Deep Space Network

The Origin and Evolution of the Deep Space Network 

NA SA’s system for communication with solar-system exploration 

spacecraft began as a Cold War crash program, but iis evolution was 

carefully planned from the starl 

Thirty-four years ago, a single principal antenna, installed the 

previous year (1958) on a crash basis in an isolated location of the 

Mojave Desert of California, supported Pioneer 4, the first United 

States spacecraft to escape the Earth’s gravitational pull and travel 

toward another solar-system body, namely the Moon, the nearest 

such body to Earth. 1 hat lone antenna, situated near the Goldstone 

Dry Lake Bed within the Department of the Army’s Fort Irwin, would 

become the cornerstone clf NASA’s Deep Space Network, a system 

currently composed of 13 antennas of various designs and sizes that 

collectively have the capability of continuously communicating with 

spacecraft at distances ranging from high altitudes above the Earth 

to the outer edge of the solar system. 

When the Goldstone antenna was procured, however, NASA had 

not yet come into being. It was instead the Department of Defense 

that provided the funding for the procurement, fabrication, erection, 

and testing of this antenna during a relatively short eight-month 

period in 1958.  The antenna, as well as the series of early lunar- 

probe attempts of which Pioneer 4 was a part, was, as we shall  

show, approved on a crash-program basis as one aspect of the Cold 

War then raging between the United States and the Soviet Union. It 

would not have been surprising if an antenna so hurriedly

manufactured and installed for a short-term goal would 

subsequently abandoned when NASA began setting up a 

have been 

permanent 

system for later lunar and planetary probes. The fact that it was 

not was a reflection of careful planning by the procurer of the 

Goldstone antenna, a group of engineers at the Jet Propulsion 

laboratory (J Pi.), an Army facility in Pasadena, California, that 

became a part of NASA in late 1958. The early evolution of the Deep 

Space Network illustrates how a major communication system can 

be firmly established through a combination of carefully chosen 

initial elements, put in place during a period of limited time and 

funding, arid later additions, installed as requirements become more 

ciemanding and further resources (such as funding and cooperating 

agencies) become available. 

The Cold War origin of solar-system exploration 

An official requirement for a system to communicate with 

space probes developed for the first time on 7 March 1958, when 

the Eisenhower Administration, through the Department of Defense’s 

new Advanced Research Projects Agency, authorized a program of 

five lunar-probe attempts, three by the Air Force and two by the 

Army, all to be conducted within a year.  The Administration publicly 

characterized the program (shortly to be named Pioneer) as a 

scientific project an effort “to determine our capability of 

exploring space in the vicinity of the moon, to obtain useful data 

concerning the moon, and provide a close Iook at the moon. ” Archival 

records show, however that the major impetus for the program disapproval was a desire by many inside and outside of government to 

find some quick means of restoring international prestige to the 

United States, after the Soviet Union’s successful orbiting the 

previous October of WJIDK  the world’s first artificial satellite, 

had shattered a widely-held percepticln of American technological 

superiority. 

In the six months between this event and Pioneer program 

announcement, in fact, numerous proposals for immediate “Moon 

shots” had been submitted to the Pentagon, and many cited the 

perceived Soviet threat. one of the first institutions to do so was 

JP1 . In a proposal entitled “Project Fled Socks” issued on 21 October 

1957, the lab observed that the launching of Sputnik 1 ICNS than 

three wcmks earlier “has had a tremendous impact on people 

everywhere” and that it “has significance which is both technical 

and political. ” The proposal stated that it was “immediately 

imperative that the United States regain its stature in the eyes of 

the world by producing a significant technological advance over the 

Soviet Union.” Pointing out that all had “some fairly sophisticated 

instrumentation and communication” capability that would allow it 

to achieve a successful lunar flyby mission, the lab advocated that 

the country “go to the moon instead of just going into orbit.” 

JPL was not alone in perceiving a potential political benefit 

deriving from a successful lunar mission. F{amo Wooldrige’s newly 

formed Space Technology Laboratories (S1 L ), located in I os Angeles, 

California, argued, in a proposal entitled “Project 13aker” issued on 

27 January 1958, that an early lunar flight with a moderate payload 

of scientific instruments could make a determination of conditions 

on the Moon that would be valuable for planning later flights with 

much heavier payloads that were certain to come within a few years. 

The firm also suggested, however, that “Of greater national 

importance may be the prestige of sencfing the first rocket to the 

moon, with clear proof that it reached its objective. ” 

Scientists and politicians, however,  were initially not 

enthusiastic about these and other lunar-probe proposals. The 

director William  Flickering recalled that members of the Office of 

Defense Mobilization’s Scientific Advisory Committee (ODMSAC) 

“were not sure that [the Red Socks proposal] was more of a stunt, as 

it were, and were not really that enthusiastic about it from a 

scientific point of view. ” Deputy Secretary of Defense Donald A. 

