The headline number for any consumer satellite terminal is the bill of materials. A fully electronically steered phased array, agile in both azimuth and elevation, needs a dense grid of phase shifters and control electronics, and the cost scales with the element count. For a constellation that must ship millions of terminals, that cost is the whole business case.
SpaceX's grant US11239553B2 (inventor Alireza Mahanfar) is classified in H01Q 3/34, electronic beam scanning. Read the controlling claim carefully and the cleverness is in what it gives up: the array steers electronically in one dimension only. The second axis is handled by the physical pointing of the terminal at install.
This is the moat-is-in-the-trade pattern. Halving the steering dimensionality roughly halves the most expensive part of the antenna while still tracking fast-moving LEO satellites across the sky, because a Starlink terminal can be mechanically tilted once and then chase satellites mainly along their orbital track. The constraint is the feature.
Set this beside the contemporaneous antenna landscape and the contrast is instructive: Northrop Grumman's January 2021 grant US10892549B1, also a phased-array system in H01Q 3/34, comes from the defense world where full agility and performance dominate cost. SpaceX optimized the opposite corner of the design space.
The market consequence is direct. Cadence and unit cost are the two levers of constellation economics, and this patent attacks unit cost at the user edge rather than at the satellite. A cheaper terminal expands the addressable market faster than another satellite in the plane does.
What the patent does not settle is durability and yield at volume, the hard part of any consumer-hardware claim. The document tells you the design intent; the manufacturing line tells you whether the intent survived contact with reality.