A History of the Deep Space Network

Key Takeaways

• The Deep Space Network has tracked spacecraft since 1958, spanning six decades of exploration.
• Three antenna complexes on three continents provide continuous 360-degree sky coverage.
• DSN upgrades through 2030 will support Artemis, Mars missions, and deep space science.

Introduction

The story of humanity’s reach into deep space is, in large part, the story of a network of dish antennas scattered across three remote corners of the planet. Without the Deep Space Network, or DSN, the images from Voyager 1 would never have arrived, the landing of Curiosity would have gone unwitnessed in real time, and the New Horizons flyby of Pluto would have been a silent event, its data forever unretrievable. The network exists because someone in the late 1950s recognized that sending a spacecraft into deep space was only half of any problem. Getting data back was the other half, and it turned out to be an engineering challenge of a completely different magnitude.

The Problem That Created the Network

Before the DSN formally existed, American scientists and engineers were already grappling with a reality that the first satellite launches made unavoidable. Sputnik 1 launched on October 4, 1957, and the United States, scrambling to respond, quickly recognized that its ground infrastructure for tracking and communicating with spacecraft was fragmentary at best. The U.S. Army, Navy, and early civilian programs each had their own tracking setups, none of which were designed with interplanetary distances in mind.

NASA came into existence on October 1, 1958, and almost immediately inherited both the ambition to explore deep space and the logistical nightmare of figuring out how to do it. The agency’s Jet Propulsion Laboratory, known universally as JPL, had been tracking early U.S. satellites and quickly became the institution most focused on the communications problem. JPL scientists understood that radio signals from a spacecraft near the Moon or Mars would arrive at Earth extraordinarily faint, attenuated by the inverse-square law across distances that dwarfed anything communications engineers had previously considered.

To detect a signal from a spacecraft tens or hundreds of millions of miles away, you need a very large antenna, very sensitive receivers, and a way to average out noise over time. You also need ground stations positioned around the globe so that as Earth rotates, at least one antenna always has line-of-sight to the spacecraft. Those three requirements — aperture, sensitivity, and geographic distribution — defined what the DSN would become.

Pioneer Beginnings

JPL’s first serious foray into deep space communications came with the Pioneer program. Pioneer 1 launched in October 1958 and reached a distance of about 113,800 kilometers before failing. The tracking infrastructure at the time was improvised, relying on a collection of small dish antennas and radio receivers that had been assembled quickly.

The station at Goldstone in California’s Mojave Desert became the anchor of what would eventually be the DSN. The Goldstone site was chosen for practical reasons: its remote desert location minimized radio frequency interference, the terrain was flat enough for large antenna construction, and it sat at a latitude that offered good coverage of the southern sky, where most planetary trajectories would eventually lead. A 26-meter antenna was constructed at Goldstone and used to track Pioneer 4, which launched in March 1959 and flew past the Moon, becoming the first American spacecraft to escape Earth’s gravity.

That early tracking work revealed something that JPL engineers had suspected but now saw in hard data: the existing antenna capability was barely sufficient for near-Earth missions, let alone anything heading toward Mars or Venus. Plans for larger dishes began almost immediately.

JPL Takes the Lead

On December 24, 1963, NASA formally established the Deep Space Network as a separately managed entity under JPL’s authority. The date is often cited as the network’s official birthday, though the infrastructure and operational concepts that defined it had been building for years before that formalization. JPL became the managing institution, a role it has held ever since, operating the DSN on behalf of NASA and, over time, on behalf of international partner agencies.

The decision to give JPL operational control was not universally popular within NASA. The agency had other centers that were developing deep space capabilities, and there were real debates about whether a single centralized network made more sense than a collection of mission-specific antenna systems. JPL’s advocates won that debate, and the centralized model has proven durable — though it has also created periodic friction between mission teams and the network operations staff who allocate antenna time.

The Three-Station Architecture

The geographic logic of the DSN is elegant. Earth completes one rotation every 24 hours, which means a spacecraft that’s continuously transmitting will spend roughly eight hours overhead any given point on the planet’s surface. To maintain continuous contact with a spacecraft, you need ground stations separated by approximately 120 degrees of longitude, so that as one station rotates away from view, another rotates into it.

The three complexes that satisfy this geometry are Goldstone in California, Robledo de Chavela near Madrid, Spain, and Tidbinbilla near Canberra, Australia. Each complex houses multiple antennas of different sizes, giving the network flexibility to handle different mission types simultaneously. The Goldstone complex covers spacecraft visible from the Western Hemisphere; Madrid handles coverage for Europe, Africa, and much of the Middle East; Canberra takes the southern hemisphere and the Pacific region.

