Tendances futures du marché spatial

How digital transformation and disruptive innovation are challenging satellite power amplification technologies?

In recent years, the space industry has faced a disruptive breakthrough affecting its fundamentals. Digital transformation is transforming the satellite business model, creating new needs in consumer and enterprise connectivity at low cost, which implies very strong changes at all levels, involving the complete value chain. As an example, constellations or smallsats have become one of the major options for future satellite systems providing new capabilities related to high throughput connectivity demands.

Declining pricing and shifts in demand have combined with continuing ambiguity in the non-GEO constellation arena, which have impacted an already tenuous launch environment, and lack of Ex-Im funding, led to lower order rates linked to CAPEX holidays in 2015 – 2016.

The traditional GEO communications business is stagnating, calling for an operator response and new approaches : Smallsats are expanding the scope of the satellite industry. Not only are new smallsat operators expanding demand for satellite manufacturing, launch, and ancillary services, but the applications addressed and approach taken by smallsats is expanding the satellite addressable market itself. ⁽¹⁾

⁽¹⁾ Satellite Manufacturing & Launch Services, Northern Sky Research, 7th edition

Market Evolution

For decades, the in-orbit satellite communication market has been dominated by a traditional transponder leasing business model, with digital TV broadcasting as the main market, based on geostationary solutions (GEO). Satellites typically embark multi-spot payloads, with spot sizes covering zones like Western Europe, Eastern Asia or Continental US, or even larger areas. Most of them were working in C-band and Ku-band, and as TV broadcast was the main service, two-way capabilities were seldom required.
As far as RF amplifiers were concerned, the trend was to offer more and more powerful devices with increased efficiency, to cope with higher bandwidth. The main amplifier technology was - and still is - TWT (Traveling Wave Tubes) based, as the only technology capable of offering wide band (20% in Ku-band) high power, in the range of 100-150W together with high efficiency (above 65% at saturation). Of course, some amplifiers were used in back off for multicarrier operation (in VSAT for example), but most of them were used close to saturation with only a single 36MHz carrier, carrying ten to twenty TV channels in standard definition resolution.

Things started to change in the mid 2000’s with the launch of the first HTS (High Throughput Satellites) satellites such as ANIK F2, WILDBLUE and IPSTAR. With these satellites, the paradigm evolved towards increased data throughput, to compete with terrestrial internet networks where it made sense. The model is no longer to broadcast the same TV channels over a wide area, but to offer to each individual connected to the satellite a bandwidth of their own: 2Mbps, 10Mbps, 30Mbps and now 50Mbps at competitive prices. And of course a two-way solution is the rule. The only solution was to design satellites as very high throughput nodes, from 10Gbps in 2005 to 100Gbps in 2011, 300Gbps in 2017 and maybe 1Tbps in a near future.
The most efficient way to achieve this is to dramatically increase the number of spots while maintaining the same global coverage (which means dealing with smaller and smaller spots: 700km, 300km, 150km in a near future), coupled with an increased allocated bandwidth and an increased spectrum efficiency (i.e. number of Mbps per allocated MHz). The problem of bandwidth has been solved through the broad acceptance of Ka-band frequencies (between 17.2 and 21.2GHz), and the increase in the number of spots (hundreds compared to 5-10 in previous generations), which has allowed the introduction of disruptive satellite architectures.

New Architectures

HTS architecture is sharing the allocated bandwidth between forward (gateway to user) and return (user to gateway) links. To guarantee full coverage with spot beams, the antenna system is composed of four reflectors, with each dedicated to one quarter of the global bandwidth (i.e. one half of the analog bandwidth + one polarization).

But this architecture is hampered by some drawbacks, such as the difficulty to generate very small spots (100km spot size requires four five-meter reflectors), and a mediocre C/I prevents high spectrum efficiencies. For these reasons, disruptive architectures are popping up, such as phased array antennas (an array of feeders controlled by a beam forming network), which could either feed a single reflector, or radiate directly into space without an additional reflector (Direct Radiating Antenna). Also called MFB (Multi-Feed per Beam), these new architectures have a significant impact on RF power components.

Of course, these new concepts are well suited for (very) high throughput satellites, but they may be less interesting for more conventional payload missions: to addum multiple concepts, we pursue many specific requirements for power generation. To summarize, connectivity service is not new, and major connectivity services are based today on GEO satellites.

But to add to the complexity (and to investors due diligence), GEO solutions are now challenged by lower orbit constellations. Pros and cons are debated in terms of coverage and ROI, including the complexity and cost of the ground segment for users which will have to install multi-beam steerable antennas on their roofs, but at least Low Earth Orbit (LEO) constellations allow a dramatic reduction of the two-way delay of the signal, from 240ms in GEO to less than 20ms in LEO, fulfilling the need for satellite integration into 4G and 5G networks whose protocols require very low latencies (20ms in 4G).

