Connecting every home in Africa -- Starlink backhaul?

Pia! runs fiber to an area and the "last mile" is wireless.

The vision of Kenyan ISP Poa! Internet is "to bring internet to every home in Africa."

Poa! offers unlimited, uncapped, 4 Mb/s fixed wireless connectivity to homes for 1,500 KSh ($11.64) per month plus a one-time installation fee of 3,500 KSh ($27.16). But wait, there's more. They provide a dual-SSID router and the home SSID is used by trusted family members and the open street SSID is for others. Street users get 100 MB of free data each day and are charged 15 KSh ($.16) per GB with no expiry .data if they exceed 100 MB on a given day

ASR 33 Teletype

Depending upon your age, 4 Mb/s may sound exceedingly slow, but my first home Internet connection was using an ASR 33 Teletype that printed ten upper-case characters per second -- about as fast as I could read. This enabled me to use email, Usenet, Telnet, FTP, etc. and, most importantly, to collaborate with remote colleagues. My first professional home computer had an 8-bit CPU, 64 KB memory, two 8-inch floppy drives, and a 300-baud modem. It enabled me to quit my day job and program, write and consult -- I owned my own professional tools.

A low-cost Windows PC and a 4 Mb/s connection would allow all that I did at that time and so much more today -- download software, articles, books, movies, etc., stream HD 720p video, make voice-over-IP calls, listen to podcasts, read newspapers, etc. and, importantly, create content and invent applications and services that are relevant to Africa. It would be fascinating to conduct a study of Poa! users to see what they are using the Internet for and how it is impacting their lives.

One thing we know for certain is that people do different things with a computer at home than with a phone on a mobile network or at a Wi-Fi hotspot. For example, I could not write this post on my Android phone.

This point is underscored by the experience of Poa!. Poa! CTO Mike Puchol said that when they deployed about 160 outdoor Wi-Fi hotspots in Kibera, a district of Nairobi, users consumed "very little" data despite the price being 10% of the mobile network operator’s rate.  When Poa! deployed residential broadband in Kibera, average consumption hit ~140 GB/month right away and is now over 200 GB/month. (In 2021, average mobile data consumption in sub-Saharan Africa was 2.9 GB per month).

Puchol provisions 1.5 Mb/s per 4 Mb/s account and he says there is "very little" contention. No doubt that contention grows during busy hours and days, but if they can add capacity as quickly as they add customers, they would be able to maintain that contention ratio without alienating too many customers.

They have fiber loops between their data center and points of presence in each network area they serve and use wireless links to reach individual houses. The service seems to be financially viable in urban areas like Kibera. Puchol reports that they currently have 20,000 home users and last December 27,000 people used the public street service. Poa! has received over $42 million in funding 

Starlink

SpaceX recently began offering Starlink Internet service in Nigeria and Rwanda and several South American nations and will begin service in Kenya next quarter. Could they substitute Starlink connectivity for fiber and offer their service in rural Kenya? 

Today, they could not. Today, there is only one Starlink ground station in Africa and it is in Nigeria. The only way a Starlink user in Kenya can reach the Internet is via inter-satellite links from a visible satellite to a remote ground station. I just ran Mike Puchol's Starlink simulation for one hour at a random location in central Kenya and there was no available coverage six percent of the time.

Since SpaceX has committed to offering connectivity in Kenya during the second quarter of this year, I assume they will have reachable ground stations by then. The same goes for Angola, Eswatini, Mozambique, and Zambia, which will have connectivity during the second quarter, and many other African nations that are slated to connect later in 2003 and in 2004. A rural village with access to a ground station could be used for Starlink backhaul.

If there is excess constellation capacity over a village location, the marginal cost of serving a new terminal will be near zero and an ISP can count on adding users at a predictable cost. I've seen a Nigerian Speedtest result showing a download rate of 238 Mb/s, an upload rate of 45 Mb/s, and a latency of 42 ms. At those speeds, a terminal could theoretically support 188 4 Mb/s customers, but that test was run on an unloaded terminal at a particular time of day. 

During a day, the available capacity at a given location varies as the satellites move and the number of users online varies. (Large files should be transferred late at night). 

In the long run, capacity increases as more satellites are launched and technology improves. SpaceX just began launching second-generation Starlink satellites which provide four times the capacity of the first generation and the next generation is expected to more than double that. The number and capacity of ground stations will also continue to increase, and more satellites with inter-satellite laser links will make it possible to reach them from remote sites. We will also see optical links to some ground stations and terminal cost and performance are also improving.

SpaceX is the only company offering low-Earth orbit broadband connectivity today, but they will have low and medium Earth orbit competitors and several of those future competitors are designing their constellations for organizations and enterprises which may make them better suited than SpaceX for backhaul from a village. For example, OneWeb will offer service-level agreements, which will take some of the uncertainty out of pricing decisions and SES will offer software-defined beam capacities and shapes.

The one prediction I can make given all this variability is that the prices SpaceX is charging in Rwanda and Nigeria today will change over time as they have in the US and other nations.

