Disclaimer . The article is a supplemented, revised and updated translation of the publication by Nathan Hurst. Some information from the article on nanosatellites was also used in the construction of the final material.
There is a theory (or perhaps a warning tale) among astronomers called Kessler syndrome, named after NASA astrophysicist who proposed it in 1978. In this scenario, an orbiting satellite or some other object accidentally hits another and breaks into pieces. These parts revolve around the Earth at a speed of tens of thousands of kilometers per hour, destroying everything in its path, including other satellites. It launches a catastrophic chain reaction that ends in a cloud of millions of pieces of non-functional space debris that rotates endlessly around the planet.
Such an event can make near-Earth space useless, destroying any new satellites sent to it, and possibly blocking access to space as a whole.
Therefore, when SpaceX requested the FCC (Federal Communications Commission - United States) to send 4,425 satellites to low Earth orbit (LEO) to provide the global high-speed Internet, the FCC was concerned about this. For more than a year, the company has answered questions from competitors' commissions and petitions filed to refuse a statement, including the submission of a “plan to reduce orbital debris” to dispel fears of the Kessler apocalypse. On March 28, the FCC granted SpaceX's bid.
Space debris is not the only concern of the FCC, and SpaceX is not the only organization trying to build next-generation satellite constellations. A handful of companies, both new and old, use new technologies, develop new business plans and apply to the FCC for access to parts of the communication spectrum that they need to reach the Earth with a fast and reliable Internet.
Big names are involved - from Richard Branson to Elon Musk - along with big bucks. Currently, Branson’s OneWeb has raised $ 1.7 billion, and SpaceX’s president and chief operating officer, Gwynn Shotwell, has estimated the project worth $ 10 billion.
Of course, there are big problems, and history suggests that they affect quite adversely. Good guys try to overcome the digital divide in underserved regions, while bad guys try to install illegal satellites on missiles. And all this is due to the fact that there is a rapid increase in demand for data delivery: in 2016, global Internet traffic exceeded 1 sextillion bytes, according to a Cisco report, ending the era of zettabytes.
If the goal is to provide good Internet access where it did not exist before, then satellites are a smart way to achieve this. In fact, companies have been doing this for decades with the help of large geostationary satellites (GSO), which are in a very high orbit, where the rotation period is equal to the speed of rotation of the Earth, due to which they are fixed over a certain region. But with the exception of a few narrowly focused tasks, for example, capturing the Earth’s surface using 175 low-orbit satellites and transmitting 7 petabytes of data to Earth at a speed of 200 Mbps, or the task of tracking cargo or providing access to the network at military bases, this type of satellite communication was not fast and reliable enough to compete with today's fiber optic or cable Internet.
Non-GSOs include satellites that operate in mid-earth orbit (Medium Earth orbit, MEO) at altitudes from 1900 to 35000 km above the Earth’s surface, and low-orbit satellites (Low Earth orbit, LEO), whose orbits are located on altitudes less than 1900 km. Today, LEOs are becoming extremely popular, and in the near future it is expected that if not all satellites will be like that, then most certainly.
Meanwhile, regulations for non-geostationary satellites have long existed and are divided between agencies inside and outside the United States: NASA, FCC, DOD, FAA and even the United Nations International Telecommunication Union - all in this game.
However, from a technological point of view, there are some great advantages. The cost of building a satellite fell as gyroscopes and batteries improved due to the development of cell phones. Their launch also became cheaper, in part due to the smaller size of the satellites themselves. Capacity has increased, inter-satellite communications have made systems faster, and large plates pointing to the sky are already out of fashion.
11 companies have submitted applications to the FCC, along with SpaceX, each of which solves the problem in its own way.
Elon Musk announced the SpaceX Starlink program in 2015 and opened a Seattle affiliate. He told employees: "We want to revolutionize the satellite communications system in the same way as rocket science."
