emerging technologies for communication satellite payloads
TRANSCRIPT
Progress in Aerospace Sciences 50 (2012) 27–34
Contents lists available at SciVerse ScienceDirect
Progress in Aerospace Sciences
0376-04
doi:10.1
E-m
journal homepage: www.elsevier.com/locate/paerosci
Emerging technologies for communication satellite payloads
Mehmet Yuceer
Turksat Satellite Design Division, Golbas-ı, 06389 Ankara, Turkey
a r t i c l e i n f o
Available online 23 December 2011
Keywords:
TWTA
MPM
Antenna
Flexibility
Reconfigurable
Tunable
21/$ - see front matter & 2011 Elsevier Ltd. A
016/j.paerosci.2011.11.001
ail address: [email protected]
a b s t r a c t
Recent developments in payload designs will allow more flexible and efficient use of telecommunica-
tion satellites. Important modifications in repeater designs, antenna structures and spectrum policies
open up exciting opportunities for GEO satellites to support a variety of emerging applications, ranging
from telemedicine to real-time data transfer between LEO satellite and ground station. This study gives
information about the emerging technologies in the design of communication satellites’ transceiver
subsystem and demonstrates the feasibility of using fiber optic links for the local oscillator distribution
in future satellite payloads together with the optical inter-satellite link.
& 2011 Elsevier Ltd. All rights reserved.
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2. Design of power modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.1. Current status of TWTAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.2. Advances in wideband TWTA linearization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3. Antenna and feed designs for flexible coverages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.1. Mechanically reconfigurable passive antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.2. Electrically reconfigurable passive antennas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.3. Active antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4. Tunable passive components. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5. Bright future: photonics for RF front-ends and optical communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
6. Spectrum design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
1. Introduction
Space technology is conservative in terms of modification,which needs a long heritage to be approved among the industry.However the system requirements coming from the satelliteoperators such as flexible power, flexible coverage, and flexiblebandwidth drive the system developers to introduce newsolutions both in satellite bus phase and payload phase. Flexibilityneeds are shown in Fig. 1.
For advancement, fiber optic solutions provide an attractivealternative to conventional RF distribution sub-system and harnesses;they may meet the low phase noise requirements, while ensuringdrastic mass savings and suppressing isolation and EMI/EMC issues,requiring only low extra power consumption. For example, coaxial
ll rights reserved.
cables have a mass of about 53 g/m versus 4 g/m for qualified opticalfibers. Aside the improvement in distribution, another concern is thefeed design.
Space antenna feed systems are growing in terms of complexity,as mission requirements needed by customers are becoming morespecific. The widening of frequency bandwidth, the frequencyreusability, the increase of power handling as well as reduction ofmass while reaching better RF performances are the goal andconstraints in feed design. The use of efficient analysis andoptimization tools is necessary to meet these aims. Today, simula-tion tools give accurate results for RF performances of singlecomponents. Thus, it is possible to investigate new structures andto significantly reduce the bread-boarding phase as well as thedevelopment cycle. The current trend is to integrate more and morecomponents in the simulation in order to optimize the overallelectrical performances of the feed chain. Again, compactness isrequired. Reconfigurable antennas for multimedia applications
Fig. 1. Flexibility possibilities.
Table 1Comparison of TWTA and SSPA power capacity.
Frequency band Max. TWTA power (W) Max. SSPA power (W)
C �2200 �1500
X �2500 �1200
Ku �1200 �500
Ka �500 �50
Q �200 �5
M. Yuceer / Progress in Aerospace Sciences 50 (2012) 27–3428
need compact feeds with numerous integrated components; asglobal RF chain performances are critical, optimization softwareshall easily handle the partitioning of the structure and improvethose performances while reducing calculation time. Power hand-ling capability is also a rising consideration as new payloads arerequired to support more channels and output power. Multipactor,PIM and thermal management have to be taken into account duringthe design phase. Finally, mass and volume reduction remain asconstant tasks to be studied. Hence optical distribution helps inboth to achieve these reductions and the increment in the perfor-mance of the RF chain. In this study, the advancements and currentsolutions utilized by the satellite developers are presented.
