Satellite Internet access

From Wikipedia, the free encyclopedia

 

Satellite Internet

Satellite Internet Characteristics
Medium	Air or Vacuum
License	ITU
Maximum downlink rate	1000 Gbit/s
Maximum uplink rate	1000 Mbit/s
Average downlink rate	1 Mbit/s
Average uplink rate	256 kbit/s
Latency	Average 638 ms
Frequency bands	L, C, Ku, Ka
Coverage	100–6,000 km
Additional services	VoIP, SDTV, HDTV, VOD, Datacast
Average CPE price	€300 (modem + satellite dish)

Satellite Internet access is Internet access provided through communications satellites. Modern consumer grade satellite Internet service is typically provided to individual users through geostationary satellites that can offer relatively high data speeds, with newer satellites using K_u band to achieve downstream data speeds up to 506 Mbit/s.

History of satellite Internet

Following the launch of the first satellite, Sputnik 1, by the Soviet Union in October 1957, the US successfully launched the Explorer 1 satellite in 1958. The first commercial communications satellite was Telstar 1, built by Bell Labs and launched in July 1962. 

The idea of a geosynchronous satellite—one that could orbit the Earth above the equator and remain fixed by following the Earth's rotation—was first proposed by Herman Potočnik in 1928 and popularised by the science fiction author Arthur C. Clarke in a paper in Wireless World in 1945.^[4] The first satellite to successfully reach geostationary orbit was Syncom3, built by Hughes Aircraft for NASA and launched on August 19, 1963. Succeeding generations of communications satellites featuring larger capacities and improved performance characteristics were adopted for use in television delivery, military applications and telecommunications purposes. Following the invention of the Internet and the World Wide Web, geostationary satellites attracted interest as a potential means of providing Internet access. 

A significant enabler of satellite-delivered Internet has been the opening up of the K_a band for satellites. In December 1993, Hughes Aircraft Co. filed with the Federal Communications Commission for a license to launch the first K_a-band satellite, Spaceway. In 1995, the FCC issued a call for more K_a-band satellite applications, attracting applications from 15 companies. Among those were EchoStar, Lockheed Martin, GE-Americom, Motorola and KaStar Satellite, which later became WildBlue. 

Among prominent aspirants in the early-stage satellite Internet sector was Teledesic, an ambitious and ultimately failed project funded in part by Microsoft that ended up costing more than $9 billion. Teledesic's idea was to create a broadband satellite constellation of hundreds of low-orbiting satellites in the K_a-band frequency, providing inexpensive Internet access with download speeds of up to 720 Mbit/s. The project was abandoned in 2003. Teledesic's failure, coupled with the bankruptcy filings of the satellite communications providers Iridium Communications Inc. and Globalstar, dampened marketplace enthusiasm for satellite Internet development. It wasn’t until September 2003 when the first Internet-ready satellite for consumers was launched by Eutelsat.

In 2004, with the launch of Anik F2, the first high throughput satellite, a class of next-generation satellites providing improved capacity and bandwidth became operational. More recently, high throughput satellites such as ViaSat's ViaSat-1 satellite in 2011 and HughesNet's Jupiter in 2012 have achieved further improvements, elevating downstream data rates from 1–3 Mbit/s up to 12–15Mbit/s and beyond. Internet access services tied to these satellites are targeted largely to rural residents as an alternative to Internet service via dial-up, ADSL or classic FSSes.

Since 2014, a rising number of companies announced working on internet access using satellite constellations in low Earth orbit. SpaceX, OneWeb and Amazon all plan to launch more than 1000 satellites each. OneWeb alone raised $1.7 billion by February 2017 for the project, and SpaceX raised over one billion in the first half of 2019 alone and expected more than $30 billion in revenue by 2025 from its satellite constellation. Many planned constellations employ laser communication for inter-satellite links to effectively create a space-based internet backbone.

As of 2017, airlines such as Delta and American have been introducing satellite internet as a means of combating limited bandwidth on airplanes and offering passengers usable internet speeds.

WildBlue satellite Internet dish on the side of a house

Companies and market

Companies providing home internet service include ViaSat, through its Exede brand, and EchoStar, through subsidiary HughesNet.

Function

Satellite Internet generally relies on three primary components: a satellite, typically in geostationary orbit (sometimes referred to as a geosynchronous Earth orbit, or GEO), a number of ground stations known as gateways that relay Internet data to and from the satellite via radio waves (microwave), and a small antenna at the subscriber's location, often a VSAT (very-small-aperture terminal) dish antenna with a transceiver. Other components of a satellite Internet system include a modem at the user end which links the user's network with the transceiver, and a centralized network operations center (NOC) for monitoring the entire system. Working in concert with a broadband gateway, the satellite operates a Star network topology where all network communication passes through the network's hub processor, which is at the center of the star. With this configuration, the number of remote VSATs that can be connected to the hub is virtually limitless.

Satellite

Marketed as the center of the new broadband satellite networks are a new generation of high-powered GEO satellites positioned 35,786 kilometres (22,236 mi) above the equator, operating in K_a-band (18.3–30 GHz) mode. These new purpose-built satellites are designed and optimized for broadband applications, employing many narrow spot beams, which target a much smaller area than the broad beams used by earlier communication satellites. This spot beam technology allows satellites to reuse assigned bandwidth multiple times which can enable them to achieve much higher overall capacity than conventional broad beam satellites. The spot beams can also increase performance and consequential capacity by focusing more power and increased receiver sensitivity into defined concentrated areas. Spot beams are designated as one of two types: subscriber spot beams, which transmit to and from the subscriber-side terminal, and gateway spot beams, which transmit to/from a service provider ground station. Note that moving off the tight footprint of a spotbeam can degrade performance significantly. Also, spotbeams can make the use of other significant new technologies impossible, including 'Carrier in Carrier' modulation.

In conjunction with the satellite's spot-beam technology, a bent-pipe architecture has traditionally been employed in the network in which the satellite functions as a bridge in space, connecting two communication points on the ground. The term "bent-pipe" is used to describe the shape of the data path between sending and receiving antennas, with the satellite positioned at the point of the bend. Simply put, the satellite's role in this network arrangement is to relay signals from the end user's terminal to the ISP's gateways, and back again without processing the signal at the satellite. The satellite receives, amplifies, and redirects a carrier on a specific radio frequency through a signal path called a transponder.

Some proposed satellite constellations in LEO such as SpaceX's Starlink, Telesat's constellation and LeoSat will employ laser communication equipment for high-throughput optical inter-satellite links. The interconnected satellites allow for direct routing of user data from satellite to satellite and effectively create a space-based optical mesh network that will enable seamless network management and continuity of service.