Quarles testified before Congress in late November 1957 that he 

found “no cause for national alarm” in the existence of the USSF1’s 

Sputnik satellites and argued that the United States “must not be 

talked into ‘hitting the moon with a rocket’ just to be first, unless 

by doing so we stand to gain something of real scientific or military 

significance. ” Eisenhower himself tolci colleagues that he would not 

be drawn into a “pathetic race” with the Soviet Union, and he 

characterized a lunar probe as “useless. ” 

The views of the scientists and politicians regarding “Moon 

shots” gradually changed, however, especially after the United 

States’ first attempt to launch a satellite (Vanguard) on 

1957 ended in spectacular failure--the explosion of the first stage 

of the launch vehicle within seconds of liftoff was recorded on live 

television. On 17 February 1958 the Space Science Panel of the new 

President’s Scientific Advisory Committee (reorganized from the old.J- 

OLMSAC) held a meeting in the Executive Office Eluilding (next to the 

White House) at which panel member Herbert York announced, to 

attending representatives from JPL and S-IL, that “it had been 

decided to attempt a lunar mission with the objectives of: a. Making 

contact with the moon as soon as possible, but with the limitation, 

 that the contact be of a type that has significance such that the 

public can admire it. ” York further stated that the panel had 

concluded, given the second objective, that “some kind of visual 

reconnaissance” (e.g., a camera to take a picture of the back side of 

the Moon) was the most significant experiment that a lunar vehicle 

could carry. PSAC’S endorsement c)f an early lunar mission would 

lead to the aforementioned Pioneer program authorization in late 

March. 

Supporting the Pioneer probes: S71 ‘s short-term approach 

The 

of launch 

would trai 

positions, 

for withou 

Pioneer program would require simultaneous development 

vehicles, spacecraft, and ground-support stations that 

lsrnit commands to the spacecraft, determine their 

and receive data from thcm. The stations were important, 

them no close-up photograph of the Moon could be 

c] btaincd and, more fundamentally, no confirmation that tho 

spacecraft were anywhere near the Moon was possible. 

[Iut what kind of network of stations should be set. Should 

it be cicsigneci solely to support the Pioneer program and its limited 

objective of photographing the Moon? Or shou Id a more elaborato 

system be constructed that would meet not only the requirements of

the pioneer program but also the anticipated needs of future 

programs not yet authorized? 

STL, initially under the leadership of Frank Lehan, had little 

choice but to undertake the short-term approach. Because of the 

more ready availability of their iaunch vehicles (1 hor IRF3M and 

Vanguard upper stages, the. three Air Force probes would be launched 

first, beginning in nlid-August 1958. l-his situation wouid allow the 

Air Force and STL the initial opportunity to reap the glory of a 

successful first lunar mission, but it allowed the latter less than 

five months to set up a network of ground-support stations. 

By necessity, the antennas used at the two principal stations 

had to be already erected or at least manufactured, and their 

locations were governed by the roles they would play in 

communicating with the lunar probes while they were in the vicinity 

of the Moon. For example, a 60-ft-diameter parabolic antenna with a 

transmitter a modification of the l“LM-18 antenna that Foundation, 

Inc., was currently manufacturing for we in the forthcoming Air 

Force Discoverer reconnaissance-satellite program--was installed 

at South Point on the island of t Iawaii because there it would have a 

favorable look-angle at the probes at the time of their fourth-stage 

retrorocket firings. 

S-l L planned for the picture taking to occur as soon as the 

probes entered orbit, before anything might go wrong with the 

spacecraft, and this milestone was expected to occur over about the 

0° longitude, which crossed parts of Europe and Africa. Lehan and 

his colleagues knew that the quality c]f the picture taking would 

improve as the diameter of the receiving ground-based antennaincreased, but the time constraint, as well as diplomatic and funding 

considerations, did not permit the installation overseas of a new 

large antenna, possibly one 200 ft or more in diameter. l-he 

University of 

Jodrell Bank, 

an Air Force 

Bank facility, 

Manchester’s 250-f t-diameter  radio telescope at 

however, already existed. A secret meeting between 

officer and Bernard Lovell, the director of the Jodrell 

enabled STL to install temporarily an appropriate feed 

and other specialized equipment on the antenna in support of the 

picture-taking activity. 

