Constellation wars: (much) more than one low Earth orbit data satellite network?

The following article contains my own opinions and analysis and is not official Deloitte research.

What if more than one low Earth orbit satellite (LEO sat) constellation gets built? There are already six (6) different satellite navigation constellations in orbit, owned by different countries/regions. In addition to at least one commercial network, my view is that China looks highly likely to launch their own LEO sat network, and I would not be surprised if we saw other constellations from (in descending order of likelihood) Russia, the EU, India and Japan, with a Sunni Muslim consortium also a possibility.

Commercial networks: Starlink has begun its “Better than nothing” beta service in parts of North America with 844 satellites in orbit [up to 904 as of Nov 25], and is launching 60-120 satellites per month, with plans for up to 20,000 satellites. UK-headquartered OneWeb emerged from bankruptcy last week, has 74 satellites in orbit already, and plans to launch another 36 next month and aims for 648 total. Amazon-owned Kuiper has no satellites in orbit now but aims for 3,236. Canadian operator Telesat has one satellite in orbit now, but has said it wants to have a constellation of almost 300 LEO sats.

Satellite navigation constellations: There are six national/regional global navigation satellite systems in orbit right now: the American GPS constellation has 31 sats, the Russian GLONASS has 24 sats, China’s BeiDou has 35 sats, the EU’s Galileo has 22 usable sats, India’s NavIC has eight sats, and Japan’s QZSS has four in orbit now, but plans to go up to seven. The reason that there is more than one positioning system is geopolitical: everyone COULD rely on the American system…but what if the US changed its mind and denied access to their technology? (None of these navigation constellations are in LEO, they are all in higher orbits.)

Other national/regional constellations: Major powers have their own intelligence gathering and military communications sats. As of 2018 there were 320 known military or dual use sats in various orbits, with the majority owned by the US, Russia, China and India. It is also worth noting that Arabsat provides data, voice and TV services to 21 Arab League countries through six sats, all in geosynchronous orbit, not LEO.

Why is China the most likely to build their own LEO sat constellation? Well, first off, they already seem to be interested! In early October, it was reported that China has filed a spectrum application with the International Telecommunication Union (ITU) for two constellations with a combined 12,992 sats, presumably for LEO data services. I see at least six reasons why they might be thinking this way:

1.  De-globalization/anti-China bans: Multiple Chinese companies have seen products banned or removed from networks, access to state-of-the-art chips denied, and talk of forced divestitures so far this year. If LEO sat networks are seen as a critical data access technology and important strategic lever in the 21st century…China almost certainly wouldn’t want to rely on non-Chinese constellations.

2.  Remote areas: China is large and has many remote areas that currently don’t have terrestrial low latency broadband. As of March 2020, the internet penetration rate in rural China was only 46%, and a LEO sat data service would help narrow that digital divide.

3.  Censorship: The Great Firewall of China is relatively easy to maintain: there are a relatively small number of terrestrial gateways between the Chinese internet and the global internet. It would be much harder to enforce a firewall for a space-based foreign LEO sat service, so China may want their own constellation.

4.  Global trade: China has many ships at sea and planes in the air, and both could use high speed low latency data services with global coverage. China-owned ships were 10.5% of the global shipping fleet as of 2019, for example.

5.  Global military: Equally, Chinese military ships, planes and troops would find a LEO sat network useful across the globe. The US Army already has an agreement with Starlink for military communications, showing the importance of the technology to support military force projection.

6.  Global aid and trade partners: Finally, China’s global infrastructure effort (the Belt and Road Initiative) has hundreds of far flung ports, railways, highways, power stations, etc. that could use LEO sat networks for backhaul, redundancy or back-up during natural disasters. But there are other countries or groups of countries that might want to build their own constellations too, for some or all of the reasons cited for China.

But LEO sat constellations cost crazy amounts of money! At $10 billion and up, there’s no way I am building my own.

Assuming a five year launch schedule, or $2B annually, building a LEO sat constellation would be about only 1% of the annual Chinese or EU military budget, about 3% of the Russian or Indian annual military spending, and 4% of Japan’s annual military budget. Chump change for China and the EU, and definitely affordable for the others.

All of China, Russia, the EU, India and Japan have extensive space expertise, skill at building satellites, and varying levels of launch capacity. Specifically, Russia, China and the EU would easily be able to launch their own constellations.

(Worth adding: Microsoft, Alphabet/Google, Apple and Facebook all have more than $50B in cash on their balance sheets, and annual free cash flow of over $10B: any one of them could literally build their own LEO sat constellation out of a single year’s free cash flow. I am not saying this is going to happen...but it seems plausible and possible.) 

Splinternets: One sub-trend of deglobalization is the emergence of national or regional internets, also known as internet balkanization or “splinternet.” The Great Firewall of China is the best known example, but there are over two dozen countries where internet censorship or filtering was deemed to be pervasive or substantial. There are also many Muslim countries, who individually may not be large enough to build their own constellation, but as a consortium would be able to do so. I think the idea of a Muslim LEO sat constellation is not a certainty…but feels like a long-shot possibility. Building a Sunni LEO sat constellation would take 3% of the Kingdom of Saudi Arabia’s annual military budget for five years, for example.

No judgement on my part: I am not in favor or against banning Chinese tech players, or restricting free speech, or fragmenting the world into trading or military or other blocs. But these things are happening, they can't be ignored, and the trend seems to be towards increasing fragmentation.

If LEO sat technology works and is seen as an important thing to have…multiple constellations will get built.

Is there room for 100,000+ satellites in low Earth orbit? We’re about to find out. Two or more private constellations, plus 4-6 national/regional constellations raises serious questions around spectrum, orbital slots, launch capacity, and risk of Kessler Syndrome (risk of satellites hitting each other, creating orbital debris, and making it impossible to leave the planet).