JPL and Argotec’s relay satellites could deliver bandwidth for more than 90 missions
Earth’s moon is the target of more missions than at any time since the Apollo era, by both space agencies and commercial entities. NASA, for example, has plans to visit the moon using both robots and humans, and is also considering—with international collaborators—a small orbiting outpost in the next decade. This facility, known as the Lunar Gateway, would store supplies, host visiting astronauts, and facilitate communication between the moon and Earth.
Although the gateway is perhaps the most ambitious of the projects planned, it’s only one of over 90 lunar missions being considered for the years between now and 2030. Of course, not all these planned missions will materialize, but many—if not most—will happen in some form. And this is only the beginning: We anticipate that interest in the moon will accelerate, eventually culminating in a permanent human presence on the surface.
The proposed Lunar Crater Radio Telescope would turn a crater on the far side of the moon into a massive dish-shaped antenna to survey the universe, accumulating massive amounts of data that need to be sent back to Earth for analysis.NASA/JPL
If that comes to pass, lunar denizens will need to stay in touch with Earth. While direct radio communication with Earth was used during the Apollo missions, it doesn’t work in every possible situation. For example, the moon’s far side, as well as large portions of its poles, have no direct line of sight to Earth. Even on the side facing Earth, hills and crater walls can block communications.
And on the practical side, direct communication across hundreds of thousands of kilometers of space requires a powerful communications terminal with a large antenna or a high-wattage amplifier, if not both. Small robots, for example, will not have the space or the power for these large systems. A better solution to lunar connectivity is a network of relay spacecraft orbiting the moon to provide continuous coverage everywhere.
Italian aerospace company Argotec and NASA’s Jet Propulsion Laboratory (JPL) are collaborating on the concept of an orbiting relay satellite constellation called Andromeda. Argotec (at which Balossino is head of the R&D unit) is developing spacecraft concepts and JPL (at which Davarian is a project manager) is providing subsystems such as radios and antennas. The approach consists of 24 relay satellites to be placed in a constellation using 4 orbits, with 6 satellites per orbit. This configuration would provide continuous coverage to the poles, and near-continuous coverage everywhere else, with only occasional slight gaps. With this relay system, missions anywhere on the lunar surface would have reliable, consistent connections to Earth.
Placing relay satellites in orbit around the moon comes with challenges. First, we would like to use orbits that are stable—meaning satellites would require little or no maneuvering. Second, orbits need to be selected with continuous or near-continuous physical line of sight to “hot spots” that will likely have considerable human or robotic activity. And third, while guaranteeing high visibility for lunar hot spots, we don’t want to deny connectivity to any other portions of the surface as a result.
Any relay-satellite network needs to provide the best possible service and coverage with the minimum number of satellites.
The moon’s south pole is one probable hot spot because its craters contain ice, at least to some extent. For longer crewed missions, the water that humans require would likely be easier to harvest from the moon rather than to haul it from Earth. Water can also, through electrolysis, provide hydrogen fuel for rockets. Another potential hot spot is the moon’s far-side equatorial region, where massive radio telescopes could one day be sited.
In addition to communications, the astronauts, rovers, and scientific instruments all need to know where they are on the moon’s surface. Relay satellites can form a sort of “lunar GPS” for navigation by timing how long it takes for signals between multiple satellites to reach a given point on the surface. In general, the more relay satellites in more orbits, the better. The trade-off is that launching and operating each additional satellite costs money. Therefore, any relay-satellite network needs to provide the best possible service and coverage with the minimum number of satellites.
Argotec’s relay network concept uses a class of stable orbits known as frozen orbits. Stable orbits make it easy to keep the satellites in their assigned orbits for the 5 years (or more) that they are expected to operate. The proposed orbits are elliptical, with a 12-hour period, a 57-degree inclination, and a distance to the moon’s surface from 720 kilometers at their closest points to 8,090 km at their farthest.
Any satellite will travel slowest at the farthest point of its orbit—called the apoapsis—and fastest when it is closest to the moon. Therefore, we want any orbit to have its apoapsis approximately above a potential hot spot in order to provide long periods of communications. With the selected orbits, the lunar poles are covered by three satellites simultaneously 94 percent of the time, with at least one satellite overhead at any given time. The equator, meanwhile, has at least one satellite overhead 89 percent of the time, and simultaneous coverage by three satellites 79 percent of the time.
