WASHINGTON, December 7, 2021 – There is concern in the telecom industry that the technology neutrality provision in the infrastructure bill, which includes $65 billion for broadband, would put a chunk of money into low-earth orbit satellites that would eventually lose its ability to uphold requirements for federal funds.
Cartesian, a consulting firm in telecom and technology, conducted a study earlier this year that was commissioned by the Fiber Broadband Association and NTCA – the Rural Broadband Association, and found that SpaceX’s Starlink LEO fleet would run out of capacity within 10 years. LEO constellations are known to require a lot of satellites for coverage and capacity, which makes it an expensive business.
As part of its obligations, SpaceX must offer 100 Mbps download and 20 Mbps upload speeds to 640,000 locations across the U.S. “That is quite a lot of capacity,” Cartesian Vice President Michael Dargue said in a recent interview with Broadband Breakfast. “We wanted to find out whether there was sufficient capacity within Starlink’s planned fleet.”
Cartesian estimated that Starlink could face a shortfall in capacity before the end of the decade in 2028. “Just over half of the RDOF subscribers wouldn’t get the full 100 Mbps that [Starlink committed to],” said Dargue.
The problem for critics of Starlink’s abilities is that Starlink continues to launch satellites into the sky at a blistering pace, which will mean the company will continue to seek an ever-growing share of federal funds. Before the Federal Communications Commission began scrutinizing winners of the $9.2-billion Rural Digital Opportunity Fund, the company had been awarded nearly $900 million from the fund for its fleet.
Now there’s concern that the technology neutrality provisions in the Infrastructure Investment and Jobs Act, signed into law in mid-November, as well as the bill’s promotion of satellite technology will mean more money going toward the nascent technology versus more proven technologies like fiber.
SpaceX did not respond to the requests for comment on these concerns. Broadband Breakfast also contacted Ligado and OneWeb to get the LEO perspective but did not hear back. When approached, the Wireless Internet Service Providers Association declined to comment.
Fiber Broadband Association President and CEO Gary Bolton said in an interview that federal funds coming down the pipe from the infrastructure bill represent a “once in a generation opportunity to get fiber to every American.
“The money is available,” Bolton said. “There is no longer a question of, ‘can we do this on the cheap?’”
A state-appointed task force for broadband in Alaska found that the federal money could allow the state famously known for difficult terrain for broadband builds to have a statewide fiber network.
“LEO satellites are great if I am climbing Mount Everest, or if I’m in some off the grid location and I need to be able to make a phone call or get on the internet,” said Bolton. “That’s great, but if you’re talking about building up the economic development for your community, that’s not so great.”
FCC needs to study Starlink
Dargue said the Cartesian study was explicitly from an “outside-in” perspective, and that the assessment was only able to go off data that SpaceX had made publicly available. The assessment noted, however, that because there is limited information regarding Starlink’s technical capabilities in the public domain, and Starlink’s technical and commercial plans seem to be constantly changing, it is difficult to truly assess the full extent of Starlink’s potential (or lack thereof).
“[The FCC] really needs to do this assessment themselves in detail,” Dargue said. “We did not have access to Starlink’s engineering data and really, if you’re going to make an award of this size, which is over a 10-year period, you need to make sure that the numbers are right. If you get to seven or eight years down the road and it does not work anymore, what do you do then?
“We were quite generous [to Starlink] in some ways,” added Dargue. The assessment assumed that served regions would not have any terrain features that would block reception, so all subscribers within range of a satellite can connect to that satellite. Additionally, the assessment assumed that the throughput of each satellite in the Starlink constellation was 20 Gbps with no pinch-points elsewhere in the network.
“Then, using demand modeling based on current demand and how Cisco and others expect that to grow over the coming decade, we look to see whether there will be enough capacity within the fleet to serve the geographic demand,” said Dargue.
Dargue said this did not mean that consumers would never see their service at 100/20, but that consumer use during peak demand hours would exceed the available capacity. He said that for consumers, this would spell a deterioration in the quality of service, resulting in buffering, scaled down resolutions, and other potential disruptions to internet services.
Proponents of LEOs say technology is important for redundancy
Though the study was not favorable to Starlink and SpaceX, Dargue is not arguing for satellite to be left out of the infrastructure equation. “It’s definitely part of the mix,” he said. “LEO satellites and other constellations are really good at serving very remote locations off the beaten track and in areas where you do not have a cluster of high demand.”
Similarly, proponents of LEO satellites and Starlink, including the Gigabit Libraries Network, have said the technology serves as an excellent way to get redundant connections in case of an outage. It is also crucial is some areas that can’t get a physical connection to the premises.