In 2016, the company filed an application with the US Federal Communications Commission, which requests permission to launch 1,600 (subsequently the number was reduced to 800) satellites from now until 2021, and then to launch the remaining satellites until 2024. These near-Earth satellites will orbit in 83 different orbital planes. The constellation, the so-called group of satellites, will communicate with each other via on-board optical (laser) communication lines, so that data can be reflected across the sky, and not returned to the ground - passing through a long "bridge", and not sent down and up.
On the ground, customers will install a new type of terminal with electronically-controlled antennas that will automatically connect to the satellite, which currently offers the best signal — similar to how a cell phone picks towers. As LEO satellites move relative to Earth, the system will switch between them every 10 minutes or so. And since thousands of people will use the system, according to Patricia Cooper, vice president of satellite management for SpaceX, there will always be at least 20 to choose from.
The ground terminal should be cheaper and easier to install than traditional satellite dishes, which should be physically oriented to that part of the sky where the corresponding geostationary satellite is located. SpaceX says that the terminal will be no larger than a pizza box (although it does not indicate what size pizza is).
Communication will be provided in two bands of the frequency spectrum: Ka and Ku. Both belong to the radio spectrum, although they use much higher frequencies than those used for stereo. The Ka band is the higher of the two, with frequencies between 26.5 GHz and 40 GHz, while the Ku band is located from 12 GHz to 18 GHz in the spectrum. Starlink received FCC permission to use certain frequencies, usually the uplink from the terminal to the satellite will operate at frequencies from 14 GHz to 14.5 GHz, and the downlink from 10.7 GHz to 12.7 GHz, and the rest will be used for telemetry, tracking and control, as well as in order to connect satellites to the terrestrial Internet.
In addition to FCC applications, SpaceX is silent and has not yet spoken about its plans. And it’s hard to find out any technical details, because SpaceX provides the entire system, starting from the components that will be used on the satellites, ending with the rockets that will bring them to the sky. But for the success of the project, this will depend on whether the service is said to be able to offer speeds comparable or better than optical fiber at the same price, along with reliability and a good user interface.
In February, SpaceX launched its first two prototypes of Starlink satellites, a cylindrical shape with solar panels in the form of wings. Tintin A and B are about a meter long, and Musk confirmed via Twitter that they successfully communicated. If prototypes continue to function, hundreds of others will join them by 2019. As soon as the system is put into operation, SpaceX will replace decommissioned satellites on an ongoing basis in order to prevent the appearance of space debris, the system will instruct them to lower their orbits at a certain point in time, after which they will begin to fall and burn out in the atmosphere. In the figure below you can see how the Starlink network looks after 6 launches.
A bit of history
Back in the 80s, HughesNet was an innovator in satellite technology. Do you know the gray antennas the size of a small dish that DirecTV mounts outside of houses? They come from HughesNet, which itself came about thanks to aviation pioneer Howard Hughes. “We invented technology that allows us to provide interactive communications via satellite,” says EVP Mike Cook.
In those days, the then Hughes Network Systems owned DirecTV and controlled large geostationary satellites that broadcast information to televisions. Then and now the company also offered services to enterprises, for example, processing credit card transactions at gas stations. The first commercial customer was Walmart, who wanted to connect employees across the country with a home office in Bentonville.
In the mid-90s, the company created a hybrid Internet system called DirecPC: a user’s computer sent a dial-up connection request to a web server and received a response via satellite, which sent the requested information down to the user's plate at a much faster speed than dial-up communication could provide .
Around the year 2000, Hughes began offering bidirectional network access services. But keeping the cost of the service, including the cost of customer equipment, low enough for people to buy, was a daunting task. For this, the company decided that it needed its own satellites and in 2007 it launched Spaceway. According to Hughes, this satellite, used so far, was especially important at launch, since it was the first to support onboard packet switching technology, in fact it was the first space switch to remove an additional hop in the form of a ground station for communication subscribers among themselves. Its capacity is over 10 Gbit / s, 24 transponders of 440 Mbit / s each, allowing individual subscribers to have up to 2 Mbit / s for transmission and up to 5 Mbit / s for download. Spaceway 1 was manufactured by Boeing on the basis of the Boeing 702 satellite platform. The launch weight of the device was 6080 kg. Currently, Spaceway 1 is one of the heaviest commercial spacecraft (SC) - it broke the record of the Inmarsat 4 F1 satellite (5959 kg) launched with the Atlas 5 rocket launcher a month earlier. While the heaviest commercial GSO, according to Wikipedia, launched in 2018, has a mass of 7 tons. The device is equipped with a relay payload (PN) Ka range. PN includes a controlled 2-meter phased antenna array consisting of 1,500 elements. PN forms a multi-beam coverage to provide broadcasting of various grids of television programs in different regions. Such an antenna allows flexible use of spacecraft capabilities in changing market conditions.