2. Design of power modules
2.1. Current status of TWTAs
At millimeter-wave bands, the performance of vacuum electro-nics devices is not likely to be matched by solid-state poweramplifiers in the near future. Vacuum electronics amplifiers continueto play an important role in high-power transmitter applications. Atmicrowave frequencies, recent advances in wide bandgap semicon-ductor devices such as GaN and power combining techniques areenabling SSPAs to challenge the power advantage of vacuumelectronics amplifiers, but vacuum electronics devices still hold theefficiency edge even at these frequencies. At millimeter-wave bands,the performance of vacuum electronics devices is not likely to bematched by SSPAs in the near future. The use of linearization furtherenhances the efficiency edge of vacuum electronics amplifiers.The development of technology at the upper millimeter-wave andsubmillimeter-wave are enabled by cold cathode technologies andmicrofabrication techniques. Most microwave engineers fail toappreciate the capability, efficiency and reliability of vacuumelectronics technology. Only vacuum electronic devices meet manyof the demanding requirements for reliable performance. In thedesign of spaceborne high power transmitters, a strong emphasis isplaced on minimizing the power consumption, applied voltage, sizeand weight of the amplifiers. For such application, where
instantaneous bandwidth is also a requirement, helix and coupled-cavity TWTs are the devices of choice. The typical available powersfor commercial satellite communication TWTAs and SSPAs arecompared in Table 1.
Power advantage of TWTAs over SSPAs is significant, especiallyat the higher frequency bands. At the lower frequency bands,TWTAs are challenged by SSPAs. The linearity of a power amplifieris always of great importance. For communication applications,the TWTA has long been considered as a device with poorlinearity compared to the SSPA. The general belief is that a TWTAmust be backed-off 3–4 dB from saturation to achieve the samelevel of linearity as an SSPA. This is an incomplete statement. It ispointed out in [1] that the power consumed by a TWTA is usuallyless than that of an SSPA with only half of the rated RF power. As aresult, for two amplifiers with the same total power consumption,a TWTA generally has more available linear power than an SSPAfor most of the frequency band. The two amplifiers have the samelinearity performance for lower-power and lower-frequencyamplifiers only. Furthermore, the linearity of a TWTA can beimproved much more by use of linearization than that of an SSPA.As a result, more of the TWTA’s RF power that is lost due tooutput back-off is available as linear power.
Predistortion linearization is a simple and effective techniquefor improving the performance of both SSPAs and TWTAs [2]. Itseffectiveness on TWTAs is more significant because of slowapproach to saturation and the higher nonlinearity of the TWTA.It has been shown in [3] that by applying the fifth orderlinearization, it is possible to achieve less than 1 dB overall gaincompression at saturation so that the combined transfer curve
Fig. 3. Microwave power module (courtesy of TESAT, Germany).
M. Yuceer / Progress in Aerospace Sciences 50 (2012) 27–34 29
approaches that of an ideal linear limiter. Consequently, the needfor back-off from saturation can be greatly reduced to takeadvantage of the higher efficiency close to saturation.
Last decade has seen a leap in the performance of helix TWTs.These devices have critical applications in high data rate commu-nications. Compact (around 3.5 lb) helix TWTs at Ka-band with500 W continuous wave power and 55% efficiency and at Q-bandwith 230 W CW power and 43% efficiency have been demon-strated [4]. A major contributing factor for the millimeter-waveTWT achievements is the use of a simulation-based design andmethodology. This design methodology relies heavily on therecently developed codes for vacuum electron devices [5–7].
A coupled-cavity TWT uses a slow-wave circuit that is mechani-cally and thermally more robust than a helix. A coupled-cavitystructure is usually larger than a helix structure at the samefrequency. It can therefore provide higher power at the samefrequency or operate at much higher frequencies than a helix.In case of a low noise, high efficiency compact microwave ormillimeter wave power amplifier, the appropriate technology touse is solid state for the input amplification and vacuum electro-nics for the output. This gain partitioning between solid state andvacuum electronics is what defines the Microwave Power Modules,MPM [8,9].
When combined with a miniaturized EPC (electronic powerconditioner), a high power density RF amplifier module resultsproviding 100 W and higher in the microwave bands and 50 Wand higher in the millimeter wave bands. Due to its small size andhigh efficiency, MPM is particularly well-suited for use inresource constrained applications such as satellite. An MMPM(millimeter-wave power module) is simply an MPM designed tooperate in the millimeter wave range as the name suggests.MMPMs have been developed at the Ka and Q-bands. An exampleof Ka-band MMPM and its internal architecture are shown inFig. 2 and in Fig. 3, respectively.
