With SpaceX planning to launch another set of Starlink internet satellites on Wednesday, the company will continue the deployment of the initial constellation of approximately 1500 satellites, on the way to building their 4400 satellite Ku-/Ka-band network.
How far along is Starlink deployment?
Ahead of the latest Falcon 9 mission, SpaceX has conducted six launches of the operational Starlink v1.0 design, for a total of 360 satellites.
A handful of these suffered problems on orbit, with three being actively deorbited so far and SpaceX noting in an FCC filing that six more have had propulsion problems after they began orbit raising. The current number of functional Starlink satellites for providing service to customers is around 350, not all of which have reached their operational position yet.
Each launch of 60 satellites that have been carried so far has aimed to populate three planes of 20 satellites, each plane spaced 20 degrees apart at the equator. Once they get 18 evenly spaced planes into position, they should be able to test the system with continuous service in the northern United States.
As it takes about four months for the satellites to all reach their operational positions from launch, it appears SpaceX decided to speed up the process a little by using the seventh launch on June 3 to populate the 18th plane.
Moving forward, subsequent launches will be used to create more planes in between the existing ones until there are 72 planes spaced five degrees apart.
Animation showing the deployment of Starlink satellites into their orbital planes since January 1, 2020. Graphic created by Ben Craddock for NASASpaceflight.
In a recent interview with Aviation Week, SpaceX President Gwynne Shotwell said that they should begin beta testing the network this year and would want to complete around 14 launches before publicly promoting Starlink service. That could allow service to begin as soon as early 2021 depending on how fast launches can be performed.
In a recent ITU filing, SpaceX laid out a very aggressive schedule for continuing the Starlink deployment, with 13 launches in the May to September time period. This schedule is likely to spread out a bit as they run into normal launch cadence issues such as weather, range coordination, booster recovery operations, and booster refurbishment.
The first launch in that group (June 3 in Florida) has been delayed nearly a month for the above reasons. Regardless of exactly how long those launches end up taking, Ms. Shotwell’s comments indicate SpaceX doesn’t think satellite production will be a gating factor for their deployments in the near future.
An interesting feature of the schedule is that after this frenzy of launches, there would be a gap with only one launch in four months, followed by a period of twice-monthly launches to finish out the initial 1584 satellite shell of the constellation. SpaceX may have options to make changes to the satellites during that pause in the deployments, such as adding the optical inter-satellite links that have been mentioned as debuting later in 2020.
The terrestrial components: Gateways and User Terminals
SpaceX has now applied for a total of 27 Ka-band gateways (ground stations) in the U.S. Filings over the last six weeks have included new gateways in Alaska, California, Florida, Michigan, North Dakota (x2), Oklahoma and Texas. It’s not clear how the gateway on the north coast of Alaska would be used right now, with the possibility SpaceX is getting a head start on setting up a gateway in a more challenging area.
These gateways could also provide coverage for much of southern Canada and northern Mexico if SpaceX gets permission to operate in those countries. A recent notice on the website of Canada’s communications authority shows that SpaceX has begun the process of seeking permission in that country.
Locations for Ka-band gateway stations filed with the FCC by SpaceX. Map created by NASASpaceflight forum member Hummy. Click the image to access an interactive Google Map.
In an interview with Aviation Week, SpaceX CEO Elon Musk said that the hardest challenge for Starlink is not the satellites, but reducing the cost of the user terminals, including installation and support over 5-10 years of usage. Mr. Musk said that it may take several years for SpaceX to solve those problems. That matches up with what most reporters, analysts, and satellite industry executives have said about LEO constellations.
Current phased array antennas are far too expensive to use for providing service to individual homes. Inexpensive, easy to use terminals are going to be necessary to serve consumers with these new constellations. SpaceX is hard at work on the development of their 19″ phased array user terminals, which Musk describes as resembling a UFO on a stick.
On June 2, 2020 a talk was given by Guy Holmes, CEO of Tape Ark, who has been working with SpaceX to explore uses of Starlink to support oil and gas exploration.
Mr. Holmes said that Starlink performance was expected to be around 100Mbps down/40Mbps up speeds with the phased array user terminals, with the possibility of higher speeds around 1 Gb if users such as research vessels could use dual parabolic antennas to track the satellites. The 100/40Mbps numbers match a filing made with the Nebraska Rural Broadband Task Force in 2019 by a SpaceX Sales Engineer. Mr. Holmes also mentioned a test at SpaceX where a group of 50 engineers handed user terminals to family members and had them connect to the Starlink network.
What are the plans for the future of the constellation?
After the initial deployment, SpaceX will continue to build the rest of its 4400 satellite Ku-/Ka-band constellation, probably starting with the addition of higher latitude orbital planes. SpaceX also has another constellation approved by the FCC that involves approximately 7000 satellites in lower orbits that would communicate in V-band frequencies. (The V-band frequencies can also be added to the original 4400 satellites.)
In the last two months, SpaceX has requested a modification to the 4400 satellite constellation, and also made filings for a larger second-generation constellation.
Visualization of the orbital planes for the initial 1584 satellite Starlink deployment (left) and the full 4408 satellite Ku-/Ka-band constellation (right). Graphics by Thomas McLaughlin for NASASpaceflight
On April 17, 2020 SpaceX submitted a request to the FCC to further modify the layout of the Starlink satellite constellation. After lowering the altitude of the initial shell of satellites in their 2018 modification, SpaceX has now requested permission to lower the rest of their Ku-/Ka-band satellites from previous altitudes of 1100-1300km to new altitudes between 540 and 570km. SpaceX has also requested changes to the inclinations and structure of the orbital planes for those satellites.
To serve higher latitudes, SpaceX’s current license includes groups of satellites at inclinations of 70, 74, and 81 degrees. The new request changes those to an expanded group at 70 degrees and a new group at 97.6 degrees. The 97.6 degree inclination, 560km altitude location would be good for providing rideshare opportunities on the launches for satellites going to SSO orbits.
To improve polar coverage SpaceX has also requested permission for communications between the higher inclination satellites and high latitude gateways to occur at angles as small as 5 degrees above the horizon, as opposed to a minimum angle of 25 degrees at lower latitudes.
On May 26, 2020 SpaceX made a filing with the FCC for a second-generation constellation of up to 30,000 Starlink satellites. These satellites would communicate in Ku-, Ka-, and E-band frequencies. Launching that number of satellites would be impractical until both the new Starship launch vehicle is available and SpaceX has a large number of customers for the satellite system, but this preplanning will help clear the way for any expansion they wish to make in the future. The timing of the application was to meet an FCC deadline for coordinating new satellite filings in certain frequencies, not necessarily because SpaceX plans to act on it immediately.
What’s happening with the early Starlink satellites?
Earlier in the Starlink program, SpaceX performed two launches with development versions of the satellites. The first of these launched the two Tintin satellites, which bear little resemblance to the current design but were useful for checking the performance of the various subsystems. The Tintins are currently being deorbited. That was later followed by the 60 Starlink v0.9 satellites, which tested the new satellite bus and deployment method but lacked part of the communications payload.
It was never clear how those satellites could be used as part of the constellation, and in a recent filing with the ITU, SpaceX does not list them as being part of the constellation. Those satellites have recently begun to lower their orbits.