Radio Astronomy with NASA’s Deep Space Network

Abstract

The Deep Space Network (DSN) is the spacecraft tracking and communication infrastructure for NASA’s deep space missions. At three sites, approximately equally separated in (terrestrial) longitude, there are multiple radio antennas outfitted with cryogenic microwave receiving systems both for receiving transmissions from deep space spacecraft and for conducting radio astronomical observations, particularly in the L band (1350 MHz–1800 MHz), X band (8200 MHz–8600 MHz), and K band (18 GHz–27 GHz). In particular, the 70 m antennas at the Canberra and Madrid DSN Complexes are well-equipped to participate in international very long baseline interferometry (VLBI) observations. Over the past five years, there has been an effort to refurbish and modernize equipment such as receiving and signal transport systems for radio astronomical observations. We summarize current capabilities, on-going refurbishment activities, and possible future opportunities.

1. Introduction

The Deep Space Network (DSN) is the spacecraft tracking and communication infrastructure for NASA’s deep space missions. It consists of multiple radio antennas located at three sites, approximately equally separated in (terrestrial) longitude (

Table 1. Deep Space Network Complexes.

Name	Antennas
Canberra	DSS-43 (70 m)
Australia	DSS-34, DSS-35, DSS-36 (34 m)
Goldstone	DSS-14 (70 m)
USA	DSS-24, DSS-25, DSS-26, DSS-23, DSS-13 (34 m)
Madrid	DSS-63 (70 m)
Spain	DSS-53, DSS-54, DSS-55, DSS-56, DSS-65 (34 m)
Abbreviation: DSS = Deep Space Station
DSS-13 is the DSN’s research and development (R&D) antenna.
DSS-23 is currently under construction.

; Figure 1).

There is a long history of using the DSN antennas for radio astronomical observations. Contributions of DSN antennas to astronomical discoveries include the first identification of superluminal motion using the so-called Goldstack (Goldstone-Haystack) interferometer [1]; demonstration of space-based very long baseline interferometry (VLBI), using an antenna on a Tracking and Data Relay Satellite System (TDRSS) satellite and two ground-based antennas, one of which was located in Canberra, from which a clear indication of violation of the inverse Compton limit and constraints on the physical processes occurring in the central engines resulted [2,3,4]; the detection of the infall and the inside-out collapse process during stellar formation [5,6]; and demonstration of a continued gap in understanding of stellar structure and Galactic chemical evolution (the so-called “³He problem”) by detection of a hyperfine line of ³He⁺ in the planetary nebula IC 418 using one of the 34 m antennas in Madrid [7].

The DSN antennas also participate in international VLBI observations. Specifically, the DSN is an affiliate of the European VLBI Network (EVN), with primary participation by antennas at the Madrid Complex.1 The DSN antennas at the Canberra Complex can be used as part of the Australia Long Baseline Array (LBA),2 when observing proposals are submitted through the Australia Telescope National Facility. There also has been a limited number of observations with Canberra antennas participating in East Asian VLBI Network (EAVN) observations.

Finally, DSN antennas have played an integral role in establishing and maintaining realizations of the International Celestial Reference Frame (ICRF) [8,9]. The ICRF is not only the defining frame used for specifying the coordinates of all astronomical sources, it serves as the reference against which the plane-of-sky positions of deep space spacecraft are determined for navigation of NASA’s deep space missions.

In this manuscript, we summarize the technical capabilities of the antennas and radio astronomical systems of the DSN (Section 2 and Section 3), for both the 70 m and 34 m antennas. In Section 4, we summarize specific radio astronomical systems, and, in Section 5, we provide a brief summary of proposal submission and scheduling. We conclude with illustrations of recent work and a brief discussion of future plans and possibilities (Section 6). The DSN Radio Astronomy Users Guide [10] and the Deep Space Network Telecommunications Link Design Handbook [11] provide much more detail on these topics, to which the reader is referred.

2. Antennas

As noted above, the DSN operates one 70 m antenna and multiple 34 m antennas at each Complex. The set of 70 m antennas is known as the 70 m Subnetwork (Section 2.1), while the set of 34 m antennas is the 34 m Subnetwork (Section 2.2). This section summarizes basic properties of these antennas, and the reader is referred to Imbriale [12] for additional details about the antennas in both subnetworks.

