Arraying of Antennas

Arraying of DSN antennas is a network capability that has been employed whenever the required receiving sensitivity significantly exceeded that which could be established on a single aperture with the best efficiency and lowest practical system temperature. The overall architecture of the future DSN, as currently understood, depends upon the ability for an array of several 34-m antennas to mimic the functionality of a single 70-m antenna whenever needed.

The Galileo mission to Jupiter, with its broken large antenna, is the current motivating factor for arraying; the Voyager mission to the outer planets spurred the development of the current arraying tools, as well as many other changes to the network through the 1980s. The current arraying or signal combining subsystem, the Baseband Assembly (BBA), was developed incorporating digital signal processing techniques first employed in the experimental ranging machine (designated the Mu-II Ranging Machine), built by the Advanced Systems Program.

Other network changes included the construction of the 34-m HEF antennas, the expansion of the 34-m Standard (STD) antennas from their prior 26-m size, and the rebuilding of the 64-m antennas to become the dual-shape, high-efficiency, 70-m antennas. For these antennas, the high-efficiency illumination patterns are a product of the radio frequency "optics" analysis software toolkit developed with support from the Advanced Systems Program.

Several radio astronomy observatories outside the DSN also collaborated in the arraying for the Voyager encounters. Signals received at the Parkes Radio Telescope in Australia were arrayed into the Canberra DSN site, and signals at the NRAO VLA in New Mexico, were arrayed into the DSN site at Goldstone, California. The low-noise receiving amplifiers for the VLA were developed in collaboration with the DSN technology development efforts.

Arraying for Voyager could not have been approached with confidence and commitment were it not for significant prior efforts supported by the DSN Advanced Systems Program. In the DSN document Applications of Telemetry Arraying in the DSN , R. Stevens states that in 1970, early analyses led to experimentation by the DSN Spanish Complex personnel, who used the two 26-m antennas receiving signals from Pioneer 8. Simple baseband and bit-stream combining were employed. No special time-alignment of the signals was required due to the low data rate and short distance between antennas. A similar simplistic approach was applied operationally in combining signals within each complex in support of the ICE spacecraft during the Giacobini-Zinner Comet encounter in 1985. Another significant effort supported by the Advanced Systems Program was the combining of high data rate signals where time-alignment of the signals from the several antennas was essential for performance. The site was Goldstone, and the occasion was the 1974 second Mercury encounter of the Mariner-Venus-Mercury (MVM) spacecraft. The successful demonstration showed an appropriate 0.7-dB improvement in sensitivity due to arraying of the two 26-m antennas at DSS 12 and DSS 13 with the 64-m antenna at DSS 14. This improvement in sensitivity was consistent with predictions, but also demonstrated a unit-to-unit variation in performance of the then-current analog Subcarrier Demodulator Assembly (SDA).

The function of the SDA has since been incorporated into the current generation of arraying equipment for improved and more consistent performance. The next follow-on arraying activity was the development of a prototype operational baseband Real-Time Combiner (RTC), which would be demonstrated with the support of Voyager at Jupiter and Pioneer 11 at Saturn. Lessons from these early demonstrations were incorporated into the operational RTC configuration that was committed for support of Voyager at Saturn with a combined 64-/34-m array.

The flight time for Voyager from Saturn to Uranus (1986) was long enough to permit development and implementation of improved arraying in the form of the BBA, which combined the functions of the RTC and the demodulation/detection equipment into one digital processor. Experience gained via the Advanced Systems Program in both arraying and digital detection processes helped establish the design of the BBA with a tolerance of about 0.1dB.

Several additional antennas were constructed during this period, and the BBA was designed to handle up to four independent baseband signal inputs. In addition, a special variant of the BBA provided for the combining of signals from the Parkes Radio Telescope into the array at the DSN Canberra Complex, 200 km distant. A similar device was used to combine the signals from the NRAO VLA in New Mexico into the Goldstone Array for the 1989 Neptune encounter.

Meanwhile, the Advanced Systems Program continued to explore the arraying process to develop methods that might provide greater operational simplicity or improved performance, or both. Combining at the symbol-stream stage of processing is feasible for signals like those from Voyager and requires a lower information-transfer rate between the antenna sites; it also simplifies the time-alignment process for signals from widely separated antennas. Symbol-stream combining was demonstrated for intercontinental arrays first with the Giacobini-Zinner comet encounter in 1985 and later with Voyager. It was considered for a time to be a backup to the development of the remote baseband arraying with the radio observatories, but was released when that development was successfully demonstrated.

Arraying has now been an operational part of the DSN for the better part of a decade, and most of the effort recently applied to it has been via the Implementation or Operations Programs. Modest efforts via the Advanced Systems Program continue to explore the boundaries of performance for various alternative arraying architectures, including forms of carrier combining or full-spectrum, as well as baseband and (complex) symbol-stream techniques.