The satellite has its own set of antennas to receive communication signals from Earth and to transmit signals to their target location. These antennas and transponders are part of the satellite's "payload", which is designed to receive and transmit signals to and from various places on Earth. What enables this transmission and reception in the payload transponders is a repeater subsystem (RF (radio frequency) equipment) used to change frequencies, filter, separate, amplify and group signals before routing them to their destination address on Earth. The satellite's high-gain receiving antenna passes the transmitted data to the transponder which filters, translates and amplifies them, then redirects them to the transmitting antenna on board. The signal is then routed to a specific ground location through a channel known as a carrier. Beside the payload, the other main component of a communications satellite is called the bus, which comprises all equipment required to move the satellite into position, supply power, regulate equipment temperatures, provide health and tracking information, and perform numerous other operational tasks.

Gateways

Along with dramatic advances in satellite technology over the past decade, ground equipment has similarly evolved, benefiting from higher levels of integration and increasing processing power, expanding both capacity and performance boundaries. The Gateway—or Gateway Earth Station (its full name)—is also referred to as a ground station, teleport or hub. The term is sometimes used to describe just the antenna dish portion, or it can refer to the complete system with all associated components. In short, the gateway receives radio wave signals from the satellite on the last leg of the return or upstream payload, carrying the request originating from the end-user's site. The satellite modem at the gateway location demodulates the incoming signal from the outdoor antenna into IP packets and sends the packets to the local network. Access server/gateways manage traffic transported to/from the Internet. Once the initial request has been processed by the gateway's servers, sent to and returned from the Internet, the requested information is sent back as a forward or downstream payload to the end-user via the satellite, which directs the signal to the subscriber terminal. Each Gateway provides the connection to the Internet backbone for the gateway beam(s) it serves. The system of gateways comprising the satellite ground system provides all network services for satellite and corresponding terrestrial connectivity. Each gateway provides a multiservice access network for subscriber terminal connections to the Internet. In the continental United States, because it is north of the equator, all gateway and subscriber dish antenna must have an unobstructed view of the southern sky. Because of the satellite's geostationary orbit, the gateway antenna can stay pointed at a fixed position.

Antenna dish and modem

For the customer-provided equipment (i.e. PC and router) to access the broadband satellite network, the customer must have additional physical components installed:

Outdoor unit (ODU)

At the far end of the outdoor unit is typically a small (2–3-foot, 60-90cm diameter), reflective dish-type radio antenna. The VSAT antenna must also have an unobstructed view of the sky to allow for proper line-of-sight (L-O-S) to the satellite. There are three physical characteristic settings used to ensure that the antenna is configured correctly at the satellite, which are: azimuth, elevation, polarization, and skew. The combination of these settings gives the outdoor unit a L-O-S to the chosen satellite and makes data transmission possible. These parameters are generally set at the time the equipment is installed, along with a beam assignment (K_a-band only); these steps must all be taken prior to the actual activation of service. Transmit and receive components are typically mounted at the focal point of the antenna which receives/sends data from/to the satellite. The main parts are:

• Feed – This assembly is part of the VSAT receive and transmit chain, which consists of several components with different functions, including the feed horn at the front of the unit, which resembles a funnel and has the task of focusing the satellite microwave signals across the surface of the dish reflector. The feed horn both receives signals reflected off the dish's surface and transmits outbound signals back to the satellite.
• Block upconverter (BUC) – This unit sits behind the feed horn and may be part of the same unit, but a larger (higher wattage) BUC could be a separate piece attached to the base of the antenna. Its job is to convert the signal from the modem to a higher frequency and amplify it before it is reflected off the dish and towards the satellite.
• Low-noise block downconverter (LNB) – This is the receiving element of the terminal. The LNB's job is to amplify the received satellite radio signal bouncing off the dish and filter out the noise, which is any signal not carrying valid information. The LNB passes the amplified, filtered signal to the satellite modem at the user's location.

Indoor unit (IDU)

The satellite modem serves as an interface between the outdoor unit and customer-provided equipment (i.e. PC, router) and controls satellite transmission and reception. From the sending device (computer, router, etc.) it receives an input bitstream and converts or modulates it into radio waves, reversing that order for incoming transmissions, which is called demodulation. It provides two types of connectivity:

• Coaxial cable (COAX) connectivity to the satellite antenna. The cable carrying electromagnetic satellite signals between the modem and the antenna generally is limited to be no more than 150 feet in length.
• Ethernet connectivity to the computer, carrying the customer's data packets to and from the Internet content servers.

Consumer grade satellite modems typically employ either the DOCSIS (Data Over Cable Service Interface Specification) or WiMAX (World Interoperability for Microwave Access) telecommunication standard to communicate with the assigned gateway.

Challenges and limitations

Signal latency

Latency (commonly referred to as "ping time") is the delay between requesting data and the receipt of a response, or in the case of one-way communication, between the actual moment of a signal's broadcast and the time it is received at its destination.

A radio signal takes about 120 milliseconds to reach a geostationary satellite and then 120 milliseconds to reach the ground station, so nearly 1/4 of a second overall. Typically, during perfect conditions, the physics involved in satellite communications account for approximately 550 milliseconds of latency round-trip time.

The longer latency is the primary difference between a standard terrestrial-based network and a geostationary satellite-based network. The round-trip latency of a geostationary satellite communications network can be more than 12 times that of a terrestrial based network.

Geostationary orbits

A geostationary orbit (or geostationary Earth orbit/GEO) is a geosynchronous orbit directly above the Earth's equator (0° latitude), with a period equal to the Earth's rotational period and an orbital eccentricity of approximately zero (i.e. a "circular orbit"). An object in a geostationary orbit appears motionless, at a fixed position in the sky, to ground observers. Launchers often place communications satellites and weather satellites in geostationary orbits, so that the satellite antennas that communicate with them do not have to move to track them, but can point permanently at the position in the sky where the satellites stay. Due to the constant 0° latitude and circularity of geostationary orbits, satellites in GEO differ in location by longitude only.

Compared to ground-based communication, all geostationary satellite communications experience higher latency due to the signal having to travel 35,786 km (22,236 mi) to a satellite in geostationary orbit and back to Earth again. Even at the speed of light (about 300,000 km/s or 186,000 miles per second), this delay can appear significant. If all other signaling delays could be eliminated, it still takes a radio signal about 250 milliseconds (ms), or about a quarter of a second, to travel to the satellite and back to the ground. The absolute minimum total amount of delay varies, due to the satellite staying in one place in the sky, while ground-based users can be directly below (with a roundtrip latency of 239.6 ms), or far to the side of the planet near the horizon (with a roundtrip latency of 279.0 ms).

For an Internet packet, that delay is doubled before a reply is received. That is the theoretical minimum. Factoring in other normal delays from network sources gives a typical one-way connection latency of 500–700 ms from the user to the ISP, or about 1,000–1,400 ms latency for the total round-trip time (RTT) back to the user. This is more than most dial-up users experience at typically 150–200 ms total latency, and much higher than the typical 15–40 ms latency experienced by users of other high-speed Internet services, such as cable or VDSL.