Meanwhile, a company called Viasat spent about ten years in research and development before launching its first satellite in 2008. This satellite, dubbed ViaSat-1, has included some new technologies, such as spectrum reuse. This made it possible for the satellite to choose between different bandwidths, in order to transmit data to the Earth without interference, even if it transmitted data along with the beam from another satellite, it could reuse this spectral range in compounds that are not adjacent.
This provided greater speed and performance. According to Viasat President Rick Baldridge, when it was put into operation, the throughput was 140 Gb / s - more than all other satellites combined covering the United States.
“The satellite market really was for people who had no choice,” says Boldridge. “If you could not access in another way, it was the technology of last resort. In fact, it had widespread coverage, but in fact did not allow the transfer of a lot of data. Therefore, this technology was mainly used for tasks such as transactions at gas stations. ”
Over the years, HughesNet (currently owned by EchoStar) and Viasat have created more and more fast geostationary satellites. HughesNet launched EchoStar XVII (120 Gb / s) in 2012, EchoStar XIX (200 Gb / s) in 2017 and plans to launch EchoStar XXIV in 2021, which the company says will offer consumers 100 Mb / s.
ViaSat-2 was launched in 2017 and now has a bandwidth of about 260 Gb / s, and three different ViaSat-3s are planned for 2020 or 2021, each of which will cover different parts of the globe. Viasat said that each of these three ViaSat-3 systems is projected to have terabits per second throughput, which is twice as much as all other satellites orbiting the earth combined.
“We have so much capacity in space that it changes the whole dynamics of providing this traffic. There is no limit to what can be provided, ”says D. K. Sachdev, a satellite and telecommunications technology consultant who works for LeoSat, one of the companies launching the LEO constellation. “Today, all the flaws of satellites are being eliminated one by one.”
This whole race of speeds did not appear by chance, since the Internet (two-way communication) began to supplant television (one-way communication) as a service for which satellites are used.
“The satellite industry is in a very old frenzy, figuring out how it will go from transmitting unidirectional video to full data,” says Ronald van der Breggen, Director of Compliance at LeoSat. “There are many opinions on how to do this, what to do, which market to serve.”
One problem remains
Delay. Unlike overall speed, latency is the amount of time it takes to send a request from your computer to the destination and back. Suppose you click on a link on a website, this request should reach the server and go back (that the server has successfully received the request and is going to give you the requested content), after which the web page is loaded.
How long it takes to load a site depends on the connection speed. The time required to complete the download request is a delay. Usually it is measured in milliseconds - therefore it is not noticeable when you watch the web, but it plays a big role when you play online games. However, there are facts when users from the Russian Federation succeeded and are able to play some of the games online even when the delay (ping) rate is close to one second.
The delay in the fiber system depends on the distance, but usually a few microseconds per kilometer, the main latency is introduced by equipment, although with optical links of considerable length the delay is more significant due to the fact that the speed of light in a fiber-optic communication line (FOCL) is only 60% of the speed of light in vacuum, and also very much dependent on the wavelength. According to Baldridge, the delay when you send a request to the GSO satellite is about 700 ms - the light travels in a space vacuum faster than in a fiber, but this type of satellite is far away, and that’s why it takes so long. In addition to games, this problem is significant for video conferencing, financial transactions and the stock market, control of the “Internet of things” and other applications that depend on the speed of interaction.