Poa! staff

In Kibera, Poa! provides both the fiber link to an area and the wireless links to the homes in the community, but decentralized alternatives where the community contracts with Poa! for the satellite link but installs, owns, and operates the local network should be considered. Whether operated by the local government or a private company, local community networks could coordinate to share expertise, training, and purchase equipment and tools in quantity. There could be variations on this theme. For example, Poa! could offer discount purchasing as a service to community networks.

The orbit-inclinations and marketing efforts of SpaceX's constellation were focused on affluent nations at first, but nearly all the African and South American nations will have Starlink service next year and OneWeb and SES will be serving many of them as well. Necessity is the mother of invention, and I can't wait to see the technology, organizations, and applications invented in the global south and their social impact.

PS -- The closest I can come to an English translation of poa is our slang term cool.

Starlink comes to Africa -- markets and competition

Source

SpaceX Starlink Internet service will be available in several African nations in the second quarter of this year and the price in Nigeria has been announced -- $600 for the "residential" terminal and a monthly fee of $43. Is there a market for Starlink at that price in Nigeria and other Sub-Saharan African nations?

The IMF projects a GDP per capita of $2,580 for Nigeria this year (and $1,900 in Sub-Saharan Africa and $2,260 for Africa overall) so the market for individual consumer accounts will be much smaller than in what the IMF refers to as "advanced economies." Instead, we will see shared accounts in government and commercial Telecenters, WiFi hot spots, etc. (Some private companies may encounter payment problems since Nigeria currently has a currency conversion ban).

The first Cabina Publica (source)

The first Internet Telecenter was a "Cabina Publica" in Lima Peru, providing access to the non-profit Peruvian Scientific Network founded by Jose Soriano. Around the same time, India was offering Internet services at rural Postal Offices and experimenting with "information Shops." The Telecenter movement grew and was documented in a book by Andy Carvin in 2005, (Coincidentally, Starlink just became available in Peru).

The availability of mobile phones and laptop computers has enabled widespread WiFi access but some telecenters still exist. For example, the Cuban government operates both fixed telecenters and WiFi access points,

Note that a rural satellite Telecenter can pivot from satellite to terrestrial connectivity if it becomes available and continue operation offering Internet and other services.

A clinic in Rural India (source)

Telecenters and hotspots are shared by individual customers, but government and private organizations in rural areas -- schools, clinics, local government offices, small businesses, etc. may also use Starlink connections to the Internet. Here is an early example where Starlink is connecting a school and clinic in rural Chile. Organizations or public access facilities may also create multi-terminal and local area networks on their own using Starlink mesh routers or third-party equipment. A multi-user organization that is more complex than a single residential terminal will require professional support. Starlink can also be used for backhaul from mobile towers in remote areas.

I've been talking about the residential Starlink offering because that is all that is currently available in Nigeria, but in other nations, SpaceX also offers a business configuration that is faster, comes with a guaranteed service-level agreement, and offers some support. The current business price is much higher -- a $2,500 terminal and $500 monthly charge -- but SpaceX has adopted an affordable pricing strategy for the residential service and may do the same for the business service. They also offer maritime and mobile services and may one day introduce other configurations.

Starlink will have competitors. 

Today they are the only low-Earth orbit (LEO) broadband Internet service provider, but OneWeb, Telesat, Amazon subsidiary Kuiper Systems, and China SatNet all plan to offer LEO broadband and SES offers competing medium Earth orbit connectivity. With the possible exception of China SatNet, none of these companies plans to offer retail residential service -- they are designing for and marketing to multi-user markets such as we are discussing here and have competitive advantages.

For example, Kuiper will have integration with Amazon Web and Ground station services, OneWeb will soon be offering service and they have technology and distribution partners with experience in developing nations, OneWeb and Telesat are well positioned to offer multi-orbit service through their geostationary orbit partners. SES is already serving potential Starlink customers and has begun launching next-generation mPower satellites. China SatNet will have political and financial advantages in Digital Silk Road nations.

Sources: Price and Speed 

There will also be competition from terrestrial ISPs. Internet customers in Nigeria and sub-Saharan Africa do not have the same expectations as those in wealthier nations. That is illustrated in this table contrasting the United States with the five nations SpaceX expects to serve in the second quarter of this year Nigeria, Mozambique, Angola, Tanzania, and Kenya. Mobile speed is faster than fixed in Africa and slower in the United States and mobile service costs less in these and other African nations. 

If it turns out that a $600 terminal and a $43 monthly charge do not attract enough customers to utilize available capacity, SpaceX will lower prices, as they have done in other nations.

An early speed test in Nigeria (source)

Available capacity is of course a moving target. Capacity per user decreases as new users come online and increases as new satellites are launched. Elon Musk claims the capacity of their second-generation "Starship" satellites will be close to ten times that of their current satellites and they are testing second-generation residential terminals which Musk says will cost a lot less.

I began by asking about the market for a $600 terminal for $43 per month service. I don't have an answer, but I bet the price will have changed within a year or two.

Update 2/9/2023

The experience of Poa Internet in Kenya indicates the viability of a residential Internet market in Africa. Poa CTO Mike Puchol reports that when they deployed ~160 outdoor Wi-Fi hotspots in Kibera, a district of Nairobi, users consumed little data in spite of the price being 10% of the mobile network operator's rate and when Poa offered 4 Mbps service with no caps over wireless links to residences, average consumption hit ~140 GB/month and is now over 200 GB/mo. People share access, use desktop and laptop computers, and do different things at home than on mobile networks or Wi-Fi hotspots.