A satellite constellation-based connectivity service that can deliver its capability through existing 4G/5G telecom infrastructure, as well as directly to mobile/cellular devices, is yet to be realized in the consumer segment for connectivity services. And the solution could be a LEO/MEO-based satellite constellation with in some cases inter-satellite links that can deliver uninterrupted connectivity services irrespective of which satellite the user is in connection with.

Ka-Band and Q-Band

For at least three decades, the satellite market was satisfied with two bands allocated to space communications by International and National regulatory bodies: C-band and Ku-band. But since the early 2000’s, in some locations (in particular over the US), Ku and C-band are saturated and new frequency band allocations have been required: Ka-band.

The particularity of this new band is its wide bandwidth (2,9GHz, not including the “military” band), which creates specific demands on amplifiers. Ka-band regulation is also very complex, only one part is allocated to user beams worldwide without risk of interference. The remaining Ka-band allocation depends on local regulators and may be shared with other applications; it is primarily used for gateway connections. But the pressure for new frequencies has made the allocation of a larger part of the Ka-band to user beams, at the expense of the gateways which will need to find available frequencies elsewhere: thus Q-band now emerges.

RF Power Requirements

RF power requirements are a consequence of these trends: yesterday, when only C and Ku-band were used, amplifiers often worked close to saturation (where they benefit from higher efficiency), with limited bandwidth (36MHz) and medium power (60-80W for C-band, 100-140W for Ku-band). But things have changed dramatically: Ka-band is now fully used and becomes a significant part of the market, Q-band emerged, and there are emerging needs in L and S-band, in particular for navigation and radio broadcasting.

From an amplifier saturation power point of view, trends are contradictory: on one side, amplifying larger bandwidths, and working in back-off to guarantee a good linearity in multicarrier amplification, underlines the need for higher saturation power. On the other side, smaller and smaller spot sizes generate less RF power required per spot. As a result, the power requirements lie between 30W and 300W at saturation, from L to Ka-band. Knowing that amplifier technologies (TWT or SSPA) have a limited power flexibility range (2 to 3dB), this 10dB gap should be addressed by at least 4 to 5 different devices, for each frequency.

Efficiency and Bandwidth

With new architectures, things changed for many reasons: first the allocated bandwidth is increasing in HTS satellites, or is split in sub-bands spread over 2.5GHz or even 2.9GHz in Ka-band. Second, amplifiers are working in multicarrier mode and the instantaneous bandwidth must be large. This constraint is increased further when new architectures such as multiport amplifiers or phased arrays are used. Today, TWTAs are designed to be capable of handling the full bandwidth (2GHz in Ku-band, 2.9GHz in Ka-band).

Efficiency is the most stringent requirement that power amplifiers must comply with. It is easy to understand why, as roughly 80% of the regular payload DC power is allocated to power amplifiers (this percentage is less when a digital processor is embarked, but the consequence is to put more pressure on electrical efficiency). One has to keep in mind that improving efficiency not only reduces the required primary power from the solar cells; it also and maybe more significantly reduces the dissipation constraints of a platform, which has only one way to evacuate excess heat: thermal radiation into space.

The EPC efficiency is typically between 93 and 95%, so the main constraint rests on the shoulders of the amplifier device itself. For many years (and still today) TWT efficiency was specified at saturation level, knowing that many of them were actually used close to this configuration. Great progress has been made during the last decades for TWTs, whose saturation efficiency increased from 50% in the 1980s to the current 70%. Room for improvement is still in front of us (some devices has already being measured at 75%), and 80% should be achievable.

But efficiency at saturation doesn’t tell the full story, as more and more amplifiers are used in multicarrier configuration. It is at this new operating point, i.e. in back-off, that the efficiency must now be optimized (back-off operation is required in multi-carrier operation to guarantee the required level of linearity). Of course, the first requirement is to keep the back off as low as possible, so most of amplifiers used in space are linearized.

For many years, efficiency improvement at output back-off (OBO) followed the improvement of efficiency at saturation, but with new TWT multi-stage collector designs, improving efficiency at 2 to 4dB OBO are now possible, and very significant gains have been demonstrated by Thales in the last few years, with close to 8% more in efficiency (from 42 to 50%) at 3dB OBO. This has a tremendous impact on the global efficiency of HTS payloads.