2.1. 70 m Subnetwork

At each Complex is a 70 m diameter antenna (Figure 2), with a surface suitable for observations into the K band (≈25 GHz). Constructed originally as 64 m diameter antennas based on the design for the Parkes Radio Telescope [13], the antennas were expanded to 70 m in diameter for the 1989 Voyager 2 Neptune encounter. The antenna optics are Cassegrain, with the radiation reflecting off the main reflector and a sub-reflector [12]. The sub-reflector is rotated to direct the radiation into one of three “cones” where an appropriate receiver is located (Section 3).

Table 2. 70 m Antenna Characteristics.

Band	Center Frequency (GHz)	Frequency Range (GHz)	Antenna	Gain (K Jy−1)
L a,b	1.668	1.628–1.708	all	1.18
S	2.295	2.2–2.3	all	1.12
X c	8.42	8.2–8.6	all	1.06
K	22	18–27	DSS-43 (Canberra) DSS-63 (Madrid)	0.56 0.36

^a Left circular polarization (LCP) only; right circular polarization (RCP) may be possible, with prior arrangement to change the mechanical configuration of the feed. ^b A refurbished and upgraded L-band system has been installed and is being commissioned on DSS-43 (Canberra, Section 6.2). ^c With S-X dichroic extended to enable simultaneous, dual-band observations, there is a loss at X band that does not exceed 10%. NOTE—Gain values are representative but can vary from antenna to antenna by up to 10%. For more detail, consult the DSN Radio Astronomy Users Guide [10] and the DSN Telecommunications Link Design Handbook [15]. In the DSN Telecommunications Link Design Handbook, gains are specified in units of dBi. We use a more traditional radio astronomical unit of K Jy⁻¹, converting between the two expressions using ${10}^{(G/10-26)}{\lambda }^2/(4\pi ·2k_{\mathrm{B}})$, where G is the antenna gain in units of dBi, $\lambda$ is the observing wavelength in units of meters, and $k_{\mathrm{B}}$ is Boltzmann’s constant.

presents a summary of the performance of the 70 m antennas. While the antenna surfaces are, in principle, capable of broad-band operation at least to the K band, in practice, the DSN frequency coverage for the 70 m antennas is dictated by a combination of deep space telecommunications needs and historical radio astronomical interests, both at California Institute of Technology (Caltech) and JPL and within the larger VLBI community. Specifically, the S- and X bands are standard deep space telecommunications bands; the L-band systems were installed originally for receiving telemetry from a balloon in Venus’ atmosphere [14], but are not used currently for any deep space missions; and the K band covers, but is broader than, a spectral allocation used for missions beyond geosynchronous orbit (GEO or GSO), but not to the 2 million km “threshold” identified by the International Telecommunications Union (ITU).

Observations at other frequencies are, in principle, possible. For instance, a ultra-high frequency (UHF) band system, operating near 0.4 GHz, was once mounted for a combination of spacecraft telecommunications and potential radio astronomical observations. There also have been tests of an array feed system designed to enable spacecraft communications at 32 GHz [16].

2.2. 34 m Subnetwork

Each Complex hosts multiple 34 m diameter antennas (Table 1), with surfaces suitable for observations as high as 45 GHz. Furthermore, the DSN is in the process of increasing the number of 34 m antennas, with the objective of ensuring that each Complex has at least four 34 m antennas, as four arrayed 34 m antennas ($4\times 34$ m) is approximately the collecting area of a 70 m antenna.

Most of the DSN’s 34 m antennas are “beam wave guide” (BWG) antennas (Figure 3), in which the radiation is directed by a sub-reflector into a hollow, open-air tube (termed a “beam wave guide”) within which there are a series of mirrors that direct the beam into a subterranean pedestal room. This design allows for multiple receivers to be operated and maintained without needing access to the antenna structure itself as well as being able to locate the sensitive receivers in a more stable location (pedestal) rather than on a moving antenna.

In addition to the beam wave guide antennas, the DSN continues to operate one “high efficiency” (HEF), dual-shaped reflector antenna, DSS-65. Its description is derived from the fact that, at the time of its construction, there were lower efficiency antennas in operation. Today the gain of DSS-65 is similar to that of the 34 m BWG antennas, but its description as a “HEF antenna” persists.

Table 3. 34 m Antenna Characteristics.

