GaN Power Amplifiers Serving Satellite Industry on Multiple Levels

by Pasternack

Satellite technology has undergone many iterations of improvement over the decades; this includes the use of advanced frequency reuse schemes with the spot beam architecture, the leverage of higher frequencies (Ka-band) for larger contiguous spectrum, the implementation of digital payloads for flexible bandwidth and power allocation, and all-electric propulsion systems. These advancements have paved the way for high throughput satellite technology (HTS) offering orders of magnitude increases in throughput (100 Gbps) and finally, the potential of small satellite constellations in LEO/MEO. With that being said, the SATCOM technological evolution is continually squeezing the most out of their power amplifiers (PAs) in terms of efficiency, linearity, and size. The Gallium Nitride (GaN) substrate has been a clear contender for power amplification due to its inherently high power density, breakdown voltage, large bandwidth and high efficiency. This article attempts to illuminate the many ways GaN PAs are being implemented in the satellite industry, improving current technology and potentially future satellite platforms. 

GaN HPAs for Low- to Medium-Power Earth Stations and Satellite Transponders

For over a decade now, GaN PAs have been steadily replacing Traveling Wave Tube Amplifiers (TWTAs) in low- to medium-power (< 1kW) ground-segment (earth station uplink PAs) and space-segment (satellite transponder downlink PAs) applications where the size, weight, and reliability of TWTA cannot be justified when compared with solid state power amplifiers (SSPA). Even a 1% improvement in PA efficiency can greatly lessen the cost during launch and orbit in DC power savings not to mention the launch weight savings due to a smaller mass [1]. 

Previously, the Gallium Arsenide (GaAs) substrate was primarily leveraged for microwave applications, but GaN rapidly caught up and essentially replaced GaAs, particularly for RF PA applications. And, as more research and development is being put into GaN processes to improve cost, power density, thermal characteristics, and high frequency performance, the line at which to choose a TWT over a SSPA is beginning to get more blurred. While high-power transmissions are still dominated by TWTAs, there are several COTS GaN PAs that have output powers on the order of several hundred watts in the microwave L-, S-, and C-bands with smaller form factors, lighter weight, better linearity, and on par efficiency as compared with TWTAs. Bearing that in mind, high frequency applications beyond the X-band that require high output powers of 50W to 100W would likely leverage TWTAs as they can generally output far more power than GaN SSPAs in that spectrum (Figure 1). 

GaN for High Throughput Satellites (HTS) 

High throughput satellite (HTS) technology has minimized the cost associated with satellite Internet, with individual satellites offering hundreds of gigabits per second speeds. Table 1 illustrates a list of some of the HTS launched to date and their respective throughputs. 

The high throughputs are largely enabled by the use of the Ka-band and frequency reuse with the spot beam architecture—a method that leverages an array of small uplink and downlink beams in multiple spots. As illustrated in Figure 2, some HTS technologies utilize planar antennas with an array of radiating elements. Satellites such as the Eutelsat Quantum plan on using PAAs for simpler remote reconfigurability in order to adjust satellite capacity and coverage based upon changing customer needs. An estimated 50W to 100W is required for an antenna feed where the PAs are often shared between multiple feeds and multiple feeds are shared per satellite beam [4]. This requires a high power density PA for the HTS in orbit—a potentially expensive endeavor considering the cost of launch—multiple GaN SSPAs can potentially be cheaper and more efficient than TWTAs. 

GaN HPAs for Phased Array Radar in Space Surveillance and Tracking (SST) Systems

Phased array radar has allowed for significant progress in space situational awareness (SSA) applications where thousands of transmit/receive modules (TRMs) can actively monitor and track space debris—a looming problem for upcoming LEO/MEO satellite constellations. According to NASA, there are 21,000 objects of space debris with diameters larger than 10 cm orbiting the Earth and the number increases drastically as the diameters decrease. There are an estimated 500,000 particles between 1 cm and 10 cm and tens of millions of particles that are less than 1 cm in diameter.

