Follow the appropriation and you arrive at missile defense, the line item that the revived "Golden Dome" framing put back on front pages in 2025-26. But the public budget debate is downstream of decades of engineering, and the cleanest window into what the primes actually build is not a press release — it is their patents. Three grants, from three primes, sketch three layers of the same problem.
Start with the engagement-management layer. Raytheon's US9726460B2, "System and method for asymmetric missile defense" (issued 2017), is classified in F41H 11/00 (defense against attack) with a notable second tag: G06N 5/04 — knowledge-based reasoning. The grant addresses the decision problem in defense, where a defender facing many cheap threats with few expensive interceptors must allocate them well. "Asymmetric" is the operative word: the cost-exchange ratio — how much your defense costs versus the attacker's offense — is the whole strategic question, and this claim sits on the reasoning that manages it.
Next, the kill mechanism. Northrop Grumman's US9671200B1, "Kinetic air defense" (issued 2017), is classified across F41G 7 (guided-missile control) and F41H 11/02. "Kinetic" is the key term, and it deserves a definition because coverage uses it loosely: a kinetic kill defeats a target by colliding with it — hit-to-kill — rather than detonating a warhead nearby. The engineering burden shifts entirely onto guidance and closing geometry, because you must physically intercept the target, not merely get close. The grant's F41G 7/2213 and 7/2286 tags are exactly that guidance lineage.
Third, the counter-seeker layer — defending the defender. Lockheed Martin's US9335127B1, "System and method for defense against radar homing missiles" (issued 2016), classified in F41G 7/301 and G01S 13/883 (radar systems), addresses threats that home on a radar emission. Here the problem inverts: the thing that lets you see the threat — your radar — is the thing the threat uses to find you. The claim is on managing that exposure, the unglamorous defensive-counter-defensive layer that rarely makes a budget headline but shapes every survivable architecture.
What the documents collectively show is that "missile defense" is not one capability but a stack: sense, decide, engage, and protect the sensors doing the sensing. Each prime's grant occupies a different rung, which is also how the contracts get structured — layered programs with different primes owning different segments. The patents are the technical fingerprints of that division of labor.
Two cautions, in this desk's house style. First, a patent is not a fielded system; these grants describe methods, several of them years old, and the public record deliberately goes dark exactly where the most sensitive seeker and discrimination work lives. Second, an award is a signal, not recognized revenue — the existence of IP tells you what a prime can build, not what it has been paid to deliver. For the dollar magnitudes and program status, the place to look is the 8-Ks and 10-Ks; the patents tell you the engineering the money is buying.