Band	Frequency Range (GHz)	Antenna	Gain (K Jy−1)
S	2.2–2.3	DSS-24, DSS-26	0.24
		DSS-34, DSS-36
		DSS-54, DSS-56
		DSS-65
X a	8.4–8.5	DSS-24	0.25
	8.2–8.6	DSS-25, DSS-26
		DSS-34, DSS-35, DSS-36
		DSS-54, DSS-55, DSS-56
		DSS-65
K	25.5–27	DSS-24, DSS-26	0.2
		DSS-34, DSS-36
		DSS-54, DSS-56
Ka	31.8–32.3	DSS-25, DSS-26	0.2
		DSS-34, DSS-35, DSS-36
		DSS-53, DSS-54, DSS-55, DSS-56

^a With S-X dichroic extended to enable simultaneous, dual-band observations, there is a loss at X band of no more than about 1%. NOTE—Gain values are representative but can vary from antenna to antenna by up to 5%. For more detail, consult the DSN Radio Astronomy Users Guide and the DSN Telecommunications Link Design Handbook [17,18]. In the DSN Telecommunications Link Design Handbook, gains are specified in units of dBi. We use a more traditional radio astronomical unit of K Jy⁻¹, converting between the two expressions using ${10}^{(G/10-26)}{\lambda }^2/(4\pi ·2k_{\mathrm{B}})$, where G is the antenna gain in units of dBi, $\lambda$ is the observing wavelength in units of meters, and $k_{\mathrm{B}}$ is Boltzmann’s constant.

presents the characteristics of the 34 m antennas. At present, the DSN frequency coverage for the 34 m antennas is dictated solely by spacecraft telecommunications needs. Specifically, the S-, X-, and Ka bands are standard deep space telecommunications bands; there is also a K band capability for “near Earth” telecommunications.3 Further, the frequency coverage is not uniform at all antennas, with some antennas having frequency capabilities that others do not.

At K- and Ka bands, the antenna gains can be affected significantly by both the zenith opacity due to the amount of water vapor in the atmosphere above the antenna and the elevation angle at which the antenna is pointing, with the two effects being somewhat coupled. Under generally clear conditions, the additional zenith opacity does not exceed 5% at either band, but, under cloudy conditions, zenith opacities can be as much as 10%. In general, opacities tend to be lowest at Goldstone, due to its arid climate, and highest at Canberra, with Madrid showing intermediate values. Assuming “clear” weather conditions, i.e., minimal atmospheric water vapor, antenna gains at K band do not vary by more than 10% for elevation angles above ${15}^{\circ }$, but can drop by 25% at the lower elevations. For Ka band, antenna gains do not vary by more than 10% for elevation angles above ${20}^{\circ }$, but can drop by as much as 40% at lower elevation angles.

In contrast to the more standard radio astronomical receivers, which would be dual polarization, generally only a single polarization is available from the 34 m antenna receivers, which can be either right circular polarization (RCP) or left (LCP). The availability of only a single polarization reflects the focus on spacecraft telecommunications.

In addition to the 34 m antennas described above, the DSN also operates DSS-13 as a “research & development” antenna. This antenna was originally a prototype for the 34 m beam wave guide design and broadly has similar capabilities to other DSN 34 m antennas. Most recently, this antenna has been used for initial testing of a concept for integrating radio frequency (RF) and laser (optical) communications (a so-called RF-Optical Hybrid antenna) [19]. A number of its metal panels were removed and replaced by mirrors that are reflective for the near-infrared lasers intended for laser communications. While it may be available for radio astronomical observations, it is recommended that individuals interested in potential observations contact the authors of this document to confirm availability and performance.

3. Receiving Systems

Echoing the discussion in Section 2.1, the current suite of receivers reflects both the DSN’s mission for spacecraft tracking and historical precedent. The focus of this section is in terms of those receivers and bands developed primarily for radio astronomy observations or spacecraft telemetry, tracking, and commanding (TT&C) bands that have been utilized for radio astronomical observations. In addition to the bands discussed here, the 34 m antennas also have K band (25.5 GHz–27 GHz) and Ka band (31.8 GHz–32.3 GHz) receiving systems, though both of these bands have been used only rarely for radio astronomical observations; Bansal et al. [20] is one of the few examples of dual-frequency observations involving Ka band.

While the DSN antennas also are equipped with transmitters, there usually are few concerns regarding interference from the transmitters. First, the frequencies for transmission and reception are separated. Second, particularly for the spacecraft TT&C bands, sharp filters are used. Both of these steps are taken to ensure that the reception of weak signals from distant spacecraft do not suffer interference from the transmissions used to command them. Given that radio astronomical signals are faint as well, these practices also benefit radio astronomical observations.