Previous space surveillance and tracking (SST) technologies used massive reflectors with the bistatic radar technique, a passive radar that sends signal transmissions to various orbital spots and extrapoles location information based on echoes received by bouncing off of space junk [3]. Projects such as Lockheed Martin’s Space Fence (S-band), the German Experimental Space Surveillance and Tracking Radar (GESTRA) system (L-band), and LeoLabs Midland Space Radar (MSR) System (UHF-band) are already leveraging hundreds to thousands of TRMs to track thousands of pieces of debris per hour. 

Phased array antennas offer enhanced beam steering and beam forming capabilities with the ability to radiate multiple beams at many frequencies. The downside to this level of performance is that with every antenna element, there is a whole transmit and receive chain with its respective HPA and LNA, along with a power hungry phase shifter. This can create a scenario where much power is required in a relatively small space; GaN HPAs have already been implemented in some SSA PAAs including the Space Fence for their high power density qualities. The implementation of GaN HPA technology for PAAs can be seen on other fronts in the satellite industry as well, namely for the growing niche of small communications satellites.

GaN for Small to Medium Satellites 

Small satellite (smallsat) constellations are seemingly growing in popularity, with New Space companies such as OneWeb and SpaceX already receiving funding to launch constellations equipped with thousands of satellites. Smallsats have the potential to offer service all over the globe and are expected to reside in LEO, cutting latency down more than five times compared to HTS technologies in GEO. Smallsats may also have an edge in terms of production—smaller satellites can be more rapidly produced on an assembly line as opposed to the highly custom and proprietary satellite work done in the past. For instance, a collaboration between OneWeb and Airbus is expected to yield a first-generation of smallsats with a weight of 145 kilograms, an anticipated throughput of 10 Gbps, and an assembly rate of 15 satellites per week. 

Some stumbling blocks to the realization of this technology are the size/power constraints for the earth station and satellite as well as rapidly dodging space debris. Smallsats will naturally require their amplifiers, solar panels, battery bank, and all other equipment to be both small and light to minimize the cost of launch. While GaN does not offer the raw power capabilities that TWTAs have at high Ka-band frequencies, GaN HPAs do have the potential to be highly integratable, saving in space.

Ground terminals will also likely require seamless handoff between satellites; this is definitely a potential application for some type of smart flat panel antenna (FPA) with 2D scanning capabilities. And, as stated earlier, the GaN substrate has already been highly utilized for this type of antenna application with PAAs. 

Conclusion

Satellite technology is progressively drastically on two major fronts, with higher throughput large satellites and mass manufactured small satellites, both of which call for an optimized use of power resources. GaN processes have evolved so rapidly that current GaN HEMT technology can offer four times the power density in sizes over 80% smaller than GaN transistors fabricated in the early 2000s [5]. This brisk advancement in GaN processes can be attributed not only to the need for GaN SSPAs in the SATCOM arena, but also the need for GaN in next-generation cellular technologies with 5G. GaN-on-SiC PAs can be leveraged in PAAs for HTS, smallsats, and phased array radar in SST applications not only for their high breakdown voltage and excellent thermal properties, but also to minimize the size and weight associated with a satellite while maintaining a high level of efficiency to minimize the required prime power for transmissions. 

References:

1. https://www.erftm.eu/wp-content/uploads/2017/10/From-History-to-Future-of-Satellite-TWT-Amplifiers1.pdf

2. https://dspace.mit.edu/bitstream/handle/1721.1/110897/communication%20satellite%20power%20amplifiers.pdf?sequence=1

3. https://www.wirelessdesignmag.com/article/2018/04/phased-array-radar-provides-platform-next-generation-low-earth-orbiting-satellites

4. https://intranet.birmingham.ac.uk/eps/documents/public/emuw2/WM07.pdf

5. http://www.skyworksinc.com/downloads/press_room/published_articles/Compound_Semiconductor_052016.pdf

(1166)