The Australian station is operated by the Commonwealth Scientific and Industrial Research Organisation, known as CSIRO, under a cooperative agreement with NASA. The Spanish station is operated by the Instituto Nacional de Técnica Aeroespacial, or INTA. These arrangements are meaningful — the DSN has always been an international operation, even if it’s funded and directed primarily by the United States.

Racing Alongside Apollo

The 1960s put extraordinary pressure on the DSN. NASA’s planetary exploration program was accelerating rapidly, with Mariner spacecraft heading toward Mars and Venus, and the Apollo program building toward a Moon landing that would require communications infrastructure of its own. The DSN and the Manned Space Flight Network were separate systems — the lunar-focused infrastructure for Apollo was managed independently — but the two networks shared resources, expertise, and in some cases, antennas.

The 26-meter antenna at Goldstone that had tracked Pioneer 4 was upgraded and supplemented. A 64-meter antenna, one of the largest dish antennas in the world at the time, began construction at Goldstone in the mid-1960s and was completed in 1966. Similar 64-meter dishes were built at the Madrid and Canberra complexes. These larger antennas were essential for the outer solar system missions that JPL was already planning, even as the immediate work focused on closer targets.

The Mariner 4 mission, which flew past Mars in July 1965, gave the DSN its first real test with a distant planetary target. The spacecraft transmitted 22 photographs of the Martian surface — the first close-up images of another planet ever obtained. The data came in slowly, at 8.33 bits per second, and the DSN’s 26-meter and 64-meter antennas at Goldstone worked through the link to receive every bit of it. The images revealed a cratered, Moon-like surface that was significantly different from the canal-bearing Mars that some scientists had hoped to find.

Expanding the Network in the Late 1960s

The late 1960s saw rapid expansion of the DSN’s physical infrastructure. New stations were added at sites in Spain, Australia, and California, and a station in Johannesburg, South Africa provided additional southern hemisphere coverage, though that station was eventually closed. A facility in Woomera, Australia also operated for a period before consolidation efforts moved its functions to the Tidbinbilla complex near Canberra.

The network was also expanding its technical capabilities in ways that weren’t always visible as hardware additions. Signal processing improvements, better receiver designs, and advances in coding theory all allowed the DSN to extract more data from signals that hadn’t changed in strength. This kind of gain — squeezing more information from the same faint radio signal — would continue to drive DSN performance improvements for decades, often outpacing the gains from simply building bigger dishes.

Antenna arraying, the technique of connecting multiple dishes electronically so that their signals combine to simulate a larger single aperture, was also being developed during this period. The mathematics behind arraying had been understood for years in radio astronomy, but applying it operationally to spacecraft tracking required new engineering approaches. Early arraying experiments at Goldstone demonstrated that combining signals from two 26-meter antennas could approach the sensitivity of a single larger dish.

The Voyager Era and Its Demands

No missions in the history of planetary exploration have asked more of the DSN than the Voyager program. When Voyager 1 and Voyager 2 launched in the summer of 1977, the network’s 64-meter antennas were the primary receiving stations. As the spacecraft traveled farther from the Sun, their signals grew progressively weaker, and the DSN had to keep pace.

The Jupiter flybys in 1979 were manageable — the spacecraft were still relatively close. Saturn in 1980 and 1981 pushed harder. By the time Voyager 2 reached Uranus in January 1986, it was so far away and the transmission power so limited that heroic measures were required to maintain a data link. JPL engineers implemented a technique called data compression on the spacecraft itself, combined with improvements in the DSN’s receivers and the arraying of multiple antennas.

For the Uranus encounter, the DSN arrayed its 64-meter Goldstone antenna with an antenna at the Very Large Array radio astronomy facility in New Mexico, and with antennas at the Japanese Usuda Deep Space Center. This international arraying arrangement, stitching together dishes across different continents and agencies to form a virtual aperture larger than any single antenna, was an operational and diplomatic achievement. The Uranus encounter returned detailed images of a planet that ground-based telescopes had barely resolved.

The Neptune encounter in August 1989 pushed even harder. Voyager 2 was so distant that its signal arrived at Earth having crossed nearly 4.5 billion kilometers. The network upgraded its 64-meter antennas to 70-meter apertures — a physical extension of the dish surface area — in time for the encounter. The Goldstone 70-meter antenna, the Madrid 70-meter, and the Canberra 70-meter all participated, arrayed with additional antennas to maximize signal collection. The Neptune flyby images, including the first close-up views of Triton, arrived sharp and detailed despite the staggering distance. That outcome was the product of years of preparatory engineering work at JPL and the three DSN complexes.