3.1. L Band (1628 MHz–1708 MHz)

The 70 m antennas are equipped with an L-band feed, mounted to the side of the X-band Transmit-Receive (XTR) Cone and illuminated by directing the subreflector to point slightly offset from the normal illumination of the XTR Cone. The L-band system normally is configured to receive left circular polarization (LCP) though, with sufficient advance notice, it can be reconfigured (manually) to receive RCP. Because the antenna subreflector is rotated to illuminate the L-band feed, other frequencies are not available simultaneously.

For DSS-14 and DSS-63, the L-band feed is a so-called “Potter horn” [21], which is a relatively narrowband design optimized for 1.668 GHz and a single polarization. Experience suggests that somewhat wider bandwidths than the original design may be able to be obtained. For DSS-43, a “wide-band” feed was designed [22] and installed. This feed was removed as part of a significant refurbishment of the antenna that occurred in 2020. Section 6.2 describes the current work to reinstall and upgrade this system, with a design that is modeled on that of the Parkes Ultra-Wideband Low system [23].

3.2. S Band (2200 MHz–2300 MHz)

All 70 m antennas and at least two 34 m antennas at each Complex are equipped with S-band receiving systems, installed within the S-band Polarization Diversity (SPD) Cone. Rather than moving the subreflector, a dichroic can be inserted into the optical path to direct the radiation into the S-band receiving system, thereby enabling either S-band observations or dual-frequency (S- and X band) observations. These systems are designed primarily for spacecraft TT&C, hence their relatively narrow bandwidths relative to current radio astronomical systems.

Following the feed is an orthomode junction, enabling dual polarization observations. However, because of the TT&C needs, one of the polarizations is always fed to a diplexer resulting in a slightly higher system temperature for that polarization (≈10% higher). The signals are then amplified by an low-noise amplifier (LNA), and fed to the S-band signal distribution assembly before being downconverted to an intermediate frequency (IF) for signal transport.

Typical (zenith) system temperatures are 17 K (DSS-14 and DSS-43) and 20 K (DSS-63) and approximately 22 K (34 m beam waveguide antennas) and at least 34 K (DSS-65). Because of the S-X dichroic plate, there is also a configuration enabling dual frequency-single polarization observations between S- and X bands. The dual S-X capability increases the S-band noise temperature by approximately 5 K.

Experience suggests that the S-band systems can be affected significantly by radio frequency interference (RFI), though time-domain observations are often possible as the process of de-dispersion provides some mitigation of RFI.

3.3. X Band (8200 MHz–8600 MHz)

All 70 m antennas and at least one 34 m antenna at each Complex are equipped with X-band receiving systems, installed within the XTR Cone. These systems are designed primarily for spacecraft TT&C, hence their relatively narrow bandwidths relative to current radio astronomical systems.

The X-band feed is followed by a diplexing junction in order to allow for injection of transmissions (for spacecraft TT&C), by an orthomode junction in order to enable dual polarization observations, and by LNAs.

Typical (zenith) system temperatures are 17 K (70 m antennas). As noted above, with the S-X dichroic plate extended, dual frequency-single polarization observations between S- and X bands are feasible. The dual S-X capability increases the X-band noise temperature by approximately 1 K.

3.4. Radio Astronomical K Band (17 GHz–27 GHz)

Both DSS-43 (Canberra 70 m) and DSS-63 (Madrid 70 m) are equipped with K-band systems, installed within the Host Country Cone. The focus of this section is on the DSS-43 system, as it is the more capable. The DSS-63 system is suitable for VLBI observations, but does not have as wide of a band nor the suite of backends available at DSS-43 (Section 4).

DSS-43 is equipped with a radio astronomical system operating in the 17 GHz–27 GHz band. This frequency range contains numerous atomic and molecular lines, including hydrogen radio recombination lines (RRLs), cyclopropenylidene (C₃H₂, rest frequency 18.3 GHz), water (H₂, rest frequency 22.2 GHz), and ammonia (NH₃, both ortho- and para-rest frequencies approximately 23.7 GHz). It is also a standard band for VLBI observations, motivated in large part by the water line.

The DSS-43 K-band system has two main components, an LNA Assembly (or “front-end”) and a Downconverter. We summarize both here, and Kuiper et al. [24] provide a more detailed description of the system.

The LNA Assembly contains two feed horns, allowing for simultaneous observations of both a target source and a secondary position (e.g., for calibration purposes). Each feed horn produces linear polarization, and the two feed horns are separated by 2.1′ (equivalent to 2.8 half-power beamwidths at 22 GHz). Each linear polarization from each feed horn is connected to a cryogenic LNA. Ambient temperature microwave absorbers and phase calibration signals (with 1 MHz or 4 MHz frequency combs) can be inserted into the signal path for calibration purposes.