But how significant is the delay problem? Most of the bandwidth used worldwide is for video. Once the video is launched and properly buffered, the delay ceases to play a big role, and speed becomes much more important. Unsurprisingly, Viasat and HughesNet tend to minimize the importance of latency for most applications, although both work to minimize it in their systems. HughesNet uses an algorithm to prioritize traffic based on what users pay attention to to optimize data delivery. Viasat has announced the introduction of a medium-orbit satellite constellation (MEO) to complement its existing network, which should reduce latency and expand coverage, including at high latitudes, where equatorial GSOs have a large delay.
“We really focus on a large volume and a very, very low capital cost to deploy this volume,” says Baldridge. “Is the delay as important as the other functions for the market we support?”
Nevertheless, there is a solution, LEO satellites are much closer to users. Thus, companies such as SpaceX and LeoSat chose this path, planning to deploy a constellation of much smaller, closer satellites, with an expected delay of 20 to 30 milliseconds for users.
“The trade-off is that since they are in a lower orbit, you get less delay from the LEO system, but you have a more complex system,” says Cook.“To equip a constellation, you need to have at least hundreds of satellites because they are in low orbit, and they move around the Earth, going faster over the horizon and disappearing ... and you must have an antenna system that can track them.”
But two stories are worth recalling. In the early 90s, Bill Gates and several of his partners invested about $ 1 billion in a project called Teledesic to provide a broadband network in regions that cannot afford the network or will not see fiber optic lines soon. It was necessary to build a constellation of 840 (subsequently the number was reduced to 288) LEO satellites. Its founders talked about solving the problem of delays and in 1994 turned to the FCC to use the Ka-band spectrum. Sounds familiar?
Teledesic ate about $ 9 billion before failing in 2003.
“This idea did not work then because of the high cost of maintenance and end-user services, but it seems to be feasible now,” said Larry Press , professor of information systems at California State University Dominguez Hills, who has been monitoring LEO systems since how Teledesic appeared. “The technique was not advanced enough for this.”
Moore’s Law and improved battery technology, sensors and processors in cell phones have given LEO constellations a second chance. Increased demand makes the economy look attractive. But while the Teledesic saga was playing, another industry gained some important experience in launching communications systems into space. In the late 90s, Iridium, Globalstar and Orbcomm jointly launched more than 100 low-orbit satellites to provide coverage for cell phones.
“It takes years to create an entire constellation, because you need a whole bunch of launches, and it’s really expensive,” says Zach Manchester, an assistant professor of aeronautics and astronautics at Stanford University. “Over the past time, say, five years or so, the terrestrial infrastructure of cell towers has expanded to such an extent that coverage has become really good, and has covered most people.”
All three companies went bankrupt quickly. And although each of them found himself again, offering a smaller range of services for specific purposes, such as emergency beacons and cargo tracking, not one of them was able to replace cell phone communications based on towers. Over the past few years, SpaceX has launched satellites for Iridium under a contract.
“We've seen this movie before,” says Manchester. "I do not see anything fundamentally different in the current situation."
Competition
SpaceX and 11 other corporations (and their investors) have a different opinion. OneWeb launches satellites this year, and it is expected that services will begin to be provided as early as next year, then several more constellations will be added in 2021 and 2023 with an ultimate goal of 1000 Tbps by 2025. O3b, currently a subsidiary of SAS, has a constellation of 16 MEO satellites that have been in operation for several years. Telesat already operates GSO satellites, but is planning a LEO system for 2021, which will have optical links with a delay of 30 to 50 ms.
Upstart Astranis also has a satellite in geosynchronous orbit, and it will be deploying more in the next few years. Although they do not solve the problem of delays, the company seeks to drastically reduce costs by working with local Internet providers and creating smaller and much cheaper satellites.
LeoSat also plans to launch the first satellite series in 2019, and complete the construction of the “constellation” in 2022. They will fly around the Earth at an altitude of 1,400 km, connect to other network satellites using optical communications and transmit information up and down in the Ka-band. They have acquired the necessary spectrum internationally, says Richard van der Breggen, CEO of LeoSat, and are awaiting early approval by the FCC.