The 70-Meter Upgrade

The physical expansion of the three primary dishes from 64 meters to 70 meters in diameter, completed between 1987 and 1988, represented one of the DSN’s largest infrastructure projects. Each dish surface was extended outward, adding additional antenna panels to increase the collecting area. Increasing a circular dish from 64 to 70 meters increases the collecting area by roughly 20 percent, which translates directly into a similar gain in signal strength. On signals measured in fractions of a watt after traveling billions of kilometers, 20 percent matters.

The upgrade required careful structural analysis of the existing antenna support structures, which hadn’t been designed with the additional weight and wind loading in mind. Engineers at all three complexes had to reinforce the elevation and azimuth drive systems while keeping operational downtime to a minimum. The fact that the network had three separate complexes helped — each antenna could be upgraded in sequence while the other two maintained network coverage.

Mars and the 1990s

The 1990s brought both triumphs and frustrations to the DSN. Mars Observer, a spacecraft expected to enter Mars orbit in August 1993, fell silent three days before orbital insertion and was never heard from again. The DSN’s antennas tracked the spacecraft until the moment contact was lost, and they searched for any residual signal for weeks afterward. None came.

Mars Global Surveyor, which arrived at Mars in September 1997, restored confidence in the program. The DSN supported its aerobraking phase — a long process in which the spacecraft gradually lowered its orbit by skimming through the upper Martian atmosphere — and then provided the communications link for years of orbital mapping operations. Mars Global Surveyor returned more scientific data about the Martian surface than all previous missions combined, most of that data flowing through the DSN’s antennas.

Mars Pathfinder landed in July 1997 and deployed the Sojourner rover, which became a cultural sensation. The mission’s direct-to-Earth communication link went through the DSN, and the volume of public interest in Mars Pathfinder was so intense that JPL’s website briefly became one of the most-visited sites on the Internet. The DSN’s role was invisible to most of the public watching rover images online, but without the network’s receiving capability, those images couldn’t have existed.

That same decade also saw the Galileo mission to Jupiter straining the DSN’s capabilities in a different way. Galileo’s high-gain antenna, a large deployable dish intended to be the spacecraft’s primary communications channel, failed to open fully after launch. The mission team and DSN engineers spent years working around this limitation, relying on a small low-gain antenna that transmitted data at a fraction of the intended rate. The DSN compensated by allocating maximum antenna time to Galileo, arraying its large dishes, and implementing sophisticated data compression and coding improvements. Despite the antenna failure, Galileo returned valuable science from Jupiter and its moons over six years of operations.

Cassini and Saturn’s Rings

Cassini, which launched in October 1997 and arrived at Saturn in July 2004, became one of the most data-intensive missions the DSN had ever supported. The spacecraft orbited Saturn for 13 years, continuously transmitting scientific data. During the Huygens probe descent to Titan in January 2005, the DSN’s antennas tracked both the Cassini mothership and tried to receive the Huygens signal directly, in addition to receiving data relayed through Cassini.

The Huygens descent was a complex relay arrangement: Huygens transmitted to Cassini, which stored the data and then relayed it to Earth through the DSN. The scientific return from that descent — the first landing on a body in the outer solar system — included audio recordings of Titan’s wind and images of a surface with channels carved by liquid methane. That data came to Earth through the 70-meter antennas of the DSN, crossing nearly 1.5 billion kilometers.

Cassini’s 13-year mission at Saturn represents the longest sustained deep space communications relationship the DSN has maintained with any single spacecraft. The network’s ability to support a mission of that duration, through four different presidential administrations, multiple NASA budget cycles, and the normal attrition of ground equipment, speaks to the institutional depth of the DSN’s operational infrastructure.

New Horizons and the Edge of the Known Solar System

When New Horizons flew past Pluto on July 14, 2015, it was 4.77 billion kilometers from Earth. The spacecraft’s signal, traveling at the speed of light, took more than four and a half hours to reach the DSN’s antennas. At that distance, the data rate was agonizingly slow — roughly 1 kilobit per second at its best, under ideal conditions.

The DSN allocated its largest antennas exclusively to New Horizons during the flyby’s most critical phases. The 70-meter dishes at Goldstone, Madrid, and Canberra all participated in receiving the mission’s closest-approach data. Because New Horizons was traveling so fast and was so far away, the geometry of the flyby was precisely calculated to maximize the data volume returned in the limited window when the spacecraft was aimed at its best science targets. The DSN’s role in that calculation wasn’t passive — the network’s tracking precision informed the navigation team’s understanding of exactly where the spacecraft was.

