Table Of ContentTIlE FUTURE OF THE DEEP SPACE NETWORK: TECIINOLOGY
DEVELOPMENT FOR K.-BAND DEEP SPACE COMMUNICATIONS"
Alaudin M. Bhanji
Manager Communications Ground Systems Secti_m
Mail Stop 238-737
Jet Propulsion Laboratory
4800 Oak Grove Drive
Pasadena CA 91109
[email protected]
Introduction is summarized in this paper, isongoing to provide
Projections indicate that in the future the number of the technology and hardware required to realize high
NASA's robotic deep space missions is likely to performance communication links at these
increase significantly. A launch rate of up to 4-6 frequencies, denoted as K,-Band.
launches per year is projected with up to 25
simultaneous missions active [1]. Future high- Advances in Antennas
resolution mapping missions to other planetary In order to realize the increased signal to noise ratio
bodies as well as other experiments are likely to that the jump to Ka-Band has to offer a number of
require increased downlink capacity. These future advances in the Deep Space Network's set of
deep space communications requirements will, antennas and microwave components is required.
according to baseline loading analysis, exceed the These areas include development of new high-
capacity of NASA's Deep Space Network in its performance feedhorns and frequency selective
present form. surfaces for multi-frequency operation. While the
decreased antenna beamwidth at Ka-Band is
There are essentially two approaches for increasing responsible for the improved link margin, a tighter
the channel capacity of the Deep Space Network. requirement on the pointing capability of the
Given the near-optimum performance of the network antennas must follow. Operation of the Deep Space
at the two deep space communications bands, S-Band Network's 70m antennas efficiently at K_-Band
(uplink 2.025-2.120 GHz, downlink 2.2-2.3 GHz), presents some special challenges. In particular, while
and X-Band (uplink 7.145-7.19 GHz, downlink 8.4- the surface deformations induced by gravity as the
8.5 GHz), additional improvements bring only antenna moves in elevation are insignificant at S and
marginal return for the investment. Th'us the only X-Band they are severe at Ks-Band. Two specific
way to increase channel capacity is simply to gravity compensation systems, a deformable mirror,
construct more antennas, receivers, transmitters and and an array feed system are currently under
other hardware. This approach is relatively low-risk consideration. Details on these systems follow.
but involves increasing both the number of assets in
the network and operational costs. Feedhorns
Multi-frequency operation in the Deep Space
The approach selected by the JPL organization Network istypical. Various systems provide
responsible for providing deep space communications simultaneous downlink/uplink capabilities at S-Band
service, the Telecommunications and Mission and X-Band and inboth bands simultaneously. This
Operations Directorate, is to look to a higher trend will continue, moving to simultaneous K_ -
frequency communication link to provide higher Band/X-Band downlink with the possibility of
bandwidth and/or increased signal-to-noise ratio. simultaneous X-Band uplink. One approach for
This approach is part of the history of the DSN, providing this capability is to develop multi-
having driven the uplink and downlink frequency of frequency feedhorns that provide both diplexing
choice from S-Band to X-Band. Deep space capability and high illumination efficiency for the
communications bands are already allocated and in antenna at all bands. The requirements for these feeds
use at 31.8-32.3 GHz for downlink and 34.2-34.7 are quite unique due to the dynamic ranges involved.
GHz tbr uplink. Tt) this end significant _ork, which Typical uplink po_ers are 20 kW CW (73 dBm) and
['he research described in this paper was carried out by the Jet Propulsion Laboratt_ry. Califi_rnia Institute of
['cchnology. under a contract with the National Aeronautics and Space Administration.
the[)SN'scryc_genically-coroelceedivers
have noise identified. Both t>t"these effects are currently under
f]oors in the 20K range at K,-Band Typical investigation. Thermal disturbances, particularly
requirements call for the teed to provide as much as differential heating of the strt,ctural member
t30 dB of isolation between the transmitter and supporting the alidade portion of the antenna have
receiver ports. The challenges involved indesigning also been identified as a source of pointing errors.
this particular type of teed involve obtaining the Two approaches have been investigated to mitigate
required isolation, and return loss and efficient these thermal effects. (1) Experiments to investigate
antenna illumination in all three bands. Matching is the effect of insulating the key structural members
achieved through the use of transition regions with have been carried out. (2) An active system using
groove depth and width variation, and iris matching temperature sensors on the key members and
structures on the side-coupled ports, [21. calculations based on a structural model of the
antenna has also been used to update the antenna
Frequency Selective Surfaces pointing in real time. The flatness of the azimuth
An alternative approach for separating signals in track has also been identified as a source of pointing
different bands is by performing the filtering error. A pointing correction system using
operation in free-space using frequency-selective inclinometers on the track and a look-up table has
surfaces (FSS). Two new frequency-selective also been implemented on an experimental basis.