S1 

- 

The engineers appear to have given little thought initially as 

to what might constitute a permanent system of stations for 

supporting an ongoing program of unmanned solar-system spacecraft 

exploration, and whether any of the antennas installed or modified in 

1956 could become part of such a permanent system. JP1. engineers, 

by contrast, began planning for a permanent system even before the 

Pioneer lunar-probes authorization was issued. 

JP1. looks to the future 

Probably the strongest advocate for s ch a permanent system 

/ Y,< /,, J) 

was Lberhardt Flechtin, chief of JPL’s CEO Research S@ion. 

More aware than his colleagues in the propulsion field of the likely 

advances in electronics and the potential distances that could be 

reached in space communications (l-able xx), he strongly urged, in 

the spring of 1958, the development of a launch vehicle (Juno IV) 

capable of delivering a 550-pound payload to the Moon and a 300- 

pound payload to the planet Mars. Such a vehicle, he argued, was 

needed “to accomplish significant missions competitive with the 

USSR; lesser vehicles will only keep us to the rear in 

accomplishment of missions. ” Juncl IV’S capability of soft landing on 

the Moon, Rechtin pointed out, could eventually permit the 

establishment of “quite stable” radio and optical telescopes on the 

lunar surface. 

As for Mars, Rechtin argued that the often discussed similarity 

of this planet to the Earth would rnakc photographic exploration of it 

“one of the major goals of prestige between the United States and 

the USSR. ” Looking further into the future, he noted that 

meteorological and surface-condition instrumentation could 

determine “the practicality of putting people on Mars.” Rechtin 

predicted that “if conditions on Mars are even slightly more suitable 

than anticipated, the past success of the human race in new 

exploration will unquestionably start the drive to Mars. Based on 

human history, it will then be first come-first served on Mars. ” l-eft 

unsaid, but most likely implied, was the desire that the United 

States get there before the USSR. 

Rechtin was not alone at JP1 in perceiving Mars and other 

planets of the solar system as the ultimate goals of space 

exploration. Albert Hibbs, who became the first chief of JF)l’s new 

Space Science Division, recalled in an interview that “[W]e wanted a 

good challenge, and that was & technical challenge, getting a 

useful payload to a planet. It was really tops in engineering 

challenge--propulsion, guidance, communications, you name it. ” 

It was for this envisioned amt]itious program of lunar anc~ 

planetary missions that JP1 , and particularly F{echtin and his fellowcommunications engineers, desired in early 1958 to build a 

permanent network of stations that could transmit commands 

spacecraft, determine their positions relative to the Earth or 

to 

other 

objects, and receive scientific and cngirrcering telemetry data from 

them. Rechtin’s conception of a permanent network was based on a 

consideration of the apparent motions of space probes and a 

requirement, sure to be iniposed by any funding agency, to keep costs 

to a minimum.  

He knew that after a space probe launched from Cap& Canaveral 

completed its injection phase, during which it would move rapidly to 

the east, it would (due to a decreasing angular velocity as it gained 

altitude) have an apparent motion from east to west that closely 

approximated that of a fixed radio source. During this post- 

injection phase the greatest components 

motion will be due to the rotation of the 

 Obviously results in the probe apparently 

the eastern to the western horizon of a 

of the probe’s apparent 

Earth, and such rotation 

moving across the sky from 

particular antenna station 

once each day. Simple geometry dictates that the minimum number 

of principal antenna stations that permits continuous, overlapping 

monitoring (necessary as missions became more complex and longer 

in duration) after the injection phase is three . Because 

the world is divided into 360° of longitude, the three stations should 

ideally be located 120° apart in longitude. 

d Confid&rt that solar-system exploration would “continue in the 

coming years, ” Rechtin and his colleagues--particularly Walter K. 

Victor, head of the Electronics Research Section, and Robertson 

Stevens, head of the Guidance 1 echniques Research Section--sought 

a communication system design that would “be commensurate with 

the projected state of the art, specifically with respect to 

parametric and maser amplifiers, increased power and efficiency in 

space vehicle transmitters, and future attitude-stabilized 

spacecraft .“ Because of the later availability of the Army lunar- 

probe launch vehicles (Jupiter IF{BM and a cluster of upper stages 

employing Baby Sergeant rocket motors), they had just enough extra 

time to design and install a communication system that could not 

only support the Pioneer lunar probes, but also evolve into a 

permanent system for supporting future solar-system exploration 

spacecraft. 

Choosing an antenna design

With regard to antenna design, Rechtin, Victor, and Stevens 

ciesired an instrument with an accuracy of 2 minutes of arc or 

better, Operation on a 24-hour basis dictated that this accuracy 

would have’ to be maintained regardless of solar exposure and rapid 

ambient temperature changes. Furthermore, “since missile [la 

vehicle] firings cannot bc held up because the wind is blowing 

somewhere around the earth nor can the bird [spacecraft] be 

Inch 

whistled back from a space mission when the wind comes up,  the 

antenna would have to be usable in winds of 60 mph and be capable 

of withstanding (in a stowed position) winds of 120 mph. 

