Radio observatories like the Very Large Array have to increasingly contend with interference from satellites in additional to terrestrial sources. (credit: Adam Kimbrough)

Satellite constellations and radio astronomy

by Adam Kimbrough
Monday, April 29, 2019

In the San Augustine Plains of central New Mexico, 27 radio telescopes stand tall, operating nearly 24 hours a day, seven days a week, capturing extremely weak signals emitted from all over the universe. This flat and vast land, once a seabed, sits at an altitude of more than 2,100 meters and is surrounded by 360 degrees of mountains. Despite the ideal conditions of this location, listening to these faint radio emissions is becoming increasingly difficult as the Earth becomes “noisier” in the same direction in which these dish antennas are pointed, the sky.

At ground level, the National Radio Astronomy Observatory’s (NRAO) Very Large Array keeps a continuous live monitoring station dedicated for receiving Wi-Fi and Bluetooth radio transmissions for visitors’ cellphones and other electronic devices. Besides their inherent interference, these signals can also emit multiples of their intended frequency, causing interference in many places at once due to poor engineering. (Known as harmonics, these are signals that are multiples of a desired frequency that are expected to be minimized by the design engineer.) When “radio noisy” visitors drive into the VLA area, they are spotted on the monitor. The visitor center receives an email alert to remind the visitors to turn off their cell phones. If the visitors fail to respect the radiofrequency-quiet rules after being asked to turn off their cellphones, a portable spectrum analyzer and feed horn antenna can be used to find the direction of the sources. The consequences of such RFI (radio frequency interference) may completely destroy a scientist’s observation data, or potentially mislead with invalid data points; that’s $5,000 an hour of American taxpayer funds!

The radio spectrum consists of frequencies 3 kilohertz to 300 gigahertz, and this large area is regulated by an international body known as the International Telecommunication Union (ITU). Frequency bands are allocated accordingly to the different services: radar, amateur radio, television, telephone, etc. What distinguishes radio astronomers from everyone else who uses the radio spectrum is that radio astronomers are passive users. While others are transmitting on their assigned frequencies, using the spectrum actively, radio astronomers are listening wherever the deep space emissions may lie, such as the microwave remanence of the Big Bang.

Today, the radio environment is becoming saturated with signals, and companies are in a constant battle to purchase a piece of this precious spectrum. While the ITU does help protect small slivers of radio spectrum for radio astronomers, sharing portions of the spectrum is not an option as other signals can easily overwhelm the receiver electronics of the scientists. This incredible sensitivity is an unfortunate truth that must compete alongside commercial interests, like 5G wireless services.

While mitigating terrestrial interference (cellphones, mobile wireless networks, vehicles, etc.) for radio astronomy telescopes is a full-time job in and of itself, orbiting transmitters are significantly more dangerous. Satellites in the path of a highly sensitive radio telescope have been a concern as early as 1982 with the launch of Russian (previously Soviet) GLONASS satellites. The problem with GLONASS was its main transmit frequency (1612 megahertz) overlapped with the spectral line of the hydroxyl (OH) radical. Radio astronomers observing at this spectral line were out of luck as soon as one of these satellites cleared the horizon in its orbit. The frequencies in use were eventually removed to accommodate astronomers, but a few decades later Motorola laid plans to launch an array of 66 Personal Communication System (PCS) satellites into low earth orbit with downlinks in the 1621.25 to 1626.5 megahertz band. This became the Iridium constellation that turned radio astronomers’ heads, and their radio telescopes away.

The original fleet of Iridium satellites in 1998 was in a non-geostationary orbit with the intention of providing seamless voice and data communication all over the world. It wasn’t that Iridium satellites’ downlink frequencies directly transmitted in radio astronomy bands, but rather the sidebands of the signals, which overlapped into observing territory, that caused the problem. Iridium failed to abide by an ITU-R Recommendation, with their downlink signals causing interference of up to 30 decibels (dB) above the levels deemed harmful. This means Iridium was transmitting signals 1,000 times stronger than they should have been. Being only a recommendation, it wasn’t mandatory for Iridium to comply with ITU-R Recommendation RA.769.

To cope with this, engineers at the Very Large Array in New Mexico spent much time and effort developing “Iridium filters” that blocked a portion of the band that received the interference. Still, according to one RF engineer at the VLA, radio astronomers did not seem to like using this feature very much. RFI removal exists in both an active and post-processing form, but it could only help things minimally.

Today, radio astronomy faces a new wave of enormous satellite constellations, among which are SpaceX’s Starlink, OneWeb, and Iridium NEXT. The SpaceX Starlink satellite constellation aims to eventually launch around 12,000 satellites provide space-based Internet services. The OneWeb constellation’s ultimate plan is to have almost 3,000 satellites in orbit to also provide Internet services, but starting with a constellation of about 650 satellites. Iridium NEXT, like the original constellation, is a telecommunications satellite constellation consisting of 66 satellites.