Private or community ISPs will be able to provide wireless or wired residential service using Starlink at, say, 4 Mbps as long as there is sufficient capacity in the community. When Sanjay Bhargava was selected to head Starlink in India in late 2021, he gave a presentation in which he stated a rule of thumb of 100 Starlink terminals in 300 square kilometers. Around the same time, three academic researchers published a paper in which they estimated a mean per-user capacity of 24.94 Mbps at a density of 1 user per 10 km^2. Capacity has increased since 2021 as SpaceX has launched more satellites, some with inter-satellite laser links and perhaps greater throughput, and, as noted above, second-generation satellites promise significant capacity increases

Hole in Space -- the mother of all video chats

New technology enables new art forms and artists Kit Galloway and Sherrie Rabinowitz (K&S) began working with geostationary satellite links in 1977. Their first work was an experiment in remote dance and music. Video of dancers at The Goddard Space Flight Center in Maryland and an educational television center in California was transmitted to a central control studio where a composite was formed and sent back to monitors the dancers could see. 

First day: "They're in New York? I'm in LA, right?"

Their next project involved remote conversation rather than movement and they called it "Hole in Space" (HiS). The conversation took place in November 1980 between people at two locations, a display window at the Broadway department store at the Century City mall in Los Angeles and a window in the foyer of the Avery Fisher Concert Hall at the Lincoln Center in New York.

They made three two-hour connections. The first day was unannounced -- participants were curious passers-by who just stopped to see what was happening. They took the second day off and some word of mouth had gotten out by the third day. There was also local news television coverage before the final day, so some people made plans to meet distant friends and relatives.

The sessions began at 5:30 Los Angeles time, so illumination was needed. They did not want to attract large crowds by lighting the participants. Instead, they relied on visible light from inside the venues augmented by infrared emitters near the bottoms of the windows.

The terminals consisted of cameras developed by Cohu Industries that were sensitive to both infrared and visible light. (You can see the effect of the heat-sensitive infrared radiation in the glowing cigarettes of some participants). The displays were RCA projection TVs with custom-made 12 by 12-foot vinyl screens. Connections to the satellite ground stations were over microwave links to a mountaintop near Los Angeles and the roof of a tall building in New York.

Third day: conveying emotion as well as content

Communication was provided by Western Union's Westar satellites, which were typically used for things like the distribution of television programs and sporting events like the Los Angeles Lakers basketball games. Western Union provided three uninterruptible two-hour sessions. 

Galloway says the geostationary satellite latency was not a problem (as it had been for dance), but feedback between the speakers and microphones was, so they had operators at each location manually toggling echo cancellation on and off. There was also someone interviewing people in the crowds during the sessions. At one point a speaker failed in Los Angeles, and Rabinowitz had to run upstairs to borrow one from the store's audio-visual sales department.

K&S became aware of terrestrial networking after HiS and in 1984 began their Electronic Cafe project which supported conversations, remote collaborative drawing, and global New Year's Eve "Telebrations".

HiS was done at a time when nearly all data was text (only upper-case if your terminal was a Teletype or keypunch machine). It was done the year Usenet began and was three years before TCP/IP replaced NCP on the ARPAnet, five years before the NSFNET was established, and eight years before we saw the first text-only version of the Web. (See links to these and other milestones here).  K&S's art pieces anticipated modern services like Zoom for meeting online and JackTrip for remote musical practice and performance. 

Here are links to a short, narrated video, 4m 48s, and a longer video produced by K&S, 29m 45s.

I wish there were holes in space between Russia and Ukraine.

SpaceX launches "second generation" Starlink satellites

Why did SpaceX designate these satellites as Gen2 rather than Gen1? 

Starlink Gen1 v1.5 vs Gen2 v F9-1 satellites (source)

In interviews last Spring, Elon Musk said the data throughput of the next version of Starlink satellite (Gen2) would be almost an order of magnitude greater than that of the first generation and that the new Starship rocket would be needed to launch them. Regulatory and engineering delays slowed Starship's progress, so the Gen2 satellites Musk referred to at the time have not yet been launched.

Last Fall, SpaceX broadened the definition of Gen2 to include three configurations, designated F9-1, F9-2, and Starship. Musk was referring to the Gen2 Starship version when he described high-capacity satellites last Spring. The SpaceX Falcon rocket will launch F9-1 and F9-2 satellites, but the Starship version will require the Starship rocket. 

Last month, the FCC authorized SpaceX to launch 7,500 Gen2 satellites operating at altitudes of 525, 530, and 535 km and inclinations of 53, 43, and 33 degrees, respectively, using frequencies in the Ku- and Ka-bands. (They had applied for authorization for 29,988 satellites but only 7,500 were approved at this time -- perhaps due to concern over space debris).

Last week, SpaceX launched 54 Gen2 F9-1 satellites and they are close, if not identical, to the earlier Gen1, version 1.5 Satellites SpaceX has been launching recently. Eric Ralph referred to them as "suspiciously similar" and NASA confirmed that they were the same size and mass as the previously launched Gen1 version 1.5 satellites.

