CONSTELLATION DESIGNS: A New CAFC Decision That Provides Insight Into What Makes A Claim “Abstract,” And Some Strategy Observations
Written by Ben Esplin
On April 28, 2026, the Federal Circuit issued a highly instructive decision on patent eligibility in Constellation Designs, LLC v. LG Electronics Inc. The court evaluated claims regarding digital communication systems, striking down broad “optimization claims” as patent-ineligible abstract ideas under 35 U.S.C. § 101, while upholding structurally specific (at least on an informational level) “constellation claims” as eligible technological solutions. Since both the “optimization claims” and the “constellation claims” arose from an insight the inventors had about constellations of symbols used to transmit data in digital communications systems, the opinion provides one of the clearest statements by the CAFC to this point regarding what the differences are between claims that are patent eligible and claims that are patent ineligible.
The key insight by the inventors led to an iterative, mathematical R&D process used to discover optimal constellation points. This process, incidentally, is described in all of the patents. In essence, the court found that since the “optimization claims” sought to protect any constellations discovered using the process, they were directed to an abstract idea without significantly more. On the other hand, the “constellation claims,” which recite specifics of certain constellations yielded by the process, were found to include the necessary “structure” to elevate these claims to a technological solution that is patent eligible under § 101. Using these contrasting findings as a blueprint provides a clear roadmap for protecting underlying innovations that generate novel solutions without running afoul of § 101.
This raises an interesting thought experiment and strategy consideration: what if the optimization algorithm had been kept internal, and the patents were directed exclusively toward the concrete technological output that was ultimately found to be patentable? As previously discussed in prior writings regarding the fusion of patent and trade secret protection, an enterprise is not obligated to patent every layer of a technology. Under the enablement requirement of 35 U.S.C. § 112, a specification must teach a person of ordinary skill in the art how to make and use the claimed invention. It does not strictly require a historical accounting of how the inventors arrived at their discovery. Enabling the specific constellation claims legally requires describing the claimed features of the constellations within the appropriate technological context, detailing how they are implemented within transmitter and receiver hardware, but it does not strictly require disclosing the mathematical derivation loop that was the kernel of innovation in this particular case.
It is fair to note that the description of the R&D process in the specification is discussed in the decision as being helpful to the ultimate finding that the “constellation claims” are patent eligible. Without the narrative in the specification explaining the "why" and "how" behind the counterintuitive overlapping points in the claimed constellations, an examiner or a court might easily view the specific coordinate claims in a vacuum as mere abstract math or arbitrary data. However, this cale illustrates that the value of maintaining a repeatable R&D process as a trade secret while protecting its fruit separately via patent is worth careful consideration.
Navigating this tension requires a highly deliberate approach to intellectual property architecture. By carefully fusing trade secret protocols for internal methodologies with precise patent drafting for the outward-facing technological output, an enterprise can ensure its assets are built to withstand intense judicial scrutiny.