Rechtin assigned William Mcrrick (head of the Antenna 

Structures and Optics Group) to icientify an antenna design that could 

satisfy these demanding requirements. Confident that JPL wouldreceive an lunar-mission assignment but aware that the “ 

procurement, fabrication, and erection of the antennas would be the 

“longest lead time item” for carrying out such an assignment, 

Rechtin made this assignment on 7 February 1958, nearly seven 

weeks before the Pioneer authorization. Merrick concluded that the 

desired antenna would have to combine the best features of a 

precision radio-astronomy antenna and a precision guidance or 

tracking radar. Merrick recalled later’ that the radio astronomers 

and suppliers he consulted “questioned our sanity, competence in the 

field ar)d/or our ability to accomplish the scheduled date [initially 

November 1958] even on an ‘around the clock’ basis.” 

Merrick and his colleagues rejected many existing antenna 

designs because of foreign manufacture, high cost, inadequate 

aperture, and/or acknowledged design flaws. Others, such as the 

CSIRO’S 21 O-ft diameter antenna at F’arkes, Australia, and NRAO’S 

140-f t-diameter antenna at Green Bank, West Virginia, were 

eliminated from consideration because these prototypes would not 

be completed until 1960 or later. 1 he Jodrell Elank type c)f antenna 

was rejected because it was “too big and expensive” and its design 

and assembly had required seven years. 

Merrick and his colleagues ultimately chose a design that had 

been initiated at the Naval Research laboratory in 1953, developed 

further by t-toward W. 1 atel at the Carnegie Institution of 

Washington, and refined by the Associated Universities, Inc. (AUI), 

and that had just been completed by the E31aw Knox manufacturing 

company in Pittsburgh. the 26-m-diameter (85-ft) antenna had a 

cantilevered-equatorial mounting and very large hour-angle and

. 

dcclination drive gears that gave high driving accuracy for relatively 

low tooth accuracy and a low tooth loading during high winds. EIlaw 

Knox, which priced the antenna at about $250,000, had already 

received orders from the University of Michigan and AU I (for 

erection at Ann Arbor and Green Bank, respectively), but neither had 

been completed when JPL placed an order, with ARPA’s approval, for 

Eventually, citing national priority, the 

three antennas in April. . 

Army was able to move one of these probe-supporting antennas to 

the front of the manufacturing line, 

Choosing a station site

That first antenna was slated for a site in the United States. 

Rechtin later recalled the planned overseas stations “so rapidly 

became bogged down in approval red tape” that their earliest 

possible activation date gradually moved beyond the second Army 

lunar-probe attempt. Three stations would be essential for possible 

future long-duration flights to the planets, but the limited objective 

of the Army lunar probes allowed JPL engineers to make do 

temporarily with one antenna. Continuous around-the-clock 

monitoring of the probes was of course impossible, but JPL 

engineers could deliberately select a trajectory that would cause 

them to arrive at the vicinity of the Moon when they were in the line 

of sight of the single principle antenna. Also, they, unlike their 

counterparts at S3 L., had no need for a separately located 

transmitter station.the probes were slated to fly by the Moon (thus 

requiring no retrorocket firing commands), and the desired pictureswould be taken automatically when a photocell mechanism indicated 

that the probes were within a certain distance of the Moon. 

With the expectation that probes would eventually be sent to 

the planets and thus their received signals would be extremely 

weak, JPL communication engineers desired a site for their single 

initial principal antenna that would minimize outside radio 

interference as much as possible. In addition to avoiding areas with 

power lines, radio stations, radar transmitters, and/or considerable 

numbers of aircraft passing overhead, they sought in particular a 

natural bowl, so that the surrounding terrain could shield the 

antenna from nearby towns and passing vehicles. The underlying soil 

had to be suitable for accurate and stable support of the antenna, 

and an access road, for transport of the sizable steel components of 

the antenna, would have to be built for what was likely to be a 

remote site. Finally, the more immediate funding and time 

constraints of the Pioneer program mandated use of Govern merit- 

owneci land. 