Why did SpaceX designate these satellites as Gen2 rather than Gen1? fo

Tim Farrar attributes it to SpaceX gaining a competitive advantage over potential rival Kuiper, saying "It's very clear that SpaceX wants to launch the first Ka-band Gen2 satellites before Kuiper's test satellites to gain a band-splitting advantage." 

In 2018, the FCC granted SpaceX permission to use the 27.5-29.1 GHz and 29.5-30.0 GHz bands for Gen1 Earth-space transmission. Subsequently, SpaceX was authorized to use those bands for its Gen2 satellites and Kuiper was authorized to use 27.5-3,0 GHz for its satellites. The FCC rules say that SpaceX and Kuiper can agree to a spectrum-sharing rule that protects the first-round rights of SpaceX, but if they cannot agree, the first to launch a satellite capable of operating in the frequency band under consideration gets to choose which portion of the spectrum it will use when interference is detected.

Source

A second motivation may have been the Gen2 right to orbit at a 43-degree inclination. Starlink has had insufficient capacity to serve customers in some regions and may have calculated that the 43-degree inclination would ease that congestion in relatively high-price, high-demand regions like the U. S., where Ookla has shown declining speeds for three quarters.

Regardless, the designations Gen1 and Gen2 seem arbitrary, and we won't see a meaningful difference between them until Gen2 F9-2 satellites are launched and we won't see a major difference until Starship is flying and launching Gen2 Starship satellites. 

Update 2/27/2023

Stack of 21 V2 Mini satellites

SpaceX has posted some information about the Gen 2 F9-2 satellites. They are referring to them as "V2 Mini" satellites and say they will have four times the capacity of the Gen 1 satellites. The image shown here shows 21 satellites, presumably the number that can be launched by the current Falcon 9 rocket. As mentioned above, Elon Musk predicts the Starship satellites will be around ten times the capacity of the current satellites. 

The update also lists several steps and links to documents SpaceX is taking to avoid collisions and to reduce reflected light that interferes with astronomy.

There was speculation that the first V2 Mini would be launched on February 23rd, but that launch was postponed until today and Starship is expected to attempt an orbital launch next month. I wonder whether SpaceX will launch more Gen 2 F9-1 satellites -- the answer probably depends upon the ability to ramp up manufacturing of the Minis.

Cuban undersea cable politics

Undersea cables between the U. S. and Cuba have long been intertwined with politics. 

In 1887, The New York Times reported on the inauguration of a cable in support of the Cuban insurgents fighting for independence from Spain -- a precursor to the Spanish-American war. Phone service between the U.S. and Cuba began in 1921 with AT&T's installation of an undersea cable and AT&T dominated international telephony to Cuba until the 1990s.

In 1994, WilTe1 applied for permission to construct a 210-kilometer, 2.5-gigabit fiber optic cable that would have had roughly 41 times the then-authorized capacity, but they never received approval for an Internet cable.

In 1966 Sprint established a wireless link from Florida to Cuba, providing Cuba with its first Internet connection with funds from the U. S. National Science Foundation International Connections Program. 

When President Obama relaxed relations with Cuba, Daniel Sepulveda, who led two U. S. government delegations to Cuba, said there were at least a half-dozen proposals — from U. S. and non-U. S. companies — to construct an undersea cable between the US and Cuba. The most promising proposal was for a 56-kilometer link between the existing ARCOS cable, which connects to southern Florida, and Cuba. The proposal was submitted in August 2018 and the FCC deemed it acceptable for expedited 45-day processing, but nothing happened until Trump established a committee to consider the security risks of the cable. This month the committee advised the FCC to deny the politically hot proposal. 

Within a week, ETECSA, Cuba's state-owned telecommunication monopoly, and the French telecommunication company Orange announced an agreement to construct a 2,470-kilometer cable connecting Cuba and the Caribbean Island of Martinique. At the announcement ceremony officials of ETECSA and Orange said technical work had already begun and the permits for the cable had been granted. This was clearly planned in advance and the announcement was triggered by the decision on the ARCOS cable. (I wonder if the topic came up when President Biden and Prime Minister Macron met recently).

As shown here, Cuba currently has undersea cable connections to Venezuela and Jamaica (red) and a U. S. owned cable (green) between Guantanamo and Florida. (There has been discussion of giving the Cable to Cuba someday). 

The new Arimao cable will run from Schoelcher, Martinique to Arimao Beach in the Cienfuegos Province on Cuba's south coast and will connect to Havana and other Cuban locations over the domestic backbone. They have already begun laying the cable, which will take around three weeks and it is expected to be ready for service in 2023. Its capacity is listed as "unknown."

The Arimao cable will improve service and resilience in Cuba, but approving the ARCOS cable would have been a better solution for the following reasons:

• It would have improved the standing of the U. S. in the region and the world. The UN General Assembly has voted on a resolution calling for the end of the U. S. embargo on Cuba every year since 1992 (except in 2020 due to COVID). This year 184 nations voted for the resolution, Colombia, Ukraine, and Brazil abstained and only the U. S. and Israel voted no.
• The 56-kilometer ARCOS cable would have been cheaper and a little faster than the 2,470-kilometer cable. (I do not know how the cable is being financed, but ETECSA and Orange are listed as owners).
• The cable would already be in service.
• The cable would have landed on the north side of the island at Cojimar, a district of Havana, the source and destination of heavy traffic, lightening the load on the Cuban backbone and improving latency.
• It may have improved U. S.-Cuba relations.