Since its initial development in the 1990s, the MPM has madesignificant jumps in both performance and functionality. Theseadvances are the result of the advanced modeling and simulationin component design, as well as the steady progress in miniaturemultistage TWT collector technology and incremental improve-ments in power conditioning efficiency. Interestingly, althoughthe power of the latest generation MPMs and MMPMs hasdoubled, their size and weight have not increased accordingly.The result is an increase in RF power density, 1.4x power density(W/volume) and 1.6x specific power density (W/mass) over firstgeneration performance. With passive mechanically or electri-cally reconfigurable antennas, the antenna gain and, conse-quently, EIRP (Equivalent Isotropic Radiated Power) will changewith the coverage modification. The concept of a flexible powerMPM (Microwave Power Module) composed of the LinearizedChannel Amplifier (LCamp) associated with the TWTA is anelegant solution in order to adjust the EIRP to mission require-ments. This concept is designed to change by remote control the
Frequency range: 17.8-20.2 GHz Operational BW: 500 MHz (800 MHz optional) RF input power: -55 to -25 dBm RF output power: 60-120w (in 63 steps) Transition time: ~15 msec/step DC input voltage: 50/70/100 V Efficiency MPM: >60 % (at Pout max.) Dynamic range: 30 dB (1 dB steps) Mass:EPC + LCAMP: 1600 g TWT (CC): 750 g
Fig. 2. Flexible Ka
saturated output power of the TWT (by acting on the anodevoltage) to keep its power efficiency quasi-constant for differentoutput power levels, typically in the range of 3 dB variation.The flexible LCamp is designed to compensate the TWT gain driftand non-linearity over the saturated output power range.This concept achieves linearity improvement and a significantgain (�10%) in satellite power.
2.2. Advances in wideband TWTA linearization
During the past 7 years, there has been great progress in thelinearization of TWTAs. Linearization is still primarily by means ofpredistortion because of the wider bandwidth and higher effi-ciencies achievable by this form of linearization. Both analog anddigital linearizations are now being applied to TWTAs. Digitallinearization offers the advantage of near ideal transfer responsecorrection. A two-tone C/I (Carrier-to-intermodulation ratio) ofgreater than 50 dB can be achieved with a TWTA at 3 dB OBO
band MPM.
M. Yuceer / Progress in Aerospace Sciences 50 (2012) 27–3430
(output back-off) from saturation. The major constraint of digitallinearization is bandwidth. Although it is now possible to digitallyprocess signals over bandwidths of greater than 1 GHz, practicalbandwidths are around 100 MHz due to limitations relating tocomplexity, cost and power consumption of digital componentson the market.
There has been significant progress in extending the bandwidthof analog linearization. It was not very long ago that a 20%bandwidth was considered wideband. Today, multioctave, multi-GHz bandwidths have been achieved [10]. For amplifiers of anoctave or greater bandwidth, both even- and odd-order distortionproducts and both intermodulation and harmonic distortion arepresent. A linearizer must correct both to achieve linear perfor-mance. Even order distortion can be minimized by the use of apush-pull structure. If more suppression is required, an even-ordernonlinear generator can be used to supplement the odd-ordernonlinearity.
In many wideband high-power amplifier applications, perfor-mance is not required across the full spectrum but only at specificintervals. For such cases, multiband linearizers may provide betterperformance than a single wideband linearizer. Dual and tri-bandlinearizers have been developed for C, X and Ku bands. Thisapproach uses two or three independent predistorter modulesthat can be switched between common input and output ampli-fiers and attenuators. Using this approach, each predistorter canbe aligned for optimum performance in a particular frequencyband. In general, with TWTAs, narrower the bandwidth, higher theimprovement obtained by linearization.