The Downconverter receives the output from the LNA Assembly and produces intermediate frequencies (IFs). There are four signal chains (labeled E1, H1, E2, and H2 for the E- and H-plane polarizations of Feeds 1 and 2), and switches allow the signals to be switched in a manner to cancel gain profiles. Each signal chain is contains a set of bandpass filters, producing 2 GHz sub-bands centered at 18 GHz, 20 GHz, 22 GHz, 24 GHz, and 26 GHz. Following the bandpass filters, polarization converters allow the signals to be converted from linear polarization to circular polarization, if desired. Finally, local oscillators (LOs) are mixed with the sub-bands to produce 1 GHz IFs. These 1 GHz IFs are transported by fiber optic cables from the antenna to the Signal Processing Center for detection and post-processing.

Estimates of the system temperature, extrapolated to outside the atmosphere, are $T_{\mathrm{RX}}=29$ K, with approximately 10% uncertainty or variation from feed-polarization channel to channel. Under near-ideal conditions (clear skies and cold ambient temperatures), total system temperatures $T_{\mathrm{sys}}$ approaching 40 K have been measured.

3.5. Radio Astronomical Q Band (38 GHz–50 GHz)

A Q-band system was installed on DSS-54 (one of the Madrid 34 m antennas), in a collaboration between the DSN (both JPL and Madrid Complex) and Spanish astronomers [25]. This system was motivated by a combination of astronomical and spacecraft TT&C considerations. From an astronomical perspective, there are a variety of molecular lines in this frequency range. From a spacecraft TT&C perspective, there is a relevant spectral allocation near the high end of the Ka band or low end of the Q band that could be used.

However, subsequent issues with the DSS-54 antenna necessitated its removal. With the advent of an even broader bandwidth Q-band system at the 40 m Yebes antenna in Spain [26], a compelling justification for continuing the DSS-54 Q-band system has not been identified. There are no current plans to reinstall a Q-band radio astronomical system within the DSN.

4. Backends

The various DSN Complexes host different processing backends, dictated largely by previous scientific interests of Caltech and JPL staff and external users. Each backend is summarized briefly; in many cases, there are more detailed publications providing additional information.

For the backends at Canberra, they are typically configured to work with the radio astronomical K-band system, though, in principle, they can accept input from any of the available bands, radio astronomical or spacecraft TT&C, at the Complex.

The development of additional backends or user-provided backends is both feasible and encouraged. Interested individuals should contact DSN Radio Astronomy staff for discussions and assessment of feasibility.

4.1. DSN Radio Astronomy Spectrometer-Canberra

The DSN Radio Astronomy Spectrometer [27] is installed at the Canberra Complex and is designed specifically to accommodate the output of the Radio Astronomical K-band system, but, in principle, can process any of the outputs from any of the systems at the Canberra Complex (For instance, it has been used as part of an effort to characterize the RFI environment in the spacecraft-tracking K band).

The Radio Astronomy Spectrometer has two modes in which it can produce spectra, (1) FFT-based 32k-point spectra, or (2) polyphase filter-based 8k spectra. Data from the Radio Astronomy Spectrometer are provided in single-dish FITS (SDFITS) format [28] in a series of levels reflecting increasing data processing.

4.2. Time-Domain Processors

There are two means of conducting time-domain observations with DSN antennas, a Pulsar Processor designed explicitly for radio astronomical observations and an Open Loop Recorder (OLR) designed for spacecraft TT&C but that can be utilized for radio astronomical observations. This section summarizes these two backends briefly, with

Table 4. Time Domain Radio Astronomy Instrumentation Recommendations.

Observation	Recommended Backend
Low, Modest DM Searches	Pulsar Processor
High DM Searches
Single-Pulse Studies, slow pulsars, FRBs
Single-Pulse Studies, recycled and millisecond pulsars	Open Loop Recorder

Abbreviation: DM = Dispersion Measure

providing a high-level recommendation of which backend is recommended for which kind of observation.

4.2.1. DSN Pulsar Processor-Canberra

The DSN Pulsar Processor [29] is installed at the Canberra Complex. The Pulsar Processor is designed for high-speed, near-real time searching of pulsar and fast radio burst (FRB) signals. The system was designed to exploit the K-band system, and it can accommodate up to 16 subbands, each up to 1 GHz wide. Standard use would be to acquire 8 GHz of bandwidth in two polarizations, for which the current configuration of the system provides 16,384 spectral channels with time resolution as fine as 32 $\mathsf{\mu }$s with a 16-bit per sample dynamic range.