According to van der Breggen, the desire for a faster satellite Internet was largely based on the creation of larger and faster satellites capable of transmitting more data. He calls it a “pipe”: the larger the pipe, the more the Internet can break through it. But companies like him are finding new areas for improvement by changing the whole system.
“Imagine the smallest type of network — two Cisco routers and a wire between them,” says van der Breggen. "What all the satellites do is provide a wire between the two boxes ... we will deliver the entire set of three elements to space."
LeoSat plans to deploy 78 satellites, each the size of a large dining table and weighing about 1,200 kg. Built by Iridium, they are equipped with four solar panels and four lasers (one at each corner) for connecting to neighbors. This is the connection that van der Breggen considers the most important. Historically, satellites reflected a signal in the form of a letter V from a ground station to a satellite and then to a receiver. Since the LEO satellites are lower, they cannot project so far, but they can transfer data between themselves very quickly.
To understand how this works, it is useful to think of the Internet as something that has a real physical nature. This is not just data, it is where this data lives and how it moves. The Internet is not stored in one place, all over the world there are servers that contain some of the information, and when you access them, your computer takes data from the nearest one that has what you are looking for. Where is this important? How big does that matter? Light (information) travels in space faster than in fiber, almost twice. And when you miss the fiber connection around the planet, it should go along a detour from node to node with detours around mountains and continents. Satellite Internet is devoid of these shortcomings, and when the data source is far away, despite adding a couple of thousand miles of vertical distance,the delay with trying LEO will be less than the delay with fiber optic Internet. For example, ping from London to Singapore can be 112 ms, instead of 186, which will significantly improve connectivity.
Here is how van der Breggen describes the task, an entire industry can be seen as the development of a distributed network no different from the Internet as a whole, just in space. Delay and speed - both play a role.
Although the technology of one company may turn out to be excellent, this is not an antagonistic game; there will be no winners or losers. Many of these companies target different markets and even help each other achieve the results they rely on. For some, these are ships, planes or military bases, for others, these are rural consumers or developing countries. But in the end, companies pursue a common goal: to create the Internet where it is not, or where it is not enough, and to do it at a price low enough to maintain their business model.
“We believe that this is not really a competing technology. We believe that in a sense, both LEO and GEO technologies are needed, ”says Cook from HughesNet. “For certain types of applications, such as streaming video, for example, the GEO system is very, very cost-effective. However, if you want to use applications that require low latency ... LEO is the way to go. ”
In fact, HughesNet has partnered with OneWeb to provide gateway technology that manages traffic and communicates with the system over the Internet.
You may have noticed that the constellation proposed by LeoSat is almost 10 times smaller than SpaceX. That's fine, says van der Breggen, because LeoSat intends to serve corporate and government clients and will cover only a few specific areas. O3b sells the Internet to cruise ships, including Royal Caribbean, and works with telecommunications providers in Samoa and the Solomon Islands where there is a shortage of wired high-speed connections.
A small startup in Toronto called Kepler Communications uses tiny CubeSats (the size of a loaf of bread) to provide network access for customers who do not require latency, 5 GB of data or even more can be obtained in 10 minutes, which is important for polar research, science , industry and tourism. So, when installing a small antenna, the speed will be up to 20 Mbit / s for upload and up to 50 Mbit / s for download, but if you use a large “plate”, then the speeds are higher - 120 Mbit / s for upload and 150 Mbit / s for reception . According to Baldridge, Viasat's strong growth has been driven by the provision of the Internet to commercial airlines; they have signed agreements with United, JetBlue and American, as well as with Qantas, SAS and others.
How, then, will this profit-oriented business model bridge the digital divide and provide the Internet for developing countries and underserved populations that are probably not able to pay the same amount and are willing to pay less? This will be possible due to the format of the system. Since the individual satellites of the constellation LEO (low-orbit satellites) are in constant motion, they must be evenly distributed around the Earth, as a result of which they from time to time will cover regions in which no one lives or the population is very poor. Thus, any margin that can be obtained from these regions will be profitable.