3. Antenna and feed designs for flexible coverages
Space antenna types, for civil and military applications are thefollowing:
�
Deployable antennas, � Earth deck fixed antennas, � Steerable antennas, � Horns and direct radiating arrays.Primary feeds are versatile products, due to the need ofadapting their design to the antenna performance goals. Theytypically integrate several building blocks, each one assuming aspecific function:
�
Frequency discrimination components: filters, Tx/Rx diplexers, � Polarization discrimination components: OMT, septumpolarizer,
� Radiating components: conical, pyramidal, or corrugated horns, � RF sensing subsystems (mode extractor), � Other components: power dividers and combiners, couplers, bends.Each component is designed to meet one or more driving criteria,such as high power handling, broadband operability and compact-ness. All these feeds are designed according to an establishedmethodology for obtaining the best performances taking into accountsome specific constraints. Other new techniques can be employed forimplementing large RF antennas as well. This membrane can beemployed as a space-fed array or as a reflecting array, operating inconjunction with a feed structure that is either station kept on or nearthe axis of the film or is held there by a space tether operating inconjunction with a counterweight on the other side [11].
Each component in the feed chain must fulfill several specificneeds. As objectives and constraints are clearly identified beforethe design process, optimization can be carried out in a moreeffective way. The main objectives are found in the RF domain:
�
VSWR at specified ports, � Decoupling between ports (frequency, polarization or bothdepending on the component),
� Axial ratio for circular polarization components, � Insertion losses, � Radiating pattern and cross-polarization level for radiatingelements,
� High order modes, which could generate RF performancesdegradations.
In parallel, some constraints have to be taken into account:
�
Power handling: multipactor effect for Tx feed chains, Coronain some specific components, � PIM: to achieve a low PIM level, the number, type and positionof flanges have to be investigated,
� Weight, compactness, satellite implantation, � Mechanical: workmanship process induces several constraints asscrews position, round-offs of milling or half-shell manufacturing,
� Thermal control.Most of these constraints are not explicitly implemented in theoptimization software. The designer experience is important fordriving the optimization in a good way. A preliminary design iscarried out with all the above described parameters. When thesolution is validated (mainly for manufacturing aspects), aconvergence process is engaged. This step is time consuming. There-fore, in order to avoid too long computation time, it is recommendedthat the preliminary design give results, which meet the specifica-tions, and if possible with comfortable margins. Then, a crossedvalidation of the RF results is done with a second analysis software ina view to secure the model consistency in terms of RF performances.After a component or subsystem has been designed, some additionalanalyses are performed:
�
A sensitivity analysis to check the stability of the solution andhighlight the designed element critical parts that could besensitive to mechanical flaws, � An electrical field analysis to predict hot points localization inthe structure for a thermal control, as well as critical over-voltages for multipactor analysis,
� An RF performances shift analysis under thermal conditions(over the specified temperature range),
� Multipactor analysis, � Sensitivity analysis, � Thermal variation.If no frequency plan flexibility is required, the simplest andmost economical solution consists of keeping the classic channe-lized repeater. In this case, the antenna solutions are passive ones,mechanically or electrically reconfigurable.
3.1. Mechanically reconfigurable passive antennas
Mechanically steerable and zoomable passive antennas (Fig. 4)are already available in various operator satellites. The antenna(Gregorian architecture) generates one beam (circular or elliptic)per polarization, and beam-pointing can be achieved by the tworotation axes (N/SþE/W) of the antenna. In case of an elliptical spot,the ellipse orientation is made possible by the rotation of thesub-reflector. Zooming capability (for example, spot size extensionfrom 11 to 71) is achievable by the mechanical translation of themain reflector on the focal axis.
A new solution consists of the use of a printed reflectarray (a flatpanel reflector composed of reflecting cells fed by the antenna feed).Each cell receives radiating power from the feed, and reflects the
Fig. 5. Electrically reconfigurable passive antennas; (a) lens antenna (b) printed reflect array.
Fig. 4. Mechanically reconfigurable passive antennas; (a) 3-D view, (b) side-view.
Fig. 6. Active array-fed-shaped-reflector.
M. Yuceer / Progress in Aerospace Sciences 50 (2012) 27–34 31
wave with a given phase permitting contoured beam coverage. Onlyone contoured beam can be generated by one passive reflect array,but as the panel is flat, it is possible to superpose several panelsdedicated to particular coverage areas (like the pages in a book), andto select in orbit the panel corresponding to the desired coverage.