For geostationary satellites, there is no way to eliminate latency, but the problem can be somewhat mitigated in Internet communications with TCP acceleration features that shorten the apparent round trip time (RTT) per packet by splitting ("spoofing") the feedback loop between the sender and the receiver. Certain acceleration features are often present in recent technology developments embedded in satellite Internet equipment.

Latency also impacts the initiation of secure Internet connections such as SSL which require the exchange of numerous pieces of data between web server and web client. Although these pieces of data are small, the multiple round-trips involved in the handshake produce long delays compared to other forms of Internet connectivity, as documented by Stephen T. Cobb in a 2011 report published by the Rural Mobile and Broadband Alliance. This annoyance extends to entering and editing data using some Software as a Service or SaaS applications as well as in other forms of online work.

One should thoroughly test the functionality of live interactive access to a distant computer—such as virtual private networks. Many TCP protocols were not designed to work in high-latency environments.

Medium and Low Earth Orbits

Medium Earth orbit (MEO) and low Earth orbit (LEO) satellite constellations do not have such great delays, as the satellites are closer to the ground. For example:

• The current LEO constellations of Globalstar and Iridium satellites have delays of less than 40 ms round trip, but their throughput is less than broadband at 64 kbit/s per channel. The Globalstar constellation orbits 1,420 km above the Earth and Iridium orbits at 670 km altitude.
• The O3b Networks MEO constellation orbits at 8,062 km, with RTT latency of approximately 125 ms. The proposed new network is also designed for much higher throughput with links well in excess of 1 Gbit/s (Gigabits per second).

Unlike geostationary satellites, low- and medium-Earth orbit satellites do not stay in a fixed position in the sky. Consequently, ground-based antennas cannot easily lock into communication with any one specific satellite. As with GPS, for a receiver the satellites are only visible for a part of their orbit, therefore multiple satellites are necessary to establish a permanent internet connection, with low-Earth orbits needing more satellites than medium-Earth orbits. The network has to switch data transfer between satellites to keep a connection to a customer.

One can communicate with MEO or LEO satellites that move in the sky in three ways, using:

• more diffuse or completely omnidirectional ground antennas capable of communicating with one or more satellites visible in the sky at the same time, but at significantly higher transmit power than fixed geostationary dish antennas (due to the lower gain), and with much poorer signal-to-noise ratios for receiving the signal
• motorized antenna mounts with high-gain, narrow beam antennas tracking individual satellites
• phased array antennas that can steer the beam electronically, together with software that can predict the path of each satellite in the constellation

Ultralight atmospheric aircraft as satellites

A proposed alternative to relay satellites is a special-purpose solar-powered ultralight aircraft, which would fly along a circular path above a fixed ground location, operating under autonomous computer control at a height of approximately 20,000 meters.

For example, the United States Defense Advanced Research Projects Agency Vulture project envisaged an ultralight aircraft capable of station-keeping over a fixed area for a period of up to five years, and able to provide both continuous surveillance to ground assets as well as to service extremely low-latency communications networks. This project was cancelled in 2012 before it became operational.

Onboard batteries would charge during daylight hours through solar panels covering the wings, and would provide power to the plane during night. Ground-based satellite internet dishes would relay signals to and from the aircraft, resulting in a greatly reduced round-trip signal latency of only 0.25 milliseconds. The planes could potentially run for long periods without refueling. Several such schemes involving various types of aircraft have been proposed in the past.

Interference

A foldable Bigpond satellite Internet dish

 

Satellite communications are affected by moisture and various forms of precipitation (such as rain or snow) in the signal path between end users or ground stations and the satellite being utilized. This interference with the signal is known as rain fade. The effects are less pronounced on the lower frequency 'L' and 'C' bands, but can become quite severe on the higher frequency 'Ku' and 'Ka' band. For satellite Internet services in tropical areas with heavy rain, use of the C band (4/6 GHz) with a circular polarisation satellite is popular. Satellite communications on the K_a band (19/29 GHz) can use special techniques such as large rain margins, adaptive uplink power control and reduced bit rates during precipitation.

Rain margins are the extra communication link requirements needed to account for signal degradations due to moisture and precipitation, and are of acute importance on all systems operating at frequencies over 10 GHz.

The amount of time during which service is lost can be reduced by increasing the size of the satellite communication dish so as to gather more of the satellite signal on the downlink and also to provide a stronger signal on the uplink. In other words, increasing antenna gain through the use of a larger parabolic reflector is one way of increasing the overall channel gain and, consequently, the signal-to-noise (S/N) ratio, which allows for greater signal loss due to rain fade without the S/N ratio dropping below its minimum threshold for successful communication. 

Modern consumer-grade dish antennas tend to be fairly small, which reduces the rain margin or increases the required satellite downlink power and cost. However, it is often more economical to build a more expensive satellite and smaller, less expensive consumer antennas than to increase the consumer antenna size to reduce the satellite cost.

Large commercial dishes of 3.7 m to 13 m diameter can be used to achieve increased rain margins and also to reduce the cost per bit by allowing for more efficient modulation codes. Alternately, larger aperture antennae can require less power from the satellite to achieve acceptable performance. Satellites typically use photovoltaic solar power, so there is no expense for the energy itself, but a more powerful satellite will require larger, more powerful solar panels and electronics, often including a larger transmitting antenna. The larger satellite components not only increase materials costs but also increase the weight of the satellite, and in general, the cost to launch a satellite into an orbit is directly proportional to its weight. (In addition, since satellite launch vehicles [i.e. rockets] have specific payload size limits, making parts of the satellite larger may require either more complex folding mechanisms for parts of the satellite like solar panels and high-gain antennas, or upgrading to a more expensive launch vehicle that can handle a larger payload.) 

Modulated carriers can be dynamically altered in response to rain problems or other link impairments using a process called adaptive coding and modulation, or "ACM". ACM allows the bit rates to be increased substantially during normal clear sky conditions, increasing the number of bits per Hz transmitted, and thus reducing overall cost per bit. Adaptive coding requires some sort of a return or feedback channel which can be via any available means, satellite or terrestrial.

Line of sight

Fresnel zone. D is the distance between the transmitter and the receiver, r is the radius of the Fresnel zone.

 

An object is in your line of sight if you can draw a straight line between yourself and the object without any interference, such as a mountain or a bend in a road. An object beyond the horizon is below the line of sight and, therefore, can be difficult to communicate with. 

Typically a completely clear line of sight between the dish and the satellite is required for the system to work optimally. In addition to the signal being susceptible to absorption and scattering by moisture, the signal is similarly impacted by the presence of trees and other vegetation in the path of the signal. As the radio frequency decreases, to below 900 MHz, penetration through vegetation increases, but most satellite communications operate above 2 GHz making them sensitive to even minor obstructions such as tree foliage. A dish installation in the winter must factor in plant foliage growth that will appear in the spring and summer.