4.2.2. Open Loop Recorder

Each Complex hosts eight Open Loop Recorders (OLRs) [30], designed to process signals from the standard DSN TT&C bands. The primary intention of the OLRs is to support Radio Science experiments associated with NASA or other space agency missions, but they can also be used for radio astronomy.

An OLR receives one or more copies of the digitized IF from an antenna via a Complex-wide packet switch. An OLR can receive up to 16 IF inputs, with each IF input possibly having a different bandwidth and bit depth. Available input IF bandwidths are “narrowband,” ranging from 200 Hz to 500 kHz, or “wideband,” ranging from 1 MHz to 50 MHz. The OLRs have been validated for operational data acquisition for bandwidths up to 512 MHz (16 × 32 MHz), and users have reported success using bandwidths of 1024 MHz (16 × 64 MHz). However, the total input data rate to an OLR cannot exceed the 10 Gbps packet switch capacity.

Standard radio astronomy processing would configure an OLR to have a common configuration across all IF inputs, typically with IF bandwidths of 8 MHz, 16 MHz, or 32 MHz so as to cover the full IF, but it is possible to have simultaneous “wideband” and “narrowband” recording. If an antenna is configured in a “dichroic” mode to receive two (sky) frequencies simultaneously (e.g., S- and X band or X- and Ka band), the IF inputs can be distributed across the two (sky) frequencies.

Following recording, data are transferred off the OLR for additional processing.

5. Proposal Submission and DSN Scheduling

This information is a brief overview of the proposal process. Full details are provided in the “DSN Radio Astronomy Proposal and Scheduling Guide.”4 For reference, radio astronomical observations obtain approximately 150 hr per month, though with significant month-to-month fluctuations depending upon antenna availability and other high-priority uses of the antennas.

The DSN antennas can be used in a stand-alone capacity or as part of a very long baseline interferometry (VLBI) observation. Three long-standing principles apply to all proposals to use the DSN for radio astronomy:

• The prime responsibility of the DSN antennas is for spacecraft telemetry, tracking, and command (TT&C). While every effort will be made to accommodate projects that require time-critical observations or observations at specific epochs, such observations can be challenging to schedule given the TT&C needs of the various missions that depend upon the DSN.
• The DSN schedules time four to six months in advance. While every effort will be made to accommodate proposals submitted less than six months in advance, review and scheduling of projects will be facilitated by submission six months in advance.
• A basic requirement for all proposals to use DSN antennas for radio astronomy is that the proposal must specify how the proposed observations require some unique capability of the DSN.

There are three different categories of radio astronomical observations, which affect where proposals should be sent for evaluation and how approved projects appear on the DSN schedule (

Table 5. DSN Radio Astronomy proposal categories.

Category	Summary
Ground-Based Radio Astronomy (GBRA)	Proposals submitted to JPL, technical evaluation at JPL
European VLBI Network & Global VLBI (EGS)	Proposals for DSN antennas as part of a VLBI array, typically submitted to the EVN
Host Country Radio Astronomy (HCRA)	Proposals submitted to respective host country (Australia and Spain) entities

).

One aspect of DSN scheduling is notable as it is different than typical practice at many astronomical observatories. Rather than a strict priority-based approach, the DSN employs a negotiating process. Missions, including radio astronomical projects, that request overlapping time on an antenna endeavor to reach a mutually-agreeable solution to resolve the overlapping requests. Johnston [31] and Johnston et al. [32] provide more details about DSN scheduling and recent efforts toward automation and optimization.

6. Future Plans and Opportunities

There are three on-going efforts to modernize radio astronomy systems within the DSN. We also discuss one further, longer-term possibility.

6.1. K Band at Madrid 70 m Antenna (DSS-63)

The Radio Astronomy K-band system included a downconverter installed in the Host Country Cone on the antenna. This system had not been upgraded to a wider band system as had that on DSS-43, and subsequently components had failed. In principle, a system similar to that at DSS-43 could be implemented at DSS-63. However, in the northern hemisphere, there is a number of other radio telescopes of comparable or larger diameter with quite capable receiving and backend processing systems, including multiple such antennas in Europe. Accordingly, refurbishment and modernization focussed on ensuring that the DSS-63 K-band system could be used in the context of VLBI observations. In particular, as part of efforts to increase the wavelength coverage of the ICRF, there has been a long-standing effort to produce a K-band component of the ICRF [33]. DSS-63 is well-positioned to contribute to this effort, given its position both within Europe and at a similar longitude as the 26 m radio telescope at Hartebeesthoek, South Africa.