3.2. Electrically reconfigurable passive antennas
Two solutions are under development: a lens antenna (Fig. 5a)and a passive array-fed shaped reflector (AFSR). They are bothbased on the use of ferrite phase shifters to control the beamshaping electrically. Ferrite phase shifter technology is a maturetechnology widely used in US military communication satellites,
and selected for its low loss characteristics. The lens antennaconsists of an offset parabolic reflector fed by a single feed, and areconfigurable electromagnetic lens located between the feed andthe reflector. The lens comprises an RX and TX array of radiatingelements, with ferrite phase shifters connecting the RX and TXelements. The beam shaping is achieved according to the phasecontrol of the RF paths connecting the RX and TX elements.The passive AFSR consists of an array of feeds illuminating ashaped reflector. The reflector shaping is designed so that allradiating elements contribute to the beam shaping. The antennainput is distributed to all radiating elements. The illumination canbe controlled to form the beam according to the phase control ofeach RF path performed by ferrite phase shifters. A printed
M. Yuceer / Progress in Aerospace Sciences 50 (2012) 27–3432
reflectarray antenna (Fig. 5b) is another solution for passiveelectrically reconfigurable antennas. MEMS technology (MicroElectro-Mechanical Systems) allows low loss and miniaturizedimplementation of micro switches. They are introduced in thereflecting cells to control the phase shift, and thus the antennafoot-print.
3.3. Active antennas
If frequency plan flexibility is required (flexibility of channelbandwidth and channel spacing), it is not possible to use achannelized repeater architecture, because no solution exists forflexible output multiplexing (omux) filters. In this case, it isnecessary to select a non- channelized (wideband/multi-channelamplification) architecture. Active antennas perform wideband/multi-channel amplification by design, and are consequentlycompatible with coverage and frequency plan flexibility. In orderto deal with the poor power efficiency of Solid State PowerAmplifiers, the preferred solution for transmit applications is touse an Array Fed Shaped Reflector with one TWTA per radiatingelement (Fig. 6). In this case, the coverage flexibility is performedat a low level in the analog Beam Forming Network (BFN).
4. Tunable passive components
The design of electronically tunable microwave componentshas been a subject of growing interest in recent years, hence theimplications in space technology. There are many strategies andtechnologies to achieve this goal, including the use of semicon-ductor varactor diodes, micro-electro-mechanical systems(MEMS), ferroelectric components and liquid crystals, amongothers. In some latest works [12–15], electrically small tunablemetamaterial resonators, mainly split ring resonators and com-plementary split ring resonators are used in the design of tunablemicrowave components. Specifically, such tunable resonatorshave been applied to the design of filters.
Frequency tunable filters find many applications on satellitecommunications. There have been feasibility studies on the devel-opment of high unloaded Q narrow-band band-pass filters tunablewith RF MEMs switches. Electronically tunable 3-D cavity resonatorsand filters have been recently proposed in the quest for very high Qsas an alternative to mechanically tuned filters [16–18]. Also, firstMEMS electrostatically tunable loaded-cavity resonator thatachieves Q values of 460–530 over a very high continuous tuningrange (3.4–6.2 GHz) has been demonstrated recently [19]. The besttunable Q (300–900) filter demonstrated so far was thermallyactuated RF-MEMS metal sheets above dielectric resonator materialsin the 15.6–16 GHz frequency range [20]. That filter has a 1–1.3%bandwidth but a limited tuning range of 2.5%. A recent work hasdemonstrated a concept developed to tune the frequency of narrow-band evanescent-mode metal-cavity filters loaded with RF-MEMSdigitally tunable capacitors [21]. A two-pole filter is demonstratedwith an unloaded Q of 315–460 over 2.96–3 GHz, better than 2.1 dBinsertion loss and a 1 dB bandwidth of 1.2–1.5%. Although themeasured performance of the filter proposed shows a limited tuningrange, the full-wave model along with the identified improvementperspectives enhance the potential of that study on narrow-bandtunable filters by the spread of RF-MEMS usage. In order to achievethe predicted tuning range, it is important that the RF-MEMS switchhas both the desired isolation in the up-state and excellent ohmiccontact when pulled-down. Each RF-MEMS switch on a tuningelement requires a separate bias-line. Hence, space environment’snegative effects have to be considered on these bias-lines.