The proposed FARSIDE telescope would use unspooled antennas across an area of the moon’s surface 10 kilometers in diameter to create a large interferometric array.NASA/JPL
Even at the apoapsis, a relay satellite is fewer than 10,000 km from the surface. Compare that to the distance from the Earth to the moon, which is about 400,000 km. Even for users positioned within direct line of sight with Earth, an overhead relay satellite reduces the communication link distance by about a factor of 40.
A shorter communication distance means a person or robot on the surface does not need a powerful terminal to maintain a low-data-rate link with Earth. Instead, they can employ the relay satellites to bounce their signals to Earth using a small communications terminal.
Relay satellites also mean that humans at two different locations on the surface can talk to each other without noticeable delay. Without relay satellites, a call would have to travel to Earth and back, taking about 3 seconds round trip. Imagine the difficulty of a phone call with a 3-second delay, and you’ll quickly realize how important relay satellites are for voice or video communications on the surface.
Even for users positioned within direct line of sight with Earth, an overhead relay satellite reduces the communication link distance by about a factor of 40.
Different missions will have different communication needs. Simple text or voice communications require only a few kilobits per second, while high-definition video and radio telescopes need megabits per second. And given the number of proposed lunar missions, any relay satellite will likely need to juggle multiple simultaneous communications. For lower bandwidth applications like text and voice, one satellite will be able to collect and aggregate the many data streams for relay elsewhere. On the other hand, an individual satellite is likely to reach its capacity with the high data production of a single radio telescope.
NASA is currently studying two radio-telescope options that could be deployed on the moon’s far side. The first is the Lunar Crater Radio Telescope (LCRT), an ultralong-wavelength radio telescope proposed by JPL engineers. The LCRT would observe the universe at frequencies below 30 megahertz, which are otherwise blocked by the Earth’s ionosphere. Robots would deploy a wire mesh 1 km in diameter in the middle of a 4-km crater to create a reflector radio telescope. It would be the largest dish-shaped radio telescope in our solar system.
The second proposed telescope is the Farside Array for Radio Science Investigations of the Dark ages and Exoplanets. FARSIDE would be a low radio frequency interferometric array—meaning it would observe distant stars and other radio sources with multiple antennas. By correlating these multiple observations, it can image the source at high resolution and accurately determine its position. The system would use 128 dual-polarization antennas deployed across a roughly circular area 10 km in diameter, and tethered to a base station for central processing and power. The base station would also transmit collected data to a relay orbiter (such as our proposed Andromeda constellation).
The software-defined Universal Space Transponder radio is the foundation of a lighter and smaller radio called the UST-Lite that JPL is currently testing for use in future spacecraft.NASA/JPL
FARSIDE would be able to image the entire sky each minute, spanning frequencies from 100 kilohertz to 40 MHz. Like the LCRT, this would extend into bands below those accessible to Earth-based radio astronomy—in the case of FARSIDE, by two orders of magnitude. Both proposed telescopes would generate massive volumes of data that need to be transmitted to Earth.
After a relay satellite receives data from a far-side radio telescope or anything else on the lunar surface, it will need to send that data onward to Earth. On Earth, large antennas will need to have adequate gain and sensitivity to support a link up to at least 100 megabits per second. Ideally, each (expensive) ground antenna should be able to receive signals from multiple relay satellites at a time to reduce the number that need to be built.
NASA’s Deep Space Network (DSN) is a good example of the type of ground network needed. The DSN has three antenna complexes across the world—in California, Australia, and Spain—with several large, highly sensitive antennas at each site. However, the DSN is designed to support deep-space missions well beyond the moon, and so using it for a lunar relay system may be overkill. Besides, the DSN is already in high demand by many missions, both current and planned. So while it may be a good initial choice, over the longer term, leasing or building commercial ground stations would be cheaper and more effective.
A lunar relay spacecraft needs to be only 50 or 60 kilograms, which is small by satellite standards. We have developed a satellite concept that is 44 by 40 by 37 centimeters when the solar arrays and antennas are stowed, with a mass (including propellent) of 55 kg. It carries a four-channel radio developed at JPL, with two channels each operating in the K-band (at about 26 gigahertz) and S-band (at about 2 GHz). One K-band channel provides connectivity to Earth (100 Mb/s for satellite-to-Earth and 30 Mb/s for Earth-to-satellite). The other three channels provide connectivity to the moon. The S-band channels offer 256 kb/s connections to the lunar surface, and 64 kb/s from the surface to the satellite. The remaining K-band channel is a 100 Mb/s satellite-to-moon link and 16 Mb/s moon-to-satellite link.