surfaces have been developed for K,-Band. Both are Finally, the simple linear models used in the pointing
of the high-pass variety, where the high frequency look-up table have been upgraded to carry spherical
radiation passes through an array of tightly packed harmonic terms of up 4_ order. Work in the blind
apertures in a half-wave thick metal plate. This type pointing area is ongoing and the final decision on
of FSS is preferred over a low pass structure which improvements will be included in the
consisting of metal patches on a dielectric substrate operational systems is yet to be made.
for two reasons. (1) Thick metal plates are quite
capable of handling reflection of the high power Gravity Compensation Systems
uplink signal whereas printed structures are not. (2) The largest antennas in the DSN, the 70m diameter
Metal plates with air-filled apertures offer a low loss dual-shaped antennas, offer the highest link margin
structure for the downlink signals that must pass due to their large collection area. Unfortunately they
through the plate, no dielectric loss is encountered. were not constructed with operation at Ka -Band in
One FSS has been designed to pass both the K, -Band mind, and gravity loading on the structure causes
uplink and downlink bands while reflecting the deflections in the surface. At Ka -Band the deflections
uplink and downlink at both S and X-Band. A second cause a gain loss of as much as 8dB at high elevation
FSS has also been designed to reflect the Ka -Band angles and 2.5 dB at low elevation angles, far from
downlink and transmit the uplink. Due to the small the rigging angle of 45 degrees. In order for these
separation between these bands, 32 arid 34.5 GHz, a antennas to offer the required performance at K, -
five-layer high-pass FSS has been designed which Band agravity compensation system must be
incorporates a waveguide iris filter inside each of the employed. Work on two such systems, a deformable
apertures in the FSS. mirror and an array feed with real-time combining is
described below.
Blind Pointing Upgrades
While the increased antenna gain afforded at K,-Band Deformable Mirror
improves the link margin the accompanying decrease The first approach to gravity compensation uses an
in beamwidth demands more accurate antenna actuated flat plate in the beam path. The plate is
pointing. The full 3dB beamwidth of a 34m-diameter deformed as the antenna tips in elevation in order to
antenna at 32 GHz is approximately 16mdeg. Once a pre-distort the beam phase so that when it is reflected
signal is received from the spacecraft monopulse by the distorted main reflector flat phase is restored.
tracking may be used as a pointing aid. as discussed The shape of the deformable mirror is computed at a
below. The antenna must point unaided _blind point) number of elevation angles for which the main
It)within the pull-in range of the monopulse. In order reflector surface has been measured using microwave
to achieve blind pointing accuracy in the milidegree holography. The computation of the mirror shape is
r:*nge a number of antenna improvements are being carried out using geometrical optics. Once the ideal
investigated. Elevation encoder hysteresis has been shape is found the best possible set of actuator
identified as one of the factors limiting blind-pointing displacements is found using a finite element
capability of some of the existing antennas inthe structural model that takes into account the thickness
DSN. Likewise discontinuities in the data gear which of the plate, It)cation of the actuators, and plate
is used to provide azimuth position ha_e also been thickness. The key mechanical design parameters are
the plate thickness, the number of actuators, and their experiment wtll tn'¢t_lve the (_'assinl spacecraft. The
location. Knowledge of the actual main reflector initial development transmitter for this application
surthce isessential for this approach to be effective. provides 80 W CW at the feedhorn and is based on a
Inpractice the system would be implemented by commercially avatlable 100 W TWTA. One of the
adjusting the plate via a lookup table and thecurrent most stringent and atypical specifications placed on
elevation angle, andhence provides noreal-time the transmitter is long term frequency stability, which
correction for factorssuch aswind and temperature. is expressed as Allan variance, (_ff:/t), versus time
A deformable mirror wasdesigned andtested atthe interval. For the specific radio science experiment
34m researchanddevelopment antennaseveral years under consideration the important time interval is the
ago, [3]. Recently the same mirror was installed on round trip light time to the spacecraft. Frequency
the70m antennaon atemporary basis.Although the stability requirements of one part in 10t5for time
design wasknown to besub-optimum for the70m intervals of 100-1000 seconds are typical. Currently
antenna geometry gain improvements of asmuch as an 800 W CW transmitter based on a 1kW klystron
3.7 cLBwereobserved athigh elevation angles, and amplifier is under development to replace the 80 W
all resultswere inreasonable agreement with theory. system. In order to meet the stringent stability
Before actual implementation are-optimization of the requirements with this type of amplifier particular
mirror design for the70m application would be attention is being paid to power supply stability and
necessary. coolant temperature stability.