Fresnel zone

Even if there is a direct line of sight between the transmitting and receiving antenna, reflections from objects near the path of the signal can decrease apparent signal power through phase cancellations. Whether and how much signal is lost from a reflection is determined by the location of the object in the Fresnel zone of the antennas.

Two-way satellite-only communication

The back panel of a satellite modem, with coaxial connections for both incoming and outgoing signals, and an Ethernet port for connection

 

Home or consumer grade two-way satellite Internet service involves both sending and receiving data from a remote very-small-aperture terminal (VSAT) via satellite to a hub telecommunications port (teleport), which then relays data via the terrestrial Internet. The satellite dish at each location must be precisely pointed to avoid interference with other satellites. At each VSAT site the uplink frequency, bit rate and power must be accurately set, under control of the service provider hub.

There are several types of two way satellite Internet services, including time division multiple access (TDMA) and single channel per carrier (SCPC). Two-way systems can be simple VSAT terminals with a 60–100 cm dish and output power of only a few watts intended for consumers and small business or larger systems which provide more bandwidth. Such systems are frequently marketed as "satellite broadband" and can cost two to three times as much per month as land-based systems such as ADSL. The modems required for this service are often proprietary, but some are compatible with several different providers. They are also expensive, costing in the range of USD $600 to $2000.

The two-way "iLNB" used on the SES Broadband.

 

The two-way "iLNB" used on the SES Broadband terminal dish has a transmitter and single-polarity receive LNB, both operating in the K_u band. Pricing for SES Broadband modems range from €299 to €350. These types of system are generally unsuitable for use on moving vehicles, although some dishes may be fitted to an automatic pan and tilt mechanism to continuously re-align the dish—but these are more expensive. The technology for SES Broadband was delivered by a Belgian company called Newtec.

Bandwidth

Consumer satellite Internet customers range from individual home users with one PC to large remote business sites with several hundred PCs.

Home users tend to use shared satellite capacity to reduce the cost, while still allowing high peak bit rates when congestion is absent. There are usually restrictive time-based bandwidth allowances so that each user gets their fair share, according to their payment. When a user exceeds their allowance, the company may slow down their access, deprioritise their traffic or charge for the excess bandwidth used. For consumer satellite Internet, the allowance can typically range from 200 MB per day to 25 GB per month. A shared download carrier may have a bit rate of 1 to 40 Mbit/s and be shared by up to 100 to 4,000 end users. 

The uplink direction for shared user customers is normally time division multiple access (TDMA), which involves transmitting occasional short packet bursts in between other users (similar to how a cellular phone shares a cell tower). 

Each remote location may also be equipped with a telephone modem; the connections for this are as with a conventional dial-up ISP. Two-way satellite systems may sometimes use the modem channel in both directions for data where latency is more important than bandwidth, reserving the satellite channel for download data where bandwidth is more important than latency, such as for file transfers.

In 2006, the European Commission sponsored the UNIC project which aims at developing an end-to-end scientific test bed for the distribution of new broadband interactive TV-centric services delivered over low-cost two-way satellite to actual end-users in the home. The UNIC architecture employs DVB-S2 standard for downlink and DVB-RCS standard for uplink.

Normal VSAT dishes (1.2–2.4 m diameter) are widely used for VoIP phone services. A voice call is sent by means of packets via the satellite and Internet. Using coding and compression techniques the bit rate needed per call is only 10.8 kbit/s each way.

Portable satellite Internet

Portable satellite modem

Portable Satellite Internet Modem and Antenna deployed with the Red Cross in South Sudan.

 

These usually come in the shape of a self-contained flat rectangular box that needs to be pointed in the general direction of the satellite—unlike VSAT the alignment need not be very precise and the modems have built in signal strength meters to help the user align the device properly. The modems have commonly used connectors such as Ethernet or Universal Serial Bus (USB). Some also have an integrated Bluetooth transceiver and double as a satellite phone. The modems also tend to have their own batteries so they can be connected to a laptop without draining its battery. The most common such system is INMARSAT's BGAN—these terminals are about the size of a briefcase and have near-symmetric connection speeds of around 350–500 kbit/s. Smaller modems exist like those offered by Thuraya but only connect at 444 kbit/s in a limited coverage area. INMARSAT now offer the IsatHub, a paperback book sized satellite modem working in conjunction with the users mobile phone and other devices. The cost has been reduced to $3 per MB and the device itself is on sale for about $1300.

Using such a modem is extremely expensive—data transfer costs between $5 and $7 per megabyte. The modems themselves are also expensive, usually costing between $1,000 and $5,000.

Internet via satellite phone

For many years satellite phones have been able to connect to the Internet. Bandwidth varies from about 2400 bit/s for Iridium network satellites and ACeS based phones to 15 kbit/s upstream and 60 kbit/s downstream for Thuraya handsets. Globalstar also provides Internet access at 9600 bit/s—like Iridium and ACeS a dial-up connection is required and is billed per minute, however both Globalstar and Iridium are planning to launch new satellites offering always-on data services at higher rates. With Thuraya phones the 9,600 bit/s dial-up connection is also possible, the 60 kbit/s service is always-on and the user is billed for data transferred (about $5 per megabyte). The phones can be connected to a laptop or other computer using a USB or RS-232 interface. Due to the low bandwidths involved it is extremely slow to browse the web with such a connection, but useful for sending email, Secure Shell data and using other low-bandwidth protocols. Since satellite phones tend to have omnidirectional antennas no alignment is required as long as there is a line of sight between the phone and the satellite.

One-way receive, with terrestrial transmit

One-way terrestrial return satellite Internet systems are used with conventional dial-up Internet access, with outbound (upstream) data traveling through a telephone modem, but downstream data sent via satellite at a higher rate. In the U.S., an FCC license is required for the uplink station only; no license is required for the users. 

Another type of 1-way satellite Internet system uses General Packet Radio Service (GPRS) for the back-channel. Using standard GPRS or Enhanced Data Rates for GSM Evolution (EDGE), costs are reduced for higher effective rates if the upload volume is very low, and also because this service is not per-time charged, but charged by volume uploaded. GPRS as return improves mobility when the service is provided by a satellite that transmits in the field of 100-200 kW. Using a 33 cm wide satellite dish, a notebook and a normal GPRS equipped GSM phone, users can get mobile satellite broadband.

System components

The transmitting station has two components, consisting of a high speed Internet connection to serve many customers at once, and the satellite uplink to broadcast requested data to the customers. The ISP's routers connect to proxy servers which can enforce quality of service (QoS) bandwidth limits and guarantees for each customer's traffic. 

Often, nonstandard IP stacks are used to address the latency and asymmetry problems of the satellite connection. As with one-way receive systems, data sent over the satellite link is generally also encrypted, as otherwise it would be accessible to anyone with a satellite receiver. 