Thanks to a search two years earlier for an off-lab site to test 

rocket engines, JPL engineers were aware that an area near 

Goldstone Dry [Lake at the Army’s Fort Irwin, located in the Mojave 

Desert about 150 mi northeast of Pasadena, would meet these 

criteria. After General John B. Medaris, the head of the Army 

Ballistic Missile Agency, in mid-May 1958 overruled another general 

who wanted to use the Goldstonc area for a proposed missile firing 

range, the work needed to convert the site into the desired antenna 

station swung into high gear. Carefully avoiding unexploded 

ordinance lying in the area, workers constructed access roads, laid

the antenna foundation, and constructed support buildings during the 

late spring and early summer. Soon after he steel components 

arrived in mid-August, a crew from the Ratio Construction Company 

began erecting the antenna. After the crew completed its work two 

months later, the feed was installed and various optical and radio- 

frequency tests were conducted to establish the system tracking 

accuracy. 

Choosing an operating frequency 

Unlike their counterparts at S1-L, JPL engineers, led by Victor, 

chose not to operate at the 108 Mtlz frequency being used for the 

Vanguard and Explorer satellites. With future missions clearly in 

mind, they noted in an early report that the presence of interference 

at frequencies below 500 MHz would “seriously limit the growth 

potential of any space communication technique” using a frequency 

in this region. Victor at first favored a frequency in the region 

between 1365 and 1535 MH7, where he anticipated significant 

hardware developments for improving receiver sensitivities because 

the region bracketed the astronomically important 21 -cm hydrogen 

line. Colleagues soon convinced him that a stable, efficient 

spacecraft transmitter operating in that region could not be built in 

time for the Pioneer probe missions, however, and he instead opted 

for a 960 MHz (L-band) operating frequency. 

l-he hard work that ST1.. and JPL communications engineers 

expended in setting LJp their respective systems of antenna stations 

(which included several with smaller antennas at launch-point anddownrange locations) in relatively short time periods paid off in 

very satisfactory operation during the actual missions. Various 

rocket failures, however, prevented all but the second Army probe 

(launched on 3 March 1959) from reaching escape velocity, and this 

probe (Pioneer 4) passed too far away (37,000 mi) from the Moon to 

activate the camera system. By then, the USSR’s .luna 1, launched on 

2 January, had already passed within 6,000 mi of the lunar surface, 

l.una 3, launched on 4 October 1959, took the first photographs of 

the far side of the Moon. 

Gaining approval for a permanent system 

The expansion of JPL’s ground-support system for the Pioneer 

lunar probes into a complete worldwide three-station network was 

not inevitable. The first challenge to JPL’s plans came from STL, 

which in late June 1958 iSSLJed a proposal that called for the 

construction of three 250-f t-diameter antennas to be located in 

t{awaii, Singapore or Ceylon, and near the eastern coast of Brazil. 

The firm claimed that diSCLJSSiOnS with JPL and “a thorough analysis 

of foreseeable space programs” (including a series of new probes 

aimed at the planets Venus and Mars that STL was simultaneously 

proposing) indicated that “the long-range interests of the United 

States in high-altitude communications relay satellites and in 

interplanetary space programs could best be served” by the 

establishing of two networks of three stations each, placed at 

intervals of about 60° around the equator of the earth.

- 16 

Rechtin thought otherwise; he considered the proposal “a ploy 

to block JPL’s [network plans] by forcing a study and reconsideration 

of JP1.’s ARPA order [for three 26-m-ciiameter antennas]. ” He may 

have been right. The estimated overall cost of STL’S proposed new 

sy$tem was $34 million. How STL expected the government to 

(

approve such a large sum in so short a time (the company claimed 

that it could “realistically” complete construction of the first 

antenna in Hawaii by 15 October 1959) is unclear. The proposal, in 

any case, was not funded. 

A greater threat came in early July 1958, when Deputy 

Secretary of Defense Donald @larlOs questioned why S1’L and JPL 

were developing two separate systems for supporting the Pioneer 

lunar probes. In response, Rechtin traveled immediately to 

Washington, and in a 8 July meeting at the Pentagon with Richard 

Cesaro, chairman of an AFIPA advisory panel on tracking, he 

acknowledged that JPL was using the extra time afforded by the 

later launch dates of the Army lunar probes “to begin a longer range 

space tracking program using the proper parameters. ” These 

parameters included the 960 Mtlz operating frequency and the 26-m- 

ciiameter antennas that would be “capable of tracking all vehicles 

from a 330-mile altitude satellite to space probes to Mars. ” 

Cesaro was impressed with Flechtin’s presentation, but asked 

that JPL prepare a formal proposal for a “World Net” that would 

consider as well the communications requirements of other intended 

ARPA space programs. J P L ‘s Aetianeta . 

rv T-racking 

Network, issued on 25 July, considered (despite its title) such 

requirements for six different space programs that the UnitedStates planned to undertake--manned space flight, meteorological 

satellites, reconnaissance satellites, communications satellites 

(both geosynchronous and low-Earth-orbiting), scientific satellites, 

and space probes--as well as the detection of “noncooperative” (i.e., 

foreign) satellites. Comparing all the requirements ), 

Rechtin and several colleagues suggested that two principal 

overseas antennas could be most advantageously placed, for 

supporting space probes and certain other space programs, at 

Woomera, Australia, and somewhere in Spain. 