The complete data set from the Pluto flyby took more than 16 months to download. The final images, some showing geological features on Pluto’s far side, arrived in October 2016. The DSN maintained its patient, continuous link throughout that entire download period.

In January 2019, New Horizons flew past Arrokoth, a contact binary Kuiper Belt object at a distance of 6.6 billion kilometers from Earth. That encounter set the record for the most distant object ever explored by a spacecraft. The DSN’s 70-meter antennas again served as the primary receiving stations, working through signal delays that stretched to over six hours each way.

The Mars Science Laboratory and Seven Minutes of Terror

On August 6, 2012, the Curiosity rover landed inside Gale Crater on Mars using an elaborate landing system that included a heat shield, a parachute, retrorockets, and a “sky crane” that lowered the rover on cables. The sequence of events, from atmospheric entry to touchdown, lasted about seven minutes — a period that became famous as the “seven minutes of terror” because the signal delay from Mars meant that by the time NASA received confirmation of atmospheric entry, the landing had already either succeeded or failed.

The DSN played a quiet but essential role in the events leading up to and following that landing. The Mars Odyssey spacecraft, operating in Mars orbit, served as a relay for Curiosity’s entry and descent telemetry, and Odyssey’s own signal reached Earth through the DSN. NASA’s Mars Reconnaissance Orbiter also captured radar data from Curiosity during its descent, again relaying through the DSN.

In the years following landing, Curiosity’s primary science data has flowed through the DSN — sometimes directly from the rover to Earth, but more typically relayed through orbital assets. The network has supported Curiosity continuously for more than a decade.

Juno and Jupiter’s Magnetic Field

Juno entered Jupiter’s orbit on July 4, 2016, after a five-year cruise from Earth. Its polar orbit carries it in close to Jupiter and then far out to about 8 million kilometers at apojove — a sweeping ellipse designed to minimize the spacecraft’s total radiation exposure while still allowing close science passes. The DSN communicates with Juno primarily during the quieter far segments of each orbit, when the spacecraft has the time and attitude geometry to point its antenna toward Earth.

Juno’s data has refined understanding of Jupiter’s internal structure, magnetic field, and atmospheric dynamics in ways that ground-based observations couldn’t achieve. The DSN’s precision tracking of Juno’s position and velocity has also contributed to gravitational science, since tiny deflections of the spacecraft’s trajectory caused by variations in Jupiter’s gravity field can be detected in the Doppler shift of the radio signal received at Earth.

That kind of gravitational science is one of the DSN’s less-publicized contributions. The network doesn’t just receive telemetry — it also measures spacecraft velocity to extraordinary precision by analyzing the frequency of the received signal. A spacecraft moving toward or away from an antenna shifts the frequency of its radio transmission by an amount proportional to its radial velocity. By measuring that shift, DSN engineers can determine the spacecraft’s velocity to within fractions of a millimeter per second, data that feeds directly into navigation and gravitational science analysis.

Technical Architecture

The DSN’s antennas don’t just passively collect radio signals. Each antenna is part of a complex system that includes low-noise amplifiers operating at near-absolute-zero temperatures, receiver chains that can handle multiple frequency bands simultaneously, and signal processors that connect to a global data network for routing information back to JPL’s mission control facilities.

The network primarily operates at S-band (around 2 GHz), X-band (around 8 GHz), and Ka-band (around 32 GHz). Each frequency band has different characteristics. X-band has been the workhorse for deep space communications for decades — it offers a good balance of antenna gain, atmospheric penetration, and spacecraft hardware simplicity. Ka-band offers significantly higher data rates but is more sensitive to atmospheric water vapor, which can absorb and scatter the signal. S-band is used for some spacecraft and for certain science applications.

The antennas at each complex range in size from 26 meters to 70 meters in diameter. The 70-meter dishes — one at each complex — are the network’s heavy-lifters, used for the most distant spacecraft and for missions requiring maximum receiving sensitivity. Smaller 34-meter antennas handle the bulk of day-to-day operations for closer missions, including those in Mars orbit or the asteroid belt.

The DSN’s Role in Navigation

Mission navigation is one of the DSN’s core functions, and it’s arguably the one that mission teams depend on most immediately during critical operations. Getting a spacecraft to a precise destination — threading a 400-kilometer gap between Saturn and its innermost ring, landing within a few kilometers of a target site on Mars — requires knowing where the spacecraft is with exceptional accuracy.