In my opinion, this was a lost opportunity.

Update 12/18/2022

Cuban Foreign Minister Bruno Rodríguez tweeted the following after the Cuba-Martinique undersea cable was announced:

Establishing a submarine cable connection between Florida and Cuba would be a positive step for both countries and would expand Internet access for Cubans. Denying it contradicts the position declared by the US government in May 2022. (original in Spanish)

Why did he post that at this time? Was it to make it clear that the U. S., not Cuba, had stopped the ALBA cable link? What impact would a third international link have?

Update 2/7/2023

ETECSA president Tania Velázquez Rodríguez said the current ALBA-1 cable is saturated and the new Arimao cable has a tentative go-live date of next April. Experts on Twitter doubt that the cable will be ready for service that soon. The expected capacity of the cable is unknown, but if the ALBA-1 cable is saturated, it should improve Cuban international access, 

Justice Department recommends that the FCC deny the proposed ARCOS cable segment connecting Florida and Cuba

Proposed link between the ARCOS undersea cable and Cuba

In September 2020, I wrote a post on a proposed 56-kilometer link between the ARCOS undersea cable and the north coast of Cuba, near Havana. The Trump-appointed Justice Department Committee for the Assessment of Foreign Participation in the United States Telecommunications Services Sector was to conduct a 120-day security review of the proposal. 

Since it is US policy to "amplify efforts to support the Cuban people through the expansion of internet services," I assumed the proposal would have smooth sailing, but it stalled. I followed up with several people but got no explanation.

When President Biden was elected, I argued in favor of the proposal, pointing out that it was consistent with our stated policy, would benefit the Cuban people, would improve our standing in the region, and would be popular with many Florida voters.

Yesterday a Justice Department press release stated The Committee had finally recommended that the FCC deny the ARCOS application. Assistant Attorney General Matthew Olsen said, “as long as the government of Cuba poses a counterintelligence threat to the United States, and partners with others who do the same, the risks to our critical infrastructure are simply too great.”

The Justice Department recommendation cites three risk factors (paraphrased):

1. ETECSA, Cuba's government-run telecommunication monopoly, would own the landing station in Cuba and could therefore have access to all traffic over a segment to Cuba.
2. The Cuban government could perform a BGP hijack to misdirect non-Cuban Internet traffic to themselves for interception.
3. The Cuban government has close ties to China and Russia and could share intel learned from #1 and #2 with them.

Doug Madory, Director of Internet Analysis at Kentik refuted the claims pointing out that all traffic to/from Cuba goes through ETECSA, and nothing stops them from misdirecting or sharing information with China and Russia now. Madory concludes "I suppose this was a political decision because the rationales listed in this announcement are completely nonsensical from a technical standpoint."

The report mentions a classified annex and considerable "confidential business information" is redacted, critical information, but from what we can see, as Madory says, this seems political.

Is the Defense Innovation Unit's Hybrid Space Architecture the "ARPANET" of space and will it run on Aalyria Spacetime?

Spacetime's user interface shows changing links and topology over time (source)

The goal of the Defense Innovation Unit (DIU) of the Department of Defense is to strengthen our national security by accelerating the adoption of commercial technology throughout the military and strengthening our allied and national security innovation bases. Space is one of its areas of focus and two of the space "lines of effort" are multi-orbit operations and logistics and hardware-to-software transformation modernization.

The DIU's Hybrid Space Architecture (HSA) program seeks to provide global, ubiquitous, and secure Internet connectivity throughout the space domain for commercial, civil, and military users, including international allies and partners. Several companies are developing multi-orbit Internet service between the low-Earth orbit (LEO) and geostationary (GEO) satellites and medium-Earth orbit (MEO) satellites and GEO satellites, but the HSA is more ambitious, calling for a "robust, secure software-defined network which integrates diverse telecom systems across LEO, MEO, GEO orbits, and cislunar space." The DIU recently awarded contracts to four companies to begin work on the HSA. They expect to award more contracts and plan on-orbit demonstrations within 24 months.

 

One of the awards went to Aalyria, a startup that is marketing Spacetime. Spacetime is a multi-layer, multi-orbit, operating system for a temporospatial network that "captures the potential for software-defined controllers to utilize knowledge of physics to make predictions about the future state of the lower-level network" for end-to-end path optimization. Those predictions require simulation of both the network stack and astronomical physics that began with the integration of space and network simulators at NASA.

The practical application of Spacetime grew out of Google's early efforts at connecting rural areas and developing nations. In a recent podcast interview, Brian Barritt, Aalyria CTO, and Executive VP said Spacetime exceeds the HSA requirements since it can integrate and optimize paths across terrestrial, air, LEO, MEO, GEO, cislunar, and eventually deep-space networks. (Vint Cerf, an early proponent of delay-tolerant networking for use in a Solar-System Internet is on the Aalyria Advisory Board and considers Spacetime to be applicable to NASA's next-generation space communications architecture).