Cesaro was once again impressed with JPL’s work, and 

indicated to Rechtin his intention to recommend that “all the 

tracking and computational facilities should be handled under 

administration with JPL as the technical arm. ” Rechtin was 

delighted with this recommendation, but nevertheless cautious. 

believed  that Cesaro “may be way over optimistic” in thinking 

Army 

He 

that 

“ARPA certainly has the power to do this and would put down any 

rebel lion.” In particular, Rechtin warned a JP1 colleague that “we 

should expect considerable uproar from the Naval Research 

Laboratory who probably figures it knows more about tracking than 

anybody else. ” The NRL’s Radio lracking Branch, under the leadership 

of John 1. Mengel, had developed the Minitrack tracking system for 

the Vanguard satellite program. 

1 he basis for Rechtin’s caution was his knowledge that 

Congress in the summer had approved President Eisenhower’s 

request 

Nation 

into being 

for establishing a civilian space agency, and as a result the 

Aeronautics and Space Administration was slated to come 

 on 1 October 1958. NASA’s impending formation meant

that ARPA was gradually losing its status as the interim United 

States space agency. 

Furthermore, by early January it became clear that not only did 

the Defense Department want a station network separate from any 

set up by NASA (because their need for secrecy conflicted with the 

new space agency’s professed openness), but also those involved in 

setting up NASA’s manned-space-flight, satellite, and space-probe 

programs desired separate station networks. As Rechtin feared, 

JPL’s plans were also strongly opposed by Mengel, whose group had 

already been transferred into NASA. Mengel claimed that the 

installation of more Minitrack stations was more essential than 

than overseas space-probe-supporting stations, because “the 

satellite experiments and their associated tracking was more 

important [than space probes] as far as NASA plans were concerned.” 

Despite Mengel’s views, on 10 January 1959 NASA, which had 

supported since early November JPL.’s development of a recommended 

set of future lunar and planetary probes and had also acquired JP1. 

from the Army, signed an agreement with the Department of Defense 

that called for, among other things, installation of stations for 

deep-space probes at Woomera and in South Africa. The preference 

by NASA and JPL for South Africa as the host country for a dedicated 

probe-supporting station derived from the fact that most space 

probes would pass over southern Africa during the injection phase of 

their flights, when it was vitally important to establish their actual 

trajectories for later accurate pointing of the other probe- 

supporting antennas.Overseas expansion 

In establishing the overseas stations, Rechtin insisted that 

they be operated by local nationals rather than “displaced 

Americans. ” Desiring the best possible performance from each of 

the stations, he reasoned (and was supported by later experience) 

that this could be obtained from professionals “proud of their work, 

held responsible, and cooperatively cc)mpetitive in spirit. ” NASA and 

JP1. fortunately identified in Australia and South Africa 

organizations--the Department of Supply’s Weapons Research 

Establishment (WRE) and the Council for Scientific and Industrial 

Research’s National Institute for Telecommunication Research 

(NITR), respectively--that were eager to cooperate in the 

establishment of the network for supporting space probes. 

l-he WRE was managing the Woomera rocket range at which the 

United Kingdom and Australian governments had been conducting 

high-altitude missile firings over the past decade, and WF3E; 

superintendent Bill Boswell anticipatecj that the addition of an 26- 

m-diameter antenna could not only expand support of these firings 

but also ensure Woomera “a leading place in satellite and space 

research, ” NITR director Frank Hewitt anticipated that the antenna 

would be a “most valuable scientific tclol” that could be used 

between missions to conduct radio-astronomy research. He also 

believed that the techniques involved with the antenna would be 

fundamental to future intercontinental communications an activity 

of great performance to a country quite distant from Europe and the

United States--and that therefore the NITR should become familiar 

with them. 

NASA sent site-survey teams to Australia and South Africa in 

February and September-October 1959. With extensive assistance 

from WRE, NITR, and other local officials, NASA and JPL eventually 

identified and selected two appropriate sites: a semi circular bowl 

open to the south at the edge of a dry lake bed known as Island 

lagoon about 18 mi from the village of Woomera and about 30 mi 

south of the range head, and a Y-shaped valley near the town of 

Hartebeesthoek about 30 mi northwest of Johannesburg and 18 mi 

west of Pretoria. 