The DSN achieves this through a combination of techniques. Doppler tracking, as mentioned earlier, provides radial velocity information. Range measurements, in which a signal is transmitted to the spacecraft and timed on its return, provide distance. Very Long Baseline Interferometry, or VLBI, compares the phase of the spacecraft’s signal as received at two or more widely separated antennas, providing angular position information that’s independent of ranging data.

VLBI between DSN antennas on different continents can resolve a spacecraft’s angular position to better than a millionth of an arc-second. At the distance of Saturn, that precision corresponds to a position accuracy of a few kilometers. At interstellar distances — the kind relevant to the Voyager spacecraft, now in the interstellar medium — the position accuracy is still a few hundred kilometers, which is remarkable given that Voyager 1 is now more than 23 billion kilometers away.

Competing Demands and Scheduling Challenges

The DSN’s antenna time is a finite resource, and the competition for it has grown steadily as NASA’s active mission count has increased. In the early 1970s, the network might be tracking a handful of spacecraft at any given time. By the mid-2020s, the DSN routinely supports dozens of active spacecraft simultaneously, ranging from Mars orbiters and rovers to spacecraft in the outer solar system to Earth-science missions that use DSN antennas for certain data downlinks.

The scheduling of DSN antenna time is managed through a complex process that balances mission priorities, spacecraft operational requirements, and antenna maintenance windows. High-priority events — planetary flybys, orbital insertions, landing operations — get scheduled first, often months in advance, with secondary missions filling the gaps. The process is managed at JPL by the DSN scheduling office, which coordinates with mission operations teams around the world.

The growing demand has created real pressure on the network. Some mission teams have reported that DSN antenna allocations have become a limiting factor on their science data return, not the spacecraft’s communication hardware or the antenna’s raw receiving capability. There simply aren’t enough antenna hours to give every spacecraft as much time as its science team would prefer.

The Antenna Reliability Problem

A fact that rarely makes headlines but causes persistent concern within the deep space community: the DSN’s largest antennas are aging. The 70-meter dish at Goldstone, for instance, is the same dish that was constructed in the 1960s and extended to 70 meters in the 1980s. Its drive systems, structural elements, and feed assemblies have been maintained and replaced over the decades, but the basic infrastructure dates to a different era of engineering.

Maintenance windows for the 70-meter antennas are carefully coordinated with mission schedules to avoid leaving critical missions without coverage. But the antennas do require periodic downtime, and the maintenance requirements increase as the hardware ages. JPL has invested in refurbishment programs for these antennas, but the underlying question of when an aging dish needs to be replaced rather than repaired has no easy answer, particularly when the cost of a new 70-meter antenna would run into the hundreds of millions of dollars.

Building New 34-Meter Antennas

Rather than attempting to build additional 70-meter dishes, NASA’s strategy for expanding DSN capacity has focused on adding more 34-meter antennas and improving the network’s arraying capability. A new 34-meter antenna, designated DSS-53, was added to the Madrid complex in 2021, and similar additions have been planned or constructed at the other complexes.

The logic behind 34-meter expansion over 70-meter construction is partly economic and partly strategic. A 34-meter antenna costs significantly less to build and maintain than a 70-meter dish. Multiple 34-meter antennas arrayed together can approach or match the sensitivity of a single 70-meter dish while providing operational flexibility — individual antennas can be taken offline for maintenance without eliminating the complex’s capability entirely.

Arraying also allows the network to serve multiple spacecraft simultaneously. Two 34-meter dishes, arrayed electronically, can support one spacecraft while a third 34-meter antenna handles a separate mission. A single 70-meter dish is a monolithic resource that can only be pointed at one spacecraft at a time.

The International Partnership Model

The DSN has maintained its international partnership structure since the 1960s, and those partnerships have deepened over time. Australia’s Commonwealth Scientific and Industrial Research Organisation operates the Canberra complex under a cooperative agreement that’s been renewed repeatedly. Spain’s INTA has done the same for the Madrid complex.

Beyond the three-complex partnership, the DSN has developed data-sharing and cross-support arrangements with other space agencies. The European Space Agency, or ESA, operates its own deep space antenna network, the European Space Tracking network known as ESTRACK, which has antennas in Australia, Spain, Argentina, and French Guiana. When DSN antennas are unavailable or insufficient, ESA’s antennas have sometimes filled in for NASA missions, and vice versa.

The Japan Aerospace Exploration Agency, or JAXA, and the Indian Space Research Organisation, or ISRO, have also developed their own deep space tracking capabilities. The growing number of national deep space programs has created both competition for radio frequency spectrum and opportunities for coordination. The frequency bands used by deep space missions are managed internationally through the International Telecommunication Union, and DSN operations must stay within allocated frequency ranges agreed upon through that process.