In another study for higher power applications [22], the authorsshow that the open split ring resonators (OSRRs) and open
complementary split ring resonators (OCSRRs) can be loaded withvaractor diodes in order to implement tunable filters in coplanarwaveguide (CPW) technology. By loading a CPW with shunt-connected varactor-loaded OSRRs, a band-stop filters with tunablecentral frequency results. The electronically tunable band stop canbe switched to a pass band by replacing the shunt resonators withvaractor-loaded OCSRRs. The measured tuning range is roughly120% for the stop-band structure (for varying voltages between0 and 28 V) and over 305 for the band-pass filter (for varyingvoltages between 5 and 25 V).
5. Bright future: photonics for RF front-ends and opticalcommunication
Key challenges to satellite microwave technology are thedevelopment of higher power amplifiers, wider bandwidthdevices, flexible components such as tuneable local oscillatorsand tuneable filters, and reconfigurable antennas.
Photonic down-conversion can be a crucial step for realizing thewide bandwidth advantages of photonics. With the introduction of‘‘beyond Ka-band’’ frequencies, the opportunities arises for thephotonic implementation of RF oscillators. Here, the broad band-widths, photonic transport and some potentially advantageousphotonic architecture can provide enhanced system performance.One of the inherent strengths of electro-optic modulators is thebroad bandwidth of the process and much work has gone intoretaining that ability in the devices. However, thermal issues relatedto optical connectors shall be investigated.
Several types of applications must be considered, including thedistribution of reference frequencies at 10 MHz (Ultra-StableReference Oscillator) and around 1 GHz (Master Local Oscillator)as well as the distribution of microwave Local Oscillators atfrequencies exceeding 10 GHz. The distribution of high spectralpurity RF reference oscillators via an optical fiber network may findapplication for example in a telecom repeater architecture basedon a digital transparent processor that may require the distributionof two types of oscillators. Such repeater architecture performsfrequency down- and up-conversion, each in two consecutivesteps, and thus needs both the distribution of an USRO, forre-constructing local oscillators within some frequency-convertersand the distribution of MLO for direct use by other frequency-converters. The major requirement is that the optical network doesnot add phase noise degradation in the reference signal andguarantees a constant power level to each equipment. 10s to100s of equipment could be distributed by an optical networkdepending on phase noise specifications. Generally, mass savinghigher than 50% on sub-system is achieved for an extra-consump-tion depending on the laser transmitter and optical amplifiers.
LO signals are generated through the heterodyne multiplica-tion of two optical carriers. Such a process is particularly efficientfor high-frequency LO distribution in applications like broadbandtelecommunication payloads and remoting of multiple-beamantennas.
The compatibility of the LO optical block with frequencyconverters has been proven. Performances in terms of IF outputpower and isolation remain identical to the ones with the RFtechnology. LO distribution to a quite number of equipmentswithout any IF phase noise degradation may be provided.
Moreover, using emerging technologies like MOEMS (Micro-Opto-Electro-Mechanical System) based upon switch matriceswith optical distribution network, flexibility in frequency planscan be ensured.
The inherent benefits of microwave photonics including wideinput bandwidths, precision timing, low-loss and electromagneticinterference free signal transport, lightweight and flexible cabling
M. Yuceer / Progress in Aerospace Sciences 50 (2012) 27–34 33
and signal multiplexing allowing reduced cabling have driven thegrowth and prosperity of this field (Fig. 7).
Over the past decade, attention has focused on the creation ofopto-electronic oscillators [26,27]. Opto-electronic oscillatorshave several attributes that enable them to significantly outper-form their conventional electronic counterparts. An opto-electro-nic oscillator has an extremely high oscillator Q withoutdegradation of the phase noise at higher frequencies. As aconsequence, opto-electronic oscillators can generate spectrallypure RF signals. Recently, a 10 GHz signal with a phase noiseperformance of �163 dBc at 10 kHz offset from the carrier wasdemonstrated [28].
Another application of photonics is the inter-satellite or inter-orbit communication by optical links between the satellites. Dueto need of a real-time data transfer from a LEO (Low-Earth-Orbit)satellite to the ground station, a link via a GEO (Geostationary-Earth-Orbit) satellite can be utilized. Excessive rates of data andthe lack of interference issues are other drivers to choose anoptical link over an RF link. The lack of atmospheric turbulence orother preventive effects put Free-Space-Optical (FSO) systemahead of the RF counterpart for inter-satellite or inter-orbit links.
