on Mars, revealing the rusty surface of that planet for the first time. It all was made possible by the Deep Space Network, which by 1973 had completed its trio of 64-meter dishes around the world.
Eberhardt Rechtin, recognized by all as the true father of the network, had left JPL in 1967, but his legacy remained in the drive to continually improve and expand the network’s capabilities. Because Voyager 2 ’s upcoming Neptune encounter would require higher data rates and greater radio sensitivity, Robertson Stevens, the chief engineer of the network’s Telecommunications Division, came up with a plan to enlarge all the 64-meter dishes to 70 meters. But modifying the antennas meant taking them out of operation temporarily. Aside from Voyager at Neptune, the next big planetary mission was the Galileo probe to Jupiter, due to be launched in 1986. How could the work be done without crippling the network for the upcoming missions?
As it happened, the DSN got some unexpected breathing space, although not in a way that anyone would have ever chosen. In January 1986 the space shuttle Challenger was lost, grounding Galileo indefinitely until the shuttle could fly again. Without any new major missions or encounters scheduled until 1989, the Deep Space Network could weather the necessary interruptions.
The Goldstone 64-meter was first in line, followed by the Madrid and Canberra antennas, and despite some problems and mishaps, the upgrades were completed successfully. By 1988 the entire 70-meter network was online.
The Voyager project went out in a blaze of glory in August 1989 with the first flyby of Neptune, supported by the 70-meter network in conjunction with the Very Large Array at Socorro, New Mexico, and the Australian radio astronomy dishes. The simultaneous connection of different antennas, called arraying, increased the network’s overall signal receiving capacity. In effect, many small antennas make up an extremely large one. It wasn’t a new idea—radio astronomers had been doing it for years—but for the Deep Space Network, it hadn’t been considered very practical or cost-effective. As Charles Stelzried explains, “When you start arraying all those little antennas, you have to get them all lined up, all working together. Then you have to put receivers on each one, and if you have a hundred of these things, that’s going to be expensive. Then there’s the problem of pointing them and combining the signals. It takes big computers to do all that, and in those days we didn’t have big computers.” But now arraying made possible the data rates from Neptune that gave us the spectacular pictures of that planet.
After numerous delays and several near-cancellations, Galileo finally left Earth in October 1989. En route to Jupiter, however, it hit a major snag that threatened to end its mission. Its high-gain antenna failed to deploy, and nothing ground controllers tried could get it unstuck. Without the antenna, there would be no pictures and hardly any data. JPL and Deep Space Network engineers would have to find a way to make do with Galileo ’s small low-gain antenna. Leslie J. Deutsch, then a young mathematician working at JPL, was one of the team that helped find a way, devising “clever uses of new error-correcting codes, different modulation techniques, antenna arraying again, and a number of other techniques. It was fun because we got to apply all these new technology advances to a mission that really needed them in order to survive.”
Missions came and went, including the Mars Pathfinder, other rover missions, and then, beginning in 2004, the Cassini exploration of Saturn and its moons. At this point the Deep Space Network was becoming a victim of its own success. The problem: There are only 24 hours in a day, and some spacecraft were built almost too well.
Douglas Mudgway explains: “It turned out that the quality of construction that went into these spacecraft to make them reliable made them reliable for many times [their design lives]. These spacecraft never seemed to die unless they ran out of attitude-control gas or their batteries became depleted or there was a failure. The scientists working on them went back to NASA and said, ‘We can’t possibly neglect this space-craft, sending back great data and pushing further and further out into the solar system where we’ve never been before. How can we turn it off?’ There was a great outcry amongst the scientific community to force NASA to keep funding them.”
At the same time, all the enhancements of the Deep Space Network made it better able to stay in contact with such distant craft. Pioneer 10 remained in touch with Earth for more than 30 years, before its voice grew too faint. And late in 2004 Voyager 1 was reaching what may be the boundary between the solar system and true interstellar space when NASA had to debate pulling the funding plug.
Even when all the differing projects could be accommodated, they were at the mercy of difficulties beyond anyone’s control. “Particularly in the early days, there were always problems with analog recorders, and the helium supplies running out for the masers, and the masers warming up, and the hydrostatic bearings on the 64-meter at Goldstone,” Mudgway recalls. “Then there was always the weather at critical times. The S band [one of the frequency bands used by the network, at about 2.3 GHz] was susceptible to degradation by rain and snow. Rain at Canberra, snow at Madrid, wind at Goldstone. All this manifested itself in gaps in the data record.”
Leslie Deutsch, who now heads a strategic planning office for the Deep Space Network, says: “When you’re conceiving a new mission that’s going to fly in 15 or 20 years, chances are you’re going to dream up much more capability than what the DSN can provide today. We have to let future mission designers know what’s reasonable. The DSN is a multimission capability. It’s there to serve the entire community of missions that go out beyond geo orbit. You don’t want to be in the position where one of those missions has to be responsible for funding an entire DSN upgrade.”
As Mudgway observes, such concerns are scarcely evident in the tranquil efficiency of network operations during a mission. The most dramatic events manifest themselves with the utmost calm, as nothing more than slight changes in the data stream. The operators at the far-flung stations around the world, concentrating on maintaining the link, keeping the antennas tracking the spacecraft, and making sure that the maser amplifiers are humming along in their cryogenic compartments, see none of the pretty pictures or graphics. “The place you see it is back in the control rooms at JPL, where it’s turned into readable data,” Mudgway says. “Even then, to the uninitiated, it’s a little blip on a line graph on a computer. A good example is initial acquisition [the first time the spacecraft’s signal is picked up shortly after launch]. All you see is a green light come on at the tracking station. That’s the first indication anybody has that the spacecraft is working. All eyes are on the DSN to hear those magic words, ‘Receiver in lock.’ When the verbal report comes to the control room, there are hoots and cheers, but you wouldn’t know what it’s all about.”
But other satisfactions more than compensate for the placid nature of the work. “The whole experience was absolutely fascinating to me,” Mudgway says of his career. “Everything we did was new and had never been done before. We felt very strongly that there was nowhere else in the world where people were doing what we were doing. That was a huge incentive. Every new flight project that came into my hands was new: new people, a new technical challenge. To see it all finally working was a great experience.”
Now approaching its fiftieth anniversary, the Deep Space Network, more than ever the essential infrastructure of deep-space exploration, continues to improve and expand. But the future will probably not see any new big antennas rising above the desert. Instead, improved electronics and computer technology have opened up alternatives to the engineering hurdles of constructing huge, gravity-defying dishes. “The trend has been toward arraying since the early eighties,” Deutsch says. “You’re going to see fewer and fewer big dishes built for any application.” They are just too expensive to build and maintain.
Nonetheless, whatever form the future Deep Space Network takes—arrays of smaller dishes, relay stations in Earth orbit or interplanetary space, or even optical-laser facilities—the big dishes will probably remain standing. Just as ancient civilizations built huge monuments in the desert to communicate with their gods, the Deep Space Network’s massive antennas serve as our technological equivalent, through which we converse with our own ambassadors to the heavens.