Imagine the difficulty of a phone call with a 3-second delay, and you’ll quickly realize how important relay satellites are for voice or video communications on the surface.
Our proposed satellites would use the K-band for Earth-to-satellite connections for two reasons. First, there is more available bandwidth in the K-band than other bands used for space communications. Second, for antennas of the same size, K-band frequencies have higher antenna gain. In other words, K-band antennas more efficiently convert received signals into electrical power. The downside of using the K-band is its weather sensitivity—rain, for example, will easily attenuate the link. The relay satellites would require an additional power margin to ensure the link remains stable.
The current relay-satellite design has three antennas: A steerable, 50 cm K-band antenna for Earth-to-satellite communications; a fixed K-band “ metasurface” antenna that has a low profile with low mass, can easily be manufactured at low cost, and can tolerate the harsh environment of outer space; and a fixed S-band antenna array. We’re also considering a small antenna in the X-band (at about 7 GHz) for Earth-to-satellite communications for additional reliability and redundancy. The X-band is a good choice here because it is less susceptible to attenuation from rain than the K-band, albeit at a lower data rate.
Currently, we are finalizing the design of the spacecraft. We intend to use commercially available hardware wherever possible to lower costs. However, we still need a few new technologies to be refined to provide the desired satellite performance while still meeting requirements for mass and power. The metasurface antenna, which can be 3D printed, is a new technology developed at JPL for small-satellite applications. The transmit-only version is operational, with a measured gain exceeding 32 decibels isotropic (dBi) for a 20-cm antenna at 32 GHz. We expect a recent improvement to the design to increase the gain to 34 dBi. We’re also working on dual-frequency capability, so that the antenna will be able to simultaneously transmit and receive signals.
The three antenna complexes making up the Deep Space Network, such as the one in Canberra, Australia that includes the antenna shown here, maintain contact with spacecraft across the solar system. The Andromeda constellation would need a similar setup to bring back data from the moon.NASA/JPL
Additionally, we’d like to use a smaller and lightweight version of the software-defined Universal Space Transponder (UST) radio called UST-Lite. JPL has completed an initial thermal-testing campaign for a UST-Lite prototype, to ensure that the radio’s generated heat can be dissipated without affecting performance. We performed additional tests to better characterize the prototype’s receiver thresholds, bit error rates, transmit waveforms, and more. We continue to optimize the receiver’s parameters, as well as to develop new modules to cover K-band frequencies (We have already developed S- and X-band modules).
We’re also addressing the network’s software needs. For example, there is no current protocol standard for communications between a relay satellite and a lunar user at the S- and K-bands. We, therefore, have begun to work with the Consultative Committee for Space Data Systems to introduce such a standard.
One way to think about the goal of any lunar-communications apparatus is that it would create 5G-like capabilities for the entire moon. This would mean taking advantage of 5G technologies wherever possible, such as installing cell sites on the moon to supplement the relay arrangement. This approach would connect many additional kinds of devices to a lunar network—for example, networks of low-power Internet of Things sensors and autonomous vehicles.
Our proposed relay network would only be a first step. In a more distant future, humans on the moon should be able to send and receive texts, make phone calls, and stream data at will. Similarly, robots and sensors should be wirelessly connected just like IoT devices are on Earth. Robots would be controlled remotely, and sensors would automatically upload their measured data.
However, this vision of lunar connectivity may take generations of lunar-communication networks to emerge. Nevertheless, we believe we can look forward to a time when there will be human colonies on the moon engaged in scientific, technical, and commercial activities in a robust wireless environment.
This article appears in the April 2022 print issue as “The Moon Needs Decent Wireless Coverage.”
- 4G on the Moon: One Small Leap, One Giant Step - IEEE Spectrum ›
- NASA's Lunar Space Station Is a Great/Terrible Idea - IEEE Spectrum ›
- Lunar Pioneers Will Use Lasers to Phone Home - IEEE Spectrum ›
- Meet the Lunar Gateway’s Robot Caretakers - IEEE Spectrum ›
- Companies Vie to Build NASA’s Next Communications Network - IEEE Spectrum ›