Barritt says Spacetime was used by hundreds of thousands of users of Project Loon and is ready for adoption today. He added that Spacetime networks can interoperate, optimizing paths and potentially sharing assets across a federated network of networks of up to a combined size of fifteen million possible links (for now).

Several LEO broadband constellations are working on LEO, GEO and air integrations -- OneWeb seems to be farthest along -- but these efforts are of limited scope compared to Spacetime. Furthermore, they are proprietary solutions and Barritt says Spacetime APIs are open and available now, and they hope to make them an open standard. (Years ago, I worked on a project benchmarking Apple's local area network technology, AppleTalk, against IBM's Token-ring and Ethernet. The benchmark results were irrelevant because Ethernet was an open standard).

A multi-orbit network requires optical communication links between satellites. SpaceX has begun deploying satellites with optical inter-satellite links in their Starlink constellation and many vendors are working on space-space optical communication. DARPA is also pursuing a space-space standard that may become the "Ethernet" of inter-satellite links. Barritt says Spacetime drivers can be written to incorporate any optical terminal.

SpaceX recently moved to affordable pricing as performance faltered in some locations due to oversubscription given limited spectrum availability and gateway capacity. They will add capacity by improving technology and launching more satellites but growing traffic volume will require optical (and v-band RF) links between satellites and the ground, and those are problematic due to atmospheric signal distortion. 

Barritt says Aalyria's Tightbeam optical terminals can adapt to adverse atmospheric conditions under program control, for example cutting transmission speed to increase signal power to maintain signal integrity. Spacetime can adjust Tightbeam power, route around inclement weather, or find an RF link if that yields the optimal path. Tesat, Mynaric, Skyloom, the Space Development Agency, and Chinese companies are also working on optical space-ground communication and Spacetime drivers should be able to accommodate their terminals if they are superior to Tightbeam.

Barritt ended the interview by stating that despite the failure of Loon, high altitude platforms are not dead and spoke of the hypothetical case of a LEO broadband constellation that was over capacity in a region. He suggested that if it were running Spacetime it could federate with lighter-than-air vehicles to add capacity.

Disclaimers

Lest you run out and invest in Aalyria after reading this, I must add a disclaimer or two. Most of what I have said is based on statements from Aalyria and Brian Barrett's interview, which I encourage you to listen to.

While Project Loon logged over two million Spacetime-user hours, there have been no other adoptions as far as I know. Telesat initially planned to use Spacetime but that was canceled in favor of a network operating system from Thales Alenia Space. I don't know why that change was made, but it may have been because Telesat had been contemplating the integration of their LEO and GEO constellations and decided to keep them independent.

Tightbeam has been tested over long, challenging terrestrial links like from the ground to a mountaintop in the San Francisco Bay area and from aircraft to the ground, but not yet from space.

The commitments of today's LEO companies, particularly SpaceX and OneWeb, to proprietary operating systems might be Spacetime's biggest problem. SpaceX has around 500,000 residential subscribers and OneWeb and Eutelsat have merged and are committed to and working on LEO-GEO integration for OneWeb's version 2 satellites, which they hope to begin launching in mid-2025. 

We will know a lot more about Spacetime and Tightbeam when we see the on-orbit HSA demonstration in 24 months. 

Sci-Fi

Rather than end on a disclaimer, I'll offer some utopian science fiction. The current version of Spacetime can handle up to fifteen million links, which must require massive, parallel computation, but Barritt envisions federations of Spacetime networks if they can increase the link limit. That could enable the sharing of common resources like Amazon's ground-station service or federation of future broadband network generations. (Has Amazon settled on control software for the Project Kuiper constellation)? 

LeoLabs offers satellite and debris tracking and collision avoidance as a service that, like Spacetime, requires powerful simulation. Would there be a meaningful way for the two companies to collaborate on collision avoidance?

Last and least likely, how about a Spacetime federation that includes China's Satnet constellation? That is politically inconceivable today, but it will be many years before Satnet is ready to launch satellites and global challenges will be more pressing by the time they are.

Update 11/23/2022

.

DIU has awarded four contracts in the second phase of HSA to companies collectively pursuing the goals of an agile and resilient communications architecture that will move data across commercial, military, and allied assets while integrating multi-domain cloud-based storage and analytics. The organizations joining the DIU effort include SpiderOak Mission Systems, Amazon Web Services, and Project Kuiper, and Microsoft Azure Space.

Note that Amazon and Microsoft will be competing on cloud access and ground station service and Project Kuiper, not SpaceX Starlink, will be working on connecting the unconnected. The others are focused on battlefield status, security, and Privacy.

Update 3/13/2023

Aalyria has a commercial contract in addition to its DIU contract.

Rivada Space Networks will use Spacetime, in its planned low-Earth orbit communications constellation. The constellation will provide secure satellite networks with pole-to-pole reach, offering end-to-end latencies similar or better than terrestrial fiber. It "will operate like an optical backbone in space using lasers to interconnect satellites to deliver an ultra-secure and highly reliable global data network for business operations in the telecom, enterprise, maritime, energy and government services markets."

Aalyria, a space Internet startup with nearly a decade's worth of intellectual property from Alphabet

Has Aalyria's optical transmission technology eliminated the space-Earth communication bottleneck?

