The antenna is the architecture, and in a megaconstellation the user terminal is where antenna cost meets the customer's wallet. US10770790B1, granted to Space Exploration Technologies Corp. on September 8, 2020, claims "uni-dimensional steering of phased array antennas" — beam steering electronically constrained to a single dimension.
The CPC is pure antenna art: H01Q 3/34 (electrical steering via phase variation), H01Q 21/061 (array arrangements), and H01Q 15/02 / 15/14 (reflecting/refracting elements). No comms-protocol classifications — this is a hardware-and-geometry claim about how the beam is formed and pointed.
Why give up an axis? A fully electronically steered 2D array needs a phase shifter behind every element across both dimensions, and the element count drives cost, power, and thermal load. By steering electronically in one dimension and handling the other axis mechanically or through array geometry, you collapse the most expensive part of the bill of materials. For a constellation that has to ship millions of terminals, that trade is the whole ballgame.
This is exactly the kind of dependent design choice that looks like a compromise on a spec sheet and reads as a strategy in a claims chart. Spectrum plus geometry is the battleground, and here SpaceX is optimizing the geometry side to win on unit economics rather than chasing maximum agility.
The usual caveat: an early grant on a specific steering method is leverage, not a fence around the concept of steerable user terminals — a crowded field including Kymeta, Anokiwave, and the satellite primes. What this record reveals is the design philosophy: SpaceX treats the terminal as a cost-engineering problem and patents the shortcuts that make mass deployment affordable.