The Interstellar Era

Voyager 1 crossed what scientists define as the heliopause — the boundary between the solar wind’s influence and interstellar space — in August 2012, becoming the first human-made object to reach interstellar space. Voyager 2 followed in November 2018. Both spacecraft are still operating and still in contact with the DSN, though the communications challenges involved are now at the extreme edge of what the network can manage.

Voyager 1’s signal, at a transmitter power of roughly 22 watts, arrives at Earth more than 23 billion kilometers away. By the time it reaches a DSN antenna, its power has been attenuated to roughly 10^-16 watts — a value so small it’s difficult to intuitively grasp. To detect it, the DSN must use its largest antennas, its most sensitive receivers, and long integration times. A small amount of noise or interference can overwhelm the signal entirely.

In 2023, a software patch was transmitted to Voyager 1 after engineers at JPL discovered that one of its four flight data system computers was sending garbled telemetry. The round-trip communication time was about 45 hours. Engineers on Earth sent commands, waited 22.5 hours for them to arrive, and then waited another 22.5 hours for the spacecraft’s response. The DSN’s antennas maintained the link throughout this patient diagnostic process, and the problem was eventually resolved through a sequence of command transmissions that shifted operations to a different memory chip on the spacecraft.

It’s genuinely uncertain how much longer the Voyager spacecraft will remain communicable. Their radioisotope thermoelectric generators are losing power at a slow but inevitable rate, and at some point in the late 2020s or early 2030s, one or both spacecraft will no longer have enough power to transmit a detectable signal. When that happens, the DSN’s link to the first human objects in interstellar space will go silent.

Preparing for Artemis

NASA’s Artemis program, which aims to return humans to the Moon and eventually establish a sustained lunar presence, has created new communications requirements that the DSN is adapting to address. Crewed missions near the Moon don’t quite fit the DSN’s traditional deep space profile — the Moon is close enough that the network’s largest dishes are actually overkill for most purposes — but the specific demands of crewed operations, including high data rates, low latency requirements, and the need for continuous coverage, have required new planning.

NASA’s Lunar Gateway, a planned small space station in lunar orbit, will serve as a waypoint for Artemis missions. Communications between Gateway and Earth will use a mix of DSN assets and a new relay system called the Near Space Network. The DSN will handle the deep space segment of Gateway communications, particularly when Gateway is on the far side of the Moon or at orbital positions where direct Earth links require the network’s higher sensitivity.

The Orion spacecraft, which carried astronaut mannequins on the Artemis I test flight in late 2022, used DSN antennas for portions of its mission. The spacecraft traveled beyond the Moon to a distance of about 430,000 kilometers from Earth — farther than any crewed spacecraft had traveled since Apollo 17 in 1972, though Artemis I carried no crew. The DSN provided high-rate telemetry during key mission phases.

Mars Sample Return and the Communications Challenge

One of the most complex future scenarios for the DSN involves NASA’s Mars Sample Return campaign, which is planned to bring samples collected by the Perseverance rover back to Earth. The campaign involves multiple spacecraft, including a Mars ascent vehicle to launch samples from the Martian surface, an orbiting spacecraft to capture the samples, and an Earth return vehicle.

The DSN will need to support all of these spacecraft simultaneously during the mission’s most critical phases. The coordination challenge is substantial: multiple vehicles operating near Mars, each requiring tracking and command capability, with operations happening in a sequence that doesn’t allow for replanning delays. The sample return campaign has faced budget and schedule pressures, and its exact architecture was still being reconsidered as of early 2026.

Software-Defined Radio and Modernization

One of the most significant recent shifts in DSN technology involves the increasing use of software-defined radio, in which functions that were previously implemented in specialized hardware are instead performed by software running on general-purpose computers. Software-defined radio offers flexibility that hardware solutions can’t match: a software update can change how an antenna processes signals, add support for a new frequency band, or improve coding algorithms, without requiring physical modifications to the antenna or its associated electronics.

JPL has been implementing software-defined receiver systems across the DSN complexes as part of a broader modernization effort. The transition is gradual — replacing operational systems while keeping the network continuously available is genuinely difficult, and every change carries risk when the antennas are actively tracking spacecraft that can’t wait for a maintenance window. But the long-term benefits of a more flexible, software-driven infrastructure are substantial, particularly as new missions adopt more sophisticated waveforms and modulation schemes that older hardware receivers weren’t designed to handle.