Fig. 8. Laser communication termina
Fig. 7. RF oscillator optical distribution network associated to a microwave LO
equipment (copyright@, TAS, France).
Internal structure and the architecture of a commercial lasercommunication terminal on market are shown in Fig. 8.
6. Spectrum design
In the EHF (extremely high frequency) domain, W band(75–110 GHz) offers large bandwidth availability for the futuresatellite communications. The push towards higher frequenciescharacterizes future research on the Q/V bands (31–60 GHz) andW-band (75–110 GHz). Preliminary experimentations of exploit-ing these bands can be hindered by some areas of uncertainty andrisk, which some of them can be eliminated by the use ofmicrowave photonics mentioned in Section 7. These uncertaintiesand risks are related to non-idealities of payloads (phase-noise,linear and nonlinear distortions, timing uncertainties, etc.) andlack of knowledge of signal propagation abnormalities especiallyin W-band. The analysis of requirements of an efficient W-bandcommunication link shall be started from some known issuessuch as link- budget and the feasibility shall be checked.
Issues other than link loss are the presence of nonlineardistortions due to the necessity of using amplifiers at theirmaximum level of power efficiency and presence of increasingsymbol imbalance and phase noise with the data rate.
The study in [23] analyzed the effects of nonlinear distortionand phase noise assuming the use rectangular pulses as digitalwaveform. However due to the limited bandwidth on the satellite,band-limited pulses shall be used in contrast with the unlimitedbandwidth rectangular pulse. A well-known example to the band-limited pulse is the raised-cosine. RSC is commonly employed insatellite applications in combination with QPSK or QAM modula-tions, implementing the so-called RSC-filtered QPSK or QAM.On the other hand, the disadvantages of RSC may impose theemployment of necessary countermeasures against nonlineardistortions. The usual solution considered is to sacrifice powerefficiency by introducing an input back-off at the border of thelinear AM/AM characteristic. In fact, power is an importantresource on satellite, and the need for efficient use of availablepower increases with frequency because of the technologicalchallenge for ‘‘beyond Ka band’’ power generation. Hence somemodulation types are investigated recently to overcome thesedisadvantages. One of them is to use prolate spheroidal wave
l (copyright@ TESAT, Germany).
M. Yuceer / Progress in Aerospace Sciences 50 (2012) 27–3434
function (PSWF) based modulation, which is already consideredin indoor UWB terrestrial communications [24].
7. Conclusions
Some of the latest developments in telecommunication satel-lite payloads have been reviewed. The distribution of highspectral purity signals and microwave LO over fiber optic assem-blies and harnesses could bring more than substantial improve-ments within future telecommunication payloads. In the mediumterm, it may constitute one of the enabling technologies allowingfor the practical implementation of advanced concepts such asmicrowave photonics, flexible repeaters and multi-beam activeantennas. Implementation on board depends on the degree ofmaturity of these techniques. Introduction of MEMS switches tothe space segment opens the door to reconfigurable antennas.
Current developments will determine the future trends,namely the widening of frequency bandwidth, spectrum reuse,multi-frequency feeds and the optimization of more integratedsubsystems.
References
[1] Van Fleteren S. Travelling wave tube vs. solid state amplifiers, [Online],Available: /http://www.djmelectronics.com/articles/twt-solid-state.htmlS.
[2] Katz A. Linearization: reducing distortion in power amplifiers. IEEE Micro-wave Magazine 2001;2(4):37–49.
[3] Qiu JX, Abe DK, et al. Linearization of TWTAs using predistortion techniques.IEEE Transactions on Electron Devices 2005;52(5):718–27.
[4] Chong CK, Davis JA, et al. Development of high-power Ka-band and Q-bandhelix TWTs. IEEE Transactions on Electron Devices 2005;52(5):653–9.
[5] Petillo J, Eppley K, et al. The MICHELLE three-dimensional electron gun andcollector modelling tool: theory and design. IEEE Transactions on PlasmaScience 2002;30(3):1238–64.