Of the three, Starlink obviously grabs the most attention and instills the most fear. Harvey Liszt, astronomer and spectrum manager for the NRAO, reached out to FCC Chairman Ajit Pai in February 2018 to express concern over SpaceX’s constellation:

SpaceX, which plans to use the 10.7 – 12.7 GHz band for its downlink, has not yet fulfilled its obligations under US131. Coordination between SpaceX and the AUI observatories (together with NSF) trailed off inconclusively around the middle of 2017 after a tentative and rather preliminary treatment of radio astronomy’s concerns and the manner in which SpaceX planned to address them.

Footnote US131 of FCC’s Title 47 states that, if transmitting in the 10.7–11.7 gigahertz band, coordination between non-geostationary satellite orbit licensees and certain radio astronomy observatories is necessary to “achieve a mutually acceptable agreement.” While post-observation RFI excision does exist in some forms, current technology does not yet allow for this when dealing with continuous RFI sources such as 5G and satellite downlinks. This could change in the future, but for now the lower frequency range of 1 to 20 gigahertz is in bad shape when it comes to collecting clean data.

Given this situation, observation frequency in radio astronomy has been trending higher and higher. The Very Large Array currently observes from 1 to 50 gigahertz. NRAO’s latest project, ngVLA (Next Generation VLA), proposing to begin collecting science in the late 2020s, would have a frequency range of 1.2 to 116 gigahertz! At the same time, space constellations are also making use of these higher frequencies, specifically with intersatellite links, or ISLs. Different from the downlink frequencies (spacecraft to ground), these ISL frequencies will operate in the V-band (40–75 gigahertz) and provide just as much harm as the downlinks. On November 19, 2018, the FCC granted SpaceX an authorization to transmit in V-band.

Five months after the NRAO’s letter to the FCC, Liszt sent another similar letter, this time regarding the OneWeb constellation. Liszt explains that awareness around the 10.7–11.7 gigahertz band is crucial because of radio astronomy’s passive band of 10.68–10.7 gigahertz. Passive bands prohibit any radio emissions in protection of radio astronomy observation, like spectral lines (water, OH radical, etc.) However, within these passive bands are many bands that merely urge spectrum users to “take all practicable dispositions to protect the radio astronomy service from harmful interference.” The 10.68–10.7 gigahertz band, however, is not one of these bands where leniency exists, so it is understandable that the NRAO wishes for a “robust demonstration that they [OneWeb] can, in the aggregate, fully protect radio astronomy from their unwanted emissions into the passive band.”

Liszt also points out an incompatibility of OneWeb’s proposed ground level power flux density with ITU-R Recommendation RA 769. Essentially, the detrimental threshold for interference to radio astronomy operating around these frequencies is some “58 dB smaller” (or quieter) than what OneWeb’s filings state. The ITU-R recommendations give active services detailed calculations on what maximum levels of data loss are acceptable for radio astronomer observers. While allowing up to five percent data loss is considered generous of radio astronomers, Liszt states, “unwanted emissions into a passive service band would be a terrible precedent.” Still, the Iridium NEXT constellation was developed without any prior consultation with radio astronomy agencies. Instead of discussing with radio astronomers, they chose to operate under the assumption that they will cease transmission “during periods of observation” as recommended. Iridium requests that astronomers provide them with a three-day notice to implement RFI reduction for the specified observatories.

When it comes to RFI for radio telescopes, an array will always be less affected than a single beam; the more antennas and the bigger the baseline, the less impact RFI will have. It won’t be surprising to see a continual development of arrayed radio telescope systems in the future. It seems almost equally as likely for these future radio telescopes to be in orbit around Earth to escape the terrestrial radio clutter or even on the far side of the Moon (like the low frequency research performed as a part of China’s Chang’e-4 mission.) Or, perhaps we will see more petitions for intentional Radio Quiet Zones like the one around Green Bank Observatory or, in the extreme case of China’s FAST radio telescope, mandatory relocation of local residents. Either way, with or without cooperation from commercial entities, electromagnetic spectrum managers have a lot of work on their hands. When considering the value of continuing of high-level science provided by radio astronomy, the world’s engineers must tackle the challenge of accommodating for the desire to carry out these large commercial orbital missions while also studying what’s beyond Earth.

Adam Kimbrough is an undergraduate electrical engineering student at Virginia Tech and a former six-month co-op student at the NRAO’s Very Large Array.

Note: we are temporarily moderating all comments subcommitted to deal with a surge in spam.

Home