Many IP-over-satellite implementations use paired proxy servers at both endpoints so that certain communications between clients and servers need not to accept the latency inherent in a satellite connection. For similar reasons, there exist special Virtual private network (VPN) implementations designed for use over satellite links because standard VPN software cannot handle the long packet travel times. 

Upload speeds are limited by the user's dial-up modem, while download speeds can be very fast compared to dial-up, using the modem only as the control channel for packet acknowledgement. 

Latency is still high, although lower than full two-way geostationary satellite Internet, since only half of the data path is via satellite, the other half being via the terrestrial channel.

One-way broadcast, receive only

One-way broadcast satellite Internet systems are used for Internet Protocol (IP) broadcast-based data, audio and video distribution. In the U.S., a Federal Communications Commission (FCC) license is required only for the uplink station and no license is required for users. Note that most Internet protocols will not work correctly over one-way access, since they require a return channel. However, Internet content such as web pages can still be distributed over a one-way system by "pushing" them out to local storage at end user sites, though full interactivity is not possible. This is much like TV or radio content which offers little user interface.

The broadcast mechanism may include compression and error correction to help ensure the one-way broadcast is properly received. The data may also be rebroadcast periodically, so that receivers that did not previously succeed will have additional chances to try downloading again.

The data may also be encrypted, so that while anyone can receive the data, only certain destinations are able to actually decode and use the broadcast data. Authorized users only need to have possession of either a short decryption key or an automatic rolling code device that uses its own highly accurate independent timing mechanism to decrypt the data.

System hardware components

Similar to one-way terrestrial return, satellite Internet access may include interfaces to the public switched telephone network for squawk box applications. An Internet connection is not required, but many applications include a File Transfer Protocol (FTP) server to queue data for broadcast.

System software components

Most one-way broadcast applications require custom programming at the remote sites. The software at the remote site must filter, store, present a selection interface to and display the data. The software at the transmitting station must provide access control, priority queuing, sending, and encapsulating of the data.

Services

Emerging commercial services in this area include:

• Outernet – Satellite constellation technology

Efficiency increases

2013 FCC report cites big jump in satellite performance

In its report released in February, 2013, the Federal Communications Commission noted significant advances in satellite Internet performance. The FCC's Measuring Broadband America report also ranked the major ISPs by how close they came to delivering on advertised speeds. In this category, satellite Internet topped the list, with 90% of subscribers seeing speeds at 140% or better than what was advertised.

Reducing satellite latency

Much of the slowdown associated with satellite Internet is that for each request, many roundtrips must be completed before any useful data can be received by the requester. Special IP stacks and proxies can also reduce latency through lessening the number of roundtrips, or simplifying and reducing the length of protocol headers. Optimization technologies include TCP acceleration, HTTP pre-fetching and DNS caching among many others. See the Space Communications Protocol Specifications standard (SCPS), developed by NASA and adopted widely by commercial and military equipment and software providers in the market space.

Satellites launched

The WINDS satellite was launched on February 23, 2008. The WINDS satellite is used to provide broadband Internet services to Japan and locations across the Asia-Pacific region. The satellite to provides a maximum speed of 155 Mbit/s down and 6 Mbit/s up to residences with a 45 cm aperture antenna and a 1.2 Gbit/s connection to businesses with a 5-meter antenna. It has reached the end of its design life expectancy.

SkyTerra-1 was launched in mid-November 2010, providing North America, while Hylas-1 was launched in November 2010, targeting Europe.

On December 26, 2010, Eutelsat's KA-SAT was launched. It covers the European continent with 80 spot beams—focused signals that cover an area a few hundred kilometers across Europe and the Mediterranean. Spot beams allow for frequencies to be effectively reused in multiple regions without interference. The result is increased capacity. Each of the spot beams has an overall capacity of 900 Mbit/s and the entire satellite will has a capacity of 70 Gbit/s.

ViaSat-1, the highest capacity communications satellite in the world, was launched Oct. 19, 2011 from Baikonur, Kazakhstan, offering 140 Gbit/s of total throughput capacity, through the Exede Internet service.

Passengers aboard JetBlue Airways can use this service since 2015. The service has also been expanded to United Airlines, American Airlines, Scandinavian Airlines, Virgin America and Qantas.

The EchoStar XVII satellite was launched July 5, 2012 by Arianespace and was placed in its permanent geosynchronous orbital slot of 107.1° West longitude, servicing HughesNet. This K_a-band satellite has over 100 Gbit/s of throughput capacity.

Since 2013, the O3b satellite constellation claims an end-to-end round-trip latency of 238 ms for data services.

In 2015 and 2016, the Australian Government launched two satellites to provide internet to regional Australians and residents of External Territories, such as Norfolk Island and Christmas Island.

Starlink (satellite constellation)

From Wikipedia, the free encyclopedia

 

Logo

 

Starlink is a satellite constellation being constructed by American company SpaceX to provide satellite Internet access. The constellation will consist of thousands of mass-produced small satellites, working in combination with ground transceivers. SpaceX also plans to sell some of the satellites for military, scientific or exploratory purposes.

Starlink constellation, phase 1, first orbital shell: approximately 1,600 satellites at 550 km altitude

 

As of November 2019, SpaceX has deployed 122 satellites. They plan to deploy 60 more per launch, at a rate of one launch every two weeks beginning in November 2019. In total, nearly 12,000 satellites will be deployed by the mid-2020s, with a possible later extension to 42,000. The initial 12,000 satellites are planned to orbit in three orbital shells: first placing approximately 1,600 in a 550-kilometer (340 mi)-altitude shell, then approximately 2,800 K_u- and K_a-band spectrum satellites at 1,150 km (710 mi) and approximately 7,500 V-band satellites at 340 km (210 mi). Commercial operation could begin in 2020.

Concerns have been raised about the long term danger of space junk resulting from placing thousands of satellites in orbits above 1000 km and a possible impact on astronomy.

The total cost of the decade-long project to design, build and deploy the constellation was estimated by SpaceX in May 2018 to be about US$10 billion. Product development began in 2015, with the first two prototype test-flight satellites launched in February 2018. A second set of test satellites and the first large deployment of a piece of the constellation occurred on 24 May 2019 (UTC) when the first 60 operational satellites were launched. The SpaceX satellite development facility in Redmond, Washington, houses the Starlink research, development, manufacturing and on-orbit control operations.

History

2015–2017

The communication satellite network SpaceX envisions was publicly announced in January 2015, with the projected design capability to support sufficient bandwidth to carry up to 50% of all backhaul communications traffic, and up to 10% of local Internet traffic, in high-density cities. CEO Elon Musk said that there is significant unmet demand for low-cost global broadband capabilities.