Aalyria, a new space Internet company, just burst out of stealth mode. It is based on work done on Alphabet's "moonshot" Project Loon and Alphabet transferred almost a decade’s worth of technology IP, patents, office space, and other assets to Aalyria in return for an equity stake in the company. Spacetime is Aalyria's intelligent network orchestration technology and Tightbeam is its advanced atmospheric laser communications technology.

Spacetime

Loon balloons floated at an altitude of 18-25 km, above
birds and the weather. They navigated by moving up or
down to catch wind currents moving in different directions.

Spacetime is a multi-layer, multi-orbit, software-defined networking system that was developed for Project Loon, one of Google's early efforts at connecting rural areas and developing nations. At one time, Telesat had agreed to use the Google networking system to link their low-Earth orbit and geostationary satellites, but Telesat has not yet launched its LEO constellation.

With the demise of Project Loon, the network management software was orphaned, but development continued and Aalyria says it now "optimizes and continually evolves the antenna link scheduling, network traffic routing, and spectrum resources -- responding in real-time to changing network requirements." That sounds like a tall order with constantly moving satellites, planes, ships, and vehicles, but the foundation was laid with drifting balloons.

This is an impressive claim, but it is not unique. Others are working on multi-orbit broadband networks and OneWeb recently signed an agreement for seamless interoperability between their low-Earth orbit satellites, Intelsat geosynchronous satellites, and airplanes. 

Tightbeam

Source

Tightbeam is a different story -- optical links are beginning to be used between satellites in space, but as far as I know, no one is currently transmitting production volume optical data between satellites and Earth. Optical communication is winning out over radio frequency links in space because they are faster, more secure, and harder to jam than radio frequency and the terminals have lower mass and consume less power. What's not to like? Unfortunately, rain, clouds, dust, or heat distort and attenuate optical signals.

One can imagine building ground stations in places with dry climates and routing around bad weather when it occurs, but Aalyria says they have developed novel hardware and algorithms that correct for these distortions enabling them to transmit data through the atmosphere at speeds up to 1.6 terabits per second over hundreds of miles.

Recently capacity limitations have slowed Space Starlink, triggering a shift to affordability-based pricing, and performance has continued to decline since that time. Over-subscription in a local area or cell contributes to that decline, but, as Mike Puchol points out, the scarcity of radio frequency spectrum for traffic between satellites and terrestrial gateways is also a constraint. Gateway congestion is already a problem and Starlink and others are planning to launch many more satellites. Puchol predicts that we will have optical links between satellites and gateways and speculates that they may use ultraviolet frequencies. The Chinese are also working on optical communication and they have conducted satellite-ground high-speed laser tests.

Regardless of who does it first, we will eventually see optical links between satellites and the ground. I've not seen any description of Tightbeam technology or results of tests and demonstrations, but if Aalyria's technology lives up to its description, it is important.

Miscellaneous 

A few miscellaneous points.

• I don't know where the name Aalyria comes from. I Googled it and only got references to the company itself. (There were tons of hits -- the company is hot).
• I wonder if they plan to operate their own constellation or license the technology. I suspect that the prospective broadband licensees already have their own "Spacetime" but not their own "Tightbeam." At some point, Aalyria (or Amazon, Microsoft, or Google) will roll out optical ground stations.
• My guess is that Tightbeam was developed by Alphabet project Taara which had been working on optical communication for Loon and other applications.
• I tried for a couple of days to get more information on Tightbeam and its performance? Technical papers, experimental results, patents, etc. but email to Aalyria.com bounces.
• Finally, I notice that the Board of Advisors has eleven members, four of which have prior Defense Department experience. That may have helped Aalyria secure an initial $8m contract with the Defense Innovation Unit. Another member is Vint Cerf, co-designer of TCP/IP, a Google VP, and most relevant in this context, a long-time proponent of interplanetary networking. Only one employee is listed as an optics engineer, but Board member Dr. Donald A. Cox III is an optical communications expert.

Update 10/8/2022

Tim Deaver, VP of Strategic Solutions at Mynaric says they are working on space-ground, air-ground, and ground-ground terminals. Will they be able to write terminal drivers and have them work in a Spacetime network?

SpaceX introduces affordability based Starlink pricing

GDP per capita, 2022 (Source)

These are the first Starlink price or service changes, but they won't be the last.

When SpaceX announced the price of the Starlink service, Elon Musk said it would be the same everywhere but I wrote that eventually it would be priced to be affordable in different nations. (If you predict enough things, you are bound to get something right).

The fixed cost of a satellite Internet constellation is high -- satellites are expensive to make and launch -- but the cost of adding and servicing a new customer is low and the market is global. At the initial price of $500 for the terminal and $99 per month for the service, there would be unused capacity in poor nations and contention for limited capacity in wealthy nations.

Earlier this summer, we were seeing complaints that Starlink sales were straining capacity in parts of the U.S. and Canada, and last month SpaceX announced variable pricing and an optional service cap in France. That was dubbed a pilot study, but since then the variable pricing dam burst, and customers in many nations received notification of permanent price cuts because of "local market conditions" and "parity in purchasing power." 