NASA funded the construction of the overseas stations and 

sent field crews to erect the antennas and install the electronics, 

but it was WRE and NITR that bore the responsibility for acquiring 

the land, constructing access roads and support buildings, and hiring 

staff to operate the station. With a new program of Ranger lunar- 

impact probes scheduled to be launched beginning in mid-1961, both 

agencies worked hard with NASA and JPL to ensure that the stations 

would be ready in time. NITR’s success in doing so was made more 

difficult by delays (occasioned by the Sharpeville township 

disturbance and the Soviet Union’s downing of a United States U-2 

spyplane in March and May 1960) in the signing of a diplomatic 

agreement governing the station and other NASA facilities in South 

Africa, By the time Ranner 1 was launched on 23 August 1961, 

however, both stations were ready and the Deep Space 

Instrumentation Facility (renamed the Deep Space Network in 1963) 

at long last had become operational.Subsequent evolution 

Anticipating that space probes would become more 

sophisticated in future years and would eventually travel beyond the 

orbits of Venus and Mars (in contrast to the fixed range of Earth- 

orbiting satellites), Rechtin sought and received from NASA a 

continuing commitment that a relatively fixed portion (generally 

about 10 percent) of the Deep Space Network budget would be 

devoted to research and development. This commitment allowed the 

Network to evolve in a timely way in subsequent years, as new 

requirements were anticipated and means to meet them were 

conceived, tested, and installed. 

In early 1961, for example, Flechtin recognized that NASA’s 

deep-space program would soon be expanding very rapidly (further 

Rangers, Mariner flybys of Venus and Mars, Lunar Orbiters, Surveyor 

lunar soft-landings, and Apollo manned lunar landings). He could 

foresee occasions when “so many flights [would be] operating at any 

one time . . . that a single antenna at each DSIF station could not 

conceivably carry the load. ” Rechtin envisioned a future situation 

when project managers could be “confronted with impossible choices 

between probes measuring dangerous solar flares, observing violent 

effects c]n Mars, roving among the crevasses on tho Moon, and 

carrying men into deep space. ” 

New 26-m-diameter antennas were thus needed (to be 

accompanied by a change in operating frequency to S-band (2388 

MHz)), and the first of these antennas was installed at Goldstone in

early 1962. Although the Woomera and Hartebeesthoek stations 

would continue to operate through the early 1970s, neither was the 

site of the new overseas antennas. WRE; had difficulties fully 

staffing the Woomera station, due to its isolated location (in the 

outback about 200 mi north of Adelaide) and insufficient housing 

staff members and their families. Although the WRE gradually 

for 

resolved the staffing and housing problems at Woomera, the long- 

term solution was to find a new adequately shielded site nearer a 

center of population. Officials from JPL. and the Australian 

Departments of Supply and Interior eventually identified such a site 

in the Tidbinbilla Valley, located 11 rni sC)lJt west of Canberra 

(Australia’s capital) along the northeastern edge of the Australian 

Alps. The station constructed at this site became operational in 

March 1965. 

NASA and JPL were quite satisfied with 

the station at Hartebeesthoek, but Rechtin in 

that relations between the governments of the 

South Africa might eventually deteriorate, due 

NITR’s operation of 

particular was fearful 

United States and 

to condemnation in 

the United States and abroad of the latter’s apartheid policies, to a 

point where operations at this station would have to be sharply 

limited and curtailed. He argued that any expansion of the station 

would make it more costly for NASA to duplicate the station 

elsewhere at a later date. 

An initial survey of sites in Italy proved unsuccessful, A 

# 

survey team found natural bowls on the island of Sard~nia, but such 

,., 

a location would be difficult to support logistically. Potential sites 

near Rome would not have this difficulty, but they were less wellshielded and would create coverage gaps between a station here and 

the one at Goldstone. 

NASA and JPL ultimately chose a site in a valley near the 

village of Robledo de Chevala, 31 mi west of Madrid, Spain, for the 

location of a 26-m-diameter antenna. A second such antenna was 

subsequently installed near the town of Cebreros, 8 mi southwest of 

Robledo de Chevala, These stations became operational in July 1965 

and January 1967. Despite some misgivings about dealing with the 

Franco authoritarian government, NASA had been quite pleased with 

assistance rendered by the Spanish government’s Instituto National 

de T6cnica Aeronautic (INTA) in the operation of a Project Mercury 

station in the Canary Islands, and this organization became NASA’s 

cooperating agency for the new Deep Space Network stations in 

Spain as well. 

