Ryan Jennings
©SHUTTERSTOCK.COM/PHOTOCREO MICHAL BEDNAREK
The large investments in new satellite communications (SATCOM) constellations are driving the demand for flat panel antennas (FPAs). We are already seeing the growing adoption of the technology in commercial and military applications. As more of the new constellations become operational, this adoption is projected to explode. There are a wide range of SATCOM terminal needs across the different market segments that are also dependent on what constellation(s) they might be using. The major markets include consumer, commercial mobility, enterprise, and government. While each of these markets have unique requirements—or limitations—on size, weight, power draw, throughput performance, cost, and ruggedness, there is some overlap, as shown in Figure 1. Generally, the consumer market can be separated from the other applications, given their unique requirements. While these consumer terminals can meet some needs in all markets, they are very limited outside of their primary use. The other three markets have significantly more overlap in requirements, with ruggedness being a key differentiator from consumer terminals. With that differentiator, prices are higher, most applications can sustain higher power draw, and many demand higher throughput performances. This does not mean that each segment may have some classes of terminals with unique requirements that do not overlap any other.
Figure 1. Primary markets have overlapping requirements for terminals dominated by enterprise, government, and commercial mobility, with consumer terminals nearly standing on their own (courtesy of Anokiwave).
Next, looking at the constellations on which these terminals might operate will also have different performance capabilities and requirements put on the terminals. These orbits include the traditional geostationary Earth orbit (GEO) and the short-cut term, nongeostationary orbit (NGSO), as shown in Figure 2. The NGSOs include low Earth orbit (LEO), medium Earth orbit (MEO), and highly elliptical orbit (HEO), all with unique terminal coverage requirements. The LEO megaconstellations are projected to provide 100 times increase in bandwidth from the legacy GEO satellites and reduce latency by 10 times, while providing this to significantly more users. MEO satellites provide fewer capabilities than LEO, including less Earth coverage, but significantly more than current GEOs.
Figure 2. The diverse requirements of GEO, low Earth orbit (LEO), medium Earth orbit (MEO) and highly elliptical orbit (HEO) networks have significant impacts on terminal performance requirements, especially when there is a need to service multiples with the same terminal (Adobe Stock).
Many are now looking for terminals that can support some or all these orbits to maximize existing capacity, while taking advantage of additional capacity and coverage. Several operators are beginning to build their own multiorbit constellations to provide increased coverage from only traditional GEO operations. Inmarsat is an example of a traditional GEO provider that is branching out with their Orchestra network, that in the near term adds HEO satellites for northern latitude coverage but includes longer-term plans for LEO and even integrating terrestrial 5G into this network [1]. Intelsat has also announced a global distribution partnership to offer inflight connectivity (IFC) to combine GEO high-throughput satellites (HTS) with the OneWeb LEO network to provide the best capabilities from each network [2].
This overwhelming variety in requirements makes it very difficult to develop the perfect flat panel SATCOM terminal. However, FPA-based terminals are the best technologies to address these markets, which are the wave of the future. The convergence for volume demands, technical advances, and thus price are making them real today. The market is anticipated to grow from US${\$}$408 million in 2022 to US$1,440 million by 2027; it is projected to register a compound annual growth rate of 28.7% from 2022 to 2027 [3].
Consumer terminals are a large part of the projected growth in FPAs and to date are solely focused on specific LEO constellations. They are typically targeted to an extremely low price, low power, and moderate throughput performances. This is achieved through the implementation of half-duplex instead of full duplex, fixed sizes to address one market’s demand, designs limited to one specific constellation, and consumer-grade ruggedness. The other markets often include a GEO constellation that impacts the terminal requirements. Typically, GEO architectures require larger scan angle and higher antenna performance than the NGSO. With GEO-only constellations, the scan angle requirements can be difficult to achieve, especially in networks with a limited number of satellites. Figure 3 shows an example of how with three satellites to cover the globe, regardless of the performance of the antennas on the satellite, the terminals cannot scan far enough to cover the northern latitudes. Land mass coverage can be improved with satellite location and/or increasing the number of satellites; however, this “scalloping” of no coverage will always occur with FPAs due to the physics of maximum scan angle. But, with the addition of other orbits like LEO and HEO, this coverage can be addressed.
Figure 3. GEO-only constellations are not well suited for FPAs and can lead to significant higher latitude coverage loss due to the physics of scan capabilities (courtesy of Anokiwave).
The increases in receive and transmit antenna performance for GEO operation results in larger terminals that draw more power and require additional cooling capabilities. Ku GEO satellites also have a unique dual linear polarization, while Ka GEO and all LEO constellations use circular polarization. Many of the GEOs have some reuse of polarization requiring more stringent cross-polarization control. This drives the technical implementation of FPAs. As an example, active electronically steered antennas (AESA) require more front-end electronics to correct polarization performance across the entire scan volume, since the vertical and linear element performance varies with scan angle, thus requiring phase and amplitude correction to each of these feeds to maintain polarization isolation.