Optical Communications

Radio waves have been the medium of choice for spacecraft communications since the beginning of the space age, but they’re not the only option. Laser communications, also called optical communications, use visible or near-infrared light rather than radio waves to transmit data. Because optical wavelengths are much shorter than radio wavelengths, a laser beam can carry far more information per second for a given transmitter power.

NASA has been developing optical communications for deep space applications through programs including the Deep Space Optical Communications experiment, or DSOC, which flew aboard the Psyche spacecraft launched in October 2023. DSOC achieved first light in November 2023, successfully transmitting data via laser from a distance of about 16 million kilometers — a record for optical space communications at the time.

Optical communications won’t replace radio communications in the near term, but they represent a future expansion path for the DSN. The receiving technology is different from radio: instead of dish antennas connected to radio receivers, optical communications require large telescopes connected to photon-counting detectors. NASA has been building ground infrastructure for optical communications at existing DSN sites, and a dedicated Optical Ground Station is planned for Table Mountain in California.

There’s an important caveat worth noting about optical communications: they’re vulnerable to weather in a way that radio communications aren’t. A thick cloud layer blocks laser light completely, while a radio signal from Mars passes through clouds without significant attenuation. This means that optical communications will likely require networks of ground receiving stations in diverse geographic locations, ensuring that at least one station has clear sky when a spacecraft needs to transmit.

Perseverance and Ingenuity

The Perseverance rover, which landed in Jezero Crater in February 2021, has been collecting rock and soil samples in sealed tubes for eventual return to Earth. Its communications architecture relies heavily on orbital relays — primarily Mars Reconnaissance Orbiter and the Mars Atmosphere and Volatile Evolution spacecraft, or MAVEN — with those orbiters downlinking data through the DSN.

The Ingenuity helicopter, which became the first powered aircraft to fly on another planet in April 2021, communicated with Earth through Perseverance as a relay. The helicopter wasn’t a DSN asset in a direct sense, but every byte of its telemetry and flight data passed through the DSN’s antennas as part of the relay chain.

Ingenuity’s flights ended in January 2024 after damage to one of its rotor blades during its 72nd flight. Its entire operational history, including 128 minutes of total flight time over Mars, was documented through data that flowed back to Earth through this communication chain, with the DSN as its final terrestrial link.

The DSN Now

As of early 2026, the DSN is simultaneously tracking dozens of active spacecraft. The list includes Voyager 1 and Voyager 2 at interstellar distances; New Horizons in the Kuiper Belt; Cassini’s successor missions and orbiters throughout the solar system; multiple Mars orbiters and surface assets; and spacecraft on their way to Jupiter’s moons, asteroids, and beyond.

NASA’s Europa Clipper, launched in October 2024, is en route to Jupiter on a trajectory that will eventually carry it into a series of close flybys of the moon Europa. The DSN will serve as its primary communications link for the duration of the mission, which is expected to last through the early 2030s at Jupiter.

NASA’s DAVINCI mission to Venus is in development, targeting a launch in the late 2020s. The DSN will support it as it does all deep space missions to Venus, Mars, and beyond.

The agency’s push toward commercial partnerships has also changed the DSN’s operational context. Companies including SpaceX have developed their own tracking and communications infrastructure for Starship and commercial satellite operations, but NASA’s scientific and exploration missions continue to rely on the DSN for deep space communications. The commercial sector hasn’t yet built anything that replicates what the 70-meter dish at Goldstone can do for a spacecraft at Jupiter.

What Gets Forgotten

The history of the DSN is told most easily through the dramatic mission moments — Voyager’s Neptune flyby, Curiosity’s landing, New Horizons at Pluto. But most of what the DSN does is quiet, continuous, and invisible: daily scheduling of hundreds of antenna passes, routine telemetry downlinks from Mars orbiters, navigation tracking updates sent to spacecraft navigators, command uplinks confirming that a sequence of instructions has been received correctly.

This steady work is what keeps missions alive. Spacecraft in deep space need regular contact to confirm their health status, upload revised command sequences, and download the science data that accumulates in their onboard storage. A mission that misses too many contact opportunities can fall into a communications blackout from which recovery may be impossible. The DSN’s day-to-day reliability is what prevents those blackouts.

The network’s operational staff — engineers and technicians at Goldstone, Madrid, and Canberra, plus the scheduling and operations teams at JPL — represent a community of expertise that has accumulated over six decades. That expertise isn’t easily documented or transmitted. Knowing how to optimize antenna pointing during a spacecraft emergency, or how to work around a receiver hardware glitch while maintaining contact with a distant probe, is knowledge that lives in people as much as in procedures.