A major evolutionary step was the design and installation of 

new 21 O-ft-diameter antennas at Goldstone, Tidbinbilla, and 

Robledc) de Chevala. These were built in response to the expected 

advent of more sophisticated spacecraft (as launch vehicles became 

more powerful), which would create a requirement for an increasing 

rate in the communication of data from the spacecraft back to Earth. 

JPL considered a number of alternatives for meeting this 

requirement increasing the power of the spacecraft transmitter, 

electronic arraying of two or more 26-m-diameter antennas, and use 

of existing large radio-telescope antennas--but economics and 

availability considerations ultimately dictated the construction of 

new large antennas up to 250 ft in diameter. After two years of 

design studies and nearly four years of contract negotiation, ground

preparation, support-building construction, and antenna erection, the 

first of the Deep Space Network’s large antennas became operational 

in May 1966. Two other such antennas became operational at 

Tidbinbilla and Robledo de Chevalla in April and September 1973. 

The Deep Space Network expanded the newer of the 26-m- 

diameter antennas at Goldstone in 1978 in order to add X-band (8.4 

GHz) capability and increase the antenna gain (received signal 

strength) for the two Voyager outer-planet missions. Antennas at 

Tidbinbilla and Robledo de Chevala were similarly expanded in 1980, 

This improvement was sufficient for the Jupiter and Saturn 

encounters (1979-81) of the two spacecraft. The extension of the 

VQ2U mission to include encounters with the more distant 

planets Uranus in 1986 and Neptune in 1989, however, forced 

Voyager and Network engineers to find new means to compensate for 

a still more severe decrease in signal strength and thus avoid an 

undesirable great limitation on the science data return. 

One means was the installation of new 34-m diameter high- 

efficiency antennas (so-called because their reflector surfaces’ are 

precision-shaped for maximum signal-gathering capability) at 

Goldstone in 1984, Tidbinbilla in 1985, and Robledo de Chevala in 

1987. The 64-m diameter antenna and the two 34-m-diameter 

antennas could now form a three-element array. This combination 

(together with a reprogramming of two of the Voyager computers to 

accommodate an image data compression technique) permitted a 

higher data rate (19 kilobits/see). 

Because a still higher data rate would be needed to meet the 

imaging science requirements at Uranus and Neptune, Deep SpaceNetwork engineers sought and received permission from Australia’s 

Commonwealth Scientific and Research Organization to add 

temporarily (in 1986 and 1989) their 64-m-diameter radio 

telescope at Parkes to t 

link. The Network made 

when the twenty-seven 

 Tidbinbilla array via a ground microwave 

use of a second interagency array in 1989, 

25-m-diameter radio-telescope antennas of 

the National Radio Astronomy Observatory’s Very Large Array in New 

Mexico were linked with the Goldstone’s antennas. 

One further step taken for the Neptune encounter was the 

extension of the 64-m diameter antennas at each station to a 

diameter of 70 m and the the reshaping of their reflector surfaces 

to improve their efficiency. These improvements, which increased 

the effective signal capture of these antennas by 5:lp~r ce t where 

completed at Tidbi#Ma.and Robledo de Chevala in+) and 

Goldstone in 1988. 

The Deep Space Network continues to evolve even today. The 

original 26-m diameter antennas installed in the 1958-61 period 

are no longer in service the one at Goldstone is now a national 

monument, the one at Woomera has been scrapped, and th& one at 

Hartebeesthoek is now used by South Africans for radio-astronomy 

research. The second set of such antennas (those extended to 34 m 

in the late 1970s) are nearing the end of their usefulness. The onset 

of metal fatigue and the mechanical limitations of their late 1950s 

design do not permit further upgrades to improve performance. l-he 

Deep Space Network will therefore soon be replacing these antennas 

with 34-m-diameter multifrequency bearn-waveguide antennas. 

These new antennas will allow critical weather-sensitive

microwave components to be located in an equipment room in the 

antenna pedestal rather than on the rotating and tipping main 

reflector. The first of these new antennas was recently installed at 

Goldstone and will shortly become operational after completion of 

performance testing. 

To probe further 

The author is nearing completion of a book-length history of 

the Deep Space Network that will be based on published sources, oral 

history interviews, and unpublished dc]cuments in archives in the 

United States, Australia, South Africa, and Spain. Photocopies of 

the documentation supporting the book (and the article above) will 

be deposited in the JPL Archives. His article “Designing the United 

States’ Initial ‘Deep Space Networks . ...” jEEE Antennas and 

+Qtion ‘aaazine’ vol. 35, no. 1, February 1993, provides 

additional detail concerning the choices of antenna design, operating 

frequency, and antenna location made by STL. and JPL for supporting 

the Pioneer lunar-probe attempts of 1958-59. 

William R. Corliss’s A History gf the Deep Space Netw@