Array Feed Compensation System Low Noise Amplifiers
A second approach for gravity compensation is an Two types of low noise amplifiers are currently under
array feed compensation system [4]. In this case an development at K=-Band, solid-state amplifiers based
array of 7feeds is located in the antenna's focal on High Electron Mobility Transistors (HEMT), and
plane. Energy that isdefocused due to the main high performance maser amplifiers. HEMT
reflector distortions is captured by the additional 6 amplifiers are currently operating at physical
feeds surrounding the central feed. Signal processing temperatures of 15K and providing equivalent noise
techniques are employed to determine the optimum temperatures referenced to the feed input of 21K.
weights for recombining the 7 received signals based When antenna spillover, loss, and sky contributions
on their individual signal-to-noise ratios. The are included operational noise temperatures are in the
advantages of this system include the ability to 38-45K range for these solid state LNA systems. A
correct for real-time effects such as wind and K=-Band maser LNA is available as well, running at
temperature. No main reflector surface measurements a physical temperature of 2K, a noise temperature of
are required. Disadvantages include the need for 7 5K referenced to the feed, and an overall operating
cryogenically cooled LNAs, and the limited fill- temperature of 22-29 K. Recent maser work is
factor of the array feed in the focal plane. This focused on increasing bandwidth and moving to more
system has been extensively tested on the 34m reliable closed-cycle refrigerators. Any further
research and development antenna and was recently improvements in maser noise performance would be
installed on the 70m antenna along with the quite marginal since these devices are presently
deformable mirror. A gain improvement in the operating near the quantum limit with respect to
vicinity of 3-4 dB was observed at high elevation noise temperature.
angles, and 2dB at low elevation angles with this
system, also in good agreement with theoretical Monopulse Pointing
predictions.
As was discussed earlier pointing the DSN 34m and
Advances in Electronic Equipment 70m antennas accurately at K_ -Band isa significant
In addition to the antennas improvements a number challenge. Fortunately a downlink signal from the
of improvements/additions to the Deep Space spacecraft is generally available to aid in pointing.
Network's electronic components are also required. The monopulse tracking system planned for
Three major areas that need development are implementation is based on a corrugated horn with a
transmitters, low noise amplifiers, and monopulse cryogenically cooled TE.,_ tracking coupler, and dual
tracking systems. Some details regarding these HEMT LNAs. Error signals from the difference
systems are pr_v,ided below. channel are then used to provide pointing corrections
to the antenna controller in real time. The pull-in
Fransmltters range of the mon_pulse system isapproximately
The near-term application of the K=-Band uplink lOmdeg, _ell _ithtn the blind pointing goal for the
capability is in radio science. The first such antenna.
An Operational 34m Antenna at K.,-Band Conclusions
Figure l below depicts the layout in the pedestal area A short description of several of the key components
of one of the 34m beam waveguide antennas which is required that will make deep space communication at
already operational at K_-Band and will be used in K, -Band for the Deep Space Network areality has
conjunction with the Cassini spacecraft to pertbrm been given. As ongoing experiments and research are
radio science experiments employing both K_ -Band completed, final implementation details am under
and X-Band uplink and downlink. Shown in the consideration, with implementation of K. -Band into
leftmost position are a K_-Band transmitter and feed the Deep Space Network taking place during the next
which illuminates the left most curved mirror. This decade.
radiation the passes through the first FSS which
reflects the K, -Band downlink signal into the Ka- References
Band receive feed and LNA. Both K, -Band signals [1] C.D. Edwards, C.T. Stelzried, L.J. Deutsch, and
pass through asecond ESS that reflects both the L. Swanson, "NASA's Deep-Space Telecom-
uplink and downlink X-Band signals into an X-Band munications Roadmap", TMO Progress Report 42-
feed and waveguide diplexing system. This X-Band 136, pp. 1-20, Feb 1999.
waveguide system is then connected to ahigh power [2] J.C. Chen, P.H. Stanton, and H.F. Reilly,
X-Band transmitter and X-Band LNA. Currently one "X/X/Ka-Band Horn Design", IEEE AP Symposium
operational antenna and the 34m research and Digest, pp. 2022-2025, July 1996,
development antenna are capable of K_ -Band [3] R. Bruno, W. Imbriale, M. Moore, and S. Stewart,
operation. In addition to the radio science "Implementation of agravity compensating mirror on
experiments planned for Cassini, which iscurrently alarge aperture antenna", AIAA Multidisciplinary
in transit to Saturn, one or both of these antennas Analysis and Optimization, Bellevue, WA, Sept.
have tracked several other spacecraft atK_ -Band. 1996.
Three spacecraft, Mars Observer. Mars Global [4] V.A. Vilnrotter, E.R. Rodemich, and S.J. Dolinar,
Surveyor, and Deep Space i have flown or are flying "Real-Time Combining of Residual Carrier Array
Ka -Band downlink technology demonstration Signals using ML Weight Estimates", IEEE Trans.
systems and have been tracked by successfully these On Communications, Vol. 10, No. 4, pp. 606-615.
antennas.
[ _-'o..)()R
Figure 1.K_,-Band uplink/downlink, X-Band uplink/downlink system layout.