SpaceX satellite development facility, Redmond, Washington, in use from 2015 to mid-2018

 

The opening of the SpaceX satellite development facility in Redmond was announced by SpaceX in January 2015 with partners, to develop and build out the new communication network. At the time, the Seattle-area office planned to initially hire approximately 60 engineers, and potentially 1,000 people by 2018. The company operated in 2,800 square meters (30,000 sq ft) of leased space by late 2016, and by January 2017 had taken on a 3,800 square meters (40,625 sq ft) second facility, both in Redmond. In August 2018, SpaceX consolidated all their Seattle-area operations with a move to a larger three-building facility at Redmond Ridge Corporate Center to support satellite manufacturing in addition to R&D.

In July 2016, SpaceX acquired a 740 square meters (8,000 sq ft) creative space in Irvine, California (Orange County). SpaceX job listings indicated the Irvine office would include signal processing, RFIC, and ASIC development for the satellite program.

By January 2016, the company had publicly disclosed plans to have two prototype satellites flying in 2016, and have the initial satellite constellation in orbit and operational by approximately 2020. By October 2016, SpaceX had developed the initial satellites that they hoped to launch and test in 2017, but the satellite division was focusing on a significant business challenge of achieving a sufficiently low-cost design for the user equipment, aiming for something that can ostensibly install easily at end-user premises for approximately US$200. Overall, SpaceX President Gwynne Shotwell said then that the project remained in the "design phase as the company seeks to tackle issues related to user-terminal cost." Deployment, if carried out, would not be until "late in this decade or early in the next." The two original test satellites were not flown, and were used only in ground testing. The planned launch of two revised satellites was moved to 2018.

In November 2016, SpaceX filed an application with the Federal Communications Commission (FCC) for a "non-geostationary orbit (NGSO) satellite system in the Fixed-Satellite Service using the K_u and K_a frequency bands."

By March 2017, SpaceX filed plans with the FCC to field a second orbital shell of more than 7500 "V-band satellites in non-geosynchronous orbits to provide communications services" in an electromagnetic spectrum that has not previously been heavily employed for commercial communications services. Called the "Very-Low Earth Orbit (VLEO) constellation," it would consist of 7,518 satellites and would orbit at just 340 kilometres (210 mi) altitude, while the smaller originally-planned group of 4,425 satellites would operate in the K_a- and K_u-bands and orbit at 1,200 kilometres (750 mi) altitude. SpaceX's plans were unusual in two areas: the company intended to use the little-used V band of the communications spectrum, and also intended to operationally use a new orbital regime, the very-low Earth orbit regime of ~340 km altitude, where atmospheric drag is quite high – normally resulting in short orbital lifetimes. SpaceX has not made public the specific spaceflight technology they intend to use to deal with the high-drag environment of VLEO. The March 2017 plan called for SpaceX to launch test satellites of the initial K_a/K_u type in both 2017 and 2018, and begin launching the operational constellation in 2019. Full build-out of the ~1200 km constellation of ~4,440 satellites was not then expected to be completed until 2024.

Some controversy arose in 2015–2017 with regulatory authorities on licensing of the communications spectrum for these large constellations of satellites. The traditional and historical regulatory rule for licensing spectrum has been that satellite operators could "launch a single spacecraft to meet their in-service deadline [from the regulator], a policy seen as allowing an operator to block the use of valuable radio spectrum for years without deploying its fleet." By 2017, the FCC had set a six-year deadline to have an entire large constellation deployed to comply with licensing terms. The international regulator, International Telecommunication Union, proposed in mid-2017 a guideline that would be considerably less restrictive. In September 2017, both Boeing and SpaceX petitioned the US FCC for a waiver of the 6-year rule, but that was ultimately not granted. By 2019, the FCC had set the rule to be that half of the constellation must be in orbit in six years, with the full system in orbit by nine years from the date of the license.

SpaceX trademarked the name Starlink for their satellite broadband network in 2017; the name was inspired by the book The Fault in Our Stars. 

SpaceX filed documents in late 2017 with the FCC to clarify their space debris mitigation plan. The company will "implement an operations plan for the orderly de-orbit of satellites nearing the end of their useful lives (roughly five to seven years) at a rate far faster than is required under international standards. [Satellites] will de-orbit by propulsively moving to a disposal orbit from which they will reenter the Earth's atmosphere within approximately one year after completion of their mission." In March 2018, the FCC issued SpaceX approval with some conditions. SpaceX would need to obtain a separate approval from the ITU. The FCC supported a NASA request to ask SpaceX to achieve an even higher level of de-orbiting reliability than the standard that NASA had previously used for itself: reliably deorbiting 90% of the satellites after their missions are complete.

2018–2019

In May 2018, SpaceX expected the total cost of development and buildout of the constellation to approach US$10 billion. In mid-2018, SpaceX reorganized the satellite development division in Redmond, and fired several members of senior management.

In November 2018, SpaceX received US regulatory approval to deploy 7,518 broadband satellites, in addition to the 4,425 approved earlier. SpaceX's initial 4,425 satellites had been requested in the 2016 regulatory filings to orbit at altitudes of 1,110-kilometer (690 mi) to 1,325-kilometer (823 mi), well above the International Space Station. The new approval was for the addition of a very-low Earth orbit non-geostationary satellite orbit constellation, consisting of 7,518 satellites operating at altitudes from 335-kilometer (208 mi) to 346-kilometer (215 mi), below the ISS. Also in November, SpaceX made new regulatory filings with the US Federal Communications Commission (FCC)to request the ability to alter its previously granted license in order to operate approximately 1,600 of the 4,425 K_a-/ K_u-band satellites approved for operation at 1,150 km (710 mi) in a "new lower shell of the constellation" at only 550 km (340 mi) orbital altitude. These satellites would effectively operate in a third orbital shell, a 550-kilometer (340 mi) orbit, while the higher and lower orbits at ~1,200-kilometer (750 mi) and ~340-kilometer (210 mi) would be used only later, once a considerably larger deployment of satellites becomes possible in the later years of the deployment process. The FCC approved the request in April 2019, giving approval to place nearly 12,000 satellites in three orbital shells: initially approximately 1,600 in a 550-kilometer (340 mi)-altitude shell, and subsequently placing ~2800 K_u- and K_a-band spectrum satellites at 1,150 km (710 mi) and ~7500 V-band satellites at 340 km (210 mi).

With plans by several providers to build commercial space-Internet mega-constellations of thousands of satellites increasing likely to become a reality, the US military began to perform test studies in 2018 to evaluate how the networks might be used. In December, the US Air Force issued a US$28 million contract for specific test services on Starlink.

In February 2019, a sister company of SpaceX, SpaceX Services, Inc., filed a request with the FCC to request a license for the operation of up to 1,000,000 fixed satellite earth stations that would communicate with its non-geostationary orbit satellite (NGSO) Starlink system.

By April 2019, SpaceX was transitioning their satellite efforts from research and development to manufacturing, with the planned first launch of a large batch of satellites to orbit, and the clear need to achieve an average launch rate of "44 high-performance, low-cost spacecraft built and launched every month for the next 60 months" to get the 2,200 satellites launched to support their FCC spectrum allocation license assignment. SpaceX said they will meet the deadline of having half the constellation "in orbit within six years of authorization … and the full system in nine years."