[6] Cooke SJ, Chang CL, et al. Three dimensional modelling of AC space charge forlarge-signal TWT simulation. IEEE Transactions on Electron Devices 2005;52(5):764–73.
[7] Chernin D, Antonsen TM, et al. A three-dimensional multifrequency largesignal model for helix travelling wave tubes. IEEE Transactions and ElectronDevices 2001;48(1):3–11.
[8] Smith C, Armstrong CM, Duthie J. The microwave power module: a versatileRF building block for high-power transmitters. Proceedings of the IEEE1999;87(5):717–37.
[9] Symons RS. Modern microwave power sources. IEEE Aerospace and ElectronicSystems Magazine 2002;17(1):19–26.
[10] Katz A, Gray R, Dorval R. Wide/multiband linearization of TWTAs usingpredistortion. IEEE Transactions on Electron Devices 2009;56(5):959–64.
[11] Bekey I. Advanced space system concepts and technologies. El Segundo: CAAerospace Press; 2003. [13–25].
[12] Gil I, Bonache J, Garcia-Garcia J, Martin F. Tunable metamaterial transmissionlines based on varactor-loaded split rings resonators. IEEE Transactions onMicrowave Theory and Techniques 2006;54:2665–74.
[13] Park S, Rebeiz GM. Low-loss two-pole tunable filters with three differentpredefined bandwidth characteristics. IEEE Transactions on MicrowaveTheory and Techniques 2008;56(5):1137–48.
[14] Zhao Q, Kang L, et al. Electrically tunable negative permeability metamater-ials based on nematic liquid crystals. Applied Physics Letters 2007;90.[011112–011112-3].
[15] Gil I, Martin F, et al. Tunable stop-band filter at Q-band based on RF-MEMSmetamaterials. Electronics Letters 2007;43:1153.
[16] Rebeiz G, Entesari K, Reines I, et al. Tuning in to RF MEMS. IEEE MicrowaveMagazine 2009;10(6):55–72.
[17] Park S-J, Reines I, Patel C, Rebeiz G. High-Q RF-MEMS 4–6 GHz tunableevanescent-mode cavity filter. IEEE Transactions on Microwave Theory andTechniques 2010;58(2):381–9.
[18] Joshi H, Sigmarsson HH, Moon S, et al. High-Q fully reconfigurable tunablebandpass filters. IEEE Transactions on Microwave Theory and Techniques2009;57(12):3525–33.
[19] Xiaoguang L, Katehi LPB, et.al. A 3.4–6.2 GHz continuously tunable electro-static MEMS resonator with quality factor of 460–530’’. In: Proceedingsof the IEEE MTT-S international microwave symposium digest; 2009.p. 1149–1152.
[20] Winter DW, Mansour RR. Tunable dielectic resonator band pass filter withembedded MEMS tuning elements. IEEE Transactions on Microwave Theoryand Techniques 2007;55(1):154–9.
[21] Amadjikpe AL, Chung DJ, et al. Two-pole digitally tunable evanescent-modewaveguide narrow-band filter with radio frequency micro-electromechanicalsystems switches. IET Microwave, Antennas and Propagation 2011;5(4).
[22] Velez A, Aznar F, et al. Tunable coplanar waveguide band-stop and band-passfilters based on open split ring resonators and open complementary split ringresonators. IET Microwave, Antennas and Propagation 2011;5(3):277.
[23] Sacchi C, Rossi T, Ruggieri M, Granelli F. Efficient waveform design for high-bit-rate W-band satellite transmissions. IEEE Transactions on Aerospace andElectronic Systems 2011;47(2):974–93.
[24] Velez A, Aznar F, et al. Tunable coplanar waveguide band-stop and band-passfilters based on open split ring resonators and open complementary split ringresonators. IET Microwave, Antennas and Propagation 2011;5(3):277–81.
[26] Yao XS, Maleki L,Eliyahu D. Progress in the opto-electronic oscillator—a tenyear anniversary review. In: Proceedings of the IEEE MTT-S internationalmicrowave symposium digest, vol. 1; 2004. p. 287–290.
[27] Savchenkov AA, Matsko AB, Ilchenko VS, Maleki L. Voltage controlledphotonic oscillator. Optics Letters 2010;35:1572–4.
[28] OEwaves, Jan. 2011, [Online], available: /www.oewaves.comS.