Flexibility

Flexibility is the new paradigm for satellite payload designers; not so much a need for TV broadcasting, it becomes more and more required for data communications to optimise the satellite felling rate and return on investment. Flexibility means spectrum, bandwidth and coverage flexibility. Spectrum flexibility has been made possible by the availability of wide-band amplifiers. In the 2000s, most of the Ku-band and Ka-band TWT amplifiers were designed and tuned to handle a limited portion of the available spectrum only: between 0.5 and 1GHz typically. Since then, wideband devices with flatter responses have been offered to the market, with 2GHz (in Ku-band) and 2.9GHz (in Ka-band) now available.
	Bandwidth flexibility is the capacity for the satellite operator to adapt the digital throughput per spot to the actual user demand. It requires the amplifier to be capable of modifying the input power with only a minor impact on the amplifier efficiency. Two technologies are available today, depending on the level of output power modulation required. If it is less than 3dB, a flexible TWTA can be one solution, where the TWT saturation point is remotely tuned from the ground by the operator. For higher flexibility, MPA (Multiport Amplifiers) are proposed on the market, using wideband TWTs operated in back-off.
Coverage flexibility is a complex topic, requiring advanced antenna architectures, together with digital beam forming solutions. These technologies are now proposed by satellites manufacturers. As far as amplifiers are concerned, it requires the availability of a wider range of performance, in particular in power (from 10W to 300W at saturation), in wide band, with an increased pressure on integration to match the pitch between antenna feeds. TWTA are part of the answer, as well as SSPA for the lower power part. Of course, electrical efficiency remains the main driver to be handled by designers.

Thales MIS Strategy:
New Bands, Efficiency, SSPA

Thales Microwave & Imaging Sub-Systems (MIS) strategy can be summarized as being a major supplier of reliable RF power amplification solutions for space, in partnership with other key electronics suppliers. Proposing wideband amplifiers is a target already fulfilled, and improving efficiency remains the key driver of our road-map whatever the technology and the frequency band.

Thales MIS now offers solutions from 100W to 300W in L-band, 100W to 500W in S-band, 40W to 150W in C-band, 40W to 220W in Ku-band, 40W to 170W in Ka-band, and up to 45W in Q-band. X-band is also proposed. Investing in higher power TWTs is still one of our drivers, both with conduction and radiation cooled devices. Higher power components are foreseen on the market in the following years, in particular in Ku, Ka and Q-band.

The shift to new payload architectures for the HTS satellites and constellations generates new requirements, in particular towards lower power.

TWTs better match the higher power segment and do not propose attractive solutions if the power requirements are less than 20W. Above that value, the tradeoff is less clear, but and SSPA, with the recent availability of Gallium Nitride (GaN) HEMT technologies, will have an increased impact on payload designs. Compared to GaAs pHEMT, GaN HEMT is capable of higher power generation per millimeter of grid width: a factor above 5 is generally agreed. Thanks to this property, fewer transistors must be combined to reach a required level of output power, with an immediate positive impact on the electrical efficiency (between 5 and 10%, depending on the class of operation).

Even if GaN efficiency is still significantly lower than TWT’s, this drawback may be manageable for low power devices, and for this reason, MIS decided to add SSPA to its portfolio.

This strategy is supported by strategic partnerships, including with the European GaN foundry UMS, a JV between Thales and AIRBUS dedicated to III-V component design and foundry.

Of course, the ground segment is also part of the game, and ground-based TWTs are required in complement to the on-board solutions for gateways. Thales MIS already proposes solutions in Ku, DBS and Ka-band, and new products are planned in V-band in the near future, with the same target of proposing both high power and wideband devices.

In a changing market with an increased portfolio of technologies (TWT and SSPA), the objective of Thales MIS remains offering to the market competitive solutions with very attractive performances, in particular at bandwidth, linearity and efficiency levels.

About Thales Microwave & Imaging Sub-Systems

Thales is the world’s leading supplier of Traveling Wave Tubes (TWTs) for satellites and other spacecraft.
Our TWTs are used on satellites spanning the full range of civil and military applications including space-based TV and radio broadcasting, data transmission, telecommunications, internet, observation and navigation. Thales Microwave & Imaging Sub-Systems has two world class production facilities dedicated to space TWTs.

Located in Velizy, France and Ulm, Germany, these facilities are certified to both ISO 9001, AS 9100, AQAP2110, ISO 14001, OSHAS 18001. Thales offers a cumulative production capacity from these two facilities of more than 150 space TWTs per month.

Each facility has the highly trained personnel and specialized equipment needed to assure on-time delivery of our state-of-the-art space Traveling Wave Tubes. Our TWTs set global standards for reliability and performance in space-borne applications.