By the end of June 2019, SpaceX had communicated with all 60 satellites but lost contact with three; the remaining 57 were working as intended. 45 satellites had reached their final orbital altitude of 550 km (340 mi), five were still raising their orbits, and another five were undergoing systems checks before they raise their orbits. The remaining two satellites were intended to be quickly removed from orbit and reenter the atmosphere in order to test the satellite de-orbiting process; the three that lost contact were also expected to reenter, but will do so passively from atmospheric drag as SpaceX was no longer able to actively control them.

In June 2019, SpaceX applied to the FCC for a license to test up to 270 ground terminals—70 nationwide across the United States and 200 in Washington state at SpaceX employee homes—and aircraft-borne antenna operation from four distributed US airfields; as well as five ground-to-ground test locations.

By September 2019, SpaceX had gone back to the FCC to apply for more changes to the orbital constellation. SpaceX asked to triple the number of orbital planes in the 550 km orbital shell, from 24 to 72, arguing that they could then place satellites into multiple planes from a single launch, and provide service earlier to more areas.

Phase	Orbit shells (km)	Number of satellites	Inclination (degrees)	Half size contractual completion time	Full size contractual completion time	Current completion (11 November 2019)	Satellites actively preparing to de-orbit (11 November 2019)	Dead Satellites (11 November 2019)
1	550	1,584	53	March 2024	March 2027	122	2	3
	1,110	1,600	53.8			0
	1,130	400	74			0
	1,275	375	81			0
	1,325	450	70			0
2	335.9	2,493	42	November 2024	November 2027	0
	340.8	2,478	48			0
	345.6	2,547	53			0

Launches

View of the 60 Starlink satellites from the May 24, 2019 launch

 

The deployment of the first 1,584 satellites will be into 24 orbital planes of 66 satellites each, with a requested lower minimum elevation angle of beams to improve reception: 25 degrees rather than the 40 degrees of the other two orbital shells. SpaceX launched the first 60 satellites of the constellation in May 2019 into a 450 km orbit and expected up to six launches in 2019 at that time, with 720 satellites (12*60) for continuous coverage in 2020.

In August 2019 SpaceX expected 4 more launches in 2019 and at least 9 launches in 2020.

Starlink satellites are also planned to launch on Starship, an under-development rocket of SpaceX that will launch 400 satellites at a time.

List of launches

Flight №	Mission	Date and time (UTC)	Launch site	Launch vehicle	Orbit altitude (km)	Inclination	Number deployed	Version	Outcome
1	Tintin	22 February 2018 14:17	Vandenberg	F9 FT  B1038.2	514	97.5°	2	–	Success
	Two test satellites known as Tintin A and B (MicroSat-2a and 2b) that were deployed as co-payloads to the Paz satellite.
2	Starlink 0	24 May 2019 02:30	CCAFS SLC-40	F9 B5  B1049.3	440 to 550	53°	60	v0.9	Success
	Second launch of test satellites for SpaceX's Starlink constellation. Said to be "production design", these are used to test various aspects of the network, including deorbiting. They do not yet have the planned satellite interlink capabilities and they only communicate with antennas on Earth. A day after launch an amateur astronomer in the Netherlands was one of the first to publish a video showing the satellites flying across the sky as a "train" of bright lights. By five weeks post launch, 57 of the 60 satellites were "healthy" while 3 had become non-operational and were derelict, but will deorbit due to atmospheric drag. As of 31 October 2019, 49 satellites were in the target 550 km orbit while the others either didn't reach it or were out of it.
3	Starlink 1	11 November 2019	CCAFS SLC-40	F9 B5  B1048.4	550 (target)	53˚	60	v1.0	Success
	A SpaceX Falcon 9 rocket launched the second batch of 60 satellites for SpaceX’s Starlink broadband network, a mission designated Starlink 1.
4	Starlink 2	November 2019	CCAFS SLC-40	F9 B5			60		Planned
	A SpaceX Falcon 9 rocket is expected to launch the third batch of approximately 60 satellites for SpaceX’s Starlink broadband network, a mission designated Starlink 2.
5	Starlink 3	December 2019	CCAFS SLC-40	F9 B5			60		Planned
	A SpaceX Falcon 9 rocket is expected to launch the fourth batch of approximately 60 satellites for SpaceX’s Starlink broadband network, a mission designated Starlink 3.

Services

Global broadband Internet

SpaceX intends to provide broadband internet connectivity to underserved areas of the planet, as well as provide competitively-priced service to urban areas. The company has stated that the positive cashflow from selling satellite internet services would be necessary to fund their Mars plans.

In early 2015, two space entrepreneurs announced Internet satellite ventures in the same week. In addition to SpaceX CEO Elon Musk announcing the project that would later be named Starlink, serial-entrepreneur Richard Branson announced an investment in OneWeb, a similar constellation with approximately 700 planned satellites that had already procured communication frequency licenses for their broadcast spectrum.

After the failures of previous satellite-to-consumer space ventures, satellite industry consultant Roger Rusch said in 2015 "It's highly unlikely that you can make a successful business out of this." Musk publicly acknowledged that business reality, and indicated in mid-2015 that while endeavoring to develop this technically-complicated space-based communication system he wanted to avoid overextending the company, and stated that they are being measured in their pace of development. Nevertheless, internal documents leaked in February 2017 indicated that SpaceX expected more than US$30 billion in revenue by 2025 from its satellite constellation, while revenues from its launch business were expected to reach US$5 billion in the same year.

In February 2015, financial analysts questioned established geosynchronous orbit communications satellite fleet operators as to how they intend to respond to the competitive threat of SpaceX and OneWeb LEO communication satellites. In October, SpaceX President Gwynne Shotwell indicated that while development continues, the business case for the long-term rollout of an operational satellite network was still in an early phase.

In 2015, court documents indicate that SpaceX had engaged in collaboration with wireless chip-maker Broadcom. Five key engineers subsequently left to join SpaceX, leading to a lawsuit filed by Broadcom alleging that "SpaceX stole our best minds." In March, an Orange County judge denied Broadcom's multiple restraining order requests.

With the initial launch of the first 60 satellites of the operational constellation in 2019, SpaceX indicated that it would require 420 satellites in the constellation to achieve minor broadband coverage of Earth, and 780 of the first ~1600 to provide moderate coverage.

Use beyond Earth

In the long term, SpaceX intends to develop and deploy a version of the satellite communication system to serve Mars.

Satellite hardware

The Internet communication satellites were expected to be in the smallsat-class of 100-to-500 kg (220-to-1,100 lb)-mass, and were intended to be in Low Earth Orbit (LEO) at an altitude of approximately 1,100 kilometers (680 mi), according to early public releases of information in 2015. In the event, the first large deployment of 60 satellites in May 2019 were 227 kilograms (500 lb) and SpaceX decided to place the satellites at a relatively low 550 kilometers (340 mi), due to concerns about the space environment. Initial plans as of January 2015 were for the constellation to be made up of approximately 4,000 cross-linked satellites, more than twice as many operational satellites as were in orbit in January 2015.