Summary

The Deep Space Network was built on a simple insight that turned out to have enormous depth: sending spacecraft into deep space was only useful if you could communicate with them. That insight has driven six decades of antenna construction, receiver development, software innovation, and international cooperation, producing an infrastructure that has made modern planetary science possible.

What’s striking, looking at that history whole, is how often the DSN’s capabilities have defined the outer boundary of what missions could do. The data rate from Voyager at Neptune was set by the DSN’s ability to receive, not by the spacecraft’s transmitter. The timeline for the New Horizons Pluto data download was set by DSN antenna time, not by any limitation on the spacecraft. Missions have consistently pushed against the network’s limits, and the network has consistently been pushed to expand.

The next few decades will add optical communications to the network’s toolkit, add new antennas to its three complexes, and likely add new challenges that nobody can fully anticipate right now. What won’t change, almost certainly, is the basic architecture: antennas on three continents, pointed at spacecraft billions of kilometers away, patiently collecting signals measured in fractions of a watt and turning them into human knowledge about the solar system. That’s a remarkable thing to have built, and it’s still being built.

Appendix: Top 10 Questions Answered in This Article

When was the Deep Space Network officially established?

NASA formally established the Deep Space Network on December 24, 1963, as a separately managed entity under the authority of the Jet Propulsion Laboratory. The infrastructure and operational concepts that defined it had been developing for years prior to that official date.

Where are the three Deep Space Network antenna complexes located?

The three DSN complexes are located at Goldstone in California’s Mojave Desert, near Robledo de Chavela outside Madrid, Spain, and near Canberra, Australia at a site called Tidbinbilla. Their geographic separation of roughly 120 degrees of longitude allows continuous coverage of spacecraft anywhere in the sky.

What is the largest antenna in the Deep Space Network?

The DSN operates 70-meter diameter dish antennas at each of its three complexes. These are the network’s most sensitive receiving instruments and are used for the most distant spacecraft, including Voyager 1 and Voyager 2, which are now in interstellar space.

How does the Deep Space Network detect signals from billions of kilometers away?

The network uses very large dish antennas combined with cryogenically cooled low-noise amplifiers that operate near absolute zero temperature to maximize sensitivity. Long signal integration times and antenna arraying, which electronically combines multiple dishes to simulate a larger aperture, allow detection of signals measuring a tiny fraction of a watt.

What is antenna arraying and why does the DSN use it?

Antenna arraying connects multiple dish antennas electronically so their signals combine, simulating the receiving capability of a larger single dish. The DSN uses arraying to boost sensitivity for distant missions and to serve multiple spacecraft simultaneously, since individual antennas within an array can track different targets.

How did the DSN support the Voyager spacecraft’s encounters with Uranus and Neptune?

For Voyager 2’s Uranus flyby in 1986, the DSN arrayed its Goldstone antenna with the Very Large Array in New Mexico and antennas at Japan’s Usuda Deep Space Center. For the Neptune encounter in 1989, the DSN’s 64-meter dishes were upgraded to 70 meters in time for the flyby, and all three complexes participated in arrayed receiving configurations.

What is the role of the Deep Space Network in spacecraft navigation?

The DSN provides three types of navigation data: Doppler tracking, which measures spacecraft velocity through radio frequency shifts; range measurements, which time signal round trips to determine distance; and Very Long Baseline Interferometry, which compares signal phases at widely separated antennas to determine angular position. This navigation data enables precise trajectory corrections and mission planning.

What is the Deep Space Optical Communications experiment?

DSOC is a laser communications technology demonstration that flew on the Psyche spacecraft launched in October 2023. It achieved first light in November 2023 at a distance of about 16 million kilometers, demonstrating that optical wavelengths can transmit data at much higher rates than radio for a given transmitter power.

How does the Deep Space Network handle increasing demand from more missions?

The DSN manages growing demand through a combination of new antenna construction, expanded arraying capability, and complex scheduling processes that prioritize critical mission events. NASA has added new 34-meter antennas at existing complexes rather than building additional 70-meter dishes, since multiple smaller antennas can be arrayed while also providing operational flexibility.

How long will the Voyager spacecraft remain in contact with the Deep Space Network?

The Voyager spacecraft’s radioisotope thermoelectric generators are losing power at a slow, steady rate, and communications are expected to become impossible sometime in the late 2020s or early 2030s when transmitter power falls below the DSN’s detection threshold. The exact timeline depends on which instruments are turned off to conserve power and how the spacecraft’s systems continue to perform.