FPAs have technical challenges that impact their design and applications. As previously mentioned, scan performance is an important design criterion. As the antenna is scanned from boresight, the gain will reduce since the capture is reduced. Also, the design of the element is important for gain, as the performance changes based on mutual coupling as the array is scanned. These two sources of loss can be seen in the array gain equation, where N is the number of elements, Ge is the embedded element gain, Lohmic is the ohmic loss, and Lscan is the scan loss: \[{\text{Array gain}} = {10}\,{*}\,{\log}{(}{\text{N}}{)} + {\text{G}}_{\text{e}}{-}{\text{L}}_{\text{ohmic}}{-}{\text{L}}_{\text{scan}}{.}\]
The ohmic loss will not change with scan, so the scan loss is a combination of the embedded element gain and scan loss. This nonideal isotropic behavior of the embedded element gain and reduce capture area can be described in this equation where ${\theta}$ is the scan angle and x is a numeric value, typically in the 1.1 to 1.5 range: \[{\text{Scan loss}} = {10}\,{\ast}\,{\log}{(}{\cos}^{\text{N}}{(}\theta{)}{)}{.}\]
The antenna performance is critical to an efficient and effective FPA. Figure 4 demonstrates the additional loss that can be seen from a power cosine factor, where a 75°scan can have up to 2.3 dB. This loss at cos1.5 would result in the need for ∼60% increase in aperture size to achieve similar performance for a cos1.1 design.
Figure 4. Scan loss for 1.1 to 1.5 cosine factors demonstrating the importance of antenna performance.
There are a variety of technologies that are covered by the term FPA, but not all are equal. This includes some mechanically pointed technologies, like Thinkom’s VICTS, passive technologies like Kymeta’s nTenna®, and ESA like Starlink’s “Dishy.” AESA technologies are projected to dominate the market given the technical limitations with mechanical and passive technologies.
Mechanically steered antenna technologies have dominated the SATCOM terminal market for decades; however, most would never be considered an FPA. The existing mechanically steered FPAs, along with their legacy counterparts, require two antennas to achieve the agility needed for switching between satellites within a LEO or MEO constellation and switching between orbits. The need for two antennas drives the size, weight, and power required with this technology and, in turn, drastically limits the markets that can utilize it.
Passive antennas have projected wide market adoption, but that has not been realized to date. While they have the potential for lower cost, they also have deficiencies with efficiency and agility. These approaches typically use RF lossy metamaterials as a “lens” or phase shifter that can control the phase with electrical stimulation. Additionally, there have been concerns about the speed at which these materials can change state and managing those variations over temperature. This technology has been focused on Ku-band solutions, as its limitations only become more difficult at the Ka-band. There will be niche applications where this technology can meet some market needs for fixed and mobile applications; however, the efficiency and agility limitations have made it difficult for broad acceptance. The limitations of these technologies make them suboptimal for LEO and multiorbit solutions.
This is where the AESA antennas come in to save the day. Their ability to point near instantaneously within or between orbits using a single beam with a single antenna is what has made the NGSO constellations viable. The LEO market, multiorbit constellations, and desire of more agnostic terminals is driving market needs for AESA solutions. Historically, the combination of cost and technical performance has been an inhibitor for the widespread adoption of AESAs. They have been used for decades by high-end military systems that could afford their price. However, we have seen significant advances with technology that is enabling lower prices.
Two key aspects of AESAs that have driven cost are the RF electronic devices and how the radiating elements are packaged with these devices. Figure 5 shows a traditional defense AESA design, often for high-performance radar systems, using gallium arsenide (GaAs) and/or gallium nitride (GaN) RF-integrated circuits (ICs), often in hermetic packages with many RF connectors required to interconnect them with the antenna elements and with each other.
Figure 5. Traditional AESAs have roots in defense radar applications that do not meet the cost points of SATCOM terminals, which has brought rise to new, cost-effective techniques and technologies (Adobe Stock).
In multithousand element arrays, the cost of GaAs and GaN per square millimeter, along with high connector count drives to unaffordable solutions for communications terminals, much less commercial applications. With the latest ICs and printed circuit board (PCB) material/manufacturing capabilities, lower cost becomes achievable and, more importantly, mass production possible. Newer architectures based on a commercially viable multilayer PCBs with radiating elements on one side and surface mounted ICs on the other have shown to be cost and performance viable, since they can all be manufactured using existing technologies that are in place to build your cell phone and your Wi-Fi access point. The Starlink “Dishy” antenna is a great example of this low-cost architecture, as shown in Figure 6.