A 56% service price cut in Brazil

This change has been discussed on Reddit and a user established a crowd-sourced database of Starlink terminal and service charges by nation. As of this writing, the database contains complete records for thirty-nine nations. The average terminal price, including shipping, is $493.99, and the average monthly service charge is $72.65. In eight nations, the service charge has been cut by over 50%. In general, service charge cuts are greater than terminal price cuts since SpaceX is already subsidizing terminal purchases. (The database is currently a shared Excel spreadsheet, but I would like to see a cleaned-up version as a page on Wikipedia or better yet on the Starlink Web site).

These are the first Starlink price or service changes, but they won't be the last. The technology, the product mix, and the market will continuously change, and SpaceX will eventually encounter competition in the non-residential satellite broadband market. They will need data-driven Ph.D. marketing managers to set prices, not MBAs.

Update 1/3/2023

Source

Ookla reports crowd-sourced speed tests for Starlink every quarter. As seen here, performance in the U. S. has dropped in the last three reported quarters. 

OneWeb and Intelsat Sign the First Multi-Orbit Broadband Agreement – More to Come

A Boeing 777 connected to OneWeb LEO and Intelsat GEO satellites during this test flight. (Image from demo video).

Last week, OneWeb announced the first of what I expect to be many multi-orbit deals. 

LEO, MEO, and GEO satellites (not to scale)

Last October, I reviewed multi-orbit tests and plans of several low Earth orbit (LEO), medium Earth orbit (MEO), and Geostationary (GEO) broadband satellite companies, and quoted Neil Masterson, CEO of LEO operator OneWeb as saying "Interoperability with GEO satellites must happen -- it's common sense ... Customers don't care whether it's a LEO satellite or a GEO satellite -- all they want is connectivity." 

In June, OneWeb and Stellar demonstrated in-flight connectivity to a Boeing 777 jetliner, achieving a download speed of 260 mb/s. Ben Griffin, OneWeb vice president, says the speed is constrained by the current aircraft antenna and their satellite spot beams are capable of 500 mb/s. (777 carrying capacity is from 312-388 passengers).

Last week, OneWeb announced the first of what I expect to be many multi-orbit deals. Intelsat, an established GEO provider of in-flight connectivity, will distribute OneWeb's forthcoming LEO in-flight connectivity. The companies expect the multi-orbit solution to be in service by 2024 and they promise seamless connectivity with OneWeb LEO satellites for latency-sensitive applications and Intelsat GEO satellites for applications where latency is not critical. 

Note the emphasis on the word seamless above. The transition between constellations must be transparent to users. Automatic routing based on packet type is required and is well understood and implemented in different systems. For example, OneWeb had previously demonstrated LEO-GEO connectivity with Hughes using their ActiveClassifier technology that "automatically identifies different types of data and apps and tags them with a specific priority setting using advanced heuristics algorithms."

It is likely that OneWeb will soon be interoperating with three GEO operators. They have demonstrated seamless LEO-GEO integration with Hughes, have agreed to partner with Intelsat on in-flight connectivity, and have signed a memorandum of understanding to be acquired by GEO operator Eutelsat. This is the sort of interoperability promised by Neil Masterson. 

Will we see multi-orbit connectivity standards?

This is the first LEO-GEO broadband effort, but I expect many others because they solve problems for both LEO and GEO Internet service providers. LEO service providers are capacity constrained. SpaceX's Starlink LEO service is currently over-subscribed at some locations in the U.S. and Canada. They will add capacity by launching improved satellites and offloading latency-tolerant traffic to a GEO partner, freeing capacity for more valuable latency-sensitive traffic

GEO operators will seek LEO partners or, as in the case of Telesat, develop their own LEO constellations to provide low look-angle connectivity in polar regions and to serve latency-sensitive applications. Dan Goldberg, CEO of GEO ISP Telesat, acknowledged this when explaining why they were creating their own LEO constellation saying, “I guess it’s something we had to do. For us, you have no choice but to land at LEO.”

This post focuses on one application, in-flight connectivity, and one LEO operator, OneWeb, but this is just the start. For example, Intelsat and OneWeb have also been working on military applications and Viasat is working on multi-orbit solutions with several non-GEO operators or prospective operators and conducting a multi-layer satellite communication study for the European Space Agency. Intelsat is reportedly in merger talks with MEO operator SES. The elephant in the room is SpaceX. Elon Musk is a big fan of do-it-yourself integration, but it's hard to imagine SpaceX operating its own GEO satellites -- partnering with and launching satellites for a GEO operator seems more likely. I wonder whether China SatNet is looking for a GEO partner.

Update 3/11/2023

OneWeb considers multi-orbit interoperability a strategic capability and now Eutelsat has signed a multi-million euro deal to provide their GEO capacity and OneWeb's LEO capacity to Intelsat. 

OneWeb is expected to launch the final batch of its first-generation LEO satellites this month and be in full operation in a few months. Eutelsat operates a large constellation of GEO satellites as does Intelsat.

This gives OneWeb a lead over SpaceX and other potential LEO providers of aircraft connectivity. It will be interesting how the handoffs between LEO and GEO satellites are handled, how it affects the passenger user interface, and how well in-flight connectivity performs in a large, fully booked airliner.