The satellites will employ optical inter-satellite links and phased array beam-forming and digital processing technologies in the K_u and K_a bands, according to documents filed with the U.S. Federal Communications Commission (FCC). While specifics of the phased array technologies have been disclosed as part of the frequency application, SpaceX enforced confidentiality regarding details of the optical inter-satellite links. Early satellites are launched without laser links, in October 2019 SpaceX expected satellites with these links to be ready by the end of 2020.

The satellites will be mass-produced, at a much lower cost per unit of capability than existing satellites. Musk said, "We’re going to try and do for satellites what we’ve done for rockets." "In order to revolutionize space, we have to address both satellites and rockets." "Smaller satellites are crucial to lowering the cost of space-based Internet and communications."

In February 2015, SpaceX asked the FCC to consider future innovative uses of the K_a band spectrum before the FCC commits to 5G communications regulations that would create barriers to entry, since SpaceX is a new entrant to the satellite communications market. The SpaceX non-geostationary orbit communications satellite constellation will operate in the high-frequency bands above 24 GHz, "where steerable earth station transmit antennas would have a wider geographic impact, and significantly lower satellite altitudes magnify the impact of aggregate interference from terrestrial transmissions."

The system will not compete with the Iridium satellite constellation, which is designed to link directly to handsets. Instead, it will be linked to flat user terminals the size of a pizza box, which will have phased array antennas and track the satellites. The terminals can be mounted anywhere, as long as they can see the sky.

Internet traffic via a geostationary satellite has a minimum theoretical round-trip latency of at least 477 ms (between user and ground gateway), but in practice, current satellites have latencies of 600 ms or more. Starlink satellites would orbit at ​¹⁄₃₀ to ​¹⁄₁₀₅ of the height of geostationary orbits, and thus offer more practical Earth-to-sat latencies of around 25 to 35 ms, comparable to existing cable and fiber networks. The system will use a peer-to-peer protocol claimed to be "simpler than IPv6", it will also incorporate end-to-end encryption natively. However, no details on this have been released as of yet. 

Starlink satellites use Hall-effect thrusters with krypton gas as the reaction mass for orbit raising and station keeping. Krypton Hall thrusters tend to exhibit significantly higher erosion of the flow channel compared to a similar electric propulsion system operated with xenon, but at a lower propellant cost.

Satellite revisions

At the time of the June 2015 announcement, SpaceX had stated plans to launch the first two demonstration satellites in 2016, but the target date was subsequently moved out to 2018. SpaceX began flight testing their satellite technologies in 2018 with the launch of two test satellites. The two identical satellites were called MicroSat-2a and MicroSat-2b during development but were renamed Tintin A and Tintin B upon orbital deployment in February 2018. Two previously manufactured satellites, MicroSat-1a and MicroSat-1b were meant to be launched together as secondary payloads on one of the Iridium-NEXT flights, but they were instead used for ground-based tests.

MicroSat 1a & 1b were originally slated to be launched into 625 km circular orbits at approximately 86.4 degrees inclination, and to include panchromatic video imager cameras to film images of Earth and the satellite.

Tintin A and B were inserted into a 514 km orbit. Per FCC filings they were intended to raise themselves to an 1125 km orbit, the operational altitude for Starlink LEO satellites per the earliest regulatory filings, but stayed close to their original orbits. SpaceX announced in November 2018 that they would like to operate an initial shell of about 1,600 satellites in the constellation at about 550 km orbital altitude, at an altitude similar to the orbits Tintin A and B stayed in.

The satellites currently orbit in a circular low Earth orbit at about 500 kilometers (310 mi) altitude in a high-inclination orbit for a planned six to twelve-month duration. The satellites will communicate with three testing ground stations in Washington and California for short-term experiments of less than ten minutes duration, roughly daily.

The 60 Starlink v0.9 satellites, launched May 2019, have the following characteristics:

• Flat-panel design with multiple high-throughput antennas and a single solar array
• Mass: 227 kg (500 lb)
• Hall-effect thrusters using krypton as the reaction mass, for position adjustment on orbit, altitude maintenance and deorbit
• Star tracker navigation system for precision pointing
• Able to use Department of Defense provided debris data to autonomously avoid collision.
• Operational altitude of 550 km (340 mi)
• 95 percent of "all components of this design will quickly burn in Earth’s atmosphere at the end of each satellite’s lifecycle"

The 60 Starlink v1.0 satellites, launched November 2019, have the additional following characteristics:

• 100% of "all components of this design will quickly burn in Earth's atmosphere at the end of each satellite's lifecycle."
• K_a-band added.
• Mass: 260 kg
• Albedo reduced.
• Operational altitude of 550 km (340 mi)

Competition and market effects

In addition to the OneWeb constellation, announced nearly concurrently with the SpaceX constellation, a 2015 proposal from Samsung outlined a 4,600-satellite constellation orbiting at 1,400 kilometers (900 mi) that could provide a zettabyte per month capacity worldwide, an equivalent of 200 gigabytes per month for 5 billion users of Internet data, but by 2019, no more public information had been released about the Samsung constellation. Telesat announced a smaller 117 satellite constellation in 2015 with plans to deliver initial service in 2021. Amazon announced a large broadband internet satellite constellation in April 2019, planning to launch 3,236 satellites in the next decade in what the company calls "Project Kuiper", a satellite constellation that will work in concert with Amazon's previously-announced large network of 12 satellite ground station facilities (the "AWS Ground Station unit") announced in November 2018.

By October 2017, the expectation for large increases in satellite network capacity from emerging lower-altitude broadband constellations caused market players to cancel some planned investments in new geosynchronous orbit broadband communications satellites.

Criticism

The large number of planned satellites have been met with criticism from the astronomical community. Astronomers claim that the number of visible satellites will outnumber visible stars, and that their brightness in both optical and radio wavelengths will severely impact scientific observations. Because the Starlink satellites can autonomously change their orbits, observations cannot be scheduled to avoid them. The International Astronomical Union and National Radio Astronomy Observatory have released official statements expressing concern on the matter.

SpaceX representatives and Musk have claimed that the satellites will have minimal impact. Many professional astronomers have disputed these claims based on initial observation of the Starlink v0.9 satellites on the first launch, shortly after their deployment from the launch vehicle. In later statements on Twitter, Musk stated that SpaceX will work on reducing the albedo of the satellites and will provide on-demand orientation adjustments for astronomical experiments, if necessary.

Concerns have been raised about the long term danger of space junk resulting from placing thousands of satellites in orbits above 1,000 km, where orbital decay takes several thousand years. Professional astronomers have raised concerns with the impact of so many transmitting satellites on both optical and radio astronomy.