“I assume that they will have different connection prices for different countries, and this will allow them to make the Internet accessible everywhere, even if it will be a very poor region,” says Press. “Once the constellation of satellites is there, its cost has already been fixed, and if the satellite is above Cuba and no one is using it, then any income that they can get from Cuba is marginal and free (does not require additional investments)” .
Entering a mass consumer market can be quite difficult. In fact, much of the industry’s success has come from providing the costly Internet to governments and businesses. But SpaceX and OneWeb, in particular, are targeting regular subscribers in their business plans.
According to Sachdev, the user interface will be important for this market. You must cover the Earth with a system that is easy to use, efficient and economical. “But just that is not enough,” says Sachdev. “You need a sufficient number of capacities, and before that you need to provide affordable prices for client equipment.”
Who is responsible for the regulation?
Two big problems that SpaceX had to solve with the FCC were how the spectrum of existing (and future) satellite communications would be distributed, and how to prevent space debris. The first question is FCC competency, but the second seems more appropriate for NASA or the US Department of Defense. Both track orbital objects to prevent collisions, but none of them are regulatory bodies.
“In fact, there is no good coordinated policy on what we should do with space debris,” says Stanford Manchester. “Right now, these people are not communicating with each other effectively, and there is no consistent policy.”
The problem is even more complicated because LEO satellites pass through many countries. The International Telecommunication Union plays a role similar to the FCC, assigning spectra, but for domestic work, the company must obtain permission from that country. Therefore, LEO satellites should be able to vary the spectral ranges used depending on the country over which they are located.
“Do you really want SpaceX to have a monopoly on connectivity in this region?”, Press is interested. “It is necessary to regulate their activities, and who has the right to do so?” They are supranational. The FCC has no jurisdiction in other countries. ”
However, this does not make the FCC powerless. Late last year, a small Silicon Valley startup called Swarm Technologies was denied permission to launch four prototypes of LEO communications satellites, each smaller than a paperback book. The main objection of the FCC was that tiny satellites can be too complicated to track and therefore unpredictable and dangerous.
Swarm launched them anyway. A Seattle-based satellite orbiting company sent them to India, where they went on a rocket carrying dozens of larger satellites, according to IEEE Spectrum. The FCC discovered this and fined the company $ 900,000, which must be paid within 5 years, and now Swarm’s application for four larger satellites is in limbo, the company works secretly. However, a few days ago there was news that approval was received for 150 small satellites. In general, money and the ability to negotiate - decided. The weight of the satellites is from 310 to 450 grams, at the moment there are 7 satellites in orbit, and the full network will be deployed in mid-2020. The latest report suggests that about $ 25 million has already been invested in the company, which opens up access to the market not only for global corporations.
For other future satellite Internet companies and existing ones learning new tricks, the next four to eight years will be crucial - they will determine if there is a demand for their technology here and now, or we will see a repeat of the story with Teledesic and Iridium. But what will happen after? Mars, according to Musk, his goal is to use Starlink to provide income for the exploration of Mars, as well as conduct a test.
“We could use the same system to create a network on Mars,” he told his staff. "Mars will also need a global communication system, and there are no fiber optic lines, no wires, or anything."
A bit of advertising :)
Thank you for staying with us. Do you like our articles? Want to see more interesting materials? Support us by placing an order or recommending it to your friends, a 30% discount for Habr users on a unique analog entry-level server that we invented for you: The whole truth about VPS (KVM) E5-2650 v4 (6 Cores) 10GB DDR4 240GB SSD 1Gbps from $ 20 or how to divide the server? (options are available with RAID1 and RAID10, up to 24 cores and up to 40GB DDR4).
Dell R730xd 2 times cheaper? Only we have 2 x Intel TetraDeca-Core Xeon 2x E5-2697v3 2.6GHz 14C 64GB DDR4 4x960GB SSD 1Gbps 100 TV from $ 199 in the Netherlands! Dell R420 - 2x E5-2430 2.2Ghz 6C 128GB DDR3 2x960GB SSD 1Gbps 100TB - from $ 99! Read about How to Build Infrastructure Bldg. class c using Dell R730xd E5-2650 v4 servers costing 9,000 euros for a penny?