The refurbished K-band downconverter leveraged a design developed within the DSN for other purposes [34], thereby also allowing for a more rapid and efficient implementation. Much like the K-band downconverter at DSS-43, the DSS-63 K-band downconverter has two stages, but, for DSS-63, only one feed horn is used, meaning that only two IF signals (two polarizations) are transported from the antenna to the SPC.

The first stage is installed in the Host Country Cone, close to the LNA Assembly. It accepts two inputs (two polarizations), at sky frequencies between 18 GHz and 27 GHz. These are downconverted, by mixing with a common and tunable local oscillator (LO), to an IF centered on 10.05 GHz with a bandwidth of 600 MHz.

These two IFs are transported over fiber optic cables to the second stage, located in the Madrid SPC. In the second stage, each IF is downconverted to a 600 MHz band centered on 750 MHz. These signals are provided to the DSN’s VLBI Recording Assembly (VRA) [35], where they are recorded for subsequent shipment to the EVN correlator or other appropriate correlator.

The refurbished K-band downconverter was installed in 2024 and has been verified using shadow tracks with spacecraft that transmit at frequencies near 26 GHz. Since its installation, the refurbished DSS-63 K-band system has been used for one EVN observation.

6.2. L Band at Canberra 70 m Antenna (DSS-43)

During a significant maintenance and refurbishment of the 70 m antenna during 2020, much of the L-band system was removed from the antenna. Specifically, the L-band feed horn was mounted on the XTR Cone, which was replaced in its entirety. This replacement necessitated the removal of the L-band feed horn. Other maintenance and refurbishment removed old cables, obsolete wave guide, and other components.

A plan was developed and is being executed to reinstall a new L-band system (Figure 4). This new L-band system has several, significant improvements over the legacy system at DSS-43, and the systems that remain at DSS-14 and DSS-63. The improvements include the following:

• Introduction of an orthomode transducer (OMT), enabling both polarizations to be processed, whereas the legacy system produced only a single polarization;
• Use of WR-510 waveguide, which has a notional frequency range of 1450 MHz to 2200 MHz, rather than the legacy WR-430 waveguide (1720 MHz–2600 MHz); and
• Introduction of digitization within the antenna, immediately after amplification, rather than transport of the signal down the antenna and into the Canberra Signal Processing Center (SPC).

The LNAs and analog-to-digital converters for this new L-band system were developed and implemented by the Australia Telescope National Facility (ATNF). Both the LNAs and analog-to-digital converters leverage designs and prior work for the Ultra-Wideband Low system on Murriyang, the 64 m Parkes Radio Telescope [23].

These changes have clear scientific benefits. Most notably, the DSS-43 L-band system now can access the hyperfine transition of neutral hydrogen (H i) at 1420 MHz. The system also naturally produces both senses of polarization, resulting in higher continuum sensitivity (by a factor of $\sqrt{2}$) or the ability to conduct measurements such as of Zeeman splitting. The expansion to lower frequencies also enhances the potential for DSS-43 to participate in VLBI observations with other antennas in the Australian Long Baseline Array (LBA) or with the intermediate frequency component of the Square Kilometre Array (SKA1-Mid).

Figure 5 shows a portion of one of the “first light” spectra from the new system. Clearly visible is the hyperfine H i at 1420 MHz. Initial testing indicates that the system temperature is approximately 25 K over the frequency range of at least 1350 MHz to 1800 MHz. Observations at frequencies above 1800 MHz will be challenged by RFI from nearby cell phone towers. The system also has (cooled) filters in front of the LNAs that filters out frequencies above 2000 MHz so as to protect the LNAs from the S band (≈2100 MHz) transmitter at DSS-43 designed to send commands to spacecraft. Due to other uses of the antenna, this system likely will not be commissioned fully until early 2026.

6.3. K Band at Canberra 70 m Antenna (DSS-43)

The Radio Astronomy K-band system includes a downconverter installed in the so-called Host Country Cone [24]. The downconverter accepts four 10 GHz inputs, from the two polarizations from the two feeds installed on the Host Country Cone. Each 10 GHz input is downconverted to 10 baseband signals, each of which is 1 GHz wide and which then are transported to the Canberra SPC over fiber optic cables.

