D2D land-grab, tower premiums, 6G problems

So, we have some familiar talking points today, based on new news and views. Let’s start with the satellite stuff. The FCC has approved Amazon Leo’s $11.5bn purchase of Globalstar, announced in April, based on a review by its space bureau of the proposed transfer of 25 MHz of MSS spectrum as part of the deal. The move bolsters the tech conglomerate’s start-up space business, ostensibly to take the fight to Starlink. Amazon Leo has nearly 400 satellites in orbit; Starlink has 11,000. It filed with the FCC in July to deploy 5,105 new D2D satellites to use Globalstar’s spectrum. (The FCC has just approved SpaceX to deploy 15,000, and AST SpaceMobile to deploy 248).

The Globalstar package includes two dozen satellites, with approved plans for 54. But the deal – scheduled to close in 2027, and likely the early part – is about spectrum, really. SpaceX agreed to buy 50 MHz of spectrum from EchoStar in late 2025, approved by the FCC in May, across the AWS-4 and H-block bands, plus up to 15 MHz of AWS-3 spectrum – for a total of up to 65 MHz – in a deal worth $19.6bn; it has since agreed to buy up to 14 MHz of paired spectrum in the 800 MHz band from Grain Management for around $8bn. A crude price-per-MHz comparison makes Globalstar look expensive: about $460m per MHz, against about $300m for the complete EchoStar holdings.

But the bands are different, and come with different assets and terms. Globalstar brings a satellite network, okay, but also regulatory authorizations and a deal with Apple to connect its phones and watches, including for SOS calls. Apple invested about $1.5bn in Globalstar ​in 2024 in return for a 20% equity stake. Amazon will pick up the stake, and maintain the service. Also, the FCC has separately authorized Globalstar to operate a new 48-satellite network (C-3 System) in the US using specific uplink/downlink D2D bands (1610-1618.725 MHz and 2483.5-2500 MHz, plus others (5091-5250 MHz, 6875-7055 MHz) for comms channels with ground stations and control links to spacecraft. It all puts a premium on ground infra.

AT&T has signed as a distribution channel for Amazon Leo in the US; SpaceX, of course, is busy (talking about) building a terrestrial network to turn its D2D extensions into a full-blown carrier service – upsetting the big US telcos along the way. The FCC sounded an “aggressive” note itself, with a view to free-up more D2D spectrum. Tomorrow (October 29), it will vote to auction the 1675-95 MHz and 2020-2025 MHz bands, and also to open up 500 MHz of additional spectrum. Brendan Carr, chairman at the FCC, called the Amazon Lee authorization a “significant step towards supercharging next-generation connectivity”.

The other talking point is about 6G monetization, but we are out of time. So I’d just recall that RCR wrote (reported) way back in 2018 that 5G might be considered a testbed for 6G (Whisper it: is 5G the new 3G? – ran the headline), and also commend Christian’s write-up of a discussion with BT and Ericsson at 6G Forum about drone sensing (ISAC) and robots (or physical AI) as the likeliest money-spinners for this next-gen mobile tech – and ask: were we right about the testbed, and haven’t we seen this movie before?

A longer version of this article is available here.