Comparing LEO Constellations: How They Stack Up!

LEO constellations are set to introduce new dynamics in connectivity services. I’ve been looking to answer questions on the competitive or complementary nature of LEO with different types of terrestrial services such as fixed wireless access or long-haul transport. In process of the analysis, one needs to differentiate among LEO constellations since they are designed and optimized for certain use cases. I compared LEO constellations which I summarize here to highlight some of the differences.

LEO Use Cases

The use cases for LEO include

1. Fixed wireless access addressed by constellations such as Starlink (SpaceX) and Kuiper (Amazon).
2. Enterprise data backhaul addressed by constellations such as Telesat (planned).
3. IoT connectivity address by constellations such as Kepler, Sateliot, and Lacuna.
4. Personal connectivity addressed by constellations such as AST.
5. High-capacity data transport addressed by constellation such as LaserLight which uses free-space optics in MEO.

The use case is, of course, critical for both commercial reasons as well as technical reasons related to the design and optimization of the constellation and user terminals. OneWeb gives the example of a constellation that started on the premise of providing fixed wireless access, but changed to data backhaul services prior to bankruptcy.

LEO Differentiation

LEO satellites differentiate along many characteristics such as constellation configuration; RF and operating frequencies; and networking features such as intersatellite connectivity (ISL). The target use case provides the context in deciding on these tradeoffs in terms of cost and performance requirements.

Below are a few characteristics for SpaceX, OneWeb, Kuiper and Telesat. This is a representative sample of constellations focusing on access and backhaul applications. We counted over 90 constellations in different stages of maturity.

But a word of caution when comparing LEO constellations: some of these parameters are estimates based on regulatory filings and are likely to change. I also don’t capture here all the information because it is more complex than to fit neatly in a table. For example, OneWeb is in the midst of transformation as it emerges from Chapter 11. In recent filings OneWeb will deploy 716 satellites in phase 1, and 47,844 in phase 2. I opted only to list 716 for now. Similarly, SpaceX has a roadmap to deploy more satellites than 4,408; it plans to launch service with initial 1,440 satellites. With SpaceX in customer beta trials more information is becoming available that adds further context. For example, median user-PoP latency is reported at 30 msec with 95^th percentile at 43 msec [see here].

	SpaceX	Kuiper	OneWeb	Telesat
No. of Satellites [Deployed]	4,408 [901]	3,236	716 [110]	298
Altitude (km)	540 - 570	590; 610; 630	1,200	1,015; 1,325
Inter-Satellite Link	Version 2	Yes	No	Yes
DL throughput/satellite (Gbps)	20	16	8.8	60
DL / UL User throughput (Mbps)	100 / 40		50 / 25
Latency (msec; RTT)	20-60	30-60	30-60	30-60
User downlink / uplink band	Ku / Ku	Lower Ka / Upper Ka	Ku / Ku	Lower Ka / Upper Ka
User downlink / uplink bandwidth (MHz)	2,000 / 500	1,300 / 600	2,000 / 500	3,600 / 4,200
Coverage	57°S - 57°N	56°S - 56°N	Global	Global
Orbital planes	Inclined	Inclined	Polar and inclined	Inclined and polar
Cov. radius / satellite (km)	573.5	704.7
Lifespan (years)	5	7	10	10
Use case	Consumer access; enterprise; government & military	Consumer access; enterprise; maritime; aviation; government & military	Enterprise data backhaul; mobile backhaul; maritime; aviation; government & military	Enterprise data backhaul; mobile backhaul; maritime; aviation; government & military
Some of the information in this table, such as the number of deployed satellites, change over time. While I plan to maintain the table it may not reflect the most up-to-date data.

Cost Comparison

Comparing LEO constellation on cost is important for the context of the financial viability of these constellations as well as their competitive threat to other sectors, such as fixed wireless access [see here]. LEO constellations cost billion of dollars to deploy in a capex-heavy business model with relatively long period for breakeven. Several factors impact the cost which including the space and ground segments, and the user terminal. For example, SpaceX raised over $3.5 billion to date, while the cost of the constellation is projected at $10 billion.

To give an example, we recently compared LEO constellations to submarine cables to analyze the synergies between the two sectors. One may not associate the two sectors. However, consider high-frequency trading where LEO could provide a latency advantage over submarine fibre. Nevertheless, the low-cost structure and tremendous capacity of submarine cables leads to many synergies. For instance, LEO could act as an extension of submarine cable traffic into land-locked countries in Africa.

Stay tuned for future posts where I address synergies between LEO and various industries – e.g. submarine, fixed wireless access, wholesale, and data centres – outlining some of the consequences investors could face from the rise of LEO constellations!