Lessons Learned from NSR’s Non-GEO Constellations Analysis Toolkit

Authors: Christopher Baugh | Carlos Placido (independent adviser)

Mega-constellations are taking center stage in a pivotal time for high-speed satcom, but business-case and architectural facets will drive the multi-orbit, multi-band satellite paradigm. It is thus worth diving into the facts using NSR’s Non-GEO Constellations Analysis Toolkit 2.0 (NCAT2) to infer how LEOs may best interplay with well-established GEOs.

Today’s fast-paced environment brings unprecedented opportunities for the telecommunications industry as a whole to make the most out of innovations and investments in space networks, both Non-GEO and GEO. This article analyzes key aspects impacting the LEO-GEO complementation potential.

Recap of LEO Developments

Despite concerns on sustainability because of their capital-intensive nature, LEO satellite constellations are quickly becoming a reality. Whether FSS-band or Space-Mobile LEO architectures ultimately succeed in driving SATCOM from niche to mainstream remains to be seen, but 2021 will likely be remembered as the year when LEO high-throughput constellations materialized.

A wide range of developments took place over the past 18 months, but SpaceX undoubtedly led the race with completion of its first Starlink sub-constellation of approx. 1,600 satellites, expected to soon come out of trial mode and begin commercial operation in various countries. China’s mega-constellation filings at the ITU level, OneWeb’s launch progress and Telesat’s IPO, are also among such major developments.

SpaceX’s success in driving down the cost via reusable rockets and collapsing of the supply chain (satellite design, manufacture, launch and network operation) has clearly driven the industry to a new level. The LEO revolution was seeded many years ago as it was in the late 1990’s when Iridium and Globalstar pioneered LEO satcom at narrowband speeds; but it was not until 2013 that O3B became the trailblazer of Non-GEO high-throughput, by leveraging FSS bands. Satcom’s current fast-paced transformation is also linked to the global proliferation of GEO-HTS initiatives, via both integrated models (i.e. Viasat, Hughes, Thaicom, Inmarsat, etc.) and open architectures (i.e. Intelsat, SES, etc.). These developments led to today’s environment where LEO-HTS poses a cannibalization threat to traditional players but also complementation opportunities.

The LEO Architecture “Blessing and Curse”

An often-overlooked aspect of LEO satellite constellations is that, while constellations deploy terabits of capacity, satellites’ low altitude limits their visibility, so constellation players are unable to steer most of the globally deployed capacity to the areas with the largest target population. Simply put: The lower the altitude, the smaller the satellite’s field of view (FoV) of the Earth’s surface, which –consequently– limits the opportunities to redirect and use satellite capacity.



Scaling back the LEO satellite power to what will be needed may drive design efficiencies at a CAPEX level (lower weight and launch cost). If this is supplemented by a good choice of orbits, frequency band and antenna elevation, well-designed LEO constellations may ultimately achieve decent global fill rates and power balances but –regardless- low altitude inevitably limits addressability.

Orbital inclination is another design factor affecting both coverage (latitude reach) and supply density. As assessed in a recent Bottom Line+ article using hard data and NCAT2 analytics, over 75% of the globally addressable rural population for LEOs resides within 35 degrees of latitude north and south of the Equator. However, leading LEO-HTS operators deploy the highest capacity density (Gbps proportionally overhead) above 35 degrees N/S to be able to serve high-ARPU regions in the northern hemisphere.



Therefore, constellations’ low altitude can be considered a blessing for latency (low) and for spectrum reuse (high) but also a curse for addressability. The addressability limitation of low altitude is the main reason LEO players must target not just fixed data applications (i.e. consumer/enterprise broadband, backhaul, etc) but also energy/mining distant locations and mobility routes (maritime, aero, rail, and eventually “connected cars”). Maximizing addressability is vital for constellation players to drive the lowest possible marginal cost per (usable) Mbps in order to be competitive.

Indeed, three aspects simultaneously collude against LEO-HTS economics between +- 30 or 35 degrees of latitude:

1. Supply-Demand: Higher density of target population coincides with lower density of Gbps deployed overhead;
    

1. Interference: More instances of potential interference with GEOs reduce the degrees of freedom;

1. ARPU: Generally low telecom ARPU pushes for hybrid solutions.

Aspect 1 was analyzed in the previous BL+ article, showcasing irregular supply-demand dynamics around Equatorial regions as in the displayed heatmap; but let us now focus on diving into aspects 2 and 3 using other NCAT2 tools.

Interference Analysis (Aspect 2): Exclusion Angle Analysis

LEO satellites must not interfere with GEO services, which always have coordination priority. To accomplish this, LEOs need to comply with the equivalent power flux density (EPFD) limits defined in the ITU Radio Regulations. In practice, this means that LEOs need to establish exclusion zones (minimum discrimination angles) with respect to the GEO “Clarke Belt” arc. Given the dynamic nature of LEO constellations, operators’ orchestration systems must trigger satellite handoffs whenever the EPFD limits are about to be exceeded due to the angle differential with GEO satellite transmissions. Below is a sample simulation of exclusion angles for the Ku-band (user beam) frequencies used by the SpaceX system when Starlink satellites (shell 1, 550 km altitude) fly over the Equator.



NCAT2 Tool #11 calculates uplink and downlink discrimination angles for NGSO systems based on the Equivalent Power Flux Density (EPFD) limits of Article 22 of the ITU Radio Regulations, applicable to the FSS frequency bands.



Non-interfering operation is vital for the co-existence with the (populated) Ku-band GEO arc. This sample analysis for Starlink (using technical data extracted from FCC filings) concludes that the usefulness of satellites’ field of view is reduced substantially when satellites fly over the Equator. Indeed, up to 50% of the satellites’ FoV could be reduced (wasted) for Starlink satellites in Equatorial regions so as to avoid interference with Ku-band GEO satellites. While this does not necessarily result in wasted capacity when satellites have steerable beams (capacity could be directed elsewhere), it certainly results in fewer “degrees of freedom” for capacity relocation, a problem that worsens with lower attitudes and higher minimum-elevation angle definitions for the ground antenna. Other visible satellites with higher discrimination angles that are operating nearby the area can take turns in dynamically supplying capacity to specific spots but possibly at low look angles, which generally affect link IP throughput.

ARPU Considerations (Aspect 3): Direct vs Hybrid Distribution

Constellations’ business-case challenges are compounded by the fact that low-ARPU markets make it harder to justify –from a business case standpoint– the cost of MSA/FPA LEO terminals, which are inherently more complex and expensive than GEO VSATs. As a general rule, the lower the link data rate and service recurring rate, the more important to use inexpensive user terminals. Thus, depending on the application, a hybrid distribution approach may provide the lowest TCO for LEOs.

NSR emphasizes the importance of hybrid networks and distribution partnerships for Non-GEO, and a recent burst of announcements by major LEO players point at addressing the right pain areas. Announcements include:

• Distribution Partnerships: Verizon-Amazon Kuiper (U.S.) and NEOM-OneWeb (Middle East)

• Hybrid Distribution: The Akiak Native Community in southwestern Alaska becoming the first US community served by OneWeb via a hybrid satellite-wireless network.

The OneWeb deployment in Akiak is a prime example of the value of hybrid distribution. OneWeb worked with Pacific Dataport and Microcom to install the first OneWeb service in Alaska, making Akiak the first LEO-enabled village in the US. A single LEO terminal communicates with the satellites overhead and connects to the community-wide wireless distribution system, using Cambium technology.



The business-case crossing points for the hybrid interplay are measurable. Using wireless access point pricing information (MSRP) publicly shared by Cambium Networks, a sensitivity analysis can be run with NCAT2 (tool # 8) to identify crossing points benchmarking direct-to-consumer and hybrid distribution. The simulation illustrates that even considering a high-ARPU service; the direct-to-consumer approach is seriously challenged by the LEO-HTS terminal CAPEX.



Conclusions from this sample NCAT2 sensitivity simulation are clear: Even ignoring the fact that direct-to-consumer links tend to be less efficient in the use of satellite spectrum when compared to a service via high-end aggregation antenna system, project’s net present value (5-year NPV for the service provider) becomes negative for the direct-to-consumer case as a result of the FPA terminal cost. Charts also illustrate why it becomes vital for the (consumer-centric) Starlink service to achieve a terminal manufacturing cost below $2,000.

GEOs to the Rescue

While LEO constellations continue blanketing the globe, GEOs (and MEOs), being further away from Earth, can direct HTS spot-beam capacity more easily towards “hot spot” areas without wasting spectral resources and using low-cost VSATs. GEO-HTS systems (large to small GEOs) are thus ideally suited for capacity complementation, not only around hot spots such as airports, but also in low-ARPU populated regions within +/- 35 degrees of the Equator, where LEOs simultaneously exhibit lower density of overhead capacity and tighter interference-avoidance constraints (on top of expensive terminals).

Multi-orbit operation will certainly involve many service flavors and sweet spots that the industry will collectively find and develop solutions for, but two applications are driving the initial validation steps: Defense and mobility, as highlighted in two recent announcements:

• A multi-orbit Ku-band demonstration for the U.S. DoD involving OneWeb (polar LEO satellites), Intelsat (GEO-HTS IS-37e Epic satellite) and various user terminals including Kymeta (flat panel antenna), SatCube and Intellian.

• SES completion of multi-orbit Ka-band field tests using an Isotropic multi-link antenna, linking to a GEO satellite while simultaneously connected with an O3B satellite in medium earth orbit (MEO).

The main conclusion is that LEOs are –certainly- not precluded from being able to serve all covered regions and applications, but the described architectural and business-case aspects will encourage both Non-GEO and GEO players to gradually become more selective in “what battles to fight” and possibly partner for multi-orbit opportunities, depending on ARPU, competitiveness and applications. As an example, it will be difficult for LEOs to beat the CAPEX+OPEX service economics of GEO-VHTS satellites for contended broadband service plans, but LEOs could provide a low-latency, high-speed, low-contention premium service to high ARPU niches, both residential and enterprise. For aero in-flight connectivity (IFEC), large LEO constellations bring the benefit of high look angles for the aircraft antenna, potentially diminishing “skew angle” issues and reducing drag via low-profile, lightweight flat-panel antennas.

The Bottom Line

The satellite industry is in the midst of a major transformation with LEO constellations under the spotlight, but GEO-LEO (and GEO-MEO) complementation could ultimately become the end game. Commercial distribution partnerships and hybrid technologies for distribution will also come into play to drive both technical and commercial pipeline efficiencies. It is hard to tell how much complementation will be modulated by specific applications, but there is no doubt that such complementation will occur sooner or later, via either private-level coordination, provider partnerships, mergers and acquisitions or all.



NSR’s Non-GEO Constellations Analysis Toolkit 2.0 (NCAT2) is an assembly of flexible, configurable and easy-to-use quantitative models that x-ray and benchmark LEO/MEO high-throughput satcom constellations at architectural and business layers. The toolset provides a data-driven, unbiased vehicle for deep-diving into the intertwined technical and business aspects driving bandwidth supply, addressability and feasibility of leading mega-constellations, and their competitive standing versus terrestrial networks.

Benchmark SpaceX Starlink, SES O3B, OneWeb, Telesat Lightspeed, Amazon Kuiper, and (any) other LEO/MEO satellite system using the NSR toolbox. NCAT2 includes a multiplicity of analytical tools distributed across 15 Excel files (plus user guide and mobility datasets). Each tool has its own set of input variables, filters and calculation engines that drive output results, charts, maps, exportable data tables and visualizations.