At the time that this system was developed originally, radio frequency (RF) over fiber capabilities had not advanced to the state that transporting 10 GHz of bandwidth would be feasible. Subsequent advances have enabled such capabilities. Consequently, currently under development is a two-stage downconverter.

Stage 1. 

In this first stage, each 10 GHz RF signal would be downconverted to a 2 GHz–12 GHz intermediate frequency (IF). The four IF signals would be transported over fiber to the SPC.

Stage 2. 

For backward compatibility, the IF signals would be converted to 10 baseband signals, each 1 GHz wide. However, in principle, it also would be feasible to split each IF signal such that it fed the baseband converters as well as providing a 10 GHz-wide band for direct digitization.

One of the significant advantages of this new system is that it transfers much of the hardware out of the antenna and into the SPC, where environmental conditions can be controlled more easily and routine maintenance is easier.

6.4. C Band and Wide-Band Receivers

During the time of the VLBI Space Observatory Programme (VSOP)/Highly Advanced Laboratory for Communications and Astronomy mission (HALCA) [36,37], there was an investigation of whether it would be possible to deploy a receiving system for frequencies near 5 GHz (“C band”). This frequency range is a standard one for ground-based radio astronomical observations, and it was one of the frequencies at which VSOP/HALCA observed. The intent was to enable at least one of the DSN antennas to participate as a ground antenna during the VSOP/HALCA mission, and potentially future space VLBI missions as part of the U.S. Space VLBI Program.5 A design was developed for a C band receiver, intended to be installed into the Host Country Cone on DSS-63, but the full receiving system was not developed and never installed.

Nonetheless, C band remains a standard radio astronomical receiving band, and all DSN antennas have surfaces capable of high performance in this frequency range. There is ample volume within the Host Country Cones for a C-band system at both DSS-43 (Canberra) and DSS-63 (Madrid), as well as ample volume within the pedestals of any of the 34 m antennas. One or more DSN antennas equipped with a C-band receiving system could be valuable addition to any number of VLBI networks, such as the EVN or the Australian LBA. Both Complexes also would have mutual visibility with the SKA1-Mid.

Moreover, since the original design of a C-band DSN receiving system, there have been developments of wide-band receiving systems [38,39,40]. These receiving systems aim for spectral dynamic ranges of 5:1 or larger. Thus, it would be feasible to implement a receiving system that covered a frequency range of, say, 4 GHz to 16 GHz. Such a receiving system would enable observations at the standard radio astronomical frequencies of approximately 5 GHz (C band) and 8.4 GHz (X band) as well as observations of methanol (CH₃OH) lines at 6.7 GHz and 12.4 GHz.

6.5. W Band (≈ 90 GHz)

In approximately 2000, the capabilities of the DSN antennas to operate at W band (≈$\lambda 3\mathrm{mm}$) were investigated. The motivation was to assess the possibility of using W band to achieve even higher data rates for spacecraft telecommunications than can be achieved at Ka band [41]. The subsequent focus on laser communications, including a flight demonstration of a deep-space system [42], has led to no further consideration of W band for spacecraft TT&C.

These investigations showed that a DSN 34 m antenna had a typical efficiency of 18% at W band [43]. Given other millimeter-wavelength astronomy capabilities now available, installing a receiver system for single-dish observations at W band with a DSN 34 m antenna does not seem well-justified.

A W-band capability could be a useful augmentation to the Global mm-VLBI Array (GMVA)6 or similar global VLBI observations. In particular, the concept of a Phase 2 next-generation Event Horizon Telescope (ngEHT) evisioned a global array of antennas observing at 86 GHz, 230 GHz, and 345 GHz [44,45]. Most of the antennas envisioned for the ngEHT have been of order 10 m in diameter, smaller than the DSN 34 m antennas. Even at low efficiency, the addition of one or more DSN antennas at 86 GHz could be a valuable addition to the ngEHT Phase 2 concept. However, there has been no characterization of DSN 34 m antennas at the higher frequencies envisioned for an ngEHT (230 GHz or 345 GHz), and atmospheric transmission at the DSN Complexes would have to be assessed.

Similarly, the Black Hole Explorer (BHEX) [46] concept envisions a single space-based antenna operating in conjunction with a network of ground-based antennas, at a similar set of frequencies as for the ngEHT Phase 2 concept. Augmentation of the BHEX ground antenna network with one or more DSN 34 m antennas could be compelling, particularly in the context of a NASA Astrophysics Explorer mission.