Figure 6. Starlink’s “Dishy” antenna is a great example of a PCB-based AESA architecture that demonstrates a cost structure for the masses (courtesy of Branch Education).
In addition to the cost-effective PCB and assembly technologies, we now see high-performance silicon-based RF ICs in the market. This technology allows for highly integrated circuits where power, control, and RF capabilities are in the same, compact IC. Silicon technologies get their price advantages over traditional GaAs and GaN through lower material costs, larger wafers providing more parts per fabrication cycle, and high yields. Most AESA architectures use analog beamforming technologies where phase and amplitude are managed in the RF domain. There are some solution sets that partially or fully implement this in the digital domain; however, these solutions result in higher cost and power unless they can be custom, single-constellation implementations. There are a number of companies making commercially available analog beamforming ICs (BFICs), including Analog Devices, Anokiwave, Sivers Semiconductors, and Renesas. Most of these BFICs are designed to support each band as well as transmit in receive in different ICs. Most are architected where a single IC supports the dual polarization feeds of four antenna elements. Some require external low-noise amplifiers or power amplifiers to meet the performance needed, which can drive additional part and/or integration costs. Si devices use a single low-voltage supply with integrated logic control, making integration into large, phased array antenna terminals within the lattice straightforward, which is very difficult to achieve with GaAs or GaN device technologies.
LEO constellations are driving demand in the market. After the failures of LEO constellations in the mid-1990s with Iridium, Globalstar, and Teledesic, many have been skeptical of their success now. Lower-cost satellites and launch costs have made it possible for SpaceX and OneWeb to have nearly global coverage with their Ku-band constellations. Amazon and Telesat will soon begin to deploy their Ka-band constellations, with several others in the early planning stages. The projected and demonstrated capabilities with these constellations has many markets salivating over what they can provide to their customers. Besides the home consumer market, IFC appears to be the next big adopter of this capability. The LEOs are opening new markets that have not been accessible before, like commercial and business aviation platforms that are smaller than narrow-body jets.
The increases in data communications across all parts of our society is a consumer behavior driving the need for the LEOs and their increased capacity to multiple markets, as shown in Figure 7. The people of the world have the desire to connect with each other, and to information and entertainment everywhere they go. Approximately 40% of the world’s population does not have access to efficient Internet connections [4]. Only a small portion of that demand is met through direct satellite connections, but more often use the connection as a backhaul.
Figure 7. SATCOM integration with 5G and private networks is highly complex, with many insertions and applications (courtesy of Anokiwave).
The previously mentioned IFC application is a good example of using the satellite’s large data throughput to service multiple users simultaneously. The next big application is for 5G backhaul, which is well suited given the low latency (20 times improvement over GEO) [5] and throughput capabilities, enabling better connectivity options in rural areas. This is a great complement to, and many believe alternative to, ground-based fiber connections. Many more applications could be envisioned, from smart cars to global shipping; all are driving volume demands, which ultimately drive the price of FPAs down. The integrated data roadmap across all of these sectors is highly complex.
The consumer market is driving their terminals to be very different from the rest of the markets. Many of these providers are developing or already producing their LEO-only ESAs to support the consumer terminal market—including Starlink, Hughes (OneWeb), and Amazon—that are specific to their constellations. The latest version of the Starlink home Internet terminal is shown in Figure 8. They will not interoperate with any other constellation and thus do not have the ability to support existing multiorbit architectures. Another unique aspect of the consumer terminals is typically operating in half-duplex mode to minimize overall cost, and they don’t require the higher throughput that full-duplex solutions provide. Their driving performance requirements are different than the traditional GEO satellites. LEOs all operate at circular polarization, have less-stringent cross-polarization isolation, typically require less scan volume, and have lower transmit power and receive sensitivity. Their AEAS designs provide the agility needed for microsecond switching from satellite to satellite that can be horizon-to-horizon handoff every 3–10 min. Before these networks were deployed, many thought dual beams were required to support a make-before-break connection. However, we now see the dual beam architecture is not needed and can be managed within the network. Multiple tests, demonstrations, and now operational consumer terminals have validated single-beam architectures. This comes with a significant savings to the terminals, where it is not required to integrate a second analog beamforming network or move to more complicated and more costly digital beamforming. These savings are not only in the devices, but also in the complexity of the PCB and result in lower power draw.
Figure 8. Starlink’s “Dishy” antenna is a great example of the home consumer terminals, being the first to market, and they are already on their third revision (Adobe Stock).
The other markets have more convergence in requirements, but still have varied needs. The Department of Defense (DoD) is a great place to start the discussion, as they cover airborne, maritime, land-mobile, and deployable terminals that overlap enterprise and commercial mobility requirements. One thing all of these markets have in common is the need for more rugged construction than the consumer terminals. Again, this makes the one perfect FPA solution difficult to achieve for these markets and applications, where some solutions are for LEO only, MEO only, GEO only, and any combination thereof with a variety of throughput requirements.
The DoD is interested in the many opportunities for government-owned and commercial-integrated SATCOM capabilities at GEO, MEO, and LEO to deliver diversity to their portfolio, ensuring reliable communications. National security is highly dependent on space access, so the vulnerabilities and resiliency can be address with a diversity in constellation and orbit. This drives a need for a terminal that is agile to support the different constellation requirements and all orbits.
We also see the commercial markets looking for multiorbit capabilities, with aviation in the forefront. The ability to use GEO in airline hub-cities for capacity and across the oceans is a great complement to the capacity of LEO and their polar coverage. While we currently see this coming in the Ku-band, as Ka LEOs and HEOs are launched, this will become common in those frequencies as well. Intelsat recently announced their AESA IFC product (shown in Figure 9) that will operate over their current GEO network as well as with the OneWeb LEO network. This is a great example of the need for flexibility within these markets.
Figure 9. Intelsat’s multiorbit airborne IFC solution is based on AESA FPA technology (courtesy of Intelsat).
There have been many approaches to meet the varied needs of these applications. One that seems to have good traction is using modular solutions that can be scaled up or down to meet the needs of a variety of applications, as shown in Figure 10, from Ball Aerospace. Since the antenna is the most expensive and complicated component for a SATCOM terminal, one would not want to design many different sized antenna boards due to the investment cost, but also due to manufacturing costs. This is where a modular approach can provide the best value where a single manufacturing line can be set up to support the antenna building-block fabrication, maximizing volumes of a smaller quantity of sku numbers. This modular approach allows companies to drive volumes of common parts. Ball Aerospace is an example company that is implementing this concept, with separate Ku and Ka building blocks in both transmit and receive arrays.
Figure 10. Ball Aerospace’s modular AESA technology allows common antenna subcomponents of the terminal to be combined to achieve the desired terminal performance (courtesy of Ball Aerospace).
Unfortunately, there is no perfect terminal solution for all market areas. For all applications, the historical pricing has limited widespread adoption of flat panel AESAs. This is being addressed through multiple aspects. First is implementing commonality to achieve needed volumes in manufacturing, which brings down the cost. PCB cost can be reduced by minimizing the number of layers and drill cycles while leveraging new circuit board materials that have good high-frequency RF performance. Traditional GaAs and GaN technologies are nonstarters, but leveraging commercially available, Si-based BFICs provide the needed performance at an affordable price. Finally, implementing a modular approach appears to be a winning architecture to achieve the best price, as this can drive the highest volumes through addressing the most markets. It is an exciting time to be in the satellite communication industry, with the explosive growth in LEO constellations driving the demand for FPAs and the need for the perfect solution. In 2020 FPA sales were ∼5,000 and are projected to be 100 times that by 2026, at greater than 500,000. This is a market worth over US$100 million, with lots of opportunities for innovative solutions.
[1] “World’s most advanced commercial communications satellite begins electrically-powered journey to geostationary orbit.” Inmarsat.com. Accessed: Jan. 10, 2023. [Online] . Available: https://www.inmarsat.com/en/news/latest-news/corporate/2022/worlds-most-advanced-commercial-communications-satellite-begins-geostationary-journey.html
[2] “Intelsat and OneWeb partnership brings multi-orbit connectivity to airlines worldwide.” Intelsat.com. Accessed: Jan. 10, 2023. [Online] . Available: https://www.intelsat.com/newsroom/intelsat-and-oneweb-partnership-brings-multi-orbit-connectivity-to-airlines-worldwide/
[3] “Global flat panel antenna market (2022 to 2027) - Demand for high data rate transmission presents opportunities.” GlobeNewsWire.com. Accessed: Jan. 10, 2023. [Online] . Available: https://www.globenewswire.com/news-release/2022/03/28/2410810/28124/en/Global-Flat-Panel-Antenna-Market-2022-to-2027-Demand-for-High-Data-Rate-Transmission-Presents-Opportunities.html
[4] I. Ivanov, “Optic fiber and LEO satellites – Competition or convergence?” Mission Crit. Mag. Accessed: Jan. 10, 2023. [Online] . Available: https://www.missioncriticalmagazine.com/articles/94053-optic-fiber-and-leo-satellites-competition-or-convergence
[5] F. Rayal. “Latency in LEO satellites vs. terrestrial fiber.” Frank.rayal.com. [Online] . Available: https://frankrayal.com/2021/07/07/latency-in-leo-satellites-vs-terrestrial-fiber/
Digital Object Identifier